acta periodica technologica - početna · acta periodica technologica apteff, 48, 1-323 (2017)...

UNIVERSITY OF NOVI SAD FACULTY OF TECHNOLOGY NOVI SAD

ACTA PERIODICA TECHNOLOGICA

APTEFF, 48, 1-323 (2017)

ACTA PERIODICA TECHNOLOGICA (formerly Zbornik radova Tehnološkog fakulteta and Proceedings of Faculty of Technology) publishes articles from all branches of technology (food, chemical, biochemical, pharmaceutical), process engineering and related scientific fields. Articles in Acta Periodica Technologica are abstracted by: Chemical Abstracts Service – Columbus, Ohio; Referativnyi zhurnal – Khimija, VINITI, Moscow; listed in Ulrich’s International Periodical Directory, and indexed in the Elsevier Biblio-graphic databases – SCOPUS. ISSN 1450-7188 (Print) CODEN: APTEFF ISSN 2406-095X (Online) UDC 54:66:664:615 Publisher University of Novi Sad, Faculty of Technology Novi Sad Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia For Publisher Prof. Dr. Radomir Malbaša, Dean Editor-in-Chief Prof. Dr. Sanja Podunavac-Kuzmanović Editorial Board From Abroad Prof. Dr. Živko Nikolov Texas A and M University, Biological and Agricultural Engineering Department, College Station, TX, USA Prof. Dr. Erika Békássy-Molnár University of Horticulture and Food Industry, Budapest, Hungary Prof. Dr. Željko Knez University of Maribor, Faculty of Chemistry and Chemical Technology, Maribor, Slovenia Dr. T.S.R. Prasada Rao Indian Institute of Petroleum, Dehra Dun, India Prof. Dr. Đerđ Karlović Margarine Center of Expertise, Kruszwica, Poland Dr. Szigmond András Research Institute of Hungarian Sugar Industry, Budapest, Hungary Dr. Andreas Reitzmann Institute of Chemical Process Engineering, University Karlshruhe, Germany From Serbia Prof. Dr. Vlada Veljković Prof. Dr. Jonjaua Ranogajec Prof. Dr. Gordana Ćetković Prof. Dr. Lidija Petrović Prof. Dr. Ljubica Dokić Prof. Dr. Branka Pilić Prof. Dr. Jelena Dodić

ACTA PERIODICA TECHNOLOGICA APTEFF, 48, 1-323 (2017)

CONTENT

Zahida Ademović, Snježana Hodžić, Zarka Halilić Zahirović, Darja Husejnagić, Jasna Džananović, Broza Šarić-Kundalić, Jasmin Suljagić PHENOLIC COMPOUNDS, ANTIOXIDANT AND ANTIMICROBIAL

PROPERTIES OF THE WILD CHERRY (Prunus avium L.) STEM .......................... 1

Vojislav V. Banjac, Radmilo R. Čolović, Lato L. Pezo, Dušica S. Čolović, Jasmina M. Gubić, Olivera M. Đuragić CONDUCTOMETRIC METHOD FOR DETERMINING WATER

STABILITY AND NUTRIENT LEACHING OF EXTRUDED FISH FEED ........... 15

Sefiu A. Bello, Johnson O. Agunsoye, Jeleel A. Adebisi, Suleiman B. Hassan EFFECT OF ALUMINIUM PARTICLES ON MECHANICAL AND

MORPHOLOGICAL PROPERTIES OF EPOXY NANOCOMPOSITES ................ 25

Mirjana Bojanić Rašović POTENTIAL OF INDIGENOUS LACTOBACILLI

AS STARTER CULTURE IN DAIRY PRODUCTS ................................................. 39

Sandra N. Bulut, Vera L. Lazić, Senka Z. Popović, Nevena M. Hromiš, Danijela Z. Šuput INFLUENCE OF STORAGE PERIOD ON PROPERTIES OF

BIOPOLYMER PACKAGING MATERIALS AND POUCHES .............................. 53

Dragana D. Četojević-Simin, Aleksandra S. Ranitović, Dragoljub D. Cvetković, Siniša L. Markov, Milica N. Vinčić, Sonja M. Djilas BIOACTIVITY OF BLACKBERRY (Rubus fruticosus L.) POMACE:

POLYPHENOL CONTENT, RADICAL SCAVENGING, ANTIMICROBIAL AND ANTITUMOR ACTIVITY ............................................................................... 63

Nina M. Đapić, Mihailo S. Ristić CHEMICAL PROFILE OF Taxodium distichum WINTER CONES ......................... 77

Jelena S. Filipović, Vladimir S. Filipović APPLICATION OF TWO RHEOLOGICAL METHODS

FOR FLOUR TESTING TO PREDICT PASTA QUALITY ..................................... 85

Jadranka L. Fraj, Lidija B. Petrović, Jelena R. Milinković, Jaroslav M. Katona, Sandra Đ. Bučko, Ljiljana M. Spasojević PROPERTIES OF WATER IN OIL EMULSIONS (W/O) STABILIZED WITH

MIXTURES OF PGPR AND POLYGLYCEROL FATTY ACID ESTERS ............. 95

Mirela D. Iličić, Spasenija D. Milanović, Katarina G. Kanurić, Vladimir R. Vukić, Svetlana S. Popović, Dajana V. Vukić CONTENT OF SUGAR, ORGANIC ACIDS AND ETHANOL IN

FERMENTED MILK BEVERAGES OBTAINED WITH DIFFERENT TYPES OF KOMBUCHA INOCULUM .................................................................. 109

Milica Ž. Karadžić, Strahinja Z. Kovačević, Lidija R. Jevrić, Sanja O. Podunavac-Kuzmanović CHEMOMETRIC AND QSAR ANALYSIS OF SOME THIADIAZINES

AS POTENTIAL ANTIFUNGAL AGENTS ........................................................... 117

Predrag S. Kojić, Vesna V. Vučurović, Nataša Lj. Lukić, Milica Ž. Karadžić, Svetlana S. Popović CONTINUOUS ADSORPTION OF METHYLENE BLUE DYE

ON THE MAIZE STEM GROUND TISSUE .......................................................... 127

Davor J. Korčok, Nada A. Tršić-Milanović THE APPLICATION OF CLEANING VALIDATION PRINCIPLES

ON DIETARY SUPPLEMENTS PRODUCTION EQUIPMENT ........................... 141

Strahinja Z. Kovačević, Milica Ž. Karadžić, Lidija R. Jevrić, Sanja O. Podunavac-Kuzmanović

MOLECULAR DOCKING ANALYSIS OF NEWLY SYNTHESIZED 2-MORPHOLINOQUINOLINE DERIVATIVES WITH ANTIFUNGAL POTENTIAL TOWARD Aspergillus fumigatus ...................................................... 155

Denis S. Kučević, Dragan M. Glamočić, Snežana J. Trivunović, Igor M. Jajić, Bilјana Č. Perišić, Saša Z. Krstović

VALIDATION AND APPLICATION OF FTIR SPECTROSCOPY IN RAW MILK ANALYSIS .................................................................................... 167

Tatjana A. Kuljanin, Vladimir S. Filipović, Biljana Lj. Lončar, Milica R. Nićetin, Violeta М. Knežević

COMPARISON OF METHODS OF ZETA POTENTIAL AND RESIDUAL TURBIDITY OF PECTIN SOLUTIONS USING CALCIUM SULPHATE/ALUMINIUM SULPHATE AS A PRECIPITANT ........................... 177

Jasmina M. Lazarević, Tatjana A. Tasić, Sanja J. Popović, Vojislav V. Banjac, Olivera M. Đuragić, Bojana M. Kokić, Ivana S. Čabarkapa

CHANGES IN MILK COMPOSITION OF DOMESTIC BALKAN DONKEYS’ BREED DURING LACTATION PERIODS ...................................... 187

Dаvоr М. Lоnčаr, Nаtаšа P. Мilоšеvić, Jasmina S. Vitas, Milica Ž. Karadžić, Lidija R. Jevrić, Goran Benedeković

CHROMATOGRAPHIC LIPOPHILICITY AND PHARMACOKINETIC BEHAVIOR OF SOME NEWLY SYNTHESIZED STYRYL LACTONE STEREOISOMERS .................................................................................................. 197

Nataša Lj. Lukić, Marija Božin-Dakić, Jovana A. Grahovac, Jelena M. Dodić, Aleksandar I. Jokić

MULTI-OBJECTIVE OPTIMIZATION OF MICROFILTRATION OF BAKER’S YEAST USING GENETIC ALGORITHM ..................................... 211

Jelena R. Milinković, Milijana V. Aleksić, Lidija B. Petrović, Jadranka L. Fraj, Sandra Đ. Bučko, Jaroslav M. Katona, Ljiljana M. Spasojević

INTERFACIAL PROPERTIES OF CHITOSAN/SODIUM DODECYL SULFATE COMPLEXES ........................................................................................ 221

Ivana Ž. Mitrović, Jovana A. Grahovac, Jelena M. Dodić, Mila S. Grahovac, Siniša N. Dodić, Damjan G. Vučurović, Vanja R. Vlajkov EFFECT OF AGITATION RATE ON THE PRODUCTION OF

ANTIFUNGAL METABOLITES BY Streptomyces hygroscopicus IN A LAB-SCALE BIOREACTOR ......................................................................... 231

Zorana Z. Rončević, Jovana J. Đuran, Jovana A. Grahovac, Siniša N. Dodić, Aleksandra S. Ranitović, Jelena M. Dodić OPTIMIZATION OF THE MEDIUM COMPOSITION FOR PRODUCTION

OF ANTIMICROBIAL SUBSTANCES BY Bacillus subtilis ATCC 6633 ............. 245

Vanja N. Šeregelj, Gordana S. Ćetković, Jasna M. Čanadanović-Brunet, Vesna T. Tumbas Šaponjac, Jelena J. Vulić, Slađana S. Stajčić

EXTRACTION AND ENCAPUSLATION OF BIOACTIVE COMPOUNDS FROM CARROTS .......................................................................... 261

Igor B. Tomašević, Ilija V. Đekić, Milica G. Aćimović, Slaviša B. Stajić, Vladimir M. Tomović

THE QUALITY DIFFERENCE BETWEEN FRANKFURTERS SEASONED WITH CONVENTIONAL AND ORGANIC SPICES ....................... 275

Vesna Vasić, Aleksandar Jokić, Marina Šćiban, Jelena Prodanović, Jelena Dodić, Dragana Kukić

FLUX INTENSIFICATION DURING MICROFILTRATION OF DISTILLERY STILLAGE USING A KENICS STATIC MIXER ........................... 285

Ana M. Vidaković, Olja Lj. Šovljanski, Aleksandra S. Ranitović, Dragoljub D. Cvetković, Siniša L. Markov

DETERMINATION OF CULTURE MEDIUM COMPOSITION FOR MAXIMIZING THE BIOMASS PRODUCTION OF Pseudomonas stutzeri .......... 295

Vesna M. Vučurović, Vladimir S. Puškaš, Uroš D. Miljić

REMOVAL OF ACRIDINE ORANGE DYE FROM AQUEOUS SOLUTION BY ADSORPTION ONTO DRIED SUGAR BEET PULP ................. 307

Yurii G. Zmievskii, Volodymyr V. Zaharov, Olexandra S. Rudenko, Iryna M. Biletskaya, Valeriy G. Myronchuk

OZONATION OF NANOFILTRATION PERMEATE OF WHEY BEFORE PROCESSING BY REVERSE OSMOSIS ............................................... 315

EDITORIAL POLICY INSTRUCTION FOR MANUSCRIPT PREPARATION

APTEFF, 48, 1-323 (2017) UDC: 634.23+588.144.3]:547.56:615.279 https://doi.org/10.2298/APT1748001A BIBLID: 1450-7188 (2017) 48, 1-13

Original scientific paper

1

PHENOLIC COMPOUNDS, ANTIOXIDANT AND ANTIMICROBIAL PROPERTIES OF THE WILD CHERRY (Prunus avium L.) STEM

Zahida Ademović1*, Snježana Hodžić2, Zarka Halilić Zahirović3, Darja Husejnagić2,

Jasna Džananović3, Broza Šarić-Kundalić3, Jasmin Suljagić1

1 University of Tuzla, Faculty of Technology, Univerzitetska 8, 75000 Tuzla, Bosnia and Herzegovina

2 University of Tuzla, Faculty of Natural Sciences, Univerzitetska 4, 75000 Tuzla, Bosnia and Herzegovina 3 University of Tuzla, Faculty of Pharmacy, Univerzitetska 8, 75000 Tuzla, Bosnia and Herzegovina

The aim of this study was to determine the total phenolic content, evaluate antioxidant propertie and antimicrobial potential, and identify phenolic compounds in alcoholic and aqueous extracts of the wild cherry (Prunus avium L.) stems collected in Bosnia and Herzegovina. Alcoholic extracts had higher contents of phenolic and flavonoid compo-nents, as well as the antioxidant and ferric reducing antioxidant capacity in comparison to aqueous extracts. All extracts were characterized by HPLC analysis. Furthermore, for the first time, the antimicrobial properties of wild cherry stem extracts were evaluated. Quercetin and (+)-catechin were the main compounds identified in the alcoholic extract, followed by chlorogenic acid and rutin. Quercetin was also the major component de-tected in aqueous extracts. Besides, alcoholic extract showed better antibacterial pro-perties against Staphylococcus aureus as a representative gram-positive bacteria than infusion, whereas none of the samples showed antibacterial activity against Escherichia coli and fungus Candida albicans. KEY WORDS: cherry stems, bioactive components, antioxidant activity, antimicrobial

properties

INTRODUCTION Plants have been used as natural sources of medicinal agents from the beginning of human civilization. Medicinal usage of plants has increased in recent years because of their antioxidant, antiviral, antibacterial and antitumor activity. Fruits are considered a natural source of antioxidants, containing anthocyanins and polyphenols, compounds that can reduce the risk of diseases caused by oxidative stress, such as cancer and cardio-vascular diseases (1, 2). The vernacular name "cherry" refers to the fruits of Prunus, an arborescent genus of the Rosaceae family native to Asia and Eastern Europe. The species Prunus avium L. (sweet cherry) is geographically distributed around the world, especially in areas with a moderate climate. Sweet cherries are important commercially as a table

* Corresponding author: Zahida Ademović, University of Tuzla, Faculty of Technology, Univerzitetska 8,

75000 Tuzla, Bosnia and Herzegovina, [email protected]



2

fruit. For medicinal and therapeutic purposes all parts of the plant are used – fruit, stem and bark of the cherry tree. Consumption of sweet cherry has been associated with beneficial health effects (3,4). Cherry fruits exhibit relatively high antioxidant activity, low glycemic response, COX 1 and 2 enzyme inhibition and anti-carcinogenic effects in vitro and in animal experiments (5). Previous studies suggested that the health benefits of cherries are due to their antioxi-dant and anti-inflammatory activities. It has been reported that the consumption of sweet cherry fruits alleviates arthritis and gout-related pain (6). Also, reduction of the prolife-ration of human colon cancer cells has been associated with the consumption of cherry fruits (7). It was reported that extracts of wild cherry bark exhibit anti-proliferative acti-vity in human colorectal cancer cells (8). Polyphenolics, which are plant secondary meta-bolites, are believed to provide those benefits (9). From a therapeutic perspective, the stem is an extremely valuable part of the cherry. In folk medicine, cherry stems have long been recognized as a natural diuretic when pre-pared as tea or dried and encapsulated (10). It is assumed that anti-inflammatory and diuretic properties result from natural antioxidants (flavonoids). In a cross-sectional questionnaire study conducted in Turkey, in the population which have taken up alterna-tive treatments, 6.2% of hypertensive patients were consuming cherry stems in the form of tea as a part of their treatment for diuresis (11). The diuretic activity of powdered cherry stem in 13 healthy volunteers was also evaluated (12). The study found that mean levels of urine calcium, sodium, chloride, and urine volume increased, but the amount of urine potassium and urine osmolality did not change after administration of cherry stem. Although authors observed no adverse reaction, they stressed that because of rising calcium excretion, it should be used with caution in patients with urolithiasis. Tea preparation of the wild cherry stems has been used since centuries in traditional medicine of Bosnia and Herzegovina as diuretic agent that helps promoting proper kidney function. It is also used for breaking down and clearing stones from the bladder and kidneys (13). In this study, the name "wild cherry" is applied to species of Prunus avium growing in wildness. It is widely found in various regions of Bosnia and Herzegovina. It occurs only as a single tree, endangered and protected by law. The fruits of wild cherry are clearly smaller and fewer in number than by the cultivated varieties, red to black purple in color. The fruit is slightly acidic and bitter compared to the domestic varieties. Most of the previous research has been conducted on cherry fruits, whereas cherry stems, which have been traditionally used in folk medicine as sedatives, diuretics and draining, have not been fully tested and characterized in respect to their bioactive ingre-dients and bioactive properties (12). Chemical composition and antioxidant and antitumor effects of the sweet cherry (P. avium L.) fruits and stems were recently compared (14). The authors reported that the stem extract revealed higher antioxidant potential than the tested extract from fruits, whereas the fruits extract was the only one showing antitumor potential. The aim of this study, in addition to the quantification of the total phenolic content, was also aimed determine the flavonoid content and characterize phenolic composition of alcoholic and aqueous extracts from the wild cherry stem. Additionally, antioxidant pro-



3

perties were determined using 2,2-diphenyl-1-picrylhydrazyl (DPPH) method and Ferric Reducing Antioxidant Power (FRAP) assay. In vitro antimicrobial activities against gram-positive bacteria Staphylococcus aureus, gram-negative bacteria Escherichia coli and fungus Candida albicans were evaluated. To our best knowledge, this is the first study on the evaluation of antimicrobial activity of the Prunus avium stem. Based on the results and knowledge gained from this project the possibility of further testing of bioactive ingredients of wild cherry stems, isolation and characterization of individual components and testing their effects on certain physiological processes will be provided.

MATERIALS AND METHODS

Plant material and preparation of extracts Wild cherry stems were collected from trees growing in natural habitat from three locations in Northern Bosnia in June 2015 and the biological authentication was carried out by Prof. dr. Senida Osmanović, biologist at the Faculty of Science, University Tuzla. Voucher specimens were deposited at the Herbarium of the Department of Pharma-cognosy, University of Tuzla. The samples were air dried and stored in desiccator pro-tected from light. Before extraction, stems were ground to a powder and mixed to obtain a homogenous sample. For preparation of the alcoholic extract, 5 g of the stem powder was extracted in an ultrasonic bath with 50 mL of ethanol at 25 °C. After 30 min the so-lution was filtered and the residue on the filter paper was extracted again in the same way. The combined extracts were evaporated at 40 °C using rotary evaporator (Büchi R-210, Switzerland) under reduced pressure to remove ethanol. The extraction yield was 19±2.5%. For the aqueous extract, 5 g of cherry stems were added to 50 mL of boiling distilled water and left to stand at room temperature for 10 min. The solution was filtered and evaporated. The extraction yield was 8.8±1.4%. Extraction was repeated at least three times, and the results are given as mean ± standard deviation (SD).

Standards and reagents Phenolic compound reference standards, (+)-catechin and rutin trihydrate, were pur-chased from Sigma-Aldrich (Germany), quercetin dihydrate from HWI Analytik GmbH (Germany) and chlorogenic acid from Aeros Organic (USA). 2,2-Diphenyl-1-picrylhyd-razyl (DPPH) and Folin-Ciocalteu reagent were obtained from Sigma-Aldrich (Germa-ny). 2,4,6-Tris(2-pyridyl)-s-triazine (TPTZ) ((≥99.9%) was obtained from Himedia Labo-ratories (India) and Methanol HPLC grade (≥99.9%) from Roth (Germany).

Microbial strains The antimicrobial activity of alcoholic and aqueous extracts was evaluated using labo-ratory control strains obtained from the American Type Culture Collection: gram-positive bacteria Staphylococcus aureus ATCC 6538 P, gram-negative bacteria Escherichia coli



4

ATCC 10536 and one fungus Candida albicans ATCC 10231. Positive controls were Ka-namycin for S. aureus and Chloramphenicol for E. coli. Nystatin was positive contol for C. albicans.

Evaluation of bioactive properties and phenolic compounds Total phenolic content (TPC) in the alcoholic and aqueous extracts was determined spectrophotometrically after reaction with the Folin-Ciocalteu phenol reagent (15). The extracts were dissolved in methanol to a final concentration of 0.25 mg/mL. 50 l of ex-tracts, 450 l deionized water and 2.5 mL of Folin-Ciocalteu reagent are mixed and in-cubated for 5 min. 2 mL of 7% sodium bicarbonate solution was added, filled with water up to 100 mL and incubated for 1.5 hours at 30 °C. Absorbance of the resulting blue co-lored liquids was measured at 765 nm using a Shimadzu UV-mini-1240 UV-Vis Spec-trophotometer. Quantitative analysis was performed based on the standard calibration curve of gallic acid in methanol. The concentrations of gallic acid in the solution from which the curve was prepared were 50, 100, 150, 250 and 500 mg/L (y=0,0014x+0.0244, R2=0.9912). The result was expressed as mg of gallic acid equivalent per gram of dry weight of sample (mg GAE/g). Total flavonoids content (TFC) was determined by 24 h precipitation reaction with formaldehyde (16). In a 50 mL flask 5 mL of the extract (1 mg/mL) in methanol, 5 mL 1:4 HCl and 2.5 mL of formaldehyde solution were added, incubated for 24 hours at room temperature and filtered. The remaining phenolic compounds, evaluated as non-fla-vonoid content (TNFC), were determined according to the previously mentioned proce-dure for TPC determination with the Folin-Ciocalteu method. TFC was calculated as subtraction of TPC and TNFC. Determination of antioxidant capacity using DPPH method. 2,2-diphenyl-1-picryl-hydrazyl (DPPH) method was performed according to the method described earlier (17). Samples were dissolved in methanol at a final concentration of 1 mg/mL, filtered and kept in the dark. The reaction solution was prepared by mixing 10-150 L of the alcoho-lic extract and 50-1000 L aqueous extracts with 3 mL of 0.135 mM solution of DPPH, and filled by methanol to 2 mL. The mixture was incubated for 30 min in the dark at room temperature. The absorbance is measured against the blank (methanol) at 517 nm. For the DPPH control solution, 3 mL of 0.5 mM DPPH solution was dissolved in 1 mL of methanol. The radical scavenging effect (%) or percent inhibition of DPPH radical was calculated according to the equation: [(Acontrol-Asample)/Acontrol] x100, where Asample is the absorbance of the solution containing the sample at 515 nm and Acontrol is the absor-bance of the DPPH solution. The results are expressed as the IC 50 value (mg/mL) or the concentration of extract that caused 50% neutralization of DPPH radicals. The Ferric Reducing Antioxidant Power (FRAP) assay. The method is based on the ability of the extract to reduce Fe3+ ions to Fe2 + ions. The resulting Fe2+ ions form with the reagent 2,4,6-Tris(2-pyridyl)-s-triazine (TPTZ) a blue colored complex, which rea-ches the absorption maximum at 593 nm. The reaction is performed in an acidic medium in order to maintain good solubility of iron. The determination of ferric reducing antioxi-dant power or ferric reducing ability (FRAP assay) was performed as described earlier



5

(18). Briefly, the working FRAP reagent was prepared freshly by mixing 10 volumes of 300 mM acetate buffer pH 3.6, with 1 volume of 10 mM TPTZ (2,4,6-tri(2-pyridyl)-s-triazine) in 40 mM HCl and with 1 volume of 20 mM ferric chloride. A volume of 0.1 mL of the sample of the concentration 1 mg/mL and 3 mL of fresh prepared FRAP rea-gent were added, and after 30 min the absorbance was measured at 593 nm compared to the blank (3 mL of FRAP reagent + 0.1 mL of solvent that was used to dissolve the samp-les). Aqueous solutions of FeSO4x7H2O (200–1000 M) were used for the calibration (y=0.0007x+0.117, R2=0.9947), and the results were expressed as FRAP value mg Fe2+/g of sample. High Performance Liquid Chromatography (HPLC). Phenolic compounds were de-termined by High-Performance Liquid Chromatography (HPLC, Shimadzu LC-20AT, Ja-pan) gradient controlled system consisting of the solvent delivery system, autosampler, pump, and UV-VIS detector. The separation method for the analysis was described pre-viously (19). The phenolic components were detected at 280 nm. The separation was car-ried out using a Selectra C18 column (4.6x250 mm, 5 m) at room temperature. The elu-tion solvent A was 0.01 M phosphoric acid and solvent B was 100% methanol. The fol-lowing binary gradient was applied: the gradient was started with 5% B in order to reach 50% B at 10 min, 70% B at 15 min, 80% B at 20 min and 100% B at 25 min. The inject-tion volume was 20 l and flow rate 1 mL/min. Samples for HPLC analysis were filtered through membrane filter prior to injection. The phenolic compounds were identified by comparing their retention times with those obtained for the standard compounds. Antimicrobial testing. The antibacterial activity of the cherry stems alcoholic and aqueous extract was tested by the agar diffusion method according to the American Cli-nical Laboratory Standards Institute (CLSI) (20) using 100 L of suspension of the tested microorganisms containing 2.0 × 106 colony forming units (cfu/mL). Mueller-Hinton agar (15 mL), sterilized in a flask and cooled to 45–50 °C, was distributed to sterilized Petri dishes with a diameter of 9 cm. The sterile swab was used to transfer the bacterial suspension and inoculate the bacteria on the surface of Mueller-Hinton agar. The wells with a diameter of 10 mm were cut with the sterile stainless steel borer down to the plas-tic into Mueller-Hinton agar plates and then filled with a volume of 25 and 50 L of the extracts samples dissolved in water at a concentration of 50 mg/mL. Two wells were pre-pared per plate. The Petri dishes were kept incubated at 37 °C for 24 h. The antimicrobial activity was measured on the basis of the diameter of the growth inhibition as follows: (<10 mm) – no antimicrobial activity; (10-15 mm) – weak antimicrobial activity; (16-20 mm) – moderate antimicrobial activity; (>20 mm) – certain antimicrobial activity (21). Kanamycin, chloramphenicol and nystatin (10.0 mg/mL) were used as reference.

Statistical analysis

All experiments were carried out in triplicates. The results are presented as mean ± standard deviation and analyzed by SPSS v. 20.0 program.



6

RESULTS AND DISCUSSION

The total phenolic content of the wild cherry stem using the Folin-Ciocalteu assay is expressed in terms of gallic acid equivalent (GAE). The determination was based on the standard calibration curve of different concentrations of gallic acid and expressed as mg of GAE/g of extract. The results in Table 1 show that the alcoholic extract had a higher total phenolic content (121.3 mg GAE/g) than the aqueous extract (74.1 mg GAE/g). The content of flavonoid was also higher for the alcoholic extract, 26.9 mg GAE/g, compared to 5.9 mg GAE/g for the aqueous extract. The value for total phenolic content of the alco-holic extract was in accordance with the result reported by Bursal et al. (22) (146 mg GAE/g), whereas the value for the aqueous extract was much lower than those reported by these authors (118 mg GAE/g).

Table 1. Total phenolic content, flavonoid content and antioxidant activity of the alcoholic and aqueous extract of wild cherry stems

Sample Total phenolics

(mg GAE/g)a Total flavonoid

(mg GAE/g)a DPPH

IC50% (mg/mL)b FRAP

(mg Fe2+/g)c

Alcoholic extract

121.3±9.7 26.9±1.4 29.75±0.21 194.05±8.3

Aqueous extract

74.1±7.4 5.9±2.2 227.96±3.15 63.66±5.7

Each value in the table represents the mean ± SD (n = 3) a GAE - gallic acid equivalent b IC50% - concentration of the sample that causes 50% neutralization of DPPH radicals, higher values corres-

pond to lower antioxidant potential c FRAP - ferric reducing antioxidant power

The DPPH radical scavenging ability is one of the best-known and frequently em-ployed methods for evaluating antioxidant activity. DPPH is a stable free radical because the spare electron is delocalized over the whole molecule. The donation of H+ to the DPPH radicals changes color of the solution from violet to pale yellow. This assay deter-mines the scavenging of the stable DPPH radicals by antioxidants compounds present in the extracts. The IC50 value was determined from the plotted graph of scavenging acti-vity against various concentrations of extracts and infusion, and it is defined as the con-centration of antioxidant necessary to decrease the initial DPPH radicals concentration by 50%. The results presented in Table 1 show the higher rate of DPPH scavenging activity of the alcoholic stem extract as compared to the aqueous stem extract. This may be explained by the presence of a higher content of phenolic compounds in the alcoholic extract. Both samples were able to reduce the stable violet DPPH radical to the yellow DPPH-H, reaching 50% of reduction with the IC 50 value of 39.58 g/mL for the alco-holic extract and 227.96 g/mL for the aqueous extract. Bursal et al (22) reported that both extracts (ethanolic and aqueous) of the cherry stem, showed significant DPPH radi-cal scavenging activity with similar IC 50 values of 17.36 and 23.38 g/mL, respectively. Contrary to that, Bastos et al. (14) reported much higher DPPH values for both extracts



7

with the values of 360 g/mL 630 g/mL for the ethanolic and aqueous extract, res-pectively. These differences could be explained by the different extraction methodologies applied in the studies.

Figure 1. DPPH scavenging activity expressed as the inhibition of DPPH radicals (%) of the alcoholic (a) and aqueous extract (b)

The ferric reducing ability of the two extracts was also tested. The presence of anti-oxidants in the sample would reduce the ferric ion (Fe3+) to the corresponding ferrous ion (Fe2+) by donating an electron. The reducing power of alcoholic and aqueous extracts are expressed as mg Fe2+ per gram of dry extract, and the results are shown in Table 1. The measured reducing power of alcoholic extract was 194.05 mg Fe2+/g dry sample, which is

y = 0,6524x + 30,588R² = 0,9917

0

20

40

60

80

100

120

0 20 40 60 80 100 120

IC %

mg/ml

a)

y = 0,2047x + 3,3372R² = 0,9931

0

10

20

30

40

50

60

70

0 100 200 300 400

IC %

mg/ml

b)



8

significantly higher than that of the aqueous extract (63.66 mg Fe2+/g). The alcoholic stems extracts exhibited higher antioxidant and ferric reducing power in comparison to the aqueous extract. This is related to the higher phenolic compounds concentration found in the alcoholic extract. A composition of the contents of phenolic compounds are of great interest to scien-tists, manufacturers and consumers due to their influence on product quality and their protective and preventive roles in the pathogenesis of certain types of cancer and several other chronic diseases. In the literature, there are several reports on the identification and quantification of phenolic compounds in P. avium L. fruits. The most commonly found compounds are phenolic acids (neochlorogenic, chlorogenic and p-coumaroylquinic acids), anthocyanins, flavonols (rutin) and flavan-3-ols (catechin, epicatechin) (23, 24). To our knowledge, only two reports have been published on chemical composition of cherry stem. Bursal et al. (22) studied phenolic acids in ethanol and aqueous extracts of cherry stems by LC-MS/MS and found pyrogallol, ferulic acid, p-coumaric acid, gallic acid, ascorbic acid and p-hydroxybenzoic acid. Bastos et al. (14) compared the HPLC phenolic profile of P. avium L. fruits and stems. They detected more phenolic compounds in the stem than in fruit. In this study, the analysis of phenolic compounds was performed using liquid gradient method developed by Escrapa and Gonzales (19). The HPLC phenolic profiles of the alcoholic and aqueous extracts of cherry stems were recorded at 280 nm. Figure 2 shows the chromatograms obtained for the mixture of four commercially available phenolic compounds standards (a), alcoholic extract (b) and aqueous extract (c). The retention ti-mes (RT) for the phenolic standard shown in the chromatogram in Figure 2a were as follows: (+)-catechin RT 11.5 min, chlorogenic acid RT 12.0 min, rutin RT 16.0 min, and quercetin RT 19.6 min. The peaks in the chromatograms of alcoholic and aqueous extracts could be tentati-vely assigned by applying the optimised chromatographic gradient elution, consisting of four elution steps. In the first gradient step (0-10 min) arbutin and gallic acid are eluted, in the second (10-15 min) catechins, chlorogenics and caffeic acids, in the third one (15-20 min) flavonoids that included flavan-3-ols, flavonols, flavones and isoflavones, whe-reas in the last one (20-25 min) are aglycones. In chromatograms of alcoholic extracts (Figure 2b) no peaks that could be related to hydroxybenzoic acids were recorded. In the second gradient, (+)-catechin and chlorogenic acid were identified based on their RT by comparison with the standards. In the third gradient, diverse flavonoids were eluted, where quercetin, as most intense peak, and rutin were identified. Also, the peak at the RT of 20.5 could be compared to the Escrapa and Gonzales (13) finding, and could belong to kaempferol. The chromatogram of the aqueous extract (Figure 2c) showed only one high-intensity peak, which based on the RT of standards was assigned to quercetin. Also, some peaks with very low intensity were detected. Bastos et al. (14) detected twenty three compounds in alcoholic extract of stems. By comparison with commercial standards they identified catechin, quercetin, rutin and kaempferol, but no chlorogenic acid. In aqueous stem extract they recorded fifteen peaks, where large amount of catechin, some quercetin, but no rutin was identified. The remaining compounds were tentatively assigned.



9

a) b)

c)

Figure 2. HPLC chromatograms at 280 nm obtained for phenolic compounds of: (a)

polyphenolic compounds standards available commercially, (b) alcoholic extract and (c) aqueous extract. Peaks: (1) (+)-catechin, (2) chlorogenic acid, (3) rutin, (4) quercetin

The difference in the chromatograms of alcoholic and aqueous extracts is due to the solvent and techniques used for the extract preparation. Ethanol as an organic solvent can better dissolve active organic components required for biological activity of the sample. Also, ultrasound extraction technique leads to the disruption of the cell wall matrix and releases components such as phenolic compounds into the extraction solvent. Many phytochemical preparations with high flavonoid content have been reported to exhibit antimicrobial activity (25). The results of this study suggest that cherry stems have strong antioxidant potential due to the presence of large quantities of phenols and flavonoids. Therefore, preliminary study on antimicrobial activity of alcoholic and aqueo-us extracts obtained from the wild cherry stems was performed against gram-positive bacteria S. aureus, gram-negative bacteria E. coli and one fungus C. albicans. Positive controls were kanamycin for S. aureus, chloramphenicol for E. coli and nystatin for C. albicans. E. coli is a common pathogenic bacteria for urinary tract infection and S. aureus is the cause of pneumonia and several infections in the gut or urinary tract. The alcoholic and aqueous extracts were prepared by dissolving extracts in water in concentration of 50 mg/mL. As can be seen from Table 2, a volume of 25 L (1,25 mg)



10

of the alcoholic extract showed a weak inhibitory effect on the gram-positive S. aureus strain, whereas 50 L (2.5 mg) showed a moderate inhibitory effect on this strain. How-ever, only a higher volume of the aqueous extract of 50 L (2.5 mg) showed a weak inhibitory effect on the S. aureus strain. None of the tested samples showed inhibitory activity on the gram-negative bacteria E. coli and fungus C. albicans.

Table 2. Antimicrobial effects of the alcoholic and aqueous extracts obtained from the wild cherry stems (P. avium)

Samples (concentration 50 mg/mL)

Volume applyed (L)

Staphylococcus aureus

Escherichia coli

Candida albicans

Alcoholic extract 25 12±0.70 <3 <2 50 16±0.30 <5 <2

Aqueous extract 25 <6 <3 <2 50 11±0.40 <5 <2

Positive controls (10 mg/mL) Kanamycin 25 >20 - - Chloramphenicol 25 - >20 - Nystatin 25 - - >20

(-) – no antimicrobial activity detected (<10 mm) (+) – weak antimicrobial activity (inhibition zone 10-15 mm) (++) – moderate antimicrobial activity (inhibition zone 16-20 mm) (+++) – certain antimicrobial activity (inhibition zone ˃20 mm)

From this study it comes clear that the solvent used for extraction influenced the degree of antibacterial activity of the extracts. Ethanol as solvent and ultrasound extrac-tion showed a higher extraction efficiency than infusion with water, which is evident from the total phenolic content and HPLC. Therefore, the alcoholic extract showed a higher antimicrobial activity than the aqueous one. The study conducted by Rauha et al. (26) on antimicrobial effects of different phenolics as pure substances showed that the growth of S. aureus was inhibited very effectively by quercetin and kaempferol, whereas these components showed only slight antibacterial activity against E.coli. Catechin and rutin failed to show any activity against S. aureus and E.coli. The yeast C. albicans. was resistant to all of these compounds. It is evident that the stem extracts did not show as marked inhibition as the pure flavonoid compounds. It should be also noted that the assays relying on diffusion of tested extract may not give a reliable quantitative measure of antibacterial activity because a potent antibacterial flavonoid may have a low rate of diffusion (27). More studies are needed to define the antimicrobial properties of stems.

CONCLUSION

In traditional medicine in Bosnia and Herzegovina, tea preparation of wild cherry stems is used because of its assumed diuretic properties. The aim of this study was to evaluate antioxidant and antimicrobial properties and identify phenolic compounds in



11

stems of wild cherry, Prunus avium. The analysis of total phenolic and flavonoid content and free radical scavenging activity showed that the extract from the stems of wild cherry fruits can be a potent source of natural antioxidants. The antioxidant and antiradical acti-vities of the aqueous extract were found to be lower than that of the alcoholic extract. The results of in vivo studies suggest that the alcoholic extract was slightly active against gram-positive bacteria, whereas the aqueous extract showed weak antibacterial proper-ties. None of the samples showed antibacterial properties against gram-negative bacteria.

Acknowledgеment This work was financially supported by the Federal Ministry of Education and Scien-ce, Bosnia and Herzegovina.

REFERENCES 1. Kaur, C,; Kapoor, HC. Antioxidants in fruits and vegetables – the millennium’s

health. Int. J. Sci. Technol. 2001, 36, 703-725. 2. Ballistreri, G.; Continella, A.; Gentile, A.; Amenta, M.; Fabroni, S.; Rapisarda, P.

Fruit quality and bioactive compounds relevant to human health of sweet cherry (Prunus avium L.) cultivars grown in Italy. Food Chem. 2013, 140, 630-638.

3. Wang, H.; Nair, M.G.; Strasburg, G.M.; Booren, A.M.; Gray. J.L. Antioxidant Polyphenols from Tart Cherries (Prunus cerasus). J. Agricult. Food Chem. 1999, 47, 840-844.

4. Serra, A.T.; Duarte, R.O.; Bronze, M.R.; Duarte, C.M.M. Identification of bioactive response in traditional cherries from Portugal. Food Chem. 2011, 125, 318-325.

5. McCune L.M.; Kubota, C.; Stendell-Hollis, N.R.; Thomson, C.A. Cherries and health: A review. Crit. Rev. Food Sci. Nutr. 2011, 51, 1-12.

6. Zhang, Y.; Neogi, T.; Chen. C.; Chaisson, C.; Hunter, D.; Choi, H.K. Cherry con-sumption and decreased risk of recurrent gout attacks. Arthritis Rheumatol. 2012, 64, 4004-4011.

7. Serra, A.T.; Seabra, I.J.; Braga, M.E.M.; Bronze, M.R.; de Sousa, H.C.; Duarte, C.M.M. Processing cherries (Prunus avium) using supercritical fluid technology. Part 1: Recovery of extract fractions rich in bioactive compounds. J. Supercrit. Fluids 2010, 55, 184-191.

8. Yamaguchi, K.; Liggett, J.L.; Kim, N-C.; Baek, S.J. Anti-proliferative effect of horehound leaf and wild cherry bark extracts on human colorectal cancer cells. Oncol. Rep. 2006, 15, 275-281.

9. Kelebek, H.; Selli, S. Evaluation of chemical constituents and antioxidant activity of sweet cherry (Prunus avium L.) cultivars. Int. J. Food Sci. Technol. 2011, 46, 2530-2537.

10. Tierra, M. The way of herbs; Pocket books: New York, 1998. 11. Balat, A. From past to present: Traditional herbs used in the treatment of nephrologic

diseases in southeast Turkey. J. Nephrology 2013, 26, S187.



12

12. Hooman, N.; Mojab, F.; Nickavar, B.; Pouryousefi-Kermani, P. Diuretic effect of powdered Cerasus avium (cherry) tails on healthy volunteers. Pak. J. Pharm. Sci. 2009, 22, 381-383.

13. Saric-Kundalic, B.; Dobes, C.; Klatte-Asselmeyer, V.; Saukel. J. Ethnobotanical survey of traditionally used plants in human therapy of east, north and north-east Bosnia and Herzegovina. J. Ethnopharmacol. 2011, 133, 1051-1076.

14. Bastos, C.; Barros, L.; Duenas, M.; Calhelha, R.C.; Joao, M.; Queiroz, R.P.; Santos-Buelga, C.; Ferreira, I.C.F.R. Chemical characterisation and bioactive properties of Prunus avium L.: The widely studied fruits and the unexplored stems. Food Chem. 2015, 173, 1045-1053.

15. Tawaha, K.; Alali, F.Q.; Gharaibeh, M.; Mohammad, M.; El-Elimat, T. Antioxidant activity and total phenolic content of selected Jordanian plant species. Food Chem. 2007, 104, 1372-1378.

16. Giusti, M.M.; Wrolstad, R.E. Anthocyanins. Characterization and measurement with UV-visible spectroscopy. In: Current protocols in food analytical chemistry; Wrol-stad, R. E., Ed; Wiley: New York, 2001; pp Fl.2.1 - F1.2.13.

17. Benvenuti, S.; Pellati, E.; Melegar, M.; Bertelli, D. Polyphenols, anthocyanins, ascor-bic acid and radical scavenging activity of Rubus, Ribes, and Aronia. J. Food Sci. 2004, 69, 164-169.

18. Jimenez-Aspee, F.; Quispe, C.; Soriano, M.C.; Gonzalez, J.F.; Hueneke, E.; Theodu-loz, C.; Schmeda-Hirschmann, G. Antioxidant activity and characterization of consti-tuents in copao fruits (Eulychnia acida Phil., Cactaceae) by HPLC–DAD–MS/MSn. Food Res. Int. 2014, 62, 286-298.

19. Escarpa, A.; Gonzales, M.C. Optimization strategy and validation of one chromato-graphic method as approach to determine the phenolic compounds from different sources. J. Chromat. A 2000, 897, 161-170.

20. CLSI Performance Standards for Antimicrobial Susceptibility Testing; 20th Informa-tional Supplement. Wayne, P.A: Clinical and Laboratory Standards Institute docu-ment 2010, M100-S20, 30.

21. Pirvu, L.; Hlevca, C.; Nicu, I.; Bubueanu, C. Comparative analytical, antioxidant and antimicrobial activity studies on a series of vegetal extracts prepared from eight plant species growing in Romania. J. Planar Chromat. 2014, 275, 346-356.

22. Bursal, E.; Koeksal, E.; Guelcin, I.; Bilsel, G.; Goeren, A.C. Antioxidant activity and polyphenol content of cherry stem (Cerasus avium L.) determined by LC–MS/MS. Food Res. Int. 2013, 51, 66-74.

23. Serrano, M.; Guillen, F.; Martinez-Romero, D.; Castillo, S.; Valero, D. Chemical Constituents and Antioxidant Activity of Sweet Cherry at Different Ripening Stages. J. Agricult. Food Chem. 2005, 53, 2741-2745.

24. Usenik, V.; Fabcic, J.; Stampar, F. Sugars, organic acids, phenolic composition and antioxidant activity of sweet cherry (Prunus avium L.). Food Chem. 2008, 107, 185-192.

25. Cushnie, T.P.; Lamb, A.J. Antimicrobial activity of flavonoids. Int. J. Antimicrob. Agents 2005, 26, 343-356.



13

26. Rauha, J.P.; Remes, S.; Heinonen, M.; Hopia, A.; Kaehkoenen, M.; Kujala, T.; Pihlaja, K.; Vuorela, H.; Vuorela, P. Antimicrobial effects of Finnish plant extracts containing flavonoids and other phenolic compounds. Int. J. Food Microbiol. 2000, 56, 3-12.

27. Zheng, W.F.; Tan, R.X.; Yang, L.; Liu, Z.L. Two flavones from Artemisia giraldii and their antimicrobial activity. Planta Med. 1996, 62, 160-162.

ФЕНОЛНА ЈЕДИЊЕЊА, АНТИОКСИДАТИВНЕ И АНТИМИКРОБНЕ ОСОБИНЕ ПЕТЕЉКИ

ДИВЉЕ ТРЕШЊЕ (Prunus avium L.) Захида Адемовић1, Сњежана Хоџић2, Зарка Халилић Захировић3, Дарја Хусејнагић2,

Јасна Џановић5, Броза Сарић Кундалић3, Јасмин Шуљагић1

1 Универзитет у Тузли, Технолошки факултет, Универзитетска 8, 75000 Тузла, Босна и Херцеговина

2 Универзитет у Тузли, Факултет природних наука, Универзитетска 4, 75000 Тузла, Босна и Херцеговина 3 Универзитет у Тузли, Фармацеутски факултет, Универзитетска 8, 75000 Тузла, Босна и Херцеговина

Циљ овог рада био је да се одреди укупни садржај фенола, процене антиокси-дативне особине и антимикробни потенцијал, као и да се идентификују фенолна једињења у алкохолним и воденим екстрактима петељке дивље трешње (Prunus avium L.) сакупљеним у Босни и Херцеговини. Алкохолни екстракти су садржавали више фенолних и флавоноидних компоненти у поређењу са воденим екстрактима. Такође, утврђено је да су имали виши антиоксидативни капацитет и способност ре-дукције фери-јона. Сви екстракти су окарактерисани HPLC анализом. Поред квер-цетина и (+)-катехина, као доминантних једињења, у алкохолном екстракту иденти-фиковани су хлорогенска киселина и рутин. Кверцентин је, као доминантна компо-нента, идентификован и у воденом екстракту. По први пут испитане су и антимик-робне особине екстраката петељки дивље трешње. Алкохолни екстракт је у поређе-њу са воденим екстрактом, показао боље антимикробне особине на Staphylococcus aureus, као представника грам-позитивних бактерија, док ниједан од испитиваних узорака није испољио активност на бактерију Escherichia coli и гљивицу Candida albicans. Кључне речи: Петељка трешње, биоактивне компоненте, антиоксидантна актив-

ност, антимикробне особине

Received: 16 July 2017 Accepted: 29 September 2017

APTEFF, 48, 1-323 (2017) UDC: 639.3.043:[628.1.032:631.81 https://doi.org/10.2298/APT1748015B BIBLID: 1450-7188 (2017) 48, 15-24


15

CONDUCTOMETRIC METHOD FOR DETERMINING WATER STABILITY AND NUTRIENT LEACHING OF EXTRUDED FISH FEED

Vojislav V. Banjac1*, Radmilo R. Čolović1, Lato L. Pezo2, Dušica S. Čolović1,

Jasmina M. Gubić1, Olivera M. Đuragić1

1 University of Novi Sad, Institute of Food Technology, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia

2 University of Belgrade, Institute of General and Physical Chemistry, Studentski Trg 12-16, 11000 Belgrade, Serbia

Water stability of eight samples of extruded salmon feeds was first determined by applying two gravimetric methods developed by the authors: gravimetric static and wet sieving method. Then, the conductometric method, primarily developed for investigation of nutrient leaching of feed into the water by the authors, was used for each sample. The aim of this study was to evaluate the potential of the conductometric measurement as a technique for determining water stability of extruded fish feed. In order to find any correlation between the results of two gravimetric tests and conductometric method, correlation analysis was employed. The results of static and wet sieving method were expressed as water stability index, which was expressed as the percent of remained dry matter of sample after being disintegrated in the water. The results of conductometric method were shown as conductivity curves for each sample, giving the insight in rate of nutrient leaching during the time. The obtained values of water conductivity showed no significant (p < 0.05) correlation with the results of static water method, while there was a negative significant (p < 0.05) correlation with the results of wet sieving method during first four hours of pellets soaking in water. The highest correlation coefficients were obtained within the first hour of conductivity measurement, demonstrating that proposed conductometric method had a potential to be applied as a rapid and simple method for determination and relative comparison of salmon feed water stability. KEY WORDS: fish feed, water stability, nutrient leaching, conductometry

INTRODUCTION Commercial fish feed is almost solely produced by an extrusion process, mostly in the form of pellets, resulting in products high in energy, with good nutritional value and high physical quality. Many different ingredients of plant and animal origin are included in fish diet formulations, in which fishmeal has been considered as a traditional protein source for both carnivorous and omnivorous species (1). Increased demand in fishmeal

* Corresponding author: Vojislav V. Banjac, University of Novi Sad, Institute of Food Technology, Bulevar

Cara Lazara 1, 21000 Novi Sad, Serbia, e-mail: [email protected]



16

by aquaculture industry together with finite marine resources for its production, increase the price of fish meal and negatively influence sustainability of the aquaculture sector (2). Thus, alternative proteins from plant, animal and microbiological sources for fishmeal partly substitution in feed formulations have been main scope of in the research field of new fish feed development (3). Physiochemical properties of ingredients, as well as pro-duction conditions affect the physical quality of extruded feed (4). Hence, the determina-tion of fish feed physical characteristics should be done, combined with determination of nutritional characteristics, to evaluate potential of new ingredient (3). An important physical property of feed for aquatic species is its water stability, defi-ned as the retention of the pellet physical integrity with minimal disintegration and nutrient leaching while it is immersed in the water and until it is consumed by animals (5). Fish feed comprises of macronutrients such as proteins, fats and carbohydrates, as well as of micronutrients such as vitamins and minerals. The losses of these valuable nutrients in the water should be minimized. Aquatic feed water stability should be taken into consideration, no matter whether instant feeders, such as finfish, or slow eating aquatic animals, such as shrimp, are fed. The criteria for the shrimp feed water stability evaluation are more strict, since, the shrimp feed has to be soaked in water for hours with least possible leaching of nutrients (3, 5). For Salmonide, such as salmon and trout, that require slow sinking pellets, water stability may be an important characteristic for pre-diction of feed degradation, not only in the water but in the gastrointestinal tract as well, since low water stability of trout feed was reported to be a potential factor for oil-belching and the abdominal distension syndrome in rainbow trout, causing oil and water separation and accumulation of free oil in the stomach of it (6). The study of Aas et al. (7) showed similar pattern where the feed with low water stability caused significantly higher level of pellet disintegration in the gastric tract of rainbow trout, which induced risk of oil-belching due to the presence of free oil and water in the stomach of the fish. This study also implied that nutritional value of a fish feed is affected by its physical proper-ties. In research by Oehme et al. (4), only long-term water stability, out of several deter-mined physical feed properties, positively affected the feed intake of Atlantic salmon, but it was concluded that further research was needed. The method for determining water stability of aquatic feed should be fast and simple, and should provide practical, accurate and reproducible results (5). There are different methods for determining water stability based on soaking feed pellets into static water (8-10) or in water with additional agitation (6), in which water stability was expressed as a percentage of remained dry matter of starting feed. In order to develop a standard method for determining water stability of shrimp feed, Obaldo et al. (5) tested three different methods with and without water agitation, in combination with water salinity and tempe-rature that were taken into accounts. Baeverfjord et al. (6) developed water stability test for determining stability of trout feed, in which mild physical stress in the form of shaking was applied to disintegrate feed while it is incubated in water at constant tempe-rature. This method clearly discriminated between extruded pellets with different water stability and additionally, obtained values of stability in water correlated very well with the consistency of the stomach contents of feeding fish. For determining water stability of trout feed in situ a method was used in which feed has been introduced in fish tanks, with



17

no fish inside, using similar water flow and temperature as employed in the digestibility experiment (11). Nutrient leaching of feed into the water during immersion was determi-ned by applying different techniques such as gravimetric (8, 9) and spectrophotometric (12) methods. The results of gravimetric determination were presented as remained pro-tein and fat content after different times of immersion, while spectrophotometry enabled construction of a curve as a function of increased nutrient concentration and time. The aim of this work was to further investigate potential of conductometric method for determining water stability of extruded fish feed and nutrient leaching of fish feed, that was previously developed and proposed by the authors (13). The results of conducto-metric measurements were compared with the water stability results of feed samples ob-tained by two gravimetric water stability tests, which were also previously developed by the authors (13). Correlations between the results were found, and the potential of conductometry as a technique for water stability determination was evaluated.

EXPERIMENTAL

Fish feed samples Eight samples of salmon feeds with different water stability were produced in a pilot plant of Feed to Food Center (Institute of Food Technology, Novi Sad, Serbia) by extru-sion processing. Diets formulation and extrusion parameters were changed during experi-mental production in order to obtain differences in the physical quality of the pellets. The diameter and the length of the pellets were targeted at 9 mm and 8 mm respectively. Two gravimetric methods for determining pellet water stability were used first: static method and wet sieving method (13). Conductometric method was then employed in or-der to determine nutrition leaching of feeds. The initial dry matter (DM) of each feed was determined by drying it at 105 °C to constant weight, prior to the water stability tests. All water stability tests were done in duplicate.

Static method

In the static method, no pellet or water agitation was involved. Around 25 g (m0) of salmon feed pellets were weighed and put in 600-ml glass beaker (height of 126 mm and a diameter of 90 mm) and 250 ml of distilled water was added. The beaker was covered with aluminum foil and put in the climate chamber (Binder KBF 240, Binder GmbH, Tuttlingen, Germany) where the feed was soaked for 24 h at 23 °C. After soaking, all pel-lets were transferred on a 2.24 mm sieve and excess of water was gently removed by ma-nually moving the sieve for 20 s. The pellets were transferred onto a plastic tray and then weighed in a previously dried and weighed aluminum plate. Soaked pellets were dried in UNB 400 oven (Memmert GmbH, Schwabach, Germany) at 105 °C for at least 18 h. The plate with pellets was again weighed after drying, in order to determine residual dry mat-ter of the samples (m). The Water Stability Index (WSI) was calculated as:



18

WSI=m

m0·initial DM·10000 (%)

Wet sieving method

The soaking part of this method was the same as described for the static method. Unlike the static method, the pellets were sieved after soaking by laboratory sieving devi-ce (Retsch AS200 Control, Haan, Germany) with a 2.24 mm sieve. The sieving was car-ried out for 10 min at an amplitude of 2 mm, with continuous washing of the pellets directly from the nozzles placed on the lid at maximum water flow in order to additio-nally promote disintegration of the pellets. After completing the wet sieving step, the steps of transferring, weighing and drying the pellets were conducted in the same manner as it is described in the static method. The results were calculated as WSI by using equation 1.

Conductometric method

An amount of 25 g of feed pellets was measured in 600-ml glass beaker and 250 ml of distilled water was added. The measuring cell of the Lab960 conductivity meter (SI Analytics, Mainz, Germany) was immediately immersed directly in the middle of the beaker to the point in which the bottom of the cell was at the 100 ml mark on the beaker. In that way, the position of the measuring cell was standardized in each trial. The pellets were let soaking in the water for 24 h at room temperature, and for that period of time the conductivity measuring device was set to automatically record values of measured water conductivity each 20 min, starting instantaneously after immersion of measuring cell. The reference temperature for the temperature compensation was preset at 20 °C during con-ductivity measurement. After 24 h, the conductivity measurement was stopped, and con-ductivity values and corresponding times were transferred from the measuring device directly to a PC using MultiLabpilot software (WTW, Weilheim, Germany). Conducti-vity curves were constructed as a function of the water conductivity and time.

Statistical analysis

Descriptive statistical analyses to calculate the means and the standard errors were performed using Microsoft Excel software (Microsoft Office 2010). All obtained results were expressed as the mean ± standard deviation (SD). The one-way analysis of variance (ANOVA) was performed for comparison of means, and significant differences between specific samples are determined according to post-hoc Tukey’s HSD test at 95% confi-dence limit, using StatSoft Statistica 12 software (14). Correlation analysis was used for finding significant correlations between the results of wet sieving and conductometric method by StatSoft Statistica 12 software.



19

RESULTS AND DISCUSSION The results of water stability of eight tested feed samples obtained by static water and wet sieving method are presented in Table 1.

Table 1. Water stability of tested salmon feeds by gravimetric methods

Sample Static water method Wet sieving method

WSI (%) WSI (%) Feed 1 90.25 ± 0.21d 58.13 ± 1.32cd Feed 2 88.03 ± 0.66ac 64.26 ± 1.10e Feed 3 89.23 ± 0.94cd 48.95 ± 0.76b Feed 4 87.25 ± 0.28a 60.07 ± 1.89de Feed 5 87.53 ± 0.00ab 42.66 ± 0.03a Feed 6 89.46 ± 0.08cd 58.47 ± 2.24cd Feed 7 88.99 ± 0.07bcd 54.25 ± 0.53bc Feed 8 90.00 ± 0.12d 61.19 ± 2.09de

Variance 1.25 50.44 CV (%) 1.05 3.82

CV − coefficient of variation; WSI − water stability index a Values with the same letter in a column, written in superscript, are not statistically different at the p < 0.05 level, 95% confidence limit, according to Tukey’s HSD test

The WSI values obtained by the static method were in the range of 87.25 − 90.25%, where Feed 4 possessed the lowest and Feed 1 had highest value of WSI. Static method showed significantly higher values of WSI in comparison with the wet sieving method, which was in accordance with the results of Obaldo et al. (5), where static water method resulted in the higher DM retention of two shrimp feeds compared with horizontal and vertical shaking methods. The authors suggested that the static method can be used for comparing the maximum water stability of aquatic feeds. The wet sieving method provi-ded much wider range of WSI values in comparison with static method, which might allow higher discrimination of the results. The WSI values obtained by the wet sieving method showed that Feed 5 possessed the lowest, and Feed 2 possessed the highest water stability. Although less precise than static method, according to the higher coefficient of variance (3.82), wet sieving method proved to be a more fitting method for determining water stability, as there were more distinctive significant differences between WSIs of tested feed samples. Obaldo et al. (5) recommended that a rapid method for water stabi-lity determination of aquatic feed has to involve vertical or horizontal shakings of the wa-ter and the pellets, in order to mimic culture conditions, and therefore provide more rea-listic results. Mild shaking of the pellets and water of constant temperature (23 °C) also simulated real conditions that occur during feed consumption by trout (6). Shaking also increased pellet degradation and to better differentiation of feed water stability (5, 6). Thus, wet sieving method that included agitation of pellets, gave a more realistic result of water stability compared to the static method.



20

The curves obtained by plotting the conductivity as a function of time are presented in Figure 1.

Figure 1. Water conductivity during 24 h long soaking of eight tested salmon feeds The extruded pellets release nutrients, primarily proteins, fat and minerals into the water right after their immersion (9). During feed nutrients leaching into the water, its conductivity increased. Obviously, a lower value of the water conductivity at a specified time of the analysis means a lower level of nutrient leaching. According to the rate of nutrient leaching, two groups of samples can be distinguished, as can be seen from Figure 1. The first group comprises the curves for the samples 1, 2 and 8, while the second group consists of curves for the samples 3, 4, 5, 6 and 7. The first group is characterized by a linear rate of nutrient leaching, during all 24 h water soaking. The samples of the second group had high rate of nutrient leaching in the first 10 hours (Feed 3, 4, 6 and 7) and 8 hours (Feed 5) of soaking, after which the loos of nutrients started in a constant manner. As can be seen, the water conductivities of the samples of the first group were signifi-cantly lower compared to those of the second group. The final water conductivity could be also suitable indicator of the feed water stability. The lowest conductivity of Feed 1 at the end of the analysis indicates its highest water stability (the lowest nutrient leaching), which was also confirmed by the static method. The low conductivities for Feeds 2 and 8 are in accordance with the results of wet sieving method, which showed that these samp-les had the highest water stability. Feed 5 contributed to the highest water conductivity, which was in accordance with its lowest water stability according to the results of the wet sieving method, and second lowest water stability according to the results of the static method. Correlation analysis was used to check the relationship between the water con-ductivity and the WSI results obtained by the static and wet sieving method, and the results are presented in Table 2.



21

Table 2. Correlations between the WSI obtained by the two methods and water conductivity values

Water immersion time

(min) Static water method Wet sieving method

20 r = -0.345 p = 0.403

r = -0.895* p = 0.003

40 r = -0.453 p = 0.260

r = -0.883* p = 0.004

60 r = -0.446 p = 0.269

r = -0.894* p = 0.03

80 r = -0.450 p = 0.263

r = -0.869* p = 0.005

100 r = -0.465 p = 0.245

r = -0.857* p = 0.006

120 r = -0.473 p = 0.236

r = -0.841* p = 0.009

140 r = -0.470 p = 0.240

r = -0.831* p = 0.011

160 r = -0.461 p = 0.250

r = -0.819* p = 0.013

180 r = -0.457 p = 0.255

r = -0.808* p = 0.015

200 r = -0.460 p = 0.252

r = -0.798* p = 0.018

220 r = -0.457 p = 0.255

r = -0.789* p = 0.020

240 r = -0.452 p = 0.260

r = -0.782* p = 0.022

1440 r = -0.431 p = 0.286

r = -0.651 p = 0.080

r − correlation coefficient * Significant at 0.05 level

The results of the correlation analysis showed that there were no significant (p < 0.05) correlations between the WSI and the water conductivities measured in the static regime. However, significant negative correlations (p < 0.05) were obtained in the wet sieving method for the first 4 hours of pellets soaking, an exception being only the final water conductivity. The highest negative correlation was observed for the results of water conductivity obtained at 20th (r = -0.895) and 60th (r = -0.894) min of the pellets soa-king. As the soaking time increased, the correlation showed a decrease. Thus, the conduc-tivity values within the first hour of conductometric measurement could be used for ap-proximation of the feed water stability. The obtained results suggest that it suffices to measure the conductuctivity just at 20 to 60 min. The conductometric test appeared to be



22

a potential alternative to the gravimetric tests for rapid determination of water stability. Since a negative correlation was established between the water conductance and WSI of the feed, for practical reasons and better comparison of feed water stability, the conducti-vity values measured in the first hour were given negative signs, the results being pre-sented in Table 3.

Table 3. Negative value of water conductance within first hour of conductometric analysis

Sample Negative value of measured water conductance (µS/cm)

20 min 40 min 60 min Feed 1 -43.3 -44.7 -50.4 Feed 2 -30.9 -40.0 -48.7 Feed 3 -439.0 -625.0 -774.0 Feed 4 -241.0 -440.0 -508.0 Feed 5 -624.0 -909.0 -1109.0 Feed 6 -347.0 -447.0 -520.0 Feed 7 -361.0 -510.0 -647.0 Feed 8 -151.4 -170.0 -181.1

By assigning negative values to the measured water conductivity, it was simpler to compare the results with WSI values, since a lower negative value of conductivity indi-cated a lower WSI. According to the assigned negative values of water conductance after 20, 40 and 60 min of soaking (Table 3), feed samples followed the same ranking pattern as the results of WSI obtained by wet sieving, excluding Feed 1. Feed 5 had a lowest negative value of water conductivity, thus possessed the lowest water stability. Thus, the conductometric test could be used for a rapid and effective relative comparison of aquafeed water stability.

CONCLUSION

Conductometric method proved to be useful primarily for quick and simple determi-nation of nutrient leaching. As the results showed that the static method is suitable for determining maximum water stability of feed, while wet sieving method is more suitable for determining actual feed water stability. The conductometric method has potential for a simple and quick relative comparison of water stability for salmon feed. This novel approach for water stability determination appears to be an interesting, rapid alternative to time consuming gravimetric methods, especially in development of new fish feeds. Further investigations with different types of aquatic feeds are needed in order to enhance the applicability of the proposed method.

Acknowledgеment

These results are part of the research within the projects III46012 and TR31011, financed by the Ministry of Education, Science and Technological Development, Republic of Serbia.



23

REFERENCES 1. Glencross, B.D.; Booth, M.; Allan, G.L. A feed is only as good as its ingredients – a

review of ingredient evaluation strategies for aquaculture feeds. Aquacult. Nutr. 2007, 13, 17-34.

2. Gatlin, D.M.; Barrows, F.T.; Brown, P.; Dabrowski, K.; Gaylord, T.G.; Hardy, R.W.; Herman, E.; Hu, G.; Krogdahl, Å.; Nelson, R.; Overturf, K.; Rust, M.; Sealey, W.; Skonberg, D.; Souza, E.J.; Stone, D.; Wilson, R.; Wurtele, E. Expanding the utilization of sustainable plant products in aquafeeds: a review. Aquacult. Res. 2007, 38, 551-579.

3. Sørensen, M. A review of the effects of ingredient composition and processing conditions on the physical qualities of extruded high-energy fish feed as measured by prevailing methods. Aquacult. Nutr. 2012, 18, 233-248.

4. Oehme, M.; Aas, T.S.; Olsen, H.J.; Sørensen, M.; Hillestad, M.; Li, Y.; ÅSgård, T. Effects of dietary moisture content of extruded diets on physical feed quality and nu-tritional response in Atlantic salmon (Salmosalar). Aquacult. Nutr. 2014, 20, 451-465.

5. Obaldo, L.G.; Divakaran, S.; Tacon, A.G. Method for determining physical stability of shrimp feeds in water. Aquacult. Res. 2002, 33, 369-377.

6. Baeverfjord, G.; Refstie, S.; Krogedal, P.; ÅSgård, T. Low feed pellet water stability and fluctuating water salinity cause separation and accumulation of dietary oil in the stomach of rainbow trout (Oncorhynchus mykiss). Aquaculture 2006, 261, 1335-1345.

7. Aas, T.S.; Terjesen, B.F.; Sigholt T.; Hillestad, M.; Holm, J.; Baeverfjord, G.; Rørvik, K.-A.; Sørensen, M.; Oehme M.; ÅSgård, T. Nutritional responses in rainbow trout (Oncorhynchus mykiss) fed diets with different physical qualities at stable or variable environmental conditions. Aquacult. Nutr. 2011, 17, 657-670.

8. Fagbenro, O.; Jauncey, K. Water stability, nutrient leaching and nutritional properties of moist fermented fish silage diets. Aquacult. Eng. 1995, 14, 143-153.

9. Ighwela, K.A.; Ahmad, A.B.; Abol-Munafi, A.B. Water stability and nutrient leaching of different levels of maltose formulated fish pellets. Global Vet. 2013, 10 (6), 638-642.

10. Umar, S.; Kamarudin, M.S.; Ramezani-Fard, E. Physical properties of extruded aqua-feed with a combination of sago and tapioca starches at different moisture contents. Anim. Feed Sci. Technol. 2013, 183, 51-55.

11. Morken, T.; Kraugerud, O.F.; Barrows, F.T.; Sørensen, M.; Storebakken, T.; Øver-land, M. Sodium diformate and extrusion temperature affect nutrient digestibility and physical quality of diets with fish meal and barley protein concentrate for rainbow trout (Oncorhynchus mykiss). Aquaculture 2011, 317, 138-145.

12. Čolović, R.; Banjac, V.; Kokić, B.; Lević, J.; Đuragić, O.; Spasevski, N.; Sredanović, S. A simple method for determination of relative leaching losses of water soluble components of fish feed, Aquaculture Europe 2014, Donostia-San Sebastian, 14−17 October 2014, pp. 264-265.

13. Banjac, V.; Čolović, R.; Vukmirović, Đ; Čolović, D.; Đuragić, O.; Palić, D.; Pezo, L. A proposal of methods for determining water stability of extruded fish feed, VII Inter-national Conference Water & Fish, Belgrade, 10−12 June 2015, pp. 170-175.



24

14. STATISTICA (Data Analysis Software System), v.12.0 (2013). Stat-Soft, Inc, USA (www.statsoft.com).

КОНДУКТОМЕТРИЈСКИ ПРИСТУП ЗА ОДРЕЂИВАЊЕ СТАБИЛНОСТИ У

ВОДИ И ОТПУШТАЊА НУТРИЈЕНАТА ЕКСТРУДИРАНЕ ХРАНЕ ЗА РИБЕ

Војислав В. Бањац1, Радмило Р. Чоловић1, Душица С. Чоловић1, Лато Л. Пезо2,

Јасмина М. Губић1, Оливера М. Ђурагић1

1 Универзитет у Новом Саду, Институт за прехрамбене технологије, Булевар цара Лазара 1, 21000 Нови Сад, Србија

2 Универзитет у Београду, Институт за општу и физичку хемију, Студентски Трг 12-16, 11000 Београд, Србија

Стабилност у води осам различитих узорака екструдиране хране за лососа испитана је применом две сопствене гравиметријске методе: статичке методе и методе мокрог просејавања. Затим је примењена и кондуктометријска метода, претходно развијена од стране аутора пре свега за одређивање отпуштања нутријената из хране за рибе у воду током њеног потапања. Истраживање је за циљ имало да се испита потенцијал кон-дуктометријског приступа као технике за одређивање стабилности у води екструдиране хране за рибе. Корелациона анализа је коришћена у циљу проналажења односа између резултата гравиметријских тестова и кондуктометријске методе. Резултат статичке ме-тоде и методе мокрог просејавања изражен је као индекс стабилности у води, који представља проценат преостале суве материје хране за рибе након распадања у води. Показано је да статичка метода може да послужи за одређивање максималне вредности стабилности у води хране за рибе, док метода мокрог просејавања даје реалне резултате стабилности у води хране за рибе. Кондуктометријска метода резултовала је у кондук-тометријским кривама за сваки испитани узорак, које су омогућиле одређивање брзине отпуштања нутријената у воду. Добијене вредности проводљивости воде нису статис-тички значајно корелирале са резултатима статичке методе, док су показале значајне негативне корелације са резултатима методе мокрог просејавања током прва четири сата натапања у води. Највише вредности степена корелације добијене су током првог сата кондуктометријcког мерења, показајући тако да кондуктометријски метод има по-тенцијал као брза и једноставна метода за одређивање и релативно поређење стабил-ности у води екструдиране хране за лососа. Кључне речи: храна за рибе, стабилност у води, отпуштање нутријента, кондукто-

метрија

Received: 22 August 2017 Accepted: 22. September 2017

APTEFF, 48, 1-323 (2017) UDC: 66.017/.018+678.86]:546.62 https://doi.org/10.2298/APT1748025B BIBLID: 1450-7188 (2017) 48, 25-38


25

EFFECT OF ALUMINIUM PARTICLES ON MECHANICAL AND MORPHOLOGICAL PROPERTIES OF EPOXY NANOCOMPOSITES

Sefiu A. Bello1,2*, Johnson O. Agunsoye2, Jeleel A. Adebisi2,3, Suleiman B. Hassan2

1 Department of Materials Science and Engineering, Kwara State University, Malete, Nigeria

2 Department of Metallurgical and Materials Engineering, University of Lagos, Nigeria 3 Department of Metallurgical and Materials Engineering, University of Ilorin, Ilorin, Nigeria

Bumper is a front or rear part of automobiles. It is designed and shaped to be impact absorbing and protecting automobiles from damage in low impact collisions. Initially, they were made from heavy steels, increasing the weight of automobiles and fuel consumption. Also, high impacts of steel bumpers on pedestrians during accidental collision cause fatalities and or disabilities. An effort to enhance fuel efficiency, safety, freedom of design and shape detailing, heavy alloys for automobile applications are now being replaced with polymeric composites. Aluminium micro particles and nanoparticles were prepared from aluminium cans through sand casting, lathe machine spinning, and ball milling techniques. Both types of aluminium particles were incorporated into a mixture of diglycidyl ether of bisphenol A (DGEBA, epoxy resin) cured with amine base hardener (ABH). Phases of the epoxy polymer and composites were identified using X-ray Diffraction (XRD). Spatial arrangement of the phases within the matrix and their elemental composition were examined using Scanning Electron Microscope with attached energy dispersive X-ray spectroscopy (SEM/EDX). Tensile, impact and micro hardness tests were conducted on the prepared epoxy/aluminium composites. Results of the XRD showed the presence of aluminium compounds/phases due to chemical reactions between aluminium particles and DGEBA/ABH system. SEM confirmed a homogeneous distribution of the phases within the epoxy matrix, and that there is a strong adhesion between the epoxy matrix and aluminium particles. Correlation between the mechanical properties of the prepared nanocomposite and the procured bumper materials exhibited a fair suitability of the prepared nanocomposites for automobile applications. KEY WORDS: Aluminium; nanoparticles; correlation, mechanical properties, bumper

material

INTRODUCTION Nanotechnology is an area of research which has led to material restructuring, resul-ting in the enhancement of material properties. Nanoscience is the study of materials

* Corresponding author: Sefiu A. Bello, Department of Materials Science and Engineering, Kwara State

University, Malete, Nigeria, e-mail: [email protected] or [email protected]



26

where some critical properties are attributable to an internal structure with at least one dimension less than 100 nm (1, 2). Nanoscience primarily deals with the synthesis, characterization, exploration, and exploitation of nanostructured materials (3). Nanostruc-tured materials are low-dimensional materials comprising building units of a submicron or nanoscale size at least in one direction and exhibiting size effects. They are the products obtained owing to the research into nanotechnology. One of the nanostructured materials is the dispersion strengthening composite or nanocomposites. Nanocomposites are obtained through incorporation of nanoparticles in a matrix. Many studies have been focused on the development of particles reinforced composites to replace fibre reinforced counterparts because of the difficulties in their manufacturing and their anisotropic properties. However, some limitations in the properties enhancement have been found in the literature due to weak interfacial adhesion leading to poor mechanical and wear resistance properties of microparticle reinforced composites (4, 5). The development of dispersion strengthening composites/nanocomposites has been proposed to address the identified challenges noticed with microparticle reinforced composites (6). Since a nanoparticle is much finer than a micro particle, ultrahigh fineness known with nanoparticle could enhance its interaction with a matrix molecule leading to fortified adhesion, which is a basis for improvement in the mechanical properties of engineering composites. An effort to enhance fuel efficiency, safety, freedom of design, and shape detailing, heavy alloy for automobile applications are now being replaced with polymeric composites (7-10). Due to high cost of glass, carbon or aramid fibre reinforced plastics, the evolution of automobile materials is limited to a high extent, especially in exotic cars (2). To expand the use of polymeric composites for automobile applications and their sustainability, adoption of cost effective recycling strategy is very important (11). Many reports have been found dealing with the development of ecofriendly reinforced plastics using particles obtained from agro and metallic wastes (12-19). However, reports on processing of aluminium cans to obtain aluminium nanoparticles (using ball milling technique) as reinforcement for development of epoxy/aluminium nanocomposites for automobile bumper application are very scarce if found at all. This work is aimed at studying the effect of aluminium particles on morphological and mechanical properties of epoxy/aluminium particulate nanocomposites. This attempt is very important since the potential benefits are related to not only environmental cleanliness, elimination of green gas emission and reduction in power consumption associated with the primary route for aluminium production, but also to the development of light weight materials for auto-mobile applications.

MATERIALS AND METHODS Epoxy resin, diglycidyl ether of bisphenol A (DGEBA) and amine based hardener (ABH) were obtained from Polymer Composite Institute (PCI), Ontario Canada through a local vendor in Lagos. Aluminium cans were obtained from the University of Lagos Management Centre, Lagos Nigeria. Aluminium cans were melted at 660±5 °C and cast



27

into bar using green sand mould. The bar was spun into particle/fibres using Col-chester/triumph lathe machine, model 2000 at the Physics Department, University of Lagos. During spinning, jets of water were intermittently applied on the aluminium surface to cool and harden the surface. This aided the aluminium breakage. Aluminium particles/fibres were further broken using steel mortar and pestle. Prior to ball milling of aluminium, milling balls/carbonised coconut shell was charged into the 87002 LIMOGES planetary mill (model: 28A20-92), in the Federal Industrial Institute of Research Oshodi (FIIRO) Lagos, and the machine was operated for 30 minutes to coat the ball surfaces with carbonised coconut shells used as solid lubricants. Then aluminium particles were added and the ball milling was carried out for 40 hours at 8.5 charge ratios (CRs) and 192 rpm. An amount of carbonised coconut shell equivalent to 0.1 % of aluminium particles was used. Aluminium powders obtained at 40 hours were classified using a set of sieves in a descending order of grain fineness (2000-75 μm), vibrated mechanically for 30 minutes with the aid of a sine shaker in accordance with BS 1377, standard 1990 (14). The finest aluminium powders collected in a pan below 75 μm sieve were used as a precursor for the synthesis of Alnp. The precursor was further milled for 70 hours at 8.5 CRs according to the known procedure (20). The obtained particles were characterized for size determination using XRD and Transmission Electron Microscope (TEM). Al micro particles (Almp) of the size of 56 µm and Al nano particles (Alnp) of the size 55.5 nm were incorporated into a DGEBA/ABH system in a 2:1 volume ratio, using mechanical shear stirring technique. The mixture was poured into a steel die open mould and allowed to set and harden at room temperature for 48 hours. Initially, 2 wt% of Almp was added to the DGEBA/ABH. The E/Almp composite samples were gently removed from the mould and then postcured at 130 °C for four hours using Shel lab vacuum oven, model SVA S2E, at the Department of Materials Science and Engineering, Kwara State University, Malete, Nigeria. The process was repeated with an increment in wt% of Almp additions, up to 10 % at an interval of 4 % in each case in line with (6). The same techni-ques and wt% of reinforcement additions were used to produce epoxy/aluminium nano-particle (E/Alnp). The phases of the developed epoxy/aluminium composites were determined using XRD, spatial configuration and elemental composition were examined using SEM/EDX. Mechanical tests such as tensile, impact and micro hardness were car-ried out on the prepared composites. Figure 1 presents some of procedural steps taken in the development of E/Alnp nanocomposites.



28

Figure 1. Procedural steps for the development of epoxy aluminium nanocomposites



29


Structure of the prepared aluminium nanoparticles The X-ray diffractometry (XRD) profile of Alnp is shown in Figure 2. The three observed major peaks reveal Al at the diffraction angles (2θ) of 45°, 52.5° and 77.33°, while the minor peaks show Al2O3 at 29.48°, 41.03°, 50.08°, 61.85°and 67.95° and SiO2 at 31.21°. The presence of SiO2 can be linked to the carbonised coconut shell used as solid lubricant to prevent sticking of aluminium particles to the milling balls. Moreover, presence of Al2O3 suggests the oxidation of aluminium particle by cooling agent (water) during machine spinning and or milling environment during particle refinement process. The TEM image in Figure 3 displays relatively coarse Alnp surrounded by many fine Alnp. Average sizes of Alnp determined by XRD aided with Scherrer’s equation and TEM supported with software are 55.5 and 70.4 nm, respectively.

Figure 2. XRD profile of aluminium nanoparticles



30

Figure 3. TEM images showing size distribution of aluminium nanoparticles

Structure of the synthesized composites Figure 4a-b displays the XRD profiles of epoxy polymer and E/Alnp nanocomposites. The major peaks observed in Figure 4a are due to C5H9NO2, C12H12N2, C5H14ClN at 18.94, 18.94/42.34 and 42.34°, respectively. Their inter-planar spacing is 4.69, 4.49/2.13 and 2.13 Å. These phases are organic compounds, justifying chemical reaction between DGEBA and ABH that led to the formation of epoxy molecules. When Alnp was added to the DGEBA/ABH system, the stoichiometric balance between DGEBA and ABH was disturbed, leading to different cross linking reactions. This resulted in the formation of new phases such as C4H9NO4, C4H9NO3, C5H4N4 and BrH4N at 36.36, 36.36, 36.36, 36.36/44.54/64.92°, respectively, with the residual Al at 64.92° (see Figure 4b). Their inter-planar spacing varies between 1.2 and 4.7 Å, meaning that the orientation of the phases within E/Alnp nanocomposites has to do with mechanical properties of the nanocomposites (21). The observed phase orientations of E/Alnp nanocomposites are different from those of epoxy polymer having inter-planar spacing from 2.13 to 4.69 Å. The presence of second phase particles within the matrix of E/Alnp nanocomposites has reduced the spacing of the epoxy molecules. Since the enhancement in mechanical properties of dispersion strengthening composites depends on orientation of pha-ses/molecules, the particles wt% and sizes, interfacial adhesion, nature of stress transfer from a matrix to the filler, the differences in inter-planar spacing ranges of epoxy polymer and E/Alnp nanocomposite is an indication that the orientation of the phases present in the matrix of E/Alnp nanocomposites plays a significant role in enhancing



31

mechanical properties of the E/Alnp nanocomposites. However, the enhancement in the mechanical properties could also be ascribed to other factors mentioned above.

Figure4. XRD profiles of (a) epoxy polymer (b) E/Alnp nanocomposites

Spatial configuration of the epoxy composites Figure 5a-b depicts the spatial configuration of the epoxy polymer and E/Alnp com-posites. Figure 5a reveals bonded segments representing infusible cross-linked structure

a

b



32

of epoxy molecules. Both white and dull epoxy molecules phases are identified. These can be linked to epoxy compounds confirmed by XRD in Figure 2a. The EDX spectrograpms in Figure 5a indicate C (major peak), O and K, which are trace elements. The observed configuration and phases in Figure 5b are different from those in Figure 5a. This difference is linked to the addition of Alnp as a reinforcement to DGEBA/ABH system. The microstructure is homogenous with even distribution of second phase particles within the epoxy matrix. This indicates a good interfacial adhesion between the matrix and the reinforcement. With this structural integrity, a rapid load transfer from the matrix to the phase particles, which enhances the load bearing capacity of the E/Alnp composite, is expected.

Figure 5. SEM/EDX of (a) epoxy polymer (b) E/Alnp nanocomposite

Mechanical properties of the epoxy nanocomposites Figure 6 reveals an increase in the tensile strength of the epoxy/aluminium particulate composites with an increase in the wt% of aluminium particles. A progressive increase in tensile strength was noticed with Alnp additions while a decrease in tensile strength was observed above 6 wt% of Almp additions. Generally, the increase in the tensile strength can be attributed to the presence of second phase particles within the epoxy matrix. These phase particles are ceramic in nature and very rigid. Their strong adhesion to the matrix allowed effective and rapid stress/load transfer from the matrix to the particles which



33

acted as load bearers by their rigidity. Therefore, their presence within the epoxy matrix enhanced the load bearing capacity of the epoxy/aluminium composites. Moreover, the decrease in tensile strength found above 6 wt% of Almp could be ascribed to saturation of the matrix with Almp. This implied that above 6 wt% of Almp addition, the epoxy matrix was unable to join all Almp together. Therefore, there were domains of freely existing Almp within the epoxy matrix, which created region of discontinuity. Under uniaxial loading, the free existing particles interfered with mobility of the composite segment, creating voids through the movement from one point to another. These voids acted as stress raisers and caused the composite deformation at a stress level below the expectation. Similar explanation was found in literature (12, 15-17). A slight decrease in the percentage elongation with increased particle additions was noticed with E/Alnp while E/Almp displayed lower percentage elongation. This shows better mechanical behaviour of E/Alnp than E/Alnp which can be linked to smaller size of Alnp than Alnp which aided its better interaction with epoxy than Alnp. Figure 7 illustrates a progressive increase in the impact energy with E/Alnp and E/Almp. This signifies an increase in the impact absorbing ability of the synthesized composites with a rise in wt% of particle additions. This increase can be linked to the softness of the epoxy matrix which relieved/damped the loading constraint and the rigidity of the second phase particles, which resisted the composite deformation. Higher impact toughness of E/Alnp than E/Almp can be associated with the fineness of the second phase particles found in the epoxy matrix of E/Alnp nanocomposites. Similar enhancement in the impact toughness of particle reinforced epoxy is found in the literature (22, 23). In addition, Figure 8 displays a small increase in the micro hardness of E/Almp micro composites, while enhancement in micro hardness of E/Alnp nanocom-posites is highly noticeable up to 6 wt% of Alnp addition to epoxy while above this level, no significant improvement was noticed. This behaviour is expected because of the progressive increase in impact absorbing capacity of E/Alnp nanocomposites. It implies that up to 6 wt% of Alnp addition, there was an increase in the impact absorbing ability and resistance to indentation of E/Alnp, while above this level of reinforcement, the decrease in the resistance to surface indentation was compensated for an increase in the impact energy of the E/Alnp. The observed increase in micro hardness values is in line with the literature (23). Correlation of mechanical properties of the prepared epoxy aluminium composites

with those of existing automobile materials The obtained results clearly demonstrated better mechanical behaviour of E/Alnp than E/Almp. Moreover, the tensile strength and impact energy of the developed E/Alnp composites were compared with Toyota (Sienna) and Nissan (Almera) bumper materials bought locally in Nigeria and tested using the same ASTM standards (ASTM D 3039 for tensile test, ASTM D 3763 for impact toughness) used for testing the prepared E/Alnp nanocomposites. The peak tensile strength was observed at 18.58 Nmm-2 for the 10 % Alnp addition to the epoxy. This value is greater than the tensile strengths of the Toyota and Nissan bumper materials (12.23 and 6.44 Nmm-2, respectively). The impact energy of



34

the Toyota bumper material was 21.18 J, that of Nissan bumper material 25.54 J, while the impact energy of the prepared E/Alnp nanocomposite was 14.01 J. This value is lower than that of either Toyota or Nissan bumper materials. In addition, this value is within the range of 10-18 J specified for Glass Material Thermoplastic for automobile bumper by Azdel (24). The micro hardness value of E/10% Alnp is 12.03 HV which is less than 442 HV peak hardness value specified in United Nation’s Economic Commission for Europe (ECE)/324 Regulation No. 42 (25) for automobile bumper materials. This shows a fair suitability of the developed E/Alnp nanocomposites for automobile applications. The current research is focused on the development of epoxy/aluminium coconut shell parti-culate hybrid nanocomposite for automobile bumper application, which is just completed, and the results will be presented in another publication.

Figure 6. Tensile strength of the epoxy/aluminium composites

Figure 7. Tensile strain of the epoxy/aluminium composites



35

Figure 8. Impact energy of the epoxy/aluminium particulate composites

Figure 9. Micro hardness of the epoxy/aluminium composites

CONCLUSION In conclusion, polymeric nanocomposites intended for automobile bumper application were developed from epoxy resin and disposable aluminium cans. Addition of aluminium particles to epoxy resulted in the improvement of mechanical properties of the synthe-sized epoxy composites. An increase of 37.84 % in the tensile strength of epoxy alu-minium nanocomposites was obtained at 10 wt% of aluminium nanoparticle, while at the same wt% of aluminium microparticles addition to epoxy, a deterioration in tensile strength of epoxy aluminium micro composites was observed. Generally, better mecha-



36

nical behaviour of synthesized epoxy aluminium nanocomposites than their micro counterparts, is attributed to higher fineness of aluminium nanoparticle than aluminium microparticles which gives room for better interaction of aluminium nanoparticles with epoxy. Moreover, this study has given birth to an ecofriendly polymeric nanocomposite, whose large/industrial scale production is expected to lower cost constraint preventing widespread applications of polymeric composites in engineering structures.

Acknowledgement Authors appreciate staff of the Ceramics Department of Federal Industrial Institute of Research Oshodi, Nigeria; Department of Materials Science and Engineering, Kwara State University, Malete, Nigeria; Department of Metallurgical and Chemical Engi-neering, Pretoria University, South Africa and the Department of Metallurgical and Ma-terials Engineering, University of Lagos for their support in making this work a reality.

REFERENCES 1. Lane, R. Special Issue: Nanotechnology; AMPTIAN Newsletter, 2002, pp 37-42. 2. Bello, S.A.; Agunsoye, J.O.; Hassan, S.B.; Zebase Kana, M.G.; Raheem, I.A. Epoxy

Resin Based Composites, Mechanical and Tribological Properties: A Review. Tribology in Industry, 2015, 37 (4), 500-524.

3. Pokropivny, V.; Lohmus, R.; Hussainova, I.; Pokropivny, A.; Vlassov, S. Introduc-tion to Nanomaterials and Nanotechnology; Tartu University Press, 2007, pp 45-100.

4. Srivastava, V.K.; Verma, A. Mechanical Behaviour of Copper and Aluminium Par-ticles Reinforced Epoxy Resin Composites. American Journal of Materials Science, 2015, 5 (4), 84-89.

5. Vasconcelos, P.; Lino, F.J.; Neto, R.J.L.; Teixeira, A. Glass and Carbon Fibre Reinforced Hybrid Composites for Epoxy Tooling. Structure, 2003, 40, 3-5.

6. Bello, S.A. Development and Characterisation of Epoxy-Aluminium-Coconut shell Particulate Hybrid Nanocomposites for Automobile Applications. Ph.D. Thesis, University of Lagos, October 2017.

7. Tammy, M. What Materials are Used for a Automobile Bumpers; First Choice Reconditioning: Lakewood, 2016, pp 1-5.

8. America's Plastic Maker. Plastic Car and Bumper Systems; America's Plastic Maker: America, 2016, pp 1-2.

9. Gerard, K. High Performance Plastics Used in the Automotive Industry; CRAFTECH BLOG, 2014, pp 1-6.

10. Yu, T.X.; Tao, X.M.; Xue, P. The energy-absorbing capacity of grid-domed textile composites. Composites Science and Technology, 2000, 60 (5), 785-800.

11. The United Kingdom composites Industry. The United Kingdom composites industry – turning ideas into investments. Reinforced Plastics, 2013, 57 (3), 43-46.



37

12. Agunsoye, J.O.; Bello, S.A.; Adetola, L.O. Experimental Investigation and Theoretical Prediction on Tensile Properties of Particulate Reinforced Polymeric Composites. Journal of King Saud University - Engineering Sciences, 2017, in press.

13. Agunsoye, J.O.; Bello, S.A.; Bello, L.; Idehenre, M.M. Assessment of mechanical and wear properties of epoxy based hybrid composites. APEM. 2016, 11 (1), 5-14.

14. Sarki, J.; Hassan, S.B.; Aigbodion, V.S.; Oghenevweta, J.E. Potential of using coconut shell particle fillers in eco-composite materials. Journal of Alloys and Com-pounds, 2011, 509 (5), 2381-2385.

15. Agunsoye, J.O.; Aigbodion, V.S. Bagasse Filled Recycled Polyethylene Bio-composites: Morphological and Mechanical Properties Study. Results in Physics, 2013, 3, 187-194.

16. Ashori, A.; Nourbakhsh, A. Reinforced polypropylene composites: effects of chemical compositions and particle size. Bioresour Technol. 2010, 101 (7), 2515-9.

17. Asuke, F.; Aigbodion, V.S.; Abdulwahab, M.; Fayomi, O.S.I.; Popoola, A.P.I.; Nwoyi, C.I.; Garba, B. Effects of bone particle on the properties and microstructure of polypropylene/bone ash particulate composites. Results in Physics, 2012, 2, 135-141.

18. Atuanya, C.U.; Aigbodion, V.S.; Nwigbo, S.C. Experimental study of the thermal and wear properties of recycled polyethylene/breadfruit seed hull ash particulate composites. Materials & Design, 2014, 53, 65-73.

19. Dhaliwal, J.S.; Kapur, G.S.; Shashikant. Chicken Eggshell as a Biofiller in Polypro-pylene; Composites Society of Plastics Engineers (SPE), 2013.

20. Bello, S.A.; Agunsoye, J.O.; Adebisi, J.A.; Anyanwu, J.E.; Bamigbaiye, A.A.; Hassan, S.B. Potential of Carbonised Coconut Shell as a Ball-Milling Interface for Synthesis of Aluminium (1xxx) Nanoparticles. Annals of Faculty of Engineering, 2017, 15 (2), 149-157.

21. Kinloch, A.J.; Taylor, A.C. The mechanical properties and fracture behaviour of epoxy-inorganic micro- and nano-composites. Journal of Materials Science, 2006, 41 (11), 3271-3297.

22. Vasconcelos, P.V.; Lino, F.J.; Magalhães, A.; Neto, R.J.L. Impact fracture study of epoxy-based composites with aluminium particles and milled fibres. Journal of Materials Processing Technology, 2005, 170 (1–2), 277-283.

23. Abdullah, S.I.; Ansari, M.N.M. Mechanical properties of graphene oxide (GO)/epoxy composites. HBRC Journal. 2015, 11 (2), 151-156.

24. MatWeb Material Property Data. Azdel Plus C401-B01 40% Chopped Fiber Mat/PP Resin Matrix.

http://www.matweb.com/search/datasheet.aspx?MatGUID=c9cea89b3274415d892ea6af98f39647 (Accessed 2010).

25. Balamurugan, A.G., Automotive Bumper and Energy Absorber. 2009.



38

УТИЦАЈ ЧЕСТИЦА АЛУМИНИЈУМА НА МЕХАНИЧКЕ И МОРФОЛОШКЕ КАРАКТЕРИСТИКЕ НАНОКОМПОЗИТА

Сефиу A. Бело1,2, Џонсон O. Агунсој2, Џелил A. Aдебиси2,3, Сулејман Б. Хасан2

1 Департман наука о материјалима и инжењерству, Државни универзитет Куаре, Малет, Нигерија 2 Департман металургије и машинства, Универзитет у Лагосу, Нигерија

3 Департман металургије и машинства, Универзитет у Илорину, Илорин, Нигерија

Браник је предњи или задњи део аутомобила. Дизајниран је и обликован тако да апсорбује ударе и и штити аутомониле од оштећења у сударима ниског интен-зитета. У почетку су били израђивани од тешких челика, што је повећавало тежину аутомобила и потрошњу горива. Такође, снажни удари челичних браника у слу-чајним сударима са пешацима резлтују у фаталним исходима и/или повредама. У настојању да се повећа ефикасност горива, безбедност, слобода у дизајнирању и де-таљи у обликовању, тешке легуре за аутомобиле се сада замењују полимерним ком-позитима. Алуминијумске микро- и нано-честице су добијене применом техника одливања у песку, обраде на стругу и млевења у млину са куглама. Честицe алу-минијума оба типа су биле инкорпориране у смесу диглицидил етар бисфенола (DGBA, епокси смола) и базног аминског очвршћивача (АBH). Фазе епокси поли-мера и композита су биле идентификоване помоћу дифракције X-зрака (XRD). Просторна структура фаза у матриксу и њихов елементарни састав су испитивани скенирајућом електронском микроскопијом у комбинацији са дисперзијом енергије X-зрака (SEM/EDX). Припремљени епокси/алуминијум композити су испитивани на истезање, отпорност на удар и микро чврстоћу. Резултати добивени помоћу XRD су показали присуство једињења/фаза алуминијума насталих реакцијом чес-тица алуминијума са DGEBA/ABH системом. Примена SEM је потврдила хомогену дистрибуцију фаза у матриксу епокси смоле и да постоји јака адхезија између епок-си матрикса и честица алуминијума. Корелација између механичких карактеристи-ка припремљених нано-композита и коришћених материјала показала је прихват-љиву погодност припремљених нано-композита за примену у аутомобилској индус-трији. Кључне речи: aлуминијум, нано-честице, корелација, механичке карактеристике,

матријал за бранике

Received: 06 May 2017. Accepted: 13 July 2017.

