characterization and implications of enterobacter cloacae strains, isolated from italian table...

M:Fo

odMi

crob

iolog

y&

Safet

y

JFS M: Food Microbiology and Safety

Characterization and Implicationsof Enterobacter cloacae Strains, Isolated fromItalian Table Olives “Bella Di Cerignola”ANTONIO BEVILACQUA, MARIANNA CANNARSI, MARIANGELA GALLO, MILENA SINIGAGLIA, AND MARIA ROSARIA CORBO

ABSTRACT: Enterobacter cloacae can be recovered in the spontaneous fermentations of Italian table olives. In thisstudy, the effects of salt (20 to 100 g/L), temperature (10 to 37 ◦C), pH (4 to 5 and 8 to 10), p-coumaric and vanillicacids (0.5 to 2 g/L), and the acidification of the medium through lactic, citric, and ascorbic acids were investigated on15 strains of E. cloacae, isolated from Italian table olives “Bella di Cerignola.” Finally, a confirmatory experiment insynthetic brine was run. The strains were inhibited only by an NaCl amount of 70 to 80 g/L and by p-coumaric acid;on the other hand, they showed the ability to grow also at low temperatures (10 to 15 ◦C). The confirmatory experi-ment highlighted their ability to survive both at 15 ◦C and at pH 5. Enterobacter cloacae could be a real problem forthe fermentation of table olives in southern Italy; some hurdles could be used (salt or brine acidification), but someenvironmental conditions (for example, the temperature) should be controlled carefully to maintain olive safety atacceptable levels.

Keywords: Enterobacter cloacae, organic acids, pH, phenols, salt, table olives

Introduction

Table olives are the most important fermented vegetables in thewestern countries and the Intl. Olive Oil Council estimated that

world production of this kind of fermented food reached around1,823,000 tons in the year 2006/07 (IOOC 2009).

There are 3 main styles for olive processing: Spanish (or Sevil-lian) method for green olives; Greek (or Natural) style for naturallyblack olives, and Californian method for black ripe olives. A de-scription of each method could be easily found in the literature(Garrido-Fernandez and others 1997).

The fermentation of table olives both in Spanish and Natu-ral styles relies on a complex microflora. In particular, in Span-ish style the fermentation course could be divided into 3 or4 different phases; the 1st stage involves the growth of En-terobacteriaceae and some other Gram-negative microorganisms(Garrido-Fernandez and others 1997). During the period of ele-vated pH (1st stage of fermentation), a high risk of spoilage exists,especially if that period is prolonged (Holpzapfel 2001); Enterobac-teriaceae, in fact, could cause spoilage through the formation ofgas pockets in the olive surface or from the production of metabo-lites, that affect the aroma of the product (Garrido-Fernandez andothers 1997). In order to prevent these problems, Chorianopoulosand others (2005) suggested to control their growth at the 1st stagesof fermentation by rapidly dropping the pH below 5.

“Bella di Cerignola” is a traditional variety of Apulian region(southern Italy) that is gaining the favor of Italian consumers. It isappreciated, in fact, for the size of the olives and the green/blackcolor of the surface; in the last years, this olive variety received

MS 20090336 Submitted 4/15/2009, Accepted 10/22/2009. AuthorsBevilacqua, Cannarsi, Gallo, Sinigaglia, and Corbo are with Dept. of FoodScience, Faculty of Agricultural Science. Authors Bevilacqua, Sinigaglia,and Corbo are with Istituto per la Ricerca e le Applicazioni Biotecnologicheper la Sicurezza e la Qualita, Univ. of Foggia, Italy. Direct inquiries toauthor Corbo (E-mail: [email protected]).

the protected denomination of origin (PDO) from the EuropeanUnion (EU) (the full name is Bella della Daunia variety Bella diCerignola). Nowadays, the fermentation of table olives Belladi Cerignola relies on a system of small-scale producers, usuallyperforming it in an artisanal way, without the addition of startersand sugars; therefore, the knowledge of the microorganisms in-volved in spontaneous fermentation is the 1st step for the improve-ment of this product.

As reported by Jasche-Ferreri (2008), Bella di Cerignola olives aremanually harvested; the picked up fruits are released into plasticbug boxes (20 kg), ready for the transport to the factories, wherethey are processed within 48 h. At the arrival, olives are subjected tosorting and size grading. The association of PDO Bella di Cerignolaproducers defined the following size scale: GP, 181 to 280 fruits/kg; GM, 121 to 180 fruits/kg; G, 91 to 120 fruits/kg; GG, 81 to90 fruits/kg; GGG, 71 to 80 fruits/kg.

Then, the olives are treated with lye (1.3% to 2.6%, dependingon the size and fruit ripeness) for 12 to 15 h; this step is generallyfollowed by 2 washing treatments (the 1st for 2 to 3 h and the 2nd for10 to 12 h). Finally, olives are brined and the NaCl at the beginningis approximately 60 to 80 g/L and increases gradually up to 100 g/Luntil the end of fermentation (30 to 60 d).

In a preliminary phase, we studied the evolution of the natu-rally occurring microflora of Bella di Cerignola olives (Campanielloand others 2005; Bevilacqua and others 2008a, 2008d) and recov-ered in brines a cell number of Enterobacteriaceae ranging between2 and 6 log CFU/mL, both at the beginning and after 30 to 40 d offermentation. The identification at species level pointed out thatenterobacteria belonged mainly to the species E. cloacae (data notpublished).

