8/12/2019 Prez et al., 2014

1/7

Effect of pH on the effectiveness of whey protein/glycerol ediblelms containing potassium sorbate to control non-O157 shigatoxin-producingEscherichia coliin ready-to-eat foods

Leonardo M. Prez a, Marina del V. Soazo a ,c, Claudia E. Balagu b, Amelia C. Rubiolo c,Roxana A. Verdini a , *a Instituto de Qumica Rosario (IQUIR, UNR e CONICET), Suipacha 570, 2000 Rosario, Santa Fe, Argentinabrea Bacteriologa Clnica, Facultad de Ciencias Bioqumicas y Farmacuticas, UNR, Suipacha 531, 2000 Rosario, Argentinac Instituto de Desarrollo Tecnolgico para la Industria Qumica (INTEC, Universidad Nacional del Litorale CONICET), Gemes 3450, 3000 Santa Fe,

Argentina

a r t i c l e i n f o

Article history:Received 22 March 2013Received in revised form30 August 2013Accepted 14 September 2013

Keywords:Whey protein edible lmsAntimicrobial propertiesNon-O157Escherichia coliPotassium sorbate

Film microstructureTransparency

a b s t r a c t

Potassium sorbate (PS) (0.5%, 1.0% and 1.5% w/w) was included into whey protein concentrate (WPC)/glycerol (Gly) edible lms at pH 5.2 and 6.0. The lms inhibited or retarded the growth of Shiga toxin-producingEscherichia coli (STEC) pathogens in both diffusion and barrier tests. Bacterial growth in-hibition was dependent on PS content at both pH values. PS release was not affected by pH. Scanningelectron microscopy (SEM) was used to analyze the microstructure of the lms and gain a betterunderstanding of their optical parameters. Acidic controllms (pH 5.2) prepared without PS were theleast transparent. SEM micrographs conrmed the greater structural heterogeneity of these lms,coinciding with opacity. The incorporation of PS into WPC/Gly lms improved transparency andproduced a smoother surface than acidic control ones. The utilization of active packaging based onwhey proteins and organic acids to control and prevent the dissemination of STEC pathogens may be

an effective, safe, ecological and relatively inexpensive alternative to be used in the food packagingindustry.2013 Elsevier Ltd. All rights reserved.

1. Introduction

Active packaging is dened as packaging in which subsidiaryconstituents are deliberately included in or on either the packagingmaterial or the package headspace to enhance the performance ofthe package system (Robertson, 2006). In order to control undesir-able microorganisms on food surfaces, natural or synthetic antimi-crobial agents canbe incorporated into polymercoatings (Appendini

& Hotchkiss, 2002; Kuorwel, Cran, Sonneveld, Miltz, & Bigger, 2011).Several compounds have been proposed as antimicrobial agents infood packaging, including organic acids, enzymes, fungicides andnatural compounds such as spices and essential oils (Kuorwel et al.,2011;Seydim & Sarikus, 2006; Tharanathan, 2003). Sorbic acid, p-aminobenzoic acid, lactic acid, and acetic acid have a long history asgenerally recognized as safe (GRAS) food preservatives that havebeen extensively used as fungistatic and bacteriostatic agents for

foods. Several studies have proved the effectiveness of some foodpreservative addition into edible lms to control microbial growth(Cagri, Ustunol, & Ryser, 2001, 2004; Vsconez, Flores, Campos,Alvarado, & Gerschenson, 2009; Ye, Neetoo, & Chen, 2008). How-ever, the effectof ediblelm matrix pH on the antimicrobial activityof GRAS organic acids has not examined.

Potassium sorbate (PS) is the potassium salt of sorbic acid thatcan effectively restrain the activity of mold, yeast and areophilic

bacteria. The use of PS is effective up to pH 6.5 but effectivenessincreases as pH decreases, consequently the pH of the lm and thefood to which the lm is applied are important parameters.

