poly (dl-lactide-co-glycolide) (plga) nanoparticles with entrapped trans-cinnamaldehyde and...

N:NanoscaleFoodScience

Poly (DL-lactide-co-glycolide) (PLGA)Nanoparticles with Entrappedtrans-Cinnamaldehyde and Eugenol forAntimicrobial Delivery ApplicationsCarmen Gomes, Rosana G. Moreira, and Elena Castell-Perez

Abstract: Eugenol and trans-cinnamaldehyde are natural compounds known to be highly effective antimicrobials; how-ever, both are hydrophobic molecules, a limitation to their use within the food industry. The goal of this study was to syn-thesize spherical poly (DL-lactide-co-glycolide) (PLGA) nanoparticles with entrapped eugenol and trans-cinnamaldehydefor future antimicrobial delivery applications. The emulsion evaporation method was used to form the nanoparticles inthe presence of poly (vinyl alcohol) (PVA) as a surfactant. The inclusion of antimicrobial compounds into the PLGAnanoparticles was accomplished in the organic phase. Synthesis was followed by ultrafiltration (performed to eliminate theexcess of PVA and antimicrobial compound) and freeze-drying. The nanoparticles were characterized by their shape, size,entrapment efficiency, and antimicrobial efficiency. The entrapment efficiency for eugenol and trans-cinnamaldehyde wasapproximately 98% and 92%, respectively. Controlled release experiments conducted in vitro at 37 ◦C and 100 rpm for72 h showed an initial burst followed by a slower rate of release of the antimicrobial entrapped inside the PLGA matrix.All loaded nanoparticles formulations proved to be efficient in inhibiting growth of Salmonella spp. (Gram-negative bac-terium) and Listeria spp. (Gram-positive bacterium) with concentrations ranging from 20 to 10 mg/mL. Results suggestthat the application of these antimicrobial nanoparticles in food systems may be effective at inhibiting specific pathogens.

Keywords: antimicrobial efficiency, controlled release, Listeria spp., PLGA nanoparticle, Salmonella spp.

Practical Application: Nanoencapsulation of lipophilic antimicrobial compounds has great potential for improving theeffectiveness and efficiency of delivery in food systems. This study consisted of synthesizing PLGA nanoparticles withentrapped eugenol and trans-cinnamaldehyde. By characterizing these new delivery systems, one can understand thecontrolled-release mechanism and antimicrobial efficiency that provides a foundation that will enable food manufacturersto design smart food systems for future delivery applications, including packaging and processing, capable of ensuringfood safety to consumers.

IntroductionNatural compounds that do not have any significant medical

or environmental impact could potentially serve as effective alter-natives to conventional antibacterial or antifungal agents. In thiscontext, the activity of several essential oils against both impor-tant human pathogenic and food spoilage microorganisms shouldbe studied (Burt 2004; Hammer and others 1999). Although themajority of the essential oils are classified as Generally Recognizedas Safe (GRAS), their use in foods as preservatives is often limiteddue to flavor considerations, since effective antimicrobial dosesmay exceed organoleptically acceptable levels.

In antimicrobial action of essential oil components, thelipophilic character of their hydrocarbon skeleton and the hy-drophilic character of their functional groups are of great impor-

MS 20100971 Submitted 8/27/2010, Accepted 11/1/2010. Authors are with theDept. of Biological & Agricultural Engineering, College Station, TX 77843-2117,U.S.A. Direct inquiries to author Gomes (E-mail: [email protected]).

tance. The activity rank of essential oil components is as follows:phenols > aldehydes > ketones > alcohols > hydrocarbons. Thephenols include thyme, savory, and oregano oils containing thymoland carvacrol as well as clove oil containing eugenol. Cinnamonoil with cinnamaldehyde as the main component is also a memberof this group (Knobloch and others 1989; Kalemba and Kunicka2003).

Cinnamaldehyde (cinnamic aldehyde or 3-phenyl-2-propenal)is the main component in cassia oil as well as cinnamon barkoil, and is a GRAS for food use based on 21 CFR (Codeof Federal Regulation) part 172.515 (CFR 2009). It has beenshown to be the major antimicrobial compound in cinnamon.In addition to exhibiting antibacterial activity, cinnamic aldehydealso inhibits mold growth and mycotoxin production (Beuchat1994).

Eugenol (2-methodxy-4-[2-propenyl] phenol) is a major con-stituent in clove oil and present in considerable amounts in theessential oil of allspice. It is also a GRAS for food use basedon 21 CFR part 172.515 (CFR 2009). Karapinar and Aktug(1987) reported that eugenol was more effective against Salmonella

C© 2011 Institute of Food TechnologistsR©

N16 Journal of Food Science � Vol. 76, Nr. 2, 2011 doi: 10.1111/j.1750-3841.2010.01985.xFurther reproduction without permission is prohibited

N:Na

nosc

aleFo

odSc

ience

Antimicrobial PLGA nanoparticles . . .

Typhimurium, Staphylococcus aureus, and Vibrio parahaemolyticusthan were thymol, anethol, and menthol (Beuchat 1994).

Eugenol and cinnamaldehyde are cyclic terpene alcohols(Figure 1) with low solubility in water. The predominant ap-plication for these compounds is in the flavoring and fragranceindustries. The restrictive factors for these compounds are theirhigh volatility, sparingly water solubility and irritant effect thatmake difficult to formulate preparations containing them.

In the last decades, micro- and nano-sized colloidal carriers havereceived a growing scientific and industrial interest (Thies 1996).These vectors may be capsules (with liquid core surrounded by asolid shell), particles (polymeric matrices), vesicles or liposomes,multiple or single emulsions and have found a wide range ofapplications. They may be loaded by living cells, enzymes, flavoroils, pharmaceuticals, vitamins, adhesives, agrochemicals, catalysts,and offer considerable advantages at use. Liquids can be handled assolids, odor, or taste can be effectively masked in a food product,sensitive substances can be protected from deleterious effects of thesurrounding environment, toxic materials can be safely handled,and drug delivery can be controlled and targeted (Robinson 1997).