Some data are available on the technological characterizationof lactic acid bacteria or on the spoiling impact of yeasts on Belladi Cerignola olives (Campaniello and others 2005; Altieri and oth-ers 2006; Bevilacqua and others 2008b, 2008c); on the other hand,no paper could be found in the literature regarding the impact of

C© 2009 Institute of Food Technologists R© Vol. 75, Nr. 1, 2010—JOURNAL OF FOOD SCIENCE M53doi: 10.1111/j.1750-3841.2009.01445.xFurther reproduction without permission is prohibited

M:FoodMicrobiology&

Safety

Enterobacter cloacae in Italian table olives . . .

Gram-negative microflora, namely the Enterobacteriaceae, on thesafety and organoleptic characteristics of this kind of food.

It has been reported that E. cloacae, as well as other speciesbelonging to the genus Enterobacter, is widespread and has beenisolated from vegetables (Bennik and others 1998; Hamilton-Miller and Shah 2001; Soriano and others 2001; Coulin and others2006).

The isolation of E. cloacae from vegetables, in particular fromBella di Cerignola olives, raises the issue of its impact for the qual-ity and the safety of this kind of food. It is well known that someE. cloacae strains could be opportunistic pathogens (Kanemitsuand others 2007); in addition, it has been reported that enterobac-teria colonize olives within the 1st fermentation days and theirnumber decreases when the lactic acid bacteria prevail. However,this could be not true for not-controlled fermentation processes, asreported elsewhere (Bevilacqua and others 2008a, 2008d). Anotherelement to keep in mind is that for some foods EU considers enter-obacteria as the target microorganism for assessing quality (Regu-lation nr 2073/2005. EU 2005).

Therefore, this research was aimed to study the technologicalcharacteristics of some E. cloacae strains, isolated from Bella diCerignola table olives, in order to give an answer to the followingquestion: does E. cloacae grow really under the environmental con-ditions used for the fermentation of olives?

This is the 1st step for assessing its role in the fermentation oftable olives, in particular for Bella di Cerignola table olives. Thestudy was divided into 2 phases: in the 1st phase, the assays wereperformed in a laboratory medium on 15 different strains; then,1 strain (the most vigorous one) was used to confirm the results inbrine.

Materials and Methods

StrainsThis study focused on 15 strains of E. cloacae, isolated from Ital-

ian table olives Bella di Cerignola. The strains were isolated in aprevious phase from a spontaneous fermentation of table olives,processed with lye (Spanish style); after some preliminarytest (Gram staining, catalase, and oxidase activity, respiration/fermentation of glucose, motility, ability to produce spores), thestrains were identified through the miniaturized system API 20E(Biomerieux Marcy L’Etoile France). The profile did not show not-typical tests and was classified as “good” or “excellent” by the iden-tification software; the percentage of identification was at least99%. The identification at species level was confirmed throughsome complementary assays, that is, lysine decarboxylase, or-nithine decarboxylase, arginine dihydrolase, glucose dehydrolase,gluconate dehydrolase, production of a yellow pigment, as reportedby Grimont and Grimont (2006).

The microorganisms were labeled with a numeric code andmaintained at 4 ◦C on Plate Count Agar slants (PCA) (Oxoid, MilanItaly). The strains were revitalized in Nutrient Broth (NB) (Oxoid),incubated at 37 ◦C for 24 h.

Technological characterizationThe experiments were performed into test tubes, containing

10 mL of NB, inoculated separately with approximately 105 CFU/mL of each strain. Different assays were performed, as follows:

1. Influence of salt. NB composition was modified through theaddition of NaCl (20 to 100 g/L).

2. Influence of pH . The pH of NB was adjusted to 4, 4.5, 5, 8, 8.5,9, 9.5, and 10 with HCl or NaOH 1N; the pH of the medium was ad-justed under sterile conditions after autoclaving. Nor color changes

neither precipitation of the ingredients of the medium wereobserved in the pH range used throughout this study.

3. Growth at different temperature. The experiments were per-formed in not-modified NB, incubated at 10, 15, 20, 25, and 37 ◦C.

4. Influence of secondary phenolic compounds. The assays wereperformed in NB added separately with 0.5, 1, 1.5, or 2 g/L ofp-coumaric (p-hydroxycinnamic acid, Sigma-Aldrich, Milan, Italy)and vanillic acids (p-hydroxycinnamic acid, Sigma-Aldrich).

5. Acidification with different organic acids. The assay was per-formed in NB, acidified to pH 4.5 and 5 with lactic, citric, and ascor-bic acids (J.T. Baker, Milan). The acids were dissolved in distilledwater in order to achieve a 1 : 1 concentration (w/w). Table 1 re-ports the volumes of acid solutions added to 100 mL of NB.

The samples were incubated at 37 ◦C or at 10 to 37 ◦C (assay 3)and the growth of the microbial tests were evaluated periodicallyby measuring the absorbance at 600 nm using a UV-VIS BECKMANDU 600 spectrophotometer (Fullerton, Calif., U.S.A.). Aliquots ofnot-modified NB, inoculated with the different strains of E. cloacaeand incubated at 37 ◦C, were used as controls.