The proteins in cheese whey, a by-product in the cheese makingprocess, have excellent nutritional and functional properties inaddition to their capacity to form lms (Chen, 1995; Javanmard,2009; Prez-Gago & Krochta, 2002). Use of whey proteins tomanufacturelms has received a great deal of attention since theyare edible and biodegradable (Krochta, & de Mulder-Johnston,1997; Ramos, Fernandes, Silva, Pintado, & Malcata, 2012).

Cagri et al. (2001) reported that incorporating 0.5%e1.5% ofsorbic acid or p-aminobenzoic acid into acidic whey protein isolate

* Corresponding author. Tel./fax: 54 341 4372704.E-mail address:verdini@iquir-conicet.gov.ar(R.A. Verdini).

Contents lists available at ScienceDirect

Food Control

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . co m / l o c a t e / f o o d c o nt

0956-7135/$ e see front matter 2013 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.foodcont.2013.09.018

Food Control 37 (2014) 298e304

mailto:verdini@iquir-conicet.gov.arhttp://www.sciencedirect.com/science/journal/09567135http://www.elsevier.com/locate/foodconthttp://dx.doi.org/10.1016/j.foodcont.2013.09.018http://dx.doi.org/10.1016/j.foodcont.2013.09.018http://dx.doi.org/10.1016/j.foodcont.2013.09.018http://dx.doi.org/10.1016/j.foodcont.2013.09.018http://dx.doi.org/10.1016/j.foodcont.2013.09.018http://dx.doi.org/10.1016/j.foodcont.2013.09.018http://www.elsevier.com/locate/foodconthttp://www.sciencedirect.com/science/journal/09567135http://crossmark.crossref.org/dialog/?doi=10.1016/j.foodcont.2013.09.018&domain=pdfmailto:verdini@iquir-conicet.gov.ar

8/12/2019 Prez et al., 2014

2/7

lms (pH 5.2) can inhibit the growth of Listeria monocytogenes,S. typhimuriumand Escherichia coliO157:H7 in trypticase soy agarsupplemented with yeast extract (TSAYE) acidied to pH 5.2 but noinhibition was observed in TSAYE adjusted to pH 6.5. These resultsreveal the importance of the medium pH for a superior effective-ness of the organic acids. Moreover, some restrictions in the pHlimit for the elaboration of whey protein edible lms are related tothe isoelectric point of whey proteins.

Casting lms at pH values lower than the isoelectric pointproved to be inefcient since the reactivity of SH groups decreasessignicantly (Ferreira, Nunes, Delgadillo, & Lopes-da-Silva, 2009).In addition, at acidic pH values whey proteins aggregate and lmsproperties could decrease (Prez-Gago & Krochta, 1999). Hence,changes in the pH of the whey protein/Gly system may affect so-lution stability, and thus, the antimicrobial characteristics of thelms containing PS. To our knowledge, no comparative studiesregarding the differences in the pH of the edible lms based onwhey protein and sorbate have been reported.

Shiga toxin-producingE. coli(STEC) is recognized worldwide asone of the most important causes of food-borne infections(Armstrong, Hollingsworth, & Morris, 1996). Clinical presentationof STEC infection varies from an asymptomatic state to bloody

diarrhea and life-threatening complications such as hemolyticuremic syndrome (Johnson et al., 1996; Williams et al., 1999). STECposses the Shiga toxin 1 and/or Shiga toxin 2 (stx1andstx2) genesconsidered the critical virulence factors in disease. Environmentalconditions present in foods (nutrients, pH, humidity, etc.) aresuitable for a rapid colonization by STEC strains to harmful levelsfor human health. Non-O157 STEC is now recognized as animportant group of bacterial enteropathogens (Gilmour et al.,2009; Gyles, 2007). Several outbreaks caused by non-O157 STECwere described (Johnson et al., 1996; Williams et al., 1999),although data implicating these STEC in some outbreaks werescanty and the source of infection was not always known. In theUnited States, most STEC outbreaks were traced to beef containingE. coli O157:H7 and for that reason most epidemiological studies

