International Journal of Food Microbiology 139 (2010) 41–47

Contents lists available at ScienceDirect

International Journal of Food Microbiology

j ourna l homepage: www.e lsev ie r.com/ locate / i j foodmicro

Differences in attachment of Salmonella enterica serovars to cabbage andlettuce leaves☆

Jitendra Patel ⁎, Manan SharmaUSDA, Agricultural Research Service, Environmental and Microbial Food Safety Laboratory, 10300 Baltimore Avenue, BARC-East, Bldg. 201, Beltsville, Maryland 20705-2350, USA

☆ Mention of trade names or commercial products doeendorsement to the exclusion of other products by the U.⁎ Corresponding author. Tel.: 301 504 7003; fax: 301

E-mail address: jitu.patel@ars.usda.gov (J. Patel).

0168-1605/$ – see front matter. Published by Elsevierdoi:10.1016/j.ijfoodmicro.2010.02.005

a b s t r a c t

a r t i c l e i n f o

Article history:Received 28 September 2009Received in revised form 1 February 2010Accepted 9 February 2010

Keywords:SalmonellaProduceAttachmentLettuceCabbageBiofilm

This study investigated the ability of five Salmonella enterica serovars to attach to and colonize intact and cutlettuce (Iceberg, Romaine) and cabbage surfaces. Biofilm formation and attachment of Salmonella serovars tointact and cut leaves were determined. Populations of loosely and strongly attached Salmonella wereobtained to calculate the attachment strength (SR). Biofilm formation, as determined by microtiter plateassay, varied with strain and growth medium used. Salmonella Tennessee and S. Thompson producedstronger biofilms compared to S. Newport, S. Negev, and S. Braenderup. Biofilm formation was also strongerwhen Salmonella spp. were grown in diluted TSB (1:10). S. Tennessee, which produced strong biofilms,attached to produce surfaces at significantly higher numbers than the populations of S. Negev. Overall, S.Tennessee displayed more biofilm formation in vitro and attached more strongly to lettuce than otherserovars. All Salmonella serovars attached rapidly on intact and cut produce surfaces. Salmonella spp.attached to Romaine lettuce at significantly higher numbers than those attached to Iceberg lettuce orcabbage. Salmonella attached preferentially to cut surface of all produce; however, the difference betweenSalmonella populations attached to intact and cut surfaces was similar (PN0.05). Salmonella attachment toboth intact and cut produce surfaces increased with time. The overall attachment strength of Salmonella wassignificantly lower on cabbage (0.12) followed by Iceberg (0.23) and Romaine lettuce (0.34). Cabbage, intactor cut, did not support attachment of Salmonella as well as Romaine lettuce. Understanding the attachmentmechanisms of Salmonella to produce may be useful in developing new intervention strategies to preventproduce outbreaks.

s not imply recommendation orS. Department of Agriculture.504 8438.

B.V.

Published by Elsevier B.V.

1. Introduction

From 1996 to 2008, eighty-two foodborne illness outbreaks wereassociated with the consumption of fresh produce. Of these produce-related outbreaks, 28 (34%) were linked to the consumption of leafygreens. During this time period, leafy greens-associated outbreaksaccounted for 949 illnesses and 5 deaths (FDA 2009). Salmonella spp.,traditionally associated with consumption of poultry origin, haveincreasingly been linked to fresh produce (Sivapalasingam et al.,2004). An August 2008 outbreak of Salmonella Saintpaul was linked toconsumptionof hot peppers andpossibly tomatoes in 43 states in theUSand Canada, resulting in 1442 cases of illnesses (CDC, 2006). An ongoingoutbreak involving 124 infections with the consumption of lettuce wascaused by Salmonella Typhimurium (Anonymous, 2009). RecentSalmonella outbreaks linked to fresh produce in other countries includeSalmonella Thompson associated with lettuce in the United Kingdom,SalmonellaAnatum linked tobasil inDenmark, Salmonella Typhimurium

DT204B in lettuce in several European countries, and SalmonellaSenftenberg associated with imported basil affecting the UK, Denmark,the Netherlands and the United States (Berger et al., 2009).