Synthetic polymers and natural macromolecules have been ex-tensively researched as colloidal materials for nanoparticles pro-duction designed for drug delivery. Synthetic polymers have theadvantage of high purity and reproducibility over natural poly-mers; with the polyester family (that is, poly (lactic acid) (PLA),poly(e-caprolactone) (PCL), and poly (glycolic acid) (PGA) be-ing of great interest in the biomedical area because of their bio-compatibility and biodegradability properties. In particular, poly(lactide-co-glycolide) (PLGA) has been FDA approved for humantherapy (Anderson and Shive 1997).

Along with other physical characteristics, the size and sizedistribution of the PLGA nanoparticles, are affected by thetechnique used for nanoparticle production and the pertinentsynthesis parameters, that is, PLGA molecular mass, the additionof active components, surfactants, and other additives (Barrat andothers 2000; Westesen and others 2001; Brigger and others 2002;Chorny and others 2002; Hans and Lowman 2002; Nakache andothers 2000). Emulsion evaporation, emulsion diffusion, saltingout, nanoprecipitation, and solvent displacement are some of thecommon methods used to form nanoparticles from preformedpolymers (Astete and Sabliov 2006).

Spices and herbs have been used for millennia to provide distinc-tive flavor for foods and beverages around the world. In additionto contributing flavor to foods, many spices also exhibit antimi-crobial activity. Essential oils are the odorous, volatile products ofplant secondary metabolism, normally formed in special cells orgroups of cells or as glandular hairs, found on many leaves andstems (Bernfield 1967). The antimicrobial action of essential oilsin model food systems or in real food is well documented in theliterature (Koutsoumanis and others 1998; Lambert and others2001).

Figure 1–Eugenol (left) and trans-cinnamaldehyde (right) chemicalstructures.

Despite the development of antibiotics, bacterial and fungalinfections are still a major issue in medicine, and the presenceof numerous drug-resistant strains poses a new challenge. Herbaldrugs have been extensively used in this field for many centuries.Recently, there has been a growing interest in natural productsdue to their availability, fewer side effects or toxicity as well asbetter biodegradability as compared to the available antibioticsand preservatives (Kalemba and Kunicka 2003).

Delivery of eugenol and trans-cinnamaldehyde entrapped inpolymeric nanoparticles has definite advantages over the deliv-ery of nonentrapped antimicrobials. The release rate of thesecompounds can be controlled and the dose frequency reduced.Furthermore, the bioactivity and stability of the active sub-stance entrapped in the nanoparticles is protected by encapsulation(Uhrich and others 1999; Lamprecht and others 2001; Zigoneanuand others 2008). Also, it will prevent the loss of volatiles andimprove their solubility in a hydrophilic medium, which broadensits application. In the food industry, it has become apparent thatinclusion of lipophilic bioactives in food matrices is one of themajor problems that manufacturers have when developing health-and-wellness-promoting foods.

There has been very little published on the release mecha-nism and kinetics of antimicrobial agents from nanoparticles. Moststudies have measured the release of drugs from micro- and nano-particles (Anderson and Shive 1997; Astete and Sabliov 2006).To understand the release of antimicrobial agents, the parametersof trans-cinnamaldehyde and eugenol diffusion kinetics must bedetermined.

Most of the models used for describing the controlled release ofsubstances have been based on solutions of the Fickian diffusionequation (Siepmann and Peppas 2001). For 1-dimensional radialantimicrobial content in a sphere of radius r, under perfect sinkinitial and boundary conditions, with constant diffusion coefficientD, the series expansion of Fick’s law for unsteady state diffusion is(Crank 1975):

M(t )Mo

= 6π2

α∑n=1

1n2

exp[− Dn2π2t

r 2

](1)

where, M(t) is the nanoparticles’ antimicrobial content at time t,and Mo the initial antimicrobial content. A simplification of thesolution of Eqn. (1) was used to predict the antimicrobial releasefrom PLGA nano-systems. Instead of an infinite number of terms,only the first term of Eqn. (1) was employed to calculate the releaserate:

M(t )Mo

= 6π2

exp(

− Dπ2tr 2

)= 6

π2exp (−Kt ) (2)

Equation (1) and (2) give significantly different results only forsmall values of t. The difference between the series solution ofEq. (1) and the 1-term solution of Eqn. (2) is less than 5% if thedimensionless ratio K = Dπ2t/r 2 > 0.52 (that is, for larger times> 6500 s assuming average D of 6.12 × 10−20 m2/s and averager of 8.6875 × 10−8 m). Shorter times can be assumed if moreterms are added to Eqn. (2), (that is, 1000 s for the same D and rvalues for the 2-term solution); however, Eqn. (2) often predictsthe shape of drug releasing curves from polymers poorly (Ritgerand Peppas 1987a; Ritger and Peppas 1987b). The initial releaserate is too low and the approach to the final drug content in thepolymeric delivery system is too rapid.

Vol. 76, Nr. 2, 2011 � Journal of Food Science N17



A simple semi-empirical equation also used to describe theamount of drug released from a wide variety of matrix deliverysystems is the so-called power law (Costa and others 2001; Zulegerand Lippold 2001; Siepmann and Peppas 2001; Xiong and others2005; Gamsiz and others 2008). The power law model is similarin form to the short-time approximation of the exact solution ofFick’s second law of diffusion (Ritger and Peppas 1987a; Ritgerand Peppas 1987b). So, it does not provide a good fitting forantimicrobial release data.

Thus, the objectives of this study were: (1) to synthesize PLGAnanoparticles with entrapped trans-cinnamaldehyde and eugenolwith poly (vinyl alcohol) (PVA) as surfactant; (2) to character-ize the size, size distribution, morphology, entrapment efficiency,and controlled release profile of the synthesized nanoparticles; (3)to determine the minimum inhibitory concentration (MIC) toestablish their antimicrobial efficiency; and (4) to determine thediffusion kinetics of trans-cinnamaldehyde and eugenol from thenanoparticles.