have focused on the prevalence of this serotype in beef and beefcattle (Gill & Gill, 2010; Hussein, 2007; Hussein & Sakuma, 2005 ).Interestingly, undercooked ground beef and other beef productsare now considered reservoirs of O157 and non-O157 STEC(Hussein, 2007). However, worldwide, additional STEC serotypes,including members of the recently named big six serogroups (O26,O45, O103, O111 and O121), have been isolated from foods otherthan beef and caused human illnesses (Mathusa, Chen, Enache, &Hontz, 2010).Balagu et al. (2006) reported the phenotypic andgenotypic characteristics and virulence properties of non-O157strains isolated from ready-to-eat food samples obtained fromsupermarkets and shop selling in Argentina were hemolytic ure-mic syndrome is endemic (Ibarra et al., 2008; Reilly, 1998).

Because of the global nature of the food supply, safety concerns

and new challenges facing the food industry are increasing both atthe production and processing levels. Considering that antimicro-bial edible packaging is a novel technology with the potential tohelp food preservation, the purpose of the present study was toevaluate the inhibitory effects of WPC/Gly ediblelms incorporatedwith PS against non-O157 STEC strains isolated from ready-to-eatfood samples and its relationship with the pH of the lm. Howev-er, since pH modications can negatively alter other relevantproperties of protein-based lms, inuence of both pH and organicacid presence on the optical properties of WPC/Gly lms was alsoanalyzed. One goal of this work was to provide an array of data tosupport comparative studies on the molecular structure, opticaland biocide properties of acidic WPC/Gly lms for rationalimprovement of such lms toward their eventual application as

edible packaging.

2. Materials and methods

2.1. Materials

WPC 80% (Arla Food Ingredients S.A., Buenos Aires, Argentina)was used to prepare thelm forming solutions. Gly (Cicarelli, SantaFe, Argentina) was added to all lm forming solutions as a plasti-cizer; PS (Sigma Chemical Co., St. Louis, Mo., U.S.A.) was included inthe formulations at different concentrations to evaluate the biocideproperties of the WPC/Gly lms. Trypticase Soy Broth (TSB), CystineLactose-Electrolyte-Decient (CLDE) and MuellereHinton culturemediums were purchased from Britania (Buenos Aires, Argentina).2,3,5-triphenyltetrazolium chloride (TTC) was obtained from Merck(Darmstadt, Germany).

2.2. Film preparation

Edible lms (casting solution 11.5% total solid) were obtainedwith a modication of the method described by Soazo, Rubiolo, andVerdini (2010). Briey, WPC and Gly (in proportion WPC/Gly 3:1 w/w dry solid basis) were dissolved in distilled water. After mixing,the solution was heated at 90 C for 30min ina waterbath (TDS-40,

Tecno Dalvo, Santa Fe, Argentina). Finally, the solution was ho-mogenized (4 min; 20,000 rpm) with an Omni GLH homogenizer(Omni International Inc., Warrenton, Virginia, U.S.A.). After ho-mogenization, the solutions were placed in an ice bath to preventfurther denaturation of the whey proteins and rapidly cooled toroom temperature. After incorporating 0.5%, 1.0% or 1.5% (w/w) ofPS, the pH was adjusted to 5.2 or 6.0 with 1.0 N HCl using a Met-rohm 713 pH-meter (Metrohm Ltd., Herisau, Switzerland). The pH-adjusted lm-forming solutions were degassed at room tempera-ture with a vacuum pump. Following degassing, the lm formingsolutions (8 g/plate) were casted by pipetting the solution intosterile 90 mm diameter plastic plates. The plates were dried for 2 hat 45 C plus 24 h at 25 C and 48 4% relative humidity, afterwhich the lms were peeled from the plates and stabilized during

24 h at 25 C and 48 4% relative humidity. The lms used in thedifferent tests were selected based on the lack of physical defectssuch as cracks, bubbles and holes.