APTEFF, 48, 1-323 (2017) UDC: 637.146.1:577.86 https://doi.org/10.2298/APT1748039B BIBLID: 1450-7188 (2017) 48, 39-52

Review paper

39

POTENTIAL OF INDIGENOUS LACTOBACILLI AS STARTER CULTURE IN DAIRY PRODUCTS

Mirjana Bojanić Rašović*

University of Montenegro, Biotechnical Faculty, Mihaila Lalića 1, 20000 Podgorica, Montenegro

Traditional production of fermented dairy products involves lactic acid bacteria that are normally present in the milk and production environment. These lactic acid bacteria represent the niche microbiota of the geographical area and they are responsible for local types of fermented products. In order to standardize indigenous products, the basic requirement is the application of the determined indigenous lactic acid bacteria as starter cultures affecting their specific characteristics by performing fermentation and influencing the ripening process. In the process of cheese fermentation usually partici-pate bacteria of the genus Lactococcus and homofermentative lactobacilli. However, the process ripening is influenced mainly by the so-called nonstarter lactic acid bacteria – lactobacilli and secondary microflora. Lactobacilli during ripening of cheese continue to breakdown the rest of lactose, but they are primarily important in the process of protein breakdown. During metabolism of sugars and amino acids, lactobacilli produce aromatic compounds which have a positive effect on the flavor of the product. Some species of lactobacilli are available as probiotics. Some lactobacilli produce bacteriocins, which prevent the growth of pathogens, as well as many spoilage microorganisms. Indigenous lactobacilli have application especially in the production of typical local dairy products that are well accepted by the local population. Besides that, the use of indigenous lactic acid bacteria as starter cultures allows the production of cheese with designated geogra-phical origin that could be placed on the international market. Consequently, indigenous lactic acid bacteria are a challenge for further research and possible their practical application in the dairy industry. KEY WORDS: lactic acid bacteria, autochtonous cheese, lactobacillus, cheese ripening

INTRODUCTION Indigenous foods are an important feature of the culture and habits of a nation and thus the most important wealth of each country. The indigenous cheeses have a versatile taste, aroma and consistency when compared to industrially produced cheeses, where technologies are strictly defined and production conditions are controlled. During the pro-duction of indigenous cheeses the milk is neither pasteurized nor starter cultures are * Corresponding author: Mirjana Bojanić Rašović, University of Montenegro, Biotechnical Faculty, Mihaila

Lalića 1, 20000 Podgorica, Montenegro, e-mail: [email protected]


Review paper

40

applied (Table 1, Figure 1,) (1). This way of production involves lactic acid bacteria (LAB) that are normally present in the production environment, as well as in milk. The study of the technological and probiotic properties of autochthonous lactic bacteria is the first step towards their practical application in obtaining indigenous dairy products. The diversity of lactic acid bacteria in traditional dairy products and production environment represents a great potential in biotechnology (2, 3). In order to standardize indigenous products, the basic requirement is to employ the determined indigenous LAB as starter cultures which affect their specific characteristics by performing fermentation and influ-ence the ripening process (4-8).

Diversity of lactobacilli in indigenous dairy products

In artisanal cheeses, the microbiota is quite heterogeneous. Its composition changes during the cheese ripening. The dominant species in fresh cheese are Lactobacillus pa-racasei, Lb. plantarum and Lb. brevis, but only Lb. paracasei dominates in the ripened cheese (9, 10). Lb. paracasei ssp. paracasei represents a dominant strain in the micro-flora of traditionally homemade Bukuljac cheese (11). Mesophilic lactobacilli are the do-minant organisms in mature Cheddar cheese (12). Strains Lb. paracasei ssp. paracasei, Lb. plantarum, Lb. brevis, Lactococcus lactis ssp. lactis, Enterococcus faecium and Enterococcus faecalis were the main groups present in the Zlatar cheese during ripening (13, 5, 14). In this cheese are also isolated Lb. casei and Lb. parabuchneri (15). In Irish Cheddar cheese the main nonstarter lactic acid bacteria (NSLAB) are Lb. paracasei, Lb. plantarum, Lb. curvatus and Lb. brevis. In fermented yak milk are present Lb. fermentum, Lb. helveticus and Lb. curvatus, showing that lactobacilli could play an important role in the fermentation of yak milk. Lb. paracasei, Lb. brevis and Lb. plantarum are dominant NSLAB of Spanish artisanal goat cheese (16). Lb. paracasei, Lb. plantarum, Lb. pen-tosus, Lb. rhamnosus and Lb. curvatus are isolated from Pecorino Siciliano cheese du-ring ripening (17). All of these LAB also were frequently found in the various traditional fermented products in many countries of Balkan, Europe and in the world (1, 18-29).

Importance of indigenous lactobacilli in dairy industry

Lactobacilli have a great industrial significance, since they are widely used as starter cultures for a variety of fermented products (30, 31). They are present in raw milk and dairy products such as cheeses, yoghurts and fermented milks (32). Strains of lactobacilli from non-dairy foods can also be a potential source of novel starter and adjunct culture for cheese making (33). Lactobacilli during ripening of cheese continue breakdown of rest lactose, but are primarily important in the process of protein breakdown. Proteins are readily degradate by lactobacilli, because they previously breakdown by peptidases and proteinases and enzymes released upon autolysis lactococci (10). Some species of lacto-bacilli are traditionally used in the production of various fermented milk, meat and other products. Lb. helveticus and Lb. delbrueckii subsp. bulgaricus are used in the production of cheeses such as Parmesan, Mozzarella, Provolone and Swiss cheeses. In addition to


Review paper

41

protective roles achieved by lowering the pH, the used lactobacilli positively influence the taste and other organoleptic properties of the final product. Lactobacilli during meta-bolism of sugars and amino acids produce aromatic compounds such as diacetyl, acetate, benzaldehyde, cresol, skatole and various amines, carbonyl, and sulfur compounds, which have a positive effect on the flavor of the product (34). Also, very important is the activity of extracellular proteinases through the production or removal of various peptides that directly influence the taste and texture of the product (35). The action of proteinases may produce bioactive peptides with different favorable characteristics (e.g., anti-hyper-tensive, anti-microbial, antioxidant peptides) (36). Lactobacilli significantly influence the final characteristics of the product because of tolerance to low pH (a pH lower limit of 3.6 to 4), and that the frequently dominant bacteria in food in the last stages of fermenta-tion (37). The Lactobacillus genus has widespread use as probiotics and is generally re-cognized as safe organisms (GRAS) (38). Lb. brevis, Lb. buchneri, Lb. casei, Lb. kefiri, Lb. paraplantarum, Lb. plantarum have QPS (qualified presumption of safety) status of EFSA (39, 40). Their favorable activity is indicated in the treatment of various types of diarrhea, inflammatory bowel disease and syndrome, lactose intolerance, different and even allergic and cardiovascular diseases, and some cancers (41). Lb. plantarum, Lb. bre-vis, Lb. casei and Lb. parabuchneri isolated from Brazilian ovine cheese show probiotic potential, resistance to gastric juices and bile salts and inhibition of food pathogens Liste-ria monocytogenes, Staphylococcus aureus, Bacillus cereus, Escherichia coli and Salmo-nella typhimurium (42). Lactobacilli possess inhibitory activity towards the multiplication of enteropathogens, also spoilage causing bacteria, because they produce antimicrobial substances (43, 44). Antimicrobial capability of lactobacilli has also been reported (45, 46, 47). They also showed antagonistic activity against Helicobacter pylori (48). Lacto-bacilli from Zlatar cheese showed good proteolytic activity, as production of bacteriocin-like substance (13). Technological and flavor formation abilities of the wild strains (Lb. casei, Lb. plantarum) can be used as starters (24). The lactobacilli community is impor-tant in the production of Mozzarella cheese (26). The role of Lb. bulgaricus, together with S. thermophilus is in the milk acidification, synthesis of aromatic compounds, development of texture and viscosity of dairy products. Aroma formation is based mainly on the production of acetaldehyde, acetoin and diace-tyl, exopolysaccharides, bacteriocins, proteinases and peptidases. Bacteriocin "bulgari-can" created by Lb. bulgaricus operates on Bacillus, Streptococcus, Staphylococcus, Sar-cina, Pseudomonas, Serratia and Escherichia (49). Lb. acidophilus stabilizes the microflora of the gastrointestinal tract of humans and animals. Fermented milk, combining Lb. acidophilus and Lb casei, is available in the pre-vention of antibiotic-associated diarrhea (50). Lb. helveticus is found predominantly in the fermented of dairy products and mainly used for the manufacture of cheeses such as Grana and Provolone. Some interesting cha-racteristics of this microorganism are its ability to produce high quantities of lactic acid in milk, its acid tolerance, and the capacity to express a complex proteolytic enzyme system. This enzyme system, which comprises proteinases, endo peptidases and exopeptidases, makes an important contribution to the decrease of ripening time, the acceleration of fla-vor development and the reduction of bitterness (51). Lb. helveticus is obligately homo-


Review paper

42

fermentative LAB. Consumption of cheese containing a Lb. helveticus - exhibits im-munoregulatory actions, including an increase in the regulatory T cell population and reduction in proinflammatory cytokine production in mice (52). Lb. plantarum is a ubiquitous microorganism that is able to colonize several eco-logical niches, including vegetables, meat, dairy substrates, and the gastro-intestinal tract (53). It has metabolic capacity and industrial applications (54, 55). Lb. plantarum showed activity against Salmonella spp. and Pseudomonas aeruginosa. It is resistant to nisin-additive for fermented dairy products (56). It has a strong inhibitory activity against Salmonella typhi, E. coli, S. aureus, E. faecalis and Citrobacter spp. (57). Lb. plantarum, Lb. pentosus, Lb. paraplantarum and Lc. lactis showed bacteriophage resistance on 41 phages from the Chr. Hansen phage collection, as well as good acidification speed and EPS production (58). Lb. paraplantarum survives simulated passage through the gastrointestinal tract when resuspended in the milk (40). It produces bacteriocins that inhibited the growth of Liste-ria monocytogenes, Listeria innocua and several lactic acid bacteria. It was also observed that Lb. paraplantarum tolerated exposure to the pH 3.5 and 0.3% bile salts for up to 180 min (59). It is facultative heterofermentative LAB, probiotic starter isolated from fermented sausage, and it is a great producer of bacteriocins (60). Lb. plantarum and Lb. paracasei are resistant to acid and bile salts, and they can be used as potentially probiotic bacteria (61). Lb. plantarum, Lb. paracasei ssp. paracasei isolated from Suero costeño, can fermente sugars D-Lactose, D-Glucose and D-Galactose. Also, they use other carbo-hydrates, and it is possible to use these strains as starter cultures for other fermentations (62). Lb. plantarum and Lb. paracasei isolated from Istrian cheese have a strong aci-dification ability, as well as proteolitic and antimicrobial activity (63). Lb. plantarum and Lb. brevis isolated from Brasilian food could be used as probiotic, due to the tolerance to low pH and bile salts. They exhibited antagonistic activity towards the pathogens L. mo-nocytogenes and S. aureus, and adhere to the human intestinal epithelial cell line (Caco-2) (64). Antifungal activity of phenyllactic acid and 4-hydroxyl-phenyl-lactic acid isola-ted from Lb. plantarum was found (65), as well as the antifungal activity of Lb. brevis on carcinogenic, toxigenic and allergenic fungi of genera Aspergillus, Fusarium, Penicillium and Trichoderma (66). Lb. paracasei ssp. paracasei produces bacteriocins, exhibits auto- and co-aggrega-tion, and possesses surface polysaccharides and proteins that interact with collagen or si-milar molecules of eukaryotic cells. The probiotic potential of lactobacillus depends on the surface characteristics of bacterial cells (40). Lb. paracasei ssp. paracasei formed coaggregate with L. innocua, E. coli or with Salmonella enterica ser. typhimurium (67). Lb. paracasei ssp. paracasei isolated from Bukuljac cheese showed a high proteolytic activity and hydrolyzed alpha(s1)- and beta-caseins. This species produces diacetyl, and exhibited antimicrobial activity (11). Lb. paracasei ssp. paracasei produce bacteriocin, exhibited antimicrobial activity against S. aureus and B. cereus and grew well in simula-ted gastrointestinal conditions (68). It is effective in the regulation of blood cholesterol and pressure, prevention of gastric mucosal lesion, immunomodulation and alleviation of allergies, anti-osteoporosis, and inhibition the fat tissue accumulation (69). Also, it pos-sess strong antagonistic properties against Salmonella enterica ssp. enterica. It is sensi-


Review paper

43

tive to antibiotics and did not produce biogenic amine except for tyramine (70). Lb. casei and Lb. plantarum efficiently hydrolyzed milk protein (71). Most isolates of Lb. casei are ribose positive (72). Lb. coryniformis subsp. coryniformis produces proteinaceous antifungal compounds, with a maximum activity at the pH between 3.0 and 4.5, which then decreased to the pH 6.0, and was lost at higher pH values (73). Some of the strains Lb coryniformis grew at 450C, also in 6.5% NaCl, produce lactate, also diacetyl and acetoin (74). Consumption of Lb. coryniformis and Lb. fermentum strains enhances the immune response and may have a clinical benefit in protection from infections (75). Lb. coryniformis is suitable as a star-ter in food manufacturing processes, and has a role in eliminating food originated car-cinogens, such as sodium nitrite and ethyl carbamate (76). Lb. coryniformis was an un-common species found in Churpi cheese. Its isolates can be potentially used for the deve-lopment of defined strain starter for Churpi cheese (77). Lb. curvatus, heterofermentative lactobacillus, produces bacteriocin with a strong antilisterial activity, and antifungal activity against Cladosporium and Fusarium ssp. was detected. Bacteriocin is heat and pH stable and its mode of action is bacteriostatic (78, 79). Lb. curvatus is able to grow at low temperature (2-4 °C), and this is important for fermentation of vegetables, meat product, and sake (80). Lb. rhamnosus prevents the appearance of antibiotic-associated diarrhea (81). Lb. rhamnosus was originally isolated from human intestinal flora. It can shorten the duration and ameliorates the symptoms of infantile rotavirus diarrhea, and has some ef-fect on preventing atopic diseases among infants and to modulate immune responses (82, 83). Lb. rhamnosus and Lb. plantarum strains isolated from artisanal Coalho cheese are probiotic candidates to be used in fermented dairy products. They have bacteriocino-genic potential (84). Lb pentosus creates exopolysaccharides (EPSs) which improve the texture of yogurt by increasing the viscosity and binding water, and interacting with other milk constitu-ents, such as proteins and micelles, to strengthen the rigidity of the casein network. EPSs produced by LAB have beneficial effects on human health as immunomodulators, anti-tumor prebiotic effects and cholesterol-lowering ability. EPS can positively affect gut health (85). Lb. buchneri has a high cholesterol-reducing rate, acid and bile tolerance, and anti-microbial activity (86). Lb. sake in combination with Lc. lactis ssp. lactis produced a good quality cheese. This bacterium plays an inhibitory effect against food-spoiling bacteria and food-borne pathogens, including L. monocytogenes, a gram-positive and pathogenic bacterium (87). Lb. fermentum exhibits probiotic characteristics and has the potential to be a candidate as a probiotic. It produced inhibitory compound including H2O2, bacteriocin and bio-surfac-tants, to inhibit the growth of intestinal and urogenital pathogens. It is effective in de-creasing the intestinal pathogens and increasing the ratio of probiotic bacteria in healthy humans (88). Lb. kefiri isolated from a dairy product have a great potential as a probiotic and can be used also for producing functional foods. Gram positive pathogens, L. monocytogenes, B. cereus, E. faecalis, S. aureus, showed sensibility to Lb. kefiri strains, as well as to


Review paper

44

Gram negative bacteria (P. aeruginosa, Salmonella enteritidis, Shigella flexneri) (89). Lb. kefiri is used to control gut inflammatory disorders (90). Secretion products and surface proteins from Lb. kefiri have a protective action against the invasion of Salmonella ente-rica serovar enteritidis to Caco-2 cells and also against the cytotoxic effects of clostridial toxins on Vero cells (91). Lb. kefiri prevents S. enteritidis interaction with epithelial cells (92).

Table 1. Characteristics of some indigenous cheeses of some countries of the Balkans

Name of cheese Country Raw milk Type of cheese Senzory characteristics

Njeguški cheese Montenegro sheep, cow

and goat semi–hard to hard,

full-fat sheep cheese

round, low cylinder, compact surface, crust is golden–yellow, a few ‘cheese eyes’ evenly distributed, round, shiny, pleasant, slightly sour–milky and moderately taste.

Pljevaljski cheese Montenegro sheep and

cow

white, brined, soft to semi–hard, fat to full fat sheep

cheese

uniform triangle or rectangle, slices without crust, uniform shape, very slight porosity, brittle, curd gentle, porcelain–looking, milky–sour taste, moderately salty with pleasant, clearly expressed milky–sour smell

Pirotski cheese Serbia sheep soft, full–fat brined cheese

square slice with regular dimensions and shapes (triangular), surface is smoothly, color is white, shiny and homogeneous few small-sized cheese eyes and technological cavities filled with whey solid cheese body, slightly crumbly and soft after ripening, moderately to strongly salty and sour, clearly expressed, taste of sheep milk, pleasant taste, typical.

Sjenički cheese Serbia sheep soft, full–fat brined cheese

triangular shaped slices with flat sides, one side is convex, surface is smooth with markedly white color equal, ivory white color, few lentil sized cheese eye solid cheese body, brittle, well linked, not crumbly pleasant – of sheep milk, typical lactic acid odor, clearly expressed, aromatic, moderately to strongly salty and slightly piquant.

Livanjski cheese Bosnia and

Herzegovina cow and

sheep hard, full–fat

golden–yellow, a couple of medium–sized cheese eyes firm, not too hard, moderately elastic full, moderately salty, piquant pleasant, typical of sheep cheeses

Travnički/Vlašićki cheese

Bosnia and Herzegovina

sheep white, full–fat brined

cheese

white, typical of sheep milk cheeses none to few irregular shaped cheese eyes firm, not too hard, well connected, porcelain–like breaks, spreadable, easy to cut and crumbly milky sour taste, clean, moderately salty, typical of sheep cheeses clearly pronounced, pleasant and without foreign additives, typical of sheep cheeses.

Dolenjski cheese Slovenia sheep hard, full–fat upper and peripheral sides are slightly convex; cheese is evenly grey–beige in color curd is hard, firm, compact but not crumbly aromatic, clean, full to slightly piquant, typical.

Kraški cheese Slovenia sheep hard, full–fat upper side of the cheese is flat, peripheral side is slightly convex, cheese is grey–beige in color curd is hard, firm and granular but not crumbly aromatic, intensive.

Istarski cheese Croatia sheep hard sheep cheese

cylindrical, smooth skin, dark yellow to light brown in color on the section of the cheese a few cheese stitches size 0.5–1 mm are allowed ‘rocky’ structure full flavor, characteristic of sheep cheeses

Krčki cheese Croatia sheep hard, full–fat

the rind of the cheese is smooth, cylindrical shape and the color varies from light yellow to golden the body is compact and bound with even shell–like surface at braking, but not crumbly, on the cut surface a handful of tiny cheese eyes can be seen ‘rocky’ structure depends on the duration of ripening of cheese, and ranges from moderate to strongly acidic and strong, characteristic of hard type sheep cheese.

Paški cheese Croatia sheep hard, full-fat Crust is hard and smooth, yellow to light brown, on the sec-tion with rarely spaced, tiny circular cheese eyes, or without them, piquant, typical of sheep sheese, with fine “crystals”.


Review paper

45

Figure 1. Njeguški cheese

CONCLUSION The diversity of lactic acid bacteria in traditional dairy products and environment re-presents a great potential for application in dairy industry. The utilization of indigenous lactic acid bacteria can lead to potential starter cultures with the necessary properties for typical local products that are well accepted by the local population. Lactobacilli can fer-ment sugars and decompose proteins. During the metabolism of sugars and amino acids they also produce aromatic compounds such as diacetyl, acetate, benzaldehyde, cresol, skatole, and various amines, carbonyl, and sulfur compounds, which have a positive effect on the flavor of the dairy product. They also produce bacteriocins and have anta-gonistic activity against pathogen bacteria, and most of them have a probiotic activity. In addition, the use of indigenous lactic acid bacteria as starter cultures would allow the production of cheese with designated geographical origin that could be placed on the in-ternational market. Consequently, they are a challenge for further research and their possible practical application in the dairy industry.

REFERENCES 1. Havranek, J. An Atlas of sheep cheeses of the Countries of the Western Balkans,

University of Zagreb, Faculty of Agriculture, Zagreb, 2014; pp 1-44, ISBN 978-953-7878-22-1.

2. Bojanic Rasovic, M.; Mayrhofer, S.; Martinovic, A.; Dürr, K.; Domig, J. K. Lactococci of Local Origin as Potential Starter Cultures for Traditional Montenegrin Cheese Production. Food Technol. Biotechnol. 2017, 55 (1), 55-66

3. Klijn, N.; Weerkamp, A.H.; Willem, M. De Vos, N. Detection and characterization of lactose-utilizing Lactococcus spp. In natural ecosystems. Appl. Environ. Microbiol. 1995, 61 (2), 788-792.

4. Topisirović , Lj.; Kojić, M.; Fira, D.; Golić, N.; Strahinić, I.; Lozo, J. Potential of lactic acid bacteria isolated from specific natural niches in food production and pre-servation. Int. J. Food Microbiol. 2006, 112, 230-235.


Review paper

46

5. Terzic-Vidojevic, A.; Vukasinovic, M.; Veljovic, K.; Ostojic, M.; Topisirovic, L. Characterization of microflora in homemade semi-hard white Zlatar cheese, Int. J. Food Microbiol. 2007, 114(1), 36-42.

6. Ostojić, M.; Topisirović, Lj. Značaj autohtonih sojeva bakterija mlečne kiseline u zaštiti geografskih oznaka mlečnih proizvoda. Prehrambena Ind. 2008, 1-2, 12-17.

7. Abd El Gawad, I.A.; Abd El Fatah A.M.; Al Rubayyi K.A. Identification and characterization of dominant lactic acid bacteria isolated from traditional Rayeb milk in Egypt. J Am Sci. 2010, 6, 728-35.

8. Azadnia P., Zamani M.H., Ahmad Ghasemi S., Khalegh Babaki A., Karimi Jashni M. Taarof N. Isolation and Identification of Thermophilic Lactobacilli from Traditio-nal Yoghurts of Tribes of Kazerun. Int. J. Anim. Vet. Adv. 2011, 10, 774-776.

9. Parente, E., Cogan, TM. Starter cultures: general aspects. Fox PF et al.(eds). Cheese: chemistry, physics and microbiology, vol. 1. General aspects. Amsterdam, NL: Else-vier, 2004, p 123-147.

10. Robinson, K. R. Dairy microbiology handbook, Wiley-Interscience, New York, 2002; p 765.

11. Nikolic, M.; Terzic-Vidojevic A.; Jovcic B.; Begovic J.; Golic N.; Topisirovic L. Characterization of lactic acid bacteria isolated from Bukuljac, a homemade goat's milk cheese. Int. J. Food Microbiol. 2008, 122 (1-2), 162-70.

12. Jordan, T.K.N., Cogan, T.M. Heat resistance of Lactobacillus spp. isolated from Cheddar cheese. Lett. Appl. Microbiol. 1999, 29, 136-140.

13. Veljovic K., Terzic-Vidojevic A., Vukasinovic M., Strahinic I., Begovic J., Lozo J., Ostojic M., Topisirovic L. Preliminary characterization of lactic acid bacteria isolated from Zlatar cheese, J. Appl. Microbiol. 2007, 103, 2142-2152.

14. Topisirović, Lj.; Veljović, K.; Terzić Vidojević, A.; Strahinić, I.; Kojić, M. Compara-tive analysis of antimicrobial and proteolytic activity of lactic acid bacteria isolated from Zlatar cheese. Genetika, 2007, 39 (2), 125 -138.

15. Terzić-Vidojević A.; Veljović K.; Tolinacki M.; Nikolić M.; Ostojić M.; Topisirovic Lj. Characterization of lactic acid bacteria isolated from artisanal Zlatar cheese pro-duced at two different geographical location. Genetika, 2009, 41 (1), 117 -136.

16. Sancheza I.; Sesenab S.; Povedaa J.M.; Cabezasa L.; Palopb L. Phenotypic and genotypic characterization of lactobacilli isolated from Spanish goat cheeses. Int. J. Food Microbiol. 2005, 102, 355-362.

17. Vernile, A.; Giammanco, G.; Spano, G.; Thomas, P.; Beresford, P.; Fox F.; Massa, S. Genotypic characterization of lactic acid bacteria isolated from traditional Pecorino Siciliano cheese. Dairy Sci. Technol. 2008, 88, 619-629.

18. Golić, N., Cadež ,N.; Terzić-Vidojević, A.; Suranská, H.; Beganović, J.; Lozo, J.;Kos B.; Sušković, J.; Raspor, P.; Topisirović, L. Evaluation of lactic acid bacteria and yeast diversity in traditional white pickled and fresh soft cheeses from the mountain regions of Serbia and lowland regions of Croatia. Int. J. Food Microbiol. 2013, 166(2), 294-300

19. Jokovic N, Nikolic M, Begovic J, Jovcic B, Savic D, Topisirovic L. A survey of the lactic acid bacteria isolated from Serbian artisanal dairy product kajmak. Int. J. Food Microbiol. 2008, 127, 305-11.


Review paper

47

20. El-Soda, M.; El-Ziney, M.; Awad, S.; Osman, G.; Omran, N., Gamal, G.; Ezzat, N.; El-Shafei, H. A culture collection of lactic acid bacteria isolated from raw milk and traditional Egyptian dairy products. Egypt. J. Dairy Sci. 2003, 31, 217-226.

21. El-Baradei, G.; Delacroix-Buchet, A.; Ogier, J.C. Biodiversity of bacterial ecosystems in traditional Egyptian Domiati cheese. Appl. Environ. Microbiol. 2007, 73(4), 1248-1255.

22. Gori, K.; Ryssel, M.; Arneborg, N.; Jespersen, L. Isolation and identification of the microbiota of Danish farmhouse and industrially produced surface-ripened cheeses. Microb Ecol. 2013, 65(3), 602-615.

23. Pogačić, T.; Mancini, A.; Santarelli, M.; Bottari, B.; Lazzi, C.; Neviani, E.; Gatti, M. Diversity and dynamic of lactic acid bacteria strains during aging of a long ripened hard cheese produced from raw milk and undefined natural starter. Food Microbiol. 2013, 36 (2), 207-215.

24. Randazzo, C.L.; Pitino, I.; De Luca, S.; Scifò, G.O.; Caggia, C. Effect of wild strains used as starter cultures and adjunct cultures on the volatile compounds of the Pecorino Siciliano cheese. Int. J. Food Microbiol. 2008, 122(3), 269-278.

25. Bhardwaj, A.; Malik R. K.; Chauhan P. Functional and safety aspects of enterococci in dairy foods. Indian J. Microbiol. 2008, 48(3), 317-325.

26. De Angelis, M.; de Candia, S.; Calasso, M.P.; Faccia, M.; Guinee, T.P.; Simonetti, M.; Gobbetti, M. Selection and use of autochtonous multiple strain cultures for the manufacture of high moisture. Int. J. Food Microbiol. 2008, 125, 123-132.

27. Wouters, J.T.M.; Ayad, E.H.E.; Hugenholtz, J.; Smit, G. Microbes from raw milk for fermented dairy products. Int. Dairy J. 2002, 12, 91-109.

28. Quadghiri, M.; Vancanneyt, M.; Vandamme, P.; Naser, S.; Gevers, D.; Lefebvre, K.; Swings, J.; Amar, M. Identification of lactic acid bacteria in Moroccan raw milk and traditionally fermented skimmed milk ‘lben’. J. Appl. Microbiol. 2009, 106, 486-495.

29. Ali, AA. Isolation and Identification of lactic acid bacteria isolated from traditional drinking yoghurt in Khartoum State, Sudan. Curr. Res. Bacteriol. 2011, 4, 16-22.

30. Vujcic, M.; Topisirović L. Molecular Analysis of the Rolling-Circle Replicating Plas-mid pAl of Lactobacillus plantarum A112 M. Appl. Environ. Microbiol. 1993, 59 (1), 274-280.

31. Magdoub, M. N. I.; Hassan, Z. M. R.; Effat, B.A. M.; Sadek, Z. I. M.; Tawfik, N. F.; Mabrouk, A.M. M. Probiotic Properties of Some Lactic Acid Bacteria Isolated from Egyptian Dairy Products. Int. J. Curr. Microbiol. Appl. Sci. 2015, 4 (12), 758-766.

32. Coeuret, V.; Dubernet S.; Bernardeau, M.; Gueguen, M.; Vernoux, J. Isolation, cha-racterisation and identification of lactobacilli focusing mainly on cheeses and other dairy products. Lait. 2003, 83 (4), 269-306.

33. Ciocia, F.; Mcsweeney, P.; Piraino, P.; Parente, E. Use of dairy and non-dairy Lacto-bacillus plantarum, Lactobacillus paraplantarum and Lactobacillus pentosus strains as adjuncts in cheddar cheese. Dairy Sci. Technol. 2013, 93 (6), 623-640.

34. Tanous, C.; Kieronczyk, A.; Helinck, S.; Chambellon, E.; Yvon, M. Glutamate de-hydrogenase activity: a major criterion for the selection of flavour-producing lactic acid bacteria strains. Antonie van Leeuwenhoek. 2002, 82(1-4), 271-278.


Review paper

48

35. Savijoki, K.; Ingmer, H.; Varmanen, P. Proteolytic systems of lactic acid bacteria. Appl. Microbiol. Biotechnol. 2006, 71(4), 394-406.

36. Kirin, S. Bitter taste - defect of cheese. Mljekarstvo, 2001, 51(4), 327-337. (in croatian) 37. Hammes, C., Hertel W. Genus Lactobacillus. de Vos W. et al.(eds.). Bergey’s Manual

of Systematic Bacteriology: Volume 3: The Firmicutes. Springer Science & Business Media, 2009; pp 465-511.

38. Hoque, M.Z.; Akter, F.; Hossain, K.M.; Rahman, M.S.M.; Billah, M.M.; Islam, K.M.D. Isolation, Identification and Analysis of Probiotic Properties of Lactobacillus spp. from Selective Regional Yoghurts, World J. Dairy Food Sci. 2010, 5 (1), 39-46.

39. EFSA. Scientific opinion on the maintenance of the list of QPS microorganisms inten-tionally added to food or feed. EFSA Journal, 2009, 7, 1-92.

40. Nikolić, M.: Characterization of surface molecules from bacterial cells involved in the potential probiotic activity of natural lactobacilli isolates: PhD Thesis, University of Belgrade, 2012.

41. Pandey, K.R.; Naik, S.R.;Vakil, B.V. Probiotics, prebiotics and synbiotics - a re-view. J. Food Sci. Technol. 2015, 52(12), 7577-7587.

42. Meira, S.M.; Helfer, V.E.; Velho, R.V.; Lopes, F.C.; Brandelli, A. Probiotic potential of Lactobacillus spp. isolated from Brazilian regional ovine cheese. J. Dairy Res. 2012, 79(1), 119-27.

43. Verdenelli, M.C.; Ghelfi, F.; Silvi, S.; Orpianesi, C.; Cecchini, C.; Cresci, A. 2009. Probiotic properties of Lactobacillus rhamnosus and Lactobacillus paracasei isolated from human faeces. Eur. J. Nutr. 2009, 48(6), 355-363.

44. Vasiee, A. R.; Tabatabaei Yazdi, F.; Mortazavi, A.; And Edalatian, M. R. Isolation, identification and characterization of probiotic Lactobacilli spp. from Tarkhineh, Int. Food Res. J. 2014, 21(6), 2487-2492.

45. Chowdhury, A.; Hossain, N. Md.; Mostazir, NJ.; Fakruddin, Md.; Morsaline, B. Md., Ahmed, M. M. Screening of Lactobacillus spp. from Buffalo Yoghurt for Probiotic and Antibacterial Activity. J. Bacteriol. Parasitol. 2012, 3 (8), 1-5, doi:10.4172/2155-9597.1000156.

46. Murry, A.C. Jr.; Hinton A. Jr.; Morrison, H. Inhibition of growth of Escherichia coli, Salmonella typhimurium and Clostridium perfringens on chicken feed media by Lactobacillus salivarius and Lactobacillus plantarum. Int. J. Poult. Sci. 2004, 3(9), 603-607.

47. Osuntoki, A.A.; Ejide, O.R.; Omonigbehin, E.A. Antagonistic effects on Enteropatho-genic and plasmid analysis of Lactobacilli isolated from fermented Dairy products. Biotechnology, 2008, 7, 311-316.

48. Hütt, P.; Shchepetova, J.; Lõivukene, K.; Kullisaar, T.; Mikelsaar, M. Antagonistic activity of probiotic lactobacilli and bifidobacteria against entero- and uropathogens. J. Appl. Microbiol. 2006, 100, 1324-1332.

49. Zourari, A.; Accolas, Jp.; Desmazeaud, Mj. Metabolism and biochemical characteris-tics of yogurt bacteria. A review. Lait. 1992, 72 (1), 1-34.

50. Beausoleil, M.; Fortier, N.; Guénette, S.; L’Ecuyer, A.; Savoie, M.; Franco, M.; Lachaîne, J.; Weiss, K. Effect of a fermented milk combining Lactobacillus acido-philus CL1285 and Lactobacillus casei in the prevention of antibiotic-associated diar-


Review paper

49

rhea: A randomized, double-blind, placebo-controlled trial. Canadian Journal of Gastroenterology, 2007, 21(11), 732-736.

51. Fortina, M.G.; Nicastro, G.; Carminati, D.; Neviani, E.; Manachini, P.L. Lacto-bacillus helveticus heterogeneity in natural cheese starters: the diversity in phenotypic characteristics. J. Appl. Microbiol. 1998, 84, 72-80.

52. Yamashita, M.; Ukibe, K.; Uenish, H.; Hosoya, T.; Sakai, F.; Kadooka, Y. Lacto-bacillus helveticus SBT2171, a cheese starter, regulates proliferation and cytokine production of immune cells. J. Dairy Sci. 2014, 97 (8), 4772-4779.

53. Siezen, R.J.; Tzeneva, V.A.; Castioni, A.; Wels, M.; Phan, H.T.; Rademaker, J.L.; Starrenburg, M.J.; Kleerebezem, M.; Molenaar D.J.E; van Hylckama, V. Phenotypic and genomic diversity of Lactobacillus plantarum strains isolated from various envi-ronmental niches. Environ Microbiol. 2010, 12(3), 758-773.

54. Siezen, R.J.; van Hylckama Vieg, J.ET. Genomic diversity and versatility of Lacto-bacillus plantarum, a natural metabolic engineer. Microb. Cell Fact. 10 Suppl 1:S3. 2011, doi: 10.1186/1475-2859-10-S1-S3. Epub 2011 Aug 30.

55. Martinović A. The use of different strains of lactic acid bacteria in the industrial production of Njeguški cheese. PhD Thesis, University of Belgrade, 2005.

56. Lozo, J.; Vukasinovic, M.; Strahinic, I.; Topisirovic, L. Characterization and anti-microbial activity of bacteriocin 217 produced by natural isolate Lactobacillus para-casei subsp. paracasei BGBUK2-16. J. Food Prot. 2004, 67(12), 2727-2734.

57. Gharaei-Fathabad, E.; Eslamifar, M. Isolation and Applications of one strain of Lacto-bacillus paraplantarum from tea leaves (Camellia sinensis). Am. J. Food Technol. 2011, 6, 429-434.

58. Martinović, A.; Radulović, Z.; Wind, A.; Janzen, T.; Obradović, D. Isolation and cha-racterization of bacterial flora from farmhouse fermented milk products of Serbia and Montenegro. Acta Vet. Belgrade. 2005, 55 (4), 307-318.

59. Tulini, F.L.; Winkelströter, L.K.; De Martinis E.C. Identification and evaluation of the probiotic potential of Lactobacillus paraplantarum FT259, a bacteriocinogenic strain isolated from Brazilian semi-hard artisanal cheese. Anaerobe, 2013, 22, 57-63.

60. Liu, L.; Li, P. Complete genome sequence of Lactobacillus paraplantarum L-ZS9, a probiotic starter producing class II bacteriocins. J. Biotechnol. 2016, 222, 15-16.

61. Niazi Amraii, H.; Abtahi, H.; Jafari, P.; Mohajerani, H.R.; Fakhroleslam, MR.; Akbari, N. In Vitro Study of Potentially Probiotic lactic Acid Bacteria Strains Isolated From Traditional Dairy Products. Jundishapur J. Microbiol. 2014, 7(6), e10168.

62. Cueto, C.; García, D; Garcés, F.; Cruz J. Preliminary studies on the microbiological characterization of lactic acid bacteria in suero costeño, a Colombian traditional fermented milk product. Rev. Latinoam. Microbiol. 2007, 49(1-2),12-8.

63. Hulak, N.; Žgomba Maksimović, A.; Kaić, A.; Skelin, A., Mrkonjić Fuka, M. Indige-nous strains of Lactobacillus isolated from the Istrian cheese as potential starter cul-tures. Mljekarstvo, 2016, 66(4), 282-292.

64. Ramos, C. L.; Thorsen, L.; Schwan, R. F.; Jespersen, L. Strain-specific probiotics properties of Lactobacillus fermentum, Lactobacillus plantarum and Lactobacillus brevis isolates from Brazilian food products. Food Microbiol. 2013, 36 (1), 22-29.


Review paper

50

65. Lavermicocca, P.; Valerio, F.; Evidente, A.; Lazzaroni, S.; Corsetti, A.; Gobbetti, M. Purification and characterization of novel antifungal compounds from the sourdough Lactobacillus plantarum strain 21B. Appl. Environ. Microbiol. 2000, 66(9), 4084-4090.

66. Tropcheva, R.; Nikolova, D.; Evstatieva, Y.; Danova, S. Antifungal activity and iden-tification of Lactobacilli, isolated from traditional dairy product "katak". Anaerobe. 2014, 28, 78-84. doi: 10.1016/j.anaerobe.2014.05.010. Epub 2014 Jun 2.

67. Lozo, J.; Jovcic, B.; Kojic, M., Dalgalarrondo, M.; Chobert, J.M.; Haertlé, T.; Topi-sirovic, Lj. Molecular Characterization of a Novel Bacteriocin and an Unusually Large Aggregation Factor of Lactobacillus paracasei subsp. paracasei BGSJ2-8, a Natural Isolate from Homemade Cheese. Curr. Microbiol. 2007, 55(3), 266-271.

68. Radulović, Z.; Petrović, T.; Nedović, V.; Dimitrijević, S.; Mirković, N.; Petrušić, M.; Paunović, D. Characterization of autochthonous Lactobacillus paracasei strains on potential probiotic ability. Mljekarstvo. 2010, 60 (2), 86-93.

69. Chiang, S.S.; Pan, TM. Beneficial effects of Lactobacillus paracasei subsp. paracasei NTU 101 and its fermented products. Appl. Microbiol. Biotechnol. 2012, 93(3), 903- 916.

70. Zheng, Y.; Lu, Y.; Wang, J.; Yang, L.; Pan, C.; Huang, Y.. Probiotic Properties of Lactobacillus Strains Isolated from Tibetan Kefir Grains. PLoS ONE. 2013, 8(7), e69868.

71. Vukotić, G.: Proteinases of lactic acid bacteria: diversity in mesophilic species lacto-bacilli and impact on bacteriocin activity. PhD Thesis, University of Belgrade, 2016.

72. Martinovic, A.; Moe, K.M.; Romeih, A.; Vogensen, F.K.; Østlie.; Skeie, S. Growth of adjunct Lactobacillus casei in Cheddar cheese differing in milk fat globule membrane components. Int. Dairy J. 2013, 31 (2), 70-82.

73. Magnusson, J.; Schnürer, J. Lactobacillus coryniformis subsp.coryniformis Strain Si3 Produces a Broad-Spectrum Proteinaceous Antifungal Compound. Appl. Environ. Microbiol. 2001, 67(1), 1-5.

74. Hegazi, F.Z.; Abo-Elnaga, I.G. Characters of Lactobacillus coryniformis, isolated from an Iraqi cheese. Zentralbl. Bakteriol. Naturwiss. 1980, 135(3), 205-211.

75. Redondo, N.; Nova, E.; Gheorghe, A.; Díaz, L. E.; Hernández, A.; Marcos, A. Eva-luation of Lactobacillus coryniformis CECT5711 strain as a coadjuvant in a vaccina-tion process: a randomised clinical trial in healthy adults. Nutr.Metab. 2017, 14, 2. doi.org/10.1186/s12986-016-0154-2.

76. Fang, F.; Feng, T.; Du, G.; Chen, J. Evaluation of the impact on food safety of a Lactobacillus coryniformis strain from pickled vegetables with degradation activity against nitrite and other undesirable compounds. Food Addit. Contam. Part A. 2016, 33(4), 623-630.

77. Prashant, T. S.K.; Singh, R.; Rameshwar, S.; Subhash, C. G.; Dilip, K. A.; Balwindar, K. J.; Dinesh, K. Phenotypic and genotypic characterization of lactobacilli from Chur-pi cheese. Dairy Sci. Technol. 2009, 89 (6), 531-540.

78. Ahmadova, A.; Dimitrov Todorov, S.; Hadji-Sfax, I.; Choiset Y.; Rabesona, H.; Messaoudi, S.; Kuliyev, A.; de Melo Franco, B.D.G.; Chobert, J.M.; Haertlé, T. Anti-


Review paper

51

microbial and antifungal activities of Lactobacillus curvatus strain isolated from homemade Azerbaijani cheese. Anaerobe. 2013, 20, 42-49.

79. Vogel, R. F.; Pohle, B. S.; Tichaczek, P.S.; Hammes, W. P. The competitive advan-tage of Lactobacillus curvatus LTH 1174 in sausage fermentations is caused by for-mation of curvacin A. Syst. Appl. Microbiol. 1993, 16, 457-462.

80. Erkman, O., Faruk Bozoglu, T. Food Microbiology: Principles into Practice, John Wiley and Sons Ltd, 2016, p 994.

81. Hickson, M. Probiotics in the prevention of antibiotic-associated diarrhoea and Clostridium difficile infection. Ther. Adv. Gastroenterol. 2011, 4(3), 185-197.

82. Salminen, M.; Tynkkynen, S.; Rautelin, H.; Saxelin, M.; Vaara M.; Ruutu, P.; Sarna, S.; Valtonen, V.; Jarvinen, A. Lactobacillus Bacteremia during a Rapid Increase in Probiotic Use of Lactobacillus rhamnosus GG in Finland. Clin. Infect. Dis. 2002, 35, 1155-1160.

83. Yan, F.; Polk, D.B. Lactobacillus rhamnosus GG: An Updated Strategy to Use Microbial Products to Promote Health. Functional Food Reviews. 2012, 4(2), 77-84.

84. dos Santos, K.M.O.; Vieira, A.D.S.; Buriti, F.C.A.; Frutuoso do Nascimento, J.C.; Silva de Melo, M.E.; Bruno, L. M.; de Fátima Borges, M.; Costa Rocha, C.R.; de Souza Lopes A.C.; de Melo Franco, B.D.G.; Dimitrov Todorov, S. Artisanal Coalho cheeses as source of beneficial Lactobacillus plantarum and Lactobacillus rham-nosus strains. Dairy Sci. Technol. 2015, 95, 209-230.

85. Elbanna, K.; Metry, W.; Elgarhy H. Exopolysaccharide from Lactobacillus pentosus Strain H2 and Its Impact on Rheological Properties and the Sensory Evaluation of Low Fat Yoghurt and UF-Soft Cheese. Int. J. Nutr. Food Sci. 2015, 4(5), 555-564.

86. Zeng, XQ.; Pan, D.D; Guo, Y.X. The probiotic properties of Lactobacillus buchneri P2. J. Appl. Microbiol. 2010, 108(6): 2059-66.

87. Hayaloglu, A.A.; Ozer, B.H.; Fox, P.F. Cheeses of Turkey: 2. Varieties ripened under brine. Dairy Sci. Technol. 2008, 88 (2), 225-244.

88. Tulumoglu, S.; Kaya, H.I.; Simsek, O. Probiotic characteristics of Lactobacillus fer-mentum strains isolated from tulum cheese. Anaerobe. 2014, 30, 120-125.

89. Carasi, P.; Díaz, M.; Racedo, S.; De Antoni, G.; Urdaci, M.; de los Angeles Serradell, M. Safety Characterization and Antimicrobial Properties of Kefir-Isolated Lactoba-cillus kefiri. BioMed Research International, 2014, 1-7.

90. Carasi, P. S.; Racedo, M.; Jacquot, C.; Romanin, D. E.; Serradell, M. A.; Urdaci, M. C. Impact of Kefir Derived Lactobacillus kefiri on the Mucosal Immune Response and Gut Microbiota J. Immunol. Res. 2015, 1-12.

91. Carasi, P.; Trejo, F. M.; Perez, P. F.; de Antoni, G. L.; Serradell, M. D. L. A. Surface proteins from Lactobacillus kefir antagonize in vitro cytotoxic effect of Clostridium difficile toxins.” Anaerobe. 2012, 18 (1), 135-142.

92. Golowczyc, M.A.; Mobili, P.; Garrote, G.L.; Abraham, A.G.; De Antoni, G.L. Protec-tive action of Lactobacillus kefir carrying S-layer protein against Salmonella enterica serovar enteritidis. Int. J. Food Microbiol. 2007, 118(3), 264-73.


Review paper

52

ПОТЕНЦИЈАЛ АУТОХТОНИХ ЛАКТОБАЦИЛА КАО СТАРТЕР КУЛТУРA У МЛЕЧНИМ ПРОИЗВОДИМА

Мирјана Бојанић Рашовић

Универзитет Црне Горе, Биотехнички факултет, Михаила Лалића 1, 20000 Подгорица, Црна Гора

Традиционална производња ферментисаних млечних производа укључује бакте-рије млечне киселине које су нормално присутне у млеку и производном окруже-њу. Ове бактерије млечне киселине представљају микробиолошку нишу географ-ског подручја и одговорне су за локалне врсте ферментисаних производа. У циљу стандардизације аутохтоних производа, основни захтев је примена одређених ауто-хтоних бактерија млечне киселине као стартер културa које утичу на њихове спе-цифичне карактеристике ферментацијом и утицајем на зрење. У процесу фермента-ције сира најчешће учествују бактерије рода Lactococcus и хомоферментативни лактобацили. Међутим, на процес зрења највише утичу тзв. нестартер бактерије млечне киселине - лактобацили и секундарнa микрофлорa. Лактобацили током зрења сира настављају разградњу остатка лактозе, али су првенствено важни у про-цесу разградње протеина. Током метаболизма шећера и аминокиселина, лактобаци-ли производе ароматична једињења која позитивно утичу на укус производа. Неке врсте лактобацила сe користе као пробиотици. Неки лактобацили производе бакте-риоцине, који спречавају раст патогена, као и многе микроорганизме кварењa хра-не. Аутохтони лактобацили би имали примену у производњи типичних локалних млечних производа који су добро прихваћени од стране локалног становништва. Осим тога, примена аутохтоних бактерија млечне киселине као стартер култура омогућило би производњу сирева са географским пореклом који би сe могли пласи-рати на међународном тржишту. Због тога су аутохтоне бактерије млечнe киселинe изазов за даље истраживање и могућу њихову практичну примeну у млечној индустрији. Кључне речи: бактерије млечне киселине, аутохтони сир, лактобацили, зрење сира

Received: 10 July 2017. Accepted: 11 September 2017.

APTEFF, 48, 1-323 (2017) UDC: 621.798.15:577.11]:621.7 https://doi.org/10.2298/APT1748053B BIBLID: 1450-7188 (2017) 48, 53-62


53

INFLUENCE OF STORAGE PERIOD ON PROPERTIES OF BIOPOLYMER PACKAGING MATERIALS AND POUCHES

Sandra N. Bulut*, Vera L. Lazić, Senka Z. Popović, Nevena M. Hromiš,

Danijela Z. Šuput

University of Novi Sad, Faculty of Technology Novi Sad, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia

Bilayer biodegradable films based on pumpkin oil cake (PuOC) and zein, as well as pouches made from these materials, were prepared, and the changes of their mechanical, physicochemical and barrier properties were analyzed during four weeks of storage. Heat seal quality of formed pouches and composition of the gas atmosphere in the pouches were also monitored. The results showed that the bilayer film had a thickness of 300 ± 10 (μm), and no its changes were observed during the storage time. The tensile strength of the tested film increased slightly in the third week, but the elongation at break showed a decreasing trend during the whole storage period. The decreases in the moisture content, total soluble matter and swelling of the obtained film, were also observed. After one month of storage, the O2 transmission rate of tested films, increased from 27 to 64 (ml/m2 24h 1 bar), and the CO2 gas transmission rate from 147 to 188 (ml/m2 24h 1 bar). The heat seal strength of the PuOC/Zein pouches decreased during the whole storage period. The percentage of O2 in PuOC/Zein pouches increased up to 7 times during the storage period; however, the percentage of CO2 decreased up to 18 times already after one week, and then remained stable in the rest of the storage period. These results are, to a smaller extent, due to the gas transmission rate through the material, especially for CO2 , and to a greater extent, due to the low heat seal strength, which decreased through the storage period, and probably influenced the content of the gases in the pouches. KEY WORDS: biopolymer films, pumpkin oil cake, zein, properties, storage period

INTRODUCTION Packaging is a necessary part of all food products. For each product, the role of packaging is to provide its protection from mechanical, physicochemical, microbiological and biological changes caused by the influence of environmental factors during storage time (1, 2). Today, polymer packaging materials are dominant in food packaging, primarily be-cause of their good physical, mechanical and barrier properties, protecting packed food * Corresponding author: Sandra N. Bulut, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




54

from adverse environmental factors, such as moisture, oxygen, microorganisms, as well as due to their easy processing, great possibilities of combining, and relatively low prices. However, due to their long service life, or non-degradability, as well as consumption of non-renewable raw materials (oil), science is more focused on finding an alternative for those materials. One of the ways to improve the ecological status of polymer packaging is to increase the use of natural biopolymers or biodegradable synthetic polymers (3). Bio-polymer materials obtained from renewable raw materials and agro-industrial waste can be considered as an alternative to conventional synthetic materials, due to their low price, possibility of obtaining from natural sources, and the possibility to solve the problem of waste disposal (4). However, the application of biopolymer materials has certain restrictions and disad-vantages, which include the rate and extent of degradation under various conditions, changes in the mechanical properties during storage, susceptibility to microbiological contamination, as well as the contamination of the packed content by individual migrants (1, 5). One of the major disadvantages of biopolymer films is aging, or changes of the film properties over time, so that their commercial application is still limited (6). Aging of films may be the result of physical or chemical reaction of the polymer matrix that occurs in the material after exposure to environmental conditions over a period of time (7). The most common changes that lead to aging of the films are: physical aging due to migration of plasticizer from the matrix (6-8) and chemical oxidation (5, 9-11). Chemical changes, such as oxidation, may cause degradation of the polymer chains, while physical changes, such as migration of low molecular weight additives, affect the film (7). Natural polymers are less stable compared to synthetic materials. The use of biopolymers as a material for food packaging is difficult to implement since the basic properties of edible films mainly depend on the environmental conditions, especially temperature and hu-midity (5, 12). In this paper, the bilayer biodegradable films based on pumpkin oil cake (PuOC) and zein, as well as pouches made from these materials, were prepared, and the changes of their mechanical (thicknesses, tensile strength and elongation at break), physicochemical (moisture content, total soluble matter and swelling), and barrier (gas transmission rate) properties, were analyzed during four weeks of storage at room conditions. Heat seal quality of formed pouches and composition of the gas atmosphere in the pouches were also monitored.

EXPERIMENTAL

Materials and Reagents

The grounded hull-less pumpkin (Cucurbita pepo L. c.v. ‘Olinka’) oil cake (PuOC) was collected from „Agrojapra“, Donji Agič, Bosnia and Herzegovina and stored at the temperature of 4 °C. Other reagents used in this study were: glycerol, zein, polyethylene glycol (PEG), ethanol, NaOH.



55

Preparation of bi-layer film The bilayer film based on pumpkin oil cake (PuOC) and zein, were prepared by “casting” method. First, PuOC suspension was prepared and cast onto Teflon-coated Petri dish. After PuOC film was dried, the zein solution was cast on the dried PuOC film, and left to dry at room conditions (23 ± 2 °C, 50 ± 5% RH). The film-forming suspension of PuOC (10%, w/w) in deionized water was produced by adding 30% glycerol (v/w, per weight of PuOC), and the film-forming solution of zein (10% w/v) was prepared by suspending zein in 85% ethanol with the addition of PEG 400 (50% w/w, per weight of zein); detailed procedure is described in Bulut et al., (13).

Preparation of bi-layer pouches When the bilayer films were dry, pouches were formed with zein layer inside, using laboratory vacuum sealer (Audion Elektro, Swissvac), the pouches were loaded with the modified atmosphere consisting of: 20% CO2 and 80% N2, and sealed.

Methods Thickness. Film thickness was measured using a micrometer (Digico 1, Tesa, Swiss Made, Renens, Switzerland), with a sensitivity of 0.001 mm. The results were expressed as the mean of ten measurements, from which an average was obtained along with the standard deviation (SD). Mechanical properties. Tensile strength (TS, MPa) and elongation at break (E, %) of the films were measured on an Instron Universal Testing Instrument Model No 4301 (Instron Engineering Corp., Canton, MA), according to the ASTM standard method D882-01 (14). The rectangular film strips (15x80 mm) were used. The initial grip separa-tion was set at 50 mm, and crosshead speed was set at 50 mm/min. The measurement of tensile properties was repeated at least three times, resulting in an average and SD. Physicochemical properties. Moisture content, swelling and total soluble matter of films were determined according to Hromiš et al., (15). At least three measurements were conducted and average values were calculated along with SD. Barrier properties. The determination of gas (CO2, N2, O2) transmission rate was conducted using isostatic gas-chromatographic method (DIN 53380 (16)), using a Lyssy GPM-200 apparatus with the belonging Gasukuro Kogyo GC-320 gas chromatograph and the HP 3396A integrator. Heat seal quality of formed pouches. Tensile strength (TS, MPa) of heat seal of formed pouches was measured on an Instron Universal Testing Instrument Model No 4301 (Instron Engineering Corp., Canton, MA). Composition of the gas atmosphere in pouches. The composition of protective atmosphere in pouches was determined using Oxybaby, Gasetechnik WITT GmbH & Co KG, United Kingdom. Films and pouches were stored at room conditions (23 ± 2 °C, 50 ± 5% RH) for four weeks, and their properties were analyzed once a week.



56


Properties of the obtained bilayer material during four weeks of storage Mechanical properties. The thickness of the obtained bilayer film based on PuOC/Zein was 300 ± 10 (μm), and no its changes were observed during the storage. Me-chanical properties, tensile strength (TS, MPa) and elongation at break (E, %), of the obtained bilayer films based on PuOC/Zein during four weeks, are presented in Figure 1. According to the obtained results, a slight increase of TS in the third week of storage was observed (Figure 1a). However, E decreased during the whole storage time. After one month, E of obtained film was 3 times lower than in the initial week (Figure 1b). The results show that the material based on PuOC/Zein became more brittle (lose elasticity) during storage at room conditions.

Figure 1. Mechanical properties of the obtained bilayer film based on PuOC/Zein (a) tensile strength (TS, MPa) and (b) elongation at break (E, %), during four weeks of

storage

Physicochemical properties The main problem in the application of biopolymer films based on protein and poly-saccharide is their hydrophilic nature, especially considering the needs of food packaging materials to be resistant to moisture. Sensitivity to humidity may lead to spontaneous, rapid and uncontrolled decomposition of the material (5, 17). Physicochemical properties of the obtained bilayer film based on PuOC/Zein, mois-ture content, total soluble matter and swelling, during four weeks of storage are presented in Figure 2. Slight decrease of the moisture content (Figure 2a) during the storage time was observed. The percentage of total soluble matter decreased after the first week and then remained stable to the rest of the storage time (Figure 2b). The percentage of swelling also decreased during the storage period, and after four weeks it was three times lower than in the initial week (Figure 2c). These results show that during the storage period at room conditions, the films based on PuOC/Zein probably lose moisture, leading to the formation of new bonds and crosslinking.

0

0,5

1

1,5

2

2,5

3

3,5

Ten

sile

str

ngt

h (M

Pa)

Week

zerofirstsecondthirdfourth

a)

0

5

10

15

20

25

30

35

40

Elo

nga

tio

at b

reak

(%

)

Week


b)



57

Figure 2. Physicochemical properties of the obtained bilayer film based on PuOC/Zein (a) moisture content, (b) total soluble matter and (c) swelling, during four weeks of

storage

Barrier properties The gas transmission rate of the obtained bilayer film based on PuOC/Zein, during four weeks of storage is presented in Figure 3. The results indicate that the obtained material present a very good barrier to oxygen in the initial week (27 ml/m2 24h 1 bar), but after one month of storage, the increase of gas transmission rate was observed (64 ml/m2 24h 1 bar). However, this material shows a moderate barrier to CO2 in initial week, 147 ml/m2 24h 1 bar, and CO2 gas transmission rate increased after forth weeks of storage to 188 ml/m2 24h 1 bar. Although, studies have shown that the presence of CO2 may have an important role in preventing microbial spoilage of food, a high permeability to CO2 of packaging materials can be in some cases an advantage, e.g. for packing products with high respiration rate.

0

5

10

15

20

25

30

Moi

stu

re c

onte

nt

(%)

Week


a)

0

5

10

15

20

25

30

35

40

45

Tot

al s

olu

ble

mat

ter

(%)

Week


b)

0

5

10

15

20

25

30

35

40

45

Sw

ellin

g (%

)

Week


c)



58

Figure 3. Gas transmission rate of the obtained bilayer films based on PuOC/Zein, during

four weeks of storage

Properties of the obtained bilayer pouches during four weeks of storage

Heat seal strength The results obtained for the seal strength of the pouches based on PuOC/Zein material during four weeks of storage are presented in Figure 4. As can be seen, the heat seal strength decreased up to three times during the storage period.

Figure 4. Heat seal strength of pouches based on PuOC/Zein materials during four weeks of storage

0

20

40

60

80

100

120

140

160

180

200

CO2 O2 N2 air

Gas

tra

nsm

issi

on r

ate

(ml/

(m2

24h

1ba

r))

zero week

fourth week

0

0,5

1

1,5

2

2,5

3

3,5

Hea

t se

al s

tren

gth

(N

/15

mm

)

Week




59

Composition of the gas atmosphere in the pouches

Composition of the gas atmosphere (CO2 and O2 contents), in the pouches based on PuOC/Zein material is presented in Figure 5. In the initial week, the content of CO2 was approximately 18% (Figure 5(a)), and of O2 approximately 2% (Figure 5(b)). Through the storage period, the percentage of O2 increased up to 7 times, but the percentage of CO2 decreased up to 18 times already after one week, and then remained stable through the storage period. These results are probably due to a smaller extent, to gas transmission through the material, especially for CO2, and to a greater extent, they are due to the low heat seal strength, which decreased through storage period and, probably, to the influence of the gas contents in the pouches.

Figure 5. Composition of the gas atmosphere in the pouches based on PuOC/Zein material - (a) CO2 content and (b) O2 content, during four weeks of storage

CONCLUSION

The results obtained in this study show that the bilayer film based on PuOC/Zein became brittle during storage at room conditions. During the storage period, the PuOC/Zein film was losing moisture, which influenced its physicochemical properties, leading to a decrease in the total soluble matter and swelling. According to the results obtained for gas transmission rate, the packaging material based on PuOC/Zein, showed a very good barrier to O2, but moderate to CO2. These findings, as well as the decrease of the heat seal strength during four weeks, influenced the composition of gas atmosphere, (percentages of CO2 and O2,) in the PuOC/Zein pouches, and led to an increase of the O2 percentage and a decrease of the CO2 percentage during the storage period. The tested materials show promising properties as a packaging material, and can form pouches for food packaging. However, the influence of storage time on the properties of the tested material and pouches was observed, and further research should be focused on possible modifications in order to avoid or reduce this influence.

0

2

4

6

8

10

12

14

16

18

20

22

CO

2

Week

zero

first

second

third

fourth

a)

0

2

4

6

8

10

12

14

16

18

20

22

O2

Week


b)



60

Acknowledgеment This work was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia, Project Number 46010.

REFERENCES 1. Coles, R.; McDowell, D.; Kirwan, M. Food packaging technology; Blackwell pub-

lishing ltd: Oxford, UK, 2003. 2. Lazić, V.; Novaković, D. Ambalaža i životna sredina; Faculty of Technology, Univer-

sity of Novi Sad: Novi Sad, Serbia, 2010. 3. Plackett, D. Biopolymers - New materials for sustainable films and coatings; John

Wiley and Sons, Ltd: 2011. 4. Popović, S. Istraživanje dobijanja i karakterizacija biorazgradivih kompozitnih fil-

mova na bazi biljnih proteina. Ph.D. Thesis, University of Novi Sad, Serbia, 2013. 5. Lazić, V.; Popović, S. Biorazgradivi ambalažni materijali; Faculty of Technology,

University of Novi Sad: Novi Sad, Serbia, 2015. 6. Olabarrieta, I.; Cho, SW.; Gällstedt, M.; Sarasua, JR.; Johansson, E.; Hedenqvist,

M.S. Aging properties of films of plasticized vital wheat gluten cast from acidic and basic solutions. Biomacromolecules, 2006, 7, 1657-1664.

7. Anker, M.; Stading, M.; Hermansson, AM. Aging of whey protein films and the effect on mechanical and barrier properties. Journal of Agricultural and Food Chemistry, 2001, 49, 989-995.

8. Hernández-Muñoz, P.; López-Rubio, A.; Del-Valle, V.; Alemnar, E.; Gavara, R. Me-chanical and water barrier properties of glutenin films influenced by storage time. Journal of Agricultural and Food Chemistry, 2004, 52, 79-83.

9. Morel, M.H.; Bonicel, J.; Micard, V.; Guilbert, S. Protein insolubilization and thiol oxidation in sulfite-treated wheat gluten films during aging at various temperatures and relative humidities. Journal of Agricultural and Food Chemistry, 2000, 48, 186-192.

10. Micard, V.; Belamri, R.; Morel, M.H.; Guilbert, S. Properties of chemically and phy-sically treated wheat gluten films. Journal of Agricultural and Food Chemistry, 2000, 48, 2948-2953.

11. Popović, S.; Peričin, D.; Vaštag, Ž.; Popović, Lj.; Lazić, V. Evaluation of edible film-forming ability of pumpkin oil cake; effect of pH and temperature. Food Hydro-colloids, 2011, 25 (3), 470-476.

12. Somanathan, N.; Naresh, M. D.; Arumugam, V.; Ranga-nathan, T. S.; Sanjeevi, R. Mechanical Properties of Alkali Treated Casein Films. Polymer Journal., 1992, 24, 603-611.

13. Bulut, S.; Lazić, V.; Popović, S.; Šuput, D.; Hromiš, N.; Popović, Lj. Possibility to maintain modified atmosphere in pouches made from biopolymer materials, Pro-ceedings of III International Congress “Food Technology, Quality and Safety” and



61

XVII International Symposium “Feed Technology” (FoodTech2016), Novi Sad, 25-27. October 2016, Book of Abstract, pp 122-127.

14. ASTM. (2001). Standard test method for tensile properties of thin plastic sheeting. Annual book of ASTM standards. Designation D882-01. Philadelphia: ASTM, Ame-rican Society for Testing Materials.

15. Hromiš, N.; Lazić, V.; Markov, S.; Vaštag, Ž.; Popović, S.; Šuput, D.; Džinić, N.; Ve-lićanski, A.; Popović, Lj. Optimization of chitosan biofilm properties by addition of caraway essential oil and beeswax. Journal of Food Engineering, 2015, 158, 86-93.

16. DIN 53380. (1969). Prüfung Kunstoffolien, Elastomerfolien, Bestimmung der Gas-durchlässigkeit.

17. Olabarrieta, I. Strategies to improve the aging, barrier and mechanical properties of chitosan, whey and wheat gluten protein films. Ph.D. Thesis. University of Stock-holm, Sweden, 2005.

УТИЦАЈ ПЕРИОДА СКЛАДИШТЕЊА НА ОСОБИНЕ БИОПОЛИМЕРНИХ

АМБАЛАЖНИХ МАТЕРИЈАЛА И КЕСИЦА

Сандра Н. Булут, Вера Л. Лазић, Сенка З. Поповић, Невена М. Хромиш, Данијела З. Шупут

Универзитет у Новом Саду, Технолошки факултет Нови Сад,

Булевар Цара Лазара 1, 21000 Нови Сад, Србија

Биополимери представљају самоуништиве биоразградиве амбалажне материја-ле, који потичу из обновљивих извора. Иако ови материјали имају велики еколош-ки значај, промене особина током времена, представља њихов недостатак. Најчеш-ће промене које доводе до старења филмова су: физичко старење услед миграције пластификатора и матрикса и хемијско старење оксидацијом. У раду је синтетисан двослојни биоразградиви филм на бази погаче уљане тикве голице (PuOC) и зеина, и од овог материјала формиране су кесице. Након синтезе су испитиване промене механичких, физичко-хемијских и баријерних особина овог материјала, током че-тири недеље складиштења на собним условима. Јачина вара и састав гаса у кесица-ма током складиштења, су такође праћене. Добијени резултати су показали да дво-слојни филм има дебљину 300 ± 10 (μm), а промена дебљине током складиштења није запажена. У трећој недељи складиштења, дошло је до порасти затезне јачине испитиваног филма, док се издужење при кидању смањивало током читавог перио-да складиштења. Смањење садржаја влаге, растворљивости и бубрења добијених филмова, је такође запажено. Након недељу дана складиштења, дошло је до пове-ћања степена преноса O2 кроз испитивани материјал са 27 на 64 (ml/m2 24h 1 bar), као и до повећања степена преноса CO2 са 147 на 188 (ml/m2 24h 1 bar). Јачина вара PuOC/Zein кесица се смањивала током целог периода складиштења. Проценат O2 у PuOC/Zein кесицама се повећао и до 7 пута у току складиштења, док се проценат CO2 смањио 18 пута већ након недељу дана, и потом остао стабилан током скла-диштења. Ови резултати су, вероватно, у малој мери последица преноса гаса кроз



62

материјал, посебно за CO2, и у још значајној мери, услед слабе јачине вара која опада у току целог периода складиштења, и вероватно утиче на садржај гасова у кесицама. Кључне речи: биополимерни филмови, погача уљане тикве голице, зеин, особине,

период складиштења

Received: 04 May 2017. Accepted: 21 September 2017

APTEFF, 48, 1-323 (2017) UDC: 634.71:547.562:615.27/.28 https://doi.org/10.2298/APT1748063C BIBLID: 1450-7188 (2017) 48, 63-76


63

BIOACTIVITY OF BLACKBERRY (Rubus fruticosus L.) POMACE: POLYPHENOL CONTENT, RADICAL SCAVENGING, ANTIMICROBIAL AND

ANTITUMOR ACTIVITY

Dragana D. Četojević-Simin1*, Aleksandra S. Ranitović2, Dragoljub D. Cvetković2, Siniša L. Markov2, Milica N. Vinčić2, Sonja M. Djilas2

1 University of Novi Sad, Faculty of Medicine, Oncology Institute of Vojvodina,

Dr Goldmana 4, 21204 Sremska Kamenica, Serbia 2 University of Novi Sad, Faculty of Technology Novi Sad, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia

A large volume of industrial waste of biological origin is produced annually world-wide, causing a serious disposal problem even though it is a huge source of beneficial compounds which may be used for their high nutritional and bioactive properties. The bio-potential of blackberry pomace (waste) obtained after juice separation from the Ča-čanska bestrna and Thornfree cultivars was evaluated. Higher amounts of total and mo-nomeric anthocyanins, total phenolics, and total flavonoids were found in Thornfree po-mace extract and demonstrated stronger 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity (2.12 mmol TEAC g-1) than Čačanska bestrna pomace extract (1.03 mmol TEAC g-1). Both extracts highly increased apoptosis/necrosis ratios in all investi-gated cell lines. The highest cell growth inhibition effects (EC50=52.5 - 64.7 μg mL-1) and the highest increase of apoptosis (AI=12.2) were obtained in breast adenocarcinoma cell line. Both blackberry pomace extracts inhibited the growth of all tested microorganisms. In the reference and wild bacterial strains MIC and MBC/MFC were achieved in the 0.39-25 mg mL-1 and 0.78-25 mg mL-1 range, respectively. Blackberry cultivar pomaces are rich source of phytochemicals with significant health promoting properties that could be further utilized as beneficial food ingredients. KEY WORDS: blackberry pomace; DPPH radical scavenging; cell growth; bactericidal

activity; fungicidal activity

INTRODUCTION During the production of juices, concentrates, jams, and jellies, the generated berry fruit pomace consists of the pulp, peel and the seeds (1). Huge volumes of waste are produced in the food industry after fruit processing worldwide, causing a serious disposal problem (2). Berry fruit pomace contains high levels of essential minerals, vitamins C

* Corresponding author: Dragana Četojević-Simin, University of Novi Sad, Faculty of Medicine, Oncology Intitute of

Vojvodina, Dr Goldmana 4, 21204 Sremska Kamenica, Serbia, e-mail: [email protected]



64

and E, fatty acids, fibers and polyphenols such as anthocyanins, phenolic acids, flavanols, and tannins (1, 3), which could be further utilized due to their proven health benefits. Blackberry pomace can even be used after extended frozen storage (1). Rubus fruticosus (Linnaeus) is a perennial shrub famous for its fruit which is traded globally due to its delicious taste. The leaves are used in traditional phytotherapy in Mediterranean countries for the treatment of wounds, sores, scratches, gum inflamma-tions, ulcers and sore throat (4-6). Berry fruits have a high content of bioactive com-pounds that possess antioxidant, anticarcinogenic, vasodilatory and antimicrobial activity (7, 8). Compared to other berries, blackberries contain high levels of ellagitannins and abundance of healthy antioxidants such as flavonoids, salicylic acid, ellagic acid, and fiber that have also been recognized for their antitumor effects (9-11). Beside their nutritional value, fruits, vegetables, herbs, and spices have been traditio-nally used to cure various human ailments (6). Increased consumption of fruits and vege-tables is recommended in dietary guidelines worldwide for their beneficial health effects (9, 12, 13). Further, broad use of antibiotics to cure microbial infections has made bac-teria and fungi resistant to common commercial antibiotics (14). Natural products as anti-microbials are a promising alternative to control foodborne bacterial infections. Previous results concerning the antibacterial activity of berry fruits showed that raspberry fruit extract inhibited growth of Salmonella typhimurium, Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, Bacillus cereus, Staphylococcus saprophyticus and Listeria monocytogenes (15, 16). On the other hand, blackberry juice stimulated growth of beneficial Lactobacillus strains (10). Apart from the inhibition of bacteria growth, blackberry extract inhibited the early stages of herpes simplex virus type 1 (HSV-1 repli-cation and had potent virucidal activity (17). In this study, the juice processing by-product, the pomace extracts from two black-berry cultivars, Čačanska bestrna and Thornfree were used to determine the contents of total phenolics, total flavonoids, and total and monomeric anthocyanins. Radical scaven-ging activity was determined using DPPH (2,2-diphenyl-1-picrylhydrazyl) assay. Anti-tumor in vitro activity, as well as cell death, were also evaluated. Antimicrobial activity was evaluated in a panel of reference and wild strains.

MATERIAL AND METHODS

Pomace extraction and sample preparation The blackberry cultivars Čačanska bestrna and Thornfree were obtained from ″Alfa RS″, Lipolist, Serbia and were stored at -20 °C prior to analysis. The blackberry pomace was prepared by pressing the unfrozen fruits through a cheesecloth. The yields of the Čačanska bestrna and Thornfree pomaces were 228 g kg-1 and 276.4 g kg-1, respectively. They pomaces were extracted with 80% methanol aqueous solution containing 0.05% acetic acid, and lyophilized according to a previously reported procedure (16). The yields of the lyophilized blackberry pomaces from Čačanska bestrna and Thornfree cultivars were 86.5 g kg-1 and 90.3 g kg-1, respectively.



65

The lyophilized extracts were re-dissolved in sterile distilled water to obtain 10 mg mL-1 stock solution for the evaluation of antioxidant and antimicrobial activity, or in DMSO (dimethyl sulphoxide) to obtain 500 mg mL-1 stock solution for the evaluation of cytotoxic activity and cell death. The extracts were investigated in the concentration ranges from 0.0017-0.0833, 0.005-2.5 and 0.195-25 mg mL-1 in the radical scavenging, cell growth and antimicrobial assay, respectively.

Polyphenol antioxidants content and DPPH radical scavenging activity Total phenolics content (TPh) in blackberry pomace extracts was determined spectro-photometrically according to the Folin-Ciocalteu method (18), by calibrating against gal-lic acid and expressing the results as gallic acid equivalents (GAE) in g kg-1 of dried ex-tracts and fresh pomace. Total flavonoids (TFd) content was measured by the aluminum chloride spectrophotometric assay (19), determined from the regression equation of the rutin calibration curve, and expressed as rutin equivalents (RE) in g kg-1 of dried extracts and fresh pomace. The total and monomeric anthocyanins content (TAc, MAc) was esti-mated spectrophotometrically using the pH single and differential method according to Lee et al. (20). Anthocyanins were quantified as cyanidin-3-glucoside equivalents (CyGE) using an extinction coefficient of 26,900 in L mol-1cm-1, and resulting values were expressed in terms of cyanidin-3-glucoside equvalents in g kg-1 of dried extracts and fresh pomace. The DPPH radical scavenging activity (SA) of blackberry pomace extracts was deter-mined according to a previously described procedure (16).

Cell lines, cell growth activity and cell death detection

Cell growth activity was evaluated in vitro in human cell lines: HeLa (cervix epithe-loid carcinoma), MCF7 (breast adenocarcinoma), HT-29 (colon adenocarcinoma) and MRC-5 (fetal lung). Cell growth activity was determined after 48 h exposition time according to the pre-viously described procedure (16), using colorimetric sulforhodamine B (SRB) assay (21).

Apoptosis and necrosis were detected using the Cell Death Detection ELISAPLUS kit, Roche (Version 11.0). The cell death detection experiments were performed in all cell lines and for the most active extracts obtained in the cell growth experiments (16). De-pending on the extract and the cell line used, the concentrations were in the 420-700 g mL-1 range (for the 2 h exposition time).

Test microorganisms and antimicrobial assays The antimicrobial activity was evaluated using the reference and wild strains (w) of Gram-positive and Gram-negative bacteria (16) and yeasts (Saccharomyces cerevisiae 112 Hefebank Weihenstephan and Candida albicans ATCC 10231). The bacterial and yeast strains were stored at -80 °C in Nutrient broth supplemented with 20% (v/v) glyce-



66

rol or Sabouraud Maltose broth (both Himedia, Mumbai, India) with 10% (v/v) glycerol, respectively. Screening of the antimicrobial activity was performed by disk diffusion method (22). The minimal inhibitory (MIC), minimal bactericidal (MBC) and Minimal fungicidal (MFC) concentrations were determined by microdilution method as previously described (16). The yeasts were grown at 25 °C on Sabouraud maltose agar (Himedia, Mumbai, India). After 48 h incubation, the colonies were enumerated by viable count. The antibiotic (30 µg cefotaxime and 10 µg clavulanic acid per disc, Bioanalyse®, Ankara, Turkey) and antifungal agent (cycloheximide solution, 0.03 g mL-1, Sigma-Al-drich, Co. St. Louis, USA) were used as controls.

Statistical analysis The results of polyphenol antioxidants content and radical scavenging activity were expressed as mean ± SD of three experiments (n=3). The results of cell growth activity were obtained in two independent experiments, each performed in quadruplicate (n = 8). In the cell death experiments the results for treatments and negative control were obtai-ned from the pooled quadruplicates (n=4) that were evaluated in duplicate (n=2) accor-ding to manufacturer’s recommendation. The antimicrobial assays were performed in du-plicate, i.e. with two tested bacterial or yeast suspensions prepared for each strain, in three repetitions (n=6). A comparison of the group means and the significance between the groups were veri-fied by one-way ANOVA using OriginPro 8 SRO (OriginLab Corporation, Northampton, USA). Statistical significance was set at p < 0.05, unless noted otherwise.


Polyphenol antioxidants of blackberry pomace As mentioned above, two blackberry cultivars (Čačanska bestrna and Thornfree), which are usually consumed as fresh and processed fruits in Serbian diet, were used to prepare pomace. As a by-product in juice processing, the blackberry pomace consists of the skin, seeds and part of pulp and contains significant amount of polyphenol antioxi-dants especially anthocyanins (23). The screening of the polyphenolic antioxidants of blackberry pomace extracts - the total phenolics (TPh), total flavonoids (TFd), total and monomeric anthocyanins (TAc and MAc) content was carried out using spectrophoto-metric assays (Table 1). The obtained results show that the pomace of both blackberry cultivars are a rich source of antioxidant compounds. Significantly higher amounts of TAc and MAc (p<0.001), TPh (p<0.01) and TFd (p<0.001) were found in the Thornfree pomace extract. This is in accordance with the results of Da Fonseca Machado et al. (24), who reported that the content of total phenolics and monomeric anthocyanins were 7.36 g GAE kg-1 and 1.02 g CyGE kg-1 fresh residue of blackberries. Also, Laroze et al. (25) established that high amounts of total polyphenols were recovered from pressing pomace of black-



67

berry (270 g GAE kg-1 extract), blueberry (172 g GAE kg-1 extract) and cranberry (138 g GAE kg-1 extract) (27). Compared to raspberry pomace extracts (16) the blackberry po-mace extracts (Table 1) showed higher contents of total phenolics (26-44 vs. 75-88 g GAE kg-1), total flavonoids (22-25 vs. 40-45 g RE kg-1), and total anthocyanins (2-4 vs. 7-12 g CyGE kg-1). Table 1. Contents of total phenolics, total flavonoids, total and monomeric anthocyanins

(TAc and MAc) in pomace extracts (PE) and fresh pomace (FP) of the blackberry cultivars

Means in the same column differ significantly at level * p < 0.05, ** p < 0.01 or *** p < 0.001

DPPH radical scavenging activity

In the present study, a method based on the DPPH radical scavenging was applied to assess the in vitro antioxidative activity of blackberry pomace extracts. Lower concen-trations (0.0017-0.0167 mg mL-1) of the Thornfree blackberry pomace extract exhibited significantly high scavenging activity (p<0.001) (Figure 1).

*SA values are means of three replicate analysis ± SD. Means for the same concentration of blackberry pomace

extracts applied differ significantly at the level * p < 0.05 or *** p < 0.001

Figure 1. DPPH radical scavenging activity* of blackberry pomace extracts

Blackberry cultivar

TPh (g GAE kg-1) TFd (g RE kg-1) TAc (g CyGE kg-1) MAc (g CyGE kg-1)

PE FP PE FP PE FP PE FP

Thornfree 88.28±3.48** 7.97±0.31** 45.51±2.16* 4.11±0.20* 12.61±0.48*** 1.14±0.04*** 11.49±0.48*** 1.04±0.04***

Čačanska bestrna 75.50±3.25** 6.53±0.28** 39.89±1.73* 3.45±0.15* 6.81±0.32*** 0.59±0.03*** 5.90±0.27*** 0.51±0.02***



68

Based on DPPH radical scavenging activity, TEAC (trolox equivalent antioxidant capacity) value of the pomace extract from both blackberry cultivars was determined (Table 2). The Thornfree extract expressed significantly higher (p<0.001) TEAC value than that of the Čačanska bestrna.

Table 2. Antioxidant activity of pomace extracts and fresh pomace of the blackberry cultivars

Blackberry cultivar

DPPH

PE (mmol TEAC g-1) FP (μmol TEAC g-1)

Thornfree 2.12 ± 0.07*** 191.71 ± 6.57***

Čačanska bestrna 1.03 ± 0.03*** 89.19 ± 3.01*** Means in the same column differ significantly at the level ***p < 0.001

It has been shown that berry fruits are rich in anthocyanins and other antioxidants, and have high antioxidant activity. Among berry fruits, blackberry stands out by exhibiting the highest antioxidant activity and the highest levels of phenols, flavonoids, anthocya-nins and carotenoids (26, 27). Also, many studies reported the fruit pomace contained abundant phenolic compounds which indicated that this by-product might be useful as a raw material for creating new value-added food, pharmaceuticals, and cosmetic formula-tions. According to the DPPH assay results, the apple pomace presented good antioxidant activity (6.36 g TE kg-1 dry weight) (28). Zhou et al. (29) reported that the antioxidant activity of the pomaces from five cultivars of bayberries (Myrica rubra) contained high amounts of phenolic compounds, being in the range 65.7-91.0 g TEAC kg-1 dry weight. In the study of Galván D’Alessandro et al. (30) high amounts of total polyphenols (>70 g GAE kg-1 dry basis), and especially of anthocyanins (>13 g CyGE kg-1 dry basis) were recovered from Aronia sp. waste, and the obtained extracts exhibited very high antioxi-dant capacity (DPPH, >450 μmol TEAC g-1 dry basis). Da Fonseca Machado et al. (24) analyzed the blackberry juice industrial residues and found that the antioxidant activity, measured by DPPH method, was in the range 29.04-76.03 μmol TE g-1 fresh residue.

In vitro antitumor activity The most pronounced cell growth effects, lowest EC50 values, and the most favorable non-tumor/tumor ratios were observed in the breast adenocarcinoma cell line (MCF7). In the breast cell line the EC50 values reached 52.5 and 64.7 μg mL-1 for Čačanska bestrna and Thornfree, respectively (Table 3). The EC50 values were slightly higher indicating lower activity in the cervix and colon tumor cells, as well as in the cells derived from the healthy tissue (in the range of 66-87.3 μg mL-1). The Čačanska bestrna extract was slight-ly more active than Thornfree (p0.05) in all cell lines (Table 3).



69

Table 3. EC50 values* (μg mL-1) of blackberry pomace extracts obtained after 48 h in human cell lines

Extract HeLa NT/T Hela MCF7 NT/T MCF7 HT-29 NT/T HT-29 MRC-5

Thornfree 80.14±8.44 b,d 1.09 64.71±11.66 b,c 1.35 85.997.78 b,e 1.02 87.28±8.35 b,e

Čačanska bestrna 69.31±7.77 a,d 1.03 52.48±10.59 a,c 1.37 66.029.60 a,d 1.09 71.70±8.74 a,d

*Values represent mean ± SD of eight measurements (n=8). NT/T non-tumor/tumor EC50 ratio Means with different letters within each column (a-b) and within each row (c-e) differ significantly (p<0.05)

Compared to raspberry pomace extracts investigated for the same activity (16), the blackberry pomace extracts from Čačanska bestrna and Thornfree cultivars showed hig-her activity (p0.01) towards colon carcinoma, but also for the cell line derived from healthy tissue, while the activity towards breast and cervix carcinoma was comparable or slightly lower (p0.05). The non-tumor/tumor EC50 ratios that were obtained for both blackberry pomace extracts in the breast (NT/T = 1.3) and colon (NT/T=1.1) adenocarci-noma cells (Table 3) were more favorable than for raspberry (NT/T=0.8) pomace extracts (16). The blackberry press residues from the Čačanska bestrna and Thornfree cultivars are rich source of anthocyanins (especially cyanidin 3-glucoside), and possess good antioxi-dative activity, reducing power and - glucosidase inhibition potential that correlated highly with the contents of anthocyanin compounds (23). Anthocyanins have been re-ported to induce phase II enzimes which may inactivate carcinogens activated by phase I enzymes, and inhibit possible DNA damage by the carcinogens (31, 32). Blackberry anthocyanins are thought to suppress cancer cell growth by modifying cell signaling pathways and expression of activated protein 1 (AP-1) and nuclear factor B (NFB) that regulate cell proliferation and cell cycle control (32). The inhibition of the cancer cell proliferation by berry juices did not involve caspase-dependent apoptosis, but appeared to involve cell cycle arrest that was demonstrated by down-regulation of the expression of cdk4, cdk6, cyclin D1 and cyclin D3 (33).

Apoptosis and necrosis In the MCF7 and HeLa cell lines both extracts increased apoptosis (EFA) and decrea-sed necrosis (EFN), while in the HT-29 and MRC-5 cell lines they decreased both pa-rameters (Table 4). The overall increase in the A/N ratio compared to the control was the most pronounced for the Thornfree extract in MCF7 (A/N 31), and was favorable for both extracts in all other cell lines (A/N 1) (Table 4). Compared to raspberry pomace extracts for the same activity (AI=1.0-3.1) (16), the blackberry pomace extracts Čačanska bestrna and Thornfree showed non-selective action towards all investigated cell lines and significantly higher (p<0.01) apoptotic increases (AI=1.4-12.2). The very high apoptotic increase was obtained by Thornfree pomace extract in the breast carcinoma cell line (A/N=32 and АI=12.2), and it was two-fold higher than in the cell line derived from healthy tissue (T/NT=1.82; Table 4).



70

Table 4. Apoptosis and necrosis obtained after 2 h in the human cell lines

Thornfree Čačanska bestrna

DMSO

MCF7

EFA 4.56b 1.23a -

EFN 0.37a 0.89a -

A/N 31.82c 3.65b 2.61a

AI 12.19b 1.40a -

NT/T 1.82b 0.25a -

HT-29

EFA 0.75a 0.80a -

EFN 0.11a 0.12a -

A/N 3.91b 3.96b 0.58a

AI 6.74a 6.83a -

NT/T 1.01a 1.26a -

HeLa

EFA 1.90a 2.28a -

EFN 0.36a 0.34a -

A/N 1.53b 1.95b 0.29a

AI 5.28a 6.72b -

NT/T 0.79a 1.24a -

MRC-5

EFA 0.72a 0.65a -

EFN 0.11a 0.17a -

A/N 1.47b 1.19b 0.22a

AI 6.68b 5.41a - Means with different letters (a-c) within each row differ significantly (p<0.05). EFA - enrichment factor for apoptosis, EFN - enrichment factor for necrosis, A/N - apoptosis/necrosis ratio, AI - apoptotic increase, NT/T - non-tumor / tumor apoptotic increase, DMSO - Dimethyl sulfoxide The blackberry pomace extracts were significantly more active towards colon adenocarcinoma cell line (IC50= 70-90 µg mL-1; Table 3) and showed significantly higher apoptotic increases (up to AI=12; Table 4) compared to raspberry pomaces (IC50=160-190 µg mL-1 and AI=3) (16) in all investigated cell lines. Although the cell growth experiments showed that the Čačanska bestrna pomace extract exhibited a slightly higher activity towards all cells lines, it should be pointed out that higher non-tumor/tumor ratios (NT/T; Table 3) and more favorable proapoptotic properties in comparison with the activity towards cells derived from the healthy tissue (T/NT ratio; Table 4) were obtained for the Thornfree extract in breast adenocarcinoma cells.



71

Antimicrobial activity The results of antimicrobial activity obtained using disk diffusion method (Table 5) show that the blackberry pomace extracts from Čačanska bestrna and Thornfree cultivars inhibited the growth of all tested bacteria, but did not show activity against tested yeasts. The blackberry extracts displayed comparable antibacterial potential against majority of tested bacteria except against E. coli ATCC 10536 and P. aeruginosaw, which were more susceptible toward Čačanska bestrna than Thornfree (inhibition zones were significantly different; p<0.05). Among bacterial strains the most resistant were wild strains of Salmo-nella sp., S. saprophyticus and Bacillus sp. The zones with reduced growth appeared around the disks for these strains, indicating bacteriostatic activity of the extracts. The inhibition zones of positive controls (antibiotics and antifungal agents) were two to three times higher than the zones of pomace extracts (Table 5). The values for MIC, MBC and MFC ranged from 0.39 to >25 mg mL-1, from 0.78 to >25 mg/mL, and from 9.38 to >25 mg mL-1, respectively (Table 5). The strains of Sta-phylococcus and P. aeruginosaw were the most susceptible, whereas C. albicans was the most resistant to the Čačanska bestrna extract. The MIC and MBC values for the Čačan-ska bestrna cultivar were lower than for Thornfree in all tested strains except for E. coli, referring to its higher antimicrobial potential. E. coli was the most susceptible, while yeasts and Salmonella strains were the most resistant to the Thornfree extract (MBC was ≥12.5 mg mL-1). The most resistant strain to the activity of both extracts were Bacillus sp., for which either high growth inhibition concentrations were needed (25 mg mL-1) or MBC values were not reached in the tested concentration range (Table 5). It is worth to emphasize that both blackberry pomace extracts inhibited growth of the strains of Sal-monella and L. monocytogenes, which are major causative agents for over half of the foodborne illnesses. Control of these foodborne enteric pathogens is a challenge for pub-lic health agencies and food industry (10, 34). Although there are several studies referring to antimicrobial potential of blackberry fruit, leaves and stem extracts (5, 10, 14), only a few studies of blackberry pomace extracts have been published to date. In those studies blackberry pomace extract showed moderate activity only against Serratia marcenscens and Bacillus strains (35) or did not show any antibacterial activity (36). The antibacterial potential of raspberry pomace ex-tracts from two cultivars grown in Serbia (MBC were in the range 0.29-0.78 mg mL-1) was demonstrated by Četojević-Simin et al. (16), and exhibited higher potential com-pared to blackberry pomaces against the same panel of bacterial strains. Different mechanisms and synergistic or additive effects between organic acids, phenolic acids, tannins, and other components of the berry extracts are proposed to be responsible for their antibacterial activity (34, 37). The hydrophobicity of the components - polymeric phenolics or organic acids (citric, benzoic, malic, etc.) enables them to stack and accumulate in the microbial cell membrane. Organic acids, in their undissociate state can diffuse through the cell membrane, where they dissociate. That causes the increase of permeability and weakening of microbial membrane integrity, acidification of the cyto-plasm, inhibition of the enzymes, damage of structural proteins and DNA, and leakage of the intracellular constituents, leading to cell death (34, 38, 39). Another potential mecha-



72

nism is the dissociation of phenolics at the plasma membrane, which leads to hyper-acidification at the plasma membrane interface and inhibition of ATP syntesis (34).

Table 5. Antimicrobial activity of the blackberry pomace extracts

* Means in the same row differ significantly at level p<0.05 (disk diffusion method); a Mean of diameter of the inhibition zone (mm; including disc, 6 mm) with standard deviation in parentheses; b Since all replicates for MIC /MBC (MFC) had equal values, standard deviations were not presented; c Zone of reduced growth; nd - inhibition zone was not detected; w - wild strain

CONCLUSION

The obtained results indicate that the blackberry pomace is a rich source of polyphe-nol antioxidants possessing high antioxidant activity. Higher amounts of total and mono-meric anthocyanins, total phenolics and total flavonoids were found in the Thornfree pomace extract. The highest non-tumor/tumor ratios and more favorable proapoptotic properties were obtained for the Thornfree extract towards breast adenocarcinoma cells. Both blackberry pomace extracts inhibited the growth of all tested microorganisms. The blackberry processing by-products of the Čačanska bestrna and Thornfree cultivars are a rich source of phytochemicals with significant health promoting properties that could be used as beneficial food ingredients, and make impact on the consumer confidence on the safety of the products and profitability of the food processing industry.

Acknowledgement

This study was funded by the Ministry of Education, Science and Technological De-velopment of the Republic of Serbia (grant number TR 31044).

Group Tested strain Disk diffusion methoda MIC/MBC(MFC)

(mg mL-1)b Controls

(antibiotic/ antifungal

agent)a Čačanska bestrna

Thornfree Čačanska bestrna

Thornfree

G (–) bacteria

Escherichia coli 17* (0.82) 15* (1.41) 0.78/1.56 0.78/1.56 30 (1.0)

Escherichia coliw 13.5 (3.11) 13.5 (1.29) 0.78/1.56 3.125/6.25 25.33 (0.58)

Salmonella typhymurium 16.75 (3.2) 14.25 (1.71) 1.17/1.56 9.38/12.5 33.6 (0.85)

Salmonella sp.w 19.25 (3.3)c 15 (2.45)c 1.17/1.56 6.25/12.5 27.83 (0.76)

Pseudomonas aeruginosa 14.75 (1.26) 13 (0.82) 0.39/0.78 4.69/6.25 20.17 (0.76)

Pseudomonas aeruginosaw 16* (0.82) 14* (0.82) 0.78/1.56 1.56/3.125 15.33 (0.58)

G (+) bacteria

Staphylococcus aureus 11 (3.56) 10.25 (3.3) 0.39/0.78 1.56/3.125 34.3 (0.75)

Staphylococcus saprophyticusw 16.75 (0.5)c 12 (2.94)c 0.59/0.78 3.125/6.25 24.0 (0.5)

Bacillus cereus 19.5 (3.7) 16.75 (0.96) 25/>25 18.75/25 34.5 (1.5)

Bacillus sp.w 17.5 (2.89)c 14.25 (0.96)c 18.75/25 >25 39.33 (0.58)

Listeria monocytogenesw 15.25 (1.71) 13.5 (1.29) 0.78/1.56 4.69/6.25 14.25 (0.55)

Yeast Saccharomyces cerevisiae nd nd 4.69/6.25 9.38/12.5 >38

Candida albicans nd nd >25 >25 15.33 ( 0.58)c



73

REFERENCES

1. Radočaj, O.; Vujasinović, V.; Dimić, E.; Basić Z. Blackberry (Rubus fruticosus L.) and raspberry (Rubus idaeus L.) seed oils extracted from dried press pomace after longterm frozen storage of berries can be used as functional food ingredients. Eur. J. Lipid. Sci. Technol. 2014, 116, 1015-1024.

2. Schieber, A.; Stintzing, F.C.; Carle, R. By-products of plant food processing as a source of functional compounds - recent developments. Trends Food Sci. Technol. 2001, 12, 401-413.

3. Szajdek, A.; Borowska, E.J. Bioactive compounds and health-promoting properties of berry fruits: A review. Plant Foods Human Nutr. 2008, 63, 147-156.

4. Martini, S.; D’Addario, C.; Colacevich, A.; Focardi, S.; Borghini, F.; Santucci, A. et al. Antimicrobial activity against Helicobacter pylori strains and antioxidant proper-ties of blackberry leaves (Rubus ulmifolius) and isolated compounds. Int. J. Anti-microb. Agents 2009, 34, 50-59.