Materials and Methods

MaterialsPLGA, with a copolymer ratio of DL-lactide to glycolide of

65 : 35 (Mw 40000 to 75000 g/mol, P2066, Sigma-Aldrich, St.Louis, Mo., U.S.A.) was used in this study. The surfactant usedin the emulsification process was PVA (98% to 99% hydrolysisdegree and average Mw 30000 to 50000 g/mol, 363138, Sigma-Aldrich). Trans-cinnamaldehyde, 99% (C80687), and eugenol,99% (E51791), were also purchased from Sigma-Aldrich. Sodiumchloride, 99.5%, sodium dodecyl sulfate (SDS), 99%, acetoni-trile (high-performance liquid chromatography [HPLC] grade),D(+) trehalose, 98% were obtained from VWR International(West Chester, Pa., U.S.A.). Nanopure water was obtained usingNanopure Diamond (Barnsted Intl., Dubuque, Iowa, U.S.A.).

Synthesis of nanoparticles with entrappedtrans-cinnamaldehyde and eugenol

The nanoparticles were prepared by an emulsion evaporationmethod (Figure 2) similar to Zigoneanu and others (2008). Un-loaded PLGA nanoparticles were also prepared and used as con-trol systems. Briefly, a solution of 50 mg of PLGA in 2 mLof dichloromethane with or without trans-cinnamaldehyde oreugenol (16%, w/w—relative to PLGA) was prepared, whichconstituted the organic phase. The aqueous phase (20 mL) wasprepared with 0.3% (w/v) PVA in nanopure water. The organicphase was added in droplets to the aqueous phase, and this mix-ture was homogenized for 2 min at 95000 rpm (Ultra-TurraxT25 basic Ika

R©-Works Inc., Wilmington, N.C., U.S.A.) form-

ing an oil-in-water emulsion. Next, the o/w emulsion was son-icated in an ice bath at 2 ◦C and 70 W of energy output (ColeParmer 8890 ultrasonic cleaner, Vernon Hills, Ill., U.S.A.) during10 min. The organic phase (dichloromethane) was evaporated for20 min using a rotoevaporator (Buchi R-210 Rotavapor, BuchiCo., New Castle, Del., U.S.A.) under vacuum (operating at 28inHg [50 torr]) (Zigoneanu and others 2008). Unloaded (control)nanoparticles were prepared using the same procedure describedpreviously, without addition of trans-cinnamaldehyde or eugenolto the organic solvent.

Following synthesis, the nanoparticles were purified by ultra-filtration to remove the excess of surfactant (PVA) and trans-cinnamaldehyde or eugenol. A Millipore-LabscaleTM TFF systemequipped with a 10 kDa cutoff Pellicon XL-Millipore (MilliporeCo., Kankakee, Ill., U.S.A.) was used. The nanoparticles wereultrafiltered with 300 mL of water and 50 mL was collected (re-tentate). Inlet pressure was 40 psi and outlet pressure in the systemwas 40 psi and the outlet pressure was 10 psi. After ultrafiltration,nanoparticles were kept for 2 h at –80 ◦C and freeze-dried at–50 ◦C under 5 mtorr (9.67 × 10−5 psi) vacuum for 48 h in aLabconco Freeze Dry-5 unit (Labconco, Kansas City, Mo.,U.S.A.). D(+) trehalose was added to the nanoparticles suspension

Figure 2–Flow diagram of nanoparticlesynthesis with entrapped trans-cinnamaldehydeand eugenol by the emulsification–evaporationmethod using PVA as surfactant.

N18 Journal of Food Science � Vol. 76, Nr. 2, 2011

N:Na

nosc

aleFo

odSc

ience


before precooling at a ratio of 1 : 1 (w/w) relative to the amountof nanoparticles. Finally, the lyophilized samples were stored indesiccator placed in the freezer (–20 ◦C) until further use (Asteteand Sabliov 2006).

Size and size distributionThe size distribution of the nanoparticles was determined in

nano-pure water at 25 ◦C temperature using a particle size ana-lyzer (90 Plus particle size analyzer, Brookhaven Instruments Co.,Holtsville, N.Y., U.S.A.). For the measurements, 200 μL of thenanoparticles suspension were dispersed in 2 mL. The water waspreviously filtered using a 0.02-μm filter (Anotop 10, Whatman,Piscataway, N.J., U.S.A.). The analysis was performed at a scatter-ing angle of 90◦, refractive index of 1.590, and at a temperature of25 ◦C. A standard of 100 nm (nanosphere size standards, Duke Sci-entific Co., Palo Alto, Calif., U.S.A.) of a polymer microsphere inwater (polymer density of 1.05 g/cm3, refractive index of 1.590 at589 nm) was used to verify equipment calibration. Measurementsof size and polydispersity index (particle size distribution [PDI])were made for freshly synthesized nanoparticles (prior ultrafiltra-tion), ultrafiltered nanoparticles, and re-suspended nanoparticles(after freeze-drying). For each sample, the mean diameter and thestandard deviation of 9 determinations were calculated.

Nanoparticle morphology characterizationThe morphology of nanoparticles was examined by transmission

electron microscopy (TEM) using a JEOL 1200 EX TEM (JEOLUSA Inc., Peabody, Mass., U.S.A.) system at the Microscopy Imag-ing Center of the Biology Dept. at Texas A&M Univ. A dropletof the aqueous phase containing nanoparticles was placed on acopper grid of 400 mesh with carbon film. The sample excesswas removed with a filter paper. Uranium acetate 2% was usedas a stain; the excess uranium acetate was removed after 1 minwith a filter paper. The sample was dried before analysis by TEM.Observations were performed at 60 kV (Panyam and others 2003;Astete and Sabliov 2006).