Survival of E. cloacae in synthetic brineThe strain 3 was used as the microbial test and the assays were

performed in a brine, prepared with sterile tap water and contain-ing 70 g/L of NaCl and 1 g/L of glucose (J.T. Baker). The experimentswere performed in 50-mL sterile tubes, containing 35 mL of brine,added with p-coumaric or vanillic acids or adjusted to pHs 5, 6, and10 through NaOH 1N or lactic acid (1 : 1) (Table 2). Then, the sam-ples were inoculated with an appropriate volume of the microbialtest, in order to achieve a final inoculum of 5 to 6 log CFU/mL, andincubated at 15, 25, or 37 ◦C, as reported in the Table 2.

The samples were periodically analyzed to evaluate the viablecell count on Violet Red Bile Glucose Agar (VRBGA) (Oxoid), incu-bated at 37 ◦C for 18 to 24 h.

Statistical analysis and data modelingAll the analyses were performed in duplicate on 2 different

batches, labeled as A and B; for each batch the experiments wereperformed twice (n = 4). Data were submitted to one-way analysisof variance (ANOVA) and Fisher’s least significant difference (LSD)test (P < 0,05), through the software Statistica for Windows, version6.0 (Statsoft, Tulsa, Okla., U.S.A.).

Table 1 --- Volumes (μL) of acid solutions added to 100 mLof NB to achieve pH values of 4.5 and 5.0.

pH 4.5 pH 5.0

Ascorbic acid∗ 186 63Citric acid 80 29Lactic acid 108 60∗Acid solutions were prepared in distilled water; the concentration was 1 : 1 (w/w).

Table 2 --- Combinations of synthetic brine (70 g/L NaCl +glucose 1g/L).

Temperature p-coumaric VanillicSamples (◦C) pH acid (g/L) acid (g/L)

A 15 6 - -B 25 6 - -C 25 10 - -D 15 5 - -E 25 5 - -F 15 6 0.1 -G 25 6 0.1 -H 15 6 - 0.1I 25 6 - 0.1L 15 5 0.1 -M 37 6 - -

M54 JOURNAL OF FOOD SCIENCE—Vol. 75, Nr. 1, 2010

M:Fo

odMi

crob

iolog

y&

Safet

y


The results of the technological assays were modeled as GrowthIndex (GI) (Blaszyk and Holley 1998), modified as follows:

G I = (Abss/Absc) × 100

where for each time, Abss is the absorbance value of the sampleswith different amounts of NaCl, pHs or incubated at 10 to 25 ◦C;otherwise Absc is the absorbance value of the control samples.

Results

Effect of the saltFigure 1 reports GI values experienced by E. cloacae strains in NB

added with 50 (Figure 1A) and 70 g/L of NaCl (Figure 1B), after 24and 48 h. An amount of NaCl of 50 g/L caused a significant inhi-bition of the microbial targets after 24 h, as evidenced by GI values,ranging from 34.68% (strain 6) to 68.49% (strain 52); in a longer stor-age time (48 h), GI did not undergo significant changes, except forstrain 52 that expressed a higher value (88.52%). An increase of thesalt content of the medium up to 70 g/L (Figure 1B) caused a strong

inhibition of E. cloacae growth, as evidenced by GI (15% to 20% af-ter 24 h and approximately 30% to 35% after 48 h of incubation); nosignificant differences were recovered amongst the tested strains.

Figure 2 shows growth profile of the strains 6 and 52 (respectivelythe most sensitive and the most resistant strains to NaCl), com-pared with the salt amount in the medium.

Strain 52 showed a growth similar to that observed for the controlsample up to 40 g/L of NaCl; a significant inhibition was observedin NB + 50 g/L of NaCl, with a dose-dependant effect of the salt:a further increase of the NaCl amount, in fact, resulted in a signif-icant reduction of GI value, until to attain the complete inhibitionof the target in NB + 80 g/L of NaCl. On the other hand, Strain 6was affected also at the lowest amount of NaCl (20 g/L), showing anegative correlation with the concentration of the salt.

Effect of pHIn Spanish style processing, the treatment of olives with lye is a

critical step for the quality of the product and for a correct courseof the lactic fermentation; throughout this phase, pH of olives andbrines is approximately 9.5 to 10 (Campaniello and others 2005). A

0102030405060708090

100

3 4 4* 6 6* 7* 12 12* 17 27 33 40 46 50 52

Strains

Gro

wth

Ind

ex (

%)

24 h

48 h

A

0102030405060708090

100

3 4 4* 6 6* 7* 12 12* 17 27 33 40 46 50 52

Strains

Gro

wth

Ind

ex (

%)

24 h48 h

B

Figure 1 --- Growth index experiencedby E. cloacae strains after 24 and48 h in NB added with 50 (A) and70 g/L (B) of NaCl. Data are theaverage (n = 4) ± standard deviation.

0

20

40

60

80

100

120

140

160

0 20 40 60 80 100 120

NaCl (g/L)

Gro

wth

Ind

ex (

%)

strain 6

strain 52

Figure 2 --- Growth Index experiencedby strains 6 and 52 compared withNaCl amount after 24 h of incubation.Data are the average (n = 4) ±standard deviation.