2.3. Film thickness

The thicknessesof three replicates of each lm formulation weremeasured with an electronic digital disk micrometer (Schwyz,China) at nine locations on the lm to the nearest 0.001 mm.Average lm thickness was 0.137 0.030 mm.

2.4. Transparency

Film transparency was determined according to ASTM D1746

(ASTM, 1997) with modications of the method described byOzdemir and Floros (2008). The lms were cut into rectangularpieces (10 mm 30 mm) and placed on the internal side of aspectrophotometer cell. Transparency oflms was measured usinga spectrophotometer (Model V-530, Jasco International, Tokyo,

Japan) at 560 nm. Five replicates of each lm were tested. Thetransparency (% Transparency) was calculated as the percentualrelationship between the light intensity with the specimen in thebeam and the light intensity with no specimen in the beam.

2.5. Scanning electron microscopy

In order to study the structure of WPC/Gly lms and to assestheir homogeneity, SEM experiments were carried out. Film sam-

ples were cryo-fractured by immersion in liquid nitrogen and

L.M. Prez et al. / Food Control 37 (2014) 298e304 299

8/12/2019 Prez et al., 2014

3/7

mounted on bronze stubs perpendicularly to their surface. Theportions were coated with gold during 15 min at 70e80 mTorr.Micrographs of lms cross-section were taken with a scanningelectron microscope (AMR 1000, Leitz, Wetzlar, Germany) using anaccelerating voltage of 20 kV. Magnications of 500 and 1000 were

used.

2.6. Bacterial strains and growth conditions

Eight non-O157 STEC strains isolated from ready-to-eat foodsamples obtained from supermarkets and shops selling in Rosario(Argentina) were kindly provided byBalagu et al. (2006)as testmicroorganisms (Table 1). E. coli O157:H7 ATCC 43895 from theAmerican Type Culture Collection which expresses the Shiga toxingenesstx1and stx2was used as a control. All strains were main-tained at 70 C in TSB containing 10% (v/v) glycerol and sub-cultured overnight in CLDE at 37 C before use.

2.7. Inhibitory activity of PS in liquid media

Minimum inhibitory concentration (MIC) was estimated ac-cording to the National Committee for Clinical LaboratoryStandards-recommended macrodilution broth method (NCCLS,2009) as described byPrez, Balagu, Rubiolo, and Verdini (2011).Briey, an overnight culture of each bacterial strain was adjusted toMcFarland 0.5 standard in saline solution (approximately5 107 cfu/ml) and used as test inoculum. One ml of the bacterialinoculum was added and mixed with 1 mL of each PS solution inMueller-Hinton broth adjusted at pH 5.2 or 6.0. The PS concentra-tions evaluated were 0.3125, 0.625, 1.25, 2.5, 5.0, 10 and 20 mg/ml.A control tube without PS was inoculated to test microbial growthand a tube containing only broth medium was evaluated to discardpossible contaminations. All tubes were overnight incubated

(w

18 h) at 37

C. The MIC values were estimated as the lowestconcentration of PS that completely inhibits growth of the micro-organism in the tubes as detected by the unaided eye. Growthcontrol tubes to assess MIC end points were also evaluated.

2.8. Inhibition zone assay of WPC/Glylms containing PS in agarmedia

The inhibition zone assay in solid media was used to determinethe antimicrobial potential oflms. WPC/Gly lms were asepticallycut in 12 mm diameter discs using a sterile cork borer. The discswere then aseptically transferred to pour plates containing 10 mL ofMueller-Hinton agar broth acidied to pH 5.2 with 1.0 N HCl, whichhad been previously seeded with each bacterial suspension

adjusted to McFarland 0.5 standard in saline solution. After

overnightincubation (w18h)at37 C thediameterof theinhibitionzone represented by a clear area of non-growth or a decreasedgrowth around the lm disc was measured perpendicularly to thenearest millimeter with a caliper. Control lms without antimicro-bial were also evaluated. Experiments were performed in triplicate.