5. Radovanović, B.C.; Milenković Anđelković, A.S.; Radovanović, A.B.; Anđelković, M.Z. Antioxidant and Antimicrobial Activity of Polyphenol Extracts from Wild Berry Fruits Grown in Southeast Serbia. Trop. J. Pharm. Res. 2013, 12, 813-819.

6. Zia-Ul-Haq, M.; Riaz, M.; De Feo, V.; Jaafar, H.Z.E.; Moga, M. Rubus Fruticosus L.: Constituents, Biological Activities and Health Related Uses. Molecules 2014, 19, 10998-11029.

7. Jimenez Garcia, S.N.; Guevara-Gonzalez, R.G.; Miranda-Lopez, R.; Feregrino-Perez, A.A.; Torres-Pacheco, I.; Vazquez-Cruz, M.A. Functional properties and quality cha-racteristics of bioactive compounds in berries: Biochemistry, biotechnology, and ge-nomics. Food Res. Int. 2013, 54, 1995-1207.

8. Bobinaite, R.; Viškelis, P.; RimantasVenskutonis. P.R. Variation of total phenolics, anthocyanins, ellagic acid and radical scavenging capacity in various raspberry (Ru-bus spp.) cultivars. Food Chem. 2012, 132, 1495-1501.

9. Kaume, L.; Howard, L.R.; Devareddy, L. The Blackberry fruit: A Review on Its Composition and Chemistry, Metabolism and Bioavailability, and Health Benefits. J. Agric. Food Chem. 2012, 60, 5716-5727.

10. Yang, H.; Hewes, D.; Salaheen, S.; Federman, C.; Biswas, D. Effects of blackberry juice on growth inhibition of foodborne pathogens and growth promotion of Lacto-bacillus. Food Control, 2014, 37, 15-20.

11. Stoner, G.D.; Wang, L.-S.; Cordell, B. Laboratory and clinical studies of cancer chemoprevention by antioxidants in berries. Carcinogenesis, 2008, 29, 1665-1674.

12. Nile, S.H.; Park, S.E. Edible berries: Bioactive components and their effect on human health. Nutrition, 2014, 30, 134-144.

13. Lee, J.; Dosset, M.; Finn, C.E. Rubus fruit phenolic research: The good, the bad, and the confusing. Food Chem. 2012, 130, 785-796.

14. Riaz, M.; Ahmad, M.; Rahman, N. Antimicrobial screening of fruit, leaves, root and stem of Rubus fruticosus. J. Medicinal Plants Res. 2011, 5, 5920-5924.



74

15. Velićanski, A.S.; Cvetković, DD; Markov, S.L. Screening of antibacterial activity of raspberry (Rubus idaeus L.) fruit and pomace extracts. Acta Periodica Technologica 2012, 43, 305-313.

16. Četojević-Simin, D.D.; Velićanski, A.S.; Cvetković, D.D.; Markov, S.L.; Ćetković, G.S.; Tumbas Šaponjac, V.T. et al Bioactivity of Meeker and Willamette raspberry (Rubus idaeus L.) pomace extracts. Food Chem. 2015, 166, 407-413.

17. Danaher, R.J.; Wang, C.; Dai, J.; Mumper, R.J.; Miller, C.S. Antiviral effects of blackberry extract against herpes simplex virus type 1. Oral Surgery Oral Medicine Oral Pathology Oral Radiology Endodontology 2011, 112, 31-35.

18. Singleton, V.L.; Orthofer, R.; Lamuela-Raventos, R.M. Analysis of total phenols and other oxidation substrates and antioxidants by means of Folin-Ciocalteu reagent. In Methods in Enzymology, Oxidant and Antioxidant (Part A); Packer, L., Eds.; Aca-demic Press: San Diego, 1999; pp 152-178.

19. Zhishen, J.; Mengcheng, T.; Jianming, W. The determination of flavonoid contents in mulberry and their scavenging effects on superoxide radicals. Food Chem. 1999, 64, 555-559.

20. Lee, J.; Durst , R.W.; Wrolstad, R.E. Determination of total monomeric anthocyanin pigment content of fruit juices, beverages, natural colorants, and wines by the pH differential method: collaborative study. J AOAC Int. 2005, 88, 1269-1279.

21. Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon, J.; Vistica, D. et al New colorimetric cytotoxicity assay for anticancer-drug screening. J. Nat. Cancer Inst. 1990, 82, 1107-1112.

22. Mayo, W.J. Chemical methods of control: Antimicrobial drugs. In: Laboratory expe-riments in microbiology; Johnson, T.R., Case, C.L., Eds.; The Benjamin/Cummings Publishing Company; San Francisco, 1998; pp 179-181.

23. Tumbas Šaponjac, V.; Gironés-Vilaplana, A.; Djilas, S.; Mena, P.; Ćetković, G.; Mo-reno, D.A. et al Anthocyanin profiles and biological properties of caneberry (Rubus spp.) press residues. J. Sci. Food Agric. 2014, 94, 2393-2400.

24. Da Fonseca Machado, A.P.; Pasquel-Reátegui, J.L.; Fernández Barbero, G.; Martínez, J. Pressurized liquid extraction of bioactive compounds from blackberry (Rubus fruti-cosus L.) residues: a comparison with conventional methods. Food Res. Int. 2014, 77, 675-683.

25. Laroze, L. E.; Díaz-Reinoso, B.; Moure, A.; Zúñiga, M.E.; Domínguez , H. Extrac-tion of antioxidants from several berries pressing wastes using conventional and su-percritical solvents. Eur. Food Res. Technol. 2010, 231, 669-677.

26. de Souza, V. R.; Pereira, P.A.P.; da Silva, T.L.T.; de Oliveira Lima, L.C.; Pio, R.; Queiroz, F. Determination of the bioactive compounds, antioxidant activity and che-mical composition of Brazilian blackberry, red raspberry, strawberry, blueberry and sweet cherry fruits. Food Chem. 2014, 156, 362-368.

27. Pantelidis, G.E.; Vasilakakis, M.; Manganaris, G.A.; Diamantidis, G.G. Antioxidant capacity, phenol, anthocyanin and ascorbic acid contents in raspberries, blackberries, red currants, gooseberries and Cornelian cherries. Food Chem. 2007, 102, 777-783.

28. Wijngaard, H.; Brunton, N. The optimization of extraction of antioxidants from apple pomace by pressurized liquids. J. Agric. Food Chem. 2009, 57, 10625-10631.



75

29. Zhou, S.-H.; Fang, Z.-X.; Lü, Y.; Chen, J.-C.; Liu, D.-H.; Ye, X.-Q.) Phenolics and antioxidant properties of bayberry (Myrica rubra Sieb. et Zucc.) pomace. Food Chem. 2009, 112, 394-399.

30. Galván D’Alessandro, L.; Dimitrov, K.; Vauchel, P.; Nikov, I. Kinetics of ultrasound assisted extraction of anthocyanins from Aronia melanocarpa (black chokeberry) was-tes. Chem. Eng. Res. Design, 2014, 92, 1818-1826.

31. Giusti, M.M.; Jing, P. Natural pigments of berries: functionality and application. In: Berry Fruit, Value added Products for Health Promotion; Zhao, Y., Ed.; CRC: Boca Raton, 2007; pp 105-146.

32. Dutie, S.J. Berry phytochemicals, genomic stability and cancer: evidence for chemo-prevention at several stages in the carcinogenic process. Mol. Nutr. Food Res. 2007, 51, 665-674.

33. Boivin, D.; Blanchette, M.; Moghrabi; A.; Béliveau, R. Inhibition of cancer cell proli-feration and suppression of TNF - induced activation of Nf kappaB by edible berry juice. Anticancer Res. 2007, 27, 937-948.

34. Vattem, D.A.; Lin, Y.-T.; Labbe, R.G.; Shetty, K. Phenolic antioxidant mobilization in cranberry pomace by solid-state bioprocessing using food grade fungus Lentinus edodes and effect on antimicrobial activity against select food borne pathogens. Inno-vative Food Sci. Emerging Technologies, 2004, 5, 81-91.

35. Krisch, J.; Galgóczy, L.; Tölgyesi, M.; Papp, T.; Vágvölgy, C. Effect of fruit juices and pomace extracts on the growth of Gram-positive and Gram-negative bacteria. Acta Biologica Szegediensis, 2008, 52, 267-270.

36. Ördögh, L.; Galgóczy, L.; Krisch, J.; Papp, T.; Vágvölgyi, C. Antioxidant and antimi-crobial activities of fruit juices and pomace extracts against acne-inducing bacteria. Acta Biologica Szegediensis, 2010, 54, 45-49.

37. Nohynek, L.J.; Alakomi, H.-L.; Kähkönen, M.P.; Heinonen, M.; Helander, I.M.; Oks-man-Caldentey, K.-M.; Puupponen-Pimiä, R.H. Berry Phenolics: Antimicrobial Pro-perties and Mechanisms of Action Against Severe Human Pathogens. Nutr. Cancer 2006, 54, 18-32.

38. Mani-López, E.; García, H.S.; López-Malo, A. Organic acids as antimicrobials to control Salmonella in meat and poultry products. Food Res. Int. 2011, 45, 713-721.

39. Diao, W.-R.; Hu, Q.-P.; Zhang, H.; Xu, J.-G. Chemical composition, antibacterial activity and mechanism of action of essential oil from seeds of fennel (Foeniculum vulgare Mill.). Food Control 2014, 35, 109-116.



76

БИОАКТИВНОСТ ТРОПА КУПИНЕ (Rubus fruticosus L.): САДРЖАЈ ПОЛИФЕНОЛА, АНТИРАДИКАЛСКА, АНТИМИКРОБНА И

АНТИТУМОРСКА АКТИВНОСТ

Драгана Д. Четојевић-Симин1*, Александра С. Ранитовић2, Драгољуб Д. Цветковић2, Синиша Л. Марков2, Милица Н. Винчић2, Соња М. Ђилас2

1 Универзитет у Новом Саду, Медицински факултет, Институт за онкологију Војводине, Др Голдмана 4, 21204 Сремска Каменица, Србија

2 Универзитет у Новом Саду, Технолошки факултет Нови Сад, Булевар Цара Лазара 1, 21000 Нови Сад, Србија

Огромне количине индустријског отпада се производе сваке године широм све-та, док његово одлагање представља значајно финансијско оптерећење. Индустриј-ски отпад се може искористити као извор једињења која поседују изузетну хран-љиву вредност и биоактивне особине. Одређен је биолошки потенцијал тропа купи-не, добијен након издвајања сока сорти Чачанска бестрна и Thornfree. Екстракт сор-те Thornfree садржао је више концентрације укупних и мономерних антоцијана, укупних фенола и флавоноида и показао јачу DPPH активност хватања слободних радикала (2,12 mmol TEAC g-1) у поређењу са сортом Чачанска бестрна. Оба екс-тракта су значајно повећала однос апоптоза/некроза на свим испитаним ћелијским линијама. Највећа инхибиција ћелијског раста (EC50=52,5 - 64,7 μg mL-1) и највећи пораст апоптозе (AI=12,2) добијен је на ћелијској линији аденокарцинома дојке. Оба екстракта су инхибирала раст свих испитаних микроорганизама. У референт-ним и „дивљим“ бактеријским сојевима MIC су постигнуте у опсегу од 0,39-25 mg mL-1, а MBC/MFC у опсегу од 0,78-25 mg mL-1. Троп купине је богат извор фито-хемикалија са особинама значајним у исхрани и за промоцију здравља, а које се мо-гу искористити као додаци храни. Кључне речи: троп купине; DPPH тест; ћелијски раст; бактерицидна активност;

фунгицидна активност

Received: 15 September 2017. Accepted: 25 October 2017.

APTEFF, 48, 1-323 (2017) UDC: 582.477.4:543.51+543.544.3 https://doi.org/10.2298/APT1748077D BIBLID: 1450-7188 (2017) 48, 77-83


77

CHEMICAL PROFILE OF Taxodium distichum WINTER CONES

Nina M. Đapić1*, Mihailo S. Ristić2

1 University of Novi Sad, Technical Faculty “Mihajlo Pupin”, Đure Đakovica bb, 23000 Zrenjanin, Serbia

2 Institute for Medicinal Plant Research “Dr Josif Pančić”, Tadeuša Košćuškа 1, 11000 Belgrade, Serbia

This work is concerned with the chemical profile of Taxodium distichum winter cones. The extract obtained after maceration in absolute ethanol was subjected to qualitative analysis by gas chromatography/mass spectrometry and quantification was done by gas chromatograp-hy/flame ionization detector. The chromatogram revealed the presence of 53 compounds, of which 33 compounds were identified. The extract contained oxygenated monoterpenes (12.42%), sesquiterpenes (5.18%), oxygenated sesquiterpenes (17.41%), diterpenes (1.15%), and oxygenated diterpenes (30.87%), while the amount of retinoic acid was 0.32%. Mono-acylglycerols were detected in the amount of 4.32%. The most abundant compounds were: caryophyllene oxide (14.27%), 6,7-dehydro-ferruginol (12.49%), bornyl acetate (10.96%), 6-deoxy-taxodione (9.50%) and trans-caryophyllene (4.20%). KEY WORDS: Taxodium distichum, cones, GC-MS

INTRODUCTION The Taxodium distichum tree is cultivated in urban areas (1-3). The properties and variability of ten Taxodium distichum trees grown in the Futoski Park and eighteen trees grown in the Dunavski Park are well described (2). The tree is ornamental conifer adap-tive to grow in this region. Conifers self-propagation is done by cones. The phytoche-mical cone composition, focused on essential oil, has been determined using gas chroma-tography coupled to mass spectrometry (GC-MS) (4-7). The analysis obtained revealed monoterpenoids: α-pinene (87.3%), β-pinene (1.7%), camphene (1.0%), limonene (1.3%), myrcene (2.0%) and sesquiterpenoid thujopsene (3.7%) (4). The characteristic secondary metabolites present in essential oils of the cones from the Mediterranean Basin revealed the presence of α-pinene (71.3%) as the most abundant essential oil compound (5). The cone essential oil had a monoterpene hydrocarbon limonene in the percentage of 18.7% (5). When milled cones were extracted with n-hexane, the analysis showed that 70% of the extract obtained were diterpenes (8). The essential oil of Taxodium distichum cone was obtained by hydrodistillation, the extraction technique which compromises elevated temperatures and can cause compound thermodegradation (4-7). The milled Taxodium distichum cones extracted with a non-

* Corresponding author: Nina M. Đapić, University of Novi Sad, Technical Faculty “Mihajlo Pupin”, Đure Đa-

kovića bb, 23000 Zrenjanin, Serbia, e-mail: [email protected]



78

polar solvent gave diterpenes as the most abundant compounds (8). In the literature, there are reports on the compounds present in Taxodium distichum cone essential oil obtained by hydrodistillaton and in n-hexane extract, while the compounds present in absolute ethanol extract have not been described up to now. This study was conducted to reveal the compounds present in winter Taxodium distichum cones, obtained by maceration in absolute ethanol.

EXPERIMENTAL

Plant material Coppery-red bald cypress cones were collected in January 2015 in the Futoški Park, Novi Sad, Serbia. The collected cones were air dried for several weeks. After drying, the water content was found to be 7.36±0.11%. The dried cones were ground and milled in a rotating blade coffee grinder Bosch MKM 6003. The powdered plant material (20 g) was dissolved in absolute ethanol (200 ml), and allowed to stand at room temperature for 3 days, with frequent agitation. The powdered plant material was separated, the marc was pressed, and the obtained liquids were combined and filtered. After solvent evaporation, brown oil was obtained. The oil was dissolved in acetone to obtain ~1% sample solution which was subjected to further analyses.

Analysis of compounds using gas chromatography The analysis was performed using gas chromatography followed by flame ionization detection (GC-FID) and gas chromatography-mass spectrometry (GC-MS). The GC-FID analysis was carried out on a Hewlett-Packard, HP-5890 Series II gas chromatograph (Waldbronn, Germany), equipped with a split-splitless injector, HP-5 fused silica capil-lary column (25 m x 0.32 mm id, 0.5 μm film thickness) and equipped with FID detector. The carrier gas (H2) flow rate was 1 mL/min, split ratio 1:30, injector temperature 250 °C, and detector temperature 300 °C, while column temperature was linearly program-med from 40 °-240 °C (at a rate of 4 °C). The same analytical conditions were employed for the GC-MS analysis, where an HP G 1800C GCD Series II Electron Ionization Detector (EID) analytical system (Palo Alto, CA, USA) equipped with split-splitless injector and automatic liquid sampled (ALS) were used, and the carrier gas was helium. The injector was heated to 250 °C, and the transfer line (MSD) to 280 °C. The column temperature was linearly raised from 40 to 260 °C (at a rate of 4 °C). The mass spectra were acquired in electron ionization (EI) mode (70eV), in an m/z range of 40-400. An HP-5MS column (30 m x 0.25 mm, 0.25 μm film thickness) was used. Identification of the individual oil components was accomplished by their retention times compared to analytical standard substances of available terpenoids, by matching mass spectral data with those kept in the mass spectra library (Wiley 275.1), using computer search and the literature (9, 10). The confirmation was done using a calibrated AMDIS program for determination of experimental values for retention indices of recorded constituents and comparing them with those from the literature.



79

RESULTS AND DISCUSSION The chromatogram obtained by the GC-MS analysis revealed the presence of 53 constituents, of which 33 were identified, and they are listed in Table 1. The predominant compounds in the extract analyzed were: caryophyllene oxide (14.27%), 6,7-dehydro-fer-ruginol (12.49%), bornyl acetate (10.96%), 6-deoxy-taxodione (9.50%), trans-caryophyl-lene (4.20%), humulene epoxide II (1.31%), trans-ferruginol (4.23%), sugiol (2.97%), 1-glyceryl oleate (2.88%), 2-glyceryl stearate (1.44%), while the other components were present in less than 1%. The compounds that were not identified made in total 13.70%. The oxygenated monoterpenes identified in the extract were: camphene hydrate, α-terpi-neol, α-campholenol, bornyl acetate, trans-verbenyl acetate, trans-pinocarvyl acetate, iso-dihydro carveol acetate and trans-carvyl acetate. Among these compounds, bornyl acetate was most abundant, while the other oxygenated monoterpenes were present in less than a half percent. Of sesquiterpenes trans-caryophyllene and α-humulene were identified. The amount of trans-caryophyllene in the extract was 4.20% and that of α-humulene less than one percent. The oxygenated sesquiterpenes were: caryophyllene oxide (14.27%), three humulene epoxide stereoisomers were detected where the dominant among them was hu-mulene epoxide II, then caryophylla-4(12),8(13)-dien-5β-ol (0.35%), 14-hydroxy-(Z)-ca-ryophyllene (0.81%) and iso-longifolol (0.22%). The diterpenes detected were: pimara-diene (0.39%), abietatriene (0.47%) and abietadiene (0.29%). There were six oxygenated diterpenes in the extract: manool oxide (0.75%), pimarinal (0.91%), 6,7-dehydro-ferrugi-nol (12.49%), trans-ferruginol (4.23%), 6-deoxy-taxadione (9.50%) and sugiol (2.97%). Retinoic acid was detected in the amount of 0.32%. Two monoacylglycerols were present in the extract: 1-glyceriyl oleate (2.88%) and 2-glyceryl stearate (1.44%). The previous report on the volatile oil chemical compounds in Taxodium distichum seed cones mentioned that the most abundant compound was dextro pinene (85%), along with: dextro limonene (5%), carvone (3%), tricyclic sesquiterpene (3%) and a “pseudo terpene alcohol” (2%) (6). Monoterpenes were the dominant compounds detected in early work on the chemical composition of Taxodium distichum cones. The essential oil compound analysis of Taxodium distichum cones, from the Mediterranean Basin, where climate is mild with rainy winters and hot and dry summers, showed that α-pinene was present in 71.3% and limonene in 18.7% (5). The North African Taxodium distichum cones essential oil contained α-pinene (87.3%), β-pinene (1.7%), camphene (1.0%), myrcene (2.0%), limonene (1.3%) and thujopsene (3.7%) (4). The Taxodium distichum cones from West Africa contained 60.5% α-pinene, 17.6% thujopsene and 29 other com-pounds (7). The compound present in all essential oils analyzed was α-pinene, while li-monene was the next compound present in the analyzed samples. The thujopsene is pre-sent in African Taxodium distichum cones essential oil in different quantities. The Taxo-dium distichum cones grown in North America, Mediterranean Basin, North and West Africa have some compounds in common, although their quantity varies. The different composition can be attributed to the different environmental effects (11). The Taxodium distichum winter cones extract obtained after maceration in absolute ethanol differ in chemical composition from previously described essential oils. The winter cones extract contained oxygenated monoterpenes, where the most abundant was bornyl acetate



80

(10.96%), whereas 14 components were with the sesquiterpenoid structure (~ 23%), with caryophyllene oxide being the most abundant (14.27%), while 63% comprised mainly diterpenoids with a small amount of other highly volatile substances.

Table 1. Chemical composition of Taxodium distichum winter cones macerated in absolute ethanol

Peak Constituents KIE KIL RT/MS %

m/m RRT CI

1 α-Campholenal 1116.9 1122 12.78 0.37 0.338 26 2 Camphene hydrate 1138.4 1145 13.52 0.29 0.357 20 3 α-Terpineol 1183.6 1186 15.09 0.16 0.399 11 4 α-Campholenol 1197.0 1190 15.55 0.41 0.411 29 5 Bornyl acetate 1275.3 1287 18.20 10.96 0.481 768 6 trans-Verbenyl acetate 1284.5 1291 18.52 0.25 0.490 17 7 trans-Pinocarvyl acetate 1288.3 1298 18.65 0.21 0.493 15 8 iso-Dihydro carveol acetate 1323.8 1326 19.82 0.29 0.524 21 9 trans-Carvyl acetate 1327.3 1339 19.93 0.46 0.527 32

10 trans-Caryophyllene 1403.5 1417 22.39 4.20 0.592 294 11 α-Humulene 1437.4 1452 23.43 0.98 0.620 68 12 Caryophyllene oxide 1570.2 1582 27.38 14.27 0.724 1000 13 1-Hexadecene 1577.7 1588 27.60 0.18 0.730 13 14 Humulene epoxide I 1584.5 1593 27.77 0.17 0.734 12 15 Humulene epoxide II 1593.5 1608 28.07 1.31 0.742 92 16 Humulene epoxide III 1598.3 1620 28.20 0.28 0.746 20 17 Caryophylla-4(12),8(13)-dien-5β-ol 1622.5 1639 28.88 0.35 0.764 24 18 n. i. 1644.6 29.49 0.14 0.780 10 19 14-Hydroxy-(Z)-caryophyllene 1658.9 1666 29.88 0.81 0.790 57 20 iso-Longifolol 1721.4 1728 31.59 0.22 0.835 15 21 n. i. 1794.4 33.50 0.16 0.886 12 22 n. i. 1886.4 35.83 0.33 0.948 23 23 n. i. 1902.0 36.23 0.29 0.958 20 24 Pimaradiene 1939.0 1948 37.11 0.39 0.981 27 25 Manool oxide 1968.6 1987 37.82 0.75 1.000 52 26 Ethyl hexadecanoate 1978.1 1992 38.05 0.41 1.006 29 27 Abietatriene 2033.9 2055 39.35 0.47 1.041 33 28 Abietadiene 2057.8 2087 39.90 0.29 1.055 20 29 n. i. 2069.5 40.16 0.41 1.062 29 30 Pimarinal 2161.1 217X 42.20 0.91 1.116 64 31 n. i. 2209.7 43.26 0.25 1.144 17 32 λ-8(17),14-diene-6,13-diol* 2248.5 2248 44.07 0.97 1.166 68 33 M=356 2279.0 44.70 4.62 1.182 323 34 6,7-Dehydro-ferruginol 2314.5 2315 45.45 12.49 1.202 875 35 trans-Ferruginol 2317.7 2331 45.51 4.23 1.203 297 36 n. i. 2329.2 45.74 0.93 1.210 65 37 n. i. 2339.8 45.96 1.51 1.215 106 38 Retinoic acid 2353.8 46.25 0.32 1.223 22



81

Table 1. Continuation

Peak Constituents KIE KIL RT/MS %

m/m RRT CI

39 n. i. 2361.9 46.41 0.34 1.227 24 40 n. i. 2398.4 47.15 2.68 1.247 188 41 n. i. 2407.6 47.33 0.23 1.252 16 42 n. i. 2413.9 47.45 0.25 1.255 17 43 6-Deoxy-taxodione 2450.7 2435 48.16 9.50 1.274 665 44 n. i. 2498.2 49.08 3.45 1.298 242 45 n. i. 2559.1 50.24 0.24 1.328 17 46 n. i. 2566.2 50.37 1.67 1.332 117 47 n. i. 2573.8 50.52 0.52 1.336 36 48 n. i. 2602.3 51.05 5.56 1.350 389 49 Sugiol 2620.8 2629 51.39 2.97 1.359 208 50 1-Glyceryl oleate (monoolein) 2677.8 2714 52.42 2.88 1.386 202 51 n. i. 2695.6 52.75 0.33 1.395 23 52 2-Glyceryl stearate (2-monostearin) 2702.7 n/a 52.86 1.44 1.398 101 53 n. i. 2745.0 53.61 1.93 1.418 135

Legend: KIE = Kovats (retention) index experimentally determined (AMDIS) KIL = Kovats (retention) index - literature data (9) RT/MS = Retention time of the corresponding constituent obtained by GC/MS RRT = Relative retention time to selected constituent [Caryophyllene oxide = 1.000] % m/m = Area % obtained by MSD CI = Concentration index *= Tentative identification n.i.= not identified n/a = not available

Figure 1. The GC/MS chromatogram of the extract of Taxodium distichum winter cones



82

CONCLUSION The GC-MS analysis of the ethanolic extract of Taxodium distichum winter cones revealed the presence of 53 compounds, of which 33 compounds were identified. The constituents of the extract were as follows: eight oxygenated monoterpenes (12.42%), two sesquiterpenes (5.18%), oxygenated sesquiterpenes (17.41%), diterpenes (1.15%) and oxygenated diterpenes (30.87%). Oxygenated diterpenes were the most abundant class of compounds in the amount of 30.87%, then oxygenated sesquiterpenes (17.41%), followed by oxygenated monoterpenes (12.42%). Sesquiterpenes were present in 5.18%, diterpens in 1.15% and monoterpene in 0.37%. Retinoic acid was present in 0.32%. Monoacylglycerols were detected in the amount of 4.32%. The extract can be used for the isolation of most abundant compounds present: one oxygenated sesquiterpene caryo-phyllene oxide (14.27%), one oxygenated diterpene 6,7-dehydro-ferruginol (12.49%), and one oxygenated monoterpene bornyl acetate (10.96%).

REFERENCES

1. Drazic, D.; Batos, B. Mocvarni cempres Taxodium distichum (L.) Rich. u uslovima Beograda, 7th Symposium on Flora of Southeastern Serbia and Neighbouring Regions, Dimitrovgrad, 6-9 June 2002, Proceeding, p. 195

2. Ninic-Todorovic, J.; Ocokoljic, M. Varijabilnost taksodijuma (Taxodium distichum (L.) Rich.) u parkovima Novog Sada, 7th Symposium on Flora of Southeastern Serbia and Neighbouring Regions, Dimitrovgrad, 6-9 June 2002, Proceeding, p. 125

3. Sijacic-Nikolic, M.; Vilotic D.; Veselinovic, M.; Mitrovic, S.; Jokanovic, D. Bald cypress (Taxodium distichum (L.) Rich.) in the protected area “Veliko ratno ostrvo”, Bulletin of the Faculty of Forestry 2010, 103, 173-184.

4. El Tantawy, M. E.; El Sakhawy, F. S.; El Sohly, M. A.; Ross, S. A. Chemical Composition and Biological Activity of the Essential Oil of the Fruit of Taxodium distichum L. Rich. Growing in Egypt, J. Essent. Oil Res. 1999, 11 (3), 386-392.

5. Flamini, G.; Cioni, P. L.; Morelli, I. Investigation of the Essential Oil of Feminine Cones, Leaves and Branches of Taxodium distichum from Italy, J. Essent. Oil Res. 2000, 12 (3), 310-312.

6. Odell, A. F. The oil of the southern cypress, JACS, 1912, 34: 824-826. 7. Ogunwande, I. A.; Olawore, N. O.; Ogunmola, O. O.; Walker, T. M.; Schmidt, J. M.;

Setzer, W. N. Cytotoxic effects of Taxodium distichum oils, Pharm. Biol. 2007, 45: 106-110.

8. Kusumoto, N.; Ashitani, T.; Murayama, T.; Ogiyama, K.; Takahashi, K. Antifungal Abietane-Type Diterpenes from the Cones of Taxodium distichum Rich, J. Chem. Ecol. 2010, 36: 1381-1386.

9. Adams, R. P. Identification of Essential Oil Components by Gas Chromato-graphy/Mass Spectrometry, 4th Ed., Allured Publishing Corp., Carol Stream, IL 60188 USA, 2007.

10. Adams, R. Identification of Essential Oil Components by Gas Chromatography/Mass Spectrometry. Allured Publishing, Illinois, USA, 1995.



83

11. Clark, R. J.; Menary, R. C. Environmental Effects on Peppermint (Mentha piperita L.). II. Effects of Temperature on Photosynthesis, Photorespiration and Dark Respi-ration in Peppermint with reference to Oil composition, Funct. Plant Biol. 1980, 7 (6): 693-697.

ХЕМИЈСКА ЈЕДИЊЕЊА ПРИСУТНА ЗИМИ У ШИШАРКАМА МОЧВАРНОГ ЧЕМПРЕСА

Нина М. Ђапић1, Михаило С. Ристић2

1 Универзитет у Новом Саду, Технички Факултет “Михајло Пупин”, Ђуре Ђаковића бб, 23000 Зрењанин, Србија

2 Институт за проучавање лековитог биља „Др Јосиф Панчић“, Тадеуша Кошћушка 1, 11000 Београд, Србија

Екстракт Taxodium distichum шишарки убраних током зиме добијен је након ма-церације у апсолутном етанолу. Добијени екстракт је квалитативно анализиран употребом гасне хроматографије/масене спектрометрије, док је квантитативна ана-лиза узорка урађена употребом гасне хроматографије са пламено јонизационим детектором. У добијеном хроматограму било је 53 пика и од тога идентификовано је 33 једињења. Екстракт је садржао оксидоване монотерпене (12,42%), сесквитер-пене (5,18%), оксидоване сесквитерпене (17,41%), дитерпене (1,15%) и оксидоване дитерпене (30,87%). Ретиноинска киселина је чинила 0,32% екстракта. Два моно-ацилглицерола су детектована у количини од 4,32%. Доминантна једињења екстрак-та су била: кариофилен оксид (14,27%), 6,7-дехидро-феругинол (12,49%), борнил ацетат (10,96%), 6-деокси-таксадионе (9,50%) и trans-кариофилен (4,20%). Kључне речи: Taxodium distichum, шишарке, GC-MS

Received: 05 July 2017. Accepted: 15 October 2017

APTEFF, 48, 1-323 (2017) UDC: 664.641.1:664.69:543.92 https://doi.org/10.2298/APT1748085F BIBLID: 1450-7188 (2017) 48, 85-93


85

APPLICATION OF TWO RHEOLOGICAL METHODS FOR FLOUR TESTING

TO PREDICT PASTA QUALITY

Jelena S. Filipović1*, Vladimir S. Filipović2

1University of Novi Sad, Institute of Food Technology, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia

2University of Novi Sad, Faculty of Technology Novi Sad, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia

Two rheological methods were used for determining the quality of five different spelt flour samples as a raw material for pasta making. The measurements encompassed correlations between the mixolab and alveograph data and textural attributes for dry and cooked pasta, as well as pasta color. The flour denoted by 5 was scored as the most convenient raw material for pasta due to the following mixolab data: protein weakening (0.67±0.05), starch gelatinization (1.69±0.10), amylase activity (2.34±0.15), starch retro-gradation (3.38±0.27), as well as good alveograph data: resistance to deformation (49±2), dough extensibility (77±1) and deformation energy (121±3) and their correlation with quality of Pasta 5 (hardness 5182.06±7.48g, adhesiveness 10.25±1.16 gsec, work of shear 410.33±5.44 gsec, fracturability 24.38±1.84 g, flexibility 27.02±1.69 mm). The method of principal component analysis proved to be a useful tool for detection of struc-ture in the relationship beetween measuring rheological data and spelt pasta texture and color attributes. KEY WORDS: spelt, mixolab data, alveograph data, texture, color, pasta

INTRODUCTION The areas under spelt have been increasing recently, which is associated with the growing interest in healthy lifestyle and healthy food, as well as with the increase in organic farming development. Spelt is valued due to its health-promoting qualities and agronomic advantages, such as high pest resistance and due to the fact that it can be cultivated on low fertility, poorly drained soils (1, 2). Globally, there is an increasing interest in using organic farming systems and maximizing grain production in low-input scenarios. Today, spelt is maybe, popular due to the perception that it is a ‘healthier’, more ‘natural’, or less ‘over-bred’ grain than commercial bread wheat (3-5). The rheological properties of wheat flour dough are essential for the successful manufacturing of flour based products because they determine its behavior during mechanical handling, thereby affecting the quality of the finished products (6). For a long time, only fari-nograph of semolina has been used to predict pasta quality (7), but lately some authors

* Corresponding author: Jelena S. Filipović, University of Novi Sad, Institute of Food Technology, Bulevar




86

have also used other rheological methods (8-10) for predicting pasta quality. Though dough for either alveograph or mixolab testing relate to fully developed gluten network contrary to crumbly pasta dough, resulting data give a valuable insight into pasta raw material quality. Alveograph data can be useful in predicting gluten quality and protease activity (10). The Mixolab technique records the dough changes when subjected to large deformations and to temperature sweeps (11, 12), also experienced in pasta technology. Mixolab data may also contribute to starch quality determination. Both gluten and starch, are closely related to pasta textural attributes (8, 9), as activity of proteolytic and amylolytic enzymes cause color changes (8). Pasta has a primary role in human nutrition thanks to its complex carbohydrate profile, large global distribution, and the most popular extruded product with a relative long shelf life if properly stored (13, 14). It is also a widely consumed food because of its low cost and easy preparation (15). Pasta quality can be expressed in terms of cooking characteristics, color, taste and aroma, and these factors are of great importance for the consumer (16). The quality of spaghetti pasta cooking can be evaluated in terms of stickiness, firmness, overcooking tolerance, degree of swelling, and cooking loss of solids (17). High quality pasta is manufactured using durum wheat semolina because of its very good cooking quality and high consumer acceptance. However, in many countries the availability of durum wheat for pasta pro-duction is limited, and there have been many reports on the partial or complete replace-ment of semolina with common wheat flour or others flours (15). Having in mind that a limiting number of methods are used for pasta flour testing, the aim of this paper is to investigate the posossibility of using mixolab and alveograph data and relate them to the quality of spelt pasta, defined by its texture and color.

EXPERIMENTAL

Materials and methods Five spelt cultivars noted as: Flour 1, Flour 2, Flour 3, Flour 4, and Flour 5 (particle size of whole meal flours ranged between 200 and 300 μm), representing the average quality from producing areas in the years 2015-2016 in Serbia were used in the investiga-tion. Dough rheology testing was performed by Mixolab (Chopin, Tripette et Renaud, Paris, France) and Alveograph (Chopin-Alveograph). The Mixolab procedure was earlier described by Filipović et al. (18). The viscoelastic behavior was also determined by the alveograph test (Chopin-Alveograph), following the standard procedure (19). The moni-tored parameters were: resistence to deformation (P), dough extensibility (L), deforma-tion energy (W), and curve configuration ratio (P/L). Pasta was made using the device "La Parmigiana D45" MAC 60, (Fidenza, Italy), according to the procedure discribed by Filipović et al. (18). Textural properties of cooked pasta were measured on a Texture Analyzer TA.HD plus (Stable Micro System, U.K.) equipped with a 5-kg load cell, described previously (18). Pasta color attributes were measured instrumentally using a Chroma meter (CR-400, Konica, Minolta, Tokyo, Japan) tri-stimulus colorimeter, as described by Filipović et al, (20). Descriptive statistical analyses for the obtained rheolo-gical and technological quality parameters were expressed as the mean ± standard devia-



87

tion (SD). The evaluation of analyses of variance (ANOVA) and principal component analyses (PCA) of the obtained results were performed using StatSoft Statistica 10.0® software. The collected data were subjected to ANOVA for the comparison of means, and significant differences are calculated according to post-hoc Tukey’s HSD (honestly significant differences) test at p<0.05 significance level, 95% confidence limit.

RESULTS AND DISCUSSION Different samples of Spelt flour are used to find relationships between parametres charcterizing the main dough proteins and starch, which are responsible for pasta texture, cooking performance and also may adversely influence the pasta color. Table 1 lists the physical responses during the mixing-pasting-gelling phases of the investigated five samples, which according to mixolab data, do not differ statistically among the samples.

Table 1. Rheological properties of flour samples

Mixolab properties Flour 1 Flour 2 Flour 3 Flour 4 Flour 5

Protein weakening (PW), Nm

0.63±0.05а 0.57±0.01а 0.56±0,08а 0.58±0.09а 0.67±0.05а

Starch gelatinization (SG), Nm

1.37±0.18а 1.35±0.14а 1.54±0.06а 1.66±0.17а 1.69±0.10а

Amylase activity (AA), Nm

2.05±0.15а 2.07±0.21а 2.16±0.05а 2.20±0.07а 2.34±0.15а

Starch retrogradation (SR), Nm

2.94±0.11а 2.95±0.08а 3.10±0.13а 3.29±0.17а 3.38±0.27а

Alveograph properties Resistance to deformation (P), mm

50±2а 45±1b 44±1b 40±1c 49±2a

Dough extensibility (L), mm

79±a3 87±1b 96±2c 98±2c 77±1a

Deformation energy (W), 10E-4J

106±2a 106±1a 117±1bc 115±1b 121±3c

Curve configuration ratio (P/L)

0.63 0.52 0.46 0.41 0.64

The results are presented as mean±SD, different letters within the same row indicates the significant difference in mean values (p<0.05), according to Tukey’s HSD test. Number of repetitions: n=6.

Flour 2 is characterized with the lowest force for protein weakening (0.57±0.01) and starch gelatinization (1.35±0.14), which indicates a weak three-dimensional viscoelastic structure. Flour 1 and Flour 2 have the lowest and very similar data for amylase activity and starch retrogradation compared with other samples of flour, suggesting good gluten and starch characteristics and low enzyme activity. Flour 5 has the highest mixolab properties, viz.: protein weakening 0.67±0.05, starch gelatinization 1.69±0.10, amylase



88

activity 2.34±0.15, and starch retrogradation 3.38±0.27. The highest protein weakening and starch gelatinization data contributed positively to the protein starch network and stability of pasta, which is consistent with the previous findings of Filipović et al (15). It can be expected that these parameters will correlate with Pasta 5 quality (Table 2). The alveograph characteristics of five different flour of spelt dough are presented in Table 1. Resistance to deformation (P value) is an indicator of the dough's ability to retain air bubble (10, 21) and higher values are favorable for pasta dough. Statistically significant differences between P values of different flours (2 and 3), flours (1 and 5) and Flour 4 are noticed, and therefore the differences in the resulting pastas can be also expected. The highest values for (P) resistance to deformation (50±2) were observed for Flour 1 and Flour 5, while the lowest value was found for Flour 4 (40±1). The extensibility of dough (L), as a predictor of the processing characteristics of dough it also points out to protease activity. Statistically significant differences were observed between the dough extensibility and deformation energy (W) among the flour samples. The highest value for W (121±3) was found for Flour 5, which is far below the values (160-250) reported by Chopin (21) for durum dough, while the lowest value was noticed for Flour 1 and Flour 2 (106±2). Based on the alveograph data the most favorable raw material for pasta spelt was Flour 5.

Table 2. Quality characteristics of spelt pasta

The results are presented as means, different letter within the same row indicates the significant difference in mean values (p<0.05), according to Turkey’s HSD test. Number of repetitions: n=6.

Pasta 1 Pasta 2 Pasta 3 Pasta 4 Pasta 5

Texture attributes of dry pastas

Fractorability (g) (F)

17.75±0.75a 18.07±0.81a 20.12±1.34ab 23.12±0.68bc 24.38±1.84c

Flexibility (mm) (FB)

13.33±0.28a 16.65 ±0.74b 17.33±0,12b 26.29±1.77c 27.02±1.69c

Color of dry pastas

Brightness (L) 83.61±0.05a 83.56±0.08a 85.05±0.10b 84.52±0.15c 86.19±0.09d

Share of red (a) color

-1.16±0.02a -1.32±0.01b -1.45±0.01c -1.39±0.02d -1.38±0.02d

Share of yellow color (b)

14.29±0.05a 14.54±0.03b 14.16±0.04c 13.46±0.05d 13.09±0.06e

Texture attributes of cooked pasta

Hardness (HR), g

2320.97±8.84a 3190.01±6.49b 3752.87±5.85c 4181.32±10.02d 5182.06±7.48e

Adhesiveness (AD), gsec

36.06±3.78c 19.05±2.91b 23.70±3.71b 24.60±1.13b 10.25±1.16a

Work of Shear (WS), gsec

170.10±8,28a 176.57±2.98a 204.10±1.72b 386.20±8.34c 410.33±5.44d



89

Table 2 shows pasta quality (textural attributes of dry and cooked pasta and pasta color) of different flour samples. Statistically significant differences between the values of technological parameters were noted between the pasta samples. The best techno-logical quality (hardness 5182.06±7.48g, adhesiveness 10.25±1.16a gsec, work of shear 410.33± 5.44 gsec, fracturability 24.38±1.84 flexibility 27.02±1.69) was obtained for Pasta 5, as it was expected based on the mixolab (PW, SG, AA and SR) and alveograph (P, L and W) data. On the other hand, the worst technological quality (hardness 2320.97±8.84 g, adhesiveness 36.06±3.78c gsec, work of shear 170.10± 8.28 gsec, fractu-rability 17.75±0.75 flexibility 13.33±0.28) was observed for Pasta 1, as indicated by the mixolab data (starch gelatiniyation, amylase activity and starch retrogradation). Pasta 5 showed a statistically significant difference in texture attributes compared to other pastas, which was consistent with the alveograph data (maximum of deformation energy). On the whole, according to the textural attributes of dry and cooked pastas, Pasta 5 had the best quality. Color of the pasta is also a very important parameter for product acceptance by consumers (22, 23). The maximum of brightness (86.19±0.09) was obtained for Pasta 5, as indicated by the alveograph (L and W) and mixolab data (SG, AA and SR), while the minimum of brightness 83.61±0.05 and share of yellow color of 14.29±0.05 were observed for Pasta 1, which is directly related to the mixolab data (starch gelatinization and amylase activity). Table 3 shows the correlations established between the rheological parameters (mixolab and alveograph) (Table 1) and technological quality of pasta (Table 2).

Table 3. Correlation analysis of the rheological properties and quality characteristics of spelt pasta

Values in bold are different from 0 at a significance level alpha=0, 1

The correlation analysis showed that there is a positive correlation between the flour rheological properties (mixolab data) SG, AA and SR and the pasta technological quality parameters of HR, AD, WS, F, FB and L*, and a negative one with b*, which was statistically significant at the p<0.1 level. The alveograph data showed that only W correlated positively to HR, AD, F and L* and negatively to b*.

Rheological properties of flour samples

PW SG AA SR P L W P/L

Qu

alit

y ch

arac

teri

stic

s of

sp

elt

pas

ta

HR 0.3291 0.9046 0.9693 0.9404 -0.2025 0.0472 0.9084 -0.0722

AD 0.3395 0.8498 0.9581 0.8921 -0.1282 -0.0116 0.8991 -0.0097

WS 0.4502 0.9278 0.8938 0.9717 -0.2359 0.0181 0.7641 -0.0610

F 0.4024 0.9786 0.9547 0.9996 -0.2398 0.0724 0.8854 -0.0938

FB 0.3239 0.9206 0.8895 0.9640 -0.3462 0.1193 0.7718 -0.1716

L* 0.4731 0.8557 0.9533 0.8590 0.0793 -0.1142 0.9668 0.,1447

a* 0.3588 -0.6619 -0.5981 -0.5977 0.6154 -0.6099 -0.7568 0.6011

b* -0.5815 -0.9387 -0.9334 -0.9741 0.0638 0.1060 -0.8248 -0.0888



90

The full auto-scaled data matrix was submitted to PCA for the detection of structure in the relationship between measuring parameters and different samples. For visualizing the data trends and the discriminating efficiency of the used descriptors a scatter plot of samples using the first two principal components (PCs) from PCA of the data matrix was obtained (Fig. 1).

Figure 1. Biplot diagram of flour/pasta samples regarding flour rheological

properties and pasta technological quality As can be seen, there is a neat separation of the five flour samples, according to flour rheological properties and pasta technological quality. The distribution of flour samples locations on the biplot diagram according to different orientation of responses points at the differences in flour quality. The results show that the first two PCs account for 92.96% of the total variability, and can be considered sufficient for data representation. The observed flour rheological properties and pasta technological quality characteristics, SR (with 9.14% contribution based on the correlation), BF (9.12%), AA (8.96%), HR (8.88%), SG (8.84%), FB (8.31%), AD (8.31%), b* (8.30%), WS (8.17%) and W (8.15%) contributed the most to the first factor calculation, while P/L (23.59%), L (23.30%), P (21.50%) and PW (19.94%) contributed more to the second factor coordinate calculation. The PCA plot also shows a good correlation between SG and L*, AA, WS, AD, HR, F, FB, W, b and SR; and also between P/L and PW.



91

From the PCA plot it can also be seen that the sample of Flour 5 was characterized by the highest L, AA, WS, AD, HR, F, FB, W, SR and SG, while the sample of Flour 1 was characterised by the lowest values of these parameters.

CONCLUSIONS Based on these investigations, the mixolab (SG, AA, SR) and alveograph data (W) gave useful information about the spelt flour as a raw material for pasta as they correlated directly with the quality of pasta (HR, AD,WS, FB, F, L* and b*). The alveograph data showed statistically significant differences in quality indicators of samples as raw mate-rial for the production of pasta, contrary to the mixolab data. Regarding the high corre-lations of SG; AA and SR with HR, AD, WS, F, FB, L and b*, the mixolab data have to be combined with other rheological methods that point to gluten quality. The mixolab data for Flour 5 (protein weakening 0.67±0.05, starch gelatinization 1.69±0.10, amylase activity 2.34±0.15), starch retrogradation 3.38±0.27 are in high correlation with spelt Pasta 5 texture (hardness 5182.06±7.48g, adhesiveness 10.25±1.16gsec, work of shear 410.33±5.44gsec, fracturability 24.38±1.84 and flexibility 27.02±1.69). The application of PCA provided a better visualization of the correlations of the tested responses and the flour quality differentiation.

Acknowledgement These results are part of the project supported by the Ministry of Education and Science of the Republic of Serbia, III 46005 and TR 31029.

REFERENCES 1. Koenig, A.; Konitzer, K.; Wieser, H.; Koehler, P. Classification of spelt cultivars

based on differences in storage protein compositions from wheat, Food Chem. 2015, 168, 176-182.

2. Biel, W.; Stankowski, S.; Jaroszewska, A.; Pużyński, S.; Bośko, P. The influence of selected agronomic factors on the chemical composition of spelt wheat (Triticum aestivum ssp. spelta L.) grain. J. Integr. Agr. 2016, 15 , 1763-1769.

3. Evans, J.; Neeson, R.; Burnett, V.; Luckett, D.J.; Fettell, N.A. Phosphorus-use efficiency, growth and yield of spelt wheat (Triticum aestivum ssp. spelta) compared with standard wheat (T. aestivum ssp. vulgare) in south-eastern, J.Organic System. 2014, 9, 63-78.

4. Bojanska, T.; Frančakova, H. The use of spelt wheat (Triticum spelt L.) for baking applications. Rost. Vyroba, 2002, 48, 141-147.

5. Callejo, J.M.; Vargas-Kostiuk, M. E.; Rodríguez-Quijano, M. Selection, training and validation process of a sensory panel for bread analysis: Influence of cultivar on the



92

quality of breads made from common wheat and spelt wheat, J. Cereal Sci. 2015, 61, 55-62.

6. Agyare, K.K.; Addo, K.; Xiongo, Y.L.; Akoh, C.C. Effect of structrured lipid on alveograph characteristics, baking and textural qualites of soft wheat flour, J. Cereal Sci. 2005, 42, 309-316.

7. Kaluđerski, G.; Filipović, N. (1998). Methods for the investigation of cereals, flour and final product quality, Faculty of Technology, Novi Sad, p. 71-291.

8. Larrosa, V.; Lorenzo, G.; Zaritzky, N.; Califano, A. Dynamic rheological analysis of gluten-free pasta as affected by composition and cooking time, J. Food Eng. 2015, 160, 11-18.

9. De la Pena, E.; Manthey, F.; Patel, B.; Campanella, O. Rheological properties of pasta dough during pasta extrusion: Effect of moisture and dough formulation, J. Cereal Sci. 2014, 60, 346-351.

10. Marchylo, B. A.; Dexter, J. E. (2001). Pasta Production, In: Cereals Processing Tech-nology, G. Owens, (Ed.), 109-130, Woodhead Publishing Limited, ISBN 0-8493-1219-1, Cambridge, UK.

11. Ozturk, S.; Kahraman, K.; Tiftik, B.; Koksel, H. Predicting the cookie quality of flours by using Mixolab (R), Eur. Food Res. Technol. 2008, 227, 1549-1554.

12. Rosell, C.; Santos, E.; Collar, C. Physical characterization of fiber-enriched bread doughs by dual mixing and temperature constraint using the Mixolab, Eur. Food Res. Technol. 2010, 231, 535-544.

13. Verado, V.; Feroli, F.; Riciputi, Y.; Lafelice, G. Evaluation of lipid oxidation in spa-ghetti pasta enriched with long chain n-3 polyunsaturated fatty acids under differrent storage conditions, Food Chem. 2009, 114, 472-477.

14. Lorusso, A.; Verni, M.; Montemurro, M.; Coda R.; Giuseppe Rizzello C. Use of fer-mented quinoa flour for pasta making and evaluation of the technological and nutri-tional features, LWT - Food Sci. Technol. 2017, 78, 215-221.

15. Biernacka, B.; Dzik, D.; Gawlik-Dziki, U.; Różyło, R.; Siastała, M. Physical, sen-sorial, and antioxidant properties of common wheat pasta enriched with carob fiber, LWT - Food Sci. Technol. 2017, 77, 186-192.

16. Chillo, S.; Laverse, J.; Falcone, P. M.; Protopapa, A.; Del Nobile, M. A. Influence of the addition of buckwheat flour and durum wheat bran on spaghetti quality. J. Cereal Sci. 2008, 47, 144-152.

17. Gallegos-Infante, J.A.; Rocha-Guzman, N.E.; Gonzalez-Laredo, R.F.; Ochoa-Martí-nez, L.A.; Corzo, N.; Bello-Perez, L.A.; Medina-Torres, L.; Peralta-Alvarez, L.E. Quality of spaghetti pasta containing Mexican common bean flour (Phaseolus vulga-ris L.), Food Chem. 2010, 119, 1544-1549.

18. Filipović, J.; Pezo, L.; Filipović, N.; Filipović, V.; Bodroža-Solarov, M.; Plančak, M. Mathematical approach to assessing spelt cultivars (Triticum aestivum subsp. spelt) for pasta making, Int. J. Food Sci. Tech. 2013, 48, 195-203 .

19. AACC. (1994). Approved methods of AACC (9th ed.). St. Paul, MN: American Asso-ciation of Cereal Chemists (Methods54-30 A).



93

20. Filipović, J.; Pezo, L.; Filipović, V.; Brkljača, J.; Krulj, J.: The effects of ω-3 fatty acids and inulin addition to spelt pasta quality, LWT - Food Sci. Technol. 2015, 63, 43-51.

21. www. treaetrisan.net/flour_criteria_judging.htm 22. Fu, B.X.; Schlichting, L.; Pozniak, C. J.; Singh, A.K. Pigment loss from semolina to

dough: Rapid measurement and relationship with, pasta colour. J. Cereal Sci. 2013, 57, 560-566.

23. Cervigni, G.; Dubcovsky, J. A deletion at the Lpx-B1 locus is associated with low lipoxygenase activity and improved pasta colour in durum wheat (Triticum turgidum ssp. durum). J. Cereal Sci. 2007, 45, 67-77.

ПРИМЕНА ДВЕ РЕОЛОШКЕ МЕТОДЕ ЗА АНАЛИЗУ БРАШНА У

ПРЕДВИЂАЊУ КВАЛИТЕТА ТЕСТЕНИНЕ

Jелена C. Филиповић1*, Владимир С. Филиповић2

*1 Универзитет у Новом Саду, Институт за прехрамбене технологије, Булевар Цара Лазара 1,

21000 Нови Сад, Србија 2 Универзитет у Новом Саду, Технолошки факултет Нови Сад, Булевар Цара Лазара 1,

21000 Нови Сад, Србија

Старе сорте пшенице, као што је спелта, могу се користити за производњу тестени-не и специјаних врста хлеба. За производњу тестенине са измењеном нутритивном вредности у односу на уобичајену тестенину или за производњу органске хране, спелта се показала погодном сировином. Квалитет тестенине од спелте се може ценити на ос-нову текстуре, боје и укуса, а реолошке особине брашна се могу дефинисати миксола-бом и алвеографом. Резултати приказују примену две реолошке методе за одређивање кавлитета 5 раличитих узорака брашна спелте као сировине за производњу тестенине. Утврђена је корелације између миколабских и алвеографских показатеља и текстурал-них својстава суве и куване тестенине као и бојe тестенине. Брашно 5 има најбоље оце-не миксолабских показатеља што показују следећи параметри: слабљење протеинске мреже (0,67±0,05), желатинизација скроба (1,69±0,10), амилазна активност (2,34±0,15) и ретроградација скроба (3,38±0,27) а добри алвеографски показатељи су: отпор на рас-тезање (49±2), растегљивост теста (77±1) и рад деформације (121±3). Ови показатељи су у корелацији са текстуром тестенине 5 (тврдоћа 5182,06±7,48 g, лепљивост 10,25± ±1,16 gsec, жилавост 410,33±5,44 gsec, сила лома 24,38±1,84 g, савитљивост 27,02±1,69 mm). PCA и scatter plot су се показали корисним за детекцију структуре међузависности између мерених параметара квалитета брашна и квалитета тестенине. Кључне речи: спелта, миксолабски показатељи, алвеографски показатељи, тексту-

ра, боја

Received: 25 May 2017. Accepted: 28 September 2017.

APTEFF, 48, 1-323 (2017) UDC: 544.773.33:547.426.1:665.12 https://doi.org/10.2298/APT1748095F BIBLID: 1450-7188 (2017) 48, 95-107


95

PROPERTIES OF WATER IN OIL EMULSIONS (W/O) STABILIZED WITH MIXTURES OF PGPR AND POLYGLYCEROL FATTY ACID ESTERS

Jadranka L. Fraj*, Lidija B. Petrović, Jelena R. Milinković, Jaroslav M. Katona,

Sandra Đ. Bučko, Ljiljana M. Spasojević


Water in oil (W/O) emulsions are dispersed systems which have very wide application as a carriers in the food, pharmaceutical and cosmetic industry products. The main problem with practical application of such systems is their low stability. The emulsifiers used to stabilize this type of emulsions are with low hydrophilic-lipophilic balance va-lues. The present work examines the possibility of the application of mixtures of lipophilic emulsifiers polyglycerol polyricinoleate (PGPR) and polyglycerol fatty acid esters for stabilization of W/O emulsions. First of all, the adsorption properties of the used emul-sifiers were examined by tensiometic measurements. Based on these results, two emulsi-fiers PGPR and decaglycerol decaoleate (Caprol 10G10O) were selected for the pre-paration of the emulsions, as well as their mass ratios and total concentrations. The re-sults of the investigation of the emulsions properties (dispersion analysis and sedimen-tation stability) showed that more stable emulsions can be obtained by decreasing the Caprol 10G10O mass ratio and increasing the total concentration of emulsifiers.

KEY WORDS: lipophilic emulsifiers, PGPR, polyglycerol fatty acid esters, W/O emulsions

INTRODUCTION

There are many products in the pharmaceutical, cosmetic and food industry that con-tain W/O emulsions. The most commonly utilized products in the food industry that are in the form of W/O emulsions are butter, margarine, spreads and shortenings (1, 2). W/O emulsions can also be used to encapsulate and control the release of hydrophilic active ingredients such as water-soluble vitamins, flavors, colors, proteins (3-5). One of their applications is also as the inner (primary) emulsions in double W/O/W emulsions. The stability mechanism of these emulsions differs from the oil in water (O/W) emul-sions, which can be stabilized by both, steric and electrostatic repulsion. In the case of W/O emulsions, only steric forces are expected to stabilize the emulsion, because of the low electrical conductivity of the continuous phase (6). The main problem with practical application of such systems is generally their low stability, because of the high mobility of water droplets, which can easily sediment, flocculate or coalescence. A better under-standing of the interactions between water, oil and emulsifier at the interfaces would

* Corresponding author: Jadranka Fraj, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar Cara

Lazara 1, 21000 Novi Sad, Serbia, e-mail: [email protected]



96

allow producing stable W/O emulsions and, therefore, encourage the development of new products and applications (7). The emulsifiers used to prepare W/O emulsions have low hydrophilic-lipophilic balance (HLB) values. The development and application of food grade emulsifiers have followed the advancement in the food industry, which has increased tremendously over the last few decades. Numerous types of oil soluble emulsifiers have been used either individually or in combination to facilitate the formation and stability of W/O emulsions. Among these, polyglycerol esters of fatty acids (PGEs) are used in a number of food products. Commercial PGEs may vary considerably in composition owing to differences in the extent in polymerization of the glycerol, in the nature of the hydrophilic part, and in the degree of esterification (8). Considering the wide interval of HLB values, they may be used as a stabilizer of W/O and O/W emulsions (9). In the European Union regulation for food additives, PGEs are given the reference number E475. Polyglycerol polyricinoleate (PGPR) has been reported as the most effective emulsifier for the stabilization of W/O emulsions (10). It is oligomeric, nonionic emulsifier produced by the esterification of castor oil fatty acids with polyglycerol. It is commonly used emulsifier in chocolate manufacture, since it has been shown to inhibit thickening of chocolate due to its excellent water binding properties (11). PGPR is given the reference number E476. The aim of the present work was to investigate the possibility of formation stable W/O emulsions stabilized with the mixtures of PGPR and polyglycerol fatty acids esters. First of all, the adsorption properties of the surfactants and their mixtures were investi-gated. After this, stability parameters, such as droplet size and droplet size distribution and sedimentation index of emulsions stabilized with selected mixtures of surfactants, were measured.

EXPERIMENTAL

Materials The following oil soluble surfactants were used: polyglycerol polyricinoleate (PGPR) donated by Jaffa Crvenka, Serbia; decaglycerol decaoleate (Caprol 10G10O), triglycerol monostearate (Caprol 3GS), hexaglycerol octastearate (Caprol ET), all from Abitec, USA, whereas the oil phase was medium chain triglycerides (MCT) of caprylic/capric faty acids (Saboderm TCC) produced by Sabo Spa, Italy. Deionized water was used as aqueous phase.

Methods

Preparation of solutions. The stock solutions of lipophilic emulsifiers were prepared by the dissolution of appropriate amount of emulsifier in the oil phase. Solutions of lower concentrations were obtained by diluting the stock solutions. Binary mixtures of the emulsifiers PGPR and Caprol 10G10O and PGPR and Caprol 3GS, at different mass ratios (1:1; 1:2; 1:5 and 1:10), were prepared by mixing the stock solutions and adding a



97

desired amount of the oil phase, to obtain the concentrations of the emulsifiers of 1; 2 and 3% (w/v). The HLB values of these mixtures were calculated as follows:

HLBmix=xA·HLBA+ 1-xA ·HLBB where HLBA and HLBB are HLB are the values of emulsifiers A and B, and xA is the mass fraction of emulsifier A in the mixture. The calculated values are presented in Table 1.

Table 1. HLB values of the binary mixtures of lipophilic emulsifiers PGPR and Caprol 10G10O and PGPR and Caprol 3GS

PGPR:Caprol 3GS HLBmix PGPR:Caprol 10G10O HLBmix

1:1 4.75 1:1 3.25

1:2 5.28 1:2 3.30

1:5 5.91 1:5 3.42

1:10 6.19 1:10 3.46

Tensiometric measurements. Measurements of the interfacial tension between water and oil with various surfactant concentrations were carried out on a Sigma 703D ten-siomater (KSV Instruments, Finland), using the Du Noüy ring method (12). Prior the measurements, the ring was immersed in water phase, then the oil phase slowly aded on the top, and the interface was left for 15 min to equilibrate. The interfacial tension was measured at the point where the ring broke away from the intefacial layer between two phases. In all experiments the temperature was kept at 40 °C. The reported values of the interfacial tension were average of three measureents, at least. Preparation of W/O emulsions. The emulsions were prepared at the water-oil mass ratio 20:80, while the aqueous phase was deionized water and continuous phases were 1; 2 or 3% (w/v) solutions of the binary mixtures of lipophilic emulsifiers, at the mentioned mass ratios, in MCT. The emulsions were prepared by dispersing water phase in the continuous phase at 40 °C by means of an Ultra Turrax T25 homogeniser (Ika, Germany) at 20,000 rpm during 10 min. Emulsion stability test. For stability test, the emulsions were transferred into 10 ml graduated cylinders and stored at room temperature for 14 days. The emulsions were observed for the changes in homogenity and phase separation during storage. The oil phase separation was visually monitored at certain time intervals. The total height of the emulsion, HE, and the height of the oil layer, HO, were measured. The extent of the phase separation was characterized by the sedimentation index, H, given as:

H= HOHE

·100 %

A higher value of the sedimentation index indicates a worse emulsion stability.



98

Droplet size and polydispersity index analysis Particle size analysis of the emulsions was performed by dynamic light scattering (DLS) measurements using a Malvern Zetasizer Nano ZS (Malvern Instruments, UK) at 25 °C. Prior to the measuremens, all samples were diluted with oil, in order to yield a suitable scattering intensity. The refractive index of MCT was 1.448. The DLS data were analyzed using the general purpose mode, thus the hydrodynamic diameter (z-average) was obtained.

Statistical analysis Paired t test at the 95% confidence level was performed to compare the difference between two measurements, and for this purpose Origin Pro 8.0 software was used.

RESULTS AND DISSCUSION

Interfacial tension measurements Oil and water do not mix due to the existence of the interfacial tension, which is generally about 30-35 mN/m. Addition of the surface active molecules to the system reduces interfacial tension to 1-5 mN/m, enabling dispersion one phase into the other, i.e. formation of emulsion. Except of lowering the interfacial tesion, the adsorbed surface active molecules form steric and electroststic barier between the dispersed phase drops, preventing their coalescence. In this work we investigated first the adsorption properties of the selected lipophilic emulsifiers. The lipophilic emulsifiers, PGPR, Caprol 10G10O, Caprol ET and Caprol 3GS, differ by the fraction of glycerol and fatty acids in the molecule, as well as by the type of fatty acids. The adsorption properties were investigated by measuring the changes in the interfacial tension with the concentration of the emulsifiers in pure solutions and in their binary mixtures, which is presented in Figure 1.

As it can be observed, the interfacial tension for all investigated systems decreased with an increase in the emulsifiers concentration. Comparing the total lowering of the interfacial tension by different emulsifiers at the same concentration, it can be noticed that polyglycerol fatty acid esters significantly lower the interfacial tension by increasing their HLB values. The increase in the HLB number is due to an increase in the fraction of the hydrophilic part in the emulsifier molecule, thereby increasing its affinity to the interface at which it is oriented so that the hydrophilic part is displaced into the aqueous phase (13). On the other hand, PGPR induces higher reduction of the interfacial tension compared to Caprol 10G10O, although it has a lower HLB number. This is a consequen-ce of the branched hydrophobic chain of the PGPR molecule, which allows formation of a compact film at the interface (14, 15). Caprol ET did not show a significant effect on the interfacial tension, which is expected due to the low HLB number (2.5) and its lipo-philic properties, i.e. extremely good solubility in the oil phase. Based on these results,



99

the emulsifiers which showed significant influence on the interfacial tension, i.e. PGPR, Caprol 10G10O and Caprol 3GS, were chosen for further investigations.

Figure 1. Changes of the interfacial tension at the O/W interface with emulsifier concentration in the oil at 40 °C.

It is known that the emulsifiers in the mixtures show different behavior compared to the pure solutions (16). In some cases, the emulsifiers in binary mixtures show a synergistic effect on the interfacial tension reduction. However, it is not always the case, because sometimes the presence of different emulsifier molecules in the system can lead to their worse packing at the adsorption layer, which depends on their properties and interactions. Although the PGPR has a very wide application, especially in the food industry, there are increasing tendencies for reduction its concentration in some products. On the other hand, comparing the maximum allowed daily intakes (MADI) for PGPR (7.5 mg/kg bodyweight) and emulsifiers from Caprol group (25 mg/kg bodyweight) it can be seen that MADI for Caprol emulsifiers is about three times higher. For that reason, in the next experiments, the influence of the increasing mass ratio of Caprol 3GS and Caprol 10G10O in the mixture with PGPR, on interfacial tension was investigated. However, in the mixtures of PGPR and Caprol 3GS with a mass ratio of Caprol 3GS higher that 0.5, a solid film of the surfactant was formed at the interface, so the measurement of interfacial tension was not possible. For this reason, the investigations included only binary mixtures of PGPR and Caprol 10G10O (Figure 2).



100

Figure 2. Changes of the interfacial tension at the O/W interface with increasing Caprol 10G10O mass ratio in the binary mixture with PGPR, at total emulsifier concentration of

1% (w/v) at 40 °C. As it can be seen in Figure 2, the increase in the Caprol 10G10O mass ratio in the mixture with PGPR induces an increase in the interfacial tension at O/W interface. Although the mixtures with higher mass ratio of Caprol 10G10O have higher HLB values (Table 1), its worse emulsification characteristics are dominant here. However, the reduction of interfacial tension to a value 5.53 mN/m can be obtained with binary mixture of these two emulsifiers at their mass ratio of 1:1. Besides the emulsifiers mass ratio in the mixture, the total concentration influences the interfacial tension and stability of the adsorption layer,. For that reason, the influence of the total emulsifiers concentration, in the binary mixture of Caprol 10G10O and PGPR at their mass ratio 1:1, on interfacial tension was investigated. The results are presented on Figure 3 and Table 2, showing also the changes in the interfacial tension with the concentrations of single emulsifier solutions. It is clear from Figure 3 and Table 2 that, as in the case of the solutions with individu-al emulsifiers, in their binary mixtures an increase in the total emulsifier concentration leads to lowering of the interfacial tension. On the other hand, comparing the single emulsifier solutions with their binary mixtures, it can be noticed that the emulsifiers mix-tures, at concentrations higher than 0.1% (w/v), showed a greater reduction of the inter-facial tension, which is a consequence of a synergistic effect of the emulsifiers molecules in the mixture. The lowering of the interfacial tension by the mixture of emulsifiers in comparison with the pure emulsifier solutions were statistically significant (P<0.05). Based on these results, it is clear that a minimal emulsifier concentration required to achieve the necessary reduction of interfacial tension is 1% (w/v).



101

Figure 3. Changes of the interfacial tension at O/W interface with the emulsifier concentration in oil solutions of PGPR, Caprol 10G10O and their binary mixtures at a

mass ratio 1:1, at 40 °C.

Table 2. Changes of the interfacial tension standard deviation of PGPR, Caprol 10G10O and PGPR:Caprol 10G10O=1:1 mixture with the emulsifier concentration

c (%, w/v) (mN/m) SDa

PGPR Caprol 10G10O PGPR:Caprol 10G10O=1:1

0.0001 18.110.085 18.60.087 16.240.108

0.001 14.590.056 16.110.051 14.890.090

0.01 14.470.064 15.30.100 15.190.140

0.1 9.890.076 13.490.056 13.730.089

1 4.90.080 10.940.040 4.040.147

10 1.690.036 3.790.021 1.170.076 a

- mean standard deviation



102

Properties of the emulsions stabilized with PGPR and Caprol 10G10O mixtures Previous measurements showed good surface activity of the lipophilic emulsifiers PGPR and Caprol 10G10O, as well as a synergistic effect of their molecules on surface tension reduction at the O/W interface. However, the low value of interfacial tension is not the only precondition for the formation of a stable dispersed system. For this reason, the possibility of obtaining stable 20% W/O emulsions, stabilized with these mixtures of emulsifiers, was further investigated. The dispersed phase droplets size is the parameter which shows a high influence on the emulsions stability. A lower diameter of the dispersed phase droplets and higher uniformity of droplet size distribution increase the sedimentation stability of the system, due to the reduced tendency toward coalescence and Ostwald’s ripening. For this reason, in the next set of experiments the properties of the 20% W/O emulsions, stabilized with PGPR:Caprol 10G10O mixtures at their mass ratios: 1:0; 1:1; 1:2; 1:5; 1:10; 0:1, were investigated. The total concentrations of emulsifiers were 1, 2, and 3% (w/v).

By microscopic examination of the obtained emulsions, very small droplets were observed, so the method of light scattering on a Zetasizer Nano ZS apparatus was used for the dispersion analysis. The changes of droplets mean diameter with increasing Caprol 10G10O mass ratio in the mixtures with PGPR at different total emulsifier concentrations, are presented in Figure 4.

Figure 4. Changes of the droplets mean diameter of 20% W/O nanoemulsions with increasing Caprol 10G10O mass ratio in the mixtures with PGPR, at total

emulsifiers concentrations of 1, 2, and 3% (w/v).



103

As it is shown in Figure 4, all emulsion samples have droplets with a mean diameter lower than 280 nm, i.e. the obtained systems were nanoemulsions. Also, the increase in the Caprol 10G10O mass ratio induced an increase in the in droplets mean diameter. This is in agreement with the tensiometric measurements, which showed a worse surface activity of Caprol 10G10O molecules. Also, the increase in the total emulsifiers con-centration induced a reduction of the droplet mean diameters. The dispersion analysis could not be performed on the emulsion samples stabilized with 1 and 2% (w/v) of pure Caprol 10G10O because of their pronounced instability. Namely, these emulsions broke down right after the preparation. A very important characteristic of the emulsions from their application aspect, is their sedimentation stability. Sedimentation instability is a consequence of the density diffe-rence between the dispersed and the continuous phase. Namely, due to the effect of gra-vity force phase separation occurs, and if dispersed phase has lower density their droplets goes to the top, forming a creamy layer, i.e., if the dispersed phase has a higher density, its droplets go down and form a sediment layer (17, 18). In this sense, the stability of the investigated nanoemulsions is expressed as a change of the sedimentation index (H) over time, as shown in Figure 5. Stability was monitored for 14 days at room temperature, and the appearance of the emulsions after this time period was also shown in Figure 5. In all the nanoemulsions stabilized with PGPR and Caprol 10G10O mixtures, at a total emulsifiers concentration of 1% (w/v), separation was observed just in the first few hours after the preparation. The sedimentation index increased gradually during the first five days, but after this period no significant change of this parameter was noticed (Figure 5A). In the nanoemulsions stabilized with the mixtures of PGPR and Caprol 10G10O, at their mass ratios 1:0 and 1:1 and with total emulsifiers concentration of 2% (w/v) (Figure 5B), separation occurred two days after the preparation, while in the nanoemulsions with higher Caprol 10G10O mass ratio separation occurred just at the first few hours after preparation. A further increase in total emulsifiers concentration to 3% (w/v) (Figure 5C) led to more pronounced differences in the separation rate. Namely, at this concentration, in the nanoemulsions stabilized with PGPR and Caprol 10G10O at mass ratios of 1:0; 1:1 and 1:2 separation occurred seven days after the preparation, with a significantly lower value of sedimentation index (less than 10%), indicating their high sedimentation stability. The nanoemulsions with higher Caprol 10G10O ratios separated just at the first few hours after the preparation. In view of the fact that the separation of emulsion is a consequence of sedimentation instability, i.e. spreading of the emulsion droplets by size, the values of sedimentation index are in correlation with the results of dispersion analysis. Figure 5 it clearly shows that in the nanoemulsions stabilized with PGPG and Caprol 10G10O mixtures at the emulsifiers mass ratio 1:10 and their total concentration of 1 and 2% (w/v), as well as in the nanoemulsion stabilized with pure 3% (w/v) Caprol 10G10O, complete separation of a portion of the aqueous phase occurred at the bottom of the cylinder, which is a consequence of advanced coalescence and partial and/or complete break up of the emulsion, indicating a poor emulsifying properties of Caprol 10G10O molecules (18, 19).



104

Figure 5. Changes of the sedimentation index with time and appearance of the 20% W/O nanoemulsions stabilized with PGPR and Caprol 10G10O mixtures at different

mass ratios after 14 days of storage at room temperature. Total emulsifiers concentration was 1% (A); 2% (B) and 3% (w/v) (C).

CONCLUSION

The investigation of the interfacial behavior of the lipophilic emulsifiers PGPR, Caprol 3GS, Caprol 10G10O and Caprol ET showed that the emulsifiers from the Caprol group significantly lowered the interfacial tension with an increase in their HLB value, due to a higher hydrophilicity of the molecule and more affinity to the interface. PGPR



105

induced a higher reduction of the interfacial tension compared to Caprol 10G10O, although it has a lower HLB number, which is a consequence of the branched hydro-phobic chain of the PGPR molecule. The investigation of the interfacial behavior of the binary mixtures of PGPR and Caprol 10G10O showed a synergistic effect of their molecules on the interfacial tension at concentrations higher than 0.1% (w/v). The freshly prepared 20% W/O emulsions stabilized with PGPR and Caprol 10G10O mixtures, at their different mass ratios, had a submicron droplet diameter in the interval of 130-280 nm, so the obtained systems were nanoemulsions. An increase in the mass ratio of Caprol 10G10O induced an increase in the droplet diameter. This was confirmed also by the investigation of sedimentation stability of emulsions. An increase in the total emulsifiers concentration showed a positive effect on the emulsions stability, so the most stable emulsions were obtained at the PGPR:Caprol 10G10O mass ratios of 1:0; 1:1 and 1:2, and their total concentration of 3% (w/v).

Acknowledgement This work was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia, Project Number III46010.

REFERENCES

1. Young, N.; Wassell, P. Margarines and spreads. In Food Emulsifiers and their Applications; Hasenhuettl, G.L., Hartel R.W., Eds.; Springer: New York, 2008; pp 307-326.

2. Charcosset, C. Preparations of emulsions and particles by membrane emulsification for the food processing industry. J. Food Eng. 2009, 92, 241-249.

3. Frelichowska, J.; Bolzinger, M.; Valour, J.; Mouaziz, H.; Pelletier, J.; Chevalier, Y. Pickering w/o emulsions: drug release and topical delivery. Int. J. Pharm. 2009, 368, 7-15.

4. McClements, D.J.; Decker, E.A.; Park, Y.; Weiss, J. Structural design principles for delivery of bioactive components in nutraceuticals and functional foods. Crit. Rev. Food Sci. Nutr. 2009, 49, 577-606.

5. Zhu, Q.; Wu, F.; Saito, M.; Tatsumi, E.; Yin, L. Effect of magnesium salt con-centration in water-in-oil emulsions on the physical properties and microstructure of tofu. Food Chem. 2016, 201, 197-204.

6. Claesson, P.M.; Blomberg, E.; Poptoshev, E. Surfaces forces and emulsion stability. In Food Emulsions (4th ed.); Friberg S.E., Larsson K., Sjöblom J., Eds.; Marcel Dekker: New York, 2004; pp 257-298.

7. Ushikubo, F.Y.; Cunha, R.L. Stability mechanisms of liquid water-in-oil emulsions. Food Hydrocolloids. 2014, 34, 145-153.

8. Krog, N.J.; Sparso, F.V. Food emulsifiers and their chemical and physical properties. In Food Emulsions; Friberg S.E., Larsson K., Sjöblom J., Eds.; Marcel Dekker: New York, 2004; pp 45-91.



106

9. Stauffer C.E., Emulsifiers for food industry. In Bailey’s Industrial Oil and Fat Products; Shahidi F., Ed.; John Wiley & Sons: New York, 2005.

10. Surh, J.; Vladisavljevic, G.T.; Mun, S.; McClements, D.J. Preparation and charac-terization of water/oil and water/oil/water emulsions containing biopolymer-gelled water droplets. J. Agric. Food Chem. 2007, 55, 175-184.

11. Beckett, S.T. Industrial chocolate manufacture and use; Blackwell Science: Oxford, 1999.

12. Sovilj, V.; Milanović, J.; Petrović, L. Viscosimetric and tensiometric investigations of interactions between gelatin and surface active molecules of various structures. Food Hydrocolloids. 2013, 32, 20-27

13. Gao, B.; Sharma, M.M. A familiy of alkyl sulfate gemini surfactants. 2. Water-oil interfacial tension reduction. J. Colloid Interface Sci. 2013, 407, 375-381.

14. Varadaraj, R.; Bock, J.; Zushma, S.; Brons, N.; Colletti, T. Effect of hydrocarbon chain branching on interfacial properties of monodisperse ethoxylated alcohol surfactants. J. Colloid Interface Sci. 1991, 147, 387-95.

15. Rosen, M.J. Surfactants and Interfacial Phenomena; John Wiley & Sons, Inc.: New Jersey, 2004.

16. Djaković, Lj. Koloidna hemija; Zavod za udžbenike i nastavna sredstva: Beograd, 1990.

17. Robins, M.M. Emulsions-creaming phenomena. Curr. Opin. Colloid Interface Sci. 2000, 5, 265-272.

18. Tadros, T.F. Emulsions formation, stability and rheology. In Emulsion Formation and Stability; Tadros T.F., Ed.; Wiley-VCH Verlag GmbH & Co.: New Jersey, 2013.

19. Kabalnov, A. Thermodinamic and theoretical aspects of emulsions and their stability, Curr. Opin. Colloid Interface Sci. 1998, 3, 270-275.

ОСОБИНЕ ЕМУЛЗИЈА ВОДА У УЉУ (В/У) СТАБИЛИЗОВАНИХ СМЕШОМ ПГПР-a И ПОЛИГЛИЦЕРОЛ ЕСТАРА МАСНИХ КИСЕЛИНА

Јадранка Л. Фрај*, Лидија Б. Петровић, Јелена Р. Милинковић, Јарослав М.

Катона, Сандра Ђ. Бучко, Љиљана М. Спасојевић

Универзитет у Новом Саду, Технолошки факултет Нови Сад, Булевар Цара Лазара 1, 21000 Нови Сад

Емулзије типа воде у уљу (В/У) су дисперзни системи са веома широком применом у прехрамбеној, фармацеутској и козметичкој индустрији, као носачи активних метерија. Основни проблем код практичне примене ових система је њихова нестабилност. Емулгатори који се користе за стабилизацију овог типа емулзија су емулгатори са ниским ХЛБ бројем. У овом раду испитана је могућност примене липофилних емулгатора полиглицерол полирицинолата (ПГПР-а) и полиглицерол естара масних киселина, за стабилизацију емулзија В/У. Најпре су испитане адсорпционе особине коришћених емулгатора применом тензиометрије. На основу тих резултата одабрана су два емулгатора, ПГПР и декаглицерол дека-



107

олеат (Caprol 10G10O), за припрему емулзија, при њиховим различитим масеним односима и укупним концентрацијама. Резултати испитивања особина емулзија (дисперзна анализа и седиментациона стабилност) су показали да се стабилније емулзије добијају смањењем удела Caprol-а 10G10O у смеши, као и повећањем укупне концентрације емулгатора. Kључне речи: липофилни емулгатори, ПГПР, полиглицерол естри масних кисе-

лина, адсорпционе особине, В/У емулзије


APTEFF, 48, 1-323 (2017) UDC: 582.282.23+663.88]:664.1+663.14.031.4+661.722 https://doi.org/10.2298/APT1748109I BIBLID: 1450-7188 (2017) 48, 109-116


109

CONTENT OF SUGAR, ORGANIC ACIDS AND ETHANOL IN FERMENTED MILK BEVERAGES OBTAINED WITH DIFFERENT TYPES OF KOMBUCHA

INOCULUM

Mirela D. Iličić*, Spasenija D. Milanović, Katarina G. Kanurić, Vladimir R. Vukić, Svetlana S. Popović, Dajana V. Vukić


The aim of this study was to examine the influence of different types and concentra-tion of kombucha inoculum on content of sugar, organic acids and ethanol in the fermented beverages produced from milk of 0.9% fat content. Three different kombucha inoculums, cultivated on black tea with addition of sucrose: standard inoculum - 10% (w/w) and 15% (w/w), concentrated by microfiltration- 10% (w/w) and 15% (w/w), and concentrated by evaporation - 1.5% (w/w) and 3.0% (w/w), were applied in the manufac-ture of fermented milk. Contents of lactose, galactose, glucose, fructose, organic acids, and ethanol in the kombuha fermented milk beverages were determined by the enzyme tests. It was found that the lactose content varied from 3.30 to 4.0 g/100g. All samples showed higher content glucose than fructose. The content of L-lactic acid in the samples ranged from 0.4 to 0.7 g/100g, while significantly lower level of D-lactic and acetic acid were determined in all samples of kombucha fermented milk (<0.06g/100g). KEY WORDS: fermented milk, kombucha inoculum type, sugar, organic acids, ethanol

INTRODUCTION Kombucha inoculum is a consortium of yeasts, acetic and lactic acid bacteria which are involved in such metabolic activities that utilize substrate in different ways (1). Yeasts hydrolyze sucrose into glucose and fructose by invertase and produce ethanol via glycolysis. Acetic acid bacteria make use of glucose to produce gluconic acid and etha-nol, to produce acetic acid (2-4). Also, kombucha inoculum can be used for lactose fer-mentation. The main primary products of lactose hydrolysis are galactose and glucose, while main secondary products appeared to be lactic acid, along with a much lower con-tent of acetic acid (5). Belloso Morales and Hernandez-Sanchez investigated the manu-facture of beverage from three types of cheese whey (fresh sweet, fresh acid and reconsti-tuted sweet) using a kombucha culture as inoculum (6). They found that the sweet whey fermented beverages contained a relatively low lactose concentration (<12g/L). The pro-duction of lactic acid in three types of whey increased, and reached a maximum value of * Corresponding author: Mirela D. Iličić, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




110

5.8g/L after 96 h (acid whey). The final ethanol content was low (<5g/L) in all fermented whey’s. The acetic acid, which was converted from ethanol, increased steadily, reaching a maximum of 0.28g/L in the fresh sweet whey after 48 h of fermentation, 1g/L in the acid whey after 96 h and 0.6g/L in the reconstituted sweet whey after 96 h of fermenta-tion. The same authors concluded that the best results were obtained with the reconsti-tuted sweet whey. Besides, a possible substrate for the kombucha fermentation, as a non-conventional starter culture, is milk. The quality of kombucha fermented milks has been reported by several authors (7-17). It was found that nutritive, rheological, and sensory characteristics of fermented milks obtained by kombucha inoculum depend on the milk composition, concentration of kombucha, tea type, and process temperature. The aim of this study was to investigate the effect of the types and concentration of kombucha inoculum on the following characteristics - content of carbohydrate, organic acids and ethanol in the fermented milk beverages.

EXPERIMENTAL Homogenized and pasteurized cows' milk of 0.9 g/100g of fat content (AD Imlek, Beograd, Novosadska mlekara - division Novi Sad) was used for the production of kom-bucha fermented milk beverages. The following starter cultures were applied for fermentation: a) standard kombucha inoculum cultivated on a black tea (Camellia sinensis L. - oxidized, 1.5 g/L) with saccharose concentration of 70 g/L. The tea was cooled to room temperature, after which inoculum from a previous fermentation was added in concen-tration of 100 mL/L. The incubation was performed at 29 °C for 7 days. The obtained kombucha tea was used as a non-conventional starter culture and amounts of 10 and 15% (v/v) were added to the milk. b) microfiltered kombucha inoculum - prepared from standard kombucha inoculum (from previous kombucha fermentation). Ceramic microfiltration membrane of 200 nm pore size (Pall, Germany) was used to concentrate the standard inoculum. The feed flow rate was 1L/min and transmembrane pressure was 30 kPa, and the degree of concentra-tion of the kombucha inoculum was 2.3. Amounts of 10% (v/v) and 15% (v/v) of the microfiltered kombucha inoculum were applied to the milk. c) evaporated kombucha inoculum - standard inoculum was exposed to vacuum-eva-poration, so that 60 mL of concentrate was obtained from 300 mL of the original/standard solution. The amounts of 1.5 and 3.0% (v/v) of kombucha inoculum concentrates were added to the milk. Six fermented beverages were manufactured from the milk inoculated at 42 °C. The fermentation lasted 9 - 12 h, to reach pH 4.5. The obtained milk gels were cooled to 8 °C, homogenized by mixing, and packed in plastic containers. Each fermentation were repea-ted three times. The plan of the experiments is presented in Table 1.



111

Table 1. Plan of the experiments - fermented beverages

No. Fermented beverage Kombucha inoculum

concentration (%) Kombucha inoculum

type 1 10I 10 Standard 2 15I 15 Standard 3 10MFI 10 Microfiltered 4 15MFI 15 Microfiltered 5 1.5EI 1.5 Evaporated 6 3.0EI 3.0 Evaporated

Contents of lactose and of primary components (galactose, glucose, fructose), and se-condary products of fermentation (L- and D-lactic acid, acetic acid and ethanol), were de-tected in the kombucha fermented milk beverages using enzyme-tests, such as K-LACGAR 12/05 (lactose and D-galactose), K-FRUGL (D- glucose and D-fructose), K-DLATE 11/05 (L-lactic acid and D- lactic acid), K-ACET 11/05 (acetic acid) and K-ETOH 03/06 (ethanol). The analysis was performed according to the guidelines supplied by the producer (Megazyme, Ireland) (19). The reaction products were monitored on a T80+UV/VIS Spectrometer (PG instruments Ltd). The data were expressed as means and standard deviation, using the software program STATISTICA, version 6 (2001) (StatSoft Inc, Tulsa, OK, USA) (20).


Content of sugar Contents of lactose content and main primary products of fermentation were presen-ted in Figure 1. The level of lactose in the kombucha fermented milk beverage varied from 3.30 (15MFI) to 4.0 (1.5EI) g/100g. It was 16% to 30% less than in the milk used for the beverage manufacture (Figure 1a). These results are in accordance with the litera-ture data (10, 11, 17). The highest average quantity of galactose (0.550 g/100g) were found in the samples 10MFI and 15 MFI. Contents of glucose and fructose in the kombucha fermented milk beverages are pre-sented in Figure 1c. It is evident that all samples showed higher content of glucose than fructose. These results are in accordance with the literature data, where different authors (2, 3, 5) concluded that yeasts hydrolyse sucrose into glucose and fructose with a prefe-rence for fructose as a substrate. The highest contents of glucose of 0.24 and 0.38 g/100g, were measured in the sample 1.5EI and 3.0EI, respectively. The different ratio of micro-organisms in the kombucha inoculums significantly influenced contents of glucose and fructose in the samples (17, 21). Radulović et al. (22) found that the number of yeasts in the MFI inoculum (8x105 cfu/mL) was significantly higher compared to the number of acetic acid bacteria (2x103 cfu/mL).



112

a) b)

c)

Figure 1. Contents of sugars in the kombucha fermented milk beverages: a) lactose;

b) galactose; c) glucose and fructose

Contents of organic acids and ethanol The main secondary products of milk fermentation by kombucha inoculums appeared to be L- and D- lactic acid, along with smaller amounts of acetic acid (Figure 2). The L-lactic acid content in the samples ranged from 0.4 to 0.7 g/100g, while significantly lower level of D-lactic and acetic acid were measured in all kombucha fermented milk bevera-ges (<0.06 g/100g). These results are in accordance with the literature data (11, 12). Iličić et al. (2012) also found that the main secondary products in kombucha milk based beve-rages were lactic acid (ranging from 0.42 to 0.82 %, w/w) along with acetic acid amount of which is much smaller (< 0.05 %, w/w). A minimal quantity of ethanol was determined in the beverages produced by using evaporated kombucha inoculum (Figure 3): the average value of ethanol in the samples varied from 0.009 (1.5EI and 3.0EI) to 0.055 g/100g (10MFI).



113

Figure 2. Lactic and acetic acid content in the kombucha fermented milk beverages

Figure 3. Ethanol content in the kombucha fermented milk beverages

CONCLUSION Three different kombucha inoculums: standard inoculum, concentrated by microfiltra-tion and concentrated by evaporation were applied in fermented milk beverage manu-facture. Type of inoculum significantly influences the characteristics of the obtained me-tabolites. The results showed that decreasing of lactose content affected the increased ga-lactose in kombucha fermented milk beverages. Glucose content in all samples is higher than fructose content. The fermented milk beverages predominantly contains L-lactic acid, while the level of D-lactic, acetic acid, and ethanol were detected in low con-centrations.



114

Acknowledgement Authors want to thank the Ministry of Education, Science and Technological Deve-lopment of the Republic of Serbia for the financial support of the research presented in this article, Project No. 46009.

REFERENCES

1. Marsh, A.J.; Sullivan, O. O; Hill, C.; Ross, R.P.; Cotter P.D. Sequence-based analysis of the bacterial and fungal compositions of multiple kombucha (tea fungus) samples. Food Microbiol. 2014, 38, 171-178.

2. Jayabalan, R.; Malbaša, R.; Lončar, E.; Vitas, J.; Muthuswamy, S. A review on kombucha tea-microbiology, composition, fermentation, benefitial effects, toxicity and tea fungus. Comprehen. Rev. Food Sci. Food Saf. 2014, 13, 538-550.

3. Lončar, E.; Kanurić, K.; Malbaša, R; Ðurić, M.; Milanović, S. Kinetics of saccharose fermentation by kombucha. Chem. Ind. Chem. Eng.Q. 2014, 3, 345-352.

4. Markov, S.; Malbaša, R.; Hauk, M.; Cvetković, D. Investigation of tea fungus microbe associations. I. The yeast. APTEFF. 2001, 32, 133-138.

5. Iličić, M.; Kanurić, K.; Milanović, S.; Lončar, E.; Djurić, M.; Malbaša, R. Lactose fermentation by kombucha - a process to obtain new milk-based beverages. Rom. Biotech. Lett., 2012, 17 (1), 7013-7021.

6. Belloso-Morales, G.; Hernandez-Sanches, H. Manufacture of a beverage from cheese whey using a tea fungus fermentation. Revista Latinoamericana Microbiologia, 2003, 45 (1-2), 5-11.

7. Мilаnоvić, S.; Cаrić, М.; Lоnčаr, Е.; Pаnić, М.; Маlbаšа, R.; Dоbrić, D. Primеnа kоncеntrаtа čајnе glјivе u prоizvоdnji fеrmеntisаnih mlеčnih nаpitаkа. Prеhrаmbеnа industriја-Мlеkо i mlеčni prоizvоdi, 2002, 13, 8-13.

8. Milanović, S.; Lončar, E.; Đurić, M.; Malbaša, R.; Tekić, M.; Iličić, M.; Duraković, K. Low energy kombucha fermented milk-based beverages, APTEFF, 2008, 39, 37-46.

9. Milanović, S.; Kanurić, K.; Vukić, V.; Hrnjez, D.; Iličić, M.; Ranogajec, M.; Milano-vić, M.: Physicochemical and textural properties of kombucha fermented dairy pro-ducts. Afr. J. Biotechnol. 2012, 11 (9), 2320-2327.

10. Iličić, М.; Мilаnоvić, S.; Cаrić, М.; Đurić, М.; Теkić, М.; Vukić, V.; Durаkоvić, K.; Pоpоvić, S. Primеnа kоmbuhе u tеhnоlоgiјi funkciоnаlnih fеrmеntisаnih mlеčnih prо-izvоdа. Prehrambena industrija - Mlekoi mlečni proizvodi/Food Industry-Milk and Dairy Products. 2009, 20, 65-69.

11. Iličić, M.; Milanović, S.; Carić, M.; Vukić, V.;Kanurić, K.; Ranogajec, M.; Hrnjez, D. The effect of transglutaminase on rheology and texture of fermented milk products. J. Texture Stud. 2013, 44, 160-168.

12. Malbaša, R.; Milanović, S.; Lončar, E.; Djurić, M.; Carić, M.; Iličić, M.; Kolarov, Lj. Milk-based beverages obtained by Kombucha application, Food Chem. 2009, 112, 178-184.



115

13. Malbaša, R.; Vitas, J.; Lončar, E.; Grahovac, J.; Milanović, S. Optimisation of the antioxidant activity of kombucha fermented milk products. Czech J. Food Sci. 2014, 32 (5), 477-484.

14. Kanurić, K.; Hrnjez, D.; Ranogajec, M.;Milanović, S.;Iličić, M.; Vukić, V.; Milano-vić, M. The effect of fermentation temperature on the functional dairy product quality. APTEFF. 2011, 42, 63-70.

15. Vitas, J.S.; Malbaša, R.V.; Grahovac, J.A.; Lončar, E.S. The antioxidant activity of kombucha fermented milk products with stinging nettle and winter savory. Chem. Ind. Chem. Eng. Q. 2013, 19, 129-139.

16. Vukić, V.; Hrnjez, D.; Kanurić, K.; Milanović, S.; Iličić, M.; Torbica, A; Tomić, J. The effect of kombucha starter culture on gelation process, microstructure and rheo-logical properties during milk fermentation, J. Texture Stud. 2014, 45, 261-273.

17. Hrnjez, D.; Vastag, Ž.; Milanovic, S.; Vukić, V.; Iličić, M.; Popović, Lj.; Kanurić, K. The biological activity of fermented dairy products obtained by kombucha, and con-ventional starter cultures during storage, J. Funct. Food. 2014, 10, 223-245.

18. Popović, R.;Milanović, S.; Iličić, M.; Ranogajec, M.; Kanurić, K.; Vukić, V.; Hrnjez, D. Nutritive characteristics and market prospects of kombucha fermented milk beve-rages, Agro FOOD Hitech. 2016, 27, 48-51.