Increased consumption ofminimally processed fruits and vegetableshas lead to an increase in the number of outbreaks to these products(Sewell and Farber, 2001). The risk of produce-associated illness hasincreased with increased consumption and importation (Beuchat,2002). However, outbreaks specifically associated with leafy greencommodities in the U.S. has increased disproportionately – 39% –

compared to increase in consumption – 9% – between 1996 and 2005,indicating that other factors may account for the increased number ofoutbreaks associatedwith these commodities (Hermanet al., 2008). TheUnited States imports of fruits and vegetables doubled during thedecade from 1994 to 2004 to $12.7 billion (Aruscavage et al., 2006). Theglobalization of the producemarket also represents a potential source offoodborne illness because of the difficulties in tracing the implemen-tation of Good Agricultural Practices (GAPs) and hygienic conditions inproduction environments. Increased reporting by health officialsprovides better surveillance of produce-related illness (Suslow et al.,2003). In addition, other factors that affect produce-related outbreaksinclude, an aging population that is susceptible to foodborne illness;a more complex supply chain; improvements in epidemiological

42 J. Patel, M. Sharma / International Journal of Food Microbiology 139 (2010) 41–47

investigation; and increasingly better methods to identify pathogens(FDA 2009).

Salmonella may contaminate fresh produce during any point fromfarm to fork through incidental contact with the organism. The sourcesof Salmonella contamination may include soil, raw or improperlycomposted manure, irrigation or wash water, handling by workers,and contact with equipment surfaces (Kroupitski et al., 2009). Thepersistence and survival of Salmonella on produce is affected by itsability to adapt to thenewecological environments (Beuchat, 2002). It isessential to understand the initial stages of Salmonella attachment tovarious plant tissues so that effective intervention and mitigationstrategies could be utilized.

Attachment of enteric bacterial pathogens has been evaluatedpreviously by several authors. E. coli O157:H7 attached preferentiallyto cut edges of Iceberg lettuce compared to intact tissues (Boyer et al.,2007; Takeuchi et al., 2000). The preferential attachment to cut cabbagesurfaces over intact oneswasalso observed in24 Listeria strains (Ells andHansen, 2006). However, Takeuchi et al. (2000) did not find differencesin S. Typhimurium attachment on intact or cut lettuce surfaces. Barakand coworkers found 10- to 1000-foldmore adherence of S.Newport onalfalfa sprouts than of E. coli O157:H7 (Barak et al., 2002). Other studiesindicated stronger attachment of S. Senftenberg to basil leavescompared to the attachment of S. Typhimurium to these leaves (Bergeret al., 2009). However, the ability of Salmonella to form biofilms in vitrohas not been correlated to its ability to attach to produce surfaces untilnow. Most attachment studies of Salmonella have been performedwithIceberg lettuce, but little is known about its attachment to other types oflettuce. We compared the ability of S. enterica serovars to form biofilmsin vitro, and attach to intact and damaged surfaces of green cabbage,Iceberg and Romaine lettuce. We also investigated the relative strengthof bacterial attachment to these produce surfaces.

2. Materials and methods

2.1. Bacterial cultures and media

Five S. enterica serovars, associated with produce commodities,were used in the study. S. Thompson 2051H, S. Tennessee 2053N, andS. Negev 26 H, isolated from thyme, were provided by Tom Hammack(U.S. Food and Drug Administration, College Park, MD). S. Braenderup(CDC clinical isolate # 95-682-997) and S. Newport (CDC clinicalisolate #9113) were used from our Environmental Microbial and FoodSafety Laboratory culture collection. The strains were cultured fromstocks stored at −80 °C in tryptic soy broth (TSB, Becton Dickinson,Sparks, MD) supplemented with 10% glycerol. Frozen cultures of eachstrain were partially thawed at room temperature (~22 °C) for 15 min,and streaked onto tryptic soy agar slants (TSA, Becton Dickinson),and incubated at 37 °C for 24 h.