Determination of trans-cinnamaldehyde and eugenolentrapment efficiency

The entrapment efficiency of trans-cinnamaldehyde andeugenol in the PLGA nanoparticles was measured by reverse-phase (RP)-HPLC. A sample of 2-mg lyophilized nanoparticleswas dissolved in 2-mL acetonitrile : water (95 : 5). Each samplewas passed through a syringe filter (pore size 0.2-μm, Acrodisc,Pall Life Sciences, Port Washington, N.Y., U.S.A.) before HPLCmeasurements. The entrapment efficiency of both active sub-stances was determined by injecting 25 μL of previously preparedsamples into the HPLC column. A Dionex Summit HPLC sys-tem (Dionex, Germering, Germany) equipped with P680 pumpwith high-speed gradient, ASI-100 autosampler, SOR-100 de-gasser running Chromeleon version 6.6. software was used withphotodiode array detection (PDA-100). All wavelengths between250 and 601 nm were collected while 280-nm wavelengths weremonitored during runs. The column was a Supelco 2.1 × 250mm, 5-μm Discovery Bio Wide Pore C18 column with Supel-guard (C18) column cartridge (Supelco, Bellefonte, Pa., U.S.A.).The mobile phase consisted of acetonitrile : water : acetic acid ina ratio of 50 : 50 : 0.01, isocratic with a flow rate of 0.25 mL/minand a total run time of 15 min per sample.

A standard curve (concentrations ranged from 2 to 200 μg/mL,25-μL injection volume) was prepared under the same condi-

tions. Trans-cinnamaldehyde and eugenol entrapment efficiencywas expressed as the percentage of active substance entrapped inthe nanoparticles reported to the initial amount of each used fornanoparticles preparation. Determinations were made in tripli-cates.

Controlled release studyLyophilized resuspended nanoparticles were dissolved in the re-

lease medium, containing phosphate saline buffer (PBS, 0.15 M,pH 7.4), at a concentration of 1 mg/mL (sufficient to establishsink conditions—trans-cinnamaldehyde and eugenol concentra-tion in the medium was kept 5 times lower than their saturationsolubility in buffer) similar to Zigoneanu and others (2008). Thesample was divided into 10 vials (1.5 mL Eppendorf tube) of 1mL. The in vitro controlled release study was performed at 37◦C and 100 rpm (Orbit shaker, Lab Line Instruments, MelrosePark, Ill., U.S.A.). At predetermined time intervals, samples werewithdrawn and filtered with a 0.2-μm filter (Acrodisc) and thetrans-cinnamaldehyde and eugenol concentrations were assessedby RP-HPLC. The mobile phase was acetonitrile : water : aceticacid in the ratio 50 : 50 : 0.01, the injected volume was 25 μLand the flow rate 0.25 mL/min. The column was a Supelco 2.1 ×250 mm, 5-μm Discovery Bio Wide Pore C18 column with Su-pelguard (C18) column cartridge (Supelco) in a Dionex SummitHPLC system (Dionex) equipped with P680 pump with high-speed gradient, ASI-100 autosampler, SOR-100 degasser runningChromeleon version 6.6 software was used with photodiode arraydetection (PDA-100). All wavelengths between 250 and 601 nmwere collected while 280-nm wavelengths were monitored duringruns. Determinations were made in triplicates.

In this study, we are proposing a modified empirical equationto describe the release of trans-cinnamaldehyde and eugenol com-pounds from PLGA-nanoparticles:

M(t )Mo

= b1 ∗ exp (−k1t ) + b2 ∗ exp(−k2t ) (3)

where, b1, b2, k1, and k2 are constants. This model can be seenas a generalized form of the solution of the Fickian model whenemploying only the first 2 terms of Eqn. (2). The values k1, andk2 are defined as the antimicrobial release rate constants. Theexperimental data were fitted to the Eqn. (3) using SPSS softwarefor Windows, version 16.0 (SPSS 2007).

Minimum inhibitory concentration (MIC)Overnight cultures of Salmonella spp. (Salmonella serotype En-

teriditis and Salmonella serotype Typhimurium) and Listeria spp.(Listeria monocytogenes ATCC 15313, L. monocytogenes Scott A46, L. monocytogenes strain A, Listeria innocua NRCC B33076, L.monocytogenes ATCC 33090), used as representative of the Gram-negative and Gram-positive bacteria, respectively; were grown intryptic soy broth (TSB). Based on turbidity measurements, corre-lated with plate counts, the cultures were diluted to approximately105 CFU/mL. An appropriate volume of 105 CFU/mL dilutionwas added to sterile TSB to give approximately 103 CFU/mL levelof organism. Each of the wells of a 96-well microtiter plate (ster-ilized 300 μL volume—MicroWell, 96-Well Plates, NUNC, flat-bottom, Thermo Scientific, West Palm Beach, Fla., U.S.A.) wascharged with 125 μL of seeded TSB using a multichannel pipettor.To the first well of a row, 125 μL of the antimicrobial nanoparticles




(containing either trans-cinnamaldehyde or eugenol) were addedand mixed well. After mixing, 125 μL of the well contents weretransferred to the next well. This transfer procedure continued for10 of the 12 rows of wells. The last row of well acted as a growthcontrol. Along with each antimicrobial being screened, sterilizedwater blank was used in an identical fashion to ensure that thesterilized water (where the nanoparticles were suspended) effect,if present, was not misinterpreted. The test compound concentra-tion in the wells ranged from 20000 to 20 μg/mL of nanoparticle(Kalemba and Kunicka 2003).

The microtiter plates were placed into a microplate reader(Tecan SpectraFluor plus, Tecan US Inc., Durham, N.C., U.S.A.)equipped with MagellanTM software. The microtiter plates wereincubated at 35 ◦C with agitation (100 rpm) and at every 3 for36 h absorbance of each well was measured at 550 nm. The MICwas defined as the concentration of the last row of well where nogrowth occurred (no change in optical density within time).

Statistical analysisData analysis was performed using the SPSS software for Win-

dows, version 16.0 (SPSS 2007) for size, PDI of nanoparticles, andentrapment efficiency. Differences between variables were testedfor significance by one-way analysis of variance (ANOVA). Sig-nificantly different means (P < 0.05) were separated by the Tukeytest. The kinetic data (controlled release study) were analyzed bythe nonlinear procedure to determine k1 and k2 and correlationcoefficient, R2.