19. Bergmeyer, H. U. Methods of EnzymaticAnalysis. 3 rd ed., Vol. VI, VCH Publishers (UK) Ltd., Cambridge UK, 1150 (1988).

20. STATISTICA version 6 (StatSoftInc, Tulsa, OK, USA (2001). 21. Iličić, M. Optimisation of functional fermented dairy beverage technology. Ph.D.

Thesis, University of Novi Sad, Faculty of Technology Novi Sad, Serbia, May (2010). 22. Radulović, Z.; Iličić, M.; Radin, D.; Paunović, D.; Mitrović, N.; Petrušić, M.; Obra-

dović, D.; Karakterizacija mikroflore kombuhe u fermentisanim mlečnim napicima, Prehrambena industrija-Mleko i mlečni proizvodi, 2009, 20, 106-109.

САДРЖАЈ ШЕЋЕРА, ОРГАНСКИХ КИСЕЛИНА И ЕТАНОЛА У ФEРMEНTИСAНИМ MЛEЧНИМ НАПИЦИМА ПРОИЗВЕДЕНИМ

ПРИМЕНОМ РАЗЛИЧИТИХ КОМБУХА ИНОКУЛУМА

Mирeлa Д. Иличић, Спaсeниja Д. Mилaнoвић, Кaтaринa Г. Кaнурић, Влaдимир Р. Вукић,Светлана С. Поповић, Дajaнa В. Вукић


Булевар Цара Лазара 1, 21000 Нови Сад, Србија У рaду je испитaн утицaj различитих врста и концентрација инокулума комбухе на одабране карактеристике фeрмeнтисaних млeчних напитака произведених из млeкa сa 0,9% млeчнe мaсти. У комбуха ферментисаним млечним напицима анали-зиран је садржај шећера (лактоза, галактоза, глукоза, фруктоза), органских кисели-на (Л- и Д-млечна киселина, сирћетна киселина) и садржај етанола. Рeзултaти анализе пoкaзуjу дa садржај лактозе у комбуха ферментисаним напи-цима варира од 3,3g/100g до 4, 0 g/100g. Сви узорци имају већи садржај глукозе у



116

односу на фруктозу. Садржај Л-млечне киселине у испитиваним напицима креће се 0,4 до 0,7 g/100g, док је садржај Д-млечне киселине, сирћетне киселине и етанола нижи од 0,06g/100g. Кључне речи: ферментисани млечни напици, комбуха инокулум, шећери, органске

киселине, етанол

Received: 24 July 2017. Accepted: 29 September 2017.

APTEFF, 48, 1-323 (2017) UDC: 543.086:66.011:547.869 https://doi.org/10.2298/APT1748117K BIBLID: 1450-7188 (2017) 48, 117-126


117

CHEMOMETRIC AND QSAR ANALYSIS OF SOME THIADIAZINES AS POTENTIAL ANTIFUNGAL AGENTS

Milica Ž. Karadžić*, Strahinja Z. Kovačević, Lidija R. Jevrić,

Sanja O. Podunavac-Kuzmanović

Univeristy of Novi Sad, Faculty of Technology Novi Sad, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia

Quantitative structure-activity relationship (QSAR) analysis has been performed in order to predict the antifungal activity of dihydroindeno and indeno thiadiazines against toxigenic fungus Aspergillus flavus. The studied compounds were classified according to their lipophilicity using the principal component analysis (PCA). The partial least square regression (PLSR) was used to distinguish the most important molecular descriptors for non-linear modeling. Artificial neural networks (ANNs) were applied for the antifungal activity prediction. The best QSAR models were validated by statistical parameters and graphical methods. High agreement between the observed and predicted antifungal activity values indicated the good quality of the derived QSAR models. The obtained QSAR-ANN models can be used to predict the antifungal activity of dihydroindeno and indeno thiadiazines and of structurally similar compounds. The modeling of the antifun-gal activity can contribute to the synthesis of new antifungal agents with better ability to protect food and feed from the mycotoxins. KEY WORDS: artificial neural networks, mycotoxins, partial least square regression,

QSAR, thiadiazines

INTRODUCTION Food and feed quality is very important in terms of human and animal health upkeep and it is necessary to permanently control raw materials and final products. The different types of fungus are frequent contaminants in food and feed products, and this is a natural phenomenon. Of significant importance among them are those species which produce metabolites harmful for humans and animals - mycotoxins. Food and feed contamination can occur due to the raw materials contamination or mycotoxin generation can appear during the production or product storage in inadequate conditions. In animals, myco-toxins can lead to the reduced feed intake, weight loss, drop in milk production, and reduced immune system functioning (1). Besides of reducing animal performance, dif-ferent mycotoxins can also concentrate in milk, meet, and eggs, and affect the human health. Mycotoxins can cause serious problems with the gastrointestinal system, skin irritation, renal damage and carcinogenic changes (2-5). The most important mycotoxige- * Corresponding author: Milica Ž. Karadžić, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




118

nic fungi are Aspergillus, Fusarium, Penicillium, Alternaria, Claviceps and Stachybotrys (6). Toxigenic fungus Aspergillus flavus is one of the four toxigenic fungus that produces Aflatoxin B1, Aflatoxin B2, Aflatoxin G1, Aflatoxin G2, and it is considered to be very important from an economic point of view (4). Economic losses refer to animal pro-ductivity, impact on human health, and global trade in the field of raw materials, food and feed. Differently substituted thiadiazine derivatives are remarkably effective compounds regarding their antifungal activity (7-10). Besides the already known thiadiazines, the development of strategies for design and synthesis of new thiadiazines with antifungal activity is of considerable interest. Quantitative structure-activity relationship (QSAR) approach is considered a very useful tool for the prediction of the bioactivity of molecules. The main task of the present study was to characterize the already synthesized dihy-droindeno and indeno thiadiazines using in silico molecular descriptors and chemometric tools to correlate structural characteristics with antifungal activity toward Aspergillus flavus. The artificial neural networks (ANNs) approach with partial least square reg-ression-variable importance in projection (PLSR-VIP) descriptors selection method was applied in order to achieve reliable prediction of antifungal activity of new thiadiazine derivatives before their synthesis. As there is a big threat and high risk for human health due to indirect exposure (through animal products) and direct exposure to mycotoxins (through food products), the modeling and synthesis of new thiadiazines as potential anti-fungal agents are of crucial importance. The presented results presented should provide guidelines for further molecular design and synthesis of more active antifungal thiadia-zines and structurally similar agents.


The studied thiadiazines and their antifungal activity The investigated thiadiazines include a set of twenty two dihydroindeno and indeno thiadiazines (9). The antifungal activity of these compounds against Aspergillus flavus was taken from the corresponding literature (9). The structures of the investigated di-hydroindeno and indeno thiadiazines are presented in Figure 1. The antifungal activity of the studied thiadiazines was expressed as mycelial growth inhibition - MGI (%).

Figure 1. The set of studied thiadiazines (Et - ethyl; Me - methyl; Ph - phenyl; Pr - propyl)



119

In silico modeling and molecular descriptors calculations The molecular descriptors of the studied thiadiazines have been calculated on the ba-sis of their molecular structures. The 2D structures have been drawn using MarvinSketch v.17.3.13 (11). The 3D molecular structures were subjected to energy minimization using molecular mechanics force field (MMFF) method - MM2, in order to generate 3D molecular descriptors. Lipophilicity, physico-chemical, topological and ADMET in silico descriptors were calculated using the following software: ChemBioDraw Ultra v.12.0 (12), ChemBio3D Ultra v.12.0 (12) ALOGPS 2.1 (13), Molinspiration online program (14) and PreADMET online program (15). Lipophilicity descriptors used for studied compounds classification are presented in Table 1.

Table 1. In silico lipophilicity descriptors of the studied compounds

Classification and QSAR analysis Principal component analysis (PCA) presents one of the most used multivariate tech-niques for the reduction of the data set that consists from the great number of the mutual-ly correlated variables (16, 17). It defines principal components that present the com-

Com

p.

AL

OG

Ps1

AC

logP

1

miL

ogP

1

AL

OG

P1

ML

OG

P1

XL

OG

P21

XL

OG

P31

Av.

logP

1

AL

OG

pS1

AC

logS

1

Av.

logS

1

PC

1

logP

1

CL

ogP

2

1 1.42 2.26 1.72 2.11 3.45 2.79 1.99 2.25 -2.50 -5.31 -3.91 6.55 1.16 0.67 2 1.69 2.47 1.81 2.37 3.73 3.02 2.39 2.50 -2.75 -4.94 -3.84 7.01 1.29 0.94 3 1.91 2.82 2.38 3.04 4.00 3.28 2.82 2.89 -3.02 -5.10 -4.06 7.47 1.99 1.46 4 2.58 3.29 2.89 3.50 4.26 3.63 3.18 3.28 -3.26 -5.37 -4.32 7.94 2.43 1.99 5 3.13 3.95 3.84 4.04 5.05 4.03 3.65 3.94 -3.68 -7.45 -5.57 9.06 3.19 2.76 6 1.32 2.58 2.12 2.60 3.73 3.22 2.35 2.56 -2.89 -5.65 -4.27 7.01 1.68 1.17 7 1.97 2.78 2.21 2.86 4.00 3.12 2.70 2.76 -3.08 -5.80 -4.18 7.47 1.80 1.43 8 2.35 3.14 2.79 3.53 4.26 3.71 3.18 3.28 -3.28 -5.44 -4.36 7.94 2.50 1.96 9 2.76 3.60 3.29 3.98 4.52 4.28 3.54 3.71 -3.51 -5.71 -4.61 8.40 2.95 2.49 10 2.76 3.60 3.29 3.98 4.52 4.28 3.54 3.71 -3.51 -5.71 -4.61 8.40 2.95 2.49 11 3.28 4.26 4.24 4.52 5.28 5.18 4.01 4.40 -3.87 -7.80 -5.83 9.52 3.70 3.26 12 1.47 2.39 2.63 2.46 2.83 3.19 2.15 2.44 -2.34 -5.81 -4.08 6.37 2.20 1.68 13 2.00 2.59 2.72 2.72 3.11 3.42 2.55 2.73 -2.66 -5.44 -4.05 6.83 2.32 1.95 14 2.44 2.95 3.29 3.38 3.38 3.67 2.98 3.16 -2.97 -5.60 -4.28 7.30 3.02 2.48 15 2.86 3.41 3.79 3.84 3.65 4.24 3.34 3.59 -3.21 -5.87 -4.54 7.76 3.47 3.01 16 3.36 4.07 4.75 4.38 4.44 5.14 3.81 4.28 -3.58 -7.95 -5.77 8.88 4.22 3.78 17 1.77 2.71 3.03 2.94 3.11 3.62 2.51 2.81 -2.75 -6.15 -4.45 6.83 2.71 2.18 18 2.17 2.91 3.12 3.20 3.38 3.86 2.91 3.08 -3.03 -5.78 -4.40 7.30 2.83 2.45 19 2.68 3.27 3.69 3.87 3.65 4.11 3.34 3.52 -3.21 -5.94 -4.57 7.76 3.53 2.98 20 3.11 3.73 4.19 4.33 3.90 4.68 3.70 3.95 -3.36 -6.21 -4.78 8.22 3.98 3.51 21 3.11 3.73 4.19 4.33 3.90 4.68 3.70 3.95 -3.36 -6.21 -4.78 8.22 3.98 3.51 22 3.55 4.39 5.15 4.87 4.67 5.58 4.17 4.63 -3.76 -8.30 -6.03 9.34 4.74 4.28

1ChemBio3D Ultra; 2ChemBioDraw Ultra



120

bination of the original variables. The total number of principal components is equal to the number of the analyzed variables. The in silico lipophilicity descriptors from different softwares were gathered in order to conduct the PCA. The aim of the PCA procedure was to group the studied dihydroindeno and indeno thiadiazines according to their lipophi-licity respectively to their antifungal potential. Partial least square regression uses linear combination of the predict variables with the independent variable. In this type of the regression, variables highly correlated with dependent variable additionally gain the importance. As a result of the PLSR analysis, VIP values for all analyzed variables can be obtained and further processed appropriately. The PLSR analysis was used in order to distinguish the input variables for the non-linear modeling of the antifungal activity of the studied compounds. ANNs are very useful when the complex relationships between variables occur. Artificial neuron has a regression equation that transfers input data into non-linear output data and so this method is considered as non-linear (18). In some cases it is necessary to conduct the input data normalization (19). The established QSAR-ANN models were statistically estimated and their predictive ability was expressed through statistical para-meters: correlation coefficient (R) for training, test and validation set, root mean square error (RMSE) for training, test and validation set. The presented classification and regres-sion modeling analysis were conducted by Statistica 13 software (20).


PCA classification The PCA results are presented through the loadings and scores plot in Figure 2. The obtained PCA model is expressed through two principal components, which describe 95.29% of the total variance, to which PC1 contributes 87.55% and PC2 7.74%. From the loadings plot, it can be noticed that the position of the studied compounds on the scores plot is influenced by the majority of lipophilicity descriptors regarding the PC1 axis. The position on the scores plot regarding the PC2 axis is influenced most by MLOGP. The majority of descriptors have negative coefficients of latent variables regarding PC1 axis except for AClogS, Average logS and ALOGpS. Regarding the PC2 axis, the number of descriptors that have negative and positive coefficient of latent variables is equal. From the scores plot, it can be seen that the studied dihydroindeno and indeno thiadiazines are grouped according to their lipophilicity, regarding the PC1 axis. The compounds with lower lipophilicity are positioned at the positive end of the PC1 axis, and the compounds with higher lipophilicity at its negative end. According to PC2, the studied compounds are grouped in accordance their type (dihydroindeno or indeno).



121

Figure 2. The PCA results of the lipophilicity descriptors of studied

compounds - loadings and scores plots

QSAR modeling The first step in QSAR modeling was the PLSR analysis and determination of the influence of the independent variables to the dependent variable. The VIP values were calculated (Figure 3) and the variables which were described by the VIP > 1.9 were taken as the most important, and used in the further non-linear modeling as the input variables. For the ANN modeling, three bioactivity and ADMET descriptors were used - PPB (in vitro plasma protein binding), EI (enzyme inhibition) and PI (protease inhibition).

Figure 3. VIP values of the variables, obtained by applying PLSR modeling



122

For the QSAR-ANN models, the training (70% of compounds: 1-7, 9-11, 13, 15, 17-20), test (15% of compounds: 12, 21, 22) and validation (15% of compounds: 8, 14, 16) sets were used. The number of hidden neurons was varied from 2 to 200 and ANNs were trained using the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm. The activation functions for the hidden and output neurons were as follows: identity (Idt), logistic (Lgt), tangent (Tanh), exponential (Exp) and sinusoidal (Sine). The observed MGI values were subjected to the min-max normalization. The architecture and statistical parameters for two best ANNs are presented in Table 2. High values of correlation coefficient (R) for training, test and validation set, as well as low values of root mean square error (RMSE) for training, test and validation set confirm statistical validity of the presented models.

Table 2. Statistical parameters for the established models

Parameter ANN1 ANN2 Architecture MLP 13-15-1 MLP 13-2-1Rcalib 0.9850 0.9784 Rtest 0.9898 1.0000 Rvalid 0.9838 0.9914 RMSEcalib 0.0013 0.0021 RMSEtest 0.0062 0.0058 RMSEvalid 0.0106 0.0217 Algorithm BFGS 43 BFGS 38 Hidden activation Lgt Lgt Output activation Exp Exp

Additionally, the statistical significance of the established QSAR-ANN models was confirmed by two graphical methods - comparing the observed and predicted values, as well as the observed values and the residuals (Figure 4). The presented graphs indicate a very good concurrence of the data according to the high correlation coefficients, as well as the residuals which were randomly distributed around the zero values on the y axis. Also, the intercept very close to zero and the slope close to 1 are the indicators of the significant linear relationship between the variables.



123

Figure 4. The observed versus predicted values and observed values versus residuals for

the established ANNs Global sensitivity analysis (GSA) describes the influence of the input variable on the variation of the obtained model parameters. The influences of the input variables are ex-pressed through GSA coefficients. All input variables have a significant influence accor-ding to GSA > 1 criterion (Figure 5). In the ANN1 model the most influential variable is in vitro plasma protein binding (PPB) and in ANN2 it is enzyme inhibition (EI).

Figure 5. The GSA coefficients of the input variables figuring in the established ANN

models



124

Cross-validation

In order to confirm the predictivity of the established ANN models, the leave-one-out cross-validation procedure was applied. The cross-validation (CV) parameters are presen-ted in Table 3. The calculated CV parameters, including cross-validation determination coefficient (R2

cv), adjusted determination coefficient (R2adj), predicted residuals error sum

of squares (PRESS), total sum of squares (TSS) and PRESS/TSS ratio, showed that the established ANN models can be considered statistically acceptable for prediction of anti-fungal activity of studied thiadiazine derivatives, since the R2

cv is significantly above 0.60 threshold and the R2

adj parameter is above 0.80. The PRESS, TSS and PRESS/TSS ratio values are in the acceptable range as well, supporting the hypothesis of the satisfactory predictivity of the obtained ANN models.

Table 3. Cross-validation (CV) parameters of the ANN models

CV Parameter ANN1 ANN2 R2

cv 0.9234 0.8674 R2

adj 0.9338 0.8832 PRESS 85.9849 149.4673 TSS 1123.135 1127.249 PRESS/TSS 0.07656 0.13259

CONCLUSION

The artificial neural networks modeling was applied on the set of twenty two dihydro-indeno and indeno thiadiazines. The PLSR-VIP approach used for the selection of descriptors targeted PPB, EI and PI as the most appropriate ones for the prediction of the antifungal activity toward fungus Aspergillus flavus. The selected descriptors describe the ability of the molecule to pass through the biological membrane and reach its site of interest in the cell. In order to make guidelines for further synthesis of these dihydroin-deno and indeno thiadiazines with antifungal activity against the toxigenic fungus Asper-gillus flavus, PPB, EI and PI molecular descriptors should be taken into consideration. The presented ANNs with the used molecular descriptors can provide good prediction of the antifungal activity of the studied compounds although their predictive power should be treated as preliminary due to the small set of molecules. The modeling of the anti-fungal activity contributes to the further molecular design and synthesis in order to find new antifungal thiadiazines and structurally similar compounds that can improve protec-tion of the food and feed from the different types of mycotoxins.

Acknowledgement This study was financially supported by the research project of the Provincial Secre-tariat for Higher Education and Scientific Research of the AP Vojvodina (Project No. 142-451-2604/2017-01/01).



125

REFERENCES 1. Lawloy, P. G.; Lynch, P. B. Mycotoxins in pig feed 1: Source of toxins, prevention

and management of mycotoxicosis. Irish. Vet. J. 2011, 54, 67-71. 2. Baggiani, C.; Giraudi, G.; Vanni, A. A molecular imprinted polymer with recognition

properties towards the carcinogenic mycotoxin ochratoxin A. Bioseparation. 2001, 10, 389-394.

3. Deohate, P. P.; Berad, B. N. Synthesis and antimicrobial activity of 1,3,5-thiadiazines and their isomerism into 1,3,5-triazines. Indian J. Chem. 2005, 44B, 638-642.

4. Iheshiulor, O. O. M.; Esonu, B. O.; Chuwuka, O. K.; Omede, A. A.; Okoli, I. C.; Ogbuewu, I. P. Effects of mycotoxins in animal nutrition: A review. Asian J. Anim. Sci. 2011, 5, 19-33.

5. Antonissen, G.; An Martel, A.; Pasmans, F.; Ducatelle, R.; Verbrugghe E.; Vandenbroucke, V.; Li, S.; Haesebrouck, F.; Van Immerseel, F.; Croubels, S. The impact of Fusarium mycotoxins on human and animal host susceptibility to infectious diseases. Toxins. 2014, 6, 430-452.

6. Zain, M. E. Impact of mycotoxins on humans and animals. J. Saudi Chem. Soc. 2011, 15, 129-144.

7. Hussein, M. A.; El-Shorbagi, A-N.; Khallil, A-R. Synthesis and antifungal activity of 3,3'-ethylenebis(5-alkyl-1,3,5-thiadiazine-2-thiones). Arch. Pharm. 2001, 334, 305-308.

8. Foroumadi, A. R.; Mirzaei, M.; Emami, S.; Salari, P.; Ghaffari, F.; Amini, M.; Shafiei, A. Synthesis and antifungal activity of some new 1,2,4-triazolo [3,4-b] [1,3,4] thiadiazines. DARU. 2002, 10, 34-37.

9. Prakash, O., Aneja, D. K.; Hussain, K.; Lohan, P.; Ranjan, P.; Arora, S.; Sharma, C.; Aneja K. R. Synthesis and biological evaluation of dihydroindeno and indeno [1,2-e] [1,2,4]triazolo [3,4-b] [1,3,4]thiadiazines as antimicrobial agents. Eur. J. Med. Chem. 2011, 46, 5065-5073.

10. Čačić, M.; Pavić, V.; Molnar, M.; Šarkanj, B.; Has-Schön, E. Design and synthesis of some new 1,3,4-thiadiazines with coumarin moieties and their antioxidative and antifungal activity. Molecules. 2014, 19, 1163-1177.

11. Chem Axon, Ltd. <http://www.chemaxon.com/>. 12. ChemBioOffice 2012. Perkin Elmer Informatics. <http://www.cambridgesoft.com/>. 13. VCCLAB, 2005. Virtual Computational Chemistry Laboratory. <http://www.vcclab.org>. 14. Molinspiration Cheminformatics, http://www.molinspiration.com/. 15. PreADMET software, https://preadmet.bmdrc.kr/. 16. González-Díaz, H.; Saíz-Urra, L.; Molina, R.; González-Díaz, Y.; Sánchez-González,

A. Computational chemistry approach to protein kinase recognition using 3D sto-chastic van der Waals spectral moments. J. Comput. Chem. 2007, 28, 1042-1048.

17. González-Díaz, H.; Saíz-Urra, L.; Molina, R.; Santana, L.; Uriarte, E. A model for the recognition of protein kinases based on the entropy of 3D van der Waals interactions. J. Proteome Res. 2007, 6, 904-908.

18. Kovačević, S. Z.; Podunavac-Kuzmanović, S. O.; Jevrić, L. R.; Djurendić, E. A.; Ajduković, J. J. Non-linear assesment of anticancer activity of 17-picolyl and 17-



126

picolinylidene androstane derivatives - Chemometric guadlines for further syntheses. Eur. J. Pharm. Sci. 2014, 62, 258-266.

19. Jayalakshmi, T.; Santhakumaran, A. Statistical normalization and back propagation for classification. IJCTE. 2011, 3, 89-93.

20. StatSoft Inc., STATISTICA (Data Analysis Software System), Version 13. http://www.statsoft.com/.

ХЕМОМЕТРИЈСКА И QSAR АНАЛИЗА ОДАБРАНИХ ТИАДИАЗИНА КАО ПОТЕНЦИЈАЛНИХ АНТИФУНГАЛНИХ АГЕНАСА

Милица Ж. Караџић*, Страхиња З. Ковачевић, Лидија Р. Јеврић,

Сања О. Подунавац-Кузмановић

Универзитет у Новом Саду, Технолошки факултет Нови Сад, Булевар Цара Лазара 1, 21000 Нови Сад, Србија

Спроведена је анализа квантитативне везе између структуре и активности (QSAR) у циљу предвиђања антифунгалне активности дихидроиндено и индено тиадиазина према токсигеној плесни Aspergillus flavus. Испитивана једињења су класификована на основу своје липофилности помоћу анализе главних компоненти (PCA). Регресија методом парцијалних најмањих квадрата (PLSR) употребљена је да би издвојила најважније дескрипторе за нелинеарно моделовање. Вештачке неу-ронске мреже (ANN) примењене су за предвиђање антифунгале активности. Најбо-љи QSAR модели су валидирани помоћу статистичких параметара и графичким методама. Висока сагласност између посматраних и предвиђених вредности анти-фунгалне активности указују на добар квалитет изведених QSAR модела. Добијени QSAR-ANN модели могу се користити за предвиђање антифунгале активности дихидроиндено и индено тиадиазина и структурно сличних једињења. Предвиђање антифунгалне активности може да допринесе синтези нових антифунгалних агена-са са бољом способношћу заштите хране и хране за животиње од микотоксина. Кључне речи: вештачке неуронске мреже, микотоксини, регресија методом парци-

јалних најмањих квадрата, квантитативна веза између структуре и активности молекула, тиадиазини

Received: 18 August 2017.

Accepted: 06 October 2017.

APTEFF, 48, 1-323 (2017) UDC: 541.183:667.284:[633.15+581.823 https://doi.org/10.2298/APT1748127K BIBLID: 1450-7188 (2017) 48, 127-139


127

CONTINUOUS ADSORPTION OF METHYLENE BLUE DYE ON THE MAIZE

STEM GROUND TISSUE

Predrag S. Kojić*, Vesna V. Vučurović, Nataša Lj. Lukić, Milica Ž. Karadžić, Svetlana S. Popović


Continuous adsorption of methylene blue from aqueous solutions onto maize stem ground tissue in column mode was investigated. The study encompassed the effects of important parameters such as flow rate, initial concentration of methylene blue, and bed depth on methylene blue removal from model solutions. The maximum adsorption capa-city of the maize stem was 45.9 mg/g at the initial methylene blue concentration of 20 mg/L, bed height of 6.5 cm and flow rate of 8 mL/min. It was found that the breakthrough time for reaching saturation increased with a decrease in the flow rate, and also occurred earlier for a higher influent concentration. The breakthrough times increased with the bed depth, thus allowing a larger volume to be treated. The Adams-Bohart, Yoon-Nelson, Clark and artificial neural network models were used to predict the breakthrough curves. These models gave excellent approximations of the experimental behavior. KEY WORDS: biosorption, fixed-bed column, breakthrough curve, maize stem, model-

ling, ANN

INTRODUCTION

Many industries, such as paper making, plastics, food, cosmetics, textile, etc., use dyes in order to color their products. The presence of these dyes in water, even at very low concentrations, is highly visible and undesirable (1). Adsorption is one of the pro-cesses that used for dye removal in wastewater treatment (2). Activated carbon is the most widely used adsorbent because it has excellent adsorption efficiency for organic compounds, but its use is usually limited due to high costs (3). This method will become less expensive if the adsorbent is made from cheap and easily available materials. Methylene blue (MB) was extensively used as the model adsorbate in the previous studies (3-10). The MB is a thiazine (cationic) dye that is generally used for dyeing cotton, coloring paper, wool, silk, hair, etc. Though MB is not strongly hazardous, it can cause some harmful effects (7). The search for a low cost and easily available adsorbent led to the investigation of materials of agricultural and biological origin, along with industrial by-products, as ad- * Corresponding author: Predrag S. Kojić, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




128

sorbents (11). In the recent years, many researchers have proved several low cost ma-terials for the removal of MB from its aqueous solutions, such as wheat shells (12), chaff (8), Indian Rosewood sawdust (3), phoenix tree leaf powder (9), rice husk (1, 7), Neem leaf powder (13). The list of numerous adsorbents for MB removal were presented by Ralatullah et al. (10). Bhatnagar et al. (14) reported a review of agro-industrial and mu-nicipal waste materials as potential adsorbents for water treatment. In the present study, maize stem ground tissue has been used as an adsorbent for the removal of MB from its aqueous solutions. The material is extremely cheap, non-toxic, chemically inert and easily available biodegradable waste. To our knowledge, this is the first study about dye adsorption on this adsorbent in the column mode. The aim of this work was to promote our previous study (15) on the development of a convenient and economic method for MB removal on maize stem grand tissue as a low-cost and easily available biodegradable adsorbent. The effect of adsorptive time, flow rate, MB concentration, and bed depth on MB removal from aqueous solutions in the column mode was investigated. The Adams-Bohart, Yoon-Nelson, Clark and artificial neural network models were used to predict the breakthrough curves.

EXPERIMENTAL

Preparation of the adsorbent

Maize stems of Gold Cup maize hybrid were collected from ready-to-harvest maize fields from Budisava, Serbia. The details on the preparation and characterization of the adsorbent can be found in the our previous paper (15).

Preparation of the MB solutions All of the MB solutions were prepared with distilled water. Three different concentra-tions of the MB solutions were prepared and used in this work: 5 mg/L, 10 mg/L and 20 mg/L.

Experimental methods A known adsorbent mass of 0.35, 0.45 and 0.65 g was packed into a plastic column of 1.2 cm internal diameter and 10 cm in height, to achieve a bed depth of 3.5, 4.5 and 6.5 cm, respectively. The column was operated in a downflow mode using gravity force. Three different influent flow rates were used: 8 mL/min, 20 mL/min and 40 mL/min. Flow rate was regulated by a valve located above the column. Samples of the effluent (10 mL aliquot) were collected at regular intervals. Concen-trations of MB in the samples were determined spectrophotometrically on a SECOMAM Prim: Light and Advance spectrophotometer, in the visible range of the spectrum. Dis-tillated water was used as the reference sample. The maximum absorbance vawelength was 668 nm.



129

Fixed bed column adsorption studies The performance of the column was described through the concept of the break-through curve. The breakthrough curve is expressed in terms of a normalized concentra-tion defined as the ratio of the outlet to inlet MB concentration (C/C0) as a function of time or volume effluent. The effluent volume (Veff) was calculated as:

Veff=Qttotal [1]

where ttotal and Q were the time of exhaustion (min) and volumetric flow rate (mL/min). The total adsorbed MB quantity qtotal, in mg, (maximum column capacity) was expres-sed as:

qtotal=QA

1000= Q

1000Caddtt=ttotal

t=0 [2] where Cad was the concentration of MB removal in mg/L. qtotal represents the area below the breakthrough curve multiplied with the flow rate. The amount of MB at equilibrium or biosorption capacity, qe (mg of sorbated MB/g of biosorbent), was determined using the following equation,

qe=qtotal

X [3]

where X is the mass of the biosorbent in g. Various fixed bed models have been developed to predict the dynamic behavior of the column (16).

The Adams-Bohart model for rectangular isotherms In the column mode, the adsorption kinetics for the MB can be presented by the Adams-Bohart model. It assumes a rectangular isotherm with a quasi-chemical rate ex-pression. Axial dispersion and finite resistance to mass transfer are neglected in this model. The Adams-Bohart model (Eq. 4) is one of the most general and widely used mo-dels in the column performance theory. It is commonly, but mistakenly referred in the environmental sorption and biosorption literature as the Thomas model (17). The Adams-Bohart model in this paper was used, as follows:

C

C0=

1

1+exp ((kThQ

)(q0X-C0Veff) [4]

where kTh is the rate constant (mL/min mg) and q0 is the maximum solid-phase con-centration of the solute (mg/g).



130

The Yoon-Nelson model

This model is based on the assumption that the rate of decrease in the probability of adsorption for each adsorbate molecule is proportional to the probability of adsorbate adsorption and the probability of adsorbate breakthrough on the adsorbent (11). In this work, the Yoon-Nelson model was used, as follows:

ln(C

C0-C)=kYNt-τkYN [5]

where kYN is the rate constant (1/min), τ the time required for 50% adsorbate break-through (min) and t is breakthrough time (min).

Clark model This model combines the Freundlich isotherm and the mass transfer concept. The Clark model is presented in the following equations (18):

C0n-1

1+C0

n-1

Cbn-1

-1 ertb e-rt

1

n-1

= C [6]

or

1

(1+Ae-rt)

1

n-1= C

C0 [7]

where:

A=C0

n-1

Cbn-1

-1 ertb [8]

Cb, tb and n are the outlet concentration at breakthrough, the time at breakthrough and the Freundlich constant, respectively.

Artificial neural network Recently, traditional breakthrough curve models have been progressively replaced with artificial neural network (ANN) models. Witek-Krowiak et al. (19) summarized the studies that used ANN for modeling of batch and fixed bed biosorption process. The ANN predicts the output on the basis of the input data without the need to explicitly define the relationship between them, which is important in the complex process such as biosorption (20).



131


Effect of the flow rate on the breakthrough curve The effect of different flow rates on the shape of the breakthrough curves is shown in Figure 1. As can be seen, the breakthrough curves became steeper as the flow rate in-creased. The breakthrough time for reaching saturation increased with a decrease in the flow rate. It was found that at the lower flow rate, higher uptake values were observed for MB to maize stems. This phenomenon could be explained through a longer contact time of the MB with the adsorbent. Hence lower flow rates are desirable for the effective removal of MB.

Figure 1. Comparison of the experimental and predicted curves obtained at different flow rates for C0 = 5 mg/L and Z = 6.5 cm

Effect of the influent MB concentration on the breakthrough curve

The adsorption performances of maize stems were tested at various initial concen-tration of MB (Figure 2). The breakthrough time decreased with the increasing inlet concentration as the binding sites became more quickly saturated in the system. Also, the



132

increase in the influent concentration resulted in a sharper breakthrough. The slope of the curve indicates high adsorption rates and shortened mass transfer zone.

Figure 2. Comparison of the experimental and predicted curves obtained at different initial concentration for Q = 8 mL/min and Z = 6.5 cm

The effect of the different bed depth on the breakthrough curve

The MB breakthrough curves, obtained from the experiments at different bed heights are presented in Figure 3. The figure shows that the breakthrough curve became steeper as the bed height decreased. The breakthrough times increased with the bed heights, thus allowing a larger volume to be treated. With a higher bed depth column, the mass transfer zone is extended and the MB had more time to be in contact with the adsorbent.



133

Figure 3. Comparison of the experimental and predicted curves obtained at different bed depth for C0 = 5 mg/L and Q = 8 mL/min

Error analysis

To describe the column breakthrough curves obtained at the different flow rates, MB inlet concentrations, and bed depths, three different models and ANN were used. The breakthrough curves showed the superposition of the experimental results (points) and the theoretically calculated points (lines). The coefficient of determination (R2) shows a crude measure of how well a model describes the variation of the data. This is an intuitive, but crude, method to compare the two models, especially nonlinear models. In order to avoid this problem, an error analysis was performed according to the mean squared error (MSE). All models used in our paper have two unknown parameters, so in the discussion there is no need for complex error criteria that include the parameters of the models. The relative mathematical formula of MSE is:

MSE=∑ C

C0 e-

CC0 c

2

n [9]

where (C/C0) e is the ratio of the effluent and influent MB concentrations obtained from experiment and (C/C0)c is the ratio of the effluent and influent MB concentrations



134

obtained from dynamic models, and n is the number of experimental data. Furthermore, the residual analysis of the developed model was also performed. The skewness parameters showed small deviations from a normal distribution, -1.37 to 2.34, while the kurtosis parameter showed a difference in "peakedness" compared to normal distribution, 2.83 to 7.21. However, the developed models showed statistically insignificant deviations from the experimental values of the model, which confirmed their suitability.

The Adams-Bohart model for the rectangular isotherm The breakthrough curves (Figures 1-3) were described by the Adams-Bohart model for various experimental conditions. The Quasi-Newton method was used to evaluate the Adams-Bohart model parameters of kTh and q0 (Table 1).

Table 1. Parameters predicted from the Adams-Bohart model

C0 (mg /L)

Q (mL/min)

Z (cm)

kTh (mL/min mg-1)

q0 (mg g-1)

R2 MSE (10-3) qe(exp) (mg g-1)

5 8 6.5 5.95 31.92 0.99 0.51 31.88

10 8 6.5 4.73 39.23 0.99 1.62 39.33

20 8 6.5 2.85 45.57 0.99 0.56 45.93

5 20 6.5 11.61 30.07 0.99 1.02 30.45

5 40 6.5 43.94 28.04 0.99 0.87 28.22

5 8 4.5 5.7 28.19 0.99 2.07 28.65

5 8 3.5 8.84 24.25 0.99 0.71 24.56

It can be concluded from Table l that the increase in the initial concentration resulted in an increase in the value of q0 and in a decrease in the value of kTh. This is because the driving force for adsorption, i.e. the difference in the concentration of MB on the maize stems and in the solution, decreases with increasing initial concentration (21). The bed capacity q0 slightly decreased, and the coefficient kTh increased with increasing flow rate. Furthermore, with the increase in the bed volume, the value of q0 increased. The Adams-Bohart model gave a bed capacity that was very close to the value determined in the batch process (15).

The Yoon-Nelson model The Yoon-Nelson model was applied to investigate the breakthrough behavior of MB onto fixed bed of maize stems (Figures 1-3). The experimental breakthrough curves are very similar to those predicted by the Yoon-Nelson model. The values of kYN and τ were determined using Quasi-Newton method and listed in Table 2. As is evident, the kYN



135

increased and τ decreased with increasing both, flow rate and MB influent concentration. The increase in the bed volume caused an increase in τ and a decrease in kYN.

Table 2. Parameters predicted from the Yoon-Nelson model

C0 (mg /L)

Q (mL/min)

Z (cm)

kYN (min-1)

Τ (min)

R2 MSE (10-3)

5 8 6.5 0.018 475.75 0.99 0.59

10 8 6.5 0.015 232.54 0.99 0.42

20 8 6.5 0.033 161.14 0.99 0.85

5 20 6.5 0.052 195.48 0.99 0.98

5 40 6.5 0.099 79.48 0.99 0.82

5 8 4.5 0.033 322.51 0.99 2.03

5 8 3.5 0.044 157.05 0.99 0.44

The Clark model The parameters of the Clark model and the determined coefficients for various conditions were obtained by the Levenberg-Marquardt model. The Freundlich constant n, obtained in our earlier batch equilibrium study, was used to calculate the remaining two parameters in the Clark model (15). The calculated parameters are given in Table 3. It is clear that if the flow rate and influent dye concentration increased, the value of r also increased. The determined coefficients and Figures 1-3 show a good agreement of the Clark model with the experimental data.

Table 3. Parameters predicted from the Clark model

C0 (mg /L)

Q (mL/min)

Z (cm)

ln(A) R

(min-1) R2 MSE (10-3)

5 8 6.5 8.93 0.021 0.99 0.48 10 8 6.5 7.2 0.03 0.99 0.39 20 8 6.5 6.68 0.05 0.99 0.26 5 20 6.5 6.77 0.045 0.99 0.38 5 40 6.5 7.69 0.1 0.99 6.7 5 8 4.5 5.05 0.022 0.99 1.22 5 8 3.5 2.86 0.032 0.99 0.21



136

Artificial neural network The feed-forward back propagation ANN was used to predict the experimental values for the breakthrough curve prediction. The Matlab neural network toolbox was used to build the ANN (The Math Works Inc. License Number 1108951). The ranges of the ANN inputs were: 0 < t < 687.5, 5 < C0 < 20, 8 < Q < 40 and 3.5 < z < 6.5. In order to prevent the influence of the initial assumptions of weights and bias, the different topologies were used, in which the number of hidden neurons were varied from 1 to 20, and the training process of each network was run ten times with random initial values of weights and biases. The numbers of neurons were 13 and 1 in the hidden and the output layer, respec-tively (Figure 4). The network was trained using the Bayesian regulation back propaga-tions algorithm. The others training algorithms (Levenberg-Marquardt, BFGS Quasi-Newton, Scaled Conjugate Gradient, etc.) were tested too, but they did not give satisfac-tory outputs. The transfer function was the linear transfer function (purelin) at the output layer, and the tangent sigmoid transfer function (tansig) at the hidden layer. All 235 data points were randomly used to train and develop the ANN; 70% of data points for training, 15% of data for validations, and 15% of data for testing the process. The optimum ANN topology was determined based on the MSE. The error analysis is shown in Table 4 for all experimental conditions.

Table 4. Error analysis for ANN

C0 (mg /L)

Q (mL/min)

Z (cm)

R2 MSE(10-4)

5 8 6.5 0.99 1.6910 8 6.5 0.99 2.620 8 6.5 0.99 2.725 20 6.5 0.99 0.895 40 6.5 0.99 1.235 8 4.5 0.99 3.65 8 3.5 0.99 0.82

Figure 4. Topology of the ANN



137

CONCLUSIONS The present study showed that the maize stems of Gold Cup could be used efficiently as a low cost and easily available biodegradable adsorbent for continuous removal of methylene blue from aqueous solutions. The breakthrough curves showed that the breakthrough time increased with the increase in the bed height and with a decrease in the initial concentration and flow rate of methylene blue. Three models and the artificial neural network were applied to the experimental data obtained from dynamic studies in order to predict the breakthrough curves and determine the column kinetic parameters. According to the MSE, the ANN model was the best fitting model compared to the Adams Bohart, Yoon-Nelson and Clark models. Further work needs to be done to establish the optimal conditions of flow rates, bed height, and initial methylene blue concentration to satisfy economic and technological requirements.

Acknowledgement

This research was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia. (Project No. 172025).

REFERENCES

1. Han, R.; Ding, D.; Xu, Y.; Zou, W.; Wang, Y.; Li, Y.; Zou, L. Use of rice husk for the adsorption of congo red from aqueous solution in column mode. Bioresour. Technol. 2008, 99 (8), 2938-2946.

2. Gupta, V.K.; Suhas. Application of low-cost adsorbents for dye removal - A review. J. Environ.Manage. 2009, 90 (8), 2313-2342.

3. Garg, V.K.; Amita, M.; Kumar, R.; Gupta, R. Basic dye (methylene blue) removal from simulated wastewater by adsorption using Indian Rosewood sawdust: a timber industry waste. Dyes Pigm. 2004, 63 (3), 243-250.

4. McKay, G.; Porter, J.F.; Prasad, G.R. The Removal of Dye Colours from Aqueous Solutions by Adsorption on Low-cost Materials. Water, Air, Soil Pollut. 1999, 114 (3), 423-438.

5. Otero, M.; Rozada, F.; Calvo, L.F.; García, A.I.; Morán, A. Kinetic and equilibrium modelling of the methylene blue removal from solution by adsorbent materials produced from sewage sludges. Biochem. Eng. J. 2003, 15 (1), 59-68.

6. Kumar, K.V.; Ramamurthi, V.; Sivanesan, S. Modeling the mechanism involved du-ring the sorption of methylene blue onto fly ash. J. Colloid Interface Sci. 2005, 284 (1), 14-21.

7. Vadivelan, V.; Kumar, K.V. Equilibrium, kinetics, mechanism, and process design for the sorption of methylene blue onto rice husk. J. Colloid Interface Sci. 2005, 286 (1), 90-100.



138

8. Han, R.P.; Wang, Y.F.; Han, P.; Shi, J.; Yang, J.; Lu, Y.S. Removal of methylene blue from aqueous solution by chaff in batch mode. J. Hazard. Mater. 2006, 137 (1), 550-557.

9. Han, R.; Wang, Y.; Zhao, X.; Wang, Y.; Xie, F.; Cheng, J.; Tang, M. Adsorption of methylene blue by phoenix tree leaf powder in a fixed-bed column: experiments and prediction of breakthrough curves. Desalination 2009, 245 (1-3), 284-297.

10. Rafatullah, M.; Sulaiman, O.; Hashim, R.; Ahmad, A. Adsorption of methylene blue on low-costadsorbents: A review. J. Hazard. Mater. 2010, 177 (1-3), 70-80.

11. Aksu, Z.; Gönen, F. Biosorption of phenol by immobilized activated sludge in a con-tinuous packed bed: prediction of breakthrough curves. Process Biochem. 2004, 39 (5), 599-613.

12. Bulut, Y.; Aydın, H. A kinetics and thermodynamics study of methylene blue adsorp-tion on wheat shells. Desalination 2006, 194 (1), 259-267.

13. Bhattacharyya, K.G.; Sharma, A. Kinetics and thermodynamics of Methylene Blue adsorption on Neem (Azadirachta indica) leaf powder. Dyes Pigm. 2005, 65 (1), 51-59.

14. Bhatnagar, A.; Sillanpää, M. Utilization of agro-industrial and municipal waste mate-rials as potential adsorbents for water treatment - A review. Chem. Eng. J. 2010, 157 (2-3), 277-296.

15. Vučurović, V.M.; Razmovski, R.N.; Miljić, U.D.; Puškaš, V.S. Removal of cationic and anionic azo dyes from aqueous solutions by adsorption on maize stem tissue. J. Taiwan Inst. Chem. Eng. 2014, 45 (4), 1700-1708.

16. Ruthven, D.M. Principles of adsorption and adsorption processes; John Wiley & Sons: 1984.

17. Chu, K.H. Fixed bed sorption: Setting the record straight on the Bohart-Adams and Thomas models. J. Hazard. Mater. 2010, 177 (1-3), 1006-1012.

18. Clark, R.M. Evaluating the cost and performance of field-scale granular activated carbon systems. Environ. Sci. Technol. 1987, 21 (6), 573-580.

19. Witek-Krowiak, A.; Chojnacka, K.; Podstawczyk, D.; Dawiec, A.; Pokomeda, K. Application of response surface methodology and artificial neural network methods in modelling and optimization of biosorption process. Bioresour. Technol. 2014, 160, 150-160.

20. Tovar-Gómez, R.; Moreno-Virgen, M.R.; Dena-Aguilar, J.A.; Hernández-Montoya, V.; BonillaPetriciolet, A.; Montes-Morán, M.A. Modeling of fixed-bed adsorption of fluoride on bone charusing a hybrid neural network approach. Chem. Eng. J. 2013, 228, 1098-1109.

21. Chen, S.; Yue, Q.; Gao, B.; Li, Q.; Xu, X.; Fu, K. Adsorption of hexavalent chro-mium from aqueous solution by modified corn stalk: A fixed-bed column study. Bioresour. Technol. 2012, 113, 114-120.



139

КОНТИНУАЛНА АДСОРПЦИЈА БОЈЕ МЕТИЛЕНСКО ПЛАВО НА ПАРЕНХИМСКОМ ТКИВУ СТАБЛА КУКУРУЗА

Предраг С. Којић*, Весна В. Вучуровић, Наташа Љ. Лукић, Милица Ж. Караџић,

Светлана С. Поповић

Универзитет у Новом Саду, Технолошки факултет Нови Сад, Булевар цара Лазара 1, 21000 Нови Сад, Србија

У раду је испитана континуална адсорпција катјонске боје метиленско плаво на паренхимском ткиву стабла кукуруза у колони. Испитани су основни параметри који утичу на уклањање метиленског плавог из водених модел раствора а то су за-премински проток и почетна концентрација адсорбата и висина слоја адсорбента. Максимални адсорпциони капацитет од 45,9 mg/g је остварен при почетној концен-трацији метиленског плавог од 20 mg/L, висини слоја адсорбента од 6,5 cm и про-току од 8 mL/min. Установљено је да се време потребно за адсорпциону равнотежу повећава са смањењем запреминског протока и да се до равнотеже раније долази ако се повећа почетна концентрација адсорбата. Повећање висине слоја адсорбента омогућило је уклањање боје из веће запремине адсорбата. Модели Adams-Bohart-а, Yoon-Nelson-а и Clark-а као и вештачка неуронска мрежа су коришћени за пред-виђање адсорпционих кривих. Ови модели су показали одлично слагање са експе-рименталним резултатима. Кључне речи: биосорпција; адсорпционa кривa; паренхимско ткиво стабла кукуру-

за; моделовање, неуронске мреже




141

THE APPLICATION OF CLEANING VALIDATION PRINCIPLES ON DIETARY SUPPLEMENTS PRODUCTION EQUIPMENT

Davor J. Korčok, Nada A. Tršić- Milanović*

Abela Pharm d.o.o., Viline vode b.b., 11000 Belgrade, Serbia

Cleaning validation for pharmaceutical production equipment is a documented proof of the efficient cleaning, and one of prerequisites of good manufacturing practice in medicine production. Successful validation confirms the efficiency of the procedures of cleaning, washing, and disinfecting of the manufacturing equipment, and records results of the chemical and microbiological analyses, which are a prerequisite for a safe final dietary product. The main goal of this study was to improve the cleaning process of the production equipment by using cleaning validation procedures on the solid form production line (capsules) in the Abela Pharm d.o.o. The validation principles that are used in manu-facturing of medicines can be applied to determine more efficient cleaning methods that will ensure longer periods of the status clean in the production of dietary supplements. The outcome is a practical analysis of the production equipment in view of regulatory demands, confirming that the cleaning validation measures ensure prevention of unwanted microbial growth or removal of contamination from the production equipment in order to preserve the activity, efficacy, and safety of the final dietary product. KEY WORDS: microbiological criteria, cleaning validation, dietary supplements

INTRODUCTION Modern day pharmaceutical manufacture meets strict regulatory and practical de-mands regarding employee hygiene standards, work environment and cleanliness of production equipment (1, 2, 3). Expert teams that organize and manage pharmaceutical production implement hygiene principles, and afterwards monitor chemical and micro-biological parameters that document efficacy of cleaning and disinfection processes in practice. All hygienic precautionary measures are taken in order to achieve the produc-tion of safe pharmaceutical product of standard quality that contains active components together with excipients without any impurities or microorganisms that might disturb its activity or affect badly on consumer health (4).

* Corresponding author: Nada A. Tršić-Milanović, Abela Pharm d.o.o., Viline vode b.b., 11000 Belgrade,

Serbia, e-mail: [email protected]



142

In pharmaceutical practice, hygienic standards are implemented by the use of Good Manufacturing Practice, and are realized with the appliance of cleaning validation prin-ciples for the production of medicines. Cleaning validation in pharmaceutical manufac-turing is a documented proof that the cleaning procedures repeatedly remove residues of previous products or cleaning agents up to maximum allowable scientifically and regula-tory confirmed limits (1, 5). Cleaning validation principles are obligatory in the manufacture of medicines, and they are also used in the production conditions that possess HVAC system (heating, ven-tilation and air-purification systems). In this research, validation principles are used as an additional data in regard to relevant literature, and are applied for non-HVAC production line system for the production of dietary supplements (5). All preliminary activities and practical procedures of validation are used to define adequate routine activities that will in practice guarantee hygienic sustainment on the used equipment. Therefore, they represent prerequisites of continual work that conforms to the relevant regulatory demands, and have safe and efficient final pharmaceutical pro-duct as the main result. In order to obtain reliable data by analyzing chemical and micro-biological parameters of equipment cleaning process, often needs the implementation of certified cleaning and disinfecting agents. If the process of harmonizing with the regulatory demands leads to the larger savings of the production resources, (bigger and more efficient use of equipment, space and employees), then multiple goals of safe and efficient production are achieved together with the preservation of natural resources. These multiple goals, as defined, serve as a starting point for the introduction of cleaning validation on multifunctional capsule production line in pharmaceutical company Abela Pharm d.o.o. Research conducted during August, 2015. was supposed to show the effects of implementation of cleaning protocols with the use of certified cleaning and disinfec-ting agents, in order to create a more efficient cleaning method and a longer lasting duration of status “clean” on the production equipment.

EXPERIMENTAL Risk analysis for Bulardi ® Probiotik capsules was performed together with sampling plan, and alongside with a validation cleaning protocol for the multifunctional capsuli-zing equipment, prior to experimental work. Risk analysis for all products manufactured on a multifunctional solid dosage form equipment for the manufacture of capsules inclu-ded comparing results for: solubility as a risk factor, pharmacology as a risk factor and formulation as a risk factor (2, 6, 7). Moreover, different sampling places were chosen for direct sampling (microbiological smear) and indirect sampling (chemical washings) on the production equipment in order to determine status “clean/dirty”. Considering above mentioned risk factors for products manufactured on the production equipment for solid dosage form, a worst case was discovered to be Bulardi® Probiotik, capsules as it contains Saccharomyces boulardii, a probiotic yeast insoluble in water, with many resi-dues which could lead to microbiological contamination of equipment and therefore con-tamination of the next product manufactured on the same production line.



143

Because of this reason, new work procedures were devised for cleaning and disin-fecting of equipment. They included reports and recommendations from certified manu-facturers of cleaning and disinfecting agents used in pharmaceutical industry, as well as the results of the try-out experiments of cleaning and disinfection of samples manufactu-red on the production line for solid dosage forms- capsules (8, 9).

This study compares results of cleaning and disinfection of multipractical capsulizing equipment using ethanol, 70% solution (Ethanol assessment) and using certified cleaning and disinfecting agents (Validation assessment) while performing defined cleaning validation principles (8,9). In order to properly carry out the comparison, both exami-nations included tests for worst case scenario previously discovered by risk analysis on the same capsulizing machine.

Three consecutive batches of the worst case product of the same size were analyzed in both experiments. The same sampling places were used on the equipment for micro-biological smears and chemical washings (Table 1). Samplings were defined to be repre-sentative for cleaning process of all parts of the equipment (10). Parts of equipment especially exposed to product or hard to reach were chosen for microbiological smears, whereas chemical sampling was performed from hardly accessible points on the equip-ment.

Table 1. Microbiological and chemical sampling spots on production equipment for solid dosage forms

Equipment name

Number of sample points

Type of sample Microbiological smear

(MB) Chemical washing

(HI) Mixing machine

3 MB1, MB2 HI1

Blistering machine

3 MB3,MB4, MB5 n/a

Capsulating machine

10 MB6, MB7, MB8, MB9,

MB10, MB11, MB12 HI2, HI3, HI4

Tools 2 MB13 HI5 Total: 18 13 5

First phase of testing included the analysis of the sampling places after cleaning/ disinfection of equipment with ethanol, 70% solution, performed immediately before the production (up to 1h) of three consecutive batches of Bulardi® Probiotik capsules (Ethanol assessment). Validation plan for the second phase of testing was devised with certified cleaning agents- alkaline detergent (potassium hydroxide with nonionic surfactants), alkaline di-sinfectant without aldehydes (dodecyl dimethyl ammonium chloride and n-(3-amino-propyl)-n-dodecylpropane-1,3-diamine) and acidic disinfectant based on hydrogen pero-xide and peracetic acid (8). Parameters of analysis were determined during the method of cleaning validation, as well as the acceptable limits and status time of “clean” and “dirty”. A competitive cleaning validation (during the routine production of product in-



144

tended for market placement) (11) was performed on three consecutive bathes of Bulardi® Probiotik, capsules, with the same batch size and with careful documentation of work phases. During the second phase of testing, samples with the status “dirty”, that lasted for 8h±2h, were collected, as well as samples with status “clean”, which lasted 12h±4h after the application of determined validation principles of cleaning (Validation assessment). This research provided comparison of results with status “clean” during both phases of testing.


Material and methods of visual inspection Before the sampling process of equipment parts began, dry and well-lit surface was observed from multiple angles to assess work surface (12).

Material and methods of ATP monitoring Identification of biological and/or organic material on equipment was preliminary tested using the fast ATP bioluminescence method, via ATP smears (ATP apparatus Lu-mitester PD-20, LuciPac Pen, Kikkoman, Japan). ATP smears (bioluminescence test for adenosine triphosphate) determine total microbiological burden or in other words, the presence of bacteria, yeasts, molds, biofilm formations or organic residues. Cleanliness status by bioluminescence was assessed based on results of ATP measurements and the result was given in relative light units (RLU). The results of ATP smears confirm that the presence of total microbiological burden: bacteria, yeasts (such as Saccharomyces bou-lardii), and other organic material after cleaning and disinfection process are within ac-ceptable limits (13, 14). This non-selective preliminary analysis quantifies any microbio-logical burden on the production equipment. Therefore, it is used as a primary method for the assessment of the efficacy of cleaning and disinfection processes.

Material and methods of the assessment of purified water chemical washings (HI) A total of 100 ml of purified water was sampled from final washings of inaccessible parts of the equipment, and was tested for (10): 1. pH value- determined by potentiometric technique (15)

2. electroconductivity- determined (µS/cm at 20º C) by conductometric method (16)

3. total and composite alkaline properties- determined by titrimetric method (17)

Material and methods of microbiological testing (MB)

1. Sterile cotton wool or synthetic material- smears were used for microbiological sampling (alginate or rajone) (18) after the addition of sterile saline solution. After the sampling, a smear was placed inside a test tube containing sterile sali-



145

ne solution. With each set of samples, a negative control was also delivered to the laboratory (blank sample). Surface area used for sampling was 100 cm2. A total count of aerobic mesophile bacteria was done at 30 °C, using horizontal method for determination of microorganism count via surface inoculation tech-nique (19, 9, 14).

2. Determination of the count of Enterobacteriaceae was done by horizontal method for determination and identification of Enterobacteriaceae via colony-count approach (20).

RESULTS AND DISCUSION

Ethanol assesment

During the first assessment, visually inspected spots were sampled on the production line treated with ethanol, 70% solution. Microbiological burden was determined after-wards by ATP smears, and chemical washings and microbiological smears were tested immediately before the production process began (status “clean” after 1h- Table 1). Visual inspection previously determined that the surface equipment was clean, whereas the results of ATP assessment were ≤200 RLU. The results of chemical sampling and average values obtained after the production of three consecutive batches of Bulardi® Probiotik, capsules are graphically presented in Figures 1 and 2.

Figure 1. Graphical view of pH values of chemical washings HI1-HI5 during Ethanol

assessment (average values of three consecutive results)

5,8

6

6,2

6,4

6,6

6,8

7

7,2

Washing 1 Washing 2 Washing 3 Washing 4 Washing 5

pH

val

ue

Mean SD



146

Figure 2. Graphical view of conductivity results of chemical washings HI1-HI5 during

Ethanol assessment (average values of three consecutive results) Immediately before production began (up to 1h) microbiological smears were samp-led and aerobic mesophile bacteria count was assessed. Obtained results are shown as a graphical image in Figure 3. Results for the samples of microbiological smears for Ente-robacteriaceae are not graphically presented, and they all equaled less than 1 CFU/cm2.

Figure 3. Graphical view of Aerobic mesoph. bacteria count in microbiological smears

MB1-MB13 during Ethanol assessment (average values of three consecutive results)

0

0,5

1

1,5

2

2,5

3


Con

du

ctiv

ity

(µS

/cm

)

Mean SD

0123456789

Aer

obic

mes

oph

. bac

teri

a (C

FU

/CM

2)

Mean SD



147

Validation assessment Second phase of analysis included the effects of cleaning after the production of three consecutive batches of Bulardi® Probiotik, capsules. Same sampling places from the first testing were used (Table 2). However, certified cleaning agents were used this time (8)

after the status “clean” that lasted 12h± 4h. Visual inspection of the equipment was found to be clean. The results of chemical washings are shown in Figures 4 and 5, whereas the results of microbiological smears- aerobic mesophile bacteria count are shown in Figure 6. Results for the samples of micro-biological smears for Enterobacteriaceae are not graphically presented, and they all equaled less than 1 CFU/cm2 (2, 14).

Figure 4. Graphical view of pH values of chemical washings HI1-HI5 during Validation

assessment (average values of three consecutive results)

Figure 5. Graphical view of conductivity results of chemical washings HI1-HI5 during

Validation assesment (average values of three consecutive results)

6,3

6,4

6,5

6,6

6,7

6,8

6,9


pH

val

ue

Mean SD

0

1

2

3

4


Con

du

ctiv

ity

(µS

/cm

)

Mean SD



148

Figure 6. Graphical view of Aerobic mesoph. bacteria count in microbiological smears MB1-MBI3 during Validation assessment (average values of three consecutive results)

Testing of final products Samples of final products for three consecutive batches of Bulardi® Probiotik, capsu-les during validation assessment, as well as samples of the following final product (Her-biko PropoMucil, capsules) manufactured on the same production line, were tested for microbiology (bacteria count, yeast and mold content) and heavy metal content, accor-ding to the regulations for dietary supplements in the Republic of Serbia (5). Cleaning validation principles were applied in this research during the manufacturing conditions without HVAC system. All of the obtained results, as stated in literature, were analyzed and assessed accor-ding to the regulatory requirements for acceptable contamination levels (2, 5, 10, 12,). Table 2 describes the acceptance criteria for visual and analytic method used during the first and second phase of testing.

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

2

Aer

obic

mes

oph

. bac

teri

a (C

FU

/cm

2)

Mean SD



149

Table 2. Acceptance criteria for visual and analytic method

Acceptance criteria Acceptance criteria limits

Visual acceptance criteria cleaned, dry and well illuminated area ob-served from different sides

ATP monitrong criteria fast determination of microbiological bur-den with ATP smears: ≤200 RLU (21).

Chemical acceptance criteria

acidic and alkaline properties: no change in color after the addition of indicators: methyl orange or bromthymol blue (or pH value between 5.0-7.0) (22)

conductivity: less than 4,3 µS/cm, on 20°C (22).

Microbiological acceptance criteria aerobic mesophile bacteria: ≤ 10 CFU/cm2

and Enterobacteriaceae from 0 to 1 CFU/cm (2, 3, 14).

Comparing the results obtained during Ethanol assessment and Validation assessment of chemical and microbiological parameters the following was concluded: - the results obtained in Ethanol assessment and Validation assessment show large differences regarding the duration period of status “clean”. With Ethanol assessment status clean equaled 1h, while in Validation assessment it equaled 12h ± 4h. - chemical burden samples (acceptance criteria for pH value equaled from 5,0 to 7,0 and acceptance criteria for conductivity equaled no more than 4,3 µS/cm, measured on 20°C) present slightly higher results of conductivity for samples obtained during Vali-dation assessment that points out the need for a more detailed rinsing of equipment due to the difficult removal of certified cleaning agent residues in contrast to the easy removal of residues of ethanol. - microbiological burden of smears for aerobic mesophile bacteria (acceptance criteria of less than 10 CFU/cm2) and for Enterobacteriaceae (acceptance criteria of less than 1 CFU/cm2) were within specified limits on all samples, after performed cleaning process and as well as after defined time of status “clean”. These results comply with the rele-vant literature (14). MB1 sampling place during Ethanol assesment presented a relatively higher microbiological burden (bioburden) probably due to insufficient action of used disinfecting agent (10). - the results of chemical washing and microbiological smears obtained from all three batches were inside the allowable limits. The absence of any residues of active compo-nents of Saccharomyces boulardii was also confirmed, as well as the absence of residues of cleaning agents, lubricants, and other potential contaminants on the manufacturing equipment. Cleaning procedure with certified cleaning agents was deemed appropriate for the worst-case final product, and for all other products manufactured on the same production line for solid dosage forms without HVAC system. It was also confirmed that



150

cross-contamination could not be achieved after performing the defined manufacturing procedures. - microbiological results obtained during Ethanol assessment presented higher values of standard deviation, while the results obtained during Validation assessment show a slightly smaller values of standard deviation (first phase of testing showed 6 results to be above 1 CFU/ cm2, whereas the second phase of testing showed only three results above 1 CFU/ cm2). These results led to a conclusion that higher microbiological efficacy of cleaning was presented after the appliance of specified validation protocol, and also - better uniformity of the process itself. (10). - microbiological burden, or aerobic mesophilic bacteria count in Validation asses-sment, presented the highest values in the following sampling places: MB4 , MB6, and MB10, most probably due to a slightly lesser bactericidal and bacteriostatic actions of the disinfecting agents on less accessible places od sampling.(10) - using validated process of cleaning and disinfection with certified cleaning agents has confirmed that the status period of “clean”, lasted longer than previously defined time period. Microbiological burden of sampling places was slightly improved. - one-time wash and disinfection achieved its goal of production of safe products, together with a more rational resource management in production process. (10) - the obtained results confirmed that the addition of certified detergents and disin-fectants contributes to a more efficient cleaning process, avoids the cleaning and disin-fection of the equipment immediately before the production and extends the status time of “clean”. Status time of “clean” was extended from 1 hour to 12 hours ± 4 hours. - Samples of final products of three consecutive batches of the product Bulardi®

Probiotik, capsules during Validation examination, as well as samples of the next final product (Herbiko PropoMucil, capsules) manufactured on the same production line pre-sented microbiological results (bacteria contents, yeast and mold count) and heavy metal content to be within limits of defined regulatory aspects in the Republic of Serbia; the absence of any residues of active components of Saccharomyces boulardii, as well as the absence of any residues of the cleaning agent, lubricants, and other potential contami-nants on the manufacturing equipment.

CONCLUSION Cleaning validation was implemented by a multidisciplinary team of Abela Pharm d.o.o that contributed to a better understanding of the process in details, thus enhancing the production process for capsulated dietary supplements. The efficacy and status time of “clean” after the cleaning and disinfection process of the equipment with 70% ethanol was compared with cleaning and disinfection process with certified pharmaceutical in-dustry agents. The primary objective of the cleaning validation process has been achieved. Cleaning procedure and status time of “clean” and “dirty” were validated after three consecutive batches for the selected worst case product during which every documented potential contaminants were reduced to predetermined levels of acceptance (18). The obtained



151

results confirmed that the addition of certified detergents and disinfecting agents con-tribute to a more efficient cleaning process and status time “clean” was extended which has proved to be a contributing factor to a more rational management of resources in the production without HVAC system and equipment, space and employee engagement were better utilized. Quality and safety of all dietary supplements manufactured on the production line for solid dosage forms was confirmed in this research. Routine use of this validated process will be continued in order to produce pharmaceutically active, efficient and safe final pro-ducts. Abela Pharm d.o.o. professional team will continue to implement cleaning valida-tion and revalidation procedures on the production line for the manufacture of capsulated dietary supplements.

Acknowledgement Authors declare no funding. There is no conflict of interest.

REFERENCES

1. European Commission, Health and consumers directorate- general, Guidelines to GMP for Medicinal Products for Human and Veterinary Use - Annex 15: Qualifica-tion and Validations; Brussels; 2015; pp 1-16.

2. European Medicines Agency, Guideline on setting health based exposure limits for use in risk identification in the manufacture of different medicinal products in shared facilities; London; 2014; pp 1- 11.

3. Republika Srbija, Ministarstvo poljoprivrede, trgovine, šumarstva i vodoprivrede, Vodič za primenu mikrobioloških kriterijuma za hranu- prvo izdanje; Beograd; 2011; pp 1- 63.

4. Both, C. Harmonized microbial limits testing validation strategies. Peer reviewed: Microbiology, Journal of Validation Technology, 2013, 19(4).

5. Lodhi, B.; Padamwar, P.; Patel. A. Cleaning validation for the pharmaceuticals, bio-pharmaceuticals, cosmetic and neutraceuticals ibndustries, J. Innovations Pharm. Biol. Sci. 2014, 1, 27-38.

6. European Parliament and European council. Regulation EC 1272/2008 of the European Parliament and of the council; Brussels; 2008.

7. The Comission of the European Communities. Commission Directive 2003/94/EC; Brussels; 2003; pp 1-5

8. CLEAN Report Abela Pharm-Borer Chemie AG (July 2014); unpublished work 9. Sandle, T. A practical Approach to the Selection of Cleanroom Disinfectants;

http://www.pharmafocusasia.com/articles/practical-approach; 17th January, 2017. 10. Shashikant B.; Bagade, V.; Siva R.K. Cleaning Validation:Unimportant Aspects in

Accesing Pharmaceutical Residues-Res. J. Pharm., Biol. Chem. Sci.; 2014, 1572-1579.



152

11. Pharmaceutical Inspection Convention. Recommendations on Validation Master Plan, Installation and Operational Qualification, Non Sterile Process Validation, Cleaning Validation; Geneva, 2007; 1-29.

12. Nassani, M. Cleaning Validation in the Pharmaceutical Industry; Journal of Vali-dation Technology; 2005, 38-58.

13. Turner, D.E.; Daugherty, E. K.; Altier, C.; Maurer, K.J. Efficacy of Limitations of an ATP-Based monitoring System, J. Am. Assoc. Lab. Anim. Sci. 2010, pp 190-195

14. Goransson, A.; Petersson K. A systematic approach to food safety, Journal of Hygienic Engeenering and Design, 2012, pp 179-181.

15. International Organization for Standardization; ISO 10523:2008 Water quality-deter-mination of pH; 2008; pp 1- 13.

16. Institut za standardizaciju Srbije; SRPS EN ISO 27888:2009 Kvalitet vode-Određivanje električne provodljivosti; 2009; pp 1- 10.

17. Institut za standardizaciju Srbije; SRPS EN ISO 9963-1:2007 Kvalitet vode-Određi-vanje alkaliteta-Deo 1.Određivanje ukupnog i kompozitnog alkaliteta; 2007; pp 1-13

18. Sutton, S. The Contamination Control Plan in Facility Validation; Journal of Valida-tion Technology, 2012, pp 43- 47.

19. Institut za standardizaciju Srbije; SRPS EN ISO 4833-2 2014 Mikrobiologija lanca hrane-Horizonatalna metoda za određivanje broja mikroorganizama-deo 1:Brojanje kolonija na 30 °C tehnikom nalivanja ploče; 2014; pp 1-21.

20. Institut za standardizaciju Srbije; SRPS ISO 21528-2 2009 Mikrobiologija hrane i hrane za životinje-Horizontalna metoda za otkrivanje i određivanje broja Enterobac-teriacae-Deo 2. Metoda brojanja kolonija; 2009; pp 1- 16.

21. Alfa, M.J.; Iram, F.; Olson, N. The adenosine triphosphate test is a rapid and reliable audit tool to assess manual cleaning adequacy of flexible endoscope channels, Am. J. Infect. Control, 2013, 249-253.

22. European Directorate for the Quality of Medicines and HealthCare Council of Europe. European Pharmacopoeia 8.0th Edition; 2013; pp 1-1456.

ПРИМЕНА ПРИНЦИПА ВАЛИДАЦИЈЕ ЧИШЋЕЊА НА ОПРЕМИ ЗА ПРОИЗВОДЊУ ДИЈЕТЕТСКИХ ПРОИЗВОДА

Давор Ј. Корчок, Нада А. Тршић- Милановић

Abela Pharm д.о.о, Вилине воде б.б, 11000 Београд, Србија

Валидација чишћења опреме у фармацеутској производњи је документовани доказ успешног чишћења и представља један од захтева добре произвођачке праксе у производњи лекова. Успешна валидација потврђује поступке прања, чишћења и дезинфекције опреме и документује резултате хемијских и микробиолошких ана-лиза који су предуслов чисте опреме и безбедног финалног дијететског производа. Циљ овог рада је унапређење чишћења процесне опреме применом поступака ва-лидације чишћења на производној линији за производњу дијететских производа тј.



153

чврстих форми- капсула у Abela Pharm д.о.о. Принципи валидације чишћења се пре свега примењују у производњи лекова. Резултати валидације добијени у овом истраживању представљају практичну анализу процеса чишћења производне опре-ме које је усклађено са регулаторним захтевима. Финална потврда успешно обав-љене валидације чишћења су резултати који потврђују превенцију нежељеног мик-робног раста и уклањање потенцијалне контаминације са прозводне опреме у циљу очувања активности, ефикасности и безбедности дијететског производа. Кључнe рeчи: микробиолошки критеријуми, валидација чишћења, дијететски

производи




155

MOLECULAR DOCKING ANALYSIS OF NEWLY SYNTHESIZED 2-MORPHOLINOQUINOLINE DERIVATIVES WITH ANTIFUNGAL

POTENTIAL TOWARD Aspergillus fumigatus

Strahinja Z. Kovačević*, Milica Ž. Karadžić, Lidija R. Jevrić, Sanja O. Podunavac-Kuzmanović


The present paper is concerned with the molecular docking analysis of newly synthesized 2-morpholinoquinoline derivatives with antifungal potential toward Asper-gillus fumigatus. The purpose of the molecular docking analysis was to determine potential interactions between the analyzed compounds and the active site of the enzyme glucosamine-6-phosphate synthase, as well as to reveal which molecular features (the presence of different substituents or isomers) could be responsible for significant anti-fungal activity of the tested compounds. The compounds with the highest antifungal potential toward pathogenic and saprotrophic fungus Aspergillus fumigatus were docked, and the results were compared with the docking of griseofulvin, which is an antifungal drug used in the treatment of various types of dermatophytoses. Newly discovered anti-fungal agents are very important in medicine, as well as in agriculture. The prevention of the presence of mycotoxins in food and feed products is one of the urgent tasks. Therefore, every effort which leads to discovery of their mechanism of action and deter-mination of side effects is precious. The present study can be considered a contribution to the analysis and selection of newly discovered antifungals from the 2-morpholino-quinoline family, and one step forward to their practical use in medicine and agriculture. The obtained results reveal the possible reason why the newly synthesized 2-morpho-linoquinoline expresses even higher antifungal activity toward Aspergillus fumigatus than griseofulvin, a commercially available antifungal therapeutic.

KEY WORDS: antifungal activity, Aspergillus fumigatus, fungi, molecular docking, molecular modeling.

INTRODUCTION

The presence of mycotoxins in food and feed products has become a global problem. These secondary metabolites of different fungi (usually from the genera Aspergillus, Penicillium and Fusarium) have an enormous influence on human health and agriculture. Aspergillus fumigatus is a fungus often isolated from decomposed plant material. Also, it is isolated from cereals, including wheat, rice, barley, soybeans and beans, and also from * Corresponding author: Strahinja Kovačević, University of Novi Sad, Faculty of Technology Novi Sad,

Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia, e-mail: [email protected]



156

meat products, milk, and dairy products (1). The optimal temperature range for the growth of this fungus is between 40 and 42 °C (1). Fumitoxin, helvolic acid, fumagillin, gliotoxin, fumitremorgins, monotrypacidin, fumigaclavines, fumigatin, tryptoquivaline are the secondary metabolites commonly found in Aspergillus fumigatus (2). Some of the toxins produced by this fungus have immunosuppressive features. They can induce apop-tosis in particular cells of the immune system. Aspergillus fumigatus has become one of the most important airborne pathogens in developed countries (2). It can cause some allergic reactions in humans as well (1). Due to high toxicity of the metabolites produced by Aspergillus fumigatus, it is highly important to prevent its presence in food and feed products. Molecular docking and computational chemistry, including different chemometric tools, are crucial in Computer-Aided Drug Design (CADD). They have become a very important factor in the design and development of novel compounds with wide spectrum of biological activity, including anticancer, antibacterial and antifungal activities (3-5). The application of CADD approaches in early phases of development of biologically active compounds can make the search for leading compounds faster. Molecular docking is a CADD tool aimed to predict the possible interactions and predominant binding mode between a ligand and the virtual screening of a huge number a key protein. Its role in modern drug design is very significant in the virtual screening a huge number of molecu-les and their ranking (6). The molecular docking approach has been successfully applied in several studies to analyze antifungal activity of different compounds, such as triazole derivatives (7), 2-(2,4-difluorophenyl)-2-hydroxy-3-(1H-1,2,4-triazol-1-yl)propyl dithio-carbamates (8), 1,3-thiazole derivatives (9), Ni2+, Co2+ and Cu2+ complexes of porphyrin core macromolecular ligand (10). The present paper describes the application of the molecular docking procedure in the analysis of interactions between the newly synthesized 2-morpholinoquinoline deriva-tives and active site of glucosamine-6-phosphate synthase enzyme, which is considered to be the potential target for antifungal and antibacterial activity (11, 12). The main aim was to determine optimal positions of the analyzed ligands into the binding pocket of the target enzyme, to analyze the important interactions between the ligands and the enzyme, and to gain an overview on the molecular characteristics that contribute to their anti-fungal activity.


The studied 2-morpholinoquinoline analogs and their antifungal activity against Aspergillus fumigatus

The analyzed 2-morpholinoquinoline analogs, which contain 1,2,4-oxadiazole, were synthesized by Karad et al (13). The results of antifungal activity of the studied analogs toward Aspergillus fumigatus have been taken from the literature (13). The antifungal ac-tivity was expressed as minimum inhibitory concentration (MIC) in mM units. The com-pounds of the highest interest were selected on the basis of the lowest MIC values. Ac-



157

cording to this criterion, the analysis included the following compounds: (1) 2-(morpho-lin-4-yl)-3-{5-[4-(trifluoromethyl)phenyl]-1,2,4-oxadiazol-3-yl}quinoline, MIC = 0.146 mM (13), (2) 3-[5-(4-chlorophenyl)-1,2,4-oxadiazol-3-yl]-6-methyl-2-(morpholin-4-yl)quinolone, MIC = 0.246 mM (13) and (3) 2-(morpholin-4-yl)-3-{5-[3-(trifluorometh-yl)phenyl]-1,2,4-oxadiazol-3-yl}quinolone, MIC = 0.234 mM (13). The analysis included griseofulvin (MIC = 0.283 mM) (13) as well, as a commercially available antifungal drug. The molecular structures of the studied compounds are presented in Figure 1. The compounds 1 and 3 are structural isomers which differ in the position of 4-trifluoro-methyl group in the phenyl group. By comparing the compound 2 with the compounds 1 and 3, it can be seen that the compound 2 has an additional methyl group in the position 6 and chlorine in the position 4 in the phenyl group. The above compounds can be con-sidered to have high antifungal potency toward Aspergillus fumigatus fungus (13).

Figure 1. Molecular structures of the analyzed 2-morpholinoquinoline analogs (1 - 3) and griseofulvin

Molecular docking analysis of 2-morpholinoquinoline analogs and griseofulvin

The molecular docking modeling was done by using SwissDock software based on a web server EADock DSS (14). The CHARMM force field was used for the calculations (15), while the ranking was done on the basis of the most favorable binding energies. The X-ray structure of glucosamine-6-phosphate synthase enzyme was retrieved from the Protein Data Bank (PDB: http://www.rcsb.org/, PDB code 2VF5) (16). The analyzed ligands were energetically minimized by MM2 method, applying ChemBio3D software (17). The UCSF Chimera program was used for the visualization of the docking results (18). In order to validate the docking results, the molecular docking procedure was carried out in triplicate. The ligand-enzyme interactions were described by the following energies: Gibbs free energy (ΔG), van der Waals energy (ΔGvdw), protein-solvent non-



158

polar energy (ΔGprotsolvnonpol), ligand-solvent non-polar energy (ΔGligsolvnonpol), ΔGelec ener-gy, ligand-solvent polar energy (ΔGligsolvpol), protein-solvent polar energy (ΔGprotsolvpol), compound-solvent non-polar energy (ΔGcompsolvnonpol), surface-full energy, extra-full ener-gy, protein-solvent polar energy (ΔGcompsolvpol), intra-full energy, solvent-full energy, full-fitness energy, inter-full energy, total energy, and simple-fitness energy. According to the literature (19), in molecular docking studies the focus on ΔG is still considered “the safest bet”. Therefore, the results were interpreted by means of the ΔG values.

Glucosamine-6-phosphate synthase (PDB: 2VF5) According to the X-ray analysis, in the binding pocket of glucosamine-6-phosphate synthase there are twelve amino acids (VAL399, GLN348, LYS603, GLY301, ALA602, ALA400, THR352, THR302, SER303, SER349, SER347 and CYS300) (11). This enzy-me is known as L-glutamine : D-fructose-6-phosphate amidotransferase or under com-monly used name glucosamine-6-phosphate synthase. It is a key enzyme in the catalysis of the reaction which involves ammonia transfer from L-glutamine to fructose-6-phos-phate and following isomerization of the formed fructosamine-6-phosphate to glucosa-mine-6-phosphate (12). Earlier studies have shown that even a short-time inactivation of this enzyme in the cells of fungi is lethal due to induction of the lysis, agglutination, and certain morphological changes (12). The structure of glucosamine-6-phosphate synthase in the complex with glucosamine-6-phosphate, obtained by X-ray crystallography, is presented in Figure 2.

Figure 2. The structure of glucosamine-6-phosphate synthase in the complex with

glucosamine-6-phosphate (PDB: 2VF5), which is marked by cyan color, retrieved from Protein Data Bank - PDB (http://www.rcsb.org)



159


The molecular docking analysis of griseofulvin

The results of the docking of griseofulvin are given in Figure 3. It was determined that there are two main strong interactions between the ligand and the active site. The first one is a strong hydrogen bond between H-atom (HN) of THR302 and O-atom (O5) of the ligand. The length of this intra-molecular bond is 2.438 Å. The second one is a stronger H-bond than the previous one, with the length of 1.938 Å, between H-atom (HG1) of THR352 and O-atom (O6) of the ligand. The ΔG value of this interactions is -6.98 kcal/mol, followed by full fitness energy of -1643.10 kcal/mol and total energy of 25.26 kcal/mol. Also, there are certain lipophilic interactions present between the ligand and the active site; however, the hydrophilic interactions are predominant due to the chemical nature of the ligand, which can be considered more hydrophilic due to the presence of polar groups in its structure, such as metoxy, oxo and chlorine substituents.

Figure 3. The docking of griseofulvin (cyan color) in the active site of glucosamine-6-

phosphate synthase

The molecular docking analysis of 2-(morpholin-4-yl)-3-{5-[4-(trifluoromethyl)phenyl]-1,2,4-oxadiazol-3-yl}quinoline (1)

The compound 1 was docked in the active site of glucosamine-6-phosphate synthase and it was determined that it forms several strong bonds with the active site (Figure 4). The presence of p-trifluoromethyl group enables several strong interactions between the ligand and active pocket of the enzyme. It can possibly form six strong H-bonds with the enzyme, concretely:

three H-bonds between the F2-atom of the ligand and H-atom of THR352 (HG1) with the length of 3.120 Å, H-atom of SER349 (HN) with the length of 2.450 Å and H-atom of GLN348 (HN) with the length of 2.523 Å;

two H-bonds between the F3-atom of the ligand and H-atom of THR352 (HG1) with the length of 2.543 Å, and H-atom of GLN348 (HN) with the length of 2.857 Å;



160

one H-bond between the F1-atom of the ligand and H-atom of GLN348 (HN) with the length of 2.495 Å.

a) b)

Figure 4. The docking of the compound 1 (cyan color) in the active site of glucosamine-6-phosphate synthase (a), and the zoomed interactions between the p-trifluoromethyl

group and corresponding residues.

Besides the H-bonds formed between p-trifluoromethyl group and the binding site, there are two more significant interactions: (1) relatively weak H-bond formed between O-atom (O2) of the ligand and H-atom of SER401 (HN) with the length of 3.685 Å; (2) H-bond between N-atom (N3) of the ligand and H-atom (HN) of ALA602 of the length of 3.188 Å. The ΔG value of the interactions between the compound 1 and the binding site are described by a ΔG value of -7.54 kcal/mol, full fitness energy of -1609.35 kcal/mol, and total energy of 36.36 kcal/mol. There are some weak lipophilic interactions between the non-polar rings of the ligand and certain residues of the binding pocked; however, the hydrophilic interactions are predominant. The p-trifluoromethyl group plays a very significant role in these interac-tions, enabling strong connections between the ligand and the residues of the binding pocket.

The molecular docking analysis of 3-[5-(4-chlorophenyl)-1,2,4-oxadiazol-3-yl]-6-methyl-2-(morpholin-4-yl)quinoline (2)

The compound 2 fits very well into the binding pocket of glucosamine-6-phosphate synthase; however it forms less inter-molecular bonds than the compound 1 (Figure 5). The compound 2, compared with the compound 1, has a chlorine substituent instead of the trifluoromethyl group in the position 4 of the 5-phenyl ring. Therefore, no significant interactions between the chlorine substituent and the neighboring residues have been determined. Nevertheless, there are some significant interactions between the ligand and the binding pocket which should be mentioned:

strong H-bond between the O-atom (O2) of the ligand and H-atom (HN) of ALA602, with the length of 2.218 Å;

relatively strong H-bond between the O-atom (O1) of the ligand and H-atom (HG1) of THR352, with the length of 3.317 Å.



161

The ΔG value of the present interactions between the compound 2 and the binding pocket is -7.32 kcal/mol. The full fitness energy is -1628.06 kcal/mol, and the total energy is 36.19 kcal/mol. As in the case of the compound 1, some weak lipophilic inte-ractions between the non-polar rings of the ligand and certain residues of the binding pocket can be observed, but the hydrophilic interactions are predominant.

Figure 5. The docking of the compound 2 (cyan color) in the active site of glucosamine-6-phosphate synthase

The molecular docking analysis of 2-(morpholin-4-yl)-3-{5-[3-(trifluoromethyl)phenyl]-1,2,4-oxadiazol-3-yl}quinoline (3)

Despite the fact that the compound 1 and compound 3 are structural isomers, which differ only in the position of trifluoromethyl group in 5-phenyl ring, the difference between their antifungal activities against Aspergillus fumigatus is evident. This could be explained by the assumption that the p-trifluoromethyl group has much more possibility to interact with the residues than the m-trifluoromethyl group, as it is shown in Figure 4b. Figure 6 indicates that the compound 3 forms four strong H-bonds between:

F1-atom of the ligand and H-atom of THR352 (HG1) with the length of 2.805 Å; F2-atom of the ligand and H-atom of GLN348 (HG1) with the length of 2.402 Å; F3-atom of the ligand and H-atom of GLN348 (HE22) with the length of 2.773 Å; O2-atom of the ligand and H-atom of ALA602 (HN) with the length of 2.884 Å.

The docking of the compound 3 is described by a ΔG value of -7.33 kcal/mol. The full fitness energy is -1609.27 kcal/mol, and the total energy is 38.61 kcal/mol. The lipo-philic interactions between the non-polar rings of the ligand and certain residues of the binding pocket are observable; however, in this case as well, the hydrophilic interactions are the most significant.



162

Figure 6. The docking of the compound 3 (cyan color) in the active site of glucosamine-6-phosphate synthase

CONCLUSION

The presented results of the molecular docking analysis of the newly synthesized 2-morpholinoquinoline derivatives with antifungal potential against Aspergillus fumigatus are in agreement with the experimental findings. Namely, the analyzed compounds 1 - 3 expressed higher antifungal activity than griseofulvin, which is a commercially available antifungal therapeutic. This could be explained by the interactions which the newly synthesized derivatives realize with the binding pocket of the corresponding enzyme (glucosamine-6-phosphate synthase). Therefore, on the basis of the obtained results it can be concluded: The compound 1 expresses the highest antifungal potential against Aspergillus

fumigatus, possibly due to the numerous strong H-bonds particularly achieved between the p-trifluoromethyl group and corresponding residues of the binding pocket;

The compound 3 is a structural isomer of the compound 1; however, its antifungal activity is much lower than the antifungal activity of the compound 1. This could be explained by a higher possibility of the p-trifluoromethyl group in the com-pound 1 to form more H-bonds with the binding pocket than the m-trifluoromethyl group in the compound 3;

The compound 2, due to the nature of the substituent in the position 4 in the 5-phenyl ring, has the lowest antifungal activity, compared to the compounds 1 and 3, but still higher than the antifungal activity of griseofulvin;

The calculated ΔG energies are in agreement with the experimental results of antifungal activity. The docking of the compound 1, as the compound with the highest antifungal activity, showed the lowest ΔG value, and the docking of gri-seofulvin, as the compound with the lowest antifungal activity, was characterized by the highest ΔG energy. Therefore, the most active compound occupies the most energetically favorable position in the binding pocket.



163

The interactions between the ligands and the active pocket are predominantly hydrophilic;

The present results can be considered to be guidelines for further syntheses of 2-morpholinoquinoline derivatives structurally similar to the studied ones, as well as a con-tribution to the search of novel potent antifungal compounds which could find a wide spectrum of practical applications in medicine and agriculture.

Acknowledgement

This study was financially supported by the research project of the Provincial Secretariat for Higher Education and Scientific Research of the AP Vojvodina (Project No. 142-451-2604/2017-01/01).

REFERENCES

1. Škrinjar, M.; Tešanović, D. Hrana u ugostiteljstvu i njeno čuvanje; Univerzitet u Novom Sadu, Prirodno-matematički fakultet: Novi Sad, 2007; pp 42-43.

2. Latgé, J. P. Aspergillus fumigatus and Aspergillosis. Clin Microbiol Rev. 1999, 12, 310-350.

3. Karadžić, M. Ž.; Jevrić, L. R; Mandić, A. I.; Markov, S. L.; Podunavac-Kuzmanović, S. O.; Kovačević, S. Z.; Nikolić, A. N.; Oklješa, A. O.; Sakač, M. N.; Penov Gaši, K. M. Chemometrics approach based on chromatographic behavior, in silico characterization and molecular docking study of steroid analogs with biomedical importance. Eur J Pharm Sci. 2017, 105, 71-81.

4. Kovačević, S. Z.; Podunavac-Kuzmanović, S. O.; Jevrić, L. R.; Vukić, V. V.; Savić, M. P.; Djurendić, E. A. Preselection of A- and B- modified D-homo lactone and D-seco androstane derivatives as potent compounds with antiproliferative activity against breast and prostate cancer cells - QSAR approach and molecular docking analysis. Eur J Pharm Sci. 2016, 93, 107-113.

5. Kovačević, S. Z.; Podunavac-Kuzmanović, S. O.; Jevrić, L. R. Multivariate regression modelling of antifungal activity of some benzoxazole and oxazolo[4,5-b]pyridine derivatives. Acta Chim Slov. 2013, 60, 756-762.

6. Morris, M. M.; Lim-Wilby, M. Molecular Docking. In Molecular Modeling of Proteins; Kukol, A., Ed.; Methods Molecular Biology™, vol 443; Humana Press, 2008; pp 365-382.

7. Wang, N.; Chai, X.; Chen, Y.; Zhang, L.; Li, W.; Gao, Y.; Bi, Y.; Yu, S.; Meng, Q. Synthesis, Antifungal Activity, and Molecular Docking Studies of Novel Triazole Derivatives. Med Chem. 2013, 9, 384-388.

8. Zou, Y.; Yu, S.; Li, R.; Zhao, Q.; Li, X.; Wu, M.; Huang, T.; Chai, X.; Hu, H.; Wu, Q. Synthesis, antifungal activities and molecular docking studies of novel 2-(2,4-difluorophenyl)-2-hydroxy-3-(1H-1,2,4-triazol-1-yl)propyl dithiocarbamates. Eur J Med Chem. 2014, 74, 366-374.



164

9. Łączkowski, K. Z.; Misiura, K.; Biernasiuk, A.; Malm, A; Paneth, A.; Plech, T Syn-thesis, Antimicrobial Activity and Molecular Docking Studies of 1,3-Thiazole De-rivatives Incorporating Adamantanyl Moiety. J Heterocycl Chem. 2016, 53, 441-448.

10. Singh, U.; Malla, A. M.; Bhat, I. A.; Ahmad, A.; Bukhari, M. N.; Bhat, S.; Anayutullah, S.; Hashmi, A. A. Synthesis, molecular docking and evaluation of anti-fungal activity of Ni(II),Co(II) and Cu(II) complexes of porphyrin core macromo-lecular ligand. Microb Pathog. 2016, 93, 172-179.

11. Agrawal, R.; Tarannum, N.; Mishra, S.; Kumar Soni, R. Antimicrobial Activity and Molecular Modelling of Aminolysed Derivative of Poly(ethyleneterephthalate). Pharma Chem. 2016, 8, 132-139.

12. Wojciechowski, M.; Milewski, S.; Mazerski, J.; Borowski, E. Glucosamine-6-phos-phate synthase, a novel target for antifungal agents. Molecular modelling studies in drug design. Acta Biochim Pol. 2005, 52, 647-653.

13. Karad, S. C.; Purohit, V. B.; Thummar, R. P.; Vaghasiya, B. K.; Kamani, R. D.; Thakor, P.; Thakkar, V. R.; Thakkar, S. S.; Ray, A.; Raval, D. K. Synthesis and biological screening of novel 2-morpholinoquinoline nucleus clubbed with 1,2,4-oxadiazole mofits. Eur J Med Chem. 2017, 126, 894-909.

14. Grosdidier, A.; Zoete, V.; Michielin, O. SwissDock, a protein-small molecule docking web service based on EADock DSS. Nucleic Acids Res. 2011, 39, 270-277.

15. Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.; Kundu, S.; Zhong, S.; Shim, J.; Darian, E.; Guvench, O.; Lopes, P.; Vorobyov, I.; Mackerell Jr, A. D. CHARMM General Force Field (CGenFF): A force field for drug-like molecules compatible with the CHARMM all-atom additive biological force fields. J Comput Chem. 2010, 31, 671-690.

16. Mouilleron, S.; Badet-Denisot, M. A.; Golinelli-Pimpaneau, B. Ordering of C-Termi-nal Loop and Glutaminase Domains of Glucosamine-6-Phosphate Synthase Promotes Sugar Ring Opening and Formation of the Ammonia Channel. J Mol Biol. 2008, 377, 1174-1185.

17. ChemBioOffice 2012, Perkin Elmer Informatics, http://www.cambridgesoft.com/ 18. Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.;

Meng, E. C.; Ferrin, T. E. UCSF Chimera - A Visualization System for Exploratory Research and Analysis. J Comput Chem. 2004, 25, 1605-1612.

19. Bissantz, C.; Kuhn, B.; Stahl, M. A Medicinal Chemist’s Guide to Molecular Inte-ractions. J Med Chem. 2010, 53, 5061-5084.



165

АНАЛИЗА МОЛЕКУЛСКОГ ДОКИНГА НОВОСИНТЕТИСАНИХ 2-МОРФОЛИНОХИНОЛИНСКИХ ДЕРИВАТА СА АНТИФУНГАЛНИМ

ПОТЕНЦИЈАЛОМ ПРЕМА Aspergillus fumigatus

Страхиња З. Ковачевић, Милица Ж. Караџић, Лидија Р. Јеврић, Сања О. Подунавац-Кузмановић

Универзитет у Новом Саду, Технолошки факултет Нови Сад, Булевар цара Лазара 1,

21000 Нови Сад, Србија У овом раду представљена је анализа молекулског докинга новосинтетисаних 2-морфолинохинолинских аналога, који показују антифунгалну активност према As-pergillus fumigatus. Циљ ове анализе је да се установе потенцијалне интеракције из-међу анализираних аналога и активног места ензима глукозамин-6-фосфат синтазе, као и да се утврде које молекулске карактеристике (присуство различитих супсти-туената или изомерија) могу бити одговорне за значајну антифунгалну активност испитиваних једињења. Једињења са највишом антифунгалном активношћу према патогеној плесни Aspergillus fumigatus докингована су у активно место ензима, а ре-зултати су поређени са резултатима молекулског докинга грисеофулвина, антифун-галног терапеутика који се примењује за третман различитих типова дерматофи-тоза. Новооткривена антифунгална једињења су значајна како за медицину, тако и за пољопривреду. Превенција појаве микотоксина у прехрамбеним производима и храни за животиње један је од ургентних задатака. Стога, сваки допринос открива-њу механизма деловања новооткривених једињења и могућих штетних ефеката је драгоцен. Ова студија даје допринос анализи и селекцији новосинтетисаних анти-фунгалних једињења из групе 2-морфолинохинолина и пружа корак напред ка њи-ховој практичној примени у медицини и пољопривреди. Резултати представљени у овом раду откривају могући узрок јаче антифунгалне активности анализираних 2-морфолинохинолинских аналога према плесни Aspergillus fumigatus од антифунгал-не активности грисеофулвина, као комерцијално доступног антифунгалног тера-пеутика. Кључне речи: aнтифунгална активност, Aspergillus fumigatus, молекулски докинг,

молекулско моделовање, плесни

Received: 29 August 2017. Accepted: 05 October 2017.