2.9. Barrier assay of WPC/Glylms containing PS in agar media

In order to study the lm performance to prevent microbialcontamination a microplate barrier assay was developed. Mueller-Hinton agar adjusted at pH 5.2 was poured into 24 well microtiterplates (Greiner Bio-One, Frickenhausen, Germany). Inocula wereprepared from an overnightculture of each bacterial strain adjustedto McFarland 0.5 standard in saline solution and then diluted 1/10,000 (approximately 5 103 cfu/ml). Discs of 16 mm diameterWPC/Glylms were aseptically cut from and applied on the surfaceof the wells lled with Mueller-Hinton agar (pH 5.2). Then, 10 mlinoculum for each strain were seeded on the lm discs. Next, 10mlof TTC (30 mg/ml) were added into each well to color the bacterialcolonies. Microplates were incubated at 37 C for 24 h until nakedeye observation. Experiments were performed in triplicate.

2.10. Determination of sorbate release from WPC/Gly edible lms

To analyze sorbate release, WPC/Gly lms containing 1.5% PS (toenhance detection sensibility) were cut in 12 mm diameter discsusing a cork borer. The discs were placed on the surface of Petridishes containing 10 mL of Mueller-Hinton agar broth acidied topH 5.2 with 1.0 N HCl. After 1 h, 3 h, or 18 h of incubation at 37 Clms disc were removed and the change in PS concentration withtime in the lm was determined. The initial amount of PS per discwas also evaluated. Each PS-containing WPC/Gly disc wasimmersed in 4 mL 0.01 N HCl solution, homogenized (1 min;20 000 rpm) with an Omni GLH homogenizer (Omni InternationalInc., Warrenton, Virginia, U.S.A.) and 10 min sonicated. Samples(30 mL) were diluted to 3.0 mL with 0.01 N HCl. Potassium sorbate

concentration of the collected samples were measured at 254 nmwith a UVeVis spectrophotometer (Model V-530, Jasco Interna-tional, Tokyo, Japan). Calibration of the extraction procedureshowed that these conditions enabled stable and reproducible re-covery (98% withr2 0.993). Percentage of PS remaining in discwere calculated and plotted as a function of time. Experimentswere performed in triplicate.

2.11. Statistical analysis

All experiments were replicated using a complete randomizeddesign. Analysis of variance (ANOVA) was used and when the effectof the factors was signicant (p

8/12/2019 Prez et al., 2014

4/7

in the solution during drying. Films incorporated with PS, at bothpH, were more easily separated from the casting plate than controllms without preservative. Furthermore, when PS concentrationwas increased in the lm-forming solutions, WPC/Gly lms exi-bility improved. These observations may be explained because thisadditive has an additional effect as plasticizer on lm matrixes(Fam, Rojas, Goyanes, & Gerschenson, 2005). The plasticizermolecules lead to decreases in intermolecular forces along thepolymer chains, thus improving exibility, and allowing an easierremoval from the forming support (Hernandez-Izquierdo &Krochta, 2008; Karbowiak et al., 2006). Films showed an averagethickness of 0.137 0.030 mm, which is similar to those reportedby other authors using similar formulations (Ramos et al., 2013;Soazo et al., 2010).