Vol. 75, Nr. 1, 2010—JOURNAL OF FOOD SCIENCE M55

M:FoodMicrobiology&

Safety


sufficient number of washings can assure a quick decrease of thepH and the following colonization of brines and olive surface bylactic acid bacteria. What happens if the washing is not correct (fewwashings for a short time, as managed by some olive producers inItaly due to the lack of water)?

The lactic acid bacteria could have some problems to colonizethe brines and different microorganisms could prevail; therefore,the ability of E. cloacae to grow at high pHs was investigated.Figure 3 shows GI values after 24 and 48 h of incubation in NB, ad-justed to pH 10. As expected, the high pH is a stressful element forcells; it caused, in fact, a significant reduction of GI, ranging be-tween 30% and 70%.

The growth profiles of 3 selected strains (3, 6, and 52, chosenas representative of the 15 isolates) compared with the pH are re-ported in Figure 4. Strain 52 appeared as the most resistant isolate,as it experienced a GI of 100% both at pH 5.5 and 8; otherwise, theisolates 6 and 3 were more sensitive. It is important to underlinethat growth was completely inhibited at pH 4 and inhibited in a sig-nificant manner at pH 4.5: the effect was strong for the strain 3 andmoderate for the isolates 6 and 52.

Regarding the trend for alkaline pHs, the increase of the pHcaused a reduction of GI and the minimum value of this parameter(25% and 46% for the strains 3 and 52, respectively) was achieved atpH 10.

Growth at different temperaturesTemperature profiles were quite similar for all the strains; there-

fore, as an example, the profile of strain 52, at different incubationtimes, is shown in the Figure 5. This figure pointed out some inter-

esting data: (1) the microbial test was completely inhibited at 10 ◦C(GI, 0%) only for a short storage time (24 h), as GI increased up to52% after 96 h; (2) after 96 h the strain 52 experienced a GI of 100%at 15 ◦C and higher value at 20 and 25 ◦C (125% and 138%, respec-tively).

Effect of phenolsFigure 6 shows the GI after 24 h of incubation in NB added with

0.5 g/L of p-coumaric or vanillic acids. The effect of p-coumaricacid could be considered quite strong, as the strains experiencedGI values ranging between 3.69% (strain 4) and 16% (strain 4∗); onthe other hand, the effect of the vanillic acid was strain-dependant,with a GI ranging from a minimum of 16% (strong effect) (strain 6)to a maximum of approximately 71% (moderate effect) (strain 12).The growth of the 15 strains was completely inhibited when higheramounts of phenolic compounds were added to NB.

Acidification with different acidsTable 3 reports GI values experienced after 24 h of incubation at

37 ◦C by E. cloacae strains in NB acidified to pH 4.5, through dif-ferent acid solutions (ascorbic, citric, and lactic acids). GI values of49% to 112.96% and 52.21% to 81.70% were recovered when ascor-bic and citric acids were used; a further decrease was observed withthe lactic acid (8.60% to 44.60%).

As the lactic acid resulted in a stronger inhibition of the studiedstrains, GI values in NB acidified through lactic acid were used topoint out a susceptibility hierarchy.

Data were submitted to one-way ANOVA and Fisher’s LSD test;the homogeneous group approach was used (Altieri and others

0102030405060708090

100

3 4 4* 6 6* 7* 12 12* 17 27 33 40 46 50 52

Strains

Gro

wth

Ind

ex (

%)

24 h

48 h

Figure 3 --- Growth index experienced byE. cloacae strains after 24 and 48 h in NBadjusted to pH 10. Data are the average(n = 4) ± standard deviation.

0

20

40

60

80

100

120

4 5 6 7 8 9 10

pH

Gro

wth

Ind

ex (

%)

strain 3

strain 6

strain 52

Figure 4 --- Growth index of the strains3, 6, and 52 compared with pH after24 h of incubation. Data are theaverage (n = 4) ± standard deviation.


M:Fo

odMi

crob

iolog

y&

Safet

y


2008). The statistical analysis underlined that GI values were dis-tributed in a continuous way, as a superposition of the statisticalgroups was observed (Table 4); we could suggest, however, that GIvalues were distributed between a minimum of 8.60% (strain 3, in-cluded in a well-defined group statistically different from the oth-ers) and a maximum of 44.60%; moreover an intermediate groupof susceptibility was identified, with a GI ranging from 22.81% and36.51%. Therefore, the following susceptibility hierarchy could besuggested: strain 3 (high susceptibility, GI of 8.60%) > strains 40,50, 46, 4, 7∗, 27, 6, 33, 52, 17, 6∗, and 12∗ (moderate susceptibility, GIranging from 22.81% and 36.51%) ≥ strains 4∗ and 12 (moderate-to-low susceptibility) (GI of approximately 43% to 44%).