APTEFF, 48, 1-323 (2017) UDC: 637.12:543.4 https://doi.org/10.2298/APT1748167K BIBLID: 1450-7188 (2017) 48, 167-175


167

VALIDATION AND APPLICATION OF FTIR SPECTROSCOPY IN RAW MILK ANALYSIS

Denis S. Kučević*, Dragan M. Glamočić, Snežana J. Trivunović, Igor M. Jajić,

Biljana Č. Perišić, Saša Z. Krstović

University of Novi Sad, Faculty of Agriculture, Trg Dositeja Obradovića 8, 21000 Novi Sad, Serbia

The aim of this study was to investigate whether FTIR spectroscopy is an accurate and valid technique for the assessment of quality parameters in raw cow's milk: fat, protein, lactose, and total solids. The assessment was based on calibration series and comparison with reference material. Furthermore, it takes into account the results obtai-ned in the inter-laboratory comparisons (proficiency testing). The calibration samples were purchased from the accredited regional reference laboratories. The validation pa-rameters included linearity, accuracy, repeatability, reproducibility, and robustness. The linearity ratio was 0.95%. The biases calculated for the fat, protein, lactose and dry matter were -0.33, 0.31, -0.25, and 0.06 respectively. The F value from the F-test was used to determine the significant differences between two independent sets of the results. The obtained results were as follows: 1.469 for fat, 1.634 for protein, 1.192 for lactose, and 0.528 for dry matter. The intra-laboratory reproducibility calculated as the Horwitz Ratios for all parameters were within the criterion limits (0.5 to 0.8). The data obtained for carry-over were 0.27% for fat, 0.52% for protein, 0.47% for lactose, and 0.47% for dry matter. Based on the obtained results it can be concluded that the FTIR spectroscopy is a reliable instrumental technique for the determination of fat, protein, lactose and total solids in raw cow's milk. KEY WORDS: FTIR spectroscopy, milk, chemical parameters, validation

INTRODUCTION Fourier transform infrared (FTIR) spectroscopy is routinely used by laboratories spe-cialized in milk analysis, whether for payment or for the purpose of Dairy Herd Improve-ment (DHI) program, because it is a fast, nondestructive, and easy procedure that enables simultaneous measurement of several components in a complex natural media (1, 2). FTIR spectroscopy is a rapid, inexpensive, and sensitive technology used for a high-throughput analysis of food components. This technique is also widely utilized for rapid quality control of numerous dairy products, since it provides fingerprint spectral informa-tion about the composition of complex samples (3, 4). In particular, FTIR spectroscopy is * Corresponding author: Denis S. Kučević, University of Novi Sad, Faculty of Agriculture, Trg Dositeja Ob-

radovića 8, 21000 Novi Sad, Serbia, e-mail: [email protected]



168

commonly used in milk analysis to obtain information related to composition and confor-mation of their components. However, this is an indirect method that requires instrument calibration with milk samples that have reference values established by reference chemis-try methods (5-7). By simultaneously exposing the sample to a beam with several diffe-rent frequencies in the mid-infrared zone, it is possible to significantly reduce the time needed to analyze the sample. By analyzing certain points in the spectrum, the instrument is able to determine the amount of each component in the sample (2). The FTIR spectro-scopy is not only fast, but also enables additional parameter analysis, just by providing reliable calibration samples. In addition, up to 1060 different wavelengths can be selected from the full spectrum, to open an entire range of possibilities for new parameters and accurate analysis. In milk analysis, mid-IR is routinely used for payment, dairy herd recording, and qua-lity control. Analytical speed is a function of the number of components analyzed such as fat, protein, lactose, and other solid components (8). Taking approximately 10 seconds per sample, FTIR spectroscopy have the potential for routine analysis if proper calibrati-on and validation procedures along with data acquisition protocols can be established (9). The aim of this paper was to validate FTIR spectroscopy and its appropriateness in milk analysis. For the validation purpose, secondary reference materials of milk purcha-sed from regional reference laboratories, were used. The validation parameters were tes-ted and then confirmed by analyzing routine samples. After that, the FTIR spectroscopy was used to analyze milk samples from three farms.

EXPERIMENTAL

Material, apparatus and samples

All the solutions used (Zero, Rinse and FTIR Equalizer), were obtained from Foss Analytical. Deionized water was produced using a Barnstead Smart2Pure Water Purifica-tion System (Thermo Scientific, USA). The FTIR measurements were performed on a MilkoScan FT+ (Foss Analytical, Den-mark), equipped with a Conveyer 4000 and Pipette case 4000. Water bath (Elektron, Banja Koviljača, Serbia) was used for sample heating. For calibration purposes, the milk samples (used as a secondary reference material) preserved with bronopol, were purchased from the regional reference laboratory (Institute of Dairy and Probiotics, Biotechnical Faculty, Slovenia). The calibration sample set con-sisted of 10 samples for chemical parameters, covering the entire range of measurement. For accuracy validation purposes, the samples set up by the German laboratory were used (Muva, Kempten, GmbH). Samples for precision testing were made from milk samples collected on commercial farms. The samples were analyzed in duplicates for the selection of representative values for the observed parameter. Finally, composite samples with similar values were formed and then analyzed. The validated method was applied for the determination of fat, protein, lactose, and total solids in a total of 1494 raw milk samples collected from different field sources.



169

Procedure

Prior to measurement, the milk samples (whether calibration samples or natural milk samples) were heated up to 40±2 °C in a water bath. After heating, the samples were in-serted in a rack placed on a conveyer belt. The conveyer forwarded the rack to a stirrer and further to the sampling pipette. A 5-ml sample was taken and thoroughly homogeni-zed by a homogenizer, passed through a cuvette and analyzed. Before calibration, it was necessary to carry out standardization, which included an analysis of the FTIR Equalizer solution. All procedures were performed in accordance with the manufacturer's recommenda-tions, and an international standard (10) and ICAR guidelines, section 12 (11).

Method validation

Tested and validated analytical procedures are important tools for the quality assuran-ce and providing “fit for purpose” analytical measurements. For this purpose, the follow-ing parameters were tested: linearity - using calibration sample sets, accuracy - using re-ference material, precision, and robustness, both with the representative natural samples.

Linearity

The purpose of linearity testing was to document the linear range of the method, the range in which simple calibration can be used. Every single calibration was performed by triple measurement of calibration standards. For this study, the object of the analysis was one sample set with 10 solutions of known concentration, covering the typical range for the specific component. The obtai-ned data were presented with the expected values on the x-axis and the measured values on the y-axis. Additionally, the residuals were calculated. The linear regression and resi-duals data from the curve were expressed as the ratio r. The ratio value should be less than 2% (10).

Accuracy

Accuracy may be defined as the proximity of agreement between the measured quantity value and the true quantity value of a reference material provided by a reference laboratory. It can be presented as a recovery percent or as bias. The bias is a quantitative term describing the difference between the average of measurements made on the same object and its true value. The bias was calculated during the study. These parameters were estimated by analyzing the samples intended for Proficiency testing (Muva Kempten, Germany). The analysis was carried out fourteen times in two days.

Precision

The precision procedure and acceptance criteria were done according to Rivera and Rodríguez (12). The precision was estimated by measuring the repeatability and the



170

reproducibility. Both precision parameters were evaluated using representative raw milk samples collected from the commercial farms in the province of Vojvodina, as previously described. Repeatability is defined as the absolute difference between two independent single test results. It involves the determination using the same test method, on an identical ma-terial, in the same laboratory, by the same operator, using the same equipment within a short interval of time. For the determination of repeatability, 30 samples were analyzed in duplicate. The results were processed using Statistica data analysis software system, ver-sion 12 (StatSoft, Inc., 2013) by performing the F test. Intra-laboratory reproducibility was performed with two analyzers, using the same method, on the identical test material in the same laboratory, during a longer period. For the determination of repeatability, 60 samples were analyzed in triplicate.

Carry-over Carry-over is a measure of the extent of contamination of a sample caused by the previous sample. In terms of the method robustness, carry-over was tested on six series of representative samples, collected from commercial dairy farms in the region. One series consisted of two milk samples and two blank samples; six series were analyzed in se-quence and the carry-over was calculated according to ISO standard (10). The carry-over factor is used automatically to compensate for carry-over from sample to sample during the analysis. That means how much sample volume is carried over from sample to samp-le, specified as a percentage.


Linearity Fat content in the externally obtained calibration sample set with 10 solutions, and the values of the slope, intercept and calibration ratio for fat are given in Table 1.

Table 1. Average fat content in the analyzed externally obtained calibration sample set and obtained linearity parameters.

No Average Reference Proposal Residual 1 2.12 2.10 2.10 -0.02 2 2.57 2.56 2.56 -0.01 3 3.11 3.10 3.11 0.00 4 3.48 3.48 3.47 -0.01 5 4.30 4.30 4.30 0.00 6 3.49 3.50 3.49 0.00 7 4.40 4.40 4.40 0.00 8 4.21 4.22 4.22 0.01 9 4.52 4.55 4.53 0.01 10 5.27 5.26 5.28 0.01

r = 0.95%



171

The linearity and the coefficient of regression between reference and measured values were calculated. The determined average milk fat values ranged from 2.12 to 5.27%. These values correspond to the reference values, which ranged from 2.10 to 5.26%. The calculated r ratio was 0.95, which meets the requirement of less than 2% (10).

Accuracy After the analysis of the reference samples (Muva, Germany), the average value and bias for all parameters were calculated according to Magnusson et al. (13), and the results are presented in Table 2. Table 2. Bias values of the chemical parameters (fat, protein, lactose and dry matter) in

milk (N=14)

As it is shown in Table 2, the bias among average and reference values (-0.33 for fat, 0.31 for protein, -0.25 for lactose and 0.06 dry matter) can be considered as acceptable according to Little (14).

Precision After statistical analysis of the results for repeatability testing, the obtained F critical values were 1.469 for fat, 1.634 for protein, 1.192 for lactose and 0.528 for total solids (data not shown). The obtained data, 0.15, 0.09, 0.32 and 0.05 for fat, protein, lactose and total solids, respectively, show that the variances were not statically different (p < 0.05). Also, the correct F value (variance 1 and variance 2 ratio) was calculated (F = 0.579). The acceptance criterion, where the calculated F is smaller than the critical F, was met in all four tests. Statistical analysis of the results for intra-laboratory reproducibility testing are shown in Table 3. The HORRAT values are 0.5 for fat, 0.8 for protein, 0.7 for lactose, and 0.7 for dry matter. Since all values are between 0.5 and 2, the intra-laboratory reproducibility can be considered as acceptable (12).

Parameter Fat (%) Protein (%) Lactose (%) Dry matter (%) Minimum 4.07 3.59 4.59 9.21 Maximum 4.10 3.61 4.62 9.22 Average value 4.084 3.602 4.605 9.215 Reference value 4.097 3.591 4.616 9.210 Bias -0.33 0.31 -0.25 0.06



172

Table 3. Reproducibility for the determination of fat, protein, lactose and dry matter in routine raw milk samples (N=60)

SD - standard deviation; RSD - relative standard deviation; PRSD - predicted relative standard deviation; HORRAT(R) - Horwitz Ratio (RSD/PRSD).

Regardless of the number of samples processed per time unit, a high inter-laboratory reproducibility observed in the analysis of routine samples.

Carry-over As can be seen in Table 4, the carry-over values for fat, protein, lactose and dry mat-ter were 0.27%, 0.52%, 0.47% and 0.47%, respectively, and they are all considered ac-ceptable, since they were within the criteria of less than ±1% (10).

Table 4. Robustness of the carry-over

Info Fat Protein Lactose Total solids Water 0.17 0.25 -0.14 1.46 Water 0.17 0.26 -0.15 1.46 Milk 3.88 3.13 4.59 12.36 Milk 3.88 3.16 4.61 12.43 Water 0.21 0.31 -0.1 1.61 Water 0.18 0.27 -0.13 1.49 Milk 3.88 3.14 4.58 12.38 Milk 3.89 3.17 4.61 12.45 Water 0.21 0.31 -0.09 1.61 Water 0.18 0.28 -0.13 1.52 Milk 3.89 3.15 4.58 12.38 Milk 3.9 3.17 4.62 12.45 Water 0.21 0.34 -0.15 1.81 Water 0.19 0.29 -0.14 1.53 Milk 3.88 3.15 4.6 12.38 Milk 3.89 3.17 4.61 12.45 Water 0.22 0.33 -0.1 1.64 Water 0.19 0.29 -0.15 1.57 Milk 3.88 3.15 4.6 12.38 Milk 3.89 3.17 4.62 12.46 Carry Over 0.27% 0.52% 0.47% 0.47%

Parameter Fat Protein Lactose Dry matter Criterion Average value 3.983 3.288 4.375 11.137 - SD 0.034 0.041 0.051 0.104 - RSD 0.854 1.247 1.166 0.934 - PRSD 1.62 1.67 1.60 1.39 RSD<PRSD HORRAT(R) 0.5 0.8 0.7 0.7 0.5<HORRAT<2



173

To confirm the applicability of FTIR spectroscopy method, raw milk samples from the three different farms were analyzed, and the results are shown in Table 5.

Table 5. Results of milk parameters in samples from three different farms

Milk Parameters FARM A Fat (%) Protein (%) Lactose (%) Total solids (%) Total No of samples 275 MIN 2.02 2.63 3.47 10.02 MAX 8.75 4.81 5.21 16.46 AVERAGE 4.06 3.41 4.68 12.35 SD 0.95 0.31 0.26 0.95 RSD 0.23 0.09 0.06 0.08 MEDIAN 3.96 3.42 4.73 12.37

FARM B Fat (%) Protein (%) Lactose (%) Total solids (%) Total No of samples 403 MIN 2.24 2.47 3.14 9.74 MAX 6.92 5.05 5.07 16.32 AVERAGE 4.39 3.51 4.62 13.37 SD 0.75 0.45 0.27 0.99 RSD 0.17 0.13 0.06 0.07 MEDIAN 4.27 3.47 4.67 13.24

FARM C Fat (%) Protein (%) Lactose (%) Total solids (%) Total No of samples 816 MIN 2.41 2.37 3.34 10.79 MAX 6.96 4.73 5.24 15.88 AVERAGE 4.50 3.49 4.60 13.46 SD 0.76 0.38 0.26 0.91 RSD 0.17 0.11 0.06 0.07 MEDIAN 4.48 3.49 4.64 13.37

CONCLUSION In approximately 10 seconds per sample, FTIR spectroscopy provides the ability to process up to 400 samples per hour. The achievement of the study is a reliable method for the quantification of fat, protein, lactose, and dry matter parameters in the routine milk analysis. This method, with appropriate and frequent calibration, offers highly specific, accurate, precise, and linear data across the analytical range. FTIR is a rapid and precise analytical technique, characterized by the reduced application of reagents and user-friendliness. The values obtained by this technique do not depend on the analyst, as with most methods; the technique also enables simultaneous analysis of multiple components in each sample. Finally, it should be noted that testing of samples by this method is very



174

cheap, despite the fact that the method proved to be exceptional in terms of precision, accuracy, and brevity. The study has shown that the FTIR method is indeed a reliable and acceptable as an instrumental technique for analyzing the given chemical parameters (fat, protein, lactose and total solids) in raw cow's milk.

REFERENCES 1. Andersen, S. K.; Hansen, P. W.; Andersen, H. V. Vibrational Spectroscopy in the

Analysis of Dairy Products and Wine, Handbook of Vibrational Spectroscopy; John Wiley & Sons Ltd: West Sussex, UK, 2002; pp 3672-3681.

2. Griffiths, P.R.; De Haseth, J.A.; Winefordner, J.D. Fourier Transform Infrared Spec-trometry, John Wiley InterScience (2nd Edition): New York, NY, 2007. pp 20-55.

3. Albanell, E.; Cáceres, P.; Caja, G.; Molina, E.; Gargouri, A. Determination of fat, protein and total solids in bovine milk by near-infrared spectroscopy. J. AOAC Int. 1999, 82, 753-758.

4. Nurrulhidayah, A.F.; Che Man, Y.B.; Rohman, A.; Amin, I.; Shuhaimi, M.; Khatib, A.: Authentication analysis of butter from beef fat using Fourier Transform Infrared (FTIR) spectroscopy coupled with chemometrics. Int. J. Res. 2013, 20, 1383-1388.

5. Jung, C.J. Insight into protein structure and protein-ligand recognition by Fourier transform infrared spectroscopy. J. Mol. Recognit. 2000, 13 (6), 325-351.

6. Etzion, Y.; Linker, R.; Cogan, U.; Shmulevich, I. Determination of Protein Concen-tration in Raw Milk by Mid-Infrared Fourier Transform Infrared/Attenuated Total Reflectance Spectroscopy. J. Dairy Sci. 2004, 87 (9), 2779-2788.

7. Solís-Oba, M.; Teniza-García, O.; Rojas-López, M.; Delgado-Macuil, R.; Díaz-Reyes, J.; Ruiz, R. Application of Infrared Spectroscopy to the Monitoring of Lactose and Protein from Whey After Ultra and Nano Filtration Process. J. Mex. Chem. Soc. 2011, 55 (3), 190-193.

8. Van de Voort, F.R. Fourier transform infrared spectroscopy applied to food analysis. Food Res. Int. 1992, 25 (5), 397-403.

9. Budinova, G.; Salva, J.; Volka, K. Application of Molecular Spectroscopy in the Mid-Infrared Region to the Determination of Glucose and Cholesterol in Whole Blood and in Blood Serum. Appl. Spectrosc. 1997, 51 (5) 631-635.

10. ISO 9622:2013/IDF 141. Milk and liquid milk products-Guidelines for the application of mid-infrared spectrometry, 2013.

11. ICAR guidelines. Section 12 - Guidelines for Milk Analysis, 2017. http://www.icar.org/Guidelines/12-Milk-Analysis.pdf (accessed 13 October 2017).

12. Rivera, C.; Rodríguez, R. Horwitz equation as quality benchmark in ISO/IEC 17025 testing laboratory. Private communication. 2014.

13. Magnusson, B.; Näykki, T.; Hovind, H.; Krysell, M. Handbook for Calculation of Measurement Uncertainty in Environmental Laboratories; Nordic Innovation: Oslo, Norway, 2012.

14. Little, T. A. (2016): Establishing Acceptance Criteria for Analytical Methods: http://thomasalittleconsulting.com/publications/articles/Acceptance_Criteria.pdf visited 30.06.2017.



175

ВАЛИДАЦИЈА И ПРИМЕНА FT-IR СПЕКТРОСКОПИЈЕ ЗА АНАЛИЗУ СИРОВОГ КРАВЉЕГ МЛЕКА

Денис С. Кучевић*, Драган М. Гламочић, Снежана Ј. Тривуновић, Игор М. Јајић,

Биљана Ч. Перишић, Саша З. Крстовић

Универзитет у Новом Саду, Пољопривредни факултет, Трг Доситеја Обрадовића 8, 21000 Нови Сад, Србија

Циљ овог истраживања је био да се испита да ли FT-IR спектроскопија пред-ставља тачну и валидну технику за процену хемијских параметара у сировом крав-љем млеку: млечне масти, протеина, лактозе и укупне суве материје. Процена је заснована на калибрацијским серијама које су настале на основу поређења рефе-рентног материјала са утврђеним вредностима. Поред тога, у обзир су узети и ре-зултати добијени у међулабораторијским испитивањима (тестирање оспособље-ности). Калибрацијски узорци су набављени од регионалних акредитованих рефе-рентних лабораторија. Параметри који су обухваћени валидацијом су: линеарност, тачност, поновљивост, репродуктивност и робусност. Удео линеарности је износио 0,95%. Биас израчунат за млечну маст је био -0,33, за протеине 0,31, за лактозу -0,25 и суву материју 0,06. За одређивање статистички значајних разлика између две серије мерења у циљу одређивања прецизности методе, примењен је F-тест. Добијени су следећи резултати: за млечну маст 1,469, за протеин 1,634, за лактозу 1,192 и за суву материју 0,528. Резултати добијени за мерење учинака преноси-вости су износили: за млечну маст 0,27%, за протеин 0,52%, за лактозу 0,47% и суву материју 0,47%. Узимајући у обзир наведене резултате валидације, може се закључити да је FT-IR спектроскопија поуздана инструментална техника за одре-ђивање млечне масти, протеина, лактозе и укупне суве материје у сировом крављем млеку. Кључне речи: FT-IR, спектроскопија, млеко, хемијски параметри, валидација

Received: 06 July 2017. Accepted: 14 October 2017.

APTEFF, 48, 1-323 (2017) UDC: 546.41`221.1+546.62`226]:547.458.88 https://doi.org/10.2298/APT1748177K BIBLID: 1450-7188 (2017) 48, 177-185


177

COMPARISON OF METHODS OF ZETA POTENTIAL AND RESIDUAL TURBIDITY OF PECTIN SOLUTIONS USING CALCIUM

SULPHATE/ALUMINIUM SULPHATE AS A PRECIPITANT

Tatjana A. Kuljanin*, Vladimir S. Filipović, Biljana Lj. Lončar, Milica R. Nićetin, Violeta М. Knežević


The affinity of calcium ion binding from CaO used in the most common process of purification of sugar beet juice is relatively low. Therefore, large amounts of this com-pound are required. This paper presents the theoretical basis of a novel sugar beet juice purification method based on the application of the binary system CaSO4/Al2(SO4) . In order to monitor the process of coagulation and precipitation of pectin in the presence of CaSO4/Al2(SO4)3, two methods were compared: measurement of the zeta potential and of residual solution turbidity. The zeta potential of pectin solution was determined by electrophoretic method, while the residual turbidity was determined by spectrophotometry. Two model solutions of pectin (0.1 % w/w) were investigated. Studies were performed with 10 different concentrations of the binary solution CaSO4/Al2(SO4)3 (50 - 500 g dm-3). The amount of the precipitant CaSO4/Al2(SO4)3 (1:1 w/w) needed to achieve the minimum solution turbidity and charge neutralization of pectin particles (zero zeta potential) were measured and compared. Colloidal destabilization occurred before a complete neutrali-zation of the surface charge of pectin particles (zeta potential ~ 0 mV). Optimal quan-tities (490 - 705 mg g-1 pectin) of the applied binary mixture, were obtained using both methods. This is much lower than the amount of CaO that is commonly used in the con-ventional process of sugar beet juice purification (about 9 g· g-1 pectin). The use of these precipitants could be important from both economic and environmental point of view. KEY WORDS: calcium sulphate/aluminium sulphate, zeta potential, residual turbidity,

pectin

INTRODUCTION

Non-sucrose matter in the sugar beet juice, mainly pectins, vary with regard to acidi-ty, molecular weight, and charge density. The negative charge usually occurs due to the dissociation of carboxylic functional groups on the surface of the pectin and protein (1-3). Separation and precipitation of these compounds in the sugar beet juice are done mostly by limestone. In the classical procedure, in addition to CaO, CO2 is introduced (first

* Corresponding author: Tatjana Kuljanin, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




178

saturation - transforming the excess Ca(OH)2 into CaCO3). This provides aggregation of negatively charged macromolecular compounds with positively charged crystals CaCO3 and coagulation colloids (4). However, the quantities of used limestone, in this stage of purification of sugar beet juice are very high, and range from 1 % to 3 % w/w, calculated on the beet. In order to reduce this amount, many compounds with divalent and trivalent cations that chemically cause discharge of undesirable macromolecules of sugar beet juice have been studied (5-9). Of the total of non-sucrose matter contained in the sugar beet juice, approximately 60% are pectins. The binding affinity of divalent cations with pectins of plant origin follows the selectivity order: Cu2+ ~ Pb2+ >> Zn2+ > Cd2+ ~ Ni2+ > Ca2+ > Mg2+ (10, 11). Cu ions show the best binding effect (mechanism of surface com-plexation), however, they can have an undesirable effect in food processing (5). On the other hand, Ca ions with weaker binding effect, particularly in the form of CaSO4, in hydrophilic pectin macromolecules cause significant decrease of hydration. This is, in addition to the charge neutralization, a precondition for more rapid coagulation and precipitation (7, 8, 10, 12). Also, there are many studies of the application of the hydrolyzing salt Al2(SO4)3 for treatment of fresh or waste water. In this case, cationic species predominate. The order of trivalent Al3+ ion bonding strength with humic materials in water, compared with divalent metal ions is as follows: Al3+ > Cu2+ > Pb2+ > Cd2+ > Ca2+ (13, 14). In accordance with this, it is clear that the application of Al3+ ions will be probably be more efficient in pectin separation compared to the use of a compound with Ca2+ ions (15). Also, priority is given to the use of Al salts from economical and ecological point of view (14, 16). In the previous work (12, Figure 1) the influence of a binary system CaSO4/Al2(SO4)3 on the efficiency of sugar beet pectin separation by the method of measuring of zeta potential was studied. In this work, the efficiency of pectin separation was monitored by two methods: measurement of the zeta potential of pectin solution and of the residual turbidity of the solution. The zeta potential measurement is the most common technique. It is based on the fact that each particle in a colloid solution is surrounded with an electric double layer. A most important characteristic of the double layer is the plane of shear. The shear plane represents the interface between the portion of the double layer that remains connected to the particle and the portion which remains stationary as the particle moved. The potential on this surface is easily measurable and is called zeta potential. The addition of a polyvalent cation leads to a decrease in the negative charge on the particle surface and decreases the absolute value of zeta potential. By keeping zeta potential close to zero, the colloidal particles such as pectin will discharge, and the conditions for effective coagu-lation and sedimentation will be achieved (17). However, the monitoring of the zeta po-tential alone will not provide all the information required to operate the coagulation pro-cess. For instance, the removal of negligibly charged components, such as low molecular weight hydrophilic organic material which is less amenable to removal by conventional coagulation methods is unlikely to be optimized by zeta potential monitoring (1, 18). Furthermore, although zeta potential can be used to optimize the treatment processes, it cannot provide a definition of the optimum in terms of the achievable residuals post-treatment, such as turbidity. Residual solution turbidity is dependent on non-charged



179

components and the operational parameters which are likely to influence the collision efficiency (1, 19). At a minimum solution turbidity, the most favourable conditions for coagulation and sedimentation process should occur (13, 20). The objective of this study was to compare the methods of measuring the zeta po-tential and minimum solution turbidity in the pectin separation from the sugar beet juice. The proposed precipitant was a mixture of CaSO4 and Al2(SO4)3 (1:1 w/w).

EXPERIMENTAL

Materials and methods Two pectin preparations were extracted from sugar beet cossettes obtained in the industrial processing of sugar beet (Žabalj Sugar Mill, Žabalj, Serbia). The hydrated salts CaSO4 x 2H2O and Al2(SO4)3 x 18H2O (Zorka Pharma, Šabac, Serbia), in the form of the (1 : 1; w/w) mixture were used for preparation of the studied solutions with de-ionized water. The pH of the solutions was regulated at 7 (pH METER Iskra MA5740, Kranj, Slovenia) before each experiment. The pectin preparation was isolated by extraction in acidic condition by standard laboratory procedure AOAC (21). The mass ratio of sugar beet cossettes to solvent was 1:10. The extraction was conducted in an extractor which was made in the laboratory of the Faculty of Technology, Novi Sad, Serbia. It is a glass vessel, volume of 2 dm3 with the built-in of a continuous mixer (TALBOYS INSTRUMENT CORP., model 102; Troem-ner Precision Weights and Laboratory Equipment, USA) and a water bath (INKO, Zag-reb, Croatia) for heating the solution. The extraction was performed in conditions, for preparation P1: pH = 1, t = 85 °C, τ = 2.5 h, for preparation P2: pH = 3.5, t = 85 °C, τ = 2.5 h. Detailed description of the extraction of pectin preparations was presented else-where (22). General extraction procedure in acidic and basic conditions has been schema-tically displayed in a previous work (8), Figure 1. The pectin preparations, after precipi-tation and cleaning, were dried in a vacuum dryer for 12 hours at 70 °C. The procedure was repeated several times and basic parameters of pectin preparation were determined according to standard methods of AOAC (21). The degree of esterification of the pectin preparation was calculated using the equivalents of free (X) and the esterified carboxyl groups (Y) (5-8), presented in Table 1. Table 2 presents the data related to the salt so-lutions used as the precipitants:

Table 1. Basic parameters of the pectin preparations

Pectin solution P1 P2

Solid content, SC (g/100g) 81.55 80.25 Equivalent of free COOH groups, X · 105 11.75 21.83 Equivalent of esterified COOH groups, Y · 105 25.74 18.55 Content of galacturonic acid, Gal.A (%) 63.45 71.31 Degree of esterification, DE (%) 68.66 45.94



180

Table 2. Data related to the pectin solutions and salts used as the precipitants

Pectin solutions P1 and P2

Solution of salts: 20 flasks (pH 7)

Volume: 50 cm3

Concentration: 1 g·dm-3 Proportion and type of salt

(% w/w) No. of

measurements Conc. of salt

(g·dm-3) 50% CaSO4 + 50% Al2(SO4)3 20 50 - 500

As shown in Table 2, twenty flasks were filled with 50 cm3 of pectin solutions: ten flasks for each preparation (P1 and P2). The pectin concentration was 1 g·dm-3. Binary solutions were prepared by dissolving 0.5 g of pure salt CaSO4 and 0.5 g of pure salt Al2(SO4)3 in 200 cm3 distilled water. Namely, after the adjustment of the pH value, an appropriate volume of the binary solution was pipetted to 50 cm3 of the pectin prepa-ration solutions to obtain the corresponding concentration of CaSO4/Al2(SO4)3, ranging from 50 to 500 mg/dm3. A more detailed plan of the experiment was given in the work (12). After adding the precipitants to the tested preparations, the solution was stirred for 30 min on a high-speed magnetic stirrer (TEHNICA, Železniki, MM-520, Slovenia) (500 rpm). Before measuring zeta potential, the pectin solutions containing coagulants were stirred manually for 5 min and left to rest for another 5 min. The zeta potential was evaluated by means of electrophoresis using a Zeta Meter ZM-77 (Zeta Meter Inc., New York, USA). The pectin solution with precipitants was poured into the electrophoretic cell with electrodes. Measurements were carried out through the stereoscopic microscope at a 6-fold magnitude. After adjusting the voltage at 150 V, an electric recording device was used to measure the time needed for a pectin particle to pass a distance of a standard micrometer division. An average value of measured time of 20 particles was used to derive the zeta potential of the tested solutions using a diagram based on the Helmoltz-Smoluchowski equation for the electrophoretic mobility of colloidal particles (23). These measurements were repeated three times. The residual turbidity measurements were performed two times: immediately after the zeta potential measurement and after the induced precipitation of pectin solution. The process of precipitation lasted 10 h, to allow the precipitants to react with pectin. For checking the beginning and the completion of sedimentation process, the change of re-sidual solution turbidity was measured. The measurements were performed by spectro-photometry using a SPEKOL 202 spectrophotometer (Carl Zeiss Jena, Jena, Germany) at the wavelength λ = 560 nm. After the first measurement of the absorbance of the soluti-on, the samples were left for 10 h at room temperature to complete the started agglome-ration of particles. After the separation of the sediment by filtration using filter paper, the absorbance of the clear part of the solution was measured again. The residual turbidity of the pectin solution with the coagulants (mixtures of CaSO4 and Al2(SO4)3; 1 : 1 w/w) was expressed as the difference between the absorbance of the solution immediately after the measurement of the zeta potential and after the aging solution in a clear part of the sample (Δ, m-1). The measurement of the residual turbidity was performed under the same conditions as the measurement of zeta potential (20). Each measurement was repea-ted more than three times only when the difference between two results was above 5%.



181

RESULTS AND DISCUSSION Figures 1 and 2 show the influence of the concentration of the binary solution CaSO4/Al2 (SO4)3 (1: 1; w / w) on the change in the zeta potential and residual turbidity of the pectin solutions P1 and P2.

Figure 1. Influence of the concentration of the CaSO4 /Al2(SO4)3 (1: 1; w /w) solution on the change in the zeta potential of pectin solutions P1 and P2 (12)

Figure 2. Influence of the concentration of the binary CaSO4/Al2(SO4)3 (1 : 1; w/w) solution on the change in the residual turbidity of pectin solutions P1 and P2



182

As can be seen from Figure 1, the zeta potential increased with the increase in the concentration of the binary solution CaSO4/Al2(SO4)3 for both pectin preparations (12). The amount of the CaSO4/Al2(SO4)3 mixture needed to bring the zeta potential to zero for pectin P1 and P2 was 465 mg∙dm-3 (705 mg∙gp

-1) and 420 mg∙dm-3 (585 mg∙gp-1), respec-

tively. These quantities are slightly larger than the required amount of pure compound CaSO4 examined in the previous work under the same conditions (7). However, with re-gards to the magnitude of the concentration of the binary solution CaSO4/Al2(SO4)3, the residual turbidity of the solutions reached a minimum value at the concentration of 400 mg∙dm-3 (550 mg∙gp

-1), for pectin P1 and 350 mg∙dm-3 (490 mg∙gp-1), for pectin P2. After

reaching these values, the residual turbidity increased (Figure 2). By comparing the re-sults shown in Figures 1 and 2, it can be concluded that the optimal amount of CaSO4/Al2(SO4)3 for achieving a minimum residual turbidity (τmin) is smaller than the optimum amount of CaSO4/Al2(SO4)3 for achieving the zero value of the zeta potential. That means, in general, that colloidal destabilization occurred before complete neutraliza-tion of the surface charge of pectin particles (zeta potential ~ 0 mV). Ratnaweera, Hiller and Bunse reported similar findings (24). According to this, it can be concluded that it is not necessary to bring the zeta potential to zero, which requires a higher coagulant dose. The use of higher coagulant dose is not a practical solution in terms of chemical feed costs and handling. The results of this investigation indicate the existence of an opera-tional interval of zeta potential (-5 to 0 mV) within successful coagulation and preci-pitation of pectin that takes place using the binary solution CaSO4/Al2(SO4)3. An increase in the number of binding sites and coagulant dose is likely to increase both charge neu-tralization and precipitation. The P2 solution showed a greater coagulation effect, and therefore a better sedimentation of pectin since it has a higher number of binding sites (smaller DE means a higher number of free COO- groups on the surface of pectin macro-molecules, Table 1). The minimum amounts of CaSO4/Al2(SO4)3 obtained using both methods (490-705 mg/g pectin) were significantly smaller than the amounts of the coa-gulant CaO, traditionally used in sugar beet juice processing (about 9 g/g pectin).

CONCLUSION The analysis of the performance of coagulation and sedimentation processes in pectin solutions by adding the binary solution CaSO4/Al2(SO4)3 revealed a clear relationship between the zeta potential and residual turbidity. The results demonstrated the existence of operational intervals with respect to the zeta potential and residual turbidity solution in which pectin removal efficiency is high. This reduces the consumption of the proposed precipitant CaSO4/Al2(SO4)3 in sugar beet juice purification, which is significant from an economic and environmental point of view. The binary precipitant CaSO4/Al2(SO4)3 is recommended as a good substitute for the traditional coagulant CaO.

Acknowledgement

The authors gratefully acknowledge the financial support from the Ministry of Education, Science and Technological Development of the Republic of Serbia, Project TR 31055.



183

REFERENCES

1. Sharp, E.L. Natural Organic Matter Coagulation. Ph.D. Thesis, Cranfield University, 2005.

2. Poel, P.W., Schiweck, H., Schwartz, T. Sugar Technology, Beet and Cane Sugar Manufacture, Berlin, 1998; pp. 187-189.

3. Hunter, R.J. Introduction to modern colloid science. Oxford University Press: New York, 2001.

4. Šušić, S., Raca, K., Radošević, N. Priručnik za industriju šećera 2. Beograd, 1980; pp 223-229.

5. Kuljanin, T., Lević, Lj., Mišljenović, N., Koprivica, G. Electric Double Layer and Electrokinetic Potential of Pectic Macromolecules in Sugar Beet. APTEFF. 2008, 39, 21-28.

6. Kuljanin, T., Mišljenović, N., Koprivica, G., Jevrić, L., Grbić, J. The effect of copper ions, aluminium ions and their mixtures on separation of pectin from the sugar beet juice. Hem. Ind. 2013, 67(1), 69-76.

7. Kuljanin, T., Jevrić L., Ćurčić, B., Nićetin, M., Filipović, V., Grbić, J. Alumi-nium and calcium ions binding to pectin in sugar beet juice: Model of electrical double layer. Hem. Ind. 2014, 68 (1) 89-97.

8. Kuljanin, T., Lončar, B., Pezo, L., Nićetin, M., Knežević, V., Jevtić-Mučibabić, R. CaSO4 and cationic polyelectrolyte as possible pectin precipitants in sugar beet juice clarification. Hem. Ind. 2015, 69(6), 617-625.

9. Kuljanin, T., Lončar, B., Nićetin, M., Knežević, V., Filipović V. Sugar beet juice clarification using calcium sulfate, copper sulfate and aluminium sulfate. II International Congress ,,Food Technology, Quality and Safety’’, 28-30.10.2014, Novi Sad, Serbia, Proceedings, 66-70.

10. Garnier, C., Axelos, M.A.V., Thibault J.F. Selectivity and cooperativity in the binding of calcium ions by pectins. Carbohyd. Res. 1994, 256, pp. 71.

11. Axelos, M.A.V., Garnier, C., Renard, C.M.G., Thibault, J.F. Pectins and Pecti-nase: interaction of pectins with multivalent cations: phase diagrams and struc-tural aspects. Ed by Visser, J. and Voragen, A. Amsterdam, 1996.

12. Kuljanin, T., Filipović, V., Nićetin, M., Lončar, B., Muzalevski, A., Jevtić-Mu-čibabić R. Separation of pectin from sugar beet juice by binary system calcium sulphate/aluminium sulphate. III International Congress “Food Technology, Quality and Safety”, 25-27.10.2016, Novi Sad, Serbia, Proceedings, 565-568.

13. Kinniburgh, D. G., Riemsdijk, W. H., Koopal, L. K., Borkovec, M., Benedetti, M. F., Avena, M. J. Ion binding to natural organic matter: competition, hetero-geneity, stoichiometry and thermodynamic consistency, Colloids and Surfaces, A: Physicochem. Eng. Aspects, 1999, P 151, 147-166.

14. Kuljanin, T., Lević, Lj., Jevrić, L., Ćurčić, B., Grbić, J., Jevtić-Mučibabić, R. Aluminium sulfate as ecological coagulant in phase of sugar beet juice clarifi-cation, XVI International Eco-Conference, Safe Food, 26-29.09.2012, Novi Sad, Serbia, Ecological Movement of Novi Sad, Proceedings, 427-435.



184

15. Lević, Lj., Gyura, J. Influence of aluminium sulphate concentration on the change of electrokinetic potential of macromolecular compounds in molasses, Nahrung, 1999, 43 (4), 288-289.

16. Sagripanti, B., Fontana, P., L’ Industria saccarif. Ital. 1992, 1, 7-11. 17. Koper, G. J. M. An Introduction to Interfacial Engineering, VSSD, Delft, 2007. 18. Owen, D.M., Amy, G.L., Chowdhury, Z.K. Characterization of natural organic

matter and its relationship to treatability. American Water Works Association Research Foundation report, Denver CO, 1993.

19. Vuorio, E., Vahala, R., Rintala, J., Laukkanen, R. The evaluation of drinking water treatment performed with HPSEC. Environ. Int. 1998, 24(5-6), 617-623.

20. Kuljanin, T., Lončar, B., Nićetin, M., Grbić, J., Jevtić-Mučibabić, R., Šobot, K. Calcium sulphate as coagulant in phase of sugar beet juice clarification - method of measurement residual solution turbidity and zeta potential. JHED, 2015, 10, 49-53.

21. AOAC - Methods of Analysis of Official Analytical Chemists, Washington, 2000.

22. Kuljanin, T., Mišljenović, N., Koprivica, G., Lević, Lj., Filipčev B. Influence of Cu2+ and Al3+ ions on zeta potential change of pectin and protein preparations extracted from sugar beet. PTEP. 2010, 14 (3), 141-144.

23. Lyklema, J. Elektrokinetics after Smoluchowski. Colloids and Surfaces, A: Physicochem. Eng. Aspects, 2003, 222, 5-14.

24. Ratnaweera, H., Hiller, N., Bunse, U. Comparison of the coagulation behavior of different Norwegian aquatic NOM sources. Environ. Int. 1999, 25(2-3), 347-355.

ПОРЕЂЕЊЕ МЕТОДА МЕРЕЊА ЗЕТА ПОТЕНЦИЈАЛА И РЕЗИДУАЛНЕ МУТНОЋЕ ПЕКТИНСКИХ РАСТВОРА УПОТРЕБОМ КАЛЦИЈУМ

СУЛФАТА/АЛУМИНИЈУМ СУЛФАТА КАО ПРЕЦИПИТАНТА

Татјана A. Куљанин, Владимир С. Филиповић, Биљана Љ. Лончар, Милица Р. Нићетин, Виолета М. Кнежевић


Булевар цара Лазара 1, 21000 Нови Сад, Србија

Калцијумови јони (пореклом из CаО) који се користе за укљањање пектина у фази чишћења сока шећерне репе имају релативно мали афинитет везивања са пектинима и осталим несахарозним материјама. Због тога су потребне велике количине креча што може изазвати негативне последице у непосредној околини фабрике шећера (алкализација земљишта). Разелектрисањем пектинских честица, тј. додавањем катјона веће валенце, створили би се услови за њихову успешну коа-гулацију и таложење. Аl2(SО4)3 у форми хидролизоване соли се често користи у третману пијаће и отпадне воде. Са друге стране, Cа2+ јони, иако имају мањи ефе-кат разелектрисања, поседују велику дехидратациону моћ што је важан предуслов



185

за успешну коагулацију и таложење пектина, протеина и осталих несахарозних је-дињења. У овом раду предложена је нова метода чишћења сока шећерне репе која се базира на употреби бинарног раствора CаСО4/Аl2(SО4)3. У циљу праћења про-цеса коагулације и таложења пектина у присуству раствора CаСО4/Аl2(SО4)3 ко-ришћене су две методе: метода мерења зета потенцијала и метода мерења рези-дуалне мутноће пектинских раствора. Електрокинетички или зета потенцијал је одређиван електрофорезом док је резидуална мутноћа раствора мерена спектро-фотометријски. Испитивана су два модел-раствора пектинских препарата (0,1% мас). Мерења су обављена са 10 различитих концентрација смеше коагуланата CаСО4/Аl2(SО4)3 у масеном односу 1 : 1, у интервалу 50-500 g · dm-3. Измерене су оптималне количине раствора CаСО4/Аl2(SО4)3 које би изазвале разелектрисање пектинских честица (нулти зета потенцијал) односно минималну мутноћу раствора. Поређењем ове две методе, уочено је да се колоидна дестабилизација (почетак коа-гулације) одвија пре комплетне неутрализације површине пектинских честица када је зета потенцијал ~ 0 мV. Оптималне количине примењеног бинарног раствора, из-мерене помоћу ове две методе, биле су у интервалу 490-705 mg по g пектина. Ове количине су вишеструко мање од количине CaО који се користи у класичном пос-тупку чишћења сока шећерне репе (око 9 g по g пектина). Одговарајућим дозира-њем бинарног раствора CаСО4/Аl2(SО4)3 уз контролу зета потенцијала и резидуалне мутноће раствора, остварио би се значајни економски ефекат. Примена ових преци-питаната била би значајна и са еколошког становишта, јер би се загађење земљиш-та свело на минимум. Кључне речи: калцијум сулфат/алуминијум сулфат, зета потенцијал, резидуална

мутноћа, пектини


APTEFF, 48, 1-323 (2017) UDC: 637.1+636.18]:612.664 https://doi.org/10.2298/APT1748187L BIBLID: 1450-7188 (2017) 48, 187-195


187

CHANGES IN MILK COMPOSITION OF DOMESTIC BALKAN DONKEYS’ BREED DURING LACTATION PERIODS

Jasmina M. Lazarević*, Tatjana A. Tasić, Sanja J. Popović, Vojislav V. Banjac,

Olivera M. Đuragić, Bojana M. Kokić, Ivana S. Čabarkapa

Univeristy of Novi Sad, Institute of Food Technology, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia

Changes in the milk composition during early, middle and late stage of lactation period were monitored in 10 individual autochthonous donkeys of the Domestic Balkan breed. Animals were grazing ad libitum during early and middle lactation, while supple-ments (meadow hay, corn and corn stalks) were added during the late lactation period. The aim of this study was to evaluate the main compounds of donkeys’ milk during different periods of lactation. Based on the obtained results it was noticed that nutritional effects on milk compo-sition were the greatest in the early and middle lactation. Chemical analysis indicated that donkeys’ milk was quite poor in dry matter (8.82-9.68%), fat (0.54-0.67%) and protein (1.44-1.79%), and conversely was high in lactose content (6.06-7.12%). Concen-tration of vitamin C varied significantly (P<0.05) during the lactation period and was very high (12.84-26.89 μg/mL) in comparison to milk from other animal species. Like-wise, the content of lysozyme (1.95-3.29 g/L) and lactoferrin (1.56-3.14 g/L) during the early and middle lactation period showed an increase. Donkeys' milk did not show any significant differences in MUFA, n6-PUFA and UFA contents, while n3-PUFA, n6/n3 and UFA/SFA ratio showed a significant variability (P<0.05) during the lactation period. KEY WORDS: Balkan donkeys’ milk, casein fractions, whey protein fractions, fatty acids

INTRODUCTION The research interest in donkeys’ milk has increased in Europe in recent years, since its composition is similar to human milk (1, 2), and that has consequently increased the price of donkeys’ milk. In the available literature, there is a lack of data regarding accurate characterization and changes in chemical composition of milk during lactation for the Balkan donkeys’ breed. In the Serbian farm "Zasavica breeding of Domestic Bal-kan donkeys" they started a project of the revitalization and preservation of this low reproductive breed, based on different diet systems consisted of feeding on pasture. Pas-ture is the major dietary source for donkeys’, but there is little information regarding its * Corresponding author: Jasmina M. Lazarević, University of Novi Sad, Institute of Food Technology, Bulevar




188

nutritional composition. It is widely recognized that pasture, during efficient grazing, is the cheapest feedstuff (3), with highly variable chemical composition. The sources of variation in pasture composition are plant species, growth stage, temperature, and light intensity. Fresh grass contains a high proportion of total fatty acids (50-75%) in the form of n3 linolenic acid (2, 4). In donkeys’ milk, total protein content is very low, in particular it shows a low level of casein (CN), and on the other hand, high level of lactose, linoleic acid (n6-PUFA), linolenic acid (n3-PUFA) and lysozyme (5, 6). The content of CN ranges from 37.56 to 39.98% of total protein content (1). Guo et al. (6) reported that the content of whey pro-teins (WP) in donkeys’ milk is in the range of 53.03 to 57.06% of total protein content. The main components of WP are -lactalbumin (-LA) and -lactoglobulin (-LG), which account from 35% to 50% of the total proteins, compared to 20% present in cows’ milk (1). Also, donkeys’ milk is a very rich source of lysozyme (LYS) (1.0 g/L) (4, 5, 7), and has significantly higher content than bovine milk (0.13 mg/L), while the level of LYS is quite similar to that in mare milk (0.4-1 g/L) (8, 9). The lipid content of donkeys’ milk is quite poor, according to Guo et al. (6), who showed significant variability (0.20-1.82%), indicating that the lipid content could be af-fected by breed, breeding area and forage, milking technique, interval between milking, and mainly by the stage and year of lactation. Similar results were reported by Fox (10). Saturated fatty acids (SFA) are the most representative fatty acids in donkeys’ milk, which was indicated by different studies (10, 20). However, it has been noticed that SFA content is lower in donkeys’ milk than in ruminant milk (11, 12). During the lactation phase, content of SFA decreases while the content of unsaturated fatty acids (UFA) shows an increase (2). In view of the above, the aim of the present study was to determine the influence of lactation period on the chemical composition of the milk from Balkan donkeys’ when diet of animals was based on pasture feeding.

EXPERIMENTAL

Animals feeding During the early (40-100 days) and middle (130-190 days) (from spring to autumn) periods of lactation. the donkeys’ were grazing ad libitum. During the winter feeding (late period of lactation; 220-280 days) meadow hay, corn and corn stalks were used as supplements (two times a day). Some of the species found in pasture vegetation included Poaceae, Fabaceae, Asteraceae and Plantaginaceae families. Water was available ad libitum.

Milk sample collection Individual raw milk samples were collected during the lactation period (between 40 and 280 day) from ten healthy female donkeys’. Sample collection was performed four



189

times per each lactation period. Milking was performed two times a day, in the morning (08:00 am) and afternoon (14:00 am), after which samples were merged for every indivi-dual animal. The samples were stored in sterile plastic bottles at 4 °C until further ana-lysis. The samples that were used for the ascorbic acid determination were stored in plastic bottles wrapped in aluminum foil. The experiment was repeated during two years.

Chemical analysis Basic chemical composition of the milk, i.e. dry matter, ash, protein, and fat were analyzed by methods of the IDF (13). Lactose content was determined by GEA Niro ana-lytical method A18b (14). The vitamin C (L-ascorbic acid) content was determined by enzymatic method (Megazyme International Ireland 2011), and the absorbance was mo-nitored using GBC CINTRA 303UV/VIS spectrophotometer.

Determination of protein profile Protein profiles (CN and whey fractions) of milk were analyzed using two-dimen-sional difference gel electrophoresis. Sample preparation was carried out according to Tidona et al. (15) with some modifications. A 1:1.5 (v/v) dilution of each milk sample was made with buffer (0.125m Tris-HCL, 4% SDS, 2% glycerol, 2% -mercaptoethanol, pH 6.8) and heated at 100 °C for 5min. The chip-based separation was performed on an Agilent 2100 bioanalyzer (Agilent Technologies, Santa Clara, CA, USA) in combination with Protein 80 Plus Lab Chip kit.

Fatty acid analysis Fat was extracted from the milk samples using Folch method (16). The preparation of fatty acid methyl esters was done by transmethylation with 14% wt. boron trifluoride/ methanol solution, which is recommended for this type of substrates. Fatty acid analysis was done on an Agilent 7890A gas chromatographer (Agilent Technologies, CA, USA) with flame ionization detector, and fused silicia capillary column (DB-WAX 30 m, 0.25 mm, 0.50 μm). The carrier gas was helium (purity 99.9997 vol%, flow rate = 1.5 ml/min).

Statistical analysis To analyze variations of the results, one-way ANOVA and Post-hoc Duncan test were used. Differences between the means with probability P<0.05 were accepted as statisti-cally significant. The analysis was performed using the software package Statistics 12.0 for Windows (Stat Soft, Tulsa, Oklahoma, USA). All measurements were made in tripli-cates, and the results were expressed as mean value ± standard deviation.



190

RESULTS AND DISCUSSION The chemical composition of Balkan donkeys’ milk samples are summarized in Table 1. Depending on the lactation period, dry matter, fat and lactose showed higher content in the early and late lactation in comparison to the middle lactation. The milk fat level did not vary significantly (P>0.05) within the period of lactation. The fat content was slightly higher in the early lactation due to the quality and availability of fresh grass in the pasture, and in the late lactation due to addition of supplements to the diet, which is consistent with the results of Ivanković et al. (17). A similar trend in fat content decrease in donkeys’ milk after 100 days of lactation was reported by Salimei et al. (1) and Guo et al. (7).

Table 1. Chemical composition of Domestic Balkan donkeys’ milk during various lactation periods

Composition Lactation period

Early (40-100 days)

Middle (130-190 days)

Late (220-280 days)

Dry matter, % 9.27ab ± 0.36 8.82a ± 0.23 9.68b ± 0.47

Ash, % 0.43a ± 0.04 0.64b ± 0.10 0.47a ± 0.02

Protein,% 1.79b ± 0.10 1.57a ± 0.08 1.44a ± 0.06

Fat,% 0.66a ± 0.27 0.54a ± 0.10 0.67a ± 0.15

Lactose,% 6.38a ± 1.88 6.06b ± 0.83 7.12c ± 0.86

Vitamin C, μg/ml 26.89a ± 0.19 24.15a ± 0.18 12.84b ± 0.20 a−b Means with different letter indexes in the same row are statistically significantly different (P<0.05).

The highest concentrations of milk protein and vitamin C were determined during the early lactation period, and gradually decreased from the beginning to the end of the lactation period. Similar trends were observed in mares’ (18), cows’ (11) and donkeys’ milk (19). The protein content did not vary significantly (P>0.05) during the middle and late lactation. In the middle period of lactation, statistically higher (P<0.05) ash content was observed compared to other stages of lactation. The obtained results of donkeys’ milk chemical composition do not appear to differ significantly from those in the litera-ture based on the other donkeys breeds, with the exception of the average content of ash, which was slightly higher than the reported values (1, 7, 19). In the present study, che-mical composition of milk was improved when donkeys’ gained access to pasture in the early and middle lactation period. The changes observed in CN and its fractions in the various lactation periods are shown in Table 2. The average content of CN showed a significant decrease after the middle period of lactation. The obtained results for CN content are similar to those reported for Amiata donkeys' breed (20, 21), but lower than for ruminant's milk (22). The CN fractions present in the analyzed samples were -CN, S1-CN and S2-CN. The chan-



191

ges in the concentration of CN fractions during the early and middle period were not statistically significant (P>0.05), while the observed decrease of all CN fractions was significant in the late lactation. The highest concentration of CN, S1-CN and -CN was observed during the middle lactation (4.82%, 1.72%, 2.88%, respectively), and for S2-CN (0.24%) during the early lactation.�

Table 2. Mean concentration of CN and casein fractions in

Domestic Balkan donkeys’ milk

Lactation period

Casein (CN) (g/L)

CN fractions S1-CN (g/L)

S2-CN (g/L)

-CN (g/L)

Early 3.84ab ± 0.58 1.37a ± 0.36 0.24a ± 0.07 2.23a ± 0.37 Middle 4.82b ± 1.19 1.72a ± 0.71 0.22a ± 0.03 2.88a ± 0.62 Late 2.63a ± 0.50 1.10a ± 0.10 0.10b ± 0.03 1.45b ± 0.20

a−b Means with different letter indexes in the same column are statistically significantly different (P<0.05).

Table 3. Mean concentration of WP and WP fractions in Domestic Balkan donkeys’ milk

Lactation period

Whey protein (WP) (g/L)

WP fractions -LA (g/L)

-LG (g/L)

LYZ (g/L)

LF (g/L)

Early 11.05a ± 0.68 2.66ab ± 0.10 2.10a ± 0.20 3.29a ± 0.28 2.06a ± 0.62 Middle 11.60a ±1.62 3.03b ± 0.65 1.97a ± 0.24 2.71a ± 0.49 3.14b ± 0.44 Late 7.49b ± 1.71 2.11a ± 0.30 1.50b ± 0.34 1.95b ± 0.39 1.56a ± 0.74

a−b Means with different letter indexes in the same column are statistically significantly different (P<0.05).

Table 3 summarizes the data of total WP and its fractions determined in donkeys’ milk during various lactation periods. The WP content was the highest during the early and middle period (11g/L), and then rapidly decreased during the late period of lactation, which was not in accordance with the study of Malacarne et al. (23), where the content of WP remained constant throughout the lactation period. The content of WP fractions -LG and LYZ slightly decreased in the middle in comparison to early lactation period, but the decrease was not statistically significant (P>0.05). In the middle of the lactation period, the content of -LA and LF was significantly higher (P<0.05) compared to the early and late periods. All WP fractions significantly decreased (P<0.05) during the late period of lactation. Table 4 shows the level of fatty acids. The fatty acids were affected by the lactation period and showed a statistically significant decrease (P < 0.05) in the SFA content, and a significant increase (P<0.05) in the PUFA content. Overall, the average SFA content was higher than the MUFA and PUFA contents. The most of UFA present in milk, whose content increased from the beginning to the end of the lactation period, were MUFA. The period of lactation did not affect (P>0.05) the MUFA content that averaged around 32%.



192

The UFA/SFA ratio increased from the early to late lactation and averaged around 1.0, which was similar to that in mare’s milk (22), and higher than the ratio found in the Ragusana donkeys’ milk (UFA/SFA ratio = 0.48) (11).

Table 4. Fatty acids composition of Balkan donkeys’ milk (% of total fatty acid)

a−b Means with different letter indexes in the same row are statistically significantly different (P<0.05).

SFA - saturated fatty acids MUFA - monounsaturated fatty acids PUFA - polyunsaturated fatty acids UFA - unsaturated fatty acids

The changes in the PUFA and n6-PUFA contents in the analyzed samples showed a tendency to increase until the end of the lactation. The maximum content of n3-PUFA was in the early lactation period. The lowest levels of n6-PUFA and n-6/n-3 ratio were observed in the early lactation period, and did not change during the middle and late lactation. The higher n6/n3 ratio may be explained by the increase in pasture feeding of donkeys on richer grass during the middle and late lactation period (2, 11).

CONCLUSION

On the basis of these results it could be concluded that Balkan donkeys’ milk represents a rich source of WP, LF and LYS, as well as essential fatty acids. During feeding of donkeys’, the fresh forage in the pasture and the addition of meadow hay, as well as of corn during the winter, increased dry matter, fat and lactose content in the milk in the early and late period of lactation. The higher n3-PUFA content and lower n6/n3 ratio is of great interest in adopting donkeys’ milk.

Fatty acid composition

Lactation period

Early (40-100 days)

Middle (130-190 days)

Late (220-280 days)

SFA 59.75a ± 4.88 57.61a ± 6.41 42.47b ± 7.49

MUFA 31.74a ± 6.87 33.12a ± 3.56 30.71a ± 3.56

PUFA 15.91a ± 3.30 21.53ab ± 6.53 25.80b ± 4.36

n3-PUFA 8.86b ± 0.53 6.80a ± 0.81 7.21a ± 1.20

n6-PUFA 9.16a ± 1.88 11.29a ± 0.83 11.18a ± 0.86

n6/n3 ratio 1.03a ± 0.17 1.67b ± 0.03 1.61b ± 0.14

UFA 47.66a ± 7.20 54.64a ± 9.71 56.51a ± 6.51

UFA/SFA ratio 0.81a ± 0.19 0.97a ± 0.25 1.37b ± 0.37



193

Acknowledgement This study was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia, Project No. III46012. Authors would like to thank Messrs. Slobodan Simić and Nikola Nilić (Special Nature Reserve “Zasavica”, Serbia) for their great cooperation in this research.

REFERENCES

1. Salimei, E.; Fantuz, F.; Coppola, R.; Chiofalo, B.; Polidori, P.; Varisco, G.: Composition and characteristics of ass’s milk. Anim. Res. 2004, 53 (1), 67-78.

2. Martemucci, G.; D'Alessandro, A. G.: Fat content, energy value and fatty acid profile of donkey milk during lactation and implications for human nutrition. Lipid Health Dis. 2012, 113 (11), 1-14.

3. Dillon, P.; Chapter 1: Achieving high dry-matter intake from pasture with grazing dairy cows. In: Fresh herbage for dairy cattle, the key to a sustainable food chain; Elgersma, A., Dijkstra, J. & Tammingna, S., Eds.; Wageningen UR Frontis Series 18; Springer, Dordrecht, Netherlands, 2006; pp 1-26.

4. Descalzo, A.M.; Rossetti, L.; Páez, R.; Grigioni, G., García, P. T.; Costabel, L.; Negri, L.; Antonacci, E.; Salado, G.; Bretschneider, G.; Gagliostro, G.; Comerón, E.; Taverna, M. A. Differential Characteristics of Milk Produced in Grazing Systems and Their Impact on Dairy Products. In Milk Production - Advanced Genetic Traits, Cellular Mechanism, Animal Management and Health, 2012; pp 339-368.

5. Vincenzetti, S.; Polidori, P.; Mariani, P.; Cammertoni, N.; Fantuz, F.; Vita, A.: Donkey’s milk protein fractions characterization. Food Chem. 2008, 106 (2) 640-649.

6. Restani, P.; Ballabio, C.; Di Lorenzo, C.; Tripod,i S.; Fiocchi, A.: Molecular aspects of milk allergens and their role in clinical events. Anal Bioanal Chem. 2009, 395 (1) 47-56.

7. Guo, H. Y.; Pang, K.; Zhang, X. Y.; Zhao, L.; Chen, S. W., Dong, M. L.: Compo-sition, physiochemical properties, nitrogen fraction distribution, and amino acid profile of donkey milk. J. Dairy Sci. 2007, 90 (4), 1635-1643.

8. Floris, R.; Recio, I.; Berkhout, B.; Visser, S.: Antibacterial and antiviral effects of milk proteins derivatives thereof. Curr. Pharm. Design. 2003, 9, 1257-1275.

9. Šarić, Lj.; Šarić, B.; Mandić, A.; Torbica, A.; Tomić, J.; Cvetković, D.; Okanović, Đ.: Antibacterial properties of Domestic Balkan donkeys’ milk. Int. Dairy J. 2012, 25, 142-146.

10. Fox, P. F. The major constituents of milk. In Dairy processing improving quality; Smit G., Eds.; Wood head Publishing Limited: Cambridge, England, 2004; pp 6-41.

11. Gastaldi, D.; Bertino, E.; Mont,i G.; Baro, C.; Fabris, C.; Lezo, A.; Medana, C.; Baiocchi, C.; Mussap, M.; Galvano, F.; Conti, A.: Donkey’s milk detailed lipid composition. Front Biosci. 2010, E2, 537-546.



194

12. Chiofalo, B.; Azzara, V.; Liotta, L.; Chiofalo, L. The chemical and physical para-meters of Ragusana ass’s milk during lactation. In Proceedings of the 6th Congress ‘New findings in equine practice; Compobasso, Italy, 2004; pp 77-84.

13. FIL-IDF, International Dairy Federation: Whole milk. Brussels, Standard: Determi-nation of milk fat, protein and lactose content; 2000, 141.

14. GEA Niro, Analytical methods for dry milk products (4th ed.). Søborg, Denmark, 1978.

15. Tidona, F.; Sekse, C.; Criscione, A.; Jacobse, M.; Bordonaro, S.; Marletta, D.; Vegarud, G. E.: Antimicrobial effect of donkeys’ milk digested in vitro with human gastrointestinal enzymes. Int. Dairy J. 2011, 21, 158-165.

16. Folch, J.; Lees, M.; Stanley, G. H. S.: A simple method for the isolation and puri-fication of total lipids from animal tissues. J. Biol Chem. 1957, 226, 497-509.

17. Ivanković, A.; Ramljak, J.; Štulina, I.; Antunac, N.; Bašić, I.; Kelava, N.; Miljenko, K.: Characteristics of the lactation, chemical composition and milk hygiene quality of the Littoral-Dinaric ass. Mljekarstvo. 2009, 59 (2) 107-113.

18. Mariani, P.; Summer, A.; Martuzzi, F.; Formaggioni, P.; Sabbioni, A.; Catalano, A. L.: Physicochemical properties, gross composition, energy value and nitrogen frac-tions of Halflinger nursing mare milk throughout 6 lactation months. Anim Res. 2001, 50, 415-425.

19. Costanzo, A. Characterization of donkey milk proteins by a proteomic approach. Ph.D Thesis, Universita degli Studi di Napoli Federico II, march 2013.

20. Martini, M.; Altomonte, I.; Salari, F.: Amiata donkeys: fat globule characteristics, milk gross composition and fatty acids. Ital. J. Anim. Sci. 2014, 13, 123-126.

21. Alabiso, M.G.; Russo, C;, Giosuė, M. L.; Torrisi, C.: La produzione di late nell 'arco di una intera lattazione in asine alleviate in Sicilia. In: Proc. 7th Congress New Findings in Equine Practices; Lodi, Italy, 2005; pp 153-160.

22. Salimei, E.; Fantuz, F.: Equid milk for human consumption. Int. Dairy J. 2012, 24, 130-142.

23. Malacarne, M.; Martuzzi, F.; Summer, A.; Mariani, P.: Protein and fat composition of mare's milk: some nutritional remarks with reference to human and cow's milk. Int. Dairy J. 2002, 12, 869-877.

ПРОМЕНЕ САСТАВА МЛЕКА МАГАРИЦЕ БАЛКАНСКЕ РАСЕ ТОКОМ ПЕРИОДА ЛАКТАЦИЈЕ

Јасмина М. Лазаревић, Татјана Т. Тасић, Сања Ј. Поповић, Војислав В. Бањац,

Оливера М. Ђурагић, Бојана М. Кокић, Ивана С. Чабаркапа

Универзитет у Новом Саду, Научни институт за прехрамбене технологије, Булевар Цара Лазара 1, 21000 Нови Сад, Србија

Промене састава млека су испитиване на 10 магарица аутохтоне балканске расе током раног, средњег и касног периода лактације. Животиње су се храниле на паш-њаку током ране и средње лактације, док су додаци исхрани (ливадско сено, куку-



195

руз и кукурузна стабљика) уведени током касног периода лактације. Циљ овог ис-траживања је било испитивање састава млека магарица током различитих периода лактације. На основу добијених резултата примећено је да су нутритивне промене млека биле највеће током раног и средњег периода лактације. Хемијском анализом је утврђено да је млеко магарице имaло низак садржај суве материје (8,82-9,68%), млечне масти (0,54-0,67%) и протеина (1,44-1,79%), док је садржај лактозе био ви-сок (6,06-7,12%). Концентрација витамина C је значајно варирала (P <0,05) током периода лактације и била је веома висока (12,84-26,89 μg/mL) у поређењу са мле-ком других врста животиња. Такође, садржај лизозима (1,95-3,29 g /L) и лактофе-рина (1,56-3,14 g/L) је био повећан током раног и средњег периода лактације. Мле-ко магарице није показало статистички значајне разлике у садржају MUFA, n6-PUFA и UFA, док је садржај n3- PUFA, n6/n3 и UFA/SFA значајно варирао (P <0.05) током периода лактације. Кључне речи: млеко магарице балканске расе, фракције казеина, фракције проте-

ина сурутке, масне киселине

Received: 15 August 2017. Accepted: 11 October 2017.

APTEFF, 48, 1-323 (2017) UDC: 543.544+544.722.123]:547.538:544.122 https://doi.org/10.2298/APT1748197L BIBLID: 1450-7188 (2017) 48, 197-209


197

CHROMATOGRAPHIC LIPOPHILICITY AND PHARMACOKINETIC BEHAVIOR OF SOME NEWLY SYNTHESIZED STYRYL LACTONE

STEREOISOMERS

Dаvоr М. Lоnčаr1*, Nаtаšа P. Мilоšеvić2, Jasmina S. Vitas1, Milica Ž. Karadžić1, Lidija R. Jevrić1, Goran Benedeković3

1 University of Novi Sad, Faculty of Technology Novi Sad, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia

2 University of Novi Sad, Faculty of Medicine, Department of Pharmacy, Hajduk Veljkova 3, 21000 Novi Sad, Serbia

3 University of Novi, Faculty of Sciences, Department of Chemistry, Biochemistry and Environmental Protection, Trg Dositeja Obradovića 3, 21000 Novi Sad, Serbia

The RP-HPLC retention constants of several newly synthesized cytotoxic styryl lacto-ne stereoisomers were determined as parameters of their lipophilicity. Stereochemistry of the compounds has an effect on the retention behavior of only tricyclic isomers. For them, the difference in retention and chromatographic lipophilic parameters are signifi-cant. The chromatographic lipophilic parameters of all compounds were correlated with a fourteen computer calculated lipophilic parameters, log P, and pharmacokinetic pa-rameters. Significant linear correlations (R2

adj>0.90) were obtained between MLog P and affinity for plasma proteins binding. The correrelations for the other pharmacokinetic parameters were improved by introducing some more molecular descriptors. Multiple linear regression analysis suggested that the volume of distribution depended on the lipopholicity and Ka

HSA (HSA-human serum albumin) and the absorption through mem-brane and permeability in the jejunum depend on the lipophilicity and hydrogen bond donors. The tested compounds showed a potent to moderate cytotoxic activity toward several human cancer cells and poor permeation through the blood brain barrier. KEY WORDS: styryl lactone isomers, chromatographic lipophilicity, pharmacokinetic

parameters

INTRODUCTION

Today, cancer is one of the major causes of death worldwide. In traditional medicine, some medicinal plants have been used for treatment of cancer. Species from the Gonio-thalamus genus (Family Annonaceae) have been used in traditional medicine to treat different diseases. Styryl lactones are a group of secondary metabolites, mainly isolated from the Goniothalamus species that have demonstrated cytotoxicity towards different human tumor cells (1-3). The synthetic chemist today have an essential role to play in the development of efficient syntheses of these compounds and their analogues, to optimize * Corresponding author: Davor M. Lončar, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




198

the biological activity. A large number of styryl lactone derivatives have been synthesi-zed and evaluated for their cytotoxicity against different human malignant cell lines (4-10). Many of the synthesized styryl lactones exhibited strong citotoxic activity. Lipophilicity is a property that has a major effect on the absorption, distribution, me-tabolism, excretion and toxicity (ADMET/Tox) properties, as well as pharmacological activity of compounds. The lipophilicity is quantitatively characterized as log P - the logarithm of the ratio of concentration of an analyte in two phases of an equilibrated 1-octanol-water system. Reversed-phase high performance liquid chromatography (RP HPLC) is often used to deter-mine lipophilicity of compounds. Chromatographic partitioning between the non-polar stationary phase and an aqueous mobile phase in liquid chromatography seems to be similar to the partitioning in the biological systems. Nowadays, chromatographic lipophilicity parameter log kw, often referred to as log ko (the extrapolated log k value to a zero organic phase concentration, i.e. neat water) is frequently regarded as an independent descriptor (11). Our attention has been focused on newly synthesized styryl lactone isomers as cyto-toxic candidates (7, 8). In this work the retention constants of this compounds by RP HPLC as alternative lipophilicity parameters were determined and quantitatively related to the in silico calculated log P and pharmacokinetic parameters (QSRR modeling).

EXPERIMENTAL

Chromatographic analysis

The newly synthesized styryl lactones isomers were analyzed by RP-HPLC using methanol-water eluent with five different volume fractions of methanol (60-70 v/v), as described in our previous paper (12). The constants of the linear relationship between the retention of compounds, log k, and volume fraction of methanol, φ (log k = log ko + Sφ), log ko and S were used for QSRR modeling.

Molecular descriptors

Fourteen log P values for the compounds tested were taken from the literature (12). Calculation of the other physicochemical characteristics of the compounds was perfor-med using Molinspiration online program (13). Pharmacokinetic parameters were calcu-lated using acd/i-lab-2.0 software package (14).

Chemometric regression Processing of the results was done using linear regression and multiple linear regres-sion. Graphing and statistical analysis were performed with the software package Ori-ginPro 8 and NCSS 2007 (15). The validation was performed using the following para-meters: squared correlation coefficient (R2) and adjusted R2 (R2

adj), Fisher test (F), root mean square error (RMSE), cross-validation coefficient (R2

cv), predicted residual error sum of squares (PRESS), total sum of squares (TSS), variance inflation factor (VIF).



199


Chromatographic lipophilicity parameters The 2D structural formulas and the names of examined compounds are presented in Table 1. The molecules 7(R)-(1-5) and 7(S)-epimers (1a-5a) have a common structural segment, furano-furone bicyclic core tetrahydro-furo[3,2-b]furan-2-one. The retention and retention parameters log ko and S of the linear relationship between the retention (log k) and volume fractions of methanol (φMeOH) in the methanol-water eluent (log k = log ko + Sφ) from our previous paper (12) are shown in Table 2 and Figure 1.

Table 1. The 2D structural formulas and the names of examined compounds

Comp. Compound structure and name Comp. Compound structure and name

1

(+)-goniofufurone

1a

7-epi-(+)-goniofufurone

2

3,6-Anhydro-2-deoxy-5,7-O-methylidene-7-C-phenyl-D-glycero-D-

ido-heptono-1,4-lactone

2a

3,6-Anhydro-2-deoxy-5,7-O-methylidene-7-C-phenyl-L-glycero-D-ido-heptono-1,4-

lactone

3

(7R)-3,6-Anhydro-2-deoxy-7-C-cinnamoyloxy-5,7-O-methylene-D-ido-

heptono-1,4-lactone

3a

(7S)-3,6-Anhydro-2-deoxy-7-C-cinnamoyloxy-5,7-O-methylene-D-ido-

heptono-1,4-lactone

4 3,6-Anhydro-5-O-(4-

methoxycinnamoyl)-2-deoxy-7-C-phenyl-D-glycero-D-ido-heptono-1,4-

lactone

4a

3,6-Anhydro-5-O-(4-methoxycinnamoyl)-2-deoxy-7-C-phenyl-L-glycero-D-ido-

heptono-1,4-lactone



200

Table 1. Continuation

Comp. Compound structure and name Comp. Compound structure and name

5

3,6-Anhydro-7-O-(4-nitrocinnamoyl)-2-deoxy-7-C-phenyl-D-glycero-D-ido-

heptono-1,4-lactone

5a

3,6-Anhydro-7-O-(4-nitrocinnamoyl)-2-deoxy-7-C-phenyl-L-glycero-D-ido-

heptono-1,4-lactone

Table 2. Retention data for the examined styryl lactone isomers (12)

Compounds logkw S R2 SD p 1 1.368 -3.068 0.9933 0.016 0,0002 1a 1.214 -2.944 0.9963 0.008 <0.0001 2 2.369 -3.592 0.9932 0.014 <0.0001 2a 1.938 -3.104 0.9959 0.009 0.0001 3 3.154 -4.428 0.9995 0.004 <0.0001 3a 2.533 -3.844 0.9977 0.008 <0.0001 4 3.236 -4.540 0.9991 0.006 <0.0001 4a 3.291 -4.544 0.9976 0.010 <0.0001 5 4.333 -5.960 0.9978 0.013 <0.0001 5a 4.295 -5.908 0.9985 0.010 <0.0001

The retention sequence of the examined compounds is based on their polarity, more exactly on their hydrophobicity. The highest polarity isomers 1 and 1a, which have OH function, are the least hydrophobic and have the smallest retention, log ko, typically when a reversed phase mechanism is present. Namely, the retention in reversed-phase liquid chromatography occurs due to non-specific hydrophobic interaction of the solute with the stationary phase. The hydrophobic nature of the molecule and retention increase by intro-ducing a cyclic ring, and/or cinnamic acid ester groups (compounds 2-5a). The constant S decreases with the increase in the hydrophobicity of the compounds (Figure 1). The isomeric pairs 1 and 1a, 4 and 4a, 5 and 5a, have an almost the same retention (Table 2). However, the retention of the isomeric pairs 2 and 2a as well as 3 and 3a are significantly different from one another (Δlog ko(2-2a)=0.4304; Δlog ko(3-3a)=0.6204. In addition, the 7(R)-configured epimers 2 and 3 have higher retention than 7(S)-epimers 2a and 3a, (Table 2). Also, the isomers 3 and 3a with cinnamoyl function at the 7C-position have higher retention than the isomers 2 and 2a, which have a phenyl function in the same position (Figure 1), suggesting that the cinnamoyl function increases the hydropho-bicity of the compounds.



201

Figure 1. Retention behavior of the styryl lactone isomers

The retention behavior of these molecules can be explained by their structural featu-res. Thus, compounds 1, 1a, 4, 4a, 5 and 5a have at least one free hydroxyl group. The-refore, they have a higher affinity toward the mobile phase, where their retention is less dependent on the stereochemistry. In contrast, the very lipophilic molecules 2, 2a, 3 and 3a have a higher affinity for the stationary phase (C18), whereby their lipophilicity de-pends significantly on the spatial orientation of the aromatic rings. As can be seen in Figure 2, the aromatic residues of molecules 2 and 3 are exo-oriented relative to the tetrahydrofurodioxine bicyclic ring system. Therefore, they are spatially more accessible for binding with the hydrophobic stationary phase, and show a higher retention. On the other hand, the endo-oriented aromatic groups in the compounds 2a and 3a are sterically more hindered, and therefore their binding to the stationary phase is weaker, resulting in a minor retention.

Figure 2. Stereochemical features of compounds 2 and 3 (exo-isomers) and of 2a and 3a (endo-isomers)

1,1a 2,2a 3,3a 4,4a 5,5a

1,0

1,5

2,0

2,5

3,0

3,5

4,0

4,5 7(R)-configured 7(S)-configured

log

k o

compounds

1,1a 2,2a 3,3a 4,4a 5,5a

-6,0

-5,5

-5,0

-4,5

-4,0

-3,5

-3,0

-2,5 7(R)-configured 7(S)-configured

s

compounds



202

Correlation of the retention constants and computer calculated log P

The retention constants log ko and S were correlated to the fourteen theoretically calculated log P (12). A best statistical quality was obtained for the relationship between log ko or S values and MLog P (Equations 1 and 2, Figure 3). MLog P = 0.6232 (±0.0539) log kо - 0.3142(±0.1595) [1]

R2adj = 0.9365; F = 133.7555; p = 2.84 x 10-6

MLog P = -0.5954 (±0.0799) S - 1.0827 (±0.3454) [2] R2

adj = 0.8582; F = 55.4825; p = 7.27 x 10-5)

a) b)

Figure 3. Relationships between: (a) log ko and MLog P; (b) S and MLog P

It is evident from Figure 3 and Equations 1 and 2 that the chromatographic retention constant log ko and S could be successfully applied to predict lipophilicity of the styryl lactone isomers since the correlation coefficient of the linear regressions was satisfactory, i.e. R2 > 0.64. These results are consistent with several literature data (16-19).

Correlation of the retention constants and pharmacokinetic predictors In order to analyze the influence of lipophilicity on the biological activity of the com-pounds, the lipophilic chromatography constants log ko and S (Table 2) are correlated with the in silico pharmacokinetic properties: volume of distribution (Vd), constant of adsorption (ka), plasma protein binding affinity (PPB), human effective permeability in jejunum (Peff), and logarithm of the blood-brain barrier partition coefficient (log BBB). The pharmacokinetic predictors, calculated using the computation program i-lab 2.0 (14), are presented in Table 3. Some physicochemical properties of the tested compounds calculated by the same program are given in Table 4.

1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5

0,5

1,0

1,5

2,0

2,5

y = 0.6232x - 0.3142

R2

adj = 0.9365

Mlo

gP

log ko

-6,0 -5,5 -5,0 -4,5 -4,0 -3,5 -3,0 -2,5

0,5

1,0

1,5

2,0

2,5y = - 0.5954x - 1.0827

R2

adj = 0.8582

Mlo

gP

S



203

Table 3. Pharmacokinetic parameters of the examined compounds (14)

Compound Vd (L/kg) ka (min-1) PPB (%) Peff (cm/s) log BBB 1, 1a 0.86 0.041 18.58 0.000599 0.12 2, 2a 1.09 0.060 64.99 0.000875 0.46 3, 3a 1.05 0.056 51.75 0.000811 0.55 4, 4a 1.37 0.051 67.85 0.000740 0.77 5, 5a 1.47 0.051 85.16 0.000742 0.46

Vd - volume of distribution; ka - constant of adsorption; PPB - plasma protein binding affinity; Peff - human effective permeability in the jejunum; log BBB - logarithm of the blood-brain barrier permeation

Table 4. In silico molecular descriptors for the investigated compounds

Clog P, MLog P, log D - calculated logarithm of octanol-water partition coefficients; *Lipinski's rule of 5; **Clark's rule; MW - molecular weight; HBD - number of hydrogen bond donors; HBA - number of hydrogen bond acceptors; N+O - sum of nitrogen and oxygen atoms; PSA(TPSA)-(total) polar surface area; Ka

HSA (HSA-human serum albumin) The apparent volume distribution in the body (Vd) is a key pharmacokinetic parameter which determines the extent of compound distribution. It represents a measure of the relative partitioning of the compound between the plasma (the central compartment) and the tissues, and depends on its lipophilicity (20). Compounds that are highly lipophilic tend to bind to the tissue components (e.g. proteins, lipids), and their blood concentration is very low. The styryl lactone isomers have a small volume of distribution (0.86-1.47 L/kg) (Tab-le 3), indicating that no serious accumulation of compounds in the fat tissue occurs. The linear correlations between Vd and chromatographic lipophilicity parameters log ko and S are presented in Equations 3 and 4. Vd = 0.1983 (±0.0292) log ko + 0.6181 (±0.2178) [3] R2

adj = 0.8336; F = 46.1022; p =1.39x10-4 Vd = -0.1802 (±0.0307) S + 0.4216 (±0.1316) [4] R2

adj = 0.7881; F = 34.4758; p = 3.74x10-4 Equations 3 and 4 show a relatively good linear correlation of Vd with the parameters log ka (R2

adj=0.8336) and S (R2adj=0.7881). But, Vd is considerably better defined as a

function of the parameters lipophilicity log ko or S, and the plasma protein binding (log Ka

HSA). The multiple linear regression models of these dependencies are shown in Equations 5 and 6 and in Figure 4.

Comp Clog P

< 5* MLog P > 4.15*

log D 1-3**

MW < 450*,**

HBD < 5*

HBA < 10*

N+O < 6**

Clog P-(N+O) > 0**

PSA < 70* Ka

HSA

1, 1a 0.019 0.41 0.81 250.25 2 5 5 -4.981 75.99 3.70 2, 2a -0.104 1.08 2.61 262.26 0 5 5 -5.104 53.99 4.11 3, 3a 0.396 1.34 2.52 332.30 0 7 7 -6.604 80.29 4.00 4, 4a 2.259 1.92 3.25 410.42 1 7 7 -4.741 91.29 4.73 5,5a 2.162 2.32 3.27 425.39 1 9 9 -6.838 130.89 4.57



204

Vd = 0.0874(±0.0220) log ko - 0.3678(±0.0601) log KaHSA - 0.6274(±0.2068) [5]

R2=0.9767; R2adj = 0.9701; R2

cv=0.9450; F = 146.7832; p = 1.93x10-6; VIF=3.14; RMSE=0.0405; PRESS=0.0271; TSS= 0.4938; PRESS/TSS=0.0549

Vd = -0.0809(±0.0163) S + 0.3941(±0.0450) KaHSA - 0.8354(±0.1446) [6]

R2=0.9832; R2adj = 0.9784; R2

cv=0.9644; F = 204.5847; p= 6.17x10-7; VIF=2.43; RMSE=0.0344; PRESS=0.0176; TSS=0.4938; PRESS/TSS=0.0356

a) b)

Figure 4. Multiple linear regression models of Vd with (a) log ko and log KaHSA ;

(b) S and log KaHSA

According to Lipinski's rule (21), which estimates the compounds absorption, satisfactory absorption or permeation can be expected, since all of the compounds ana-lyzed have less than 5 H-bond donors, 10 H-bond acceptors, the molecular weight smaller than 500, and the calculated log P (Clog P) bellow 5 (or MLog P bellow 4.15), (Table 4). The absorption constants, ka have a small value, 0.041-0.060 min-1 (Table 3), indicating a quite short absorption half-times of the test compounds (17, 18). By using the correlation ka to log ko or S and HBD we obtain the best following multiple regression model: ka = 3.70x10-4(±5.09 -4) log ko - 8.16x10-3(±7.32x610-45) HBD + 5.73x10-2(±1.79x10-3) [7] R2=0.9513; R2

adj = 0.9374; R2cv=0.9101; F = 66.8767; p = 2.6x10-5;

VIF=1.06; RMSE=1.68x10-3; PRESS=3.65x10-5; TSS=4.06x10-4; PRESS/TSS=0.0899 ka = 3.00x10-4(±5.16x10-4) S - 8.26x10-3(±7.19x610--45) HBD + 5.71x10-2(±2.34x10-3) [8] R2=0.9536; R2

adj = 0.9361; R2cv=0.9083; F = 68.3432; p = 2.7x10-5;

VIF=1.00; RMSE=1.70x10-3; PRESS=3.72x10-5; TSS=4.06x10-4; PRESS/TSS=0.0809

02

46

8

10

0,8

0,9

1,0

1,1

1,2

1,3

1,4

1,5

1,01,5

2,02,5

3,03,5

4,04,5

Vd

log k olog K HSAa

02

46

8

10

0,8

0,9

1,0

1,1

1,2

1,3

1,4

1,5

-2,5-3,0

-3,5-4,0

-4,5-5,0

-5,5-6,0

Vd

Slog K HSAa



205

The equations [7] and [8] shows that the lipophilicity of the compounds has a domi-nant influence on ka. The analysis of Linnankoski et al. (22) showed that the most im-portant parameters describing log ka of 22 structurally diverse drugs were polar surface area (PSA), the number of hydrogen bond donors (HBD) and octanol-water partition coefficient at pH 6.0 (logD6.0). Milošević et al. (18) tested also multiple linear models and the best statistical parameters were found when log ka of the seco-androstene derivatives was presented as a function of the retention constant RM

o, MW and TPSA Many cоmpоunds bind reversibly to plasma proteins, among which mainly plasma al-bumin, lipoproteins and glycoproteins are involved. Plasma protein binding (PPB) is im-portant in pharmacokinetics, since it influences the volume of distribution, degree of me-tabolism and rate of elimination. The compound molecules that are bound to plasma pro-tein cannot permeate through cell membranes by passive transcellular or paracellular per-meation. The highest degree of binding to plasma protein have the isomers 5 and 5a (Table 3). In the plasma, the isomers 1 and 1a, 4 and 4a, 5 and 5a are bound to the human serum albumin (HSA), and the isomers 2 and 2a, 3 and 3a to lipoproteins (LP), (14). For the compounds that bind to HAS, the relationships between the PPB values and cons-tants, log ko and S, are linear (Figure 5), with very good statistical parameters (Equations 9, 10).

a) b)

Figure 5. Relationships between: (a) PPB and log ko; (b) PPB and S

HAS - human serum albumin binding (compounds 1, 1a, 4-5a); LP - lipoproteins binding (compounds 2-3a)

PPB = -22.3705(±1.1627) log ko - 8.9366(±13.2480) [9]

R2adj = 0.9866; F = 370.2058; p = 4.30x10-5

PPB = -22.8754(±2.8487) S - 45.6055(±13.2480) [10] R2

adj = 0.9270; F = 64.4831; p = 0.00131

1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,510

20

30

40

50

60

70

80

90 HSA

adjR2=0.9866 LP

PP

B

log ko

-6,0 -5,5 -5,0 -4,5 -4,0 -3,5 -3,0 -2,510

20

30

40

50

60

70

80

90 HSA

adjR2 = 0.9270 LP

PP

B

S



206

Human intestinal permeability, Peff is a key drug-related parameter governing the gastrointestinal drug absorption process. Tested styryl lactones are predict to be absorbed by trancellular passive permeation, i.e. diffusion through the lipid bilayer of the cell membrane. Hence, the permeability of molecules depends on lipophilicity. As expected, the lowest values Peff have the most polar and at least a lipophilic compounds 1 and 1,a. Very good correlation is obtained when the descriptors used logko or S and HBD (Equations 11, 12). Peff = 4.413.89x10-6 (±8.26x10-6) log ko - 1.18x10-4 (±1.14x10-5) HBD + 8.35x10-4 (±2.79x10-5) [11] R2=0.9430; R2

adj = 0.9267; R2cv=0.8943; F = 57.8942; p = 4.4x10-5;

VIF=1.06; RMSE=2.62x10-5; PRESS=8.93x10-9; TSS=5.68x10-6; PRESS/TSS=1.58x10-3

Peff = -3.48x10-6 (±8.04x10-6) S - 1.19x10-4 (±1.12x10-5) HBD + 8.34x10-4 (±3.64x10-5) [12] R2=0.9422; R2

adj = 0.9257; R2cv=0.8928; F = 57.0641; p = 4.6x10-5;

VIF=1.00; RMSE=2.64x10-5; PRESS=9.06x10-9; TSS=8.45x10-8; PRESS/TSS=0.1072 The best Peff model Winiwarter et al. (23) was obtained based on 13 passively trans-cellularly absorbed compounds using the variables HBD, PSA and either log D5.5 or log D6.5 (octanol/water distribution coefficient at pH 5.5 and 6.5, respectively). Good blood-brain barrier (BBB) penetration is necessary for central nervous system (CNS) drugs. However, the drugs for therapy in peripheral tissues may cause side effects in the brain; therefore, it is desirable that they have minimal penetration into the brain. Permeation of drugs through the blood-barrier is usually described as log BBB=log (cbrain/cblood), where cbrain and cblood are the equilibrium concentration of a drug in the brain and blood, respectively. Table 3 shows that the isomers 1 and 1a have low log BBB and a limited absorption through the BBB. The isomers 2-5a have moderate log BBB (0.46-0.77) and poor penetration to the brain can be predicted for them (24). There was no sig-nificant correlation between log BBB and log ko and S. These results are expected because all isomers violate one to four Clark's rules (25), as indicated by bold in Table 4. In addition, certain of these compounds show cytotoxicity towards cancer cells of some of internal organs in the human body (7, 8).

CONCLUSION The RP-HPLC method was used to evaluate lipophilic properties of some styryl lactone isomers. The retention constants log ko i S were recommended for lipophilicity expression of the examined isomers. The influence of stereostructure of the isomers on their lipophilicity was observed only among tricyclic isomers. The lipophilicity of all compounds significantly influences the pharmacokinetic parameters Vd, ka, PPB and Peff. In addition to lipophilicity, a significant molecular descriptors which influenced Vd was KaHSA, whereas ka and Peff were significantly influenced by HBD. The examined styryl lactone isomers violate Clark′s rules that favor brain permeation.



207

Acknowledgement These results are the part of the projects No. OI 1612036, financially supported by the Ministry of Education, Science and Technological Development of the Republic of Ser-bia and project No. 114-451-347/2015-02) financially supported by the Provincial Secre-tariat for Science and Technological Development of Vojvodina.

REFERENCES

1. de Fátima, A.; Kohn, L.K.; de Carvalho, J.E.; Pilli, R.A. Cytotoxic activity of goniothalamin and analogues against human cancer cells. Bioorg. Med. Chem. 2006, 14(3) 622-631.

2. Choo, C-Y.; Abdullah, N.; Diederich, M. Cytotoxic activity and mechanism of action of metabolites from the Goniothalamus genus. Phytochem. Rev. DOI 10.1007/s11101-014-9372-2 (2014) online.

3. Seyed, M.A.; Jantan, I.; Bukhari, S.N.N.A. Emerging Anticancer Potentials of Gonio-thalamin and its Molecular Mechanisms - A Review. BioMed Res. Int., 2014, 14, ID536508, 10 pages.

4. Popsavin, V.; Srećo, B.; Benedeković, G.; Francuz, J.; Popsavin, M.; Kojić, V.; Bogdanović, G. Design, synthesis and antiproliferative activity of styryl lactones related to (+)-goniofufurone. Eur. J. Med. Chem. 2010, 45, 2876-2883.

5. Popsavin, V.; Kovačević, I.; Benedeković, G.; Popsavin, M.; Kojić, V.; Bogdanović, G. Divergent synthesis of cytotoxic styryl lactones related to goniobutenolides A and B, and crassalactone D. Org. Lett. 2012, 14, 5956-5959.

6. Barcelos, R.C.; Pastre, J.C.; Caixeta, V.; Vendramini-Costa, D.B.; Caixeta, V.; Lom-gato, G.B.; Monteiro, P.A.; de Carvalho, J.E.; Pilli, R.A. Design and synthesis N-acylated Aza-goniothalamin derivatives and evaluation of their in vitro and in vivo antitumur activity. Chem. Med. Chem. 2014, 9, 2725-2743.

7. Benedeković, G,; Francuz, J.; Kovačević, I.; Popsavin, M.; Srećo Zelenović, B.; Kojić, V.; Jakimov, D.; Bogdanović, G.; Divjaković, V.; Popsavin, V. Conformatio-nally constrained goniofufurone mimics as inhibitors of tumour cells growth: Design, synthesis and SAR study. Eur. J. Med. Chem. 2014a, 82, 449-458.

8. Benedeković, G.; Popsavin, M.; Francuz, J.; Kovačević, I.; Kojić, V.; Bogdanović, G.; Divjaković, V.; Popsavin, V. Design, synthesis and SAR analysis of antitumor styryl lactones related to (+)-crassalactones B and C. Eur. J. Med. Chem. 2014b, 87, 237-247.

9. Francuz, J.; Kovačević, I.; Popsavin, M.; Benedeković, G.; Srećo Zelenović, B.; Ko-jić, V.; Jakimov, D.; Aleksić, L.; Bogdanović, G.; Srdić-Rajić, T.; Lončar, E.; Rodić, M.V.; Divjaković, V.; Popsavin, V. Design, synthesis and in vitro antitumour activity of new goniofufurone and 7-epi-goniofufurone mimics with halogen or azido groups at the C-7 position. Eur. J. Med. Chem. 2017, 128, 13-24.

10. Kovačević, I.; Popsavin, M.; Benedeković, G.; Kesić, J.; Kojić, V.; Jakimov, D.; Srdić-Rajić, T.; Bogdanović, G.; Divjaković, V.; Popsavin, V. Synthesis and in vitro



208

antitumour activity of crassalactone D, its stereoisomers and novel cinnamic ester derivatives. Eur. J. Med. Chem. 2017, 7(134) 293-303.

11. Valkó, K.I. Lipophilicity and biomimetic properties measured by HPLC to support drug discovery. J. Pharm. Biomed. Anal. 2016, 130, 35-54.

12. Karadžić, M.Ž; Lončar, D; Benedeković, G.; Kovačević, I.; Popsavin, V.; Kovačević, S.Z.; Jevrić, L.R.; Podunavac-Kuzmanović, S.O. A comparative study of chromato-graphic behavior and lipophilicity natural styryl lactones, their derivatives and ana-logues. E. J. Pharm. Sci. 2017, 105, 99-107.

13. http://www.molinspiration.com. 14. http://www. ilab.acdlabs.com/i-Lab2/. 15. Hintze, J. NCSS and GESS, NCSS, LLC; Kaysville, Utah. http://www.ncss.com 16. Komsta, L.; Skibiński, R.; Berecka, A.; Gumieniczek, A.; Radkiewicz, B.; Radoń, M.

Revisiting thin-layer chromatography as a lipophilicity determination tool - a compa-rative study on several techniques with a model solute set. J. Pharm. Biomed. Anal. 2010, 53, 911-918.

17. Perišić-Janjic, N.; Kaliszan, R.; Milosevic, N.; Uscumlic, G.; Banjac, N. Chromato-graphic retention parameters in correlation analysis with in silico biological descrip-tors of a novel series of N-phenyl-3-methyl succinimide derivatives. J. Pharm. Biomed. Anal. 2013, 72, 65-73.

18. Milošević, N.P.; Stojanović, S.Z.; Penov-Gaši, K.; Perišić-Janjić, N.; Kaliszan, R. Reversed- and normal-phase liquid chromatography in quantitative structure reten-tion-property relationships of newly synthesized seco-androstene derivatives. J. Pharm. Biomed.l Anal. 2014, 88, 636-642.