Results and Discussion

Size and polydispersity of nanoparticlesSize and polydispersity of nanoparticles was measured after

evaporation, ultrafiltration, and freeze-drying steps, as shown inTable 1 and 2. The processing steps (that is, evaporation, ultra-filtration, and freeze-drying) affected the size of unloaded PLGAnanoparticles, which decreased (P < 0.05) from 317 to 174 nm.The same effect was observed with the entrapment of the an-timicrobial compounds, though it was not significant. Initially, theunloaded PLGA nanoparticles were significantly (P < 0.05) oflarger size than the other antimicrobial loaded nanoparticles, butthey decreased (P < 0.05) in size during the subsequent evapora-tion step. With the exception of the evaporation step, there wasno significant difference in size among the antimicrobial loaded

Table 1–Effect of treatment (unloaded PLGA, PLGA loaded withtrans-cinnamaldehyde [PLGA-C], PLGA loaded with eugenol[PLGA-E]), and processing parameter (evaporation, ultrafiltra-tion, and freeze-drying) during synthesis on size of nanoparti-cles. Standard deviation reported in parenthesis.

Size (nm)

Processing parameters

After After AfterTreatments evaporation ultrafiltration freeze-drying

Unloaded PLGA x317.1bw,x234.7a

w174.0a

(14.2) (52.8) (83.3)PLGA-C w225.3a,b

w222.6aw173.8a

(5.08) (12.87) (40.4)PLGA–E w185.2b

w167.2aw179.3a

(96.8) (47.0) (36.5)a,bMeans within a column that are not followed by a common superscript letter aresignificantly different (P < 0.05).w,xMeans within a row that are not followed by a common superscript letter are signifi-cantly different (P < 0.05).

and unloaded nanoparticles throughout the processing steps. Thenanoparticles had a mean diameter of 180 ± 4 nm (Table 1).

The presence of antimicrobial did not affect (P < 0.05) the meandiameter of the nanoparticles with the eugenol loaded particlesbeing slightly larger (Table 1). The presence of antimicrobial agentyielded a more viscous dispersed phase, making it difficult for themutual dispersion of the phases and therefore resulting in largerparticles. Similar results were observed in previous studies usingPVA as emulsifier (Mainardes and Evangelista 2005; Zigoneanuand others 2008).

The processing steps (that is, evaporation, ultrafiltration, andfreeze-drying) did not (P < 0.05) affect the polydispersity forall treatments, including whether or not the nanoparticles wereunloaded or loaded with the antimicrobial agents. The only ex-ception occurred with the unloaded PLGA nanoparticles duringthe freezing-drying steps, where a significant size decrease (P <

0.05) occurred (Table 2).All the treatments evaluated showed different families of parti-

cles with different sizes. The advantage of a monodisperse system(PDI < 0.100) is related to its ability to deliver a consistent amountof compound, as compared to a mixture of polydisperse particles,of different loading capacities (Astete and Sabliov 2006). In thisstudy, the polydispersity was higher (above 0.100) probably becauseimportant amounts of PVA were removed during the ultrafiltrationprocess and this could cause some agglomeration of nanoparticles,though it did not affect its antimicrobial capacity.

In the emulsion-solvent evaporation method, the emulsificationand stabilization of the globules are crucial factors. The higher en-ergy released during homogenization and sonication in the processleads to a rapid dispersion of polymeric organic phase as nan-odroplets of small size and monomodal distribution profile. Theemulsification can be considered one of the most important stepsof the process, because an insufficient dispersion of phases resultsin large particles with wide size distribution. A reduction of theemulsion globule size allows the formation of smaller nanoparticles(Mainardes and Evangelista 2005). Also, the amount of surfactantplays an important role in the emulsification process and in theprotection of the droplets, because it can avoid the coalescenceof globules (Quintanar-Guerrero and others 1996; Murakami andYoshino 1999). Therefore, to reduce the polydispersity of thisstudy (Table 2), one could increase the energy release duringhomogenization (by increasing rpm) and/or sonication (increas-ing power [wattage]). Another alternative would be selecting a

Table 2–Polydispersity of nanoparticles as a functionof treatments (unloaded PLGA, PLGA loaded with trans-cinnamaldehyde [PLGA-C], PLGA loaded with eugenol [PLGA-E]), and processing parameter during synthesis. Standarddeviation reported in parenthesis.

Polydispersity (−)

Processing parameters

After After AfterTreatments evaporation ultrafiltration freeze-drying

Unloaded PLGA x0.32ax0.33a

w0.25a

(0.03) (0.04) (0.05)PLGA-C w0.31a

w0.30aw0.29a

(0.01) (0.02) (0.03)PLGA–E w0.33a

w0.30aw0.33a

(0.03) (0.06) (0.04)a,bMeans within a column that are not followed by a common superscript letter aresignificantly different (P < 0.05).w,xMeans within a row that are not followed by a common superscript letter are signifi-cantly different (P < 0.05).


N:Na

nosc

aleFo

odSc

ience


different emulsifier with surface charge (it increases repulsion andconsequently avoids coalescence of globules).

Nanoparticle morphology characterizationFigure 3 shows the morphology of unloaded and loaded PLGA

particles (trans-cinnamaldehyde-PLGA). Eugenol–PLGA loadedparticles images are not shown due to their similarity (size andshape) with the other particles. All particles showed a sphericalshape and smooth surface with a broad size distribution and astrong tendency to form clusters. No distinct difference in size andshape was noticed; however, among treatments. The aggregationpresent on all particles was likely due to natural clustering due toinsufficient steric stabilization by the PVA, which is a non-ionicsurfactant. The size of the nanoparticles measured by the particleanalyzer is consistent with the particle sizes estimated from TEM(in the range of 200 nm) images.

The PVA chain is formed from alternating hydrophilic andhydrophobic segments. The hydrophobic moieties of PVA inter-connect with the PLGA chains to create a matrix, whereas thehydrophilic PVA moieties face the water phase. This arrangementis conducive to the formation of the PVA “corona,” observed byTEM (Zigoneanu and others 2008). The affinity of PVA withwater pulled the PVA to the surface of micelles by the bulk watermolecules in the emulsion process. At the same time, the hy-drophobic polylactide segments were directed into the inner coreof the micelles. Corona type microspheres were formed after sol-vent evaporation as observed by the TEM images.