Survival in synthetic brineAs a final step of this study, the survival of a selected strain

(strain 3, chosen as the representative of the 15 isolates) was as-sessed in brine aliquots, incubated at different temperatures, ad-justed to pH 5 or 10 and added with 0.1 g/L of p-coumaric acid.The results of some combinations are shown in Figure 7; this pic-ture reports the population number as CFU/mL (viable count).The viable count decreased quickly in the combination C (brineadjusted to pH 10) up to the undetectable level just after 48 h;a similar trend was evidenced in the combination M (brine incu-bated at 37 ◦C). A significant decrease of viable count was observedalso in the sample G (0.1 g/L of p-coumaric acid and storage at25 ◦C) just after 48 h of incubation (4.58 log CFU/mL); at theend of the experiment, the viable count was approximately 1.9 logCFU/mL.

In the other 2 aliquots of brine, containing the phenolic com-pound (the samples F and L, incubated at 15 ◦C), cell numbers

changed in a little way within the entire running time (many timesthe changes were not significant).

In the remaining combinations (sample B, D, E, H, and I), strain 3did not undergo significant changes within the running time (datanot shown).

Discussion

Garrido-Fernandez and others (1997) reported that Gram-negative bacteria colonized brines within 48 to 72 h and

amongst these microorganisms the most important are Pseu-domonas spp., Flavobacterium spp., and some Enterobacteriaceae(as E. cloacae, Klebsiella aerogenes, and Escherichia coli). It is well

Table 3 --- Growth index experienced by E. cloacae strainsafter 24 h in nutrient broth, acidified to pH 4.5 throughascorbic, citric, and lactic acids (1 : 1, w/w). Data are theaverage (n = 4).

Strain code Ascorbic acid Citric acid Lactic acid

3 49.00a 60.87a 8.60b

4 71.43a 59.75b 29.12c

4∗ 99.35a 59.07b 44.60b

6 85.02a 71.42b 31.19c

6∗ 69.73a 63.67a 35.05b

7∗ 70.68a 57.64b 30.60c

12 93.66a 70.60b 42.76c

12∗ 80.63a 81.70a 36.51b

17 112.16a 78.35b 34.97c

27 110.96a 73.91b 30.70c

33 83.32a 68.85b 32.25c

40 66.43a 52.21b 22.81c

46 72.14a 57.30b 28.15c

50 78.48a 60.87b 23.24c

52 77.38a 55.36b 33.20c

a,bValue in a line with different letters are significantly different (one-way ANOVAand Fisher’s LSD test) (P < 0.05).

0

20

40

60

80

100

120

140

160

10 15 20 25 30 35 40

Temperature (°C)

Gro

wth

Ind

ex (

%)

24 h48 h72 h96 h

Figure 5 --- Temperature profile of thestrain 52 at different incubationtimes. Data are the average (n = 4) ±standard deviation.

0

10

20

30

40

50

60

70

80

90

3 4 4* 6 6* 7* 12 12* 17 27 33 40 46 50 52

Strains

Gro

wth

Ind

ex (

%)

p-coumaric acid

vanillic acid

Figure 6 --- Growth index experiencedby E. cloacae strains after 24 h in NBadded with 0.5 g/L of p-coumaric orvanillic acids. Data are the average(n = 4) ± standard deviation.


M:FoodMicrobiology&

Safety


known that E. cloacae is widespread, as it was isolated from fecesof humans and animals, water, soil, plants, plant materials, insects,and dairy products: however, its role and impact in olive fermen-tation is not known (Davidson and others 2000; Neto and others2003). Tsuda and others (2001) reported that E. cloacae is an endo-phytic bacterium of spinach and that one strain (labeled as SM10)was able to suppress or reduce the Fusarium wilt; however, the san-itary implication of this species has increased since 1988, when itwas reported that it could develop resistance to antibiotics (Gaston1988). Further, some authors have showed that it is an opportunis-tic pathogen in humans (Kanemitsu and others 2007).

Therefore, its isolation and identification in Bella di Cerignolaolives should be considered as a potential sanitary risk. In the lightof this idea, the technological characterization of E. cloacae strains,proposed in this study, is the preliminary phase for the improve-ment of the sanitary quality of Bella di Cerignola olives.

The results of growth in Sab broth + NaCl showed that thesalt could be considered as an effective hurdle to control Gram-

Table 4 --- Homogenous group approach (one-way ANOVAand Fisher’s LSD test) (P < 0.05) for the growth indexvalues experienced by E. cloacae strains after 24 h in nu-trient broth, acidified to pH 4.5 through lactic acid. Dataare the average (n = 4).

Statistical groupsStrain Growthcode index I II III IV V VI VII VIII

3 8.6040 22.8150 23.2446 28.15

4 29.127∗ 30.60

27 30.706 31.19

33 32.2552 33.2017 34.97

6∗ 35.0512∗ 36.5112 42.76

4∗ 44.60

negative growth, as the target strains were inhibited in Sab + 70 g/Lof NaCl, thus confirming the data of Takahashi and others (1997),who reported the inhibition of a E. cloacae strain, isolated fromleaves of Odontioda sp., in a lab medium containing 70 g/L of NaCl.This result, however, could not assure the safety of table olives;small-scale producers of Apulian region, in fact, use brine with lowamounts of salt (50 to 60 g/L) at the beginning of the fermentationto promote the growth of lactic acid bacteria.

Another element to keep in mind is the resistance of the testedstrains to high pHs and the ability to grow at pH 5: a not-correctacidification of brine, in fact, is a common problem for small-scale producers, performing olive fermentation without starter. Ithas been reported that Enterobacter spp. show an optimal growtharound the neutrality (Takafumi and others 1994); in addition,Martens and others (1999) reported a theoretical minimal pH of 5.6for an E. cloacae strain of meat origin.