19. Marciniec, K.; Bafeltowska, J.; Maślankiewicz, M.J.; Buszman, E.; Boryczka, S. Determination of lipophilicity of quinolinesulfonamides by reversed-phase HPLC and theoretical calculations. J. Liq. Chromatogr. Related Technol. 2016, 39(15) 702-709.

20. Stepinsky, D. Use unbound volumes of drug distribution in pharmacokinetic calcula-tions. Eur. J. Pharm. Sci. 2011, 42, 91-98.

21. Lipinski, C.A.; Lombardo, F.; Dominy, B.W.; Feeney P.J. Experimental and compu-tational approaches to estimate solubility and permeability in drug discovery and development settings. Adv. Drug Deliv. Rev. 2001, 46, 3-26.

22. Linnankoski, J.; Makela, J.M.; Ranta, V.P.; Urtti, A.; Ylipertulla, M. Computational prediction of oral drug absorption based on absorption rate constants in humans. J. Med. Chem. 2006, 49, 3674-3681.

23. Winiwarter, S.; Bonham. N.M.; Ax, F.; Hallberg, A.; Lennernas, H.; Karlen, A. Cor-relation of human jejunal permeability (in vivo) of drugs with experimentally and theoretically derived parameters. A multivariate data analysis approach. J. Med. Hem. 1998, 41, 4939-4949.

24. van der Waterbeemd, H.; Gifford, F. ADMET in silico modeling: towards prediction paradise? Nat. Rev. Drug Discov. 2003, 6, 192-204.

25. Clark, D.E. In silico prediction of blood-brain barrier permeation. Drug Discov. Today. 2003, 8, 927-933.



209

ХРОМАТОГРАФСКА ЛИПОФИЛНОСТ И ФАРМАКОКИНЕТИЧКО ПОНАШАЊЕ НЕКИХ НОВО СИНТЕТИСАНИХ СТЕРЕОИЗОМЕРА

СТИРИЛ ЛАКТОНА

Давор М. Лончар1, Наташа П. Милошевић2, Jaсминa С. Витaс1, Mилицa Ж. Кaрaџић1, Лидиja Р. Jeврић1, Горан Бенедековић3

1 Унивeрзитeт у Нoвoм Сaду, Teхнoлoшки фaкултeт Нoви Сaд, Булeвaр цaрa Лaзaрa 1, 21000 Нoви Сaд, Србиja

2 Унивeрзитeт у Нoвoм Сaду, Дeпaртмaн зa фaрмaциjу, Meдицински фaкултeт, Хajдук Вeљкoвa 3, 21000 Нoви Сaд, Србиja

3 Унивeрзитeт у Нoвoм Сaду, Природно-математички факултет, Дeпaртмaн зa хемију, биохемију и заштиту околине, Доситеја Обрадовића 3, 21000 Нoви Сaд, Србиja

Реверзно-фазном течном хроматографијом под високим притиском одређене су ретенционе константе ново синтетизованих изомера стирил лактона као параметри њихове липофилности. Стерохемија има утицаја на ретенцију само код трициклич-них изомера. Код њих је значајна разлика у ретенцији и хроматографским пара-метрима липофилности. Хроматографски параметри липофилности свих испитива-них једињења корелирани су са четрнаест компјутерски израчунатих log P и фарма-кокинетичким параметрима. Значајне линеарне корелације (R2

adj>0.90) добијене су са MLog P и афинитетом везивања за протеине плазме. Релације за остале фармако-кинетичке параметре побољшане су увођењем и других дескриптора. Вишеструка линеарна регресиона анализа указује да запремина дистрибуције зависи од липо-филности и Ka

HSA (HSA-хумани серум албумин) а апсорпција кроз липидне био-мембране и пермеабилност у јејунуму зависе од липофилности и донорских водо-ничних веза. Испитиванa једињења имају цитотоксичну активност према ћелијама више хуманих канцера и показују слабу пермеацију кроз крвно-мождану баријеру. Кључне речи: изомери стирил лактона, хроматографска липофилност, фармакоки-

нетички параметри


APTEFF, 48, 1-323 (2017) UDC: 66.011:66.067.2:664.642 https://doi.org/10.2298/APT1748211L BIBLID: 1450-7188 (2017) 48, 211-220


211

MULTI-OBJECTIVE OPTIMIZATION OF MICROFILTRATION OF BAKER’S

YEAST USING GENETIC ALGORITHM

Nataša Lj. Lukić*, Marija Božin-Dakić, Jovana A. Grahovac, Jelena M. Dodić, Aleksandar I. Jokić


This paper presents a multi-objective optimization model by applying genetic algo-rithm in order to search for optimal operating parameters of microfiltration of baker’s yeast in the presence of static mixer as a turbulence promoter. The operating variables were the suspension concentration, transmembrane pressure, and feed flow rate. Two conflicting objective functions, maximizing the permeate flux and maximizing the reduc-tion of energy consumption, were considered. This multi-objective optimization problem was solved by using the elitist non-dominated sorting genetic algorithm in the Matlab R2015b software. The Pareto fronts along with the process decision variables cor-respondding to the optimal solutions were obtained. It was found that lower suspension concentrations (2-4.5 g/L), feed flow rate in the range 109-127 L/h, and transmembrane pressure of 1 bar were the optimal process parameters which yielded maximum permeate flux (177-191 L/(m2h)) and maximum reduction of energy consumption (44-50%). Final-ly, the results were compared with the previously published results obtained by applying desirability function approach. Given that genetic algorithms have generated multiple solutions in a single optimization run, the study proved that genetic algorithms are preferable to classical optimization methods. KEY WORDS: multi-objective optimization, genetic algorithm, baker’s yeast, static mixer,

microfiltration

INTRODUCTION Simultaneous optimization of two or more objective functions is quite common in most real-world engineering problems. Furthermore, objective functions are usually con-flicting and, hence, there is no single solution that is the best (global optimum) but a set of equally good (non-dominated) solutions, known as Pareto optimal solutions. Pareto optimal solutions are those solutions for which an improvement in one objective can be obtained only by worsening at least one other objective. The main purpose of every multi-objective optimization (also called multicriteria optimization, vector optimization or Pareto optimization) method is to identify a Pareto optimal solution set. Afterwards, * Corresponding author: Nataša Lj. Lukić, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




212

the so-called decision maker selects “preferred” or the “best” solution using additional in-formation which is most likely subjective (intuitive) and difficult to quantify (1). Generally speaking, there are two main approaches to solving multi-objective optimi-zation problems: classical methods and evolutionary algorithms. By using classical opti-mization methods, conflicting objectives are aggregated into a single scalar objective function using arbitrary weighting factors. These methods are quite ineffective, despite being simple and straightforward, mainly because in order to find multiple solutions they have to be applied many times. Besides, the obtained optimal solutions are highly de-pendent on the chosen values of weighting factors. Also, in the cases when the non-con-vexity of the objective function gives rise to a duality gap, some of the optimal solutions may never be found (2, 3). In the past two decades, due to the emergence of faster and more powerful computers, various multi-objective evolutionary algorithms, such as gene-tic algorithms, evolutionary programming, evolution strategies and genetic programming, have been successfully used to identify multiple Pareto optimal solutions in only one si-mulation run (4). Genetic algorithm (GA) is a heuristic optimization and search technique based on the evolutionary principles of natural selection and genetics. Because of their ability to per-form simultaneous search of different regions in a solution space, genetic algorithms are able to find a diverse set of solutions, especially in the cases when the objective functions have non-convex, discontinuous, or multi-modal characteristics (5). A wide variety of multi-objective GA methods, such as Multi-objective Genetic Algorithm (MOGA), Niched Pareto Genetic Algorithm (NPGA), Weight-based Genetic Algorithm (WBGA), Random Weighted Genetic Algorithm (RWGA), Nondominated Sorting Genetic Algo-rithm (NSGA), Strength Pareto Evolutionary Algorithm (SPEA), improved SPEA (SPEA2), Pareto-Archived Evolution Strategy (PAES), Pareto Envelope-based Selection Algorithm (PESA), Region-based Selection in Evolutionary Multi-objective Optimization (PESA-II), Elitist Non-dominated Sorting Genetic Algorithm (NSGA-II), Multi-objective Evolutionary Algorithm (MEA), Micro-GA, Rank-Density Based Genetic Algorithm (RDGA), and Dynamic Multi-objective Evolutionary Algorithm (DMOEA), have been developed so far (5). A thorough explanation of some of the above-mentioned multi-ob-jective GA methods can be found elsewhere (1, 5, 6). One of the most widespread and generalized multi-objective GA methods in use is NSGA-II. Recently, genetic algorithms have been successfully used to solve multi-objective optimization problems in various membrane separation processes (7-12). The main objective of the present research was the genetic algorithm-based simulta-neous optimization of permeate flux and reduction of energy consumption obtained in the microfiltration of baker’s yeast suspension assisted by the insertion of a Kenics static mixer.



213

EXPERIMENTAL

Experimental procedure

Baker’s yeast (Saccharomyces cerevisiae) suspension was used as the feed during microfiltration. All experiments were carried out using a monotubular ceramic membrane (250 mm long, 6 mm ID and 10 mm OD) from Tami (Germany). The membrane, with an average pore size of 200 nm, had a total filtration area of 43.3 cm2. A stainless steel Kenics static mixer was used as the turbulence promoter. The selected input variables were transmembrane pressure (0.2 bar≤TMP≤1 bar), feed flow rate (80 L/h≤Q≤180 l/h), and suspension concentration (2 g/L≤C≤10 g/L). Detailed experimental procedure, along with the specifications of the filtration unit, is given elsewhere (13).

Optimization procedure As previously mentioned, the main aim of multi-objective optimization performed in this study was to identify the process conditions (i.e. input variables), such as transmem-brane pressure, feed flow rate and suspension concentration, in order to maximize per-meate flux (Jsm) and reduction of energy consumption (ER). Hence, the optimization problem is composed of two objectives (fitness functions): permeate flux and reduction of energy consumption. Having in mind the adequacy of response surface methodology (RSM) models developed by Jokić et al. (13), which were evaluated by the coefficient of determination and model p-value, RSM models were used as expressions for objective functions. Experimental ranges were taken as limits on the input variables. Therefore, the multi-objective optimization problem can be formulated mathematically as follows:

maximize Jsm TMP, Q, C =max (-69.405 + 244.696∙x1 + 1.833∙x2 - 1.303∙x3 -0.833∙x1∙x2 - 4.278∙x1x3 + 0.016∙x2x3 - 38.141∙x1

2 -0.004∙x2 2 - 0.225∙x32) [1]

maximize ER TMP, Q, C =max (-210.135 + 27.870∙x1 + 2.806∙x2 + 27.433∙x3 -0.228∙x1∙x2 - 9.775∙x1x3 - 0.075∙x2x3 + 46.477∙x1

2 -0.010∙x22 - 1.044∙x3

2) [2]

subject to bound constraints 0.2 bar ≤ TMP ≤ 1.2 bar

80 L/h ≤ Q ≤ 180 L/h2 g/L ≤ C ≤ 10 g/L

[3]

where Jsm and ER are the objective functions in the optimization algorithm, while x1, x2 and x3 are the transmembrane pressure, feed flow rate and suspension concentration, respectively. Two sets of optimizations were carried out. In the first (set A), the values of suspen-sion concentration were set to 2 g/L, 6 g/L or 10 g/L, thus assigning feed flow rate and transmembrane pressure as decision variables. In the second (set B), all three variables (C, TMP and Q) were specified as decision variables.



214

Elitist Non-Dominated Sorting Genetic algorithm (NSGA-II) The NSGA-II algorithm proposed by Deb et al. (4) introduces the concept of elitism to non-dominated sorting genetic algorithm in order to enhance the convergence proper-ties. Due to its ability to successfully solve constrained multi-objective optimization problems, NSGA-II is one of the most widely used evolutionary algorithms. In the first step of the algorithm, an initial population is generated based on the problem and constraints; afterwards, fitness function values are evaluated based on the RSM models. Next, selection of the individuals in the population is performed as a result of ranking and crowding distance. In order to generate new population (offspring), the NSGA-II algo-rithm employs genetic operators such as crossover and mutation. Afterwards, both chil-dren and parent population are combined and the individuals are selected based on the elitism and crowding distance (14). Finally, if one of the specified stopping criteria is ful-filled, the algorithm stops; otherwise the next offspring population is generated. A flow-chart describing the NSGA-II algorithm is given in Figure 1.

Figure 1. Flowchart of the NSGA-II algorithm. All calculations were performed by using the Global Optimization Toolbox within Matlab R2015b software and its inbuilt multi-objective GA function ‘gamultiobj’, which is a variant of NSGA-II. ‘Gamultiobj’ function is applicable only for minimization problems and, therefore, the maximization problems in this paper (Eqs. 1 and 2) were converted into minimization problems by multiplying the objective functions with −1. The following GA operating conditions were set: the population size 100, the maximum number of generations 600, selection function tournament of size 3, the crossover fraction 0.75 with scattered crossover function, adaptive feasible mutation function, forward direction for migration with migration fraction set to 0.15, and the crowding distance sorting. Two stopping criteria were set: the algorithm stops if the average relative change in the spread of the Pareto solutions over 50 generations is less than 10-5 or if the maxi-mum number of generations is exceeded.



215


As expected, a sharp decline in permeate flux occurred during the first 10 to 15 minutes of the microfiltration, mainly due to the formed concentration polarization layer. Afterwards, the permeate flux gradually decreased until pseudo-steady state values were obtained. The values of permeate flux at the pseudo-steady state were dictated by opera-ting conditions. Figure 2 illustrates the change in the permeate flux with time during the microfiltration of baker’s yeast suspension with and without static mixer. With the inser-tion of static mixer, the permeate flux was up to threefold higher than the permeate flux without static mixer. The amount of permeate flux improvement is dependent upon the operating conditions (TMP and Q) and suspension concentration.

Figure 2. Permeate flux during microfiltration of baker’s yeast

(TMP=1 bar, Q=130 L/h and C=2 g/L).

The main goal of applying genetic algorithm in the optimization of microfiltration of baker’s yeast suspension in the presence of static mixer, is to maximize the permeate flux and reduce the specific energy consumption. Given that an improvement (increase) in the permeate flux can lead to a decrease in the reduction of specific energy consumption, multiple non-dominated Pareto solutions were obtained for both optimization sets. The Pareto front gives an indication of the trade-off between the conflicting objectives Jsm and ER, which later allows a decision maker to decide upon the preferred solution. The optimal Pareto front and the input decision values corresponding to the Pareto optimal solutions for the optimization set A were determined and the results are presented in Figure 3. It can be seen that the results are in good agreement with the optimization re-sults obtained by Jokić et al. (13). For the lowest suspension concentration, C=2 g/L (see Figure 3a), the best results are achieved when TMP equals 1 bar and feed flow rate is kept between 122.8 and 127.3 L/h. In this case, Jsm and ER on the Pareto front are bounded in a narrow range of 190.7-190.8 L/(m2h) and 44.2-44.4%, respectively. This indicates that



216

there is hardly any difference between the non-dominated Pareto solutions. When the concentration of suspension is kept at 6 g/L (Figure 3b), the optimal results are achieved for the transmembrane pressure of 0.2 and 1 bar, with the feed flow rate in the range 108-135 L/h. Here, a discontinuity in the Pareto front was obtained with the values of Jsm about 110 L/(m2h) and about 168 L/(m2h), which were attained with TMP values of 0.2 and 1 bar, respectively. Evidently, the left section on the Pareto front is of lesser interest since, compared to the right section, it leads to a significant loss in Jsm, while at the same time only a small gain in ER is obtained. By analyzing the right section of the Pareto front shown in Figure 3b, it can be seen that a small loss in Jsm leads to a higher gain in ER, which means that a preferred solution should be for higher values of ER (about 47%). For the highest suspension concentration of 10 g/L (Figure 3c), the multi-objective opti-mization produced a wider range of non-dominated solutions from which a single ope-rating point could be selected by the decision maker (8). As can be seen, the values of Jsm and ER on the Pareto front are in the range of 78-143 L/(m2h) and -5-50%, respectively.

Figure 3. Optimal Pareto front (left) and decision space (right) for optimization set A. Suspension concentration: a) 2 g/L, b) 6 g/L, and c) 10 g/L.

Legend: circles - GA approach (this paper); stars - desirability function approach (results from Jokić et al. (13))

The results of the optimization set B are presented in Figure 4. Figure 4a shows the optimal Pareto front when both permeate flux and the reduction of specific energy are maximized, whereas Figures 4b-d show three decision variables (TMP, Q and C) cor-



217

responding to the obtained Pareto solutions. Here, a discontinuity in the Pareto front can also be seen, as was the case for the above-discussed optimization set A when C=6 g/L. Higher suspension concentrations (about 7.5 g/L) and transmembrane pressure of 0.2 bar generate the left section of the Pareto front for which Jsm and ER are in the range of 91-111 L/(m2h) and 50-53%, respectively. On the other hand, lower suspension concentra-tions (2-4.5 g/L) and TMP value of 1 bar are responsible for the right section of the Pareto front where Jsm is between 177-191 L/(m2h), while ER is in the range 44-50%. Additionally, the values of feed flow rate are approximately in the range 109-127 L/h for both sections of the Pareto front. However, by moving from the right to the left section of the Pareto front it can be observed that a small gain in ER, and at the same time a sub-stantial loss in Jsm, could be achieved. For instance, a 5% improvement in ER is obtained by simultaneously lowering Jsm by about 45%. In other words, probably the right section of the Pareto front should be considered operationally by the decision maker (8). As can be seen in Figure 4, these findings also confirmed the optimization results from Jokić et al. (13). Nevertheless, genetic algorithm-based multi-objective optimization is preferable to classical optimization methods, such as desirability function approach, since it generates multiple solutions in one optimization run. Genetic algorithm-based multi-objective optimization allows an instant search for promising solutions, and more importantly, allows subsequent analysis on the top solutions without concern that certain optimal solutions may never be found (8, 15).

Figure 4. Optimal Pareto front (a) and process decision variables corresponding to each of the optimal solutions (b-d) for optimization set B.



218

CONCLUSION In this paper, elitist non-dominated sorting genetic algorithm, implemented within powerful commercial software Matlab, was used for multi-objective optimization of the microfiltration of baker’s yeast in the presence of a Kenics static mixer. Three decision parameters including suspension concentration, feed flow rate, and transmembrane pres-sure were selected. The insertion of static mixer improved permeate flux by up to three-fold, depending upon the operating parameters. The obtained Pareto fronts gave an indi-cation of the trade-off between two competitive objectives, permeate flux and reduction of specific energy consumption, which allows a decision maker to decide upon the pre-ferred solution. It was observed that lower suspension concentrations (2-4.5 g/L), feed flow rate in the range of 109-127 L/h, and TMP value of 1 bar yielded optimal conditions for which permeate flux was between 177 and 191 L/(m2h), while the reduction of spe-cific energy consumption was in the range 44-50%. By comparing the results with those obtained by using desirability function approach, it can be confirmed that genetic algo-rithms are more effective than classical methods because they provide multiple solutions for multi-objective optimization problems in a single optimization run.

Acknowledgement This research was financially supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia (Project No. TR-31002).

REFERENCES

1. Coello Coello, C.A. A Comprehensive Survey of Evolutionary-Based Multiobjective Optimization Techniques. Knowl. Inf. Syst. 1999, 1 (3), 269-308.

2. Chankong, V.; Haimes, Y.Y. Multiobjective decision making: theory and metho-dology; North Holland: 1983.

3. Goicoechea, A.; Hansen, D.R.; Duckstein, L. Multiobjective decision analysis with engineering and business applications; Wiley: 1982.

4. Deb, K.; Pratap, A.; Agarwal, S.; Meyarivan, T. A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE T. Evolut. Comput. 2002, 6 (2), 182-197.

5. Konak, A.; Coit, D.W.; Smith, A.E. Multi-objective optimization using genetic algorithms: A tutorial. Reliab. Eng. Syst. Safe. 2006, 91 (9), 992-1007.

6. Zitzler, E.; Deb, K.; Thiele, L. Comparison of Multiobjective Evolutionary Algo-rithms: Empirical Results. Evol. Comput. 2000, 8 (2), 173-195.

7. Yuen, C.C.; Aatmeeyata; Gupta, S.K.; Ray, A.K. Multi-objective optimization of membrane separation modules using genetic algorithm. J. Membr. Sci. 2000, 176 (2), 177-196.

8. Guria, C.; Bhattacharya, P.K.; Gupta, S.K. Multi-objective optimization of reverse osmosis desalination units using different adaptations of the non-dominated sorting genetic algorithm (NSGA). Comput. Chem. Eng. 2005, 29 (9), 1977-1995.



219

9. Badrnezhad, R.; Mirza, B. Modeling and optimization of cross-flow ultrafiltration using hybrid neural network-genetic algorithm approach. J. Ind. Eng. Chem. 2014, 20 (2), 528-543.

10. Soleimani, R.; Shoushtari, N.A.; Mirza, B.; Salahi, A. Experimental investigation, modeling and optimization of membrane separation using artificial neural network and multi-objective optimization using genetic algorithm. Chem. Eng. Res. Des. 2013, 91 (5), 883-903.

11. Murthy, Z.V.P.; Vengal, J.C. Optimization of a Reverse Osmosis System Using Ge-netic Algorithm. Sep. Sci. Technol. 2006, 41 (4), 647-663.

12. Strugholtz, S.; Panglisch, S.; Gebhardt, J.; Gimbel, R. Neural networks and genetic algorithms in membrane technology modelling. J. Water Supply Res. T. 2008, 57 (1), 23-34.

13. Jokić, A.; Zavargo, Z.; Šereš, Z.; Tekić, M. The effect of turbulence promoter on cross-flow microfiltration of yeast suspensions: A response surface methodology approach. J. Membr. Sci. 2010, 350 (1), 269-278.

14. Yusoff, Y.; Ngadiman, M.S.; Zain, A.M. Overview of NSGA-II for Optimizing Machining Process Parameters. Procedia Engineer. 2011, 15, 3978-3983.

15. Bhatti, M.S.; Kapoor, D.; Kalia, R.K.; Reddy, A.S.; Thukral, A.K. RSM and ANN modeling for electrocoagulation of copper from simulated wastewater: Multi objecti-ve optimization using genetic algorithm approach. Desalination 2011, 274 (1), 74-80.

ВИШЕКРИТЕРИЈУМСКА ОПТИМИЗАЦИЈА МИКРОФИЛТРАЦИЈЕ ПЕКАРСКОГ КВАСЦА ПРИМЕНОМ ГЕНЕТСКИХ АЛГОРИТАМА

Наташа Љ. Лукић, Марија Божин-Дакић, Јована А. Граховац, Јелена М. Додић,

Александар И. Јокић


У овом раду представљен је модел вишекритеријумске оптимизације генетским алгоритмима у циљу одређивања оптималних процесних услова за извођење мик-рофилтрације пекарског квасца коришћењем статичког мешача као промотора тур-буленције. Испитивани процесни параметри су концентрација суспензије, проток суспензије и трансмембрански притисак. Истовремено, узете су у обзир две међу-собно противречне функције циља, максимални флукс пермеата и максималнo сма-њење специфичне потрошње енергије. Решење вишекритеријумског оптимизацио-ног проблема одређено је применом недоминантног сортирајућег генетског алго-ритма друге генерације у склопу Matlab R2015b софтвера. Приказани су Парето фронтови, као и вредности процесних параметара за које се остварују добијени оптимални услови. Утврђено ја да се при нижим концентрацијама суспензије пе-карског квасца (2-4.5 g/L), протоку супензије у опсегу 109-127 L/h, као и транс-мембранском притиску од 1 bar постижу оптимални услови за извођење микро-филтрације, при чему су максимални флукс пермеата и максималнo смањење спе-



220

цифичне потрошње енергије у опсегу 177-191 L/(m2h) односно 44-50%. Добијени резултати су упоређени са претходно објављеним резултатима добијеним приме-ном принципа жељене функције. Имајући у виду да се применом генетских алгори-тама добијају вишеструка решења у току једне оптимизације, ова студија је пока-зала да су генетски алгоритми знатно ефикаснији од класичних оптимизационих метода при решавању вишекритеријумских оптимизационих проблема. Кључне речи: вишекритеријумска оптимизација, генетски алгоритам, пекарски

квасац, статички мешач, микрофилтрација


APTEFF, 48, 1-323 (2017) UDC: 678.1:677.46:546.226 https://doi.org/10.2298/APT1748221M BIBLID: 1450-7188 (2017) 48, 221-229


221

INTERFACIAL PROPERTIES OF CHITOSAN/SODIUM DODECYL SULFATE

COMPLEXES

Jelena R. Milinković*, Milijana V. Aleksić, Lidija B. Petrović, Jadranka L. Fraj, Sandra Đ. Bučko, Jaroslav M. Katona, Ljiljana M. Spasojević


Contemporary formulations of cosmetic and pharmaceutical emulsions may be achieved by using combined polymer/surfactant system, which can form complexes with different structure and physicochemical properties. Such complexation can lead to additional stabilization of the emulsion products. For these reasons, the main goal of this study was to investigate the interfacial properties of chitosan/sodium dodecyl sulfate complexes. In order to understand the stabilization mechanism, the interface of the oil/water systems that contained mixtures of chitosan and sodium dodecyl sulfate, was studied by measuring the interfacial tension. Considering the fact that the properties of the oil phase has influence on the adsorption process, three different types of oil were investigated: medium-chain triglycerides (semi-synthetic oil), paraffin oil (mineral oil) and natural oil obtained from the grape seed. The surface tension measurements at the oil/water interface, for chitosan water solutions, indicate a poor surface activity of this biopolymer. Addition of sodium dodecyl sulfate to chitosan solution causes a significant decrease in the interfacial tension for all investigated oils. The results of this study are important for understanding the influence of polymer-surfactant interactions on the properties of the solution and stability of dispersed systems. KEY WORDS: interfacial properties, chitosan, sodium dodecyl sulfate

INTRODUCTION The interactions between polymers and surfactants in aqueous media give rise to the formation of association structures, thereby modifying functional and interfacial pro-perties of the solution. The morphologies of association complexes depend on the molecular properties of the polymer and the surfactant. Polymer/surfactant assembly has received a constant interest over several decades since polymers may modify the proper-ties of surfactant solutions. Especially, oppositely charged polymer/surfactant systems were widely studied, as they bind strongly to each other and offer drastic changes in the properties of surfactant in the bulk and at the interface (1-4). The existence of a pheno-menon of the interaction in the polymer-surfactant mixtures enabled the formation of * Corresponding author: Jelena R. Milinković, University of Novi Sad, Faculty of Technology Novi Sad, Bule-

var Cara Lazara 1, 21000 Novi Sad, Serbia, e-mail: [email protected]



222

complexes, which has been widely used in the pharmaceutical and cosmetic industries, as it may greatly influence the characteristics of the product. In that manner the adsorption and orientation of polymer/surfactant complexes at various interfaces and their effect on the stability of emulsion is a subject of great importance in the area of formulation of pharmaceuticals and cosmetics (5, 6). Chitin, the second most abundant natural biopolymer (next to cellulose), can be widely found in exoskeleton of shellfish, shrimps and lobsters. Chitosan (Ch), a cationic polysaccharide, is obtained by alkaline deacetylation of chitin. Solubility, rheological properties and surface adsorption of aqueous Ch solutions depend on the average molecu-lar weight, degree of deacetylation, distribution of acetyl and amino groups along the po-lymer chain, as well as on the pH and ionic strength of the solution (7, 8). At the pH <6.5, chitosan is soluble in water and has a positive charge due to the presence of protonated amino groups. Due to the possessing the combination of properties such as biocompatibi-lity, biodegradability, nontoxicity, bioactivity and antifungal activity, chitosan has many applications in medicine, cosmetics and controlled drug delivery systems (9-13). As mentioned above, polymer may interact with some other polymer or surfactant, and it is possible to expect various phenomena caused by their physicochemical pro-perties. Since chitosan is positively charged at low pH values, it spontaneously associates with oppositely charged polymers in the solution and forms polyelectrolyte complexes (10). Also, anionic surfactants may associate with chitosan, forming chitosan/surfactant complexes. In the last decade, the interactions between chitosan and sodium dodecyl sulfate have become an increasingly interesting research area (14-18). These specific interactions and the systems elaborated on their basis are important for many practical applications, such as colloidal stabilization, wettability, and adhesion. The emulsion stabilized by the polymer/surfactant complex is generally more stable than the one stabilized by surfactant only. This is because it is stabilized not only by the electrostatic repulsion of the droplets, but also by the steric hindrance of the bulky oppo-sitely charged polymer/surfactant complexes (19). In addition, for environmental or safety reasons, the use of a large quantity of low molecular mass surfactants is not recom-mended. Furthermore, the bulk physicochemical characteristics are also important for the formation and stability of the emulsion. Also, chemical structure of the oil, such as fatty acids chains length, degree of unsaturation and molecule configuration, influence the interfacial properties as well as emulsion stability and its use (20). For these reasons, our studies have focused on the investigation of the interfacial behavior of selected surfactant, sodium dodecyl sulfate (SDS), and its mixture with chi-tosan at the oil/water interface. In this study, three different types of oils were used: me-dium-chain triglycerides (semi-synthetic oil), paraffin oil (mineral oil), and grape seed oil (natural origin).



223

EXPERIMENTAL

Materials Low molecular weight Ch (product number: 448869) was obtained from Sigma-Aldrich (Shanghai, China). The chitosan's degree of deacetylation, determined by poten-tiometric titration according to the procedure described by Yuan at al. (21), is found to be 81.8%. SDS, purity >99%, was purchased from Merck (Darmstadt, Germany). Different oils were used as oil phase: medium-chain triglycerides or caprylic/capric triglycerides (Saboderm TCC, Comcen, Zemun, Serbia), paraffin oil (Centrohem, Stara Pazova, Serbia), and grape seed oil (Olitalia, Forlì (FC), Italy). In all experiments, buffered water was used as a solvent and the pH was adjusted using 0.2 M water solution of acetic acid (Zorka-Pharma, Šabac, Serbia) and 0.2 M water solution of sodium acetate (Centrohem, Stara Pazova, Serbia).

Methods

Preparation of solutions

The experiments were carried out at pH 4. The pH measurements were performed by using an 827 lab pH-meter (Metrohm, Herisau, Switzerland). Stock solution of 1% (w/w) Ch was prepared by dissolving a given mass of the polymer in the buffered water, while stirring and after relaxation at room temperature during 24 h, the pH value of the solution was checked. The SDS stock solution of 2% (w/w) was prepared by using the same procedure, while the Ch/SDS mixtures were prepared by mixing required volumes of Ch and SDS stock solutions. Concentrations of SDS were varied from 0.00001% (w/w) up to 1% (w/w), while the Ch concentration was kept constant at 0.01% (w/w). The mixtures were left for 24 h at room temperature. Before measurements, pH value of the solutions was checked.

Interfacial tension

Measurements of the interfacial tension between the oil and water phase containing chitosan and various surfactant concentrations, were carried out on a Sigma 703D tensiometer (KSV Instruments, Helsinki, Finland) using the Du Noüy ring method (21). Prior to the measurement, the ring was immersed in deionized water (below the surface), then the oil phase was slowly added a top, and the system was left for 15 min, allow equlibriation of the interface. The interfacial tension of each oil/water system was measu-red by pulling the ring out of heavier, aqueous phase into lighter, oil phase. In all mea-surements, the temperature was kept constant at 30 °C. The reported values of the interfa-cial tension were average of three measurements, at least.



224


Interfacial tension measurements The formation of interfacial tension at the interface of two liquids which do not interfere with each other is due to different effects of their molecules at the interface. The stability of the system can be influenced by a variety of factors, especially by the pre-sence of various soluble solids, which, depending on the type and degree of solubility in the phases, can result in the changes of interfacial tension, and thereby influence the characteristics of the given system. In order to investigate the adsorption properties of chitosan, surface tension was measured at the oil/water interface. Besides the physicochemical properties of the adsor-bent, the adsorption process is influenced by the properties of the oil phase. For that reason, different types of oils were selected: medium-chain triglycerides, grape seed oil, and paraffin oil. Chitosan solution (0.01%) in buffered water (pH 4) was used as the aqueous phase. The obtained results are shown in Table 1.

Table 1. Water/oil interfacial tension at 30 °C and pH 4.

Surface tension (mN/m)

Oil Medium-chain triglycerides

Paraffin oil Grape seed oil

Buffered water 19.56 43.20 26.28 Buffered water with 0.01% Ch 15.05 43.88 26.01

The results suggest that the highest value of the interfacial tension was obtained for the paraffin oil/water system, and the lowest one for the medium-chain triglyceri-des/water system, which is in accordance with the different polarity of the oils. The addition of chitosan to buffered water decreased slightly the value of the interfacial ten-sion in the medium-chain triglycerides/water system, while practically did not have any influence on the other two systems. These results indicate a weak surface activity of the biopolymer, and are in agreement with previous studies (14). The introduction of surfactant molecules to the oil/water system leads to their adsorp-tion at the interface, which results in a reduction of the interfacial tension. The extent of changes in the interfacial tension depends on the characteristics of surfactant molecules, as well as of the oil and water phases. The results of the interfacial tension measurements in different oil/water system containing SDS are presented in Figure 1.



225

Figure 1. Interfacial tension in the oil/water system of pure SDS as a function of SDS

concentration, at 30 °C and pH 4 As expected, the increase in the SDS concentration resulted in a rapid decrease, to reach a plateau at a certain SDS concentration. It seems that the plateau value shows no dependence on the oil nature, since it was reached at 0.1% (w/w) SDS for all investigated oils. It can be thought that at this concentration the surface is completely saturated with SDS molecules and further increase of SDS concentration leads to the formation of micelles in the aqueous phase that do not have adsorption ability at the interface. The addition of the polyelectrolyte to the solution of the opposite charged surfactant most often leads to the changes in the system charge that are caused by their electrostatic and hydrophobic interactions. In that manner, the interfacial tension of the Ch/SDS mixtures in water and different oils were measured. The results are presented in Figure 2.

Figure 2. Interfacial tension in the oil/water system of Ch/SDS mixtures as a function of

SDS concentration at 30 °C and pH 4. The chitosan concentration was 0.01%.



226

As observed from Figure 2, the introduction of Ch into the system influenced signi-ficantly the oil/water interfacial tension. Namely, the changes in the interfacial tension of the Ch/SDS mixtures showed characteristic brake points at a certain SDS concentrations. At the beginning, the interfacial tension decreases gradually until reaching the first break point T1 at 0.001% SDS (Ch/SDS mass ratio 10:1) for the system medium-chain triglycerides/water and T’

1 at 0.005% SDS (Ch/SDS mass ratio 5:1) for paraffin oil/water and grape seed oil/water systems. The value of the interfacial tension of the medium-chain triglycerides/water system at T1 was 8.62 mN/m. This value for the paraffin oil/water system at T’

1 was 16.99 mN/m, and for the grape seed oil/water system at T’1 it

was 10.46 mN/m. After reaching T1, the interfacial tension approximate a plateau till second break point T2 at 0.02% (w/w) SDS (Ch/SDS mass ratio 1:2) for all investigated systems. After T2, the interfacial tension decreased significantly until reaching the third break point T3 at 0.1% (w/w) of SDS concentration (Ch/SDS mass ratio 1:10), which is the critical micelle concentration for the pure SDS solution. Since the changes in all investigated systems indicate the presence of Ch/SDS interactions, they will be explained on the medium-chain triglycerides/water system (see Figure 3).

Figure 3. Interfacial tension in the medium-chain triglycerides/water system of pure SDS

and Ch/SDS mixtures as a function of SDS concentration at 30 °C and pH 4 As can be seen from Figure 3, at low SDS concentrations surface, active Ch/SDS complexes were formed and the interfacial tension of the mixture had a lower value than the solution of pure SDS of the corresponding concentration. It can be assumed that in such conditions electrostatic binding of single SDS molecules to the amino groups of chi-tosan chain took place resulting in formation of the complexes that have synergistic effect on the interfacial tension. Upon reaching an SDS concentration of 0.001% (w/w), at T1, turbidity in the mixtures occurs (Fig. 4), which is in accordance with our previously published results (14).



227

Figure 4. Phase separation in the mixtures of 0.01% chitosan and

various SDS concentrations

The interaction continues mostly in the bulk, until reaching T2 at 0.02% SDS, by binding SDS in the form of aggregates to the Ch chains (1, 3, 14). Such phenomenon should not influence the oil/water interface. However, in the T1-T2 region, the surface tension slightly increases, indicating reorganization of the interfacial layer due to the hydrophobic SDS binding and stretching of the polymer chains, i.e. the interaction taking place simultaneously at the interface. At 0.02% SDS (Ch:SDS mass ratio 1:2) the Ch/SDS complexes fully precipitate as the coacervate phase (Figure 4). Further addition of SDS leads to desorption of Ch/SDS complexes and gradual allocation of SDS mole-cules at the interface, causing significant decrease in the interfacial tension until reaching the third break point T3 at 0.1% (w/w) SDS, which is critical micelle concentration for the pure SDS solution. After the concentration of 0.1% (w/w), the interface is saturated only with SDS molecules, and the interfacial tension remains unchanged. Once micelles of SDS abound in solution, the interfacial tension becomes essentially equal to that of a micelle, polymer-free surfactant solution.

CONCLUSION

The investigation of the interfacial behavior of Ch/SDS complexes showed that increasing concentration of SDS in a chitosan solution leads to characteristic changes in the interfacial tension. These changes indicate an electrostatic interaction between the po-sitively charged chitosan and the opposite charged SDS, i.e. formation of a chitosan/SDS complexes, which exhibit the surface activity. Surface active complexes begin to form at very low concentrations of SDS. As already demonstrated with various systems, the sur-face activity of these complexes is much higher than the one of the polymer or surfactant alone. At higher SDS concentrations, the interaction is more hydrophobic and SDS molecules bind to the polymer chains in the form of aggregates, leading gradually to pre-cipitation of the complexes as the coacervate phase. The lowest interfacial tension was achieved in the system medium-chain triglycerides/water. The results of this study show-ed that Ch/SDS complexes could be used for stabilization O/W emulsions with medium-chain triglycerides, paraffin and grape seed oils. In addition, this phenomenon suggests the possible application of Ch/SDS mixtures in the microencapsulation processes.

Acknowledgement

This work was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia, Project Number III46010.



228

REFERENCES 1. Goddard, D. E., Ananthapadmanabhan, K. Interaction of Surfactants with Polymers

and Proteins. Application of polymer-surfactant systems; Boca Raton, FL: CRC Press. 1993, 395-413.

2. Langevin, D. Polyelectrolyte and surfactant mixed solutions. Behavior at surfaces and in thin films. Adv. Colloid Interface Sci. 2001, 89, 467-484.

3. Taylor, D. J. F.; Thomas, R. K.; Penfold, J. Polymer/surfactant interactions at the air/water interface, Adv. Colloid Interface Sci. 2007, 132, 69-110.

4. Hansson, P. Self-Assembly of Ionic Surfactants in Polyelectrolyte Solutions: A Model for Mixtures of Opposite Charge. Langmuir, 2001, 17, 4167-4180.

5. Kragel, J.; O'Neill, M.; Makievski, A. V.; Michel, M.; Leser, M. E.; Miller, R. Dynamics of mixed protein-surfactant layers adsorbed at the water/air and water/oil interface. Colloids Surf. B 2003, 31, 107-114.

6. Zana R., Xia J. Gemini Surfactants Synthesis, Interfacial and Solution-Phase Behavior, and Applications; Surfactant Science Series 117, Marcel Dekker, New York, 2004, pp 65-80.

7. Helgason T.; Gislason J.; McClements D. J.; Kristbergsson K.; Weiss J. Influence of molecular character of chitosan on the adsorption of chitosan to oil droplet interfaces in an in vitro digestion model. Food Hydrocolloids. 2009, 23, 2243-2253.

8. Qun G.; Ajun W. Effects of molecular weight, degree of acetylation and ionic strength on surface tension of chitosan in dilute solution. Carbohydr. Polym. 2006, 64, 29-36.

9. Amidi, M.; Mastrobattista, E.; Jiskoot, W.; Hennink, W. E. Chitosan-based delivery systems for protein therapeutics and antigens. Adv. Drug Deliv. Rev. 2010, 62, 59-82.

10. Hamman, H. J. Chitosan based polyelectrolyte complexes as potential carrier mate-rials in drug delivery systems. Mar. Drugs 2010, 8, 1305-1322.

11. Muzzarelli, R. A. A. Chitins and chitosans for the repair of wounded skin, nerve, cartilage and bone. Carbohydr. Polym. 2010, 76, 167-182.

12. Shi, X.Y.; Tan, T.W. Preparation of chitosan/ethylcellulose complex microcapsule and its application in controlled release of vitamin D2, Biomaterials 2002, 23, 4469-4473.

13. Agnihotri S. A.; Mallikarjuna N. N.; Aminabhavi T. M. Recent advances on chitosan-based micro- and nanoparticles in drug delivery. J. Controlled Release, 2004, 100, 5 - 28.

14. Petrović L.; Milinković J.; Fraj J.; Bučko S.; Katona J. An investigation of chitosan and sodium dodecyl sulfate interactions in acetic media. J. Serb. Chem. Soc. 2016, 81(5), 575-587.

15. Onesippe, C.; Lagerge, S. Study of the complex formation between sodium dodecyl sulfate and chitosan. Colloids Surf. A. 2008, 317, 100-108.

16. Onesippe, C.; Lagerge S. Studies of the association of chitosan and alkylated chitosan with oppositely charged sodium dodecyl sulfate. Colloids Surf. A. 2008, 330, 201-206.



229

17. Thongngam, M.; Mcclements, D. J. Characterization of Interactions between Chitosan and an Anionic Surfactant. J. Agric. Food Chem. 2004, 52, 987-991.

18. Lundin, M.; Macakova, L. Interactions between Chitosan and SDS at a Low-Charged Silica Substrate Compared to Interactions in the Bulk-The Effect of Ionic Strength. Langmuir. 2008, 24, 3814-3827.

19. Vongsetskul, T.; Phaenthong, K.; Chanthateyanonth, R.; Sunintaboon, P.; Tangbori-boonrat, P. Emulsions Stabilized by Chitosan/lithium Dodecyl Sulfate Complexes. Chiang Mai J. Sci. 2015, 42(2), 393-400.

20. Ushkubo F. Y.; Cunha R. L. Stability mechanisms of liquid water-in-oil emulsions. Food Hydrocolloids. 2014, 34, 145-153.

21. Yuan, Y.; Chesnutt, B. M.; Haggard, W. O. Bumgardner, J. D. Deacetylation of Chi-tosan: Material Characterization and in vitro Evaluation via Albumin Adsorption and Pre-Osteoblastic Cell Cultures. Materials. 2011, 4, 1399-1416.

MЕЂУФАЗНО ПОНАШАЊЕ КОМПЛЕКСА

ХИТОЗАН/НАТРИЈУМ-ДОДЕЦИЛ-СУЛФАТ Јелена Р. Милинковић, Милијана В. Алексић, Лидија Б. Петровић, Јадранка Л. Фрај,

Сандра Ђ. Бучко, Јарослав М. Катона и Љиљана M. Спасојевић

Универзитет у Новом Саду, Технолошки факултет Нови Сад, Булевар цара Лазара 1, 21000 Нови Сад

Код савремених формулација козметичких и фармацеутских емулзија, стабили-зација се може постићи применом комбинованих полимер/површински активна ма-терија система, који у смеши могу да формирају комплексе различите структуре и физичко-хемијских особина, односно могу довести до додатне стабилизације емул-зионих производа. Из ових разлога, главни циљ ове студије био је испитивање осо-бина комплекса хитозан/натријум-додецил-сулфат. У циљу што бољег разумевања механизма стабилизације и међуфазних особина смеше хитозана и натријум-доде-цил-сулфата, мерен је међуповршински напон. Узимајући у обзир чињеницу да својства уљне фазе утичу на процес адсорпције, испитивана су три различита типа уља: триглицериди средње дужине ланаца (полусинтетско уље), парафинско (мине-рално) уље и уље коштица грожђа (природног порекла). Мерења површинског на-пона на граници фаза уље/вода за раствор хитозана указују на слабу површинску активност овог биополимера. Додатак натријум-додецил-сулфата у раствор са хи-тозаном доводи до значајног смањења међуповршинског напона за сва испитивана уља. Резултати ове студије су важни за разумевање утицаја интеракције поли-мер/површински активна материја на особине раствора и стабилност дисперзног система. Кључне речи: међуфазне особине, хитозан, натријум-додецил-сулфат

Received: 26 July 2017. Accepted: 06 October 2017.

APTEFF, 48, 1-323 (2017) UDC: 615.282:579.84:66.023.2 https://doi.org/10.2298/APT1748231M BIBLID: 1450-7188 (2017) 48, 231-244


231

EFFECT OF AGITATION RATE ON THE PRODUCTION OF ANTIFUNGAL METABOLITES BY Streptomyces hygroscopicus IN A LAB-SCALE

BIOREACTOR

Ivana Ž. Mitrović1*, Jovana A. Grahovac1, Jelena M. Dodić1, Mila S. Grahovac2, Siniša N. Dodić1, Damjan G. Vučurović1, Vanja R. Vlajkov1


2 University of Novi Sad, Faculty of Agriculture, Trg Dositeja Obradovića 8, Novi Sad 21000, Serbia

The application of antifungal compounds produced by microorganisms in the control of plant diseases caused by phytopathogenic fungi is a promising alternative to synthetic pesticides. Among phytopathogenic fungi, Alternaria alternata and Fusarium avenaceum are significant pathogens responsible for the storage rot of apple fruits. During storage, transport and marketing A. alternata and F. avenaceum can cause significant losses of apple fruits and their control is of great importance for the producers and consumers. In the present study, the effects of agitation rate on the production of antifungal metha-bolite(s) by Streptomyces hygroscopicus in a 3-L lab-scale bioreactor (Biostat® Aplus, Sartorius AG, Germany) against two isolates of A. alternata and two isolates of F. ave-naceum were investigated. The cultivation of S. hygroscopicus was carried out at 27 °C with agitation rates of 100 rpm and 200 rpm during 7 days. The aim was to analyze the bioprocess parameters of biofungicide production in a medium containing glycerol as a carbon source, and examine the effect of agitation rate on the production of antifungal metabolite(s). The in vitro antifungal activity of the produced metabolites against fungi from the genera Alternaria and Fusarium grown on potato dextrose agar medium was determined every 24 h using wells technique. In the experiments conducted in the bioreactor at different stirring speeds, it was found that the maximum production of antifungal metabolites occurred after 96 hours of cultivation. A higher consumption of nutrients and a larger inhibition zone diameter was registered in the experiment with an agitation rate of 200 rpm. KEY WORDS. Streptomyces hygrioscopicus, biocontrol, agitation rate, Alternaria alter-

nata, Fusarium avenaceum

INTRODUCTION During marketing of apple fruits or production of apple juice, fruit may be exposed out of storage for a number of days, under conditions favorable for fungal disease de-velopment. The spoilage of apples by fungi is often accompanied by the production of

* Corresponding author: Ivana Ž. Mitrović, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




232

mycotoxins which are a potential threat to human health (1). Beside Penicillium expansum, which is one of the most important storage pathogens of apple fruit worldwide and causes blue mold, a decay that can lead to significant economic losses (2), Alternaria spp. and Fusarium spp. are also listed among the significant storage pathogens of apple fruit (3, 4). Several Alternaria species are associated with moldy core of apples while dry core rot is mostly linked to a single species, A. alternata (5). Spores infect the open calyx of young fruits, and mycelia reach the seed and carpel wall during storage. Moldy-core is characterized by the growth of the mycelia within the locules, with or without penetration into the mesoderm. The disease may become invasive and lead to a slow, dry rot confined to the flesh immediately surrounding the core. External symptoms are rare, although in-fected fruits may color and drop prematurely (6). Approximately, 30 metabolites with possible toxicity are known from various species of Alternaria. Alternariol (AOH), alte-nariol monomethyl ether (AME), altenuene (ALT), tenuazonic acid (TeA), tentoxin (TEN), and altertoxins I, II, and III are considered to be some of the important myco-toxins (7). Fusarium species are primarily known as the causal agents of plant diseases, but fungi of this genus also produce many secondary toxic metabolites that can cause acute or chronic diseases in humans and domestic animals. The most important toxins produced by F. avenaceum are moniliformin, acuminatopyrone and chrysogine, which are most commonly found in the apple products. In addition to these mycotoxins, Sorensen et al. (8) also mentioned enniatine, chlamydosporol, fusarin, and others. Fusarium rot occurrs on apples and other fruits while they are stored and shelved. It causes brown, soft, and watery circular necrosis, that gradually spreads over infected tissue, which becomes slightly sunken, sometimes with dense whitish mycelium on the surface (9). Chemical treatment of ripened fruit has many serious side-effects, especially leaving residues and sometimes causing fruit injury, in addition to the presence of offensive odors under modified storage conditions (10). The wide usage of chemical compounds to prevent plant diseases is often expensive, has adverse effects on humans, causes environmental pollution, and can also be lethal to the beneficial organisms (11). Due to adverse toxicological properties and resistance occurrence, use of chemical fungicides is being reduced, and their application after harvest is prohibited in many countries. The increasing concern for environmental protection and demand for organic farming drives research towards alternative control measures, such as the use of natural antagonists to biologically control plant pathogens (12). Various kinds of microbial antagonists have been investigated as potential antifungal biocontrol agents for plant disease management. Actinomycetes are one of the important groups of soil microorganisms. They are important producers of bioactive compounds and constitute a potential group of biocontrol agents. It is well known that actinomycetes produce 70% to 80% of bioactive secondary metabolites, where approximately 60% of antibiotics developed for agricultural use are isolated from Streptomyces spp. Growth and secondary metabolite production depends on many factors, including soil type, aeration, salinity, relative moisture content, and temperature (13). Most of active metabolites derived from actinomycetes are produced by using a submerged culture. For the production of antifungal components by microorganisms, it is



233

necessary to provide an appropriate amount of dissolved oxygen. To achieve this, the appropriate aeration and agitation rate must be defined. When mixing rate increases and the aeration rate is constant, an increase in the amount of dissolved oxygen in the cultivating liquid is expected, but the increase in the shear force may have a negative impact on the microorganism, and further affect its productivity. The morphology of streptomycetes may have direct or indirect effects on production of target products. For example, the formation of compact pellets of S. lividans was found to be of great impor-tance for hybrid antibiotic production, while tylosin production by S. fradiae was impro-ved using a smaller pellet or clump size (14). The study of airflow rates is important in submerged bioreactors because the microorganism grows completely immersed in the culture medium without direct contact with the gas phase, which induces oxygen mass transfer limitation both inside and outside the pellets (15). In general, high agitation speeds can damage microorganisms and, as a result, the cell concentration and producti-vity do not increase (16). On the other hand, low agitation rate can lead to over-aggre-gation of cells and the formation of large pallets, which also do not represent a good form of the product. Therefore, it is very important to determine the effect of increase in agitation rate on the production of antifungal metabolites under specific conditions. In the present study, the effects of agitation rate on the production of antifungal metabolite(s) by S. hygroscopicus against two isolates of A. alternata and two isolates of F. avenaceum in a 3-L lab scale bioreactor (Biostat® Aplus, Sartorius AG, Germany) were investigated. The aim was to analyze the bioprocess parameters of biofungicide production in a medium containing glycerol as a carbon source and examine how an increase in agitation rate influences the course of cultivation under the given conditions.

EXPERIMENTAL

Fungal pathogen Isolates of A. alternata and F. avenaceum were obtained from apple fruit samples showing rot symptoms, collected during 2012, after four months storage in Ultra Low Oxygen storages in Vojvodina Province, Serbia. The pathogens were identified according to pathogenic, morphological and ecological characteristics. Two A. alternata isolates (KA10 and T1Jg3) and two F. avenaceum isolates (KA12 and KA13) were selected as representatives of the collection and used in the study. The isolates were initially grown on Potato Dextrose Agar (PDA) plates for seven days. After seven days, a small amount a mycelia plug 3 mm in diameter was taken from the margin of the colony of each isolate, and added to the flasks containing 50 ml of potato dextrose broth. The flasks were incubated for 48 hours on a rotary shaker (150 rpm) at 25°C. Before use, the culture liquid was filtered through the double layer of sterile cheesecloth.

Antifungal components production Microorganism producer, S. hygroscopicus, was isolated from the soil samples collec-ted from various locations on the territory of Novi Sad, Serbia, and stored in the Micro-



234

bial Culture Collection of the Faculty of Technology in Novi Sad, Serbia. According to pathogenic, morphological and ecological characteristics, and the molecular identifica-tion, the isolate was identified as S. hygroscopicus (17). In the previous studies, the selec-ted isolate of S. hygroscopicus exhibited strong antifungal activity against different phytopathogenic fungi, causal agents of post-harvest rots (18, 19). The medium used for the growth of microorganism producer had the following composition (g/l): glucose (15.0), soybean flour (10.0), CaCO3, (3.0), NaCl, (3.0), MgSO4, (0.5), (NH4)2HPO4, (0.5), K2HPO4, (1.0). Composition of the comparative fermentation medium was as follows (g/l): glycerol (15.0), CaCO3 (3.0), NaCl (3.0), MgSO4 (0.5), (NH4)2HPO4 (0.17), K2HPO4 (0.33). The pH of the medium was adjusted to 7 ± 0.1 (Consort C863, Turnhout, Belgium) prior to autoclaving. Production of antifungal metabolites by S. hygroscopicus was carried out in a 3-L bench-scale bioreactor (Biostat Aplus) with a 2-L working volume during 7 days. The fermentation was carried out at 27 °C at the agitation rates of 100 rpm and 200 rpm. Two liters of fermentation medium was inoculated with 10% (v/v) of a preculture after the 72-h growth on a rotary shaker (IKA KS 4000i Control Incubating Shaker) at 150 rpm. The cultivation liquid was tested every 24 h. The sample of the cultivation medium was centrifuged at 10,000 g (Rotina 380 R, Hettich, Germany) for 10 min and the supernatant of cultivation medium was used for further analysis.

Analytical methods The obtained supernatants were filtered through a 0.45-μm nylon membrane (Agilent Technologies, Germany), and then analyzed by HPLC (Thermo Scientific Dionex UltiMate 3000 series) to determine residual glycerol content. The HPLC instrument was equipped with an HPG-3200SD/RS pump, autosampler WPS-3000(T) SL (10-ll injection loop), column ZORBAX NH2 (250 mm 9 4.6 mm, 5 lm), and detector Refracto- Max520. Acetonitrile (75:25, v/v) was used as eluent at a flow rate of 1.2 ml/min, and the elution time was 20 min at the column temperature of 30 °C. The residual nitrogen in the cultivation medium was determined by Total Kjeldahl Nitrogen (TKN, (EPA 60014-79-020)) method (20). A spectrophotometric method was used to determine phosphorus content (21). The biomass from the culture filtrate, separated by centrifugation, was transferred to a preweighed dry filter paper using a clean spatula and then placed in an oven at 55 °C overnight to reach a fixed weight. The growth in terms of biomass accumulation was ex-pressed as g/l culture medium. The biomass from the cultivation medium was measured every 24 h (22).

In vitro antagonistic activity assay The in vitro antagonistic activity assay was performed in 85-mm Petri plates using wells technique (23). Two layers of PDA medium were spread in the plates. The first layer consisted of 2% PDA medium. After solidification, the second layer (5 ml), composed of 1.2% PDA and filtered fungal culture liquid (35%), was added. Three wells



235

per plate with a diameter of 15 mm were made and two plates represented one treatment. The antifungal activity of cell-free culture filtrate was also tested. For each treatment, 100 μl of cell-free culture filtrate was added in each well. Sterile distilled water was used as negative control treatment. Antifungal activity of the produced metabolites was tested every 24 h. The assessment of antagonistic activity was carried out by measuring diameter of inhibition zones (mm) - zones around wells with no visible mycelia growth.

Data analysis The data on mycelia growth inhibition of four fungal isolates by antifungal metabo-lites of S. hygroscopuicus during 7 days of cultivation were processed by factorial ANOVA using Software STATISTICA 12 (Statistica 2012).


Effect of agitation rate on substrate consumption and cell growth during the cultivation

It is well known that the nutritional sources like carbon, nitrogen and minerals, as well and the environmental factors, have profound effect on antimicotic production by actinomycetes (24). The optimal agitation rate is very important for a maximum produc-tion of any culture in submerge cultivations and need to be carefully monitored, as it has both beneficial and deleterious effects (rupture of the cells and thus slower growth, chan-ge in cell morphology, foam production at high agitation, etc.) (25). Techapun et al. (26) in their research with Streptomyces sp. Ab 106 observed that the optimal agitation rate was between 150 and 200 rpm. However, the most favorable form of the metabolite pro-duction by streptomicetes is small pellets (27). Low stirring speed can lead to aggregation of cells and the formation of very large pallets, which are not a good form for production of targets agents. On the other hand, higher agitation speed increased the amount of dissolved oxygen and dispersion of macromolecules in the medium, and contributed to the greater growth and better production of antifungal agents, observed in this study. Figure 1 demonstrates the substrate consumption and growth of S. hygroscopicus cells in submerged cultures in the 3-L bioreactor at the agitation rates of 100 rpm and 200 rpm. The time-course of cell growth and substrate consumption during 7 days of cultivation of S. hygroscopicus in the laboratory-scale bioreactor showed that the consumption of carbon and nitrogen was in correlation with the increase in the biomass production of cells. Based on the observed consumption of nutrients during 7 days of cultivation, it can be concluded that there is no lag phase because the cells begin to consume the substrate from the start of the bioprocess. Also, the biomass of cells as an indirect indicator of the cultivation course showed that the exponential phase lasted until the third day, when the stationary phase of the bioprocess, in which the consumption of nutrients is significantly reduced, begins (Figure 1c). The results of the fermentation studies at different agitation rates suggest that the agitation rate of 200 rpm led to a higher consumption of carbon, nitrogen and phospho-



236

rous sources, and consequently caused enhanced production of biomass. At the lower agitation speed (100 rpm) the cell biomass as well as substrate consumption were found to be lower. The low biomass and low activity at lower agitation speed could be attribu-ted to the dearth of oxygen being experienced by the organism due to the insufficient mixing (25).

Figure 1. Effect of agitation rates on the residual glycerol content (a), residual nitrogen content (b), biomass (c), and residual phosphates (d),

during 7 days of S. hygroscopicus cultivation in a 3-L bioreactor at 100 rpm (■), and 200 rpm (▲).

At the end of the process, the residual glycerol was lower (4.06 g/l) in the bioprocess with the agitation speed of 200 rpm, than in the bioprocess with 100 rpm (6.87 g/l) (Figure 1a). Similar results were obtained for the nitrogen and phosphate sources (Figures 1b and 1d). At the end of the cultivation process at 200 rpm, the residual amounts of nitrogen and phosphorus were 0.0330 g/l and 0.1832 g/l, while in the cultivation with stirring of 100 rpm they were 0.0566 g/l and 0.3850 g/l, respectively.

0

2

4

6

8

10

12

14

16

0 24 48 72 96 120 144 168 192

Res

idua

l gly

cero

l (g/

l)

Time (h)

a)

0,00

0,01

0,02

0,03

0,04

0,05

0,06

0,07

0,08

0,09

0 24 48 72 96 120 144 168 192

b)

Time (h)

Res

idu

alni

trog

en [

g/l]

0,00

0,50

1,00

1,50

2,00

2,50

0 24 48 72 96 120 144 168 192

Time (h)

c)

Cel

l bio

mas

s (g

/l)

0,00

0,10

0,20

0,30

0,40

0,50

0,60

0,70

0,80

0,90

0 24 48 72 96 120 144 168 192

d)

Time (h)

Res

idu

al p

hos

ph

orus



237

Effect of agitation rate on the production of antifungal metabolite(s) during the cultivation

The statistical analysis was performed to determine whether the duration of cultiva-tion at different agitation rates had an effect on the production of antifungal metabolites effective against four tested isolates.

Table 1. Results of the factorial analysis of variance: sources of variation of the inhibition zone diameter during 7 days of cultivation at 100 rpm

Source of variation

SS Degr. of -Freedom

MS F-value p-value

Cultivation Time

1043.67 7 149.1 318.07 0.00

Test fungi 14.08 3 4.69 10.01 0.00 Time*Fungi 83.58 21 3.98 8.49 0.00 Error 30.00 64 0.47

SS - sum of squares; MS - mean square

As expected, the cultivation time had statistically significant effect (p<0.05) on the inhibition zone diameter at both applied agitation rates (Tables 1 and 2). Also, significant differences between inhibition zones diameter were observed between the fungal isolates. Interaction between these two factors (cultivation time and test fungi) also significantly affected the inhibition zone diameter. However, regardless of the applied mixing speed, the most significant source of variation of inhibition zones diameter was the cultivation time. Since this factor influences significantly the production of antifungal components it is important to determine whether it is possible to reduce duration of the bioprocesses while maintaining or increasing the productivity of target components. Also, it is important to determine whether the optimal cultivation time depends on agitation rate, because this is of great importance for the techno-economic assessment of the justification of the increase in the agitation rate. Table 2. Results of factorial analysis of variance: sources of variation of inhibition zone

diameter during 7 days of cultivation at 200 rpm

Source of variation

SS Degr. of - Freedom

MS F-value p-value

Cultivation Time

1480.66 7 211.52 520.7 0.00

Test fungi 38.03 3 12.68 31.2 0.00 Time*Fungi 44.05 21 2.10 5.2 0.00 Error 26.00 64 0.41

SS - sum of squares; MS - mean square



238

Figure 2 presents the mean values of mycelia growth inhibition zone diameter of test phytopathogenic fungi caused by supernatant of S. hygroscopicus cultivation filtrate at 100 rpm stirring speed during 7 days of cultivation. The obtained results showed that the antifungal activity was the strongest after 3 days of cultivation of S. hygroscopicus under defined conditions, with inhibition zones radius over 11 mm, indicating that the applied antifungal agent is potentially highly efficient (28). Also, the fungi from the genera Alter-naria did not show sensitivity during the first 48 hours of cultivation, while the isolates of the Fusarium genus became sensitive to the produced metabolites after 24 hours of cultivation. By inspecting Figure 2 it can be concluded that after 96 hours cultivation A. alternata KA10 was the most sensitive to the produced antifungal metabolites (average inhibition zone 25.33 mm), following by F. avenaceum KA12 (25 mm), A. alternata T1Jg3 (24.33 mm), and F. avenaceum KA13 (23.66 mm). Since the tested isolates belong to different genera their different sensitivity to antifungal components could be expected.

Time*Fungi; LS Means

Current effect: F(21, 64)=8.4910, p=.00000

Effective hypothesis decomposition

T1Jg3 KA10 KA13 KA12

0 24 48 72 96 120 144 168

Time [h]

12

14

16

18

20

22

24

26

28

Inh

ibit

ion

zon

e d

iam

eter

[m

m]

Figure 2. Mean values of mycelia growth inhibition zone diameter (mm) of test

phytopathogenic fungi caused by supernatant of the S. hygroscopicus filtrate during cultivation at 100 rpm for 7 days.

The experimental results obtained at a stirring speed of 200 rpm are presented in Figure 3. It can be seen that the largest inhibition zone diameter for all test phytopatho-genic fungi occurred in 96 h of cultivation, in the stationary phase of the bioprocess. In view of the fact that the stationary phase of the bioprocess begins between the third and fourth day of cultivation and that maximum amount of antifungal metabolites was formed after the fourth day of cultivation, it can be concluded that the synthesized antifungal metabolites are secondary metabolites.



239

Time*Fungi; LS Means

Current effect: F(21, 64)=5.1636, p=.00000

Effective hy pothesis decomposition

T1Jg3 KA10 KA13 KA12

0 24 48 72 96 120 144 168

Time [h]

12

14

16

18

20

22

24

26

28

30

32

Inh

ibit

ion

zon

e d

iam

eter

[m

m]

Figure 3. Mean values of mycelia growth inhibition zone diameter (mm) of test phytopathogenic fungi caused by supernatant of the S. hygroscopicus filtrate during

cultivation at 200 rpm for 7 days.

During the first 24 hours of cultivation, no antifungal activity against the isolates was observed, while significant growth inhibition appeared after 72 hours of cultivation of S. hygroscopicus (Figure 3). The largest zone of mycelia growth inhibition was noticed in the phytopathogenic fungi A. alternata KA10 (average inhibition zone diameter 29 mm) and A. alternata T1Jg3 (27.33 mm) in 96 h of cultivation (Figures 3 and 4). Slightly smaller inhibition zones gave phythopatogenic fungi from the Fusarium genus.

Figure 4. Inhibition zones caused by 100 μl of S. hygroscopicus cultivation filtrate

at 200 rpm in 96 h of cultivation: a) A. alternata KA10, b) A. alternata T1Jg3, c) F. avenaceum KA12, d) F. avenaceum KA13



240

Since the inhibition zones caused by S. hygroscopicus cultivation filtrate at 200 rpm are larger than in the case of 100 rpm, it can be assumed that a higher mixing rate sig-nificantly affects the morphology of the cells and, consequently, the larger production of antifungal metabolites. Leda et al. (15), in their study using S. viridosporus as micro-organism producer concluded that the increase in agitation rate from 200 to 500 rpm resulted in a consistent decrease in the cell growth and production as well. It is possible that an increase in the mixing rate resulted in high shearing effects that were detrimental to mycelium integrity. Besides, the streptomycin isolates acted differently at different mixing rates.

CONCLUSION The work was concerned with the influence of different agitation rate on the substrate consumption, biomass growth, and the biosynthesis of antifungal metabolite(s) during cultivation of S. hygroscopicus. Observing the time-course of substrate consumption and the cells growth in both bioprocesses it can be concluded that there is no lag phase and the exponential phase lasts until the third day when the stationary phase begins. The re-sults of this study showed that cultivation time has a statistically significant effect on the antifungal activity of S. hygroscopicus. Since the maximum productivity was observed after 96 h of cultivation in both agitation rates, it can be noticed that the produced active components are secondary metabolite(s). In accordance with this fact, it can be concluded that at both agitation rates (100 rpm and 200 rpm), the bioprocess could be stopped after 96 h of cultivation, which is a significant fact from an economical aspect. A larger inhibition zone diameter of the test phytopathogenic fungi observed at the higher mixing rate (200 rpm) after 96 h of cultivation indicates an enhanced oxygen transfer to the cells of S. hygroscopicus and a favorable effect on the cells morphology, making them more productive. Moreover, the results of the study indicate that the isolate of S. hygroscopicus has potential as an antagonist of tested postharvest apple pathogens from the genera Alterna-ria and Fusarium. Hence, after additional studies regarding its activity in vivo under realistic production conditions, it could be used for protection of apple fruits from storage pathogens, while simultaneously helping to solve the problem of using chemical pesticides, which present a great environmental problem.

Acknowledgement

The authors gratefully acknowledge the support of the Ministry of Education, Science and Technological Development of the Republic of Serbia, Project Number: TR- 31002.

REFERENCES

1. Hasan, H.A.H. Patulin and aflatoxin in brown rot lesion of apple fruits and their regu-lation. World J Microbiol Biotechnol. 2000, 16, 607-612.



241

2. Vico, I.; Duduk, N.; Vasić, M.; Nikolić, M. Identification of Penicillium expansum causing postharvest blue mold decay of apple fruit. Pestic. Phytomed. 2014, 29 (4), 257-266.

3. Grantina-Ievina, L. Fungi Causing Storage Rot of Apple Fruit in Integrated Pest Ma-nagement System and their Sensitivity to Fungicides. Rural Sustainability Research, 2015, 34, 2-11.

4. Granado, J.; Thürig, B.; Kieffer, E.; Petrini, L.; Flieβbach, A.; Tamm, L.; Weibel, P. F.; Wyss, S. G. Culturable Fungi of Stored ‘Golden Delicious’ Apple Fruits: A One-Season Comparison Study of Organic and Integrated Production Systems in Switzer-land. Microb Ecol. 2008, 56, 720-732.

5. Serdani, M.; Kang, J.C.; Andersen, B.; Crous, P.W. Characterisation of Alternaria species-groups associated with core rot of apples in South Africa. Mycol. Res. 2002, 106, 561-569.

6. Reuveni, M.; Sheglov, D. Effects of azoxystrobin, difenoconazole, polyoxin B (polar) and trifloxystrobin on germination and growth of Alternaria alternata and decay in red delicious apple fruit. Crop Prot. 2002, 21, 951-955.

7. Lee, H.B.; Patriarca, A.; Magan, N. Alternaria in Food: Ecophysiology, Mycotoxin Production and Toxicology, Mycobiology, 2015, 43 (2): 93-106.

8. Sorensen, J.L.; Phipps, R.K.; Nielsen, K.F.; Schroers, H.J.; Frank, J.; Thrane, U. Ana-lysis of Fusarium avenaceum Metabolites Produced during Wet Apple Core Rot, J. Agric. Food Chem. 2009, 57, 1632-1639.

9. Sever, Z.; Ivić, D.; Kos, T.; Miličević, T. Identification of Fusarium species isolated from stored apple fruit in Croatia. Arh. Hig. Rada. Toksikol. 2012, 63, 463-470.

10. Batta, Y.A. Postharvest biological control of apple gray mold by Trichoderma har-zianum Rifai formulated in an invert emulsion. Crop Prot. 2004, 23, 19-26.

11. Rončević, Z.; Grahovac, J.; Vučurović, D.; Dodić, S.; Bajić, B.; Tadijan, I.; Dodić J. Optimization of medium composition for the production of compounds effective against Xanthomonas campestris by Bacillus subtilis. APTEFF. 2014, 45, 247-258.

12. Kanini, G.S.; Katsifas, E.A.; Savvides, A.L.; Hatzinikolaou, D.G.; Karagouni, A.D. Greek indigenous streptomycetes as biocontrol agents against the soil-borne fungal plant pathogen Rhizoctonia solani. J Appl Microbiol. 2013, 114, 1468-1479.

13. Priya, E.; Thenmozhi, R.; Nagasathya, A.; Praveen Kumar, D.; Thajuddin, N.; Murali-tharan, G. Antagonistic potential of Streptomyces flavomacrosporus GACMPT-57 against plant pathogens. J. Microbiol. Biotech. Res. 2014, 4, 68-73.

14. Yen, H.W.; Li, Y.L. The effects of viscosity and aeration rate on rapamycin produc-tion in an airlift bioreactor by using Streptomyces hygroscopicus. J Taiwan Inst Chem Eng. 2014, 45, 1149-1153.

15. Leda, M.F.; Gottschalk, R.N.; Elba, P.S.B. Effect of Aeration on Lignin Peroxidase Production by Streptomyces viridosporus T7A. Appl Biochem Biotechnol. 2003, 105-108, 799-807.

16. Shioya, S.; Morikawa, M.; Kajihara, Y.; Shimizu, H. Optimization of agitation and aeration conditions for maximum virginiamycin production. Appl Microbiol Biotech-nol. 1999, 51, 164-169.



242

17. Mojićević, M.; Grahovac, J.; Petković, M.; Vučković, I.; Dodić, J.; Dodić, S.; Vojno-vić, S. Production of nigericin and niphimycin by soil isolate Streptomyces sp. MS1: anti-Candida bioassay guided response surface methodology for the optimized culture medium. Scientific Journal Facta Universitatis. in press.

18. Grahovac, M.; Balaž, J.; Grahovac, J.; Dodić, J.; Tanović, B.; Hrustić, J.; Tadijan, I. Screening of antagonistic activity of selected microorganisms against apple rot patho-gens. Rom. Biotechnol. Lett. 2014, 19 (1), 8959-8965.

19. Grahovac, J.; Grahovac, M.; Dodić, J.; Bajić, B.; Balaž, J. Optimization of cultivation medium for enhanced production of antifungal metabolites by Streptomyces hygro-scopicus. Crop Prot. 2014, 65, 143-152.

20. Herlich, K. Official Methods of Analysis of the Association of Official Analytical Chemists, fifth ed.; Association of Official Analytical Chemists: Arlington, 1990; pp. 758-759.

21. Gales, M.E.J.; Julian, E.C.; Kroner, R.C. Method for quantitative determination of total phosphorus in water. J. Am. Water Works Assess. 1966, 58, 1363-1368.

22. Singh, L.S.; Mazumder, S.; Bora T.C. Optimisation of process parameters for growth and bioactive metabolite produced by a salt-tolerant and alkaliphilic actinomycete, Streptomyces tanashiensis strain A2D. J de Mycologie Medicale. 2009, 19, 225-233.

23. Segy I. Methods of Soil Microbiology. Moskva, Mir, 1983. 24. Ripa, F.A.; Nikkon, F.; Zaman, S.; Khondkar, P. Optimal Conditions for Antimicro-

bial Metabolites Production from a New Streptomyces sp. RUPA-08PR Isolated from Bangladeshi Soil. Mycobiology. 2009, 37, 211-214.

25. Nigam, V.K.; Khandelwal, A.K.; Agarwal, A.; Mohan, M.K.; Vidyarthi, A.S. Produc-tion of a Thermostable Nitrilase in a Lab Scale Stirred Tank Bioreactor. IJBSBT. 2012, 4, 81-90.

26. Techapun, C.; Poosaran, N.; Watanabe, M.; Sasaki, K. Optimization of aeration and agitation rates to improve cellulase-free xylanase production by thermotolerant Strep-tomyces sp. Ab 106 and repeated fed-batch Cultivation using agricultural waste. J Biosci Bioeng. 2003, 95, 298-301.

27. Ilić, S.; Konstantinović, S.; Savić, D.; Veljković, V.; Joković, N.; Gojgić - Cvijović, G. The influence of modified media with carboxymethyl cellulose on morphology and antibiotic production by Streptomyces hygroscopicus. Biotechnol. and Biotechnol. Eq. 2008, 578-580.

28. Tadijan, I.; Grahovac, J.; Dodić, J.; Grahovac, M.; Dodić, S. Effect of Cultivation Time on Production of Antifungal Metabolite(s) by Streptomyces hygroscopicus in Laboratory-Scale Bioreactor. J Phytopathol. 2016, 164, 310-317.



243

УТИЦАЈ БРЗИНЕ МЕШАЊА НА ПРОДУКЦИЈУ АНТИФУНГАЛНИХ МЕТАБОЛИТА Streptomyces hygroscopicus У ЛАБОРАТОРИЈСКОМ

БИОРЕАКТОРУ

Ивана Ж. Митровић1*, Јована А. Граховац1, Јелена М. Додић1, Мила С. Граховац2, Синиша Н. Додић1, Дамјан Г. Вучуровић1, Вања Р. Влајков1

1 Универзитет у Новом Саду, Технолошки факултет,

Булевар цара Лазара 1, Нови Сад 21000, Србија

2 Универзитет у Новом Саду, Пољопривредни факултет, Трг Доситеја Обрадовића 8, Нови Сад 21000, Србија

Плодови јабуке су током читаве године присутни у исхрани деце и одраслих те је квалитет и здравствена безбедност ових намирница од изузетног значаја. Током вегетације и читавог периода складиштења, транспорта и продаје плодови јабуке су подложни инфекцији различитим фитопатогеним гљивама које плод користе као супстрат за раст, развој и размножавање те узрокују губитке који могу да износе од 50 до 80%. Међу значајним проузроковачима болести ускладиштених плодова јабу-ке налазе се и гљиве из родова Alternaria и Fusarium. Alternaria врсте проузрокују промене, како на листовима, тако и на плодовима јабука, док поједине врсте (A. arborescens, A. tenuissima и A. alternata) могу паразитирати и листове и плодове. Најчешћи проузроковачи трулежи плодова јабуке из рода Fusarium су: F. avena-ceum, F. culmorum, F. lateritium и F. solani. Зараза плодова врстом F. avenaceum може изазвати контаминацију сокова јабуке и осталих производа микотоксинима, секундарним метаболитима ових гљива, штетних за људско здравље. За заштиту плодова јабуке од фитопатогених гљива данас се углавном користе синтетички фунгициди. Међутим, због штетних екотоксиколошких особина и појаве резистент-ности, употреба хемијских фунгицида након бербе плодова забрањена је у великом броју земаља. Загађење животне средине, појава резистентних врста, неселек-тивност и остаци пестицида у храни довели су до потребе проналажења и кориш-ћења природних антагониста за биолошко сузбијање биљних патогена. Актиноми-цете рода Streptomyces представљају потенцијално значајну групу микроорганизама за производњу биоактивних компоненти значајних за пољопривреду. Услови кул-тивације, пре свега интензитет мешања, имају веома важну улогу у производњи антифунгалних агенаса применом Streptomyces врста, с обзиром да значајно утичу на пренос кисеоника у субмерзним култивацијама и на морфологију ћелија произ-водног микроорганизма. У овом раду испитан је утицај различитих интензитета мешања на производњу антифунгалних метаболита актиномицете Streptomyces hyg-roscopicus на пораст два изолата врсте Alternariа alternata и два изолата врсте Fusarium avenaceum. Култивација је изведена у лабораторијском биореактору (Bio-stat® Aplus, Sartorius AG, Germany) запремине 3 l на 27 °C са брзинама мешања од 100 о/мин и 200 о/мин током 7 дана. Циљ је био да се анализирају параметри био-процеса производње антифунгалних агенаса у подлози са глицеролом и испита у којим условима је производња антифунгалних метаболита већа. Ефикасност проду-



244

кованих антифунгалних метаболита на тест гљиве A. alternata (КА10 и Т1Јg3) и F. avenaceum (КА12 и КА13) испитивана је након свака 24 h култивације in vitro, дифузионом методом бунарчића. У експериментима који су спроведени у биореак-тору при различитим брзинама мешања (100 и 200 о/мин) утврђено је да се макси-мална производња антифунгалнихг метаболита уочава у 96 сату култивације. Већи пречници зона инхибиције код тестираних фитопатогених гљива и већа потрошња субстрата који се јављају у биореактору са већом брзином мешања, показују да ме-шање са 200 о/мин омогућава боље преношење кисеоника до ћелија S. Hygrosco-picus и има позитивне ефекте на морфологију ћелија производног микроорганизма, чинећи га продуктивнијим. Kључне речи: Streptomyces hygroscopicus, биолошка контрола, интензитет меша-

ња, Alternariа alternata, Fusarium avenaceum.

Received: 07 September 2017. Accepted:04 October 2017.

APTEFF, 48, 1-323 (2017) UDC: 579.852.11:615.281:66.011 https://doi.org/10.2298/APT1748245R BIBLID: 1450-7188 (2017) 48, 245-259


245

OPTIMIZATION OF THE MEDIUM COMPOSITION FOR PRODUCTION OF

ANTIMICROBIAL SUBSTANCES BY Bacillus subtilis ATCC 6633

Zorana Z. Rončević, Jovana J. Đuran*, Jovana A. Grahovac, Siniša N. Dodić, Aleksandra S. Ranitović, Jelena M. Dodić


In the effort to overcome the increase in antimicrobial resistance of different patho-gens, natural products from microbial sources appear to be the most favorable alter-native to current antibiotics. Production of antimicrobial compounds is highly dependent on the nutritional conditions. Hence, in order to achieve high product yields, selection of the media constituents and optimization of their concentrations are required. In this research, the possibility of antimicrobial substances production using Bacillus subtilis ATCC 6633 was investigated. Also, optimization of the cultivation medium composition in terms of contents of glycerol, sodium nitrite and phosphates was done. Response sur-face methodology and the method of desirability function were applied for determination of optimal values of the examined factors. The developed model predicts that the maxi-mum inhibition zone diameters for Bacillus cereus ATCC 10876 (33.50 mm) and Pseudo-monas aeruginosa ATCC 27853 (12.00 mm) are achieved when the initial contents of gly-cerol, sodium nitrite and phosphates were 43.72 g/L, 1.93 g/L and 5.64 g/L, respectively. The results of these experiments suggest that further research should include the utili-zation of crude glycerol as a carbon source and optimization of composition of such me-dia and cultivation conditions in order to improve production of antimicrobial substances using Bacillus subtilis ATCC 6633. KEY WORDS: Bacillus subtilis, antimicrobial activity, cultivation medium, response

surface methodology, optimization

INTRODUCTION The use of antimicrobial compounds, although appropriate and conservative, contri-butes to the development of resistance, while a widespread unnecessary and excessive use makes it worse (1). So, the alternatives for many of the currently used antimicrobial agents are needed. In agreement with the fact that pathogenic microorganisms gradually develop resistance and increasing consumers' awareness of the potential negative impact of synthetic preservatives, natural metabolites produced by microorganism have the po-tential to be used for biological control of pathogens in agriculture, food safety, and * Corresponding author: Jovana J. Đuran, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




246

medicine (2). Current trend of giving value to natural, environment-friendly and renew-able resources increases the potential use of these metabolites in the mentioned areas (3). Many Bacillus species, especially Bacillus subtilis strains have been identified as pro-ducers of a wide range of compounds with antimicrobial activity against different microorganisms, including pathogenic and spoilage bacteria, fungi and yeast (4). In view of the fact that Bacillus subtilis is recognized as a safe microorganism (GRAS status), it has received considerable attention as a biological control agent (5). Natural components with antimicrobial activity produced by Bacillus subtilis include cyclic lipopeptides, polypeptides, proteins (enzymes), and non-peptide products (3). To make the production of antimicrobial components feasible, it is necessary to develop the optimum production conditions. The formulation of the cultivation medium is of critical importance because its composition affects the yield and type of the synthesized components (6). In order to achieve high yields of antimicrobial agents the proper selection of carbon, nitrogen and phosphorus sources, as the most important nutrients in the cultivation medium, is requi-red. Among these factors, carbon source has the strongest influence on the production of bioactive compounds using the bacterial strains, and the effect of this nutrient on the bioprocess efficacy has been the subject of continuous study by both industry and re-search groups (7). These investigations showed that Bacillus subtilis is capable to use dif-ferent substrates as carbon source such as low-cost industrial by-products and waste ma-terials (8, 9). Glycerol, a major by-product of biodiesel manufacturing process, could be used as an organic carbon substrate in a different biotechnological production process (10, 11). Its bioconversion to high value compounds through microbial cultivation is an environmental-friendly choice that also contributes to the reduction of waste treatment costs and decreases the economics of overall production of the desirable product (12). Thus, optimization of glycerol-based media seems to be a very promising and economical route to produce antimicrobial metabolites using Bacillus subtilis. When it comes to nitrogen and phosphorus sources, nitrites and phosphates appeared to be appropriate for the biosynthesis of antimicrobial substances by Bacillus subtilis (13). Response surface methodology (RSM) is a collection of statistical and mathematical methods that are useful for the modeling and analyzing different biotechnological pro-cess. RSM defines the relationship between the controllable input parameters and the obtained responses, and also provides simultaneously information necessary for the pro-cess design and optimization (14). Successful application of RSM to enhance the antimic-robial compounds production by optimizing the culture medium has been reported (15). The aim of this study was to examine the possibility of the production of antimicro-bial substances effective against different pathogenic microorganisms, using Bacillus subtilis ATCC 6633. In order to increase the antimicrobial activity and reduce the costs of substrate preparation further research included the optimization of the cultivation me-dium composition in terms of contents of glycerol, sodium nitrite, and phosphate.



247

MATERIALS AND METHODS

Producing microorganism Bacillus subtilis ATCC 6633 was used in the experiments for antimicrobial substan-ces biosynthesis. The strain of producing microorganism was stored at 4 °C on commer-cial nutrient agar (Torlak, Serbia) medium and subcultured at monthly intervals.

Cultivation media The inoculum preparation was performed on a nutrient broth (Torlak, Serbia). The biosynthesis of antimicrobial substances was carried out on the nutrient broth and the semisynthetic glycerol-based media formulated according to the chosen experimental design. In agreement with the defined aim of study and the applied experimental design, contents of glycerol, sodium nitrite and phosphates (in the form of K2HPO4) in the medium were varied. The media also contained (g/L): yeast extract (0.5), CaCO3 (17.0), MgSO4x7H2O (0.5) and MnSO4x4H2O (0.05), and their pH value was adjusted to 7.0 prior to sterilization by autoclaving at 121 °C and pressure of 2.1 bar for 20 min.

Inoculum preparation and biosynthesis conditions The inoculum was prepared on the growth medium in two steps, first, by refreshing the culture by incubation for 48 h at 28 °C and second, by double passage of the mic-roorganism. Second passage was inoculated with 10 % (v/v) of inoculum obtained in first, and the incubation of each passage lasted 48 h at 28 °C. The samples were spon-taneously aerated and externally mixed (laboratory shaker, 150 rpm). The inoculation of the cultivation media was performed by adding 10 % (v/v) of the prepared inoculum. Production of substances with antimicrobial activity was performed in 300 mL Erlen-meyer flasks containing 100 mL of cultivation media of the appropriate composition. The biosynthesis of antimicrobial metabolites was carried out under aerobic conditions at the temperature of 28˚C and agitation rate of 150 rpm on a laboratory shaker (Ika® Werke IKA® KS 4000i control, Germany) for 96 h.

Analytical methods The supernatants obtained by centrifugation at 10,000 rpm for 15 min (Eppendorf Centrifuge 5804, Germany) were analyzed in terms of glycerol, nitrogen and phosphorus residues. The residual glycerol content was determined by high pressure liquid chroma-tography (HPLC). The samples were filtered through a 0.22 μm nylon membrane (Agi-lent Technologies Inc, Germany) and then analyzed. The HPLC instrument (Thermo Scientific Dionex UltiMate 3000 series) was equipped with a pump HPG-3200SD/RS, autosampler WPS-3000(T)SL (10 μL injection loop), column ZORBAX NH2 (250 mm x 4.6 mm, 5 μm), and detector RefractoMax520. The eluent was 70% (v/v) acetonitrile at a flow rate of 1.0 mL/min, and elution time of 10 minutes at column temperature of 30 °C.



248

The determination of the residual content of total nitrogen was performed by the Kjeldahl method (16), while residual content of total phosphorus was determined using spectrophotometric method (17).

In vitro antimicrobial activity assay Antimicrobial activity of the obtained cultivation broths was tested in in vitro, by standard disk-diffusion method using sterile disks (18). The test microorganisms were grown on the following commercial media: Muller-Hinton agar, Torlak®, Serbia (Pseu-domonas aeruginosa ATCC 27853), nutrient agar, Torlak®, Serbia (Bacillus cereus ATCC 10876), Sabouraud-Maltose agar, HiMedia®, India (Candida albicans ATCC 10231 and Aspergillus niger ATCC 16404). Agar was melted, cooled to 50±1 °C, and mixed in sterile conditions with the prepared suspension of test microorganism in a ratio of 9:1. The test microorganisms were treated with 15 μL of cultivation broth. A disk load with commercial antibiotic was prepared as a positive control, and a disk load with only distilled water was prepared as a negative control. After the incubation at 30 °C for 48 h, diameters of the zones of inhibition were measured using a zone scale from Himedia®, and expressed in millimeters (including disk diameter). Activity of each sample was tes-ted in three replicates for each microorganism.

Experimental design and optimization by RSM The response surface methodology based on the Box-Behnken experimental design was used for the optimization of media composition for antimicrobial substances pro-duction. The independent variables and their varied contents (g/L) were: X1 - glycerol (20, 35, 50), X2 - NaNO2 (1, 2, 3) and X3 - K2HPO4 (5, 10, 15). In this regard, according to the Box-Behnken experimental design with three factors at three levels and three repetitions in the central point (15) a set of 15 experiments was carried out. The relations between the independent variables and the responses Yi (inhibition zone diameter (mm), residual glycerol (g/L), residual nitrogen (g/L) and residual phosphorus (g/L) content) were determined by the second-order polynomial equation:

Yi = bo + Σ biXi + Σ biiXi2

+ Σ bijXiXj [1]

where Yi is the selected response, b0 is the intercept, bi, bii and bij are the linear, quadratic and interaction regression coefficient, respectively, while Xi and Xj are the varied factors. The adequacy of the model was evaluated by coefficient of determination (R2) and model p-value. The significance of regression coefficients was assessed by p-values at the 0.05 significance level. The statistical software package Statistica v. 13.0 was used for the regression analysis of the experimental data, and also to generate the response surface graphs. The method of desirability function was applied for the determination of optimal values of examined variables (Design-Expert 8.1).



249

RESULTS AND DISCUSSION In order to examine the possibility of antimicrobial substances production using Bacillus subtilis ATCC 6633 a preliminary experiment was carried out, which included cultivation of producing microorganism on commercial nutrient broth. Based on the inhi-bition zone diameters of 10.0 mm and 6.0 mm for Bacillus cereus ATCC 10876 (Gram-positive bacteria) and Pseudomonas aeruginosa ATCC 27853 (Gram-negative bacteria) obtained by the disk-diffusion method, respectively, it can be concluded that the applied cultivation broth had an antibacterial effect against tested bacteria, while no antifungal effect against Candida albicans ATCC 10231 and Aspergillus niger ATCC 16404 was detected. Gram-positive and Gram-negative bacteria have differences in their cell wall structure, the most important being the presence of a lipid-rich outer membrane in the cell wall of Gram-negative bacteria (19). Compared with Gram-positive bacteria, Gram-negative bacteria are more resistant against different extreme environmental conditions and antimicrobial agents because of the complex structure of their cell wall (20). So, the reason for the lower antimicrobial activity obtained against Pseudomonas aeruginosa ATCC 27853 may be the difference in the cell wall structure between tested bacteria. On the other hand, the fungal cell wall consists of glycoproteins and polysaccharides, mainly glucan and chitin, extensively cross-linked, to form a complex network (21). The obtai-ned results indicate that the antimicrobial substances produced in applied experimental conditions do not decompose complex structure of the fungal cell wall, which may be the reason for the high resistance of Candida albicans ATCC 10231 and Aspergillus niger ATCC 16404. In a large-scale production of antimicrobial components, commercial semisynthetic broths are not economically feasible medium, and in order to reduce the costs of raw material, some cheaper substrates are to be sought. Also, from the economical aspect of the production process, production of antimicrobial compounds with high yields is essential, and can be achieved primarily by media manipulation. Therefore, the other aim of this study was to optimize the medium composition in order to increase the product yield, reduce costs of substrate preparation and content of unused nutrients, which consequently cause reduction of the organic load of waste streams that will be generated after the separation of the bioactive components. In this research, the media for the production of antimicrobial substances by Bacillus subtilis ATCC 6633 was formulated by varying content of glycerol, sodium nitrite and phosphates. The values of the examined factors were selected based on the available literature data (13, 22).

Statistical analyses and definition of the mathematical models Based on the results of experiments formulated by the Box-Behnken design and regression analysis, quadratic polynomial equations were established to identify the relation between the selected responses and examined factors. The selected responses were: the inhibition zone diameter obtained against Bacillus cereus ATCC 10876 (mm, Y1), the inhibition zone diameter obtained against Pseudomonas aeruginosa ATCC 27853 (mm, Y2), residual glycerol content (g/L, Y3), residual nitrogen content (g/L, Y4)



250

and the residual phosphorus content (g/L, Y5). The results of the statistical analyses are presented in Table 1.

Table 1. Coefficients of regression equations and their significance for the inhibition zone diameters, residual glycerol, nitrogen and phosphorus content

Bolded values represent the significant coefficients at a confidence level of 95%.

The mode of interactions between the examined factors is indicated by the regression equations coefficients. A positive sign for the values of coefficients of interaction indicates a synergistic effect, while a negative sign represents an antagonistic effect of the factors on the selected response. The p-values serve as a tool for checking the signi-ficance of each of the coefficients of the regression equations. The factors with p-values less than 0.05 are significant at a confidence level of 95%. These coefficients are bolded in Tables 1. The adequacy and significance of the quadratic model was checked using the analysis of variance (ANOVA) which was tested using Fisher’s statistical analysis and the results are shown in Table 2. All polynomial models were significant at the 95% confidence level. The model F-values of 124.593, 151.900, 40.602, 331.611 and 32.847 for the inhibition zone diameter obtained against Bacillus cereus ATCC 10876 (mm, Y1) and Pseudomonas aeruginosa ATCC 27853 (mm, Y2), residual content of glycerol (Y3), nitrogen (Y4) and phosphorus (Y5), respectively, imply that the models for the selected responses are significant. The fitness of the model was checked and confirmed by the coefficient of determination (R2). The R2 value closer to 1 denotes better correlation between the observed and predicted values. The relatively high values of the deter-mination coefficient obtained for all responses indicate a good fit of the experimental data to Eq. (1).

Responses Y1 Y2 Y3 Y4 Y5

Effects C p-value C p-value C p-value C p-value C p-value

Intercept

b0 -38.751 0.068 -9.725 0.117 23.588 0.2295 -0.038 0.765 -0.163 0.660

Linear

b1 3.445 0.002 0.409 0.064 0.428 0.4927 -0.003 0.574 0.046 0.012

b2 -2.961 0.718 14.167 0.002 -20.78 0.0486 0.181 0.024 -0.208 0.257

b3 -0.454 0.781 -0.183 0.716 -4.032 0.0534 0.029 0.052 -0.076 0.066

Quadratic

b11 -0.036 0.004 -0.006 0.003 0.008 0.3262 0.000 0.410 -0.001 0.013

b22 -2.854 0.124 -2.896 0.002 3.705 0.0679 0.003 0.776 0.015 0.658

b33 -0.049 0.463 0.004 0.835 0.124 0.1097 -0.000 0.120 0.004 0.039

Interaction

b12 0.000 0.999 0.008 0.795 -0.108 0.3392 -0.000 0.376 -0.002 0.502

b13 -0.015 0.483 0.005 0.449 0.005 0.8306 0.000 0.999 -0.001 0.185

b23 1.250 0.008 -0.175 0.113 1.030 0.0202 -0.005 0.075 0.028 0.006



251

Table 2. Analysis of variance of the modeled responses

Response DF SS MS F-value p-value R2

Y1 5r

10m 44.100r

10990.150m 8.821r

1099.015m 124.593 0.000023 0.975

Y2 5r

10m 4.167r

1265.833m 0.834r

126.583m 151.900 0.000014 0.943

Y3 5r

10m 47.036r

3819.562m 9.407r

381.956m 40.602 0.000370 0.968

Y4 5r

10m 0.003r 2.043m

0.001r 0.204m

331.611 0.000002 0.974

Y5 5r

10m 0.019r 1.274m

0.004r 0.127m

32.847 0.00619 0.956

DF - degree of freedom; SS - sum of squares; MS - mean squares. r - residual; m - model.