Determination of trans-cinnamaldehyde and eugenolentrapment efficiency

Entrapment efficiency is used to indicate the amount of com-pound entrapped into the polymeric matrix. The entrapment effi-ciency for eugenol was 98.27% and for trans-cinnamaldehyde was92.56%. This is in agreement with the data found in the litera-ture about hydrophobic components whose entrapment is usuallygreater than 90% (Brigelius-Flohe and Traber 1999; Zigoneanuand others 2008).

Controlled release studyThe trans-cinnamaldehyde and eugenol release profiles were

evaluated in vitro as a function of time. The results over 72 hreleasing time are shown in Figure 4. The release profiles for bothagents were biphasic, showing an initial burst effect in the firstcouple of hours followed by uniform release within time. Sincesolubility in water of both compounds is similar (approximately1 g/L of water), an explanation for the rapid burst observed onthe release profile of eugenol would be that a large amount of thecompound was close to or attached to the surface of the particle(Sansdrap and Moes 1997).

The 2-term exponential model (Eqn. 3) fitted well the exper-imental data for both antimicrobial releasing profiles. The valuesfor the release constant rates and the coefficient of determina-tion, R2, are shown in Table 3. The sum of the exponential seriesyields a rate that initially is steeper and then slows down at long

Table 3–Release constant values for trans-cinnamaldehyde andeugenol antimicrobial agents determined by fitting Eq. (3) to theexperimental data.

Product k1(1/s) k2(1/s) R2

trans-cinnamaldehyde 1.76 × 10−4 2.75 × 10−6 0.96Eugenol 4.10 × 10−4 1.65 × 10−6 0.97

times compared to the 1-term exponential equation; this resultemphasizes the difficulty in obtaining a sustained release effectfrom a monodisperse sphere formulation of this type, and indeedthe rapid initial release may be indistinguishable from the bursteffect observed in many systems. For trans-cinnamaldehyde, therelease rate constants were obtained by fitting the experimentaldata to Eqn. (3) were k1 = 1.76 × 10−4 1/s and k2 = 2.75 ×10−6 1/s compared to the values of k1 = 4.10 × 10−4 1/s andk2 = 1.65 × 10−6 1/s obtained for eugenol. These values showthe differences between the 2 nanoparticles release characteristics,with almost 80% of the trans-cinnamaldehyde being released in thefirst 5 h compared to only 45% of eugenol being released duringthe same time period. Even after 72 h of incubation, eugenol didnot completely release into the medium.

As illustrated by the chemical structures of trans-cinnamaldehydeand eugenol (Figure 1), the steric conformation of eugenol and thefact that eugenol is more lipophilic than trans-cinnamaldehyde, itmay be more difficult for the eugenol to diffuse from the innermostparts of the PLGA nanoparticles to the outside medium. Furtherresearch is needed in this area.

The release rate constant, k1, for eugenol was higher than fortrans-cinnamaldehyde, which could be caused by more eugenolthan trans-cinnamaldehyde being adsorbed to the outside of the

Figure 4–Antimicrobial release kinetic by PLGA nanospheres as a functionof time (h) (Top: PLGA-trans-cinnamaldehyde; Bottom: PLGA-eugenol). Seetext for equations used in the fitting procedure.




particle, consequently yielding a steeper burst effect on the dilu-tion.

The fraction of antimicrobial released in the initial burst is thusdependent on the composition of the nanoparticles. Generally,the mechanisms by which active agents can be released from adelivery system are the combination of diffusion of the activeagent passing through the polymer, polymeric erosion, swelling,and degradation. Usually, the degradation of PLGA is slow, there-fore the release mechanism of the antimicrobial substances from

the nanoparticles may depend on the substance diffusion and thePLGA surface and bulk erosions or swelling (Mu and Feng 2003).In our antimicrobial release system, the diffusion occurred whenthe substance passed through the polymeric matrix into the exter-nal environment, by passing between polymer chains. In this typeof system, the release rate normally decreases with time becausethe substance has a progressively longer distance to travel and thusrequires a longer time to escape. In the 72-h period of incubation,the average released amount of antimicrobial was approximately

Figure 3–Top: TEM images of unloaded PLGA nanoparticles (×50000 magnification). Bottom: TEM images of PLGA-trans-cinnamaldehyde loadednanoparticles (×50000 magnification).


N:Na

nosc

aleFo

odSc

ience


64% and 87% of the total antimicrobial load for eugenol and trans-cinnamaldehyde, respectively.

The rapid release (under 0.5 h approximately 20% of antimicro-bial load) can be attributed to the diffusion of the encapsulated oradsorbed antimicrobial in the external surface of the nanoparticles.The second release period (after 5 h) shows a dependency withtime and is related to the antimicrobial diffusivity inside the matrixsystem, the surface area, and, obviously, the antimicrobial agent.

Minimum inhibitory concentration (MIC)The 2 antimicrobial nanoparticles (loaded with trans-

cinnamaldehyde and eugenol) showed different degrees of growthinhibition against Salmonella spp. and Listeria spp. using thebroth micro dilution method (Figure 5 and 6). The growth ofSalmonella spp. was inhibited by both compounds, PLGA-trans-cinnamaldehyde (Top Figure 5) and PLGA-eugenol (Bottom Fig-ure 5), at 10000 μg/mL, being this concentration the minimuminhibitory concentration, or the lowest antimicrobial compoundconcentration in the broth resulting in lack of visible microorgan-ism growth changes.