The strains studied in this research appeared to be inhibited sig-nificantly only at pHs < 5, thus confirming that the isolation sourcecould play a significant role in the phenotypical characteristics of awild strain.

Temperature is a key-element for a correct fermentation courseand the ability to grow both at high and low temperatures is a re-quired characteristic for the starter to be used in olives (DuranQuintana and others 1999); on the other hand, the trend of thespoiling microflora at different temperatures should be assessed fortheir implication on the growth of lactic acid bacteria.

As regards the ranges of tolerance of E. cloacae to the tempera-ture, James and others (2006) reported that E. cloacae is able to growboth at 15 and 43◦C; moreover, it could be classified as a eurypsy-crothrophs, because it shows the ability to grow at 7 ◦C within 10 dof incubation.

The data of the growth at different temperatures confirmed thatE. cloacae could acquire the ability to grow at low temperature ina prolonged storage time (for example, after 72 and 96 h of incu-bation); moreover, for some selected temperature values (as 20 and25 ◦C), the studied strains experienced a GI higher than in the con-trol sample (GI > 100%), thus suggesting that could have an optimaltemperature lower than that reported traditionally for the species(30 to 37 ◦C) (Takafumi and others 1994).

Frequently a lack of lactic acid fermentation occurs for someolive varieties (that is, Manzanilla), due to the presence in the brine

0

1

2

3

4

5

6

7

8

0 50 100 150 200 250

Time (h)

log

CF

U/m

L

A (T, 15 °C)C (pH, 10)F (T, 15 °C; p-coum)G (T, 25 °C; p-coum)L (T, 15 °C; pH, 5; p-coum)M (T, 37 °C)

Figure 7 --- Survival of the strain 3 inbrine (70 g/L NaCl + 1 g/L glucose), atdifferent pHs, added with secondaryphenolic compounds and incubatedat 15, 25, or 37 ◦C. Data are theaverage (n = 4) ± standard deviation.


M:Fo

odMi

crob

iolog

y&

Safet

y


of some inhibitors (phenols and others antimicrobials) (Medinaand others 2008). It is well known that oleuropein and the productsof its hydrolysis could affect the trend of microflora of olives andbrines; few data are available on the effects of secondary phenoliccompounds on olive microflora. Some literature reports are avail-able on the significance of these compounds on lactic acid bacteria(Bevilacqua and others 2006, 2008b); on the other hand, the effectof these molecules on the spoiling microflora is not known.

The influence of p-coumaric and vanillic acids (chosen as theselected compounds of the hydroxycinnamic and hydroxybenzoicgroups) against E. cloacae was assessed in this article. The resultsshowed that the effect of p-coumaric acid was stronger than thatobserved for vanillic acid; it is important to underline that the in-hibition occurs at relatively high amounts of phenolic compounds(0.5 and 1 g/L, for p-coumaric and vanillic acids, respectively).

Before packing and marketing, olive producers add to the brinesome weak acids (lactic, acetic, citric, sorbic, and benzoic acids)to prevent yeast growth and achieve a “sufficient level of sani-tary risk” (Warth 1977; Liewen and Marth 1985; Luck 1990; Arroyo-Lopez and others 2006); this practice is permitted in table olives inmany countries and the effect of these preservatives on yeast mi-croflora and pathogens is known and documented (Marsilio andChichelli 1992; Garrido-Fernandez and others 1997; Turantas andothers 1999; Nakai and Siebert 2004; Arroyo-Lopez and others 2008)Moreover, some authors suggest to add to brine little amounts ofascorbic acid, as it could prevent color change within the storagetime (Arroyo-Lopez and others 2006).

The results of the assay of acidification with different acids con-firmed that their antibacterial activity did not rely only on theamounts added to the brine to achieve a pH value of 4.5 (seeTable 1); the inhibition of E. cloacae growth, in fact, seemed to berelated partially also with the pKa of the compounds (ascorbic acid,pKa1: 4.10; citric acid, Ka1: 3.07; lactic acid, Ka: 3.85), as the effective-ness of the acids was the following: lactic acid–citric acid > ascorbicacid.

After the assays of the technological characterization, performedin a laboratory medium, a confirmatory experiment was run, as-sessing the survival of the strain 3 in synthetic brine. Althoughstrain 3 appeared as the most sensitive with regard to salt and pH, itwas the most vigorous isolate, able to grow and reach the stationaryphase under optimal conditions in 15 to 18 h, otherwise the otherstrains attained this phase after 24 h.

An interesting result was the drastic decrease of E. cloacae countsat high temperature (37 ◦C), for a probable strengthening of salt tox-icity due to an increase of membrane permeability.

Further, this idea could explain the decrease of viable count inbrine added with 0.1 g/L of p- coumaric acid and incubated at25 ◦C, whereas the same amount of the phenol did not affect cellviability at 15 ◦C. Finally, the data of the other combinations shouldbe considered with attention for their implication on the sanitaryrisk, as the microbial survival at pH 5 (the pH that is usually foundin Italian table olive, fermented without starter) and the survival atlow temperature (for example, the temperature of the October andNovember in Southern Italy) could have a strong implication.