2.2. Biofilm formation

Overnight cultures of individual Salmonella serovars grown in TSBwere diluted 1:10,000 in growth media and 200 µl was deposited inwells of a sterile 96-well polystyrene microtiter plate (FisherScientific, Newark, DE) and incubated under static conditions at30 °C for 48 h. Growth media that were tested included LB (LuriaBertani, Becton Dickinson) broth, diluted (1:10) LB broth, TSB anddiluted (1:10) TSB. For each replicate experiment, eight wells wereinoculated for each serovar in each growth medium. Growth mediumdevoid of bacterial inoculum served as a negative control. After 48 hincubation in microplate, 200 μl of culture was completely removedby aspiration, and the well was washed five times with sterile distilledwater. The plates were air-dried for 45 min, and 200 µl crystal violetsolution (0.41% w/v dye, Fisher Scientific) was added per well andincubated at room temperature for 45 min. Crystal violet solution wascompletely removed from wells by aspiration and washed five times

with sterile distilled water. After allowing wells to air dry for 45min,200 µl of 95% ethanol was added to each well, and the contents of thewells were mixed to dissolve the crystal violet dye. Biofilm formationin the well was measured using optical density at 600 nm (OD600) ineach well using a microquant microplate spectrophotometer (BioTekInstruments, Winooski, VT).

2.3. General preparation of produce surfaces

Three types of produce, Iceberg and Romaine lettuce heads (Lactucasativa), and white cabbage heads (Brassica oleracea sp. capitata) wereobtained from a local retail store. For each head of produce, the twooutermost leaf layers were removed aseptically and discarded. Twotypes of produce surfaces were used in the attachment assay. For intactsurfaces, disc-shaped pieces were cut using a sterilized cork-borer (2-cm diameter). To evaluate cut surfaces, strip-shaped pieces were cutfrom the center vein of the leaf at the base of the lettuce head. The veinwas split longitudinally along the centerwith a sterile scalpel and then a2.0×0.5 cm rectangle was cut to acquire pieces with both sidespresenting cut surfaces for bacterial attachments. The upper and lowerepidermis of the leaf was removed. All pieces, disc or strips, were storedin empty 100 mm×15mm petri dishes (Fisher Scientific) with watersoaked Kimwipes® to preserve humidity. Lettuce or cabbage pieceswere stored at 4 °C until the time of inoculation.

2.4. Inoculation of intact and fresh-cut lettuce and cabbage

S. enterica strains from TSA slants were grown in TSB for 24 h at37°C after which cells were pelleted by centrifugation (5000g,15 min) and washed with phosphate buffered saline (PBS, pH 7.2)twice. The cell pellets of each strain were resuspended in anappropriate volume of PBS to obtain an OD600 (Genesys20, Thermo-spectronic, Rochester, NY) of 1. The cell density for each individualserotype was adjusted to approximately 6 log CFU/ml, and wasverified by spiral plating (WASP2, Don Whitley Scientific, Frederick,MD) on XLT4 agar (Acumedia, Lansing, MI).

Bacterial suspensions (11 ml) of each serotypewere transferred intowells of a sterile six-cell culture plate (Fisher Scientific). Discs or strips oflettuce or cabbage were aseptically submerged into Salmonella suspen-sions and incubated at 10°C for 24 h. At specific time intervals (0, 1, 4,and 24 h), pieces were removed from suspensions and dipped in a wellof a culture plate containing 11 ml sterile PBS for 2–3 s to removeresidual cells carried over from the inoculum. Populations of looselyand strongly attached bacteria were determined as described by Ellsand Hansen (2006) with some modifications. Discs or strips weretransferred into a sterile 50-ml centrifuge tube (Fisher Scientific) con-taining 25 ml sterile PBS with 0.1% Tween 20 (Fisher Scientific) andvortexed for 20 s to remove loosely attached Salmonella cells. In order torecover populations of strongly attached Salmonella, the vortexed pieceswere transferred into a 50-ml centrifuge tube containing 25 ml bufferedpeptone water (BPW, Becton Dickinson) and sonicated for 30 s using aPolyTron® homogenizer (Kinematica, Lucerne, Switzerland). ThePolyTronwas sterilized in 70% ethanol between each sample and rinsedtwice in sterile distilledwater to remove residual ethanol. Enumerationof Salmonellawas carried out by spiral plating the appropriately dilutedhomogenates of PBS containing loosely attached cells or BPWcontainingstrongly attached cells on XLT4 agar. Typical Salmonella colonies werecounted after incubation of 24h at 37 °C. Randomly selected colonieswere confirmed by latex agglutination assay (Remel Inc., Lenexa, KS).Pieces of discs or strips inoculated in sterile water served as controls.