Furthermore, Listeria spp. growth was inhibited by both com-pounds, PLGA-trans-cinnamaldehyde (Top Figure 6) and PLGA-eugenol (Bottom Figure 6), at 20000 μg/mL. At 10000 μg/mL,a delay in the lag phase, a slower growth rate, and thus a lowerfinal cell concentration of microorganism were observed for both

Figure 5–Growth of Salmonella spp. in TSB as a function of active compoundconcentration (Top: PLGA-trans-cinnamaldehyde; Bottom: PLGA-eugenol).Concentration units are microgram per milliliter.

nanoparticles. It is commonly known that Gram-positive bacteriaare more susceptible to essential oils than Gram-negative microor-ganisms (Kalemba and Kunicka 2003). However, there are someexceptions, with Vibrio vulnificus (Gram-negative) proven to bemost susceptible to the essential oil constituents, with L. monocyto-genes (Gram-positive) the least sensitive among a number of testedbacteria (Kim and others 1995), as confirmed in this study.

The concentrations of nanoparticles would correspond to aMIC of approximately 800 and 1600 μg/mL of antimicrobialcompound (eugenol and trans-cinnamaldehyde) for Salmonella spp.and Listeria spp., respectively; considering the loaded amountof antimicrobial into the nanoparticles and its entrapment effi-ciency. These results show the spectrum of eugenol and trans-cinnamaldehyde antimicrobial nanoparticles, which are able toeffectively inhibit the growth of both Gram-positive and Gram-negative bacteria.

Therefore, the use of nanoparticles to encapsulate antimicro-bial hydrophobic compounds could improve their efficacy dueto 3 main factors: improved hydrophilicity, sustained release, andsmall size, as observed in the present study, trans-cinnamaldehydeand eugenol increased their solubility in water due to microen-capsulation with PLGA (an aqueous suspension of 1 mg/mLwas translucid). Furthermore, the nanoparticles showed a sus-tained release with continuous migration of the antimicrobialwith time. In addition, their small size (approximately 200 nm)

Figure 6–Growth of Listeria spp. in TSB as a function of active compoundconcentration (Top: PLGA-trans-cinnamaldehyde; Bottom: PLGA-eugenol).Concentration units are microgram per milliliter.




would not present a barrier to access the cytomembranes ofbacteria.

ConclusionsThe emulsion evaporation method allowed the preparation of

an antimicrobial loaded system of biodegradable PLGA carriercontaining natural compounds derived from essential oils (thatis, trans-cinnamaldehyde and eugenol) incorporated in the poly-meric matrix. These nanoparticles are suitable for hydrophobicantimicrobial release carriers, with an entrapment efficiency rang-ing from 92% to 98% for both compounds and final mean sizeunder 200 nm, for a 16% (w/w) antimicrobial theoretical loading.

The in vitro release kinetics of antimicrobial compounds wasgoverned by an initial burst followed by a slow and continuousrelease profile. The 2-term exponential kinetic model predictedwell the trans-cinnamaldehyde and eugenol release. In 72 h, 64%and 87% of the initial load was released in the medium for eugenoland trans-cinnamaldehyde, respectively. The release rate constantswere k1 = 1.76 × 10−4 1/s and k2 = 2.75 × 10−6 1/s for trans-cinnamaldehyde compared to k1 = 4.10 × 10−4 1/s and k2 =1.65 × 10−6 1/s for eugenol. Furthermore, the controlled releaseprofile showed that these biodegradable PLGA/antimicrobial de-livery systems have great potential and should be given particularconsideration in the design of new active packaging or even fordirect food application.

The antimicrobial efficacy of these nanoparticles was evalu-ated against Salmonella spp. (Gram-negative bacterium) and Liste-ria spp. (Gram-positive bacterium) with MICs ranging from 10 to20 mg/mL, respectively for both nanoparticles, demonstrating abroad spectrum of application in food systems where both typesof microorganism could present a risk. These results indicate thatsuch nanoparticles could be useful antimicrobial delivery systemscapable of inhibiting growth of either Gram-positive or negativepathogenic bacteria with continuous releasing during 72 h.

AcknowledgmentThe authors thank Dr. Z. Nikolov and Dr. B. Faulkner from

Biological and Agricultural Engineering Dept. and Dr. P. Cre-mer from Chemistry Dept. at Texas A&M Univ. for allowinguse of their lab facilities and equipment to conduct ultrafiltrationand HPLC analysis (entrapment efficiency and controlled releasestudy), sonication, and size and polydispersity measurements, re-spectively.

ReferencesAnderson JM, Shive MS. 1997. Biodegradation and biocompatibility of PLA and PLGA micro-

spheres. Adv Drug Deliv Rev 28(1):5–24.Astete CE, Sabliov CM. 2006. Synthesis and characterization of PLGA nanoparticles. J Biomater

Sci Polym Ed 17(3):247–89.Barrat G, Courraze G, Couvreur C, Fattal E, Gref R, Labarre D, Legrand P, Ponchel G,

Vauthier C. 2000. Polymeric micro- and nanoparticles as drug carries. In: Dumitriu S, editor.Polymeric biomaterials. 2nd ed. New York, N.Y.: Marcel Dekker. p 753.

Bernfield P. 1967. Biogenesis of natural compounds. 2nd ed. Oxford, N.Y.: Pergamon Press.Beuchat LR. 1994. Antimicrobial properties of spices and their essential oils. In: Dillon VM,

Board RG, editors. Natural Antimicrobial systems and food preservation. Wallingford, U.K.:Cab Intl. p 328.

Brigelius-Flohe R, Traber MG. 1999. Vitamin E: function and metabolism. J Fed Am Soc ExpBiol 13:1145–455.

Brigger I, Dubernet C, Couvreur P. 2002. Nanoparticles in cancer therapy and diagnosis. AdvDrug Deliv Rev 54(5):631–51.

Burt S. 2004. Essential oils: their antibacterial properties and potential applications in foods: areview. Int J Food Microbiol 94:223–53.

[CFR] Code of Federal Regulations. 2009. Title 21, Part 172.515. Food additives permitted fordirect addition to food for human consumption: synthetic flavoring substances and adjuvants.56–63.

Chorny M, Fishbein I, Danenberg HD, Golomb G. 2002. Lipophilic drug loaded nanospheresprepared by nanoprecipitation: effect of formulation variables on size, drug recovery andrelease kinetics. J Control Release 83(3):389–400.