Three-dimensional response surfaces were plotted on the basis of the model equation, to investigate the interaction among the examined factors and to determine the range of concentration of each factor for maximum antimicrobial compounds biosynthesis by Bacillus subtilis ATCC 6633. In the response surface plots, two factors varied when the third factor was kept at a fixed level (zero level). Figure 1A shows the effects of the initial content of glycerol and sodium nitrite at a constant concentration of phosphates (10.0 g/L) on the production of compounds with antimicrobial activity against Bacillus cereus ATCC 10876. As it can be seen from the response surface plot, the increase of the initial glycerol concentration in the medium results in the increase of inhibition zone in whole investigated range of sodium nitrite content. Also, the results in this figure indicate that over the glycerol concentration of about 40.0 g/L in the medium there is no significant increase of antimicrobial activity with further increase of nitrogen content. According to the predictions of the model, the inhibition zone is maximum (about 33.0 mm) at maximum values of initial glycerol content (40-50 g/L), regardless of the sodium nitrite concentration. The response surface plot presented in Figure 1B illustrates the effects of the initial content of glycerol and phosphates on the inhibition zone diameter at a constant sodium nitrite concentration (2.0 g/L). From this plot is evident that the antimicrobial activity increased with the increase of the initial glycerol concentration in the whole range of the initial phosphates content. Therefore, a maximum inhibition zone diameter of about 35.0 mm model predicts at the maximum values of the initial glycerol concentration (40-50 g/L), regardless of the initial phosphates content. The effects of the initial contents of sodium nitrite and phosphates on the biosynthesis of compounds effective against Bacillus cereus ATCC 10876 at a constant glycerol concentration (35.0 g/L) are presented in Figure 1C. According to the predictions of the model, the inhibition zone diameter is maximum (about 35.0 mm) if the initial nitrogen and phosphorus content are 3.0 and 15.0 g/L (maximal values of both factors), respectively.



252

Figure 1. Effects of the cultivation medium composition on inhibition zone diameter for Bacillus cereus ATCC 10876 (A, B, C) and Pseudomonas aeruginosa ATCC 27853

(D, E, F)



253

Also, when medium for production of antimicrobial compounds contains minimal initial concentrations of sodium nitrite (1.0-1.4 g/L) and phosphates (5.0-9.0 g/L) the model predicts inhibition zone diameter for Bacillus cereus ATCC 10876 of about 30.0 mm. A positive interaction between the two independent variables (coefficient of interaction, b23, Table 1) points to a synergistic effect between the contents of sodium nitrite and phosphates on antimicrobial component production by Bacillus subtilis ATCC 6633 effective against Bacillus cereus ATCC 10876. The minimum response predicted by the model was around 13.0 mm at the maximal content of sodium nitrite and minimal content of phosphates, which indicates that high concentration of nitrogen source in the medium with insufficient phosphorus content have a negative effect on metabolic activity of the producing microorganism. From the analysis of the response surfaces it can be seen that the increase of glycerol content in the medium affects positively the production of desired antimicrobial substances, which makes glycerol a limiting nutrient for their biosynthesis. Furthermore, the results presented in Figure 1A-C indicate that the nitrogen and phosphorus ratio is very important for the production of antimicrobial components effective against Bacillus cereus ATCC 10876. By maintaining the appropriate ratio of these nutrients in the cultivation medium a maximal production of bioactive components with antimicrobial activity is provided. The effects of the initial content of glycerol and sodium nitrite on the inhibition zone diameter obtained for Pseudomonas aeruginosa ATCC 27853 at a constant content of phosphates (10 g/L) are presented in Figure 1D. The obtained plot shows that the antimicrobial substances production increased with an increase in the initial glycerol concentration up to the range of 30.0 g/L to 45.0 g/L, beyond which there is a decrease. Also, the increase of the initial sodium nitrite content up to about 1.8-2.4 g/L, has a po-sitive effect on antimicrobial activity. With higher content of nitrogen source in the cul-tivation medium, the antimicrobial activity shows a significant decrease. Consequently, in accordance with the model predictions, a maximum inhibition zone diameter of approximately 12.0 mm was obtained at the initial glycerol concentration ranging from 30.0 g/L to 45.0 g/L and the initial sodium nitrite concentration ranging from 1.8 g/L to 2.4 g/L. The effects of the initial contents of glycerol and phosphates on the antimicrobial substances production in cultivation medium at a constant nitrogen concentration (2 g/L) are illustrated in Figure 1E. The presented results indicate that the decrease of phosphates content in the medium leads to a maximum of antimicrobial activity. A negative influence of the phosphates content in the medium could be explained by the fact that in the most cases antimicrobial compounds are products of the secondary metabolism, thus earlier beginning of stationary phase, during which the secondary metabolites are synthe-sized, is often achieved by limiting the phosphorus source content in the production medium (23). A maximum inhibition zone diameter of about 12.0 mm, according to the model predictions, is achieved at the initial glycerol content of 25.0-50.0 g/L and minimal phosphates content (5.0-8.0 g/L). Figure 1F illustrates the effects of the initial contents of sodium nitrite and phosphates on the biosynthesis of compounds effective against Pseudomonas aeruginosa ATCC



254

27853 at a constant glycerol concentration (35.0 g/L). From this response surface plot it can be seen that increase of the initial sodium nitrite content up to about 2.2-2.6 g/L, has a positive effect on the bioactive component concentration in cultivation media, regard-less of the initial phosphates content. This positive effect of sodium nitrite values in the examined range on the production of the desired components indicates that the proposed nitrogen concentrations are sufficient for the initial biomass multiplication and pro-duction of antimicrobial substances. Further increase of nitrogen content in the cultiva-tion medium will probably cause the prolongation of the exponential phase and signifi-cant nutrients depletion of the medium before the beginning of the stationary phase, which would result in a reduced production of desired components. Also, it is evident that, in the applied experimental conditions, the production of antimicrobial compounds decreases with the increase in the phosphates content in the medium. As can be seen from the response surface plot, the inhibition zone diameter is maximum (about 12.0 mm) at the minimal content of phosphates (5.0-8.0 g/L) and sodium nitrite content of about 2.0-2.8 g/L.

Optimization of the cultivation medium composition

The optimization of the medium composition for the antimicrobial compounds pro-duction was accomplished using a multi-response method called desirability function (24). This method involves the transformation of each response variable (Yi) to an indi-vidual function of desirability (di). In this study, the concept of desirability function was used to optimize the contents of glycerol, sodium nitrite and phosphates in the medium for production of substances effective against Bacillus cereus ATCC 10876 and Pseudo-monas aeruginosa ATCC 27853 using Bacillus subtilis ATCC 6633. Two sets of optimi-zation were made. The obtained results are presented in Table 3. The demonstrated antimicrobial activity of the substances produced by Bacillus subtilis ATCC 6633 against the tested Gram-positive (Bacillus cereus ATCC 10876) and Gram-negative (Pseudomonas aeruginosa ATCC 27853) bacteria indicated the potential broad-spectrum antimicrobial activity of the obtained cultivation broth. So, if the only goal of optimization is to achieve a maximum inhibition zone diameter against Bacillus cereus ATCC 10876 and Pseudomonas aeruginosa ATCC 27853 (first set, Table 3), the desirability function is of the maximum value (0.96) at the initial contents of glycerol, so-dium nitrite and phosphates of 43.71 g/L, 1.92 g/L and 5.63 g/L, respectively. By applying media with such nutrients contents, the model predicts the inhibition zone dia-meters of 33.50 mm (Bacillus cereus ATCC 10876) and 12.00 mm (Pseudomonas aeruginosa ATCC 27853). The predicted residual content of glycerol, total nitrogen and phosphorus are 15.20 g/L, 0.32 g/L and 0.16 g/L, respectively. When defining the media composition it must be taken into account that the efficiency of the antimicrobial substances production are improved when the residual nutrients con-tent is minimal, because the unused nutrients represent losses from an economic view-point (25). One of the common concerns regarding human and environmental health is the presence of organic components in wastewater, whose inappropriate disposal is an



255

environmental problem, and therefore requires additional processing prior to their release into the environment (26).

Table 3. The optimal values of factors and predicted values of selected responses

Factors and responses Goal Optimal value Goal Optimal value

First set Second set

Glycerol (g/L) in the range 43.71 in the range 34.98

NaNO2 (g/L) in the range 1.92 in the range 1.78

K2HPO4 (g/L) in the range 5.63 in the range 5.20

Predicted values Predicted values

Inhibition zone diameter A (mm)

maximize 33.50 maximize 29.22

Inhibition zone diameter B (mm)

maximize 12.00 maximize 12.00

Residual glycerol (g/L) in the range 15.20 minimize 8.83

Residual nitrogen (g/L) in the range 0.32 minimize 0.30

Residual phosphorus (g/L)

in the range 0.16 minimize 0.20

Desirability function 0.96 0.80

A - Bacillus cereus ATCC 10879; B - Pseudomonas aeruginosa ATCC 27853

Therefore, in the second optimization set (Table 3), in addition to achieving the maximum antimicrobial activity, the minimum residual contents of the nutrients were selected as responses. According to the model predictions for the highest value of desi-rability function (0.80) the optimal values of the initial contents of glycerol, sodium nit-rite and phosphate are 34.98 g/L, 1.78 g/L and 5.20 g/L, respectively. The predicted inhibition zone diameter obtained against Bacillus cereus ATCC 10876 is 29.22 mm and against Pseudomonas aeruginosa ATCC 27853 is 12.00 mm, while the residual nutrients content are 8.83 g/L, 0.30 g/L and 0.20 g/L for glycerol, total nitrogen and phosphorus, respectively. Comparing the results of the first and second optimization sets, a reduced concentration of residual glycerol by 41.91% and a decrease of the inhibition zone diameter by 12.78% for Bacillus cereus ATCC 10876 were noticed, while the inhibition zone diameter for Pseudomonas aeruginosa 27853 remained unchanged. The changes in the concentrations of residual nitrogen and phosphorus were not significant. To validate the model developed in the second optimization set, a new series of experiments were carried out with the optimum values of examined factors. On average, the experimentally obtained values of inhibition zone diameters for Bacillus cereus ATCC 10879 and Pseudomonas aeruginosa ATCC 27853 were 28.67±1.53 mm and 11.50±1.32 mm, respectively. For the residual sugar, total nitrogen, and phosphorus the average values in the additional experiments were 8.29±0.92 g/L, 0.28±0.09 g/L and



256

0.17±0.07 g/L, respectively. These results show that the experimentally determined va-lues are in good agreement with the statistically predicted values for all modeled responses, which confirmed the adequacy of the model. The results obtained in this study indicate a considerable difference in the antimicro-bial activity of cultivation broths against tested microorganisms, probably because they belong to the different bacterial species with characteristic differences in the structure of cell wall. In comparing the inhibition zone diameter obtained for two tested bacterial pathogens, it can be assumed that the antimicrobial components produced on semi-synthetic glycerol-based media also show the different antibacterial effect against Gram-positive and Gram-negative bacteria. Therefore, a detailed study is required to explore the relationship between the antimicrobial activity of the obtained cultivation broth and its concentration. Generally, concentrated bioactive products will provide a strengthened antimicrobial effect. Hence, these results represent a basis for further research in order to produce and formulate an antimicrobial agent with increased activity against Gram-ne-gative bacteria.

CONCLUSION The results obtained in this study demonstrated that Bacillus subtilis ATCC 6633, in applied experimental conditions produce substances with antibacterial effect against Bacillus cereus ATCC 10876 and Pseudomonas aeruginosa ATCC 27853 on both examined cultivation media (nutrient broth and semisynthetic glycerol-based media). Sta-tistical methodology was employed to define the composition of alternative medium with glycerol for production of antimicrobial compounds. The RSM based on the Box-Behnken design, as well as desirability function approach, were applied to find out the optimal concentration of glycerol, sodium nitrite and phosphates. The applied statistical techniques proved to be a valuable tool for improving medium composition that leads to a higher degree of antimicrobial compound production. In addition, the results of our study represent the basis for further improvements of the production of antimicrobial substan-ces using Bacillus subtilis ATCC 6633.

REFERENCES 1. Alanis, A.J. Resistance to Antibiotics: Are We in the Post-Antibiotic Era?. Arch. Med.

Res. 2005, 36(6), 697-705. 2. Gyawali, R.; Ibrahim, S.A. Natural products as antimicrobial agents. Food Control.

2014, 46, 412-429. 3. Wang, T; Liang, Y.; Wu, M.; Chen, Z.; Lin, J.; Yang, L. Natural products from

Bacillus subtilis with antimicrobial properties. Chin. J. Chem. Eng. 2015, 23(4), 744-754.

4. Yilmaza, M.; Sorana, H.; Beyatlib, Y. Antimicrobial activities of some Bacillus spp. strains isolated from the soil. Microbiol. Res. 2006, 161(2), 127-131.



257

5. Lee, N.K.; Yeo, I.C.; Park, J.W.; Kang, B.S.; Hahm, Y.T. Isolation and characte-rization of a novel analyte from Bacillus subtilis SC-8 antagonistic to Bacillus cereus. J. Biosci. Bioeng. 2010, 110(3), 298-303.

6. Dodić, J.; Rončević, Z.; Grahovac, J.; Bajić, B.; Korolija, O. Biosynthesis of compo-nents with antifungal activity against Aspergillus spp. using Streptomyces hygroscopi-cus. Hem. Ind. 2015, 69(2), 201-208.

7. Sanchez, S.; Chavez, A.; Forero, A.; Garcıa-Huante, Y.; Romero, A.; Sanchez, M.; Rocha, D.; Sanchez, B.; Avalos, M.; Guzman-Trampe, S.; Rodrıguez-Sanoja, R.; Langley, E.; Ruiz, B. Carbon source regulation of antibiotic production. J. Antibiot. 2010, 63, 442-459.

8. San-Lang, W.; Ing-Lung, S.; Chi-How, W.; Kuo-Choan, T.; Wen-Teish, C.; Yawo-Kuo, T.; Jen-Jon, R.; Chuan-Lu, W. Production of antifungal compounds from chitin by Bacillus subtilis. Enzyme Microb. Techno. 2002, 31(3), 321-328.

9. Islam, M.R.; Jeong, Y.T.; Lee, Y.S.; Song, C.H. Isolation and identification of antifungal compounds from Bacillus subtilis C9 inhibiting the growth of plant pathogenic fungi. Microbiology. 2012, 40(1), 59-66.

10. Sousa, M.; Melo, V.M.; Rodrigues, S.; Santana, H.B.; Goncalves, L.R.B. Screening of biosurfactant-producing Bacillus strains using glycerol from the biodiesel synthesis as main carbon source. Bioprocess Biosyst. Eng. 2012, 35(6), 897-906.

11. Sousa, M.; Dantas, I.T.; Felix, A.K.N.; de Sant'Ana, H.B.; Melo, M.M.; Gonçalves, L.R.B. Crude Glycerol from Biodiesel Industry as Substrate for Biosurfactant Pro-duction by Bacillus subtilis ATCC 6633. Braz. Arch. Biol. 2014, 57(2), 295-301.

12. Abad, S.; Turon, X. Valorization of biodiesel derived glycerol as a carbon source to obtain added-value metabolites: Focus on polyunsaturated fatty acids. Biotechnol. Adv. 2012, 30(3), 733-741.

13. El-Banna, N.M.; Quddoumi, S.S. Effect of nitrogen source on the antimicrobial activity of the bacilli air flora. Ann. Microbiol. 2007, 57(4), 669-671.

14. Ferreira, S.L.C.; Bruns, R.E.; Ferreira, H.S.; Matos, G.D.; David, J.M.; Brandão, G.C.; da Silva, E.G.P.; Portugal, L.A.; dos Reis, P.S.; Souza, A.S.; dos Santos WNL. Box-Behnken design: An alternative for the optimization of analytical methods. Anal. Chim. Acta. 2007, 597(2), 179-186.

15. Tanyildizi, M.S.; Özer, D.; Elibol, M. Optimization of a-amylase production by Ba-cillus sp. using response surface methodology. Process Biochem. 2005, 40(7), 2291-296.

16. Herlich K. Official Methods of Analysis of the Association of Official Analytical Chemists. 5th ed. Washington, DC, USA: Association of Official Analytical Che-mists; 1990.

17. Gales, M.E.Jr.; Julian, E.C.; Kroner, RC. Method for quantitative determination of total phosphorus in water. J. Amer. Water Works Assoc. 1966, 58(10), 1363-1368.

18. Bauer, A.W.; Kirby, W.M.M.; Sherris, J.C.; Turck, M. Antibiotic susceptibility tes-ting by a standardized single disk method. Am. J. Clin. Pathol. 1966, 45(4), 493-496.

19. Brown, L.; Wolf, J.M.; Prados-Rosales, R.; Casadevall, A. Through the wall: extra-cellular vesicles in Gram-positive bacteria, mycobacteria and fungi. Nat. Rev. Micro-biol. 2015, 13(10), 620-630.



258

20. Espad, R.M.; Espad, F.E. Lipid domains in bacterial membranes and the action of antimicrobial agents. BBA-Biomembranes. 2009, 1788(1), 289-294.

21. Bowman, S.M.; Free, SJ. The structure and synthesis of the fungal cell wall. BioEs-says. 2006, 28(8), 799-808.

22. El-Banna, N.M. Effect of carbon source on the antimicrobial activity of the air flora. World J. Microb. Biot. 2005, 21(8):1451-1454.

23. Sansinenea, E.; Ortiz, A. Secondary metabolites of soil Bacillus spp. Biotechnol. Lett. 2011, 33(8), 1523-1538.

24. Derringer, G.; Suich, R. Simultaneous Optimization of Several Response Variables. J. Qual. Technol. 1980, 19(12), 214-219.

25. Banat, I.M.; Satpute, S.K.; Cameotra, S.S.; Patil, R.; Nyayanit, N.V. Cost effective technologies and renewable substrates for biosurfactants’production. Front. Micro-biol. 2014, 5, 1-18.

26. Berkel, V.R.; Willems, E.; Lafleur, M. The Relationship between Cleaner Production and Industrial Ecology. J. Ind. Ecol. 1997, 1(1), 51-66.

OПTИMИЗAЦИJA СAСTAВA ХРAНЉИВE ПOДЛOГE ЗA ПРOИЗВOДЊУ AНTИMИКРOБНИХ СУПСTAНЦИ ПРИMEНOM Bacillus subtilis ATCC 6633

Зoрaнa З. Рoнчeвић, Joвaнa J. Ђурaн, Joвaнa A. Грaхoвaц, Синишa Н. Дoдић,

Aлeксaндрa С. Вeлићaнски, Jeлeнa M. Дoдић

Унивeрзитeт у Нoвoм Сaду, Teхнoлoшки фaкултeт Нoви Сaд, Булевар Цара Лазара 1, 21000 Нови Сад, Србиja

Jeдaн oд нajзнaчajниjих изaзoвa дaнaшњицe прeдстaвљa oчувaњe људскoг здрaв-љa пoрeд свe чeшћe пojaвe рeзистeнциje пaтoгeних микрooргaнизaмa нa пoстojeћe aнтимикрoбнe aгeнсe. Услeд ширeњa aнтибиoтскe рeзистeнциje, штo je пoслeдицa њихoвe нeпрaвилнe и прeкoмeрнe упoтрeбe, пoслeдњих гoдинa, нeкoнвeнциoнaлни aнтимикрoбни aгeнси свe вишe дoбиjajу нa знaчajу. Mикрooргaнизми прeдстaвљajу нeисцрпaн извoр биoaктивних jeдињeњa. Изражену прoизвoдну спoсoбнoст пoсe-дуjу бaктeриje рoдa Bacillus. Oнe прoдукуjу ширoк спeктрa прирoдних мeтaбoлитa сa изрaзитим aнтибaктeриjским и/или aнтифунгaлним дeлoвaњeм. Састав хранљи-вих подлога је од посебног значаја за микробиолошку производњу aнтимикрoбних супстaнци, jeр сe њeгoвим вaрирaњeм знaчajнo утичe нa пoвeћaњe принoсa жeљe-нoг прoизвoдa. У овом раду испитaнa je мoгућнoст биoсинтeзe aнтимикрoбних суп-станци примeнoм сoja Bacillus subtilis ATCC 6633 нa кoмeрциjaлнoj пoлусинтeтич-кoj пoдлoзи (хрaнљиви буjoн). Кao тeст микрooргaнизми oдaбрaни су сojeви Pseu-domonas aeruginosa ATCC 27853, Bacillus cereus ATCC 10876, Candida albicans ATCC 10231 и Aspergillus niger ATCC 16404. Фoрмирaни прeчници зoна инхиби-циje указују дa дoбиjeнa култивaциoнa тeчнoст пoкaзуje aнтимикрoбнo дeлoвaњe прoтив Bacillus cereus ATCC 10876 (10,0 mm) и Pseudomonas aeruginosa ATCC 27853 (6,0 mm), дoк прeмa осталим тест микроорганизмима ниje уoчeнa aнтимик-рoбнa aктивнoст. У нaстaвку истрaживaњa, извeдeнa je oптимизaциja сaстaвa пoдлo-



259

гe у пoглeду сaдржaja глицeрoлa, нaтриjум-нитритa и фoсфaтa. Биосинтеза анти-микробних компоненти изведена је на различтим подлогама чији је састав форму-лисан у складу са Box-Behnken-овим eкспeримeнтaлним плaнoм. За одређивање оптималног почетног садржаја нутриjeнaтa примењена је метода жељене функције у кoмбинaциjи сa пoлинoмским зaвиснoстимa пoсмaтрaних oдзивa дoбиjeних кo-ришћeњeм пoступкa oдзивнe пoвршинe. Нajeфикaсниje дeлoвaњe култивaциoнe тeчнoсти прoтив Bacillus cereus ATCC 10876 (33,50 mm) и Pseudomonas aeruginosa ATCC 27853 (12,00 mm), модел предвиђа при почетним садржајима глицерола, натријум-нитрита и фосфата oд 43,72 g/l, 1,93 g/l и 5,64 g/l, рeдoслeдoм. Поред тога, урађен је још један сет оптимизације са циљем минимизације рeзидуaлних сaдржaja нутријената, кaкo би сe утицaлo нa смaњeнe oргaнскoг oптeрeћeњa oтпaдних тoкo-вa, али и нa укупнe трoшкoвe припрeмe хрaнљивe пoдлoгe. Дoбиjeни рeзултaти прeдстaвљajу oснoву зa дaљa истрaживaњa кojимa би сe унaпрeдилa прoизвoдњa жељених aнтимикрoбних супстaнци, a кoja би пoдрaзумeвaлa испитивaњe мoгућ-нoсти упoтрeбe сирoвoг глицeрoлa кao извoрa угљeникa у пoдлoзи зa биoсинтeзу, oптимизaциjу прoцeсних услoвa, кao и дeфинисaњe кoрaкa у биoсeпaрaциoнoм низу кaкo би сe фoрмулисao гoтoв биoлoшки прeпaрaт. Кључнe рeчи: Bacillus subtilis, антимикробна активност, хрaнљивa пoдлoгa, RSM,

oптимизaциja


APTEFF, 48, 1-323 (2017) UDC: 633.43:547.562+547.979.8 https://doi.org/10.2298/APT1748261S BIBLID: 1450-7188 (2017) 48, 261-273


261

EXTRACTION AND ENCAPUSLATION OF BIOACTIVE COMPOUNDS FROM CARROTS

Vanja N. Šeregelj*, Gordana S. Ćetković, Jasna M. Čanadanović-Brunet,

Vesna T. Tumbas Šaponjac, Jelena J. Vulić, Slađana S. Stajčić

University of Novi Sad, Faculty of Technology Novi Sad, Bulevar Cara Lazara 1, 21000 Novi Sad

Carrot is an important root vegetable rich in bioactive compounds like carotenoids and polyphenols with appreciable levels of several other functional components having significant health-promoting properties. Utilization of carotenoids is limited due to their instability. Encapsulation is one of the alternatives used to improve carotenoid stability. The objectives of this study were to optimize the extraction and encapsulation of carote-noids from carrot. Freeze-dried carrots were extracted by conventional solvent extrac-tion (CSE) using four solvents, i.e. ethanol, acetone, ethyl acetate, and hexane. Although CSE using hexane and ethyl acetate resulted in highest carotenoid contents (18.27 and 15.73 mg β-carotene/100 g), acetone and ethanol at a solid to solvent ratio 1:10 w/v with slightly lower carotenoid contents (14.52 and 11.45 mg β-carotene/100 g) were chosen for further studies due to their higher food compatibility and higher polyphenol content (88.86 and 66.21 mg GAE/100 g) than with lower solid to solvent ratio (1:5 w/v). Ethanol and acetone carrot extracts were encapsulated using different carriers by a freeze-drying method in order to obtain the optimum encapsulate with the highest carotenoid encapsu-lation efficiency (EE). Encapsulation using the maltodextrin, whey and soy protein as a wall material yielded an EE of carotenoids ranging from 41.95% to 100%. The en-capsulated β-carotene content was evaluated during two months of storage at ambient temperature under light and dark conditions. Generally, the retention of carotenoids was significantly higher in dark conditions, where maximum retention (65.94-87.32%) occur-red in the samples encapsulated in maltodextrin and soy protein. KEY WORDS: bioactive compounds, carotenoids, extraction, encapsulation, storage

INTRODUCTION In the recent years, the worldwide demand for natural pigments has been increased due to the increased awareness of their health benefits and nutritional properties, and also because of the recognized toxicity of synthetic pigments (1). In view of their provitamin A activity and antioxidant function, carotenoids are one of the most important natural * Corresponding author: Vanja N. Šeregelj, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




262

food colorants. Epidemiological studies demonstrated that the consumption of diets rich in carotenoids is associated with lower incidence of cancer, cardiovascular diseases, age related macular degeneration and cataract formation (2,3). Carotenoids are red, yellow and orange pigments, which are widely distributed in nature and are especially abundant in yellow-orange fruits and vegetables and dark green leafy vegetables (4). Among these, carrots and carrot-based products are the main source of carotenoids in the human diet. According to Chen et al. (5) β-carotene constitutes a lar-ge portion (60-80%) of the carotenoids in carrots followed by α-carotene (10-40%) and lutein (1-5%). In addition to carotenoids, carrot also contains the appreciable amount of phenolics which have been linked to reducing the risk of major chronic diseases (6). The-refore, studies on the most convenient and effective extraction methods of these bioactive compounds have been attracting the interest of many researchers. There are numerous methods used for extraction of bioactive compounds and some examples include conventional solvent extraction (CSE). However, the extraction effici-ency varies widely due to different polarities of targeted bioactive compounds and sol-vents employed in the extracting process. For example, phenolic compounds are extrac-ted more effectively with highly-polar solvents, while non-polar or low-polar solvents showed higher extraction yields of lipophilic compounds like carotenoids (7,8). Pinelo et al. (9) suggest that the extraction conditions such as solvent concentration, contact time, extraction temperature, solvent to solid ratio and particle size are also important parame-ters affecting the extraction yield of bioactive compounds content and antioxidant ca-pacity of the extracts. Nowadays, the nutritional content of food after preparation and during storage has become an important concern for consumers and manufacturers who have contributed to better food quality and, consequently, to consumers' health. Incorporation of β-carotene in various food systems is limited by its poor water solubility and instability in the pre-sence of light, heat, and oxygen (10). Encapsulation is one of the alternative techniques used for improving carotenoid stability, which entraps a sensitive ingredient inside a coa-ting material (11). The initial step in encapsulating is the selection of a suitable encap-sulation technique and coating material, especially considering their incorporation in food products. Among several methods of encapsulation, freeze drying, also known as lyo-philization, is one of the most useful processes for drying thermosensitive substances such as natural pigments and aromas (12). Different wall materials for coating can be used and a maltodextrin, soy and whey protein are often used. The aim of the present study was to elucidate the effect of different extraction parameters to obtain carrot extracts with high carotenoid and phenolic content, as well as high antioxidant capacity. To protect extracted sensitive bioactive compounds, the obtai-ned and selected extracts were encapsulated by freeze drying method. For this purpose, the effect of different encapsulation agents to obtain the powder with high carotenoid retention was evaluated. The influence of storage and light exposure on carotenoid con-tents in the encapsulates was also examined.



263

EXPERIMENTAL

Chemicals and instruments The solvents used for extraction (ethanol, acetone, hexane and ethyl acetate) were purchased from “Zorka” Chemical Co. (Šabac, Serbia). Folin-Ciocalteu reagent, 2,2-di-phenyl-1-picrylhydrazine (DPPH), gallic acid, trichloroacetic acid and Trolox were from Sigma Chemical Co. (St Louis, MO, USA). Maltodextrin was provided by Battery Nu-trition Limited (London, United Kingdom), while soy and whey protein isolate were pur-chased from Olimp Laboratories (Debica, Poland). All other chemicals and solvents used were of the highest commercial grade. Absorbances in spectrophotometrical assays were measured on a MultiskanGO microplate reader (Thermo Fisher Scientific Inc., Waltham, MA, USA) and UV-1800 spectrophotometer (Shimadzu, Kyoto, Japan).

Plant material Fresh carrots were produced by Beoflora d.o.o., Beograd, Serbia, and purchased from the local store. After washing, fresh carrots were chopped in a kitchen blender (B 800 E, Gorenje, Slovenia), dried at -40 °C for 48 h in a freeze-drier (Martin Crist Alpha 2-4, Osterode, Germany), finely ground, and stored at -20 °C until use.

Optimization of the extraction of bioactive compounds from carrots

The extraction of bioactive compounds from carrots implied optimization of the pro-cess considering different polarity of extracting solvents and solid to solvent ratio. The first set of experiments included conventional solvent extraction (CSE) with polar (etha-nol, acetone and ethyl acetate) and non-polar (hexane) solvents at solid to solvent ratio of 1:10 (w/v). The second set of experiments included solvent extraction with ethanol and acetone at a ratio of 1:5 (w/v).

Conventional solvent extraction Freeze-dried carrots were extracted with extracting solvent in two different solid to solvent ratios (1:10 and 1:5 w/v) for 10 minutes, three times with the same volume of solvents. The extraction was performed using a laboratory shaker (Unimax 1010, Hei-dolph Instruments GmbH, Kelheim, Germany) at 300 rpm, under light protection, at room temperature. The obtained three extracts were filtered (Whatman paper No.1), combined, and stored in dark bottles at -20 °C to further analysis.

Determination of bioactive compounds content in carrot extracts

Total carotenoid content (TCar). The content of carotenoids in the carrot extracts was analyzed spectrophotometrically by the method of Nagata and Yamashita (13), using extracting solvent as the blank. The content of TCar was calculated using the equation:



264

TCar (mg β-carotene/100ml) = 0.216A663 - 1.22A645 - 0.304A505 + 0.452A453 where A663, A645, A505 and A452 represent the absorbances measured at 663 nm, 645 nm, 505 nm and 453 nm, respectively. The TCar was expressed as mg of β-carotene equiva-lents per 100 g DW. Total phenolics content (TPh). The amount of total phenolics in carrot extracts was determined spectrophotometrically according to the Folin-Ciocalteau method adapted for 96 well microplate (14). The results were expressed as gallic acid equivalents (GAE) per 100 g DW.

Determination of carrot extracts bioactivity

Antiradical activity by DPPH assay (AADPPH). The free radical scavenging effect of carrot extracts on 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical was measured spectro-photometrically based on the modified method by Jiménez-Escrig et al. (15). The ability to scavenge DPPH radicals, i.e. AADPPH, was calculated following the equation:

AADPPH (%) = [(AC − AS) / AC] × 100

where AC is the absorbance of the control, and AS is the absorbance in the presence of the extract. The results were also expressed as µmol Trolox equivalents (TE) per 100 g DW. Reducing power assay (RP). The RP was determined by the method adopted from Oyaizu (16), measuring the reduction of the Fe3+/ferricyanide complexes to the ferrous (Fe2+) form. The absorbances were read at 700 nm against the control. The calibration curve was constructed using Trolox, and the results were expressed as µmol Trolox equi-valents (TE) per 100 g DW. Encapsulation optimization. Three carrier agents, including maltodextrin (MD), soy protein (SP) and whey protein (WP) isolate were used as wall materials for carotenoid encapsulation. Two criteria were considered for optimization of encapsulate formulation: the amount of core material contained in the powders and the retention of core material inside powders affected by different light exposure conditions during storage. Emulsion preparation and freeze-drying. All the powders were prepared according to the method developed by Indrawati et al. (17), with some modifications. MD and WP (7g) were dissolved in 10.5 ml water at 60 °C and kept under stirring until the tempera-ture reached 30 °C, while SP was dissolved in the same way in 40 ml of water. Tween 80 (0.1 g) was added as an emulsion stabilizer. Separately, 10 ml of each extract was com-bined with sunflower oil (1.5 ml), concentrated under reduced pressure on a rotary eva-porator set at 40 °C to remove the organic solvent, and immediately mixed with pre-viously prepared carrier solution. The solutions were vigorously homogenized at 11,000 rpm for 3 min at room temperature, frozen overnight, and subjected to freeze-drying at -40 °C for 48 h. The encapsulated powders were packed in airtight plastic bags and stored at 4 °C until further use. Contents of Surface Carptenoid (SC) and Total Carotenoid (TC) in the encapsulate and Encapsulation Efficiency (EE). SC, TC, and EE were determined by following a modified Barbosa et al. method (11). SCs were determined by direct extrac-



265

tion of 0.25 g encapsulate sample with 5 ml acetone on a vortex for 20 s, followed by centrifugation at 5000 rpm (10 min) and supernatant separation. For TC determination, 0.25 g of sample was vortexed with 0.2M PBS (pH 7) for 1 min to break the capsules, extracted with 2 ml of acetone and 3 ml of diethyl ether, and left over night at 4 °C. After separating the pigment layer, extraction was repeated with the same solvent volumes to collect total pigment. The carotenoid quantification was carried out according to the pre-viously described protocols. The encapsulation efficiency was calculated by using the equation:

% EE = [(TC-SC)/TC] x 100 Simultaneously, the control samples, i.e. the carriers without extracts, were prepared in the same way, for the correction of interfering substances originating from the carrier material.

Storage stability tests The control and pigment microcapsules were stored at room temperature in glass and amber bottles for 2 months to determine the effect of time and light exposure on the stability of carotenoids. For that purpose, total carotenoids were determined by described method every month.

RESULTS AND DISCUSSION Previous studies have indicated that carrot extracts represent a rich source of bioactive compounds (18, 19). According to the knowledge that their yield significantly depends on the different extraction conditions, solvent polarity and sample to solvent ratio were evaluated in order to increase the extraction efficiency and choose the most convenient method for further use in food formulations. The effect of different extraction conditions on the total carotenoid and phenolic yields in carrot extracts are shown in Table 1. When using the solvent to solid ratio 1:10, the carotenoid extraction yields obtained under studied conditions were in the range from 0.19 to 18.27 mg β-carotene/100g DW. As expected, the highest content of carotenoids was obtained using hexane as an extrac-ting solvent, according to the empirical rule "like dissolves like". The carotenoid yield in ethyl acetate extract was close to the one obtained with acetone (15.73 and 14.52 mg β-carotene/100 g DW, respectively). Yen et al. (20) reported 24 mg β-carotene/100g DW in freeze-dried carrots successively extracted with acetone and petroleum ether. Using con-ventional solvent extraction employing diacetone alcohol and petroleum ether, and furt-her purification with diacetone alcohol, methanolic KOH and distilled water, Dutta et al. (21) determined lower content of carotenoids (8.4 mg β-carotene/100g DW) compared to the results obtained for conventional solvent extraction in this study. However, the yields of phenolics were significantly different, depending on the sol-vent used for extraction. These varied between 34.85 and 165.52 mg GAE/100 g DW. The results showed that ethyl acetate was superior in isolating phenolic compounds com-



266

pared to the other three solvents in the CSE method. This is probably due to the higher polarity and better solubility of phenolic components present in plant material (22). Bys-tricka et al. (23) reported 6-fold lower phenolics content for carrot ethanolic extract, while Cefola et al. (24) reported five times less phenolics in carrot acetone extract (11.37 mg GAE/100g DW and 17.06 mg GAE/100g DW, respectively) compared to the contents in carrot extracts obtained with the same solvents in this study. The selection of a solvent is based on several physicochemical properties along with the cost cost and toxicity. Some solvents, such as ethanol and acetone are designated as "generally recognized and safe" (GRAS) and are the preferred solvents currently used for food products. Since our research focuses on giving priority to the food grade solvents, in order to determine the optimal ratio of plant material/solvent, the extraction of carrots with ethanol and acetone were repeated. From the results presented in Table 1 it follows that using the both ratios (1:10 w/v and 1:5 w/v, respectively) gives approximately equal carotenoid content, with no statistically significant difference (p ≤0.05) in case of etha-nol. The use of ethanol in a ratio of 1:10 w/v, resulted in 2.5% more carotenoids com-pared to the other ratio studied (1:5 w/v), while for acetone extraction the content was by 14.3% lower. In the case of phenolic compounds, a significant increase of the isolation efficiency with higher volumes of extracting solvents was observed. The total phenolic content was by 123.5% higher for ethanolic extract and 154.8% for acetone extract com-pared to the extracts obtained with lower volumes of the same extracting solvents. How-ever, for the encapsulation and storage studies, ethanol and acetone extracts obtained at a 5:1 ratio extraction were chosen because of higher contents of carotenids and the poten-tial of these encapsulates for further use as colorants.

Table 1. Total carotenoid and phenolic contents in carrot extracts

Values with the same letter in the superscript in the same column are not significantly different (p<0.05). TCar - total carotenoid content; TPh - total polyphenol content.

Since carotenoids, along with polyphenols, are the most important phytochemicals in carrots, they are considered as responsible for the antioxidant activity. For the antioxidant activity assessment of carrot extracts, the DPPH free radical scavenging activity as well as reducing power, were monitored spectrophotometrically. The relationship between the carrot extracts concentration and discoloration of DPPH radical solution, reflecting the radical scavenging activity of the analyzed extracts, is shown in Figure 1. The magnitude of the radical scavenging effect of the carrot extracts was ranked as ethanol > acetone > ethyl acetate > hexane. The polarity of the solvents used for the extraction of carrot was different, yielding different composition of the ex-

Extraction method Extraction

solvent TCar

(mgβ-carotene/100g DW) TPh

(mgGAE/100 g DW)

Conventional solvent extraction (10:1 v/w)

Ethanol 11.45±0.52a 66.21±0.15a Acetone 14.52±0.67b 88.86±0.22b

Ethyl acetate 15.73±0.65b,d 165.52±0.12c Hexane 18.27±0.73c 34.85±0.09d

Conventional solvent extraction (5:1 v/w)

Ethanol 11.17±0.51a 29.62±0.02e Acetone 16.60±0.72d 34.85±0.15d



267

tracts. The ethanol, acetone and ethyl acetate extracts had the highest total bioactive com-pounds content (i.e. carotenoids and polyphenols), which resulted in the superior DPPH radical scavenging activity in the range from 120.07 to 195.01 µmol TE/100 g DW. On the other hand, the hexane extract did not show any activity towards DPPH radical. Be-sides various antioxidant mechanisms involved in radical scavenging activity, a synergis-tic activity among present bioactive molecules may occur. Cefola et al. (22) investigated the antioxidant activity of various carrot varieties (yellow, orange and purple) on DPPH radicals. The obtained results indicated the highest antioxidant activity of purple carrots, due to the highest content of bioactive phytochemicals, where the total carotenoid and phenolic contents were four times higher than with the other two varieties. The antioxi-dant activity of yellow and orange varieties was similar. The reducing power assay may serve as a significant indicator of potential antioxidant activity, which is evidenced by measuring the ability of the extracts to reduce the Fe3+/ferricyanide complex to the ferrous form. Figure 1b shows the reducing power of the five carrot extracts, where the results of the ethanol and acetone extracts indicate the highest reducing ability (7368.07 and 3167.91 µmol TE/100 g DW, respectively), and ethyl acetate extract the lowest (71.19 µmol TE/100 g DW).

Figure 1. Antiradical activity (a) and reducing power (b) of different carrot extracts

The low bioavailability of carotenoids from natural sources and limited incorporation of β-carotene in food products due to its poor water solubility and instability in the presence of heat, light, and oxygen, has led to the development of encapsulation methods, to improve stability and achieve maximum health benefits. The EE value is considered as an important parameter for evaluating the properties of microcapsules, and it presents the ratio of the pigment bound inside the microcapsules to the total carotenoid content in the microcapsule, comprising the bound and SC. The SC, TE, and EE of carotenoids for all microcapsules are shown in Table 2. The EE value varied considerably with the wall materials and composition of encap-sulating materials. In all acetone extracts. microcapsules, surface and total carotenoid contents were higher than in the ethanol extract microcapsules, indicating better pigment isolation with acetone.



268

Table 2. Surface, total carotenoids, and encapsulation efficiency of carotenoids for ethanol and acetone extracts encapsulated in maltodextrin, whey, and soy protein

Sample SC

(mg β-carotene/100g DW) TC

(mg β-carotene/100g DW) EE (%)

ME 0.23 ± 0.02a 0.49 ± 0.04a 53.80 ± 0.29a MA 0.67 ± 0.00b 1.14 ± 0.06b 41.95 ± 2.56a WPE 0.13 ± 0.04a,c 0.64 ± 0.05a 84.87 ± 3.92b WPA 0.16 ± 0.01a,c 1.21 ± 0.06b 85.89 ± 0.88b SPE 0.00 ± 0.00c 0.46 ± 0.06a 100 ± 0.00b SPA 0.51 ± 0.02b 1.14 ± 0.06b 55.36 ± 0.43a

Values with the same letter in superscript in the same column are not significantly different (p<0.05). SC, TC - surface and total carotenoids in microcapsules; EE - encapsulation efficiency; ME, MA - etha-

nol and acetone extract encapsulated in maltodextrin, respectively, WPE, WPA - ethanol and acetone ex-tract encapsulated in whey protein, respectively; SPE, SPA - ethanol and acetone extract encapsulated in soy protein.

The encapsulation efficiency varied from 41.95% up to 100%. The difference bet-ween EE of acetone and ethanol extracts encapsulated in the same carrier was not significantly different, except for the case of soy protein. The maximum efficiency of 100% was obtained for the ethanol carrot extract with soy protein, whereas in the case of acetone extract this value was 55.36%. Since the initial concentration of carotenoids in the ethanol extract was lower, the total amount was successfully encapsulated with soy protein. Generally, both proteins (soy and whey), were found to be better carriers for carotenoids. This is consistent with the reports of Grimme and Brown (25), showing that carotenoids in plants are bound by protein. According to Jafari et al. (26), functional properties of proteins including solubility, film formation, emulsification and stabilization of emulsion droplets, exhibited many de-sirable characteristics for the wall material. These proteins changed their structure during emulsification through unfolding and adsorption at the oil-water interface, and sub-sequently formed a resistant multilayer around oil droplets. That would stabilize the oil-in-water emulsion and improve lipid microencapsulation. In addition, the maltodextrin showed the lowest encapsulation efficiency. This was expected given the fact that powder had the highest amount of unencapsulated β-carotene, which negatively affects its EE. It is well known that carotenoids are unstable against light, oxygen, moisture, and temperature. The degradation rate of carotenoids in microcapsules, under light and dark conditions at room temperature, were monitored by the change in carotenoid retention upon storage for 60 days, as shown in Table 3. As expected, an increase in the storage period in the presence of light resulted in progressive loss of β-carotene in all microcap-sules, confirming instability of carotenoids. For instance, after two months exposed to light, the carotenoids completely vanished from the acetone extract encapsulated in mal-todextrin and from the ethanol extract encapsulated in whey protein. In most of the mic-rocapsules, the decline in the carotenoid content between first and second month was significant (p<0.05). On the other hand, the degradation of encapsulated carotenoids un-der dark conditions was much slower compared to that in the opposite conditions. More-



269

over, there was no significant difference (p<0.05) in the carotenoid content after the first and second moth of storage in amber bottles, except in the case of the acetone extract encapsulated in soy protein.

Table 3. Changes in carotenoid content during storage of the microencapsulates

Values with the same letter in superscript in the same column are not significantly different (p<0.05). ME, MA - ethanol and acetone extract encapsulated in maltodextrin, respectively, WPE, WPA - ethanol

and acetone extract encapsulated in whey protein, respectively; SPE, SPA - ethanol and acetone extract encapsulated in soy protein.

n.d. - not detected

Figure 2. Carotenoid retention in the encapsulate samples after one and two months of storage exposed to light (a), and in the dark (b)

Time (months) / Light exposure

Encapsulate sample ME MA WPE WPA SPE SPA

0 0.47±0.04a 1.09±0.10a 0.59±0.09a 1.15±0.11a 0.43±0.06a 1.10±0.09a

1 Light 0.21±0.04b 0.46±0.07b 0.07±0.00b 0.48±0.07b 0.09±0.01b,d 0.53±0.02b Dark 0.41±0.04a 0.86±0.09a,c 0.18±0.01c 0.49±0.07b 0.23±0.03c 0.77±0.03c

2 Light 0.03±0.00c n.d. n.d. 0.03±0.00c 0.04±0.00b 0.29±0.01d Dark 0.37±0.04a 0.82±0.05c 0.17±0.00c,d 0.24±0.02b 0.18±0.00c,d 0.49±0.05b



270

The retention of carotenoids, presented in Figure 2, decreased more rapidly during the first 30 days due to exposure to the light for the MD, WP and SP microcapsules with ethanolic extract (to 44.42%, 11.58% and 17.96%, respectively), as well as for the micro-capsules obtained with acetone extract (to 39.94%, 36.38% and 46.57%, respectively).

CONCLUSION In the present study, the elaboration of the effect of different extraction parameters revealed that they have a significant effect on the extraction yield. Pure ethanol and ace-tone are considered as safe solvents, and the high concentration of bioactive compounds from carrots could be isolated with these solvents. Further, the antioxidant potential eva-luation showed that both ethanol and acetone extracts were promising. The major prob-lem of these ingredients is their instability in the presence of heat, light, and oxygen. Hence, the carrot extracts were mixed with three different wall materials (MD, WP and SP) for the evaluation of the encapsulation efficiency and storage stability. All wall ma-terials largely increased the half-life of the encapsulated pigments during storage in dark conditions compared with the samples exposed to light. Generally, the proteins showed the highest encapsulation efficiency, lower degradation rate in both studied conditions, and were defined as the most effective wall material in stabilizing the pigments. Hence, the significance of this research is in demonstrating successful encapsulation and exten-sion of the shelf life of carotenoid content, which can be very important for obtaining attractive and acceptable final products.

Acknowledgements

This research is part of the Project TR 31044 which is financially supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia and by COST Action Eurocaroten CA15136.

REFERENCES 1. Chaitanya Lakshmi G. Food Coloring: The Natural Way. Research Journal of Che-

mical Sciences 2014, 4 (2), 87-96. 2. Meyers, K. J.; Mares, J. A.; Igo, R. P.; Truitt, B.; Liu, Z., Millen, A. E.; Klein M.;

Johnson E.J.; Engelman, C. D.; Karki , C. K.; Blodi, B.; Gehrs, K.; Tinker, L.; Wallace, R.; Robinson, J.; LeBlanc E. S.; Sarto, G.; Bernstein, P. S.; SanGiovanni, J. P.; Iyengar, S. K. Genetic evidence for role of carotenoids in age-related macular degeneration in the carotenoids in age-related eye disease study (CAREDS). Investigative Ophthalmology & Visual Science, 2014, 55 (1), 587-599.

3. Sharoni, Y.; Linnewiel-Hermoni, K.; Khanin, M.; Salman, H.; Veprik, A.; Danilenko, M.; Levy, J. Carotenoids and apocarotenoids in cellular signaling related to cancer: A review. Molecular Nutrition & Food Research, 2012, 56 (2), 259-269.



271

4. Ji, J.; Wang, G.; Wang, J.; Wang, P. functional analysis of multiple carotenogenic ge-nes from Lycium barbarum and Gentiana lutea L. for their effects on beta-carotene production in transgenic tobacco. Biotechnol. Lett. 2009, 31 (2), 305-312.

5. Chen, B. H.; Peng, H. Y.; Chen, H. E. Changes of carotenoids, Color, and Vitamin A contents during Processing of carrot juice. J. Agric. Food Chem. 1995, 25, 515.

6. Liu, R. H. Potential synergy of phytochemicals in cancer prevention: mechanism of action. J. Nutr. 2004, 134 (12), 3479S-3485S.

7. Santos-Buelga, C.; Gonzalez-Manzano, S.; Duenas, M.; Gonzalez-Paramas, A. M. Extraction and isolation of phenolic compounds. Methods Mol. Biol. 2012, 864, 427-464.

8. Prado, J. M.; Veggi, C. P.; Meireles, M. A. A. Extraction methods for obtaining caro-tenoids from vegetables - Review. Current Analytical Chemistry, 2014, 10, 29-66.

9. Pinelo, M.; Fabro, P. D.; Manzocco, L.; Nunez, M. J.; Nicoli, M. C. Optimization of continuos phenol extraction from Vitis vinifera byproducts. Food Chem. 2005, 92, 109-117.

10. Gul, K.; Tak, A.; A. K. Singh, A. K.; Singh, P.; Yousuf, B.; Wani, A. A. Chemistry, encapsulation, and health benefits of β-carotene - A review. Cogent Food & Agriculture, 2015, 1, 1018696.

11. Barbosa, M. I. M. J.; Borsarelli, C. D., Mercadante, A. Z. Light stability of spray-dried bixin encapsulated with different edible polysaccharide preparations. Food Research International, 2005, 38, 989-994.

12. Karel, M.; Langer, R. Controlled release of food additives. Risc S, Reineccius G. In; Flavour Encapsulation. American Chemical Society, 1988, 177-191.

13. Nagata, M.; Yamashita, I. Simple method for simultaneous dermination of chloro-phyll and Carotenoids in tomato fruit. J. Jap. Soci. Food Sci. Tehnol. 1992, 39, 925-928.

14. Singleton, V. L.; Rossi, J. A. Colorimetry of total phenolics with phosphomolybdic-phosphotungstic acid reagents. American Journal of Enology and Viticulture, 1965, 16, 144158.

15. Espin, J. C.; Soler-Rivas, C.; Wichers, H. J. Characterization of total free radical sca-venger capacity of vegetable oils and oil fractions using 2,2-diphenyl-1picrylhydrazyl radical. Journal of Agricultural Food Chemistry, 2000, 48, 648-656.

16. Oyaizu, M. Studies on product of browning reaction prepared from glucose amine. Japanese Journal of Nutrition, 1986, 44, 307-315.

17. Indrawati, R.; Sukowijoyo H.; Dumilah esti Wjayantu, R.; Limantara, L. Encapsu-lation of Brown Seaweed Pigment by Freeze Drying: characterization and Its Stability During Storage. Procedia Chemistry, 2015, 14, 353-360.

18. Sun, T.; Simon, P.W.; Tanumihardjo, S.A. Antioxidant phytochemicals and antioxi-dant capacity of biofortified carrots (Daucus carota L.) of various colors. J Agric Food Chem. 2009, 57, 4142-4147.

19. Arscott, S.A.; Tanumihardjo, S.A. Carrots of many colors provide basic nutrition and bioavailable phytochemicals acting as a functional food. Comprehensive Reviews in Food Science and Food Safety, 2010, 9, 223-239.



272

20. Yen, Y. H.; Shih, C. H.; Chang, C. H. Effect of adding ascorbic acid and glucose on the antioxidative properties during storage of dried carrot. Food Chem. 2008, 265-272.

21. Dutta, D.; Raychaudhuri, U.; Chakraborty, R. Retention of β-carotene in frozen carrots under varying conditions of temperature and time of storage. Afric J.of Biot. 2005, 4 (1), 102-103.

22. Zhao, M.; Yang, B.,; Wang , J.; Li, B.; Jiang, Y. Identification of the major flavonoids from pericarp tissues of lychee fruit in relation to their antioxidant activities. Food Chem. 2006, 98, 539-544.

23. Bystricka, J.; Kavalcova, P.; Musilova, J.; Vollmannova, A.; Toth, T.; Lenkova, M. Carrot (Daucus carota L. ssp. sativus (Hoffm.) Arcang.) as source of antioxidants. Acta Agriculturae Slovenica, 2015, 105, 303-311.

24. Cefola, M.; Pace, B.; Signore, A.; Serio, F. Compositional analysis and antioxidant profile of yellow, orange and purple polignano carrots. Ital. J. Food Sci. 2012, 24, 285-291.

25. Grimme, L. H.; Brown, J.S. Functions of chlorophylls and carotenoids in thylakoid membranes. Adv. Photosyn. Res. 1984, 2, 141-144.

26. Jafari, S. M.; Assadpoor, E.; He, Y.; Bhandari, B. Encapsulation efficiency of food flavours and oils during spray drying. Drying Technology, 2008, 26 (7), 816-835.

ЕКСТРАКЦИЈА И ИНКАПСУЛАЦИЈА БИОАКТИВНИХ КОМПОНЕНТИ ИЗ ШАРГАРЕПЕ

Вања Н. Шерегељ*, Гордана С. Ћетковић, Јасна М. Чанадановић-Брунет,

Весна Т. Тумбас Шапоњац, Јелена Ј. Вулић, Слађана С. Стајчић


Шаргарепа је коренасто поврће изузетно богато биоактивним једињењима, као што су каротеноиди, полифеноли и друге функционалне компоненте које допри-носе здрављу. Употреба каротеноида је ограничена због њихове нестабилности. Инкапсулација је једна од алтернативних метода која се користи за побољшање њи-хове стабилности. Циљ овог истраживања била је оптимизација екстракције и ин-капсулације каротеноида из шаргарепе. Лиофилизирана шаргарепа екстрахована је конвенционалном методом екстракције (CSE) користећи раствараче различите по-ларности (етанол, ацетон, етил ацетат и хексан). Иако је CSE са хексаном и етил ацетатом резултирала највећим садржајем каротеноида (18,27 и 15,73 mg β-каро-тена/100 g), ацетонски и етанолни екстракти (1:10 m/V) са нешто нижим садржајем каротеноида (14,52 и 11,45 mg β-каротена/100 g), изабрани су за даља истраживања због здравствене прихватљивости и већег садржаја полифенола (88,86 и 66,21 mg β-каротена/100 g). Етанолни и ацетонски екстракти шаргарепе инкапсулирани су ме-тодом лиофилизације на различите носаче, у циљу добијања оптималног инкапсу-лата са највећом ефикасношћу инкапсулације каротеноида. Ефикасност инкапсула-



273

ције етанолног и ацетонског екстракта, примењујући малтодекстрин, протеине су-рутке и соје као носаче, била је у распону од 41,95% до 100%. Садржај β-каротена у инкапсулатима тестиран је током 2 месеца складиштења у светлим и тамним усло-вима. Генерално, ретенција каротеноида је значајно већа у мрачним условима скла-диштења, при чему је максимална ретенција (65,94-87,32%) утврђена у узорцима са малтодекстрином и протеинима соје. Кључне речи: биоактивна једињења, каротеноиди, екстракција, инкапсулација,

складиштење

Received: 15 September 2017. Accepted: 01 November 2017.

APTEFF, 48, 1-323 (2017) UDC: 637.524:635.7 https://doi.org/10.2298/APT1748275T BIBLID: 1450-7188 (2017) 48, 275-284


275

THE QUALITY DIFFERENCE BETWEEN FRANKFURTERS SEASONED WITH CONVENTIONAL AND ORGANIC SPICES

Igor B. Tomašević1*, Ilija V. Đekić2, Milica G. Aćimović3, Slaviša B. Stajić1,

Vladimir M. Tomović4

1 University of Belgrade, Faculty of Agriculture, Department of Animal Source Food Technology,

Nemanjina 6, 11080 Belgrade, Serbia 2 University of Belgrade, Faculty of Agriculture, Food Safety and Quality Management Department, Nemanjina

6, 11080 Belgrade, Serbia 3 University of Novi Sad, Institute of Food Technology, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia


Frankfurters seasoned with conventional and organic garlic or coriander were investigated for the differences in taste and odor intensities, instrumental color and oxidative stability during 35 days of cold storage. Garlic powder, both organic (0.78 mg MDA/kg) and conventional (0.71 mg MDA/kg), promoted lipid oxidation compared to control frankfurters without garlic (0.52 mg MDA/kg). Consumers assessed odor (6.68) and taste (6.18) intensities of frankfurters with organic garlic significantly higher compared to the odor (4.73) and taste (4.65) of the frankfurters with conventional garlic. Color of frankfurters with organic or conventional spices was also significantly different. At the 95% confidence level, at least 20% of the consumers could distinguish between the samples with organic and conventional garlic, and at least 14% of the consumers between the samples with organic and conventional coriander. For the first time it is suggested that seasoning with organic instead of conventionally produced spices might improve quality characteristics of meat products. KEY WORDS: spices, sausages, sensory evaluation, lipid oxidation

INTRODUCTION Edible spices serve many functions in food products. Their primary functions are to flavor food and provide aroma, texture, and color. Every spice or flavoring contains predominating chemical components that create these sensual qualities. The taste of a spice such as sweet, spicy, sour, or salty, is due to the presence of esters, phenols, acids, alcohols, chlorides, alkaloids, or sugars. The ratio of volatiles to non-volatiles varies among spices, causing flavor similarities and differences within a genus and even within a variety. They vary depending upon the species of spice, its source and environmental growing conditions (1). Quality considerations for organic spices are the most important, * Corresponding author: Igor B. Tomašević, University of Belgrade, Faculty of Agriculture, Department of

Animal Source Food Technology, Nemanjina 6, 11080 Belgrade, Serbia e-mail: [email protected]



276

and they should be superior quality-wise in respect of inherent biochemical constituents (2). The so-called 'secondary metabolites' make a major contribution to the specific odors, tastes and colors of (spice) plants (3). There is ample, but circumstantial, evidence that, on average, organic vegetables most likely contain more of these compounds than con-ventional ones (4). These findings make it only reasonable to assume that the application of organic spices in meat products would impact their color and taste significantly compared to the meat products made with conventional spices. Garlic (Allium sativum L.) is one of the most commonly used ingredients as a flavor enhancer in meat products. Coriander (Coriandrumsativum L.) is sweet and aromatic (rose like) spice used in some smoked sausages. The flavor of frankfurters is typically due to black pepper, nutmeg, and possibly, to coriander and garlic (5). From a technological standpoint, the use of garlic and coriander and their derivatives in meat and meat products has been proposed by many authors due to their antioxidant activity (6-8). Hence, the aim of the present study was to evaluate the possible differences in lipid oxidation, color and sensory properties of frankfurters made with organic and conven-tionally grown garlic and coriander.

EXPERIMENTAL

Frankfurter samples preparation The raw materials: post-rigor pork (mixture of round and shoulder muscles) and fresh back fat were purchased from a major local retailer. The meat was trimmed of visible fat and connective tissue. Frankfurters were manufactured in a pilot meat processing plant of the Faculty of Agriculture, University of Belgrade. Five different batters, first without (control) and four with (single) spices, were produced all containing 50% of pork meat, 25% of back fat and 25 % of water (ice). This was performed in duplicate using meat batters of 4 kg. All batters were produced on the same day and in identical manner: refrigerated meat and fat were chopped to 8 mm particle size in a meat grinder (Laska 82H, Austria) and then mixed with ice, nitrite-salt (1.8%), polyphosphate (0.4%), soy protein isolate (1%), sodium erythorbate (0.04%) and sugar (0.5%) in a meat cutter (Müller EMS, Germany) until the temperature reached 7 °C. Finally, 3% of the single selected type of spice was added to the mixture and the comminution process was conti-nued while ensuring that the maximum batter temperature did not exceed 12 °C. Four dif-ferent spices used in our investigation were: garlic powder (Allium sativum L.) conven-tionally produced (GC) and from organic origin (GO); ground coriander (Coriandrum sativum L.) conventionally produced (CC) and from organic origin (CO). All spices were obtained from the same producer (Lay Gewürze, Queienfeld, Germany) declaring their accordance with organic Regulation (EC) No 834/2007 where appropriate. After the emulsification, the prepared batter was stuffed into 24 mm diameter collagen casings, after which they were hanged, smoked and cooked for approximately 2 hours in a smoking/cooking chamber (Belje, Croatia), until the temperature in the central



277

part of the sausages reached 72 °C/10 min. The cooked sausages were showered in cold water and stored at 4 ± 1 °C for a period of 35 days.

Methods

Lipid oxidation (TBARS) measurement. Malondialdehyde (MDA), the compound used as an index of lipid peroxidation, was determined by a selective third-order deriva-tive spectrophotometric method according to Botsoglou et al. (9), with slight modifi-cations. In brief, 2 grams of samples with 10 ml of 5% aqueous solution of trichloroacetic acid (Merck, Darmstadt, Germany) and 5 ml of 0.8% butylated hydroxytoluene (Sigma Chemical Co, St. Louis, MO) in hexane, were briefly vortexed (CZ Classic, VelpScien-tifica, Italy) and then homogenized in an ultrasonic bathroom (ATM40-3LCD, Madrid, Spain) for 5 min before being centrifuged (Eppendorf Centrifuge 5804 R, Hamburg, Germany) at 3,000 rpm for 5 minutes. The top layer was discarded, and a 2.5-ml aliquot from the bottom layer was mixed with 1.5 ml of 0.8% aqueous 2-thiobarbituric acid (Sigma Chemical Co, St. Louis, MO) to be further incubated at 70 °C for 30 min in water bath (Memmert MB14, Germany). Following incubation, the mixture was cooled under tap water and submitted to conventional spectrophotometry (Shimadzu, Model UV-160A, Tokyo, Japan) in the range of 400-650 nm. The concentration of MDA in the analyzed samples was calculated as described by Botsoglou et al. (9). The results are expressed as mg MDA/kg frankfurter. Color. Color of frankfurters was evaluated using a computer vision system (CVS) previously described by Girolami et al. (10) with a modifications explained below. A Sony Alpha DSLR-A200 digital camera featuring a 10.2 Megapixel CCD sensor was used for image acquisition. The camera was located vertically at a distance of 30 cm from the sample. The camera setting was the following: shutter speed 1/6 s, manual operation mode, aperture Av F/11.0, ISO velocity 100, flash off, focal distance 30 mm, lens DT-S18-70 mm f 3.5-5.6. Lighting was achieved with four fluorescent lamps (Philips Master Graphica TLD 965), with a color temperature of 6500 K and a color rendering index (Ra) close to 98%. Light diffusers covered each lamp. The camera was calibrated with a 24 color chart X-Rite Colorchecker; the CCD sensor was adjusted using the standard color rendition chart Colorchecker Passport (Michigan, USA). The Colorchecker was photographed using the implemented CVS to obtain the input device RGB signals in the theoretical range of 0-255 (the RGB values are expressed as sRGB D65 and CIE Lab D50 2° observer). The camera was connected to a Toshiba Portege R830 PC equipped with a 22″ LG IPS LED external monitor. The monitor with an sRGB gamut (Standard RGB) was calibrated with X-Rite i1 Display Pro device by selecting white chromaticity at 6500 K (illuminant D65), gamma at 2.2 and white luminance at 140 cd/m2, and the i1Profiler 1.5.6 software was used to create the ICC monitor profile. The Adobe Photoshop CC (64 bit) software was used for image analysis. The L*, a*, and b* values were measured on the digital image of the sample visualized on the monitor by pointing the cursor at the centre of the area (11x11 pixels) to be



278

evaluated by clicking on it. The L*, a* and b* values from RGB images were measured from RAW photographs. Total color difference (ΔE*) was calculated according to the formula:

ΔЕ*= (Lc*-Lo

*)2+(ao

*-ac* )

2+(bo

*-bc* )

2

where the subscript (c) denotes the values for frankfurters with conventional and (o) denotes the values for frankfurters with organic spices. Triangle test. Seventy-five consumers (38 male and 37 female) participated in all the triangle tests. For one participant, the odd product was the same for the two replications and the six triads (AAB, ABA, BAA, BBA, ABB and BAB) were counterbalanced over participants at each replication (11). The sausages were cooled to 21 °C, cut into quarters (length×diameter: 10×25 mm) and served to the consumers. The tests were carried in the meat laboratory under daily light. The degree of difference between the sample assessor choice and the others was expressed by circling one of the following descriptors which most closely describes the intensity of difference („0“ = none; „1“ - very slight; „2“ - slight; „3“ - moderate; „4“ - large; „5“ - extreme). Consumer preference. Sixty consumers (24 male and 36 female, age ≤25) were invited to evaluate intensities of odor and taste of garlic and coriander in sausages. Samples were served in a sequential order, with a two-minute mandatory break between each sample. Participants rinsed their mouths with room temperature mineral water between samples to reduce potential carry-over effects (12). To evaluate taste and odor a 9-point intensity scale was used with two descriptive anchors, (1 = ‘‘very low intensity”) and (9 = ‘‘very high intensity”) in order to signify increasing sensation intensity (13). The data obtained from the 9-point scale were processed using t-test for sausages containing the same type of spice, garlic or coriander. The level of statistical significance was set at α = 0.05 (95% confidence). Statistical analysis. The data were analyzed using SPSS Statistics 17.0 (Chicago, Illinois, USA). A three-sample independent t-test was used to compare the pairs of means with P ≤0.05. When multiple effect comparison was made, factorial analysis of variance (ANOVA), and a post-hoc Tukey analysis of Honestly Significant Difference (HSD) were used. The data are presented as mean values of measurements along with standard deviations.


TBARS measurements

The antioxidative activity of garlic aroma (sulphur) compounds (allicin, diallyl disul-fide, and diallyltrisulfide) was previously confirmed (14). However, when it comes to the antioxidative activity of garlic powder in meat and meat products, limited data are available. Park et al. (15) examined the antioxidant effects of garlic powders (5%) in fresh pork belly and pork loin in the vacuum-packaged samples held at 8 °C for 28 days.



279

The authors found that garlic powders were effective at reducing TBARS values in fresh pork belly, but no significant reduction was found in the pork loin treated with garlic powder. Our results suggest that garlic powder, both conventional and organic, promoted lipid oxidation in frankfurters throughout the whole period of their shelf life. The MDA concentrations were significantly higher (P< 0.5) in samples with added garlic compared to control (Table 1). Significant difference in the lipid oxidation between the frankfurters with conventional and organic garlic could not be observed during the period of 35 days of cold storage. These results are in agreement with the findings of Mariutti et al. (16) suggesting that garlic presented no effect as antioxidant and accelerated lipid oxidation in chicken meat.

Table 1. Differences in lipid oxidation TBA-RS (mg MDA/kg) of frankfurters with conventional and organic spices during shelf life

Time (days)

Frankfurter 1 day 7 day 14 day 21 day 35 day

Control 0.28a(±0.07) 0.31a

(±0.01) 0.35a(±0.08) 0.47a,e

(±0.11) 0.52a(±0.07)

Garlic conventional

0.47b(±0.01) 0.50c,d

(±0.06) 0.56c,d(±0.06) 0.61c,d,e

(±0.05) 0.78b(±0.08)

Garlic organic 0.59b(±0.01) 0.62d

(±0.06) 0.65d(±0.07) 0.68d

(±0.05) 0.71b(±0.01)

Coriander conventional

0.31a(±0.12) 0.32 a,b

(±0.12) 0.39 a,b(±0.09) 0.42 a,b

(±0.06) 0.47a(±0.01)

Coriander organic

0.31a(±0.01) 0.39a,b,c

(±0.09) 0.47a,b,c(±0.01) 0.54a,b,c,d

(±0.07) 0.56a(±0.05)

Mean values of six replications ± standard deviation; Means within a same day lacking a common superscript letter are significantly different (P< 0.05).

Although it was previously reported that coriander significantly lowered MDA con-centrations in different types of meat products (6, 17, 18) our experiments could not confirm its antioxidant activity in frankfurters. From the first until the last (35th) day of the cold storage, MDA concentrations in control and frankfurters with conventional or organic coriander were not significantly different (P ˃ 0.5). During the same period, a significant difference in lipid oxidation between samples with conventional and organic coriander could not be observed (Table 1).

Color Although, according to Mokrzycki and Tatol (19), even unexperienced observer can notice the difference in color when total color difference is 2 < ΔE* < 3.5, it should be noted that when the difference is concentrated in one dimension, like it was the case with



280

conventional and organic garlic samples in our study, this color difference is pronounced even more. Therefore, despite the fact that the values for lightness (L*) and yellowness (b*) were not significantly different between frankfurters with conventional and organic garlic (Table 2) and the point that total color difference was only slightly greater than 2 throughout the examined period of shelf life, we should conclude that there was a clear difference in color between the samples. From the first till the last day of cold storage, the frankfurters with organic garlic were significantly greener due to the repeatedly lower redness (a*) values compared to the samples with conventional garlic.

Table 2. Differences in color of frankfurters with conventional and organic spices during shelf life

Time (days)

Frankfurter Color 1 day 7 day 14 day 21 day 35 day

Garlic conventional

(GC)

L* 73.65a (±1.13)

74.10a (±0.97)

73.95a (±1.14)

75.15a (±0.74)

76.10a (±1.11)

a* 9.25a (±0.79)

9.75a (±0.64)

9.50a (±0.55)

9.35a (±0.98)

9.25a (±0.59)

b* 5.95a (±0.89)

6.65a (±0.67)

6.10a (±0.85)

6.10a (±0.55)

6.45a (±0.87)

Garlic organic (GO)

L* 75.50b (±1.54)

76.35b (±1.26)

75.55b (±0.72)

77.25b (±0.72)

78.15b (±1.13)

a* 8.35b (±0.99)

8.30b (±0.47)

8.35b (±0.95)

8.55b (±0.35)

8.60b (±0.59)

b* 6.05a (±0.60)

5.70a (±0.97)

6.30a (±0.57)

6.15a (±0.43)

6.10a (±0.87)

ΔE*(GC-GO) 2.06 2.84 1.98 2.24 2.17

Coriander conventional

(CC)

L* 68.45c (±0.76)

69.30c (±1.41)

69.90c (±0.97)

70.25c (±0.65)

70.85c (±1.46)

a* 11.55c (±0.76)

12.55c (±0.89)

12.20c (±0.65)

11.96c (±0.85)

12.25c (±1.02)

b* 11.35C (±1.75)

11.15c (±1.08)

10.85c (±0.99)

11.15c (±1.19)

10.90c (±1.07)

Coriander organic (CO)

L* 67.85c (±1.95)

68.90c (±0.76)

69.00c (±0.85)

69.35c (±0.97)

70.45c (±1.46)

a* 11.95c (±1.23)

12.40c (±1.27)

12.10c (±0.55)

11.85c (±0.75)

12.15c (±0.98)

b* 12.00c (±2.31)

11.90c (±2.80)

11.45c (±0.95)

11.95c (±1.05)

11.30c (±1.26)

ΔE*(CC-CO) 0.97 0.86 1.08 1.20 0.58 Mean values of six replications ± standard deviation; Mean values within the same day denoted with the same small letter are not significantly different (P ˃ 0.5).

On the other hand, the total color difference (ΔE*) between frankfurters with conven-tional and organic coriander was in the range of 0.58 - 1.20 during the 35 days of cold storage (Table 2), signifying that the observer could not notice the color difference



281

between the samples, according to Mokrzycki and Tatol (19). This conclusion is sup-ported by the fact that all color components measured in our study, lightness (L*), redness (a*) or yellowness (b*), did not exhibited significant differences between the samples with conventional and organic coriander throughout the examined period of shelf life.

Triangle test The sensory analysis showed that the frankfurters with conventional and organic garlic can be distinguished as different (n = 75, nc = 42) at the level of α = 0.05. At the 95% confidence level, at least 20% of the consumers could distinguish between the samples. An average degree of difference for correct answers was 3.19 ±1.17, with no significant difference between the correct and incorrect answers (P ≥0.05). The sensory analysis also indicated that the samples with organic and conventional coriander can be distinguished as different (n = 75, nc = 39) at the level of α = 0.05. At the 95% confidence level, at least 14% of the consumers could distinguish between the samples. An average degree of difference for correct answers was 2.15 ±1.20, with no significant difference between the correct and incorrect answers (P ≥0.05). More consumers identified differences between the frankfurters with garlic than for frankfurters with coriander (42 compared to 39), and the overall intensity of difference was higher for “garlic” than for “coriander” samples (3.19 compared to 2.15). Various factors that affect the chemical composition of garlic are involved in the production process and, therefore, could be a useful means to enhance the quality and the bioactive properties of the final product. Cultivation practices, particularly the ferti-lization schedule, are important, since they not only contribute to covering the crop requirements in nutrients, but also could beneficially influence the chemical composition and quality of the final product (20). Farming systems (intensive, conventional and organic farming) have been pointed out to influence not only plant growth and yield, but also the quality of the garlic, in terms of its chemical composition and quality features, such as allicin content (21). Perhaps, this can be an explanation for the observed diffe-rences in sensory perception between the organic and conventionally grown garlic used in our experiment.

Consumer preference As can be seen from Table 3, the odor in organic spice is more intensive than the conventional one for both types of spices. This study confirmed that the sausages with organic garlic had statistically significantly higher odor intensity (6.68) compared to the sausages with conventional garlic (4.73), P <0.05. A similar situation was observed for the frankfurters with organic coriander, which received significantly higher (P <0.05) odor intensity scores (6.13) compared to the frankfurters with conventional coriander (4.73). Organic taste of garlic in sausages is more intensive (6.18) compared to the conventional one (4.65), with observed statistically significant difference (P <0.05). In contrast, the organic coriander had a less intensive, but not significantly different taste



282

(5.50) compared to the sausages with conventional coriander (5.95, P >0.05). This can be explained by the previously published observations that the oil yield in the mature stage of ripening of coriander was slightly lower than for coriander under conventional agriculture (22, 23).

Table 3. Consumer preference of the frankfurters with conventional and organic spices

Sample Odor1,2,3 Taste1,2,3

Garlic organic 6.68 ± 1.59a 6.18 ± 1.62a

Garlic conventional 4.73 ± 1.61b 4.65 ± 1.35b

Coriander organic 6.13 ± 1.68c 5.50 ± 1.95c

Coriander conventional 4.73 ± 2.18d 5.95 ± 1.95c 1 n = 60;2 The intensity scale ranking 1 = ‘‘very low intensity” and 9 = ‘‘very high intensity’;3 Items denoted with the same small letter are not significantly different (P ˃ 0.5).

CONCLUSIONS

This study reveals that garlic powder promotes lipid oxidation in frankfurters, while coriander powder does not exhibit this effect. The differences between conventional and organic spices in this matter could not be observed. On the other hand, when sensory attributes are concerned, it was clear that the frankfurters with conventional and organic spices can be distinguished as different by the consumers. Organic garlic contributed to more pronounced intensities of both odor and taste compared to conventional garlic, while for coriander the same was perceived by the consumers only for odor. It is also evident that there was a clear difference in color between the frankfurters with organic and conventional garlic. Based on the results of our study, it can be concluded that sea-soning of frankfurters with organic spices can contribute to a significantly different sensory profile of this type of meat products.

REFERENCES 1. Raghavan, S. Handbook of spices, seasonings, and flavorings. CRC Press/Taylor &

Francis: Boca Raton, FL, 2007. 2. George, C. K. Organic spices. In Handbook of Herbs and Spices, Woodhead

Publishing: 2001; pp 34-38. 3. Bennett, R. N.; Wallsgrove, R. M. Secondary metabolites in plant defence

mechanisms. New Phyto. 1994, 127 (4), 617-633. 4. Brandt, K.; Mølgaard, J. P. Organic agriculture: does it enhance or reduce the

nutritional value of plant foods? J Sci Food Agr. 2001, 81 (9), 924-931.



283

5. Knipe, C. L.Sausages, types of | Emulsion A2 - Dikeman, Michael. In Encyclopedia of Meat Sciences (Second Edition), Devine, C., Ed. Academic Press: Oxford, 2014; pp 256-260.

6. Marangoni, C.; Moura, N. F. d. Antioxidant activity of essential oil from Coriandrum Sativum L. in Italian salami. Food Sci. 2011, 31, 124-128.

7. Sallam, K.; Ishioroshi, M.; Samejima, K., Antioxidant and antimicrobial effects of garlic in chicken sausage. Lebensm Wiss Technol. 2004, 37 (8), 849-855.

8. Bhattacharyya, D., A comparative study on the antioxidant and antimicrobial pro-perties of garlic and coriander on chicken sausage. Int Meat Sci. 2011, 1 (2), 108-116.

9. Botsoglou, N. A.; Fletouris, D. J.; Papageorgiou, G. E.; Vassilopoulos, V. N.; Mantis, A. J.; Trakatellis, A. G. Rapid, Sensitive, and Specific Thiobarbituric Acid Method for Measuring Lipid Peroxidation in Animal Tissue, Food, and Feedstuff Samples. J Agric Food Chem. 1994, 42 (9), 1931-1937.

10. Girolami, A.; Napolitano, F.; Faraone, D.; Braghieri, A. Measurement of meat color using a computer vision system. Meat Sci. 2013, 93 (1), 111-118.

11. Sauvageot, F.; Herbreteau, V.; Berger, M.; Dacremont, C. A comparison between nine laboratories performing triangle tests. Food Qual Prefer. 2012, 24 (1), 1-7.

12. Li, B.; Hayes, J. E.; Ziegler, G. R. Just-about-right and ideal scaling provide similar insights into the influence of sensory attributes on liking. Food Qual Prefer. 2014, 37, 71-78.

13. Lawless, H. T.; Heymann, H., Sensory Evaluation of Food - principles and practices. Springer, New York, USA, 2010.

14. Kim, S. M.; Kubota, K.; Kobayashi, A. Antioxidative Activity of Sulfur-containing Flavor Compounds in Garlic. Biosci Biotechnol Biochem. 1997, 61 (9), 1482-1485.

15. Park, S. Y.; Yoo, S. S.; Shim, J. H.; Chin, K. B. Physicochemical properties, and antioxidant and antimicrobial effects of garlic and onion powder in fresh pork belly and loin during refrigerated storage. J Food Sci. 2008, 73 (8), C577-84.

16. Mariutti, L. R.; Nogueira, G. C.; Bragagnolo, N. Lipid and cholesterol oxidation in chicken meat are inhibited by sage but not by garlic. J Food Sci. 2011, 76 (6), C909-15.

17. Sancho, R. A.; de Lima, F. A.; Costa, G. G.; Mariutti, L. R.; Bragagnolo, N. Effect of annatto seed and coriander leaves as natural antioxidants in fish meatballs during frozen storage. J Food Sci.2011, 76 (6), C838-45.

18. Michalczyk, M.; Macura, R.; Tesarowicz, I.; Banaś, J. Effect of adding essential oils of coriander (Coriandrum sativum L.) and hyssop (Hyssopus officinalis L.) on the shelf life of ground beef. Meat Sci. 2012, 90 (3), 842-850.

19. Mokrzycki, W. S.; Tatol, M. Color difference ΔE: a survey. Mach Graph Vis. 2011, 20 (4), 383-411

20. Martins, N.; Petropoulos, S.; Ferreira, I. C. F. R., Chemical composition and bioactive compounds of garlic (Allium sativum L.) as affected by pre- and post-harvest con-ditions: A review. Food Chem. 2016, 211, 41-50.

21. Huchette, O.; Arnault, I.; Auger, J.; Bellamy, C.; Trueman, L.; Thomas, B.; Kahane, R. Genotype, nitrogen fertility and sulphur availability interact to affect flavour in garlic (Allium sativum L.). J Hortic Sci Biotechnol. 2007, 82 (1), 79-88.



284

22. Nguyen, Q.-H.; Talou, T.; Cerny, M.; Evon, P.; Merah, O. Oil and fatty acid accumulation during coriander (Coriandrum sativum L.) fruit ripening under organic cultivation. The Crop Jour. 2015, 3 (4), 366-369.

23. Msaada, K.; Hosni, K.; Ben Taarit, M.; Chahed, T.; Hammami, M.; Marzouk, B., Changes in fatty acid composition of coriander (Coriandrum sativum L.) fruit during maturation. Ind Crops Prod. 2009, 29 (2-3), 269-274.

ИСПИТИВАЊЕ РАЗЛИКА УТИЦАЈА КОНВЕНЦИОНАЛНИХ И ОРГАНСКИХ ВРСТА ЗАЧИНА НА КВАЛИТЕТ ВИРШЛИ

Игор Б. Томашевић1*, Илија В. Ђекић2, Милица Г. Аћимовић3, Славиша Б. Стајић1,

Владимир М. Томовић4

1 Универзитет у Београду, Пољопривредни факултет, Катедра за технологију анималних производа, Немањина 6, 11080 Београд, Србија

2 Универзитет у Београду, Пољопривредни факултет, Катедра за управљање квалитетом и безбедношћу у производњи хране, Немањина 6, 11080 Београд, Србија

3 Универзитет у Новом Саду, Научни институт за прехрамбене технологије, Булевар цара Лазара 1, 21000 Нови Сад, Србија

4 Унниверзитет у Новом Саду, Технолошки факултет, Нови Сад, Булевар цара Лазара 1, 21000 Нови Сад, Србија

Виршле су зачињене конвенционалним и органским белим луком, односно ко-ријандером и испитивана је разлика у њиховом укусу и мирису, боји и оксида-тивној стабилности током 35 дана чувања на хладном. Бели лук у праху, како ор-гански (0,78 mg MDA/kg) тако и конвенционални (0,71 mg MDA/kg), промовисао је оксидацију липида у односу на контролне узорке виршли (0,52 mg MDA/kg). По-трошачи су оценили интензитете мириса (6,68) и укуса (6,18) виршли са органским белим луком као значајно израженије у односу на мирис (4,73) и укус (4,65) вир-шли са конвенционалним белим луком. Боја виршли произведених са органским, односно конвенционалним зачинима такође је била значајно различита. Са нивоом сигурности од 95%, барем 20% потрошача било је у стању да уочи разлике у боји између узорака са органским и конвенционалним белим луком, а најмање 14% по-трошача уочило је разлике у боји између виршли са органским и конвенционалним коријандером. Резултати нашег истраживања сугеришу технолозима меса да би производња кобасица са органским уместо конвенционално узгајаним зачинима могла да доведе до побољшања њиховог квалитета. Kључне речи: зачини, кобасице, сензорни квалитет, оксидација липида

Received: 26 May 2017. Accepted:02 July 2017.

APTEFF, 48, 1-323 (2017) UDC: 66.067.3:66.048.6 https://doi.org/10.2298/APT1748285V BIBLID: 1450-7188 (2017) 48, 285-293


285

FLUX INTENSIFICATION DURING MICROFILTRATION OF DISTILLERY STILLAGE USING A KENICS STATIC MIXER

Vesna M. Vasić*, Aleksandar I. Jokić, Marina B. Šćiban, Jelena M. Prodanović,

Jelena M. Dodić, Dragana V. Kukić

* University of Novi Sad, Faculty of Technology Novi Sad, Bulevar Cara Lazara 1, 21000 Novi Sad, Serbia

The present work studies the effect of operating parameters (pH, feed flow rate, and transmembrane pressure) on microfiltration of distillery stillage. Experiments were con-ducted in the presence of a Kenics static mixer as a turbulence promoter, and its influen-ce on the flux improvement and specific energy consumption was examined. Response surface methodology was used to investigate the effect of selected factors on microfiltra-tion performances. The results showed that response surface methodology is an approp-riate model for mathematical presentation of the process. It was found that the use of a static mixer is justified at the feed flow rates higher than 100 L/h. In contrast, the use of a static mixer at low values of feed flow rate and transmembrane pressure has no justifi-cation from an economic point of view. KEY WORDS: microfiltration, response surface methodology, kenics static mixer, dis-

tillery stillage

INTRODUCTION Bioethanol is widely used today as an alternative energy source, and it is believed to be one of the dominating biofuels in the future, especially in the transport sector. Enlargement of bioethanol production increases the amount of wastewater, whose pollu-tion potential is one of the most serious problems today, especially in developing countries, where the waste flows are not treated and utilized in a suitable manner. So, the bioethanol industry is forced to develop the most suitable technique for wastewater treatment and utilization, which will achieve a specified water quality and minimize capital and operating costs. In the recent years, the use of membrane separation technologies has gained great importance in many processing industries such as biotechnology, pharmaceutical, chemi-cal, metallurgical industry, etc. This is especially true for the most commonly used sold pressure-driving membrane processes, microfiltration (MF) and ultrafiltration (UF), which are widely used in the food industry, biotechnology, drinking water purification, wastewater treatment, metallurgical, and petroleum industry (1). Water reuse and recyc-ling, wastewater treatment, drinking water production and environmental protection are * Corresponding author: Vesna M. Vasić, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




286

the key challenges for the future of our planet. Membrane separation technologies for the removal of total suspended solids and a fraction of dissolved solids from wastewaters are becoming more and more promising. Also, this processes are playing a major role in wastewater purification systems because of their high potential for recovery of water and other valuable products (2,3) from many industrial wastewaters. Considering the fact that membrane separation technologies have been investigated and optimized for years, there has been a great degree of advancement in the develop-ment of membrane processes. Despite all benefits that membrane processes have, there are some limitations for their use. A major drawback of membrane processes is fouling formation on the membrane surface and flux decline. Results obtained during microfiltra-tion of distillery stillage in the system without static mixer (4) showed a fast flux decrea-se during the first few minutes of filtration, and does not decrease significantly after-words. To overcome this problem, it is necessary to find a solution for reduction of the effect of fouling of the membrane. The most efficient solution is the usage of appropriate pretreatment that would eliminate most of the foulants from the feed solution (5,6), or controlling the hydrodynamic conditions of the feed using turbulent promoters (7,8). The cross-flow microfiltration is influenced by a great number of parameters such as crossflow velocity, transmembrane pressure (TMP), membrane resistance, layer resis-tance, size distribution of the suspended solids, etc. A lot of models have been developed to describe the processes of cross-flow microfiltration, and they can be termed as physi-cal and empirical models. However, only a few are sufficient to describe the real pro-cesses (physical models) and explain the different experimental and practical results (9). Empirical modeling is very precise and useful in practice, but it is not helpful for under-standing the processes of cross-flow microfiltration. In the recent years, the development of various computer programs, empirical models based on the concept of neural networks and response surface methodology (RSM) gain in importance for describing of microfil-tration and other membrane processes. RSM was introduced by Box and Wilson in 1951 (10,11), and it can be defined as an empirical statistical technique for multiple regression analysis of data obtained from pro-perly designed experiments. It is a collection of statistical and mathematical techniques useful for developing, improving and optimizing processes. The field of RSM consists of the experimental strategy for exploring the space of the process or independent variables, empirical statistical modeling to develop an appropriate approximating relationship bet-ween the yield and the process variables and optimization methods for finding the values of the process variables that produce desirable values of the response (12). The aim of this work was to investigate the improvement of the process of cross-flow microfiltration of distillery wastewater using Kenics static mixer as turbulence promoter. The parameters considered for design of experiments were TMP, feed flow rate, and pH. The RSM was used to investigate the effect of selected factors on microfiltration performances.



287

EXPERIMENTAL

Materials and methods Distillery stillage from the ethanol factory Reachem, Srbobran, Serbia, produced from starch feedstock, was used in the experiments. The investigations were carried out in a laboratory cross-flow microfiltration unit, according to the procedure described by Jokic et al. (7). The feed suspension was concentrated to a volume concentration factor of 1.88. The pH value was adjusted by adding NaOH solution (0.1 mol/L) to the stillage. The membrane was cleaned before each experiment according to the recommendation of the manufacturer; the cleaning sequence was a classical acid-base one. The static turbulence promoter used in the experiments was the stainless steel Kenics static mixer. Its length was 23 cm and the diameter of 6 mm. It consisted of a series of helical mixing elements made from thin, flat strips, twisted through 180º to form helics. The helics are turned around their main axis by 90 against the next element. For the experimental part of the work, the Box-Behnken design was selected. The design variables and their ranges were: X1: TMP (0.3; 0.6 and 0.9 bar); X2: feed flow rate (40; 100 and 160 L/h) and X3: pH (3; 6 and 9). For three variables with three replicates at the center point the total number of experiment was 15. The experimental design in the actual and the coded (in the brackets) levels of variables are summarized in Table 1. For each experiment, the change of permeate flux during the time was monitored. The efficiency of the static mixer as a turbulence promoter was determined as the improve-ment of permeation flux defined as a relative increase of the permeate flux during the use of static mixer. The effect of the use of the static mixer on energy consumption during “cross-flow” microfiltration can be expressed through the reduction of specific energy consumption (ER).


Results for the responses obtained after cross-flow microfiltration are presented in Table 1. The response function (Y) was the permeate flux. These values were related to the coded variables by second-degree polynomial using the method of least squares:

Y=b0+ bi Xi+ bii Xii2+ bij XiXj

where b0 represents the intercept (constant), bi the linear, bii the quadratic and bij the inte-raction effect of the factors, while Y represents the response. The adequacy of the model was evaluated by coefficient of determination (R2) and model p-value. The significance of regression coefficients was assessed by p-values at the 0.05 significance level. The analysis of variance (ANOVA) tables were generated and the effect of individual, linear, quadratic and interaction term were determined (Tables 2 and 3).



288

Table 1. The Box-Behnken experimental design and responses

Experiment number

Factors1 Responses2

Q (l/h) TMP (bar)

pH Jnsm(L/m2h)3 Jsm(L/m2h) FI (%) ER (%)

1 40 (-1) 0.3 (-1) 6 (0) 4.2 10.0 136.9 -60.5 2 160 (+1) 0.3 (-1) 6 (0) 29.2 63.4 117.0 8.5 3 40 (-1) 0.9 (+1) 6 (0) 7.7 18.7 141.8 -59.7 4 160 (+1) 0.9 (+1) 6 (0) 25.5 56.8 122.7 122.6 5 40 (-1) 0.6 (0) 3 (-1) 8.8 16.5 87.2 -21.9 6 160 (+1) 0.6 (0) 3 (-1) 28.3 51.3 81.2 81.2 7 40 (-1) 0.6 (0) 9 (+1) 7.3 21.1 189.5 -51.7 8 160 (+1) 0.6 (0) 9 (+1) 23.7 62.6 164.1 32.1 9 100 (0) 0.3 (-1) 3 (-1) 16.9 45.6 170.5 1.4 10 100 (0) 0.9 (+1) 3 (-1) 17.2 46.4 169.1 101.8 11 100 (0) 0.3 (-1) 9 (+1) 16.1 44.3 175.5 -54.1 12 100 (0) 0.9 (+1) 9 (+1) 15.7 49.1 212.6 -21.8 13 100 (0) 0.6 (0) 6 (0) 15.5 46.0 195.9 -1.4 14 100 (0) 0.6 (0) 6 (0) 14.3 46.0 221.7 7.2 15 100 (0) 0.6 (0) 6 (0) 16.2 46.0 183.9 -5.3

1 TMP- transmembrane pressure, Q - feed flow rate, 2 Jnsm- permeate flux without static mixer, Jsm- permeate flux with static mixer, FI- permeate flux improvement, ER- reduction of specific energy consumption 3 Results obtained after microfiltration without static mixer [5]

Table 2. Regression coefficients for the responses

Responses JNSM (L/m2h) JSM (L/m2h) FI (%) ER (%)

Effects Coeff. t-value Coeff. t-value Coeff. t-value Coeff. t-value b0 -0,645212 -0,09692 -19,9245 -1,2292 -85,631 -0,70718 -40,3201 -0,43007 b1 0,200481 3,76737 0,8965 6,9189 3,418 3,53165 -0,0121 -0,01613 b2 6,155248 0,52987 18,8696 0,6671 77,758 0,36802 72,6890 0,44433 b3 -0,946886 -0,72536 -1,7506 -0,5507 20,159 0,84902 -7,2433 -0,39400 b11 0,000260 1,29960 -0,0024 -4,9250 -0,017 -4,66756 0,0008 0,27802 b22 4,259259 0,53234 -1,5556 -0,0798 -108,097 -0,74278 -2,5587 -0,02271 b33 0,095132 1,03574 0,0657 0,2937 -1,123 -0,67202 0,8813 0,68130 b12 -0,100139 -2,60535 -0,2122 -2,2677 0,011 0,01643 1,5745 2,90878 b13 -0,004643 -1,12742 0,0099 0,9854 -0,029 -0,38433 -0,0288 -0,49654 b23 -0,217262 -0,26379 1,2143 0,6055 11,456 0,76469 -20,2812 -1,74852

Table 3. Analysis of variance (ANOVA) for the model responses

Response Source

Residual Model DF SS MS DF SS MS F p R2

FI 5 3167.10 633.42 9 22125.71 2458.41 3.88 0.07473 0.875

ER 5 1898.60 379.72 9 44676.44 4964.05 13.07 0.00564 0.959 DF- degree of freedom; SS - sum of squares, MS - mean squares, R2 - coefficient of determination.

As can be seen from presented results, the most significant linear effect on the flux improvement has the feed flow rate, while quadratic effects are the feed flow rate and



289

TMP. The most important interaction that influences flux improvement is between the TMP and the pH value. On the other hand, the most important linear effect on the re-duction of specific energy consumption had the TMP. The most significant interactions are between the feed flow rate and the TMP, and between the TMP and the pH value. This can be explained according to the t values obtained in the model.

Effect of the static mixer on the permeate flux improvement The flux improvement was in the range between of 81% and 222%, depending on experimental conditions. The effects of operating parameters on the flux improvement by applying the Kenics static mixer are given in Figure 1.

а) b)

c) Figure 1. Effects of the operating parameters on the flux improvement in the presence of

the static mixer: a) feed flow rate and TMP, b) pH and feed flow rate, c) pH and TMP.



290

Figure 1a shows the effect of the feed flow rate and TMP on flux improvement. The increase in the feed flow rate leads to an initial increase of the flux improvement and then this value starts to decrease with the further increase of the flow. Such behavior can be noticed at all values of TMP. This can be explained by the fact that at low feed flow rate values, turbulent flow occurs as a result of inserting static mixer into the membrane chan-nel. The turbulent flow is sufficient to reduce the cake thickness which reduces resistance to the permeate flow, leading to an increase of the flux compared with the system without static mixer. However, at higher values of feed flow rate, the flux improvement starts to decrease since at this flow values turbulence can occur even without the static mixer (13). This fact indicates that there is a range of moderate flows at which the flux improvement is the highest. In this case, these are the feed flow values in the range of 100 - 120 L/h, for all values of TMP. Similar observations are presented in Figure 1b, which shows the effect of the pH value and feed flow rate on the flux improvement. As can be seen, the values of flux improvement are larger at the high pH values. Figure 1c shows the effect of the pH and TMP. With the increase of TMP, there are no significant changes in the flux improvement at lower pH values, while at higher ones, the TMP increase leads to a slight increase of improvement of the permeation flux. This can be explained by the fact that the changes in the structure of the wastewater was observed after the change of its pH value (14), which leads to a change in the cake thickness and its resistance.