Conclusions

Enterobacter cloacae strains, isolated from Bella di Cerignola ta-ble olives, appeared highly resistant to salt and pH, as they were

inhibited only at pH 4.5 and 10 or for NaCl concentrations > 70 g/L;moreover, they were able to grow at low temperature (10 to 15 ◦C).With regard to the other characteristics, they appeared greatly sen-sible to p-coumaric acid and lactic acids. These results were con-firmed by the assays in the brine, conducted on strain 3, thus

suggesting that E. cloacae could really grow during the spontaneousfermentation of Bella di Cerignola table olives; as regards its impacton the correct fermentation course, it is not known if it could influ-ence the growth and acidification kinetic experienced by lactic acidbacteria. This study is the 1st evidence reporting the growth of E.cloacae in fermented olives, emphasizing the need to give attentionto its presence.

AcknowledgmentsThis study was supported by Apulian Region through the grant “Se-lection and characterization of lactic acid bacteria with probioticcharacteristics as starters for table olives” (PE-003).

ReferencesAltieri C, Bevilacqua A, Ouoba L II, Sinigaglia M, Jakobsen M. 2006. Adhesion prop-

erties of lactic acid bacteria, isolated from Italian table olives, and Propionibacte-ria. In Proceedings of The 20th International ICFMH Symposium FoodMicro 2006(Bologna, 29 August – 2 September 2006). Bologna (Italy): Eikos sa, Bologna. 348 p.

Altieri C, Bevilacqua A, D’Amato D, Del Nobile MA, Sinigaglia M. 2008. Modelling thesurvival of starter lactic acid bacteria and Bifidobacterium bifidum in single andsimultaneous cultures. Food Microbiol 25:729–34.

Arroyo-Lopez FN, Duran Quintana MC, Garrido Fernandez A. 2006. Microbial evo-lution during storage of seasoned olives prepared with organic acids, potassiumsorbate, sodium benzoate, and ozone used as preservatives. J Food Prot 69:1354–64.

Arroyo-Lopez FN, Bautista-Gallego J, Duran-Quintana MC, Garrido-Fernandez A.2008. Effects of ascorbic acid, sodium metabisulfite and sodium chloride on fresh-ness retention and microbial growth during the storage of Manzanilla-Alorenacracked table olives. LWT 41:551–60.

Bennik MHJ, Vorstman W, Smid EJ, Gorris LGM. 1998. The influence of oxygen andcarbon dioxide on the growth of prevalent Enterobacteriaceae and Pseudomonasspecies isolated from fresh controlled-atmosphere stored vegetables. Food Micro-biol 15:459–69.

Bevilacqua A, Corbo MR, Sinigaglia M. 2006. In vitro susceptibility of Lactobacillusplantarum, isolated from Italian table olives, to secondary phenolic compounds.Adv Food Sci 28:175–80.

Bevilacqua A, Cannarsi M, Gallo M, Corbo MR, Sinigaglia M. 2008a. Microbiologicalquality of Daunia table olives processed in a small scale factory. Industria Conserve83:211–5.

Bevilacqua A, Corbo MR, Sinigaglia M. 2008b. Influence of secondary phenolic com-pounds on the metabolism and growth/survival of lactic acid bacteria isolated fromItalian table olives “Bella di Cerignola”. In Proceedings of the II International Con-ference on table olives (Dos Hermanas, Sevilla, Spain- 26–27 March 2008). Sevilla(Spain): Fundacion para el fomento y promocion de la aceituna de mesa. 49 p.

Bevilacqua A, Corbo MR, Sinigaglia M. 2008c. Technological characterization of yeastsisolated from Italian table olives “Bella di Cerignola”. In Proceedings of the II Inter-national Conference on table olives (Dos Hermanas, Sevilla, Spain- 26–27 March2008). Sevilla (Spain): Fundacion para el fomento y promocion de la aceituna demesa. 47 p.

Bevilacqua A, Gallo M, Cannarsi M, Corbo MR, Sinigaglia M, Jasche-Ferreri A. 2008d.Preliminary evaluation of the microbiological quality of “Bella di Cerignola” tableolives: influence of a not-correct storage. Industrie Alimentari XLVIII:612–9.

Blaszyk M, Holley RA. 1998. Interaction of monolaurin, eugenol and sodium citrate ongrowth of common meat spoilage and pathogenic organisms. Int J Food Microbiol39:175–83.

Campaniello D, Bevilacqua A, D’Amato D, Corbo MR, Altieri C, Sinigaglia M. 2005. Mi-crobial characterization of table olives processed according to Spanish and Naturalstyles. Food Technol Biotechnol 43:289–94.

Chorianopoulos NG, Boziaris IS, Stamatiou A, Nychas G-JE. 2005. Microbial associa-tion and acidity development of unheated and pasteurized green-table olives fer-mented using glucose or sucrose supplements at various levels. Food Microbiol22:17–124.

Coulin P, Farah, Z, Assanvo J, Spillman H, Puhan Z. 2006. Charaterisation of themicroflora of attieke, a fermented cassava product, during traditional small scalepreparation. Int J Food Microbiol 106:131–6.