Costa P, Manuel J, Lobo S. 2001. Modeling and comparison of dissolution profiles. Eur J PharmSci 13(2):123–33.

Crank J. 1975. The mathematics of diffusion. 2nd ed. Oxford, U.K.: Oxford Univ. Press.Gamsiz DE, Shah LK, Devalapally H, Amiji MM, Carrier RL. 2008. A model predicting delivery

of saquinavir in nanoparticles to human monocyte/macrophage (Mo/Mac) Cells. BiotechnolBioeng 101(5):1072–82.

Hammer KA, Carson CF, Riley TV. 1999. Antimicrobial activity of essential oils and otherplant extracts. J Appl Microbiol 29:130–5.

Hans ML, Lowman AM. 2002. Biodegradable nanoparticles for drug delivery and targeting.Curr Opin Solid State Mater Sci 6(4):319–27.

Kalemba D, Kunicka A. 2003. Antibacterial and antifungal properties of essential oils. Curr MedChem 10:813–29.

Karapinar M, Aktug SE. 1987. Inhibition of foodborne pathogens by thymol, eugenol, mentholand anethole. Int J Food Microbiol 4:161–6.

Kim JM, Marshall MR, Wei C. 1995. Antibacterial activity of some essential oil componentsagainst 5 foodborne pathogens. J Agric Food Chem 43(11):2839–45.

Knobloch K, Pauli A, Iberl B, Weigand H, Weis N. 1989. Antibacterial and antifungal propertiesof essential oil components. J Essential Oil Res 1:119–28.

Koutsoumanis K, Tassou CC, Taoukis PS, Nychas G-JE. 1998. Modelling the effectiveness ofa natural antimicrobial on Salmonella Enteriditis as a function of concentration, temperatureand pH, using conductance measurements. J Appl Microbiol 84:981–7.

Lambert RJW, Skandamis PN, Coote PJ, Nychas G-JE. 2001. A study of the minimum in-hibitory concentration and mode of action of oregano essential oil, thymol and carvacrol. JAppl Microbiol 91:453–62.

Lamprecht A, Ubrich N, Yamamoto H, Schafer U, Takeuchi H, Maincent P, KawashimaY, Lehr CM. 2001. Biodegradable nanoparticles for targeted drug delivery in treatment ofinflammatory bowel disease. J Pharmacol Exp Ther 299:755–81.

Mainardes RM, Evangelista RC. 2005. PLGA nanoparticles containing praziquantel: effect offormulation varialbes on size distribution. Int J Pharm 290:135–44.

Mu L, Feng SS. 2003. PLGA/TPGS nanoparticles for controlled release of paclitaxel: effects ofthe emulsifier and drug loading ratio. Pharm Res 20(11):1864–72.

Murakami Y, Yoshino H. 1999. Preparation of poly (DL-lactide-co-glycolide) nanoparticlesby modified spontaneous emulsification solvent diffusion method. Int J Pharm 187:143–52.

Nakache E, Poulain N, Candau F, Orecchioni A, Irache J. 2000. Biopolymer and poly-mer nanoparticles and their biomedical applications. In: Nalwa HS, editor. Handbook ofnanostructured materials and nanotechnology. San Diego, Calif., U.S.A.: Academic Press.p 577.

Panyam J, Sahoo SK, Prabha S, Bargar T, Labhasetwar V. 2003. Fluorescence and electronmicroscopy probes for cellular and tissue uptae of poly(D,L-lactide-co-glycolide) nanoparticles.Int J Pharm 262(1–2):1–11.

Quintanar-Guerrero D, Fessi H, Allemann E, Doelker E. 1996. Influence of stabilizing agentsand preparatives variables on the formation of poly(D,L-lactic acid) nanoparticles by anemulsification-diffusion technique. Int J Pharm 143:133–41.

Ritger PL, Peppas NA. 1987a. Transport of penetrants in the macromolecular structure of coals.4. Models for analysis of dynamic penetrant transport. Fuel 66(6):815–26.

Ritger PL, Peppas NA. 1987b. Transport of penetrants in the macromolecular structure of coals.7. Transport in thin coal sections. Fuel 66(10):1379–88.

Robinson JR. 1997. Controlled drug delivery. Past, present, and future. In: Park K, editor.Controlled drug delivery: challenges and strategies. Wash.: Am Chem Soc. p 1–7.

Sansdrap P, Moes AJ. 1997. In vitro evaluation of the hydrolytic degradation of dispersed andaggregated poly (DL-lactide-co-glycolide) microspheres. J Control Release 43:47–58.

Siepmann J, Peppas NA. 2001. Mathematical modeling of controlled drug delivery. Adv DrugDeliv Rev 48(2–3):137–8.

SPSS. 2007. SPSS for Windows. Chicago, Ill.: SPSS Inc.Thies C. 1996. A survey of microencapsulation processes. In: Benita S, editor. Microencapsula-

tion: methods and industrial applications. New York: Marcel Dekker. p 1–19.Uhrich K, Cannizaro SM, Langer R, Shakesheff KM. 1999. Polymeric systems for controlled

drug release. Chem Rev 99:3181–98.Westesen K, Bunjes H, Hammer G, Siekmann B. 2001. Novel colloidal drug delivery systems.

PDA J Pharm Sci Technol 55:240–7.Xiong XP, Zhang LN, Ma ZC, Li Y. 2005. Effects of molecular weight and arm number

on properties of star-shape styrene-butadiene-styrene triblock copolymer. J Appl Polym Sci95(4):832–40.

Zigoneanu IG, Astete CE, Sabliov CM. 2008. Nanoparticles with entrapped a-tocopherol:synthesis, characterization, and controlled release. Nanotechnology 19:1–8.

Zuleger S, Lippold BC. 2001. Polymer particle erosion controlling drug release. I. Fac-tors influencing drug release and characterization of the release mechanism. Int J Pharm217(1–2):139–52.


poly (dl-lactide-co-glycolide) (plga) nanoparticles with entrapped trans-cinnamaldehyde and...

Documents