Effect of the static mixer on reduction of the specific energy consumption Specific energy consumption is one of the most important factors from the economic point of view. It is defined as the power dissipated per unit volume of permeate. Two op-posite phenomena influence this quantity in the presence of the static mixer. By inserting the static mixer into the membrane channel the pressure drop increases along the mem-brane because of the increased resistance to feed flow. This leads to increased drop in the hydraulic power. On the other hand, the presence of the static mixer leads to an increase in the feed flow rate due to the changes in fluid flow through the membrane. As a result of this opposing effects, the energy efficiency of the static mixer is limited. The appli-cation of the static mixer is justified by the reduction of the specific energy consumption compared to the system without a static mixer (4). The effects of operating parameters on the reduction of specific energy consumption as a result of the presence of the static mixer are shown in Figure 2. The effects of TMP and feed flow rate on the reduction of specific energy con-sumption are shown in Figure 2a. The increase in the feed flow rate is accompanied by increased reduction of the specific energy consumption at all applied TMP, except for the low pressures. With the increase in TMP at a minimal feed flow the reduction of specific energy consumption has a constant negative value, while its highest value reaches at the maximum values of TMP and feed flow rate. Figure 2b shows the effect of the pH value and feed flow rate on reduction of the specific energy consumption. As it can be seen, the increase in pH resulted in a decreased



291

reduction of the specific energy consumption at all feed flow rates. At low feed flow rates, these values are negative, while positive values of specific energy consumption can be noticed at the feed flow rates higher than 100 L/h.

a) b)

c)

Figure 2. Effects of operating parameters on the reduction of specific energy consumption as a result of the presence of the static mixer: a) feed flow rate

and TMP, b) pH and feed flow rate, c) pH and TMP The effects of the TMP and pH value are shown in Figure 2c. The increase in the TMP leads to an increase in the reduction of specific energy consumption at all observed pH values. The largest increase was recorded at the lowest pH value and maximum TMP, while at higher pH values this increase was less pronounced and the specific energy consumption had a negative value.



292

CONCLUSIONS Based on the results of the statistical analysis it can be concluded that the influence of operating parameters on permeate flux during microfiltration of distillery stillage can be adequately described using response surface methodology. The flux improvement was in the range between 81% and 222%. The obtained results showed that there is a range of applied feed flow rates at which flux improvement has the highest value, and this range is 100 - 120 L/h. Also, it was found that the use of the static mixer is justified at the feed flow rates higher than 100 L/h. In contrast, the use of the static mixer at low values of feed flow rate and TMP has no justification from an economic point of view.

Acknowledgement This research was supported by the grant number TR 31002 from the Ministry of Education, Science and Technological Development of the Republic of Serbia.

REFERENCES

29. Wang, Z.; Cui, Y.; Yao, J.; Chu, J.; Liang, Y. The influence of various operating conditions on specific cake resistance in the crossflow microfiltration of yeast sus-pensions. Desalin.Water. Treat. 2009, 1(1-3), 237-247.

30. De Souzaa, R.R.; Bergamascoa, R.; Da Costab, S.C.; Fengc, X.; Fariaa, S.H.B.; Gi-menesa, M.L. Recovery and purification of lactose from whey. Chem. Eng. Process. 2010, 49 (11), 1137-1143.

31. Cuartas-Uribe, B.; Alcaina-Miranda, M.I.; Soriano-Costa, E.; Mendoza-Roca, J.A.; Iborra-Clar, M.I.; Lora-García, J. A study of the separation of lactose from whey ultrafiltration permeate using nanofiltration. Desalination, 2009, 241 (1-3), 244-255.

32. Vasic, V.; Jokic, A.; Sciban, M.; Prodanovic, J.; Dodic, J. Crossflow microfiltration of distillery stillage: a response surface methodology approach. Environ. Eng. Manag. J. 2016, 15 (12), 2781-2788.

33. Schäfer, A.I.; Fare, A.G.; Waite, T.D. Cost factors and chemical pretreatment effects in membrane filtration of waters containing natural organic matter. Water Res. 2001, 35 (6), 1509-1517.

34. Soffer, Y.; Aim, R.B.; Adin, A. Membrane for water reuse: Effect of pre-coagulation on fouling and selectivity. Water Sci. Technol. 2000, 42 (1-2), 367-372.

35. Jokić, A.; Zavargo, Z.; Šereš, Z.; Tekić, M. The effect of turbulence promoter on cross-flow microfiltration of yeast suspension: A response surface methodology approach. J. Membr. Sci. 2010, 350 (1-2), 269-278.

36. Krstic, D.M.; Tekic, M.N.; Caric, M.D. The effect of turbulence promoter on cross-flow microfiltration of skim milk, J. Membr. Sci. 2002, 208 (1-2), 303-314.

37. Ripperger, S.; Altmann, J. Crossflow microfiltration-state of the art. Sep. Purif. Technol. 2002, 26 (1), 19-31.



293

38. Myers, R.H.; Montgomery, C.M. Response Surfaces Methodology: Process and Product Optimization Using Designed Experiments; Wiley, New York, 1995.

39. Lazic, Z.R. Design of Experiments in Chemical Engineering: A Practical Guide; Wiley-VCH, Weinheim, 2004.

40. Carley, K.M., Kamneva, N.Y.; Reminga, J. Response Surface Methodology, CASOS Technical Report, Carnegie Mellon University, Institute for Software Research Inter-national, 2004.

41. Krsti, D. Poboljšanje “cross-flow” mikrofiltracije upotrebom statičkog mešača kao promotora turbulencije, PhD Thesis, Tehnološki fakultet, Novi Sad, Serbia, in Ser-bian, 2003.

42. Vasić, V,; Šćiban, M.; Jokić, A.; Kukić, D.; Prodanović, J. Effect of pH value on the characteristics of distillery wastewater, 4th International congress: ”Engineering, eco-logy and materials in the processing industry, Jahorina, Republic of Srpska, Bosnia and Herzegovina, 04.-06. March 2015, Proceedings, 1129-1134.

ПОВЕЋАЊЕ ФЛУКСА ТОКОМ МИКРОФИЛТРАЦИЈЕ СКРОБНЕ ЏИБРЕ

ПРИМЕНОМ KENICS СТАТИЧКОГ МЕШАЧА

Весна М. Васић, Александар И. Јокић, Марина Б. Шћибан, Јелена М. Продановић, Јелена М. Додић, Драгана В. Кукић


Булевар цара Лазара 1, 21000 Нови Сад, Србија

У овом раду испитан је утицај радних параметара (рН, проток напојне суспензи-је и трансмембрански притисак) на флукс пермеата током микрофилтрације скроб-не џибре. Експерименти су извођени на лабораторијској апаратури за микрофил-трацију, у присуству Kenics статичког мешача као промотора турбуленције. Испи-тан је утицај истог на пораст флукса пермеата и специфичну потрошњу енергије. Поступак одзивне површине коришћен је за дефинисање утицаја процесних пара-метара на перформансе микрофилтрације. На основу резултата статистичке анализе утврђено је да се применом поступка одзивне површине на адекватан начин може описати утицај одабраних радних параметара на флукс пермеата током микрофил-трације џибре у систему са присуством статичког мешача. Примена статичког мешача оправдана је при вредностима протока већим од 100 L/h. Супротно томе, употреба статичког мешача при нижим вредностима протока и трансмембранског притиска нема оправданост са економског аспекта. Кључне речи: микрофилтрација, поступак одзивне површине, Kenics статички

мешач, џибра


APTEFF, 48, 1-323 (2017) UDC: 662.63:577.84 https://doi.org/10.2298/APT1748295V BIBLID: 1450-7188 (2017) 48, 295-305


295

DETERMINATION OF CULTURE MEDIUM COMPOSITION FOR MAXIMIZING THE BIOMASS PRODUCTION OF Pseudomonas stutzeri

Ana M. Vidaković*, Olja Lj. Šovljanski, Aleksandra S. Ranitović,

Dragoljub D. Cvetković, Siniša L. Markov


In the recent years, extensive studies have been conducted on the application of bioremediation in order to prevent and solve the global problem of nitrate accumulation especially in water and soil. In the studies related to the biological denitrification, Pseduomonas stutzeri is usually used as a model microorganism due to its ability of performing a complete reduction of nitrate. The aim of this work is to establish a culture medium at a laboratory level for maximizing the biomass production of Pseduomonas stutzeri ATCC 17588 and Pseudomonas stutzeri D1, previously isolated from the Danube (Novi Sad, Serbia) by using a Box-Bhenken experimental design and response surface methodology. The selected independent variables (+, 0, - levels) were: glucose (0, 1 and 2 g/L), KNO3 (1, 2 and 3 g/L), and peptone (0.4, 2 and 4 g/L), while the number of viable cells (log/mL) was chosen as a dependent factor (response). Based on the obtained results, it can be noticed that the maximum number of P. stutzeri ATCC 17588 viable cells is achieved when the initial content of glucose, KNO3 and peptone in the medium are 0 g/L, 2 g/L and 4 g/L, respectively. In the case of P. stutzeri D1, the most appropriate media for enhancing the biomass production consists of glucose (1 g/L), KNO3 (3 g/L) and peptone (4 g/L). Тhe obtained results can be used for further techno-economic analysis and scale up of the P. stutzeri biomass production. KEY WORDS: biomass, Pseudomonas stutzeri, Box-Bhenken design, response surface

methodology

INTRODUCTION

In the past decades, a widespread use of inorganic pesticides has been registered, influencing the accumulation of nitrate in soils and underground waters (1). Excessive levels of nitrate in drinking water could potentially cause human health problems such as blue-baby syndrome in infants (2) and stomach cancer in adults (1). Moreover, a high nitrate concentration in water is commonly related to eutrophication in the aquatic en-vironment (3). In order to prevent such occurrences, the maximum allowed concentration of nitrate in drinking water recommended by the World Health Organization is 10 mg/L (NO3

--N) (4).

* Corresponding author: Ana M. Vidaković, University of Novi Sad, Faculty of Technology Novi Sad, Bulevar




296

Considering that the accumulation of nitrate is a global problem, an intensive research in the field of removal of nitrate from water and soil has been carried out (5-6). It has been pointed out that one of the possible ways to remove nitrate from contaminated areas is the usage of high performance biological processes, such as denitrification thanks to its selectivity and posibility of total nitrate elimination (2). Biological denitrification is a process of reducing nitrate and nitrite to nitrogen oxides and molecular nitrogen by me-tabolic activities of microorganisms (7). In the studies related to biological denitrification processes Pseudomonas stutzeri is usually used as a model microorganisam, due to its ability to reduce nitrate up to molecular nitrogen (8). Although the field of biological denitrification by P. stutzeri has been briefly studied, there are no available scientific papers focusing on the composition of adequate media for enhancing the biomass production of P. stutzeri, which is important for industrial scale biomass production and application on wide contaminated areas. Previous studies related to P. stutzeri have dealt with the effect of C/N ratio, pH value, temperature, dissolved oxygen and other parameters on its denitrification efficiency (9-10). Moreover, it has been proven that P. stutzeri can be successfully used for wastewater treatments (99% effi-ciency in removal of 200 mg/L NO3) (6), as well as for biocleaning of cultural heritage materials (11-12). The most common modern method for the optimization of medium composition is response surface methodology (RSM) coupled with appropriate experimental design (Placket-Burmann, Box-Bhenken, etc.). RSM is a statistical technique for modeling and optimization of multiple variables, to determine the optimum process conditions (para-meters) by coupling experimental designs with interpolation by second-order polynomial equations in sequential testing procedure (13) Therefore, the aim of this work is to establish the appropriate media for maximizing the biomass production of Pseudomonas stutzeri ATCC 17588 and Pseudomonas stutzeri D1 (wild strain) on a laboratory level by using response surface methodology combined with the Box-Bhenken experimental design.

EXPERIMENTAL

Microorganisms

The bacteria strain Pseudomonas stutzeri ATCC 17588 was used as a reference cul-ture, and Pseduomonas stutzeri D1 as a wild strain, previously isolated from the Danube and identified by Vitek®2 Compact System (BioMérieux, Craponne, France) (14). Both microorganisms were stored in Nutrient Broth (HiMedia, Mumbai, India), with the addi-tion of glycerol (Lach-ner, Neratovice, Czech Republic) as a cryoprotectant and kept in the freezer for ultra-low temperatures (Snijders Labs, Tilburg, the Netherlands) at -80 °C.

Cultivation media The production of biomass was performed in the cultivation media whose content depends on the chosen experimental design. In the media for both testing microorganisms



297

the content (g/L) of glucose (Lach-ner, Neratovice, Czech Republic; purity ≥99.5%), KNO3 (Centrohem, Stara Pazova, Serbia; purity >99%), and peptone (HiMedia, Mumbai, India) was varied. The pH value of each medium was adjusted to 7.0 prior to sterilization at 1 bar/121 °C for 15 min.

Cultivation conditions The biomass production was performed in Erlenmeyer flasks (300 mL) containing 90 mL of the medium. The suspension of microorganism was prepared from a 24-hour cul-ture by removing the biomass from the agar slant and suspending it in 9 mL of sterile sa-line suspension to reach approximately the concentration of 3x108 CFU/mL, correspon-ding to McFarland No. 1 standard. Afterwards, the suspension was further diluted to the final concentration of 104 CFU/mL. The inoculation of the media was performed by ad-ding 10% (v/v) of the prepared bacterial suspension. The cultivation was carried out at room temperature for 24 h.

Determination of viable number of bacterial cells At the end of the process, 1mL of homogenized cultivation medium was separated. The separated aliquot was further diluted and a volume of 100 μL of each dilution was spread onto previously prepared sterile Plate Count Agar (HiMedia, Mumbai, India). After the incubation at 37 °C for 24 h, the colonies were counted, and the number of viable cells per milliliter was determined.

Experimental design and optimization by RSM The Box-Bhenken design with three factors at three levels and three repetitions in the central point and RSM were used to optimize the composition of the cultivation medium and determine the influence of the media ingredients on the biomass production. A total of 15 runs, including the replicated center points, were performed. The content (g/L) of glucose (X1), KNO3 (X2) and peptone (X3) were setup as independent factors, while the number of viable cells (Y), as an indirect criterion of biomass yield, was chosen as a dependent factor (response). The three levels of each independent factor were coded as -1, 0 and +1, corresponding to the lower, middle and higher values, respectively. The complete experimental design is given in Table 1. The relations between the independent factors and the response were calculated by the second-order polynomial equation (15):

Y=b0+ ∑ biXi+ ∑ bii

2Xii2 ∑ bijXiXj [1]

where Y is the defined response; b0 is the intercept, bi, bii and bij are the linear, quadratic and interaction regression coefficient, respectively, while Xi and Xj are the varied factors.



298

Table 1. The Box-Bhenken experimental design and varied values of factors

Experiment Coded factor level Varied factor value (g/L)

X1 X2 X3 Glucose KNO3 Peptone

1 -1 -1 0 0 1 2

2 1 -1 0 2 1 2

3 -1 1 0 0 3 2

4 1 1 0 2 3 2

5 -1 0 -1 0 2 0.4

6 1 0 -1 2 2 0.4

7 -1 0 1 0 2 4

8 1 0 1 2 2 4

9 0 -1 -1 1 1 0.4

10 0 1 -1 1 3 0.4

11 0 -1 1 1 1 4

12 0 1 1 1 3 4

13 0 0 0 1 2 2

14 0 0 0 1 2 2

15 0 0 0 1 2 2

Statistical analyses of the experimental results were performed using Statistica software v. 13.2 (Dell, Round Rock, Texas, USA). The significance of the influence of each examined factor, as well as their interaction, was examined by comparing the t-values for each of the coefficients in the regression equation [1]. The response surface plots were created with the same software and drawn for a constant value of one factor and varied values of the other factors.

RESULTS AND DISSCUSION As a general rule, each organism has its own nutritional requirement for maximum biomass production (15). To our knowledge, no medium has yet been established for biomass production of P. stutzeri. In this study, the media for maximizing the biomass production of selected denitrifying bacteria (reference and wild strain of P. stutzeri) were formulated by varying contents of KNO3, peptone, and glucose. The basis for the selection of culture media ingredients was the composition of the Nitrate Broth (DifcoTM Nitrate Broth, Becton, Dickinson and Company, Le Pont-de-Claix, France). This medium is recommended for the detection of nitrate reduction with the intention of differentiating and identifying various types of bacteria. Two main components of the Nitrate Broth, peptone as a source of proteins and KNO3 as a source of nitrate, were chosen as a part of



299

the new medium for the biomass production of the P. stutzeri. The content (g/L) of peptone and KNO3 in Nitrate Broth is 4 and 3, respectively. In order to reduce cost of the new culture medium, the content of peptone and KNO3 was varied as it is shown in Table 1. Although denitrifying bacteria usually use nitrates as a terminal electron acceptor in the respiratory pathway, there are indications that the addition of external carbon source with the role of an electron acceptor and source of energy such as glucose, may accelerate the denitrification process and bacterial growth (9, 16). Based on the previously men-tioned positive results of carbon addition on the denitrifying bacteria growth, glucose (common carbon source) was chosen as a third component in the new medium, and its content was varied through the experiment as it is presented in Table 1. In order to achieve the goal of this study, the experiments formulated by the Box-Bhenken design were conducted and the obtained results are given in Table 2.

Table 2. Number of viable cells (log/mL) of bacteria P. stutzeri in dependence of the medium composition formulated by the Box-Bhenken experimental design

Experiment P. stutzeri ATCC 17588 P. stutzeri D1

1 7.8 7.2

2 7.5 6.2

3 7.6 7.6

4 5.9 6.2

5 7.2 6.0

6 5.6 6.2

7 8.5 8.5

8 8.5 7.5

9 5.8 7.7

10 5.3 6.5

11 8.4 8.0

12 8.4 8.6

13 7.0 6.9

14 7.0 6.3

15 6.6 6.7 For the defined response presented in Table 2 (number of viable cells (log/mL)), a polynomial model of the second-order [1] was established to evaluate and quantify the influence of the examined factors. The t-value is used to determine the significance of the regression coefficients. The coefficients of the regression equation and t-values at a confidence level of 95% are given in Table 3.



300

Table 3. Regression equation coefficients and t-values

Microorganism P. stutzeri ATCC 17588 P. stutzeri D1

Effects Coefficient t-value Coefficient t-value

Intercept

b0 6.76588 7.18236 8.94928 7.44287

Linear

b1 -1.05680 -1.90973 0.62697 0.88764

b2 0.22577 0.28982 -2.08125 -2.09311

b3 0.23200 0.68347 -0.68468 -1.58023

Quadratic

b11 0.42917 2.37809 -0.24292 -1.05456

b22 -0.08333 -0.46176 0.41458 1.79980

b33 0.01784 0.31541 0.17708 2.45305

Interaction

b12 -0.35250 -2.03303 -0.10000 -0.45185

b13 0.21296 2.21764 -0.16066 -1.31072

b23 0.08098 0.84329 0.24663 2.01206

As it can be seen from Table 3, the coefficients of the regression equation may have positive and negative values. When a synergistic effect between the examined factors exists, then the coefficient of regression equation has a positive sign for the value. On the other hand, a negative sign represents an antagonistic effect of the varied factors. In order to understand the interactions of the examined factors and define the optimal contents of the media ingredients, three-dimensional response surface plots were generated. Each plot represents the effect of two factors on the response (number of viable cells), whereas the third factor was maintained at the central value from the Box-Bhenken experimental design. The effects of the initial content of glucose and KNO3 on the number of cells of P. stutzeri (reference and wild strain) at the constant level of peptone (2 g/L) are shown in Figure 1a-b. Based on the obtained results (Figure 1a) it is evident that, in the case of P. stutzeri ATCC 17588, the maximum number of viable cells (about 8 log/mL) is achieved in the medium without glucose (0 g/L) and at any concentration of KNO3. In the case of P. stutzeri D1, the maximum number of viable cells (about 7.5 log/mL) is observed for the lowest (1 g/L) and highest (3 g/L) content of KNO3, when the content of glucose is in the range of 0-1 g/L.



301

a) b) Figure 1. The effects of the initial glucose and KNO3 content on the number of cells of

a) P. stutzeri ATCC 17588 and b) P. stutzeri D1 at a constant peptone concentration (2 g/L)

The influence of the initial peptone and glucose concentration on the number of cells of P. stutzeri (reference and wild strain) at constant level of KNO3 (1 g/L) is illustrated in Figures 2a-b. From the response surface plot shown in Figure 2a, it can be seen that, in the applied experimental conditions for P. stutzeri ATCC 17588, the number of viable cells increased with the increasing initial peptone content almost independently of the glucose content. On the other hand, the highest number of viable P. stutzeri D1 cells (about 8.5) is detected in the medium with a maximum content of peptone (4 g/L) and glucose in the range of 0-1g/L (Figure 2b).

a) b) Figure 2. The effects of the initial peptone and glucose content on the number of cells

of a) P. stutzeri ATCC 17588 and b) P. stutzeri D1 at a constant KNO3 concentration (2 g/L)



302

The effects of the initial content of peptone and KNO3 on the number of viable cells of the tested microorganisms at a constant glucose concentration (1g/L) are presented in Figures 3a-b. For the achievement of the maximum number of viable cells of the both examined microorganisms the highest content of peptone in cultivation medium is required. Also, it can be noticed that, in the case of P. stutzeri ATCC 17588, the number of viable cells does not depend on the KNO3 content (Figure 3a). On the contrary, the maximum number of the viable P. stutzeri D1 cells is achieved only when the highest concentration of KNO3 (3 g/L) is present in the medium.

a) b) Figure 3. The effects of the initial peptone and KNO3 content on the number of cells of

a) P. stutzeri ATCC 17588 and b) P. stutzeri D1 at a constant glucose concentration (1 g/L)

According to the results presented in Figure 1-3 it can be concluded that the maxi-mum production of P. stutzeri ATCC 17588 biomass (8.5 log/mL) is achieved in a me-dium which contains glucose, KNO3 and peptone in the following concentrations (g/L) 0, 2, and 4, respectively. The optimal medium for the P. stutzeri D1 biomass production consists of glucose (1 g/L), KNO3 (3 g/L) and peptone (4 g/L). Namely, it can be noticed that glucose in the media stimulates the biomass production of the wild strain, while, in the case of the reference strain of P. stutzeri, glucose does not have an impact on the maximization of the biomass production. To our knowledge, these are the first results related to the adequate media compo-sition for maximizing the P. stutzeri biomass production. Тhe obtained results may further be used for the scaling-up of the P. stutzeri biomass production (pilot and indus-trial production). The production of P. stutzeri biomass on an industrial scale will have an enormous impact on the bioremediation processes, that will lead to the removal of nitrates from large contaminated surfaces. In this way, the global problem of nitrate generation would be reduced, which in the long term would have a positive effect on the protection of the environment and human health.



303

CONCLUSION The response surface methodology coupled with the Box-Bhenken design was succes-sfully employed to determine the most suitable culture media for maximizing the biomass production of Pseudomonas stutzeri ATCC 17588 and Pseudomonas stutzeri D1 (strain isolated from Danube). Based on the obtained results, the most convenient culture me-dium for biomass production consists of: KNO3 (2 g/L) and peptone (4 g/L) in the case of P. stutzeri ATCC 17588 and glucose (1 g/L), KNO3 (3 g/L) and peptone (4 g/L) in the case of P. stutzeri D1. In the future work it is necessary to relate the data presented in this work with the results of the optimization of the bioprocess parameters to define the optimum conditions of P. stutzeri biomass production at the pilot and industrial scale.

Acknowledgement The financial support of the Ministry of Education, Science and Technological De-velopment of the Republic of Serbia (Contract No. III45008) is gratefully acknowledged.

REFERENCES

1. Wang, J. and Chu, L.: Biological nitrate removal from water and wastewater by solid-phase denitrification process. Biotechnol .Adv. 2016, 34,1103-1112.

2. Gomez, M.A., Gonzalez-Lopez, J. and Hontoria-Garcia, E.: Influence of carbon source on nitrate removal of contaminated groundwater in a denitrifying submerged filter. J. Hazard. Mater. 2000, B80, 69-80.

3. Wu, Y., Wang, Y., Wang, J., Xu, S., Yu, L., Philippe, C. and Wintgens T.: Nitrate re-moval from water by new polymeric adsorbeent modified with amino and quaternary ammonium groups: Batch and column adsorption study, J. Taiwan. Inst. Chem. E. 2016, 66, 191-199.

4. Fu, F., Dionysiou, D.D. and Liu, H.: The use of zero-valent iron for groundwater remediation and wastewater treatment: A review. J. Hazard. Mater. 2014, 267, 194-205.

5. Jibin, Z., Pengxia, W., Hao, B. and Yu, Z.: Heterotrophic nitrification and aerobic denitrification by bacterium Pseudomonas stutzeri YZN-001, Bioresource. Technol. 2011, 102, 9866-9869.

6. Rezaee, A., Godini, H., Dehestani, S. and Kavani S.: Isolation and characterization of a novel denitrifying bacterium with high nitrate removal: Pseudomonas stutzeri. Iran. J. Environ. Health. Sci. Eng. 2010, 7, 313-318.

7. Tsukuda, S., Christianson, L., Kolb, A., Saito, K. and Summerfelt, S.: Heterotrophic denitrification of aquaculture effluent using fluidized sand biofilters. Aquacult. Eng. 2014, 64, 49-59.

8. Lalucat, J., Bennasar, A., Bosch, R., Garcia-Valdes, E. and Palleroni, N.: Biology of Pseudomonas stutzeri. Microbiol. Mol. Biol. R. 2006, 70, 510-547.



304

9. Yang, X., Wang, S. and Zhou, L.: Effect of carbon source, C/N ratio, nitrate and dissolved oxygen concentration on nitrite and ammonium production from denitrifi-cation process by Pseudomonas stutzeri D6. Bioresource. Technol. 2012, 104, 65-72.

10. Hunag, T., Guo, L., Zhang, H., Su, J., Wen, G. and Zhang, K.: Nitrogen-removal efficiency of a novel aerobic denitrifying bacterium, Pseudomonas stutzeri strain ZF31, isolated from a drinking-water reservoir. Bioresource. Technol. 2015, 196, 209-2016.

11. Bosch-Roig, P., Regidor Ros, J.L. and Montes Estrellés, R.: Biocleaning of nitrate alterations on wall paintings by Pseudomonas stutzeri. Int. Biodeter. Biodegr. 2013, 84, 266-274.

12. Vučetić, S., Ranogajec, J., Markov S., Vidaković, A., Hiršenberger, H. and Bera, O.: Development and modeling of the effective bioactive poultices for reducing the nitrate content in building materials. Constr. Build. Mater. 2017, 142, 506-513.

13. Ferreira, S.L.C., Bruns, R.E., Ferreira H.S., Matos, G.D., David, J.M., Brandão, G.C., da Silva E.G.P., Portugal, L.A., dos Reis, P.S., Souza, A.S. and dos Santos, W.N.L.: Box-Bhenken design: An alternative for the optimization of analytical methods. Anal. Chim. Acta 2007, 597, 179-186.

14. Vidaković, A., Šovljanski, O., Velićanski, A. and Markov, S.: Isolation of denitrifying bacteria for bio-cleaning of nitrate alterations on cultural heritage materials, Inter-national Conference On Science and Technique Based on Applied and Fundamental Research, Szeged, 2 June 2016, Book of Abstracts, p. 45.

15. Elibol, M.: Optimization of medium composition for actinorhodin production by Streptomyces coelicolor A3 (2) with response surface methodology. Process. Bio-chem. 2004, 39, 1057-1062.

16. Naili, O., Benounis, M., Darbouche, A. and Kheddouma, A.: Biological Denitri-fication Using Pure Cultures Isolated from Wastewater Treatment Plant of Khenchela (Eastern Algeria). J. Biol. Environ. Sci 2015, 9, 33-39.

OДРЕЂИВАЊЕ САСТАВА ХРАНЉИВЕ ПОДЛОГЕ ЗА МАКСИМАЛНУ ПРОДУКЦИЈУ БИОМАСЕ Pseudomonas stutzeri

Ана М. Видаковић*, Оља Љ. Шовљански, Александра С. Ранитовић,

Драгољуб Д. Цветковић, Синиша Л. Марков


Последњих година се спроводе опсежна истраживања о примени биоремедија-ције за превенцију и решавања глобалног проблема акумулацијe нитрата, нарочито у води и земљишту. У оквиру истраживања процеса биолошке денитрификације, бактерија Pseudomonas stutzeri се најчешће користи као модел организам због спо-собности потпуне редукције нитрата. Циљ овог рада је формулисање хранљиве подлоге на лабораторијском нивоу погодне за максималну продукцију биомасе бак-теријских сојева Peudomonas stutzeri ATCC 17588 и Pseudomonas stutzeri D1



305

(изолован из Дунава, Нови Сад, Србија) употребом Box-Bhenken-oвог експеримен-талног дизајна и методе одзивне површине. Као независне променљиве (+, 0 и - ниво) у оквиру експеримената одабране су: глукоза (0, 1, 2 g/L), KNO3 (1, 2, 3 g/L) и пептон (0,4, 2, 4 g/L), док је број вијабилних ћелија дефинисан као зависна променљива (одзив). На основу добијених резултата може се запазити да се макси-малан број вијабилних ћелија P. stutzeri ATCC 17588 постиже у хранљивој подлози која садржи 0 g/L глукозе, 2 g/L KNO3 и 4 g/L пептона. У случају P. stutzeri D1, одговарајућа хранљива подлога за производњу биомасе се састоји од 1 g/L глукозе, 3 g/L KNO3 и 4 g/L пептона. Добијени резултати се могу користити за техно-еко-номске анализе процеса и повећање размере производње биомасе P. stutzeri. Кључне речи: биомаса, Pseudomonas stutzeri, Box-Bhenken- oв дизајн, метода

одзивне површине


APTEFF, 48, 1-323 (2017) UDC: 667.283.1:[66.021.3.081.3:664.121 https://doi.org/10.2298/APT1748307V BIBLID: 1450-7188 (2017) 48, 307-314


307

REMOVAL OF ACRIDINE ORANGE DYE FROM AQUEOUS SOLUTION BY ADSORPTION ONTO DRIED SUGAR BEET PULP

Vesna M. Vučurović*, Vladimir S. Puškaš, Uroš D. Miljić

University of Novi Sad, Faculty of Technology Novi Sad, Bulevar Cara Lazara 1, 21000 Novi Sad

A simple, low cost, and effective method for the removal of acridine orange (AO), a mutagenic cationic dye, from aqueous model solutions by adsorption onto dried sugar beet pulp (SBP) was evaluated in the present study. The AO removal was enhanced along with the increase of the initial solution pH and dye concentration. It was found that the adsorption process closely follows a pseudo-second-order chemisorption kinetics. The obtained equilibrium data obey both the Freundlich and Langmuir isotherm models. The SBP was proved to be very promising adsorbent for AO removal. Maximum adsorption capacity of the Langmuir monolayer of SBP for AO was found to be 5.37, 34.6, 89.62, 144.53 and 324.58 mg/g, at 25 °C for the solution pH of 2, 4, 5, 6, and 8, respectively. KEY WORDS: acridine orange, sugar beet pulp, adsorption, isotherm, kinetics

INTRODUCTION Today, the increased use of dyes in the textile, paper, rubber, plastics, cosmetics, pharmaceutical and food industries often causes pollution problems in the form of colo-red wastewater discharged into water environment (1). Acridine orange (AO) is a nucleic acid selective fluorescent cationic dye which is widely used in the field of printing and dyeing, leather, lithography and biological staining. AO is toxic, cell-permeable, interacts with DNA and RNA, and has photodynamic and mutagenic actions. AO is xenobiotic, visible at parts per million dilutions in aqueous solutions (2). Recent progress in the wastewater treatment technology has led to intensive development of dyes removal by adsorption onto various agricultural wastes, as a renewable replacement for costly com-mercially adsorbents such as activated carbon (3). However, only several investigations have been devoted to the study of AO adsorption onto different adsorbents such as car-boxylic functionalized superparamagnetic mesoporous silica microspheres (4), magnetic nanoparticles (5), granular kohlrabi peel (6), magnetically labeled baker’s yeast cells (7), magnetic charcoal (8), magnetic nanoparticles (9), magnetically responsive yeast-based biosorbent (10), etc. Sugar beet pulp (SBP) is a very cheap (100 US$ per metric tone), abundantly available byproduct of the table sugar industry resulting from the extraction of simple sugars from the sugar beet (Beta vulgaris L.) (11). On a dry mass basis, it con-tains 65-80% polysaccharides, consisting roughly of 40% cellulose, 30% hemicelluloses, * Corresponding author: Vesna M. Vučurović, University of Novi Sad, Faculty of Technology Novi Sad, Bule-

var Cara Lazara 1, 21000 Novi Sad, Serbia, e-mail: [email protected]



308

and 30% pectin (12). Due to large pores, heterogeneous structure, and the abundant avai-lability of active sites SBP is a highly efficient adsorbent for cations, such as cadmium and led (13), copper (14), methylene blue (15), and less effective adsorbent for anions, such as Gemazol turquoise blue-G (anionic dye) (14). To our knowledge, this would be first report describing AO removal by adsorption using SBP. Equilibrium isotherm adsorption and kinetic studies were performed in order to explore the effects of initial dye concentration, and solution pH on the rate, kinetics and mechanism of adsorption.

EXPERIMENTAL

Adsorbate. Acridine orange zinkchlorid double salt (chem. formula: C17H20 CIN3·ZnCl2; M = 319.86 g mol−1, λ max: 490 nm), cationic dye, was purchased from Merck, Germany, and was used as received. The stock solutions of the desired concentrations of AO (20, 30, 40 and 50 mg/l) were prepared with distilled water. The pH value of the solutions was adjusted with 0.1 M NaOH and 0.1 M HCl solution, using a pH meter (Consort C863, Belgium) with a combined pH electrode. Adsorbent. The pulp was obtained from the TE-TO Senta Sugar Factory, Serbia. The material was extensively washed with distilled water to remove soil and dust, boiled with water for 10 min, filtrated out and dried in an oven at 60 °C up to a constant mass. The dried material was ground on a laboratory mill achieving particle size with diameter less than 250 µm. The prepared material was stored in an airtight container for further use. Batch adsorption experiments. Batch adsorption experiments were conducted by varying the initial AO concentration (20, 30, 40 and 50 mg/l) and solution pH (2, 4, 5, 6 and 8). The adsorption studies encompassed the aspects of adsorption isotherms and adsorption kinetics. A fixed amount of the adsorbent (1 g) was poured into a definite volume (100 ml) of dye solution for each examined combination of pH and initial dye concentration in Erlenmeyer flasks. The Erlenmeyer flasks were thermostatated (25 °C) on a rotary shaker (GFL, Germany, Type 3015), at 120 rpm and shaking diameter 30 mm. The adsorption was carried out for 240 min, which was more than sufficient to reach full equilibrium. At the predetermined time, a sample (5 ml) of the mixture was taken and centrifuged for 10 min at 3000 rpm. The supernatant solution was analyzed by measuring the absorbance at λmax= 490 nm using a spectrophotometer (Carl Zeiss, Jena, Germany). The final AO concentration was determined using the constructed calibration curve. The amount of adsorption at time t, qt (mg/g) and at equilibrium, qe (mg/g) was calculated as follows:

W

VCCq to

t

)(

[1]

W

VCCq eo

e

)(

[2]

where Co and Ct and Ce are the liquid-phase concentrations of dye at initial moment, at time t, and at equilibrium, respectively; V (l) is the volume of the solution and W (g) is the mass of dry adsorbent used. The dye removal percentage was calculated by:



309

Removal (%) 100)(

o

eo

C

VCC [3]


From the dependence of the amount of AO adsorbed on SBP (qt) at a contact time (t) for various process conditions (figures not shown) it was noticed that there are two separated re-gions, of which the first straight line can be attributed to the macropore diffusion and the se-cond linear portion to the micropore diffusion. For all initial AO concentrations and pH of 4, 5, 6 and 8, the initial adsorption was fairly rapid during the first 15 min, reaching dye removal of 74.2-96.0 %, and then proceeded slowly towards the end of adsorption up to the maximum removal of 93.9-96.7 %. This indicates a very fast adsorption and strong interaction between the dye and the most readily available adsorbing sites on the adsorbent surface. The rapid ad-sorption of AO on SBP is attributed to the abundant availability of active sites, presence of multiple functional groups and porous structure which provides a large surface area and ready access for cations (13). Hence, it can be assumed that the adsorption process occurs mainly on the surface of SBP particles. The second phase may be attributed to a very slow diffusion of the adsorbent from the liquid phase to micro-pores. The difference in the values of qt at 60 min and 240 min was very small, probably due to the impossibility of dye molecules to dif-fuse deeper into the adsorbent structure. The initial AO concentration provides important dri-ving force to overcome the mass transfer limitations of AO between the aqueous solution and solid phase. Thus, a higher initial AO concentration enhances the adsorption process. The ad-sorption capacity at equilibrium (qe) increased with an increase in the initial AO concentration from 20 to 50 mg/l under the same operating conditions. On the other hand, at the pH 2, for each examined initial AO concentration, the initial adsorption process was significantly slo-wer, the first phase duration was about 60 min, reaching 57.6-68.7%. Also, the final equilibri-um adsorption capacity (qe) was the lowest at the pH 2. The qe significantly increased as the initial pH increased from the pH 2 to 4. The value of qe was slightly altered beyond the pH 4, and achieved a maximum at the pH 8. Consequently, the final removal was the lowest at the pH 2 (ranging from 60.8% to 70.6 %), sharply increased at the pH 4 (ranging from 93.9 to 94.6 %), and then slightly increased up to the pH 8 (ranging from 96.5 to 96.7%). AO exists in the aqueous solution in the form of positively charged ions, and consequently, the degree of its adsorption on the adsorbent surface is primarily influenced by the surface charge on the adsor-bent. At the pH>8, most probably AO is not any more completely protonated, therefore no ionic interactions between the AO and the adsorbent can occur. The surface charge on the ad-sorbent is in turn influenced by the solution pH. The SBP is dominated by negatively charged sites that are mostly carboxylated groups with some weaker acidic groups (15). The overall surface of SBP at the pH 2 is surrounded by hydronium ions and become increasingly less ne-gative, which decreases the interaction of dye cations with binding sites of SBP by greater re-pulsive forces, and reduces the attraction of positively charged ions of dye. At the pH values higher than 4, the carboxyl groups of SBP are deprotonated and negatively charged. Therefo-re, the maximal adsorption level is reached when the carboxyl functional groups of SBP are deprotonated and neutralized, probably via the electrostatic interaction with positively charged AO molecules. These results are also in accordance with the results of previous studies (13-15).



310

The adsorption isotherm models were used to describe the interaction between the solution and adsorbent, and the values of the constant, to express the surface properties and determine the maximum adsorption capacity of dried SBP to bind AO. The equilib-rium data were analyzed by the linear regression analysis to fit the isotherm models (Tab-le 2) given by Freundlich (16), Langmuir (17) and Temkin (18). The value of the deter-mination coefficient (R2) closer to 1 indicates that the respective equation fits better the experimental data. The linear form of adsorption isotherms equations, the obtained values of relevant parameters, along with the determination coefficients are shown in Table 1.

Table 1 Isotherm parameters for adsorption of AO onto SBP at 25 °C.

Isotherm Equation Parameter pH 2 pH 4 pH 5 pH 6 pH 8

Freundlich efe Cn

Kq log1

loglog

Kf 0.434 1.760 1.713 1.837 2.953

1/n (g-1) 0.652 0.906 0.986 0.972 0.989

n (g) 1.535 1.104 1.014 1.028 1.011

R2 0.978 0.989 0.994 0.995 1.000

Langmuir memae qCqKq

111

qm (mg/g) 5.37 34.60 89.62 144.53 324.58

Ka (l mg-1) 0.059 0.053 0.020 0.013 0.009

R2 0.964 0.993 0.994 0.996 1.000

Temkin ete CBKBq lnln Kt (l/g) 0.154 3.608 3.107 3.650 10.277

B (J/mol) 1.364 2.773 2.998 2.981 3.096 R2 0.950 0.980 0.965 0.978 0.984

The Freundlich isotherm model is the earliest known empirical equation, and is shown to be consistent with the exponential distribution of active centers and characteristic for heterogeneous adsorbent surfaces. Kf and n are the Freudlich constants, which are the indicators of adsorption capacity and adsorption intensity, respectively (16). As shown in Table 1, the Kf value increases with the increase of the solution pH, indicating relatively easy uptake of AO from the aqueous solution with a high adsorption capacity of SBP. In general, the value of 1/n ranges between 0 and 1, and indicates the degree of non-linearity between solution concentration and adsorption as follows: if the value of 1/n is equal to unity, the adsorption is linear; if the value is below unity, this implies that the adsorption process is chemical and indicates a normal Langmuir isotherm; if the value is above unity, the adsorption is a favorable physical process. The more heterogeneous the surface, the closer 1/n value is to 0 (19). The obtained values of 1/n for the pH 4-8 were in the range 0.906-0.989 (g-1), indicating linear chemical adsorption. Also, the obtained 1/n values were not close to 0, confirming that the surface of SBP is dominated by homo-geneous and negative charge with a small heterogeneity. At the pH of 2, the 1/n value



311

was significantly lower (0.652), which confirms that the surface charge of SBP was strongly affected by the presence of hydronium ions. The Freundlich isotherm model gave a good fit (the R2 ranged from 0.978 to 1.000) to the adsorption data of AO on SBP. The Langmuir isotherm model is based on the assumption of the existence of a mono-layer adsorption onto a completely homogeneous surface with a finite number of identical sites and with negligible interaction between the adsorbed molecules. The quantities qm (mg/g) and Ka (l/mg) are the Langmuir constants related to the maximum adsorption capacity and bonding energy of adsorption, respectively. The constant qm represents the limiting capacity when the adsorbent surface is fully covered with dye, particularly in the cases when the adsorbent did not reach its full saturation in the experiment (17). As shown in Table 1, the obtained qm values were 5.37, 34.6, 89.62, 144.53 and 324.58 mg/g for the solution pH of 2, 4, 5, 6, and 8, respectively. Therefore, it can be concluded that the adsorption was strongly affected by the solution pH. The Langmuir isotherm model yielded the best fit compared to the other two models, especially at the pH value 4-8 (R2=0.993-1.000). On the basis of the obtained qm values it was found that SBP is one of the most promising low-cost, abundantly available and highly efficient adsorbents for the removal of AO from aqueous solution in comparison to other adsorbents reported in the literature (5 - 10). The main characteristic of the Langmuir isotherm can be expressed in terms of a dimensionless separation factor called equilibrium parameter RL (20), defined by: RL=1/(1+KaCo), where RL is a dimensionless separation factor, Co (mg/l) is the initial AO dye concentration and Ka (l/mg) is the Langmuir constant. The value of the parameter RL indicates the shape of isotherm to be either unfavorable (RL>1), linear (RL=1), favo-rable (0<RL<1), or irreversible (RL=1). The values of separation factor RL obtained in this work were in the range between 0.252 and 0.847, indicating that the adsorption of AO onto SBP was favorable for the variety of examined process conditions. The Temkin iso-therm model was studied to explore the energy distribution. The Temkin isotherm con-tains a factor that explicitly takes into account the interactions between the adsorbing spe-cies and the adsorbent. The isotherm constant B=RT/b (J/mol) is related to heat of sorp-tion, R is the gas constant (8.314 J/ mol·K), T (K) is the absolute temperature, b and Kt (l/g) are the Temkin isotherm constants (18). The Temkin equation was found to be poor to fit the experimental adsorption equilibrium data (the R2 ranged from 0,950-0.984). The experimental data yielded excellent fits within the following isotherm order: Langmuir > Freundlich > Temkin. The results showed that the pH value is the key variable, affecting strongly the adsorption capacity of SBP for AO. The simplified linear form of pseudo-first order kinetic model (21), pseudo-second order kinetic model (22), and intraparticle diffusion model (23) were applied to test the experimental data in order to describe the dynamics and mechanism of AO adsorption on SBP (Table 2). The k1 (min-1) is the adsorption rate constant of the pseudo-first-order equation, k2 (mg/g·min) is the rate constant of the pseudo-second-order equation), while ki i s the intraparticle diffusion rate constant.



312

Table 2 Parameters of the kinetic models, equilibrium amount and linear determination coefficient for the removal of AO with SBP at 25 °C.

The adsorption of liquid/solid system based on solid capacity which proceeds by diffusion through a boundary, most likely follows the pseudo-first-order equation. The pseudo-second-order kinetic model is used to describe the chemisorption processes. From the obtained values of the squared correlation coefficients (R2) it is obvious that the adsorption of AO onto SBP closely follows a pseudo-second-order kinetics (R2 =0.974-1.000), and that chemisorption is definitely the rate controlling step. On the other hand, although the R2 values are reasonably high in some cases, and the calculated qe values obtained from pseudo first-order and intraparticle diffusion model agree well with the experimental data, it can be generally said that these models were not appropriate for description of the AO adsorption onto SBP.

CONCLUSION This work clearly indicates that the adsorption of AO on dried SBP is a simple and highly efficient method for the removal of AO from aqueous solutions, and that it may be used for treatment of the wastewater that is difficult to deal with using chemical and biological methods. The adsorption process closely follows the pseudo-second-order ki-netic model, while rate controlling step is chemisorption. The maximum Langmuir mono-layer adsorption capacity of SBP for AO was found to be 324.58 mg/g for the solution pH 8.



313

Acknowledgement

The authors acknowledge the financial support of the Project TR-31002 from the Mi-nistry of Science and Technological Development of the Republic of Serbia.

REFERENCES 1. Yener, J.; Kopac, T.; Dogu, G.; Dogu T. Dynamic analysis of sorption of Methylene

Blue dye on granular and powdered activated carbon. Chem. Eng. J. 2008, 144, 400-406.

2. Lauretti, F.; Lucas de Melo, F.; Benati, F.J.; de Mello Volotăo, E.; Santos, N.; Linha-res, R.E.C.; Nozawa C. Use of acridine orange staining for the detection of rotavirus RNA in polyacrylamide gels. J. Virol. Methods 2003, 114, 29-35.

3. Bhatnagar, A.; Minocha, A.K.; Sillanpää, M. Adsorptive removal of cobalt from aqueous solution by utilizing lemon peel as biosorbent. Biochem. Eng. J. 2010, 48, 181-186.

4. Fu, X.; Chen, X.; Wang, J.; Liu, J. Fabrication of carboxylic functionalized superpara-magnetic mesoporous silica microspheres and their application for removal basic dye pollutants from water. Micropor. and Mesopor. Mater. 2011, 139, 8-15.

5. Qadri, S.; Ganoe, A.; Haik Y. Removal and recovery of acridine orange from solu-tions by use of magnetic nanoparticles. J. of Hazard. Mater. 2009, 169, 318-323.

6. Gong, R.; Zhang, X. ; Liu, H.; Sun, Y.; Liu B. Uptake of cationic dyes from aqueous solution by biosorption onto granular kohlrabi peel. Bioresour.Technol. 2007, 98, 1319-1323.

7. Safarik, I.L.P.; Safarikova, M. Adsorption of dyes on magnetically labeled baker's yeast cells. Eur. Cells and Mater. 2002, 3, 52-55.

8. Śafařík, I.; Nymburská, K.; Śafaříková, M. Adsorption of Water-Soluble Organic Dyes on Magnetic Charcoal. J. Chem. Technol. Biotechnol. 1997, 69, 1-4.

9. Safarik, I.; Safarikova, M. Magnetic fluid modified peanut husks as an adsorbent for organic dyes removal. Phys. Procedia 2010, 9, 274-278.

10. Safarik, I.; Rego, L.F.T.; Borovska, M.; Mosiniewicz-Szablewska, E.; Weyda, F.; Safarikova, M. New magnetically responsive yeast-based biosorbent for the efficient removal of water-soluble dyes. Enzyme Microb. Technol. 2007, 40, 1551-1556

11. Dinand, E., H. Chanzy and Vignon, M.: Parenchymal cell cellulose from sugar beet pulp: preparation and properties. Cellulose 3 (1996) 183-188.

12. Sun, R.; Hughes, S. Fractional extraction and physico-chemical characterization of hemicelluloses and cellulose from sugar beet pulp. Carbohyd. Polym. 1998, 36, 293-299.

13. Pehlivan, E.; YanIk, B.H.; Ahmetli, G.; Pehlivan, M. Equilibrium isotherm studies for the uptake of cadmium and lead ions onto sugar beet pulp. Bioresour.Technol. 2008, 99, 3520-3527.

14. Aksu, Z.; Isoglu, I.A. Use of dried sugar beet pulp for binary biosorption of Gemazol Turquoise Blue-G reactive dye and copper(II) ions: Equilibrium modeling. Chem. Eng. J. 2007, 127, 177-188.



314

15. Vučurović, V.; Razmovski, R.; Tekić, M. Methylene blue (cationic dye) adsorption onto sugar beet pulp: Equilibrium isotherm and kinetic studies. J.Taiwan Inst. of Chem. E. 2012, 43, 108-111.

16. Freundlich, H.M.F. Uber die adsorption in losungen. Z. Physik. Chemie (Leipzig) 1906, 57A, 385-470.

17. Langmuir, I. The constitution and fundamental properties of solids and liquids. J. Am. Chem. Soc. 1916, 38, 2221-2295.

18. Tempkin, M.J.; Pyzhev V. Recent modifications to Langmuir isotherms. Acta Physio-chim. URSS 1940, 12, 217-222.

19. Renault, F.; Morin-Crini, N.; Gimbert, F.; Badot, P.-M.; Crini G. Cationized starch-based material as a new ion-exchanger adsorbent for the removal of C.I. Acid Blue 25 from aqueous solutions. Bioresour. Technol. 2008, 99, 7573-7586.

20. Webi, T.W.; Chakravort, R.K. Pore and solid diffusion models for fixed-bed adsor-bers. AIChE J. 1974, 120, 228-238.

21. Lagergren, S. About the theory of so-called adsorption of soluble substances. Kungli-ga Svenska Vetenskapsakademiens, Handlinger 1989, 24, 1-39.

22. Ho, Y.S.; McKay, G. Sorption of dye from aqueous solution by peat. Chem. Eng. J. 1998, 70, 115-124.

23. Weber, W. J.; Morris, J.C. Kinetics of Adsorption on Carbon from Solutions. J Sanit Eng Div ASCE 1963, 89, 31-59.

УКЛАЊАЊЕ БОЈЕ АКРИДИН ОРАНЖ ИЗ ВОДЕНОГ РАСТВОРА

АДСОРПЦИЈОМ НА СУВИМ РЕЗАНЦИМА ШЕЋЕРНЕ РЕПЕ

Весна М. Вучуровић, Владимир С. Пушкаш, Урош Д. Миљић


У овом раду је испитан једноставан, јефтин и ефикасан метод за уклањање му-тагене катјонске боје акридин оранж (АО) из водених модел раствора адсорпцијом на сувим резанцима шећерне репе (СРШР). Са повећањем вредности pH раствора и почетне концентрације боје у раствору повећава се ефикасност уклањања боје. Процес адсорпције АО се одвија према хемисорпционом кинетичком моделу дру-гог реда. Остварени равнотежни адсорпциони капацитет одговара моделима изо-терми по Freundlich-у и Langmuir-у. У раду је потврђено да СРШР представљају веома перпективан и ефикасан адсорбент за уклањање АО. Максимални Langmuir-ов адсорпциони капацитет СРШР за уклањање АО при pH вредностима раствора 2, 4, 5, 6 и 8 је износио 5,37, 34,6, 89,62, 144,53 и 324,58 mg/g, на 25 °C. Kључне речи: Акридин оранж, резанци шећерне репе, адсорпција, изотерма, кине-

тика


APTEFF, 48, 1-323 (2017) UDC: 628.166.094.3-926.214:66.067.2:637.142 https://doi.org/10.2298/APT1748315Z BIBLID: 1450-7188 (2017) 48, 315-323


315

OZONATION OF NANOFILTRATION PERMEATE OF WHEY BEFORE PROCESSING BY REVERSE OSMOSIS

Yurii G. Zmievskii1*, Volodymyr V. Zaharov1, Olexandra S. Rudenko2,

Iryna M. Biletskaya1, Valeriy G. Myronchuk1

1 National University of Food Technology of the Ministry of Education and Science of Ukraine,

Vladimirskaya str. 68, 01601, Kiev, Ukraine;

2 V.I. Vernadskii Institute of General & Inorganic Chemistry of the NAS of Ukraine, Palladin Pr. 32/34, 03142, Kiev, Ukraine

During nanofiltration processing of whey a significant amount of permeate is generated. In some cases this permeate is treated by reverse osmosis to get purified water for technological needs. Dry substances are not used, because they contain practically the same amount of organic and inorganic components. Mineral substances can be used for the mineralization of drinking water purified by reverse osmosis. However, the presence of organic compounds complicates the process of separation, as well as reduces the specific productivity of reverse osmosis membranes at the concentration stage. Therefore, the search for methods of destruction and removal of organic components is grounded. In the presentед work, experimental studies of ozonation and sorption of organic compounds by activated carbon were carried. It has been shown that ozonation improves the degree of sorption purification by six times. Sequential treatment with ozone and subsequent filtration through the layer of activated carbon improves the specific productivity of reverse osmosis membranes by 30% at the stage of treatment of the nanofiltration permeate, while their selectivity remains unchanged. KEY WORDS: ozonation, reverse osmosis, adsorption, permeate, nanofiltration

INTRODUCTION Whey is a valuable by-product produced by the manufacture of solid, dairy cheeses, and casein (1). The modern technology of its processing consists of the following sequen-ce of technological operations. First, the residues of milky fat and casein fraction are se-parated in a centrifugal separator. Then, after pasteurization and cooling to 8-12 °С, the whey is concentrated by nanofiltration to a content of dry substances of 18-22%. The final condensation takes place in vacuum evaporator plants, after which whey goes to the crystallizer, where lactose crystals are formed.

* Corresponding author: Yurii G. Zmievskii, National University of Food Technology of the Ministry of Education

and Science of Ukraine, Vladimirskaya str. 68, 01601, Kiev, Ukraine, e-mail: [email protected]



316

The treated whey is dried in a spray dryer, cooled and packaged. Such product has wide applications in the food industry, since up to 40% of minerals are removed at the nanofiltration stage (2), which influences positively organoleptic markers. However, an unresolved problem is effective use of nanofiltration permeate, which is produced in amount of about 650 dm3 per 1000 dm3 of processed whey. This is due to the low content of dry matter (0.4-1.2%) and approximately equal amounts of organic and inorganic components (1.1-2 g/dm3 of lactose and 1.8 g/dm3 of total content of inorga-nic ions, such as Na+, K+, Cl-) (1, 3, 4). However, the permeate composition is unique for each individual production and depends on the conditions of separation, composition of milky whey, quality of membranes, and duration of the nanofiltration process. At most industrial plants, nanofiltration permeate of milky whey is drained into the sewage sys-tem. Among separation techniques which allow concentration of low-molecular organic substances and inorganic ions, reverse osmosis is the most commonly used due to simple equipment, ease of scaling and soft separation conditions (5, 6). This method is applied to water desalination (7, 8), particularly in the food and beverage industry (9, 10). Some industrial plants apply reverse osmosis to concentrate whey (11). Then, the obtained puri-fied water is used further, for instance, for washing of the equipment, while the concen-trate is discharged to the sewage system. In previous works (12, 13) we investigated the processes of treatment of model solutions of nanofiltration permeate by reverse osmosis and membrane distillation in a combination with electrolysis. The purpose was to obtain concentrates of mineral salts that could be used to mineralize drinking water purified by reverse osmosis. However, the experiments with real solutions showed that the majority of organic compounds are transported through the ion exchange membranes simultaneously with inorganic ions. This decreases the quality of the obtained concentrates. Reverse osmosis was also applied to concentrating the permeate, which is formed during nanofiltration. It was also proved that the organic components in nanofiltration permeate reduce the specific productivity of the reverse osmosis membranes. The situation can be improved by modifying membranes with nanoparticles of inorga-nic ion-exchangers (14-17) or removal of organics from the permeate solution before the membrane process. The organic compounds can be removed from aqueous solutions, for instance, by means of destruction (18) and adsorption (19). One of the promising approa-ches to destruction and removal of organic compounds is the process of ozonation of so-lution, followed by filtration through the bed of activated carbon (20). Therefore, the pur-pose of this work was to apply this approach to the pre-treatment of nanofiltration per-meate that is formed during whey processing before concentrating by reverse osmosis.

EXPERIMENTAL

Milky whey, produced by Pyriatyn Cheese Plant (Group of Enterprises “Milk Allian-ce”, Ukraine), contained 6.2 % of dry matter, 0/9 % proteins, 4.6 % lactose, and 0.5 % of inorganic components. This liquid was used for preparation of nanofiltration permeate under laboratory conditions. The equipment for filtration was described in detail earlier,



317

except for the cell (21). However, in this case, the roll flow-type module produced by Keensen Technology Co. Ltd. (China) was used, the module was supplied with a nanofil-tration membrane of effective area of 0.56 m2. The filtration process was carried out for 3 h, and 4 liters of permeate was obtained at 0.5 MPa. Preliminarily, the whey was treated using a microfiltration apparatus supplied with a cartridge separating element (BCCF, Aquafilter, USA). A dead-end laboratory setup with an effective membrane area of 1.3·10-3 m2 was used for concentrating nanofiltration permeate under laboratory conditions. A reverse osmosis membrane RM Nanotech (Russia) was used after previous filtering distilled water through it at the working pressure until a steady flux was established. The chemical oxygen demand (COD) for nanofiltration permeate was determined ac-cording to Kubel's method (22), and its value was 11,400 mg/dm3. The nanofiltration permeate of milky whey obtained under industrial conditions (by Pyriatyn Cheese Plant, Ukraine) was also used in our investigations, the COD was esti-mated was to be 400 mg/dm3. In this case, the membranes produced by GE Osmonics Company (USA) were used. At the same plant, the nanofiltration permeate was treated by reverse osmosis using the membrane produced by GE Osmonics Company (USA). A pressure drop was kept at 3 MPa. Removal of the organics from the concentrate obtained during the process of re-verse osmosis, was also studied. The COD value was 5000 mg/dm3. A laboratory apparatus was used for ozonation of the nanofiltration permeate (Figure 1). The system included a generator of ozone (Ecozone 6-AW, Digidrol, Ukraine) as well as an oxygen concentrator (JAY-8, China). The efficiency of the process of ozone gene-ration was 6 g/h. The operation principle of the apparatus is as follows. After the vacuum pump is swit-ched on, air passes through a dehumidifier filled with silica. The rate of air pumping is controlled with a rotameter. The air is enriched with oxygen in a concentrator, then the gas is passed through the ozonizer, where ozone is generated due to the electrical dischar-ge. The resulting gas mixture enters the reactor, where it passes through the processed solution. Since foam can be produced in the nanofiltration permeate, a foam catcher is installed after the reactor. Further, the ozone-gas mixture gradually falls into two Drexel bottles and filled with a KI solution. The remaining mixture of gases is released into the atmosphere through a vacuum pump. After ozonation, the solution was passed through the column filled with activated carbon (Desotec Organosorb, Belgium). The height of the adsorbent bed was 16 cm, the diameter of the column was 6 cm, the superficial solu-tion velocity was 1.5-2.5 m/h. For comparison, only adsorption was used for removal of organics from the permeate solution obtained under laboratory conditions. The concentration of ozone in the gas phase was determined by iodometric method. The gas phase passed through a Drexel bottles (see Figure 1) filled with a solution of po-tassium iodide, reacting with ozone. Then, this solution was titrated with sodium thiosul-phate, and the amount of ozone that reacted was calculated. The nanofiltration permeate of whey, which was obtained under industrial conditions, was used for separation by means of reverse osmosis at 4 MPa, using a KS-C membrane produced by NanoTech (Russia). The apparatus for reverse osmosis was described earlier



318

(21). For comparison, the liquid after ozonation and adsorption was also treated by rever-se osmosis. The content of inorganic ions in the permeate and concentrate was determi-ned similarly to (16).

Figure 1. Scheme of the laboratory apparatus for ozonation.

RESULTS AND DISCUSSION First of all, the industrial samples of nanofiltration permeate and concentrate after re-verse osmosis were treated with ozone for 20 min. The rate of air flow through the sys-tem was 240 dm3/h, the concentration of ozone in the gas phase was 25 mg/dm3. The pro-cess was accompanied by significant foam formation. As a result, the turbid solution be-came transparent indicating degradation of macromolecules (Figure 2). The COD value decreased down to 2280 mg/dm3 for reverse osmosis concentrate. In other words, the concentration of organics in this liquid was lower more than two times in comparison with the pristine solution, Thus, the ozonation technique allows us to decrease sufficient-ly the content of organic substances, but additional purification is needed. Regarding the nanofiltration permeate, a decrease of the COD value is not so sufficient (from 400 down to 360 mg/dm3). However, the solution also becomes more transparent. Comparing with the unconcentrated nanofiltration permeate, the concentrate after reverse osmosis con-tains more degradable substances, and these substances evidently provide yellow color of the liquid.



319

1 - NF permeate of whey; 2 - NF permeate of whey concentrated by reverse osmosis

Figure 2. Samples before and after ozonation

Further attention was focused on the combination of ozonation and sorption proces-ses. The nanofiltration permeate of milky whey obtained under laboratory conditions was purified from organics. The large value of COD for the pristine permeate solution indica-tes lower selectivity of the nanofiltration membranes than that of the membrane used un-der industrial conditions. The results are given in Figure 3.

Figure 3. Experimental data of the ozonation and adsorption process involving

nanofiltration permeate obtained under laboratory. It can be seen that the ozonation caused no sufficient decrease of the content of orga-nic substances in the nanofiltration permeate, probably because of the impeded destruct-tion of large macromolecules. The adsorption produced more than a four-fold decrease in the COD value. At last, the minimal content of organic substances was reached after the ozonation followed by adsorption, which is in agreement with the literature (23, 24). As



320

known, the ozonation of organic substances causes breaking C-C bonds, followed by the formation of oxygen-containing groups, particularly dissociated ones (COOH) (25). On the other hand, materials based on activated carbon contain both hydrophilic and hydro-phobic pores (26). Ion exchange groups are located in the hydrophilic pores. Moreover, ozone can oxidize the surface of activated carbon (additional hydrophilization). As a re-sult, additional -OH groups appear on the surface. These groups are responsible for anion exchange. Thus, the reason of enhanced adsorption can be additional contribution of the ion exchange mechanism to adsorption. When considering such treatment as preparation of the solution before its reverse os-mosis concentration, two functions of activated carbon are important. The first one is the maximal removal of organic substances from the solution. The second function of activa-ted carbon is the destruction of ozone traces, their content in the solution can reach se-veral milligrams per liter under ambient temperature. When the nanofiltration permeate is used further for separation by means of reverse osmosis, ozone removal is necessary to prevent destruction of the membrane. However, ozone is unstable (its half-life is 20 min. (27)), and tends to transform to oxygen. Thus, the solution must be stored for some time or heated before processing by reverse osmosis. Nevertheless, under industrial conditions, the increased processing time or additional technological process (heating) creates a significant number of organizational problems. Therefore, first of all, in order to deactivate the oxidizing agent, it is necessary to apply filtering of the solution through the bed of activated carbon. Appropriate studies have been carried out to confirm the positive effect of the combi-nation of ozonation and adsorption. In this case, the process of reverse osmosis was stu-died. As found, ozonation followed by adsorption reduced the COD value from 400 to 100 mg/dm3. The results are shown in Figure 4.

Figure 4. Permeate flux through the reverse osmosis membrane as a function of time.

It is seen that the productivity of the reverse osmosis membrane is higher by 33-40% if the solution was treated with ozone. The selectivity towards inorganic ions was about 95% in both cases. The improvement of rejection ability of the membrane is evidently



321

caused by lower content of organics in the liquid after preliminarily ozonation and adsorption. It should also be noted that ozonation disinfects the solution. Under industrial condi-tions, the tank of temporary storage of the permeate before reverse osmosis has to be wa-shed with chemical reagents using hot water at least once every two days. If the ozona-tion process is properly organized, the frequency of washing should be significantly redu-ced, however, it should be checked in the production conditions and investigated separa-tely.

CONCLUSIONS Treatment of nanofiltration permeate of milky whey by ozone followed by filtration through the bed of activated carbon was investigated. The synergetic effect of the combi-nation of these processes can be due to a change of the composition of organic substances and appearance of -OH anion exchange groups on the surface of activated carbon under the influence of ozone. This allowed increasing the permeate flux through the reverse os-mosis membranes by 33-40% without deterioration of selectivity. It is assumed that the disinfectant effect of ozone will reduce the cost of chemical reagents and hot water used in dairy enterprises for the washing of temporary storage tanks for nanofiltration permeate before reverse osmosis.

REFERENCES

1. Bylund, Ed.G. Dairy Processing Handbook; Lund: Tetra Pak Processing Systems AB, 1995; pp 331-349.

2. Pan, K.; Song, Q.; Wang, L.; Cao, B. A study of demineralization of whey by nano-filtration membrane. Desalination. 2011, 267, 217-221.

3. Atra, R.; Vatai, G.; Bekassy-Molnar, E.; Balint, A. Investigation of ultra- and nanofil-tration for utilization of whey protein and lactose. J. Food Eng. 2005, 67, 325-332.

4. Bidhendi, G.N.; Nasrabadi, T. Use of nanofiltration for concentration and deminera-lization in the dairy industry. Pak. J. Biol. Sci. 2006, 9 (5), 991-994.

5. Baker, R.W. Membrane technology and applications; John Wiley & Sons, Ltd, 2004, 2nd ed; pp. 191-232.

6. Mulder, M. Basic principles of membrane technology; Dordrecht: Kluwer, 1996; pp 208-415.

7. Hisham T. El-Dessouky and Hisham M. Ettouney "Fundamentals of Salt Water De-salination", Elsevier, 2002; 439-452.

8. Chong, T.H.; Loo, S.L.; Krantz, W.B. Energy-efficient reverse osmosis desalination process. Original research article. Journal of Membrane Science. 2015, 473, 177-188.

9. Muro, C.; Francisco Riera, F.; Díaz, M. Membrane Separation Process in Wastewater Treatment of Food Industry, Food Industrial Processes - Methods and Equipment; InTech, 2012, pp 253-280.



322

10. Yulie, E. Meneses, Y.E.; Stratton, J.; Flores, R.A. Water reconditioning and reuse in the food processing industry: Current situation and challenges. Trends in Food Scien-ce & Technology. 2017, 61, 72-79.

11. Vourch, M.; Balannec, B.; Chaufer, B.; Dorange, G. Nanofiltration and reverse osmo-sis of model process waters from the dairy industry to produce water for reuse. De-salination. 2005, 172, 245-256.

12. Zmievskii, Yu.G.; Kirichuk, I.I.; Mironchuk, V.G. Membrane treatment of wastewa-ter obtained after the whey processing. Journal of Water Chemistry and Technology. 2014, 36(6), 309-316.

13. Zmievskii, Yu.G.; Kyrychuk, I.I.; Mironchuk, V.G. Using of direct contact membrane distillation for wastewater treatment obtained after whey processing. Carpathian J. Food Sci. Technol. 2016, 8(2), 5-10.

14. Myronchuk,V.G.; Dzyazko, Y.S.; Zmievskii, Y.G.; Ukrainets, A.I; Bildukevich, A.V.; Kornienko, L.V.; Rozhdestvenskaya, L.V.; Palchik, A.V. Organic-inorganic membranes for filtration of corn distillery. Acta Period. Technol. 2016, 47, 153-165.

15. Dzyazko, Yu.S.; Rozhdestvenskaya, L.M.; Zmievskii, Yu.G.; Vilenskii, A.I.; Myron-chuk, V.G.; Kornienko, L.V.; Vasilyuk, S.L.; Tsyba, N.N. Organic-inorganic mate-rials containing nanoparticles of zirconium hydrophosphate for baromembrane sepa-ration. Nanoscale Res. Lett. 2015, 10, 64. DOI https://doi.org/10.1186/s11671-015-0758-x

16. Dzyazko, Yu.: Rozhdestveskaya, L.; Zmievskii, Yu.; Volfkovich, Yu.; Sosenkin, V.; Nikolskaya, N.; Vasilyuk, S.; Myronchuk, V.; Belyakov, V. Heterogeneous Membra-nes Modified with Nanoparticles of Inorganic Ion-Exchangers for Whey Deminera-lization. Materials Today: Proceedings. 2015, 2 (6), 3864-3873.

17. Zmievskii, Yu.G.; Dzyazko Y.S.; Myronchuk,V.G.; Rozhdestvenskaya, L.M.; Vilen-skii, A.I.; Kornienko, L.V. Fouling of polymer and organic-inorganic membranes du-ring filtration of corn distillery. Ukr. Food J. 2016, 5 (4), 739-747.

18. Papageorgiou, A.; Stylianou, S.K.; Kaffes, P.; Zouboulis, A.I.; Voutsa, D. Effects of ozonation pretreatment on natural organic matter and wastewater derived organic matter - Possible implications on the formation of ozonation by-products. Chemo-sphere. 2017, 170, 33-40.

19. Björklund, K.; Li, L.Y. Adsorption of organic stormwater pollutants onto activated carbon from sewage sludge. J. Environ. Manage. 2017, 197, 490-497.

20. Lee, L.Y.; Ng, H.Y.; Ong, S.L.; Hu, J.Y.; Tao, G.; Kekre, K; Viswanath, B.; Lay, W.; Seah, H. Ozone-biological activated carbon as a pretreatment process for reverse os-mosis brine treatment and recovery. Water Res. 2009, 43, 3948-3955.

21. Myronchuk, V.G.; Grushevskaya, I.O.; Kucheruk, D.D.; Zmievskii, Yu. G. Experi-mental study of the effect of high pressure on the efficiency of whey nanofiltration process using an OPMN-P membrane. Pet. Chem. 2013, 53 (7), 439-443.

22. ISO 8467:1993 Water quality - Determination of permanganate index. 23. Rakness, K.; Gordon, G.; Langlais, B.; Masschelein, W.; Matsumoto, N.; Richard, Y.;

Robson, C.M.; Somiya, I. Guideline for measurement of ozone concentration in the Process Gas from an ozone generator. Ozone Sci. Eng. 1996, 18, 209-229.



323

24. Zhang, J.; Northcott, K.; Duke, M.; Scales, P.; Gray, S. Influence of pre-treatment combinations on RO membrane fouling. Desalination. 2016, 393(1), 120-126.

25. Steven Van Geluwe, S.V.; Braeken, L.; Bruggen, B.V Ozone oxidation for the alle-viation of membrane fouling by natural organic matter: A review. Water Research. 2011, 45(12), 3551-3570.

26. Volfkovich, Yu.M.; Filippov, A.N.; Bagotsky, V. S. Structural Properties of Porous Materials and Powders Used in Different fields of science and technology. Springer. London, Heidelberg, New York, Dordrecht, 2014, pp 135-158.

27. Goncalves, A.A. Ozone - an emerging technology for the seafood industry. Braz. Arch. Biol. Technol. 2009, 52(6), 1527-1539.

ОЗОНИЗАЦИЈА ПЕРМЕАТА НАНОФИЛТРАЦИЈЕ СУРУТКЕ ПРЕ ПРОЦЕСИРАЊА РЕВЕРСНОМ ОСМОЗОМ

Јуриј Г. Змијевски1, Володимир В. Захаров1, Олександра С. Руденко2,

Ирина M. Билецкаја1, Валериј Г. Мирончук1

1 Национални универзитет прехрамбене технологије Министарства образовања и науке Украјине, Владимирска ул. 68, 01601, Кијев, Украјина

2 Институт опште и неорганске хемије "В. И. Вернадски" Националне академије наука Украјине,

Паладин пр. 32/34, 03142, Кијев, Украјина

У току обраде сурутке нанофилтрацијом настаје знатна количина пермеата. У неким случајевима се пермеат подвргава реверсној осмози да би се добила пречиш-ћена вода за технолошке потребе. Суве супстанце се не користе из разлога што практично садрже подједнаке количине органских и неорганских компоненти. Ми-нералне супстанце се могу користити за минерализацију пијаће воде пречишћене реверсном осмозом. Међутим, присуство органских једињења усложњава процес раздвајања, а такође смањује специфичну продуктивност мембрана реверсне осмо-зе у фази концентровања. Према томе, основана су настојања да се пронађу методе разарања и уклањања органских компонената. У овом раду су представљени резултати експерименталног испитивања озони-зације и сорпције органских једињења помоћу активног угља. Показано је да озони-зација побољшава степен пречишћавања сорпцијом шест пута. Озонизација после које следи филтрација кроз слој активног угља побољшава специфичну продуктив-ност мембрана у реверсној осмози 33-40% у фази обраде пермеата нанофилтраци-јом, док њихова селективност остаје непромењена.

Кључне речи: озонизација, реверсна осмоза, адсорпција, пермеат, нанофилтрација

Received: 01 October 2017. Accepted: 08 November 2017.

EDITORIAL POLICY

The journal Acta Periodica Technologica (formerly Proceedings of Faculty of Technolo-gy) publishes reviews and original scientific papers covering from all branches of technology (food, chemical, biochemical, and pharmaceutical), process engineering and related scientific fields. Acta Periodica Technologica is an Open Access journal. Contributions to journal shall be submitted in English, with summaries in English and Serbian which is closely defined with Instruction for manuscript preparation. The Journal is issued once a year. The journal is indexed in Chemical Abstracts, Columbus, Ohio, Referativnyi zhurnal -Khimija, VINITI, Moscow, Ulrich’s International Periodical Directory, and Elsevier Biblio-graphic databases - SCOPUS.

Editorial responsibilities The Editorial Board is responsible for deciding which articles submitted to Acta Periodica Technologica will be published. The Editorial Board is guided by the Editorial Policy and constrained by legal requirements in force regarding libel, copyright infringement and pla-giarism. The Editor-in-Chief reserves the right to decide not to publish submitted manuscripts in case it is found that they do not meet relevant standards concerning the content and formal aspects. The Editorial Staff will inform the authors whether the manuscript is accepted for publication within 60-90 days from the date of the manuscript submission. Editor-in-Chief must hold no conflict of interest with regard to the articles they consider for publication. If an Editor feels that there is likely to be a perception of a conflict of interest in relation to their handling of a submission, the selection of reviewers and all decisions on the paper shall be made by the Editorial Board. Editor-in-Chief and Editorial Board shall evaluate manuscripts for their intellectual content free from any racial, gender, sexual, religious, ethnic, or political bias. The Editor and the Editorial Staff must not use unpublished materials disclosed in sub-mitted manuscripts without the express written consent of the authors. The information and ideas presented in submitted manuscripts shall be kept confidential and must not be used for personal gain. Editors and the Editorial Staff shall take all reasonable measures to ensure that the re-viewers remain anonymous to the authors before, during and after the evaluation process.

Authors’ responsibilities

Authors warrant that their manuscript is their original work, that it has not been published before and are not under consideration for publication elsewhere. Parallel submission of the same paper to another journal constitutes a misconduct and eliminates the manuscript from consideration by Acta Periodica Technologica.

The Authors also warrant that the manuscript is not and will not be published elsewhere after the publication in Acta Periodica Technologica in any language without the consent of the Publisher. In case a submitted manuscript is a result of a research project, or its previous version has been presented at a conference in the form of an oral presentation (under the same or similar title), detailed information about the project, the conference, etc. shall be provided in Acknowledgement at the end of paper, and before List of references. A paper that has already been published in another journal cannot be reprinted in Acta Periodica Technologica. It is the responsibility of each author to ensure that papers submitted to Acta Periodica Technologica are written with ethical standards in mind. Authors affirm that the article con-tains no unfounded or unlawful statements and does not violate the rights of third parties. The Publisher will not be held legally responsible should there be any claims for compensation. Reporting standards. A submitted manuscript should contain sufficient detail and referen-ces to permit reviewers and, subsequently, readers to verify the claims presented in it. The de-liberate presentation of false claims is a violation of ethical standards. Authors are exclusively responsible for the contents of their submissions and must make sure that they have permission from all involved parties to make the data public. Authors wishing to include figures, tables or other materials that have already been pub-lished elsewhere are required to obtain permission from the copyright holder(s). Any material received without such evidence will be assumed to originate from the authors. Authorship. Authors must make sure that all only contributors who have significantly contributed to the submission are listed as authors and, conversely, that all contributors who have significantly contributed to the submission are listed as authors. If persons other than au-thors were involved in important aspects of the research project and the preparation of the manuscript, their contribution should be acknowledged in a Acknowledgements section. Acknowledgment of Sources. Authors are required to properly cite sources that have sig-nificantly influenced their research and their manuscript. Information received in a private conversation or correspondence with third parties, in reviewing project applications, manu-scripts and similar materials, must not be used without the express written consent of the in-formation source. Plagiarism. Plagiarism, where someone assumes another's ideas, words, or other creative expression as one's own, is a clear violation of scientific ethics. Plagiarism may also involve a violation of copyright law, punishable by legal action. Plagiarism includes the following:

Word for word, or almost word for word copying, or purposely paraphrasing por-tions of another author's work without clearly indicating the source or marking the copied fragment (for example, using quotation marks);

Copying equations, figures or tables from someone else's paper without properly ci-ting the source and/or without permission from the original author or the copyright holder.

Please note that all submissions are thoroughly checked for plagiarism. Any paper which shows obvious signs of plagiarism will be automatically rejected and authors will be temporarily forbidden to publish in the journal.

In case plagiarism is discovered in a paper that has already been published by the journal, it will be retracted in accordance with the procedure described below under Retraction policy, and authors will be temporarily forbidden to publish in the journal. Conflict of interest. Authors should disclose in their manuscript any financial or other sub-stantive conflict of interest that might have influenced the presented results or their interpret-tation. Fundamental errors in published works. When an author discovers a significant error or inaccuracy in his/her own published work, it is the author’s obligation to promptly notify the journal Editor or publisher and cooperate with the Editor to retract or correct the paper. By submitting a manuscript the authors agree to abide by the Acta Periodica Techno-logica’s Editorial Policies.

Reviewers’ responsibilities

Reviewers are required to provide written, competent and unbiased feedback in a timely manner on the scholarly merits and the scientific value of the manuscript. The reviewers assess manuscript for the compliance with the profile of the journal, the relevance of the investigated topic and applied methods, the originality and scientific relevan-ce of information presented in the manuscript, the presentation style and scholarly apparatus. Reviewers should alert the Editor to any well-founded suspicions or the knowledge of possible violations of ethical standards by the authors. Reviewers should recognize relevant published works that have not been cited by the authors and alert the Editor to substantial similarities between a reviewed manuscript and any manuscript published or under conside-ration for publication elsewhere, in the event they are aware of such. Reviewers should also alert the Editor to a parallel submission of the same paper to another journal, in the event they are aware of such. Reviewers must not have conflict of interest with respect to the research, the authors and/or the funding sources for the research. If such conflicts exist, the reviewers must report them to the Editor without delay. Any selected referee who feels unqualified to review the research reported in a manuscript or knows that its prompt review will be impossible should notify the Editor without delay. Reviews must be conducted objectively. Personal criticism of the author is inappropriate. Reviewers should express their views clearly with supporting arguments. Any manuscripts received for review must be treated as confidential documents. Re-viewers must not use unpublished materials disclosed in submitted manuscripts without the express written consent of the authors. The information and ideas presented in submitted manuscripts shall be kept confidential and must not be used for personal gain.

Peer review The submitted manuscripts are subject to a peer review process. The purpose of peer review is to assists the Editor-in-Chief and Editorial Board in making editorial decisions and through the editorial communications with the author it may also assist the author in impro-ving the paper.

All papers submitted to the journal will be reviewed by at least two independent referees who will be asked to complete the refereeing job within 2-4 weeks. Final decision on public-cation will be made by the Editorial Board. The choice of reviewers is at the Editors' discretion. The reviewers must be knowledgeab-le about the subject area of the manuscript; they must not be from the authors' own institution and they should not have recent joint publications with any of the authors. In a main phase of peer review process, the reviewers must fill reviewer’s form which indicates which aspects to be covered in order to make decision about manuscript public-cation. In the final part of the form, reviewers must submit their observations and suggestions on how to submitted manuscript improve. All of the reviewers of a paper act independently and they are not aware of each other’s identities. If the decisions of the two reviewers are not the same (accept/reject), the Editor may assign additional reviewers. During the review process Editor may require authors to provide additional information (including raw data) if they are necessary for the evaluation of the scholarly merit of the manuscript. These materials shall be kept confidential and must not be used for personal gain. The Editorial team shall ensure reasonable quality control for the reviews. With respect to reviewers whose reviews are convincingly questioned by authors, special attention will be paid to ensure that the reviews are objective and high in academic standard. When there is any doubt with regard to the objectivity of the reviews or quality of the review, additional re-viewers will be assigned.

Procedures for dealing with unethical behaviour

Anyone may inform the editors and/or Editorial Board at any time of suspected unethical behaviour or any type of misconduct by giving the necessary information/evidence to start an investigation. Investigation. Editor-in-Chief will consult with the Editorial Board on decisions regarding the initiation of an investigation. During an investigation, any evidence should be treated as strictly confidential and only made available to those strictly involved in investigating. The accused will always be given the chance to respond to any charges made against them. If it is judged at the end of the investigation that misconduct has occurred, then it will be classified as either minor or serious. Minor misconduct. Minor misconduct will be dealt directly with those involved without involving any other parties, e.g.:

- Communicating to authors/reviewers whenever a minor issue involving misunderstanding or misapplication of academic standards has occurred.

- A warning letter to an author or reviewer regarding fairly minor misconduct. Major misconduct. The Editor-in-Chief, in consultation with the Editorial Board, and, when appropriate, further consultation with a small group of experts should make any deci-sion regarding the course of action to be taken using the evidence available. The possible out-comes are as follows (these can be used separately or jointly):

- Publication of a formal announcement or editorial describing the misconduct.

- Informing the author's (or reviewer's) head of department or employer of any misconduct by means of a formal letter.

- The formal, announced retraction of publications from the journal in accordance with the Retraction Policy (see below).

- A ban on submissions from an individual for a defined period. - Referring a case to a professional organization or legal authority for further

investigation and action. When dealing with unethical behaviour, the Editorial Board will rely on the guidelines and recommendations provided by the Committee on Publication Ethics (COPE): http://publica-tionethics.org/resources/.

Retraction policy

Legal limitations of the publisher, copyright holder or author(s), infringements of profess-sional ethical codes, such as multiple submissions, bogus claims of authorship, plagiarism, fraudulent use of data or any major misconduct require retraction of an article. Occasionally a retraction can be used to correct errors in submission or publication. The main reason for withdrawal or retraction is to correct the mistake while preserving the integrity of science; it is not to punish the author. Standards for dealing with retractions have been developed by a number of library and scholarly bodies, and this practice has been adopted for article retraction by Acta Periodica Technologica: in the electronic version of the retraction note, a link is made to the original article. In the electronic version of the original article, a link is made to the retraction note where it is clearly stated that the article has been retracted. The original article is retained unchanged, save for a watermark on the PDF indicating on each page that it is “retracted.”

Open access policy

Acta Periodica Technologica is an Open Access Journal. All articles can be downloaded free of charge from: http://www.doiserbia.nb.rs/journal.aspx?issn=1450-7188 or http://www.tf.uns.ac.rs/site/index.php/sr-lat/delatnost/publikacije and used with therms de-fined with Creative Commons licence (http://creativecommons.org/licenses/by-nc-nd/3.0/rs). The journal does not charge any fees at submission, reviewing, and production stages.

Self-archiving Policy The journal Acta Periodica Technologica allows authors to deposit Publisher’s ver-sion/PDF in an institutional repository and non-commercial subject-based repositories, such as PubMed Central, Europe PMC or arXiv, or to publish it on Author's personal website (in-cluding social networking sites, such as ResearchGate, Academia.edu, etc.) and/or depart-mental website, at any time after publication. Full bibliographic information (authors, article title, journal title, volume, issue, pages) about the original publication must be provided and a link must be made to the article's DOI.

Copyright Once the manuscript is accepted for publication, authors shall transfer the copyright to the Publisher. If the submitted manuscript is not accepted for publication by the journal, all rights shall be retained by the author(s). Authors grant to the Publisher the following rights to the manuscript, including any supplemental material, and any parts, extracts or elements thereof:

- the right to reproduce and distribute the Manuscript in printed form, including print-on-demand;

- the right to produce prepublications, reprints, and special editions of the Manuscript; - the right to translate the Manuscript into other languages; - the right to reproduce the Manuscript using photomechanical or similar means in-

cluding, but not limited to photocopy, and the right to distribute these reproductions; - the right to reproduce and distribute the Manuscript electronically or optically on any

and all data carriers or storage media - especially in machine readable/digitalized form on data carriers such as hard drive, CD-Rom, DVD, Blu-ray Disc (BD), Mini-Disk, data tape - and the right to reproduce and distribute the Article via these data carriers;

- the right to store the Manuscript in databases, including online databases, and the right of transmission of the Manuscript in all technical systems and modes;

- the right to make the Manuscript available to the public or to closed user groups on individual demand, for use on monitors or other readers (including e-books), and in printable form for the user, either via the internet, other online services, or via internal or external networks.

The authors and third parties who wish use the article in a way not covered by the Creative Common licence (http://creativecommons.org/licenses/by-nc-nd/3.0/rs) must obtain a written consent of the publisher. Contact e-mail for written consent is: [email protected]. Authors grant to the publisher the right to publish the article, to be cited as its original publisher in case of reuse, and to distribute it in all forms and media.

Disclaimer The views expressed in the published works do not express the views of the Editor-in-Chief and Editorial Board. The authors take legal and moral responsibility for the ideas expressed in the articles. Publisher shall have no liability in the event of issuance of any claims for damages. The Publisher will not be held legally responsible should there be any claims for compensation.

INSTRUCTION FOR MANUSCRIPT PREPARATION

Acta Periodica Technologica publishes reviews and scientific papers covering all branches of technology: food, chemical, biochemical, pharmaceutical, as well as process engineering and related scientific fields. Acta Periodica Technologica is published in English. The journal may include supplements from congresses, meetings or symposiums. SUBMISSION OF PAPERS All correspondence, including submission of the manuscript, notification of the Editor’s decision and requests for revision, takes place by e-mail [email protected]. Authors are expected to propose the category of manuscript (review or original sci-entific paper) and three potential reviewers. Reviewers should be experts in the field of the paper, and not associated with the institution with which the authors are affiliated. The final choice of referees will remain entirely with the Editor. Also, optionally, the authors should state any person that is not desired as a reviewer. Submission of paper implies that:

- it is prepared according this Instructions, - it has not been published previously (except in the form of an abstract or as a

whole in the proceedings of papers of a scientific meeting, or as part of a pub-lished lecture or academic thesis),

- it is not under consideration for publication elsewhere, and - it will not be published elsewhere in the same form, in English or in any other

language, without the written consent of the publisher. PREPARATION OF MANUSCRIPTS

Language: Manuscript should be written in English. Typing: Manuscript must be written in Word with a font size 10 pt, double spaced with 2,5 cm margins, on A4 pages (max. 10 pages for scientific papers). All pages of the manuscript should be numbered. Import tables and figures into the text. Abbreviations and symbols-notation should be explained at first appearing, or on a separate list at the end of manuscript. General format. The manuscript should contain the following in this order: Title page, ABSTRACT and KEY WORDS, INTRODUCTION, EXPERIMENTAL, RE-SULTS and DISCUSSION, CONCLUSION, ACKNOWLEDGEMENT and REFEREN-CES as well as ABSTRACT and KEY WORDS IN SERBIAN LANGUAGE. Title page: On the first page should be the title without symbols, formulae or abbre-viations (capital bold letters). Title should be concise and explanatory of the content of the paper. Full name (name, initial and surname) of authors (without degrees, professio-nal or official titles) should be given under the title, written in italic. Clearly indicate (with asterix) who is responsible for correspondence at all stages of refereeing and publication. Ensure that e-mail address and the full postal address are provided. Affilia-tion of authors should be given after the author’s name. Indicate all affiliations with

superscript number immediately after authors name and in front of appropriate address. If the paper was given, wholly or in part, at a scientific meeting, this should be stated in a footnote on the title page. Abstract of the paper (100-250 words, written in italic) should be given under the title and authors. Abstracts should contain the aim of investigated work, methods, results and conclusion. Key words (normal letters, max. 5 key words) should be listed afterwards. Introduction should state previous relevant work with appropriate references, the problem investigated and the aim of work. Experimental. The materials and methods used should be stated clearly in sufficient detail to permit the work to be repeated by others. Only new techniques should be de-scribed in detail; known methods must have adequately references. Results and Discussion. Results should be presented concisely, with tables or illu-strations for clarity. The significance of the findings should be discussed without repe-tition of the material in the Introduction. Adequate number of illustrations, graphs and chemical formulae used must be kept on minimum. Conclusion. This section should present the main conclusions of the study. Also, conclusions should indicate the significance of contribution and application possibilities of the obtained results. Acknowledgements: These should be kept to a minimum. References cited should be indicated in the text using Arabic numerals in brackets ( ), in the order of appearing. All publications cited in the text should be presented in a list of references given on a separate page. Abbreviations of journal titles should be given according to the Chemical Abstracts Service (CASSI Search Tool; http://cassi.cas.org). The list of references should be presented according to ACS citation style and their appearance in the text. Give names of all authors (do not use „et.al.“), with their initials after respective surnames. Include article titles in journals. The abbreviated titles should be followed by the year (bold), volume (italic), number (in brackets if exists), and first and last page numbers. Examples: Journals: Pascual, E.C.; Goodman, B.A; Yeretzian, C. Characterisation of Free Radicals in Solubile Coffee by Electron Paramagnetic Resonance Spectroscopy. J. Agric. Food Chem. 2002, 50 (21), 6114-6122.

Books: Morris, R. The Last Sorcerers: The Path from Alchemy to the Periodic Table; Joseph Henry Press: Washington, DC, 2003; pp 145-158.

Book with more chapters: Puls, J.; Saake, B. Industrially Isolated Hemicelluloses. In Hemicelluloses: Science and Technology; Gatenholm, P., Tenkanen, M., Eds.; ACS Symposium Series 864; American Chemical Society: Washington, DC, 2004; pp 24-37.

Book of Abstracts: Noe, W.; Howaldt, M.; Ulber, R.; Scheper, T. Immunobase elution assay for process control, 8th European Congress on Biotechnology, Budapest, 17-21 August 1997, Book of Abstracts WE 163, p. 246.

Thesis: Linstead, J.B.: Linstead, J.B. Effects of adding natural antioxidants on colour stability of paprika. Ph.D. (or M.S.) Thesis, University of Glasgow, November 2006.

Patent: Lenssen, K. C.; Jantscheff, P.; Kiedrowski, G.; Massing, U. Cationic Lipids with Serine Backbone for Transfecting Biological Molecules. Eur. Pat. Appl. 1457483, 2004.

Unpublished data: Should be cited with one of the following comments: in press, un-published work or personal communication.

Online citations: Should include the author, title, website and date of access. Example: Wright, N.A. The Standing of UK Histopathology Research 1997-2002. http://pathsoc.org.uk (accessed 7 October 2004).

Abstract and key words in Serbian language should be given at the end of manu-script (after references), in extended form (max. length 1 page), printed in Cyrillic (nor-mal letters) with the title (capital letters), full name(s) of each author(s) and affiliation(s) (italic letters). For authors outside Serbia, the Editorial Board will provide a Serbian translation of their English abstract. Chemical nomenclature and units. Authors are requested to use SI units and che-mical nomenclature following the rules of Chemical Abstracts whenever possible. Tables. Each Table is numbered with Arabic numeral, followed by the title (Table 1. Result...). The table width must be 12.5 cm max. Figures. Each drawing or figure should also be numbered with Arabic numerals followed by the title (Figure 1. Chromatogram of...). Graphs and charts must be prepared by Microsoft Excel or Origin. Schemes must be prepared by Microsoft Visio or Corel Draw. It is necessary to submit them as separate files in original extension (xls, xlsx, vdr, cdr). Scanned black & white schemes should be submitted in tif, wmf, or bmp form. Photographs should be submitted in jpg form. Formulae and Equations. Type formulas and mathematical equations clearly, accu-rately placing superscripts and subscripts. Equations should be indicated in the text using Arabic numerals in square brackets [ ]. Review process. All papers submitted to the journal will be reviewed by at least two independent referees who will be asked to complete the refereeing job within 2-4 weeks. Final decision on publication will be made by the Editorial Board. Manuscripts may be sent back to authors for revision if necessary. Revised manuscript submissions should be made as soon as possible (within 2 weeks) after the receipt of the referees comments. Proofs. One set of page proofs will be sent by e-mail to the corresponding Author. Please use this proof only for checking the typesetting, editing, completeness and cor-rectness of the manuscript. The author may list the corrections and return to the journal in an e-mail within 48 hours of receipt. Author service. For inquiries relating to the submission of manuscript, please send an e-mail to the Editor ([email protected]). Postal address: Acta Periodica Technologica, Editorial Board, Bulevar cara Lazara 1, 21000 Novi Sad, Serbia.

FORMER EDITORS-IN-CHIEF

Prof. Dr. Adalbert Šenborn (1967-1970) Prof. Dr. Radivoj Žakula (1972-1975)

Prof. Dr. Miroslava Todorović (1976-1994) Prof. Dr. Biljana Škrbić (1995-1998)

Prof. Dr. Sonja Đilas (1999-2016)

THIS ISSUE OF ACTA PERIODICA TECHNOLOGICA

IS FINANCIALLY SUPPORTED BY:

Ministry of Education, Science and Technological Development of Republic of Serbia

Editorial:


Phone: +381 21 485 3693

Fax:+381 21 450 413 e-mail: [email protected]

Proofreader: Prof. Dr. Luka Bjelica Prepress: Branislav S. Bastaja Cover design: Živojin Katić

Printed by Futura d.o.o., Petrovaradin Copies: 200

Articles published in the Acta Periodica Technologica are Open-Access articles distributed under a license Creative Commons BY-NC-ND 3.0 Serbia (http://creativecommons.org/licenses/by-nc-nd/3.0/rs)

acta periodica technologica - početna · acta periodica technologica apteff, 48, 1-323 (2017)...

Documents