3.2. Film transparency

It is important to notice that when an edible lm is intended tobe used as a supercial layer on food surfaces, transparency is animportant attribute that contributes to consumer acceptability. Ingeneral, lms obtained in this study were rather transparent.However, in the absence of PS, lms obtained at the acidic pH valueof 5.2 were substantially less transparent than lms obtained at pH6.0, presumably due to a partial proteins precipitation when pH isclose to the isoelectric point of whey proteins (pIz 5) (Fig.1). Filmformation from proteins is believed to proceed through the for-mation of a three-dimensional network of protein molecules byionic, hydrogen, hydrophobic, and disulde bonds. At pH 5.2 pro-teins are mainly in the zwitterions form; i.e., a neutral moleculewith a positive and a negative electrical charge at different loca-tions within that molecule.At this state,near the pI,the solubility ofa protein is negligible because its net charge is zero and any elec-trostatic repulsion disappears. This fact supports the idea that lmtransparency could decrease at acidic pH due to protein aggrega-tion. Our results correspond well with this phenomenon in WPC/Gly lms without PS. Interestingly, the addition of 0.5%, 1.0% and

1.5% of PS signicantly improved the transparency of WPC/Glylmswhen compared to control lms without PS at pH 5.2. Under acidicconditions, protein molecules in lm-forming solutions arepartially unfolded due to the protein denaturation and their hy-drophobic groups are exposed (Hamaguchi, Yin, & Tanaka, 2007;Kristinsson & Hultin, 2003). The organic sorbate molecule couldinteract with the amino or hydroxyl groups on non-crosslinkedprotein through covalent bonds and thus lead to an increase in

transparency (Fig. 1). This effect was also observed in analogousexperiments performed by our group where incorporation of so-dium benzoate or sodium propionate in WPC/Gly acidic lm ob-tained at pH 5.2 improves transparency (David, 2012; unpublishedundergraduate thesis data). This phenomenon seems to be inde-pendent of the organic salt and its concentration, and is probablymore related to hydrophobic interactions between the organicchain and the unfolded protein matrix structure.

3.3. Scanning electron microscopy

Our observations on lms transparency are supported by SEMimages. As can be seen inFig. 2(C and D), the microstructure ofWPC/Gly lms incorporated with PS shows a smoother surfacewhen compared with that of the acidic controllms obtained at pH5.2. These observations correspond well with the fact that wheyprotein particles formed at pH values near the pI had a cauliower-like appearance giving rise to macroscopic gels with heterogeneousmicrostructure formed of larger pores (Saglam, Venema, de Vries,van Aelst, & van der Linden, 2012). Langton and Hermansson(1992)showed that whey protein gels formed between pH 4 and6 were opaque and had a particulate network which can be

described as consisting of aggregated particles. These authors alsoreported that the size of those aggregates changed to denser andwell-packed aggregates when the pH increased. When pH valuesare far from the pI the microstructure of whey proteins macro-scopic gels present a homogeneous surface with relatively smallpore sizes distributed regularly through the protein network(Langton & Hermansson, 1992; Saglam et al., 2012). SEM imagespresented inFig. 2 (A and B) agree with these previous experiences,while changes observed in the microstructure of the WPC/Gly lmsincorporated with PS suggest that interactions between proteinchains and the organic acid contribute to improve the structuralarrangement of the compounds, and thus could enhance lmtransparency.

3.4. Inhibitory activity of PS in liquid media

Our MIC data showed that the PS concentration needed toinhibit microbial growth in a liquid medium at pH 6.0 was 2e4levels higher than at pH 5.2 for all the STEC analyzed (Table 2).These results conrm that the unprotonated-protonated ratio inthe liquid medium was relevant for the antimicrobial ability of theorganic acid. PS is a weak acid (pKa 4.76) which is more effective in

a

bc

bc

bc

bc

a

0

10

20

30

40

50

60

70

0% PS 0.5% PS 1.0% PS 1.5% PS 0% PS 0.5% PS 1.0% PS 1.5% PS

pH 6.0 pH 5.2

%T

ransparency

bb

b

Fig. 1. Effect of potassium sorbate (PS) and pH (6.0 or 5.2) on the transparency of WPC/Gly edible

lms. Different letters show signi

cant differences (p