Davidson EW, Rosell RC, Hendrix DL. 2000. Culturable bacteria associated with thewhitefly, Bemisia argentifolii (Homoptera aleyrodidae). Fla Entomol 83:159–71.

Duran Quintana MC, Garcia Garcia P, Garrido-Fernandez A. 1999. Establishment ofconditions for green table olive fermentation at low temperature. Int J Food Micro-biol 51:133–43.

European Union (EU). 2005. Commission Regulation (EC) n. 2073/2005 on microbi-ological criteria for foodstuffs. Available online at: eur-lex.europa.eu. Accessed Au-gust 3rd 2009.

Garrido-Fernandez A, Fernandez-Diez MJ, Adams MR. 1997. Table olives: productionand processing. London: Chapman & Hall.

Gaston MA. 1988. Enterobacter: an emerging nosocomial pathogen. J Hosp Infect11:197–208.

Grimont F, Grimont PAD. 2006. The genus Enterobacter. In: Dworkin M, Falkow S,Rosenberg F, Schleifer K-H, Stackebrandt E, editors. The prokaryotes. A handbookof the biology of bacteria. proteobacteria: gamma sub-class. 3rd ed. Volume 6. NewYork: Springer. p 197–214.

Hamilton-Miller JMT, Shah S. 2001. Identity and antimicrobial susceptibility of enter-obacterial flora of salad vegetables. Int J Antimicrob Agents 18:81–83.

Holpzapfel WH. 2001. Appropriate starter culture technologies for small-scale fer-mentation in developing countries. Int J Food Microbiol 75:197–212.


M:FoodMicrobiology&

Safety


IOOC (International Olive Oil Council). 2009. Statistic of table olive’s world produc-tion. Available online at: www.internationaloliveoil.org. Accessed February 25th

2009.James MJ, Martin JL, David AG. 2006. Protection of foods with low-temperatures and

characteristics of psychrotrophic microorganisms. In: Modern food microbiology.7th ed. New York: Springer. p 395–414.

Jasche-Ferreri A. 2008. Table Olives in Italy. II International Conference on table olives(Dos Hermanas, Sevilla, Spain- 26–27 March 2008). Sevilla (Spain).

Kanemitsu K, Endo S, Oda K, Saito K, Kunishima H, Hatta M, Inden K, Kaku M. 2007.An increased incidence of Enterobacter cloacae in a cardiovascular ward. J HospInfect 66:130–4.

Liewen MB, Marth EH. 1985. Growth and inhibition of microorganisms in presence ofsorbic acid: a review. J Food Prot 48:364–375.

Luck F. 1990. Food application of sorbic acids and its salts. Food Addit Contam7:711–5.

Marsilio V, Chichelli A. 1992. Influencia del sorbato potassico y del benzoato sodicosobre la estabilidad de las aceitunas de mesa en salmuera. Grasas Aceites 43:66–74.

Martens DE, Beal C, Malakar P, Zwietering MH, Van’t Riet K. 1999. Modelling the in-teractions between Lactobacillus curvatus and Enterobacter cloacae. I. Individualgrowth kinetics. Int J Food Microbiol 51:53–65

Medina E, Romero C, de Castro A, Brenes M, Garcıa A. 2008. Inhibitors of lactic acidfermentation in Spanish-style green olive brines of the Manzanilla variety. FoodChem 110: 932–7.

Nakai SA, Siebert KJ. 2004. Organic acid inhibition models for Listeria innocua,Listeria ivanovii, Pseudomonas aeruginosa and Oenoccus oeni. Food Microbiol21:67–72.

Neto RJ, Yano T, Beriam LOS, Destefano SAL, Oliveira VM, Rosato YB. 2003. Compar-ative RFLP- ITS analysis between Enterobacter cloacae strains isolated from plantsand clinical origin. Arq Inst Biol 70:367–72.

Soriano JM, Rico H, Molto JC, Manes J. 2001. Incidence of microbila flora lettuce, meatand Spanish potato omelette from restaurants. Food Microbiol 18:159–63.

Takafumi I, Takashi A, Toshio Y, Shinji M, Hidemasa H, inventors; Kanagawa JP, as-signe. 1994 January 11. Enterobacter cloacae ferm BP1529 having plant growth ac-celeratory activity. U.S. patent 5360737.

Takahashi Y, Takahashi K, Sato M, Watanabe K, Kawano T. 1997. Bacterial leaf rotof Odontioda orchids caused by Enterobacter cloacae. Ann Phytopathol Soc Jpn63:164–9.

Tsuda K, Kosaka Y, Tsuge S, Kubo Y, Horino O. 2001. Evaluation of the endophyte En-terobacter cloacae SM10 isolated from spinach roots for biological control againstFusarium wilt of Spinach. J Gen Plant Pathol 67:78–84.

Turantas F, Goksungur Y, Handan Dincer A, Unluturk A, Guvenc U, Zorlu N. 1999.Effect of potassium sorbate and sodium benzoate on microbial population and fer-mentation of black olives. J Sci Food Agric 79:1197–202.

Warth AD. 1977. Mechanisms of resistance of Saccharomyces bailli to benzoic, sorbic,and other weak acids used as food preservatives. J Appl Bacteriol 43:215–30.


characterization and implications of enterobacter cloacae strains, isolated from italian table...

Documents