2.5. Attachment strength

The attachment strength (SR) was calculated as described byDickson and Koohmaraie (1989). SR value represents the percentageof the total population of bacteria associated with produce surface

43J. Patel, M. Sharma / International Journal of Food Microbiology 139 (2010) 41–47

which were strongly attached to produce surface. [SR=(stronglyattached bacteria) /(strongly attached bacteria+loosely attachedbacteria)].

2.6. Statistical analysis

A randomized complete block design was used with 3 replicatesper treatment. The populations of loosely and strongly attachedSalmonella obtained at each sampling period were converted to logCFU/cm2 for intact and cut surfaces. The data obtained from threereplicates were analyzed by a two-way ANOVA using ‘Proc Mixed’(SAS 8.2, Cary, NC) for interaction effects of the serovar, produce andsampling period. In all cases, the level of statistical significance levelwas of Pb0.05.

3. Results

3.1. Biofilm formation

The results of themicrotiter plate assay comparing biofilm formationin each of four differentmedia by S. enterica serovars are shown in Fig. 1.Biofilm formation was significantly affected by the Salmonella serovarsand the growth medium used. Overall, S. Tennessee and S. Thompsonshowed significantlymore biofilm formation than serovars Braenderup,Negev, and Newport, and were classified as strong biofilm producersaccording to the criteria suggested by Stepanovic et al. (2004). Biofilmformation was significantly lower when these serovars were grown inmedia for 24 h (data not shown) as opposed to 48 h. Biofilm formationin diluted and full strength TSB was significantly higher than in dilutedand full strength LB. Diluted TSB (1:10) promoted the most biofilmformation, followed by full strength TSB. Biofilm formation in LB anddiluted LB (1:10) was not different (PN0.05).

When grown in TSB, S. Tennessee produced significantly more bio-film (0.80±0.19) than the biofilm produced by S. Thompson (0.45±0.17). S. Braenderup (0.33±0.12), S. Negev (0.32±0.12), and S.Newport (0.31±0.09) produced less biofilm than S. Thompson.Similarly, S. Tennessee produced significantly more biofilm (0.77±0.20) than S. Thompson (0.46±0.15) and S. Braenderup (0.45±0.16)when grown in diluted TSB (1:10). Biofilms produced by S.Negev and S.Newport were not different (Pb0.05) from uninoculated dilutedTSB (1:10) control. All strains exhibited poor biofilm formation infull strength and diluted LB growth media. S. Tennessee producedsignificantly higher biofilm formation in LB (0.40±0.05) and 1:10diluted LB (0.38±0.09) compared to other serovars evaluated. Biofilmsformed by S. Braenderup, S. Negev, and S. Newport in LB or 1:10 LBdiluted media were not different (Pb0.05) from the uninoculated LB or1:10 LB controls.

Fig. 1. Biofilm formation of five Salmonella serotypes on polysterene microtiter

3.2. Effect of produce surfaces on Salmonella serovar attachment

Salmonella attachment to Romaine lettuce was significantlygreater than the attachment to Iceberg lettuce or cabbage. In general,attachment of Tennessee serovar was significantly greater than theNegev serovar attachment to produce surfaces. All Salmonella serovarsattached very rapidly on intact and cut produce surfaces. Number ofSalmonella strongly attached to intact surfaces after 5 min (time 0 h)were 1.83, 3.36, and 3.19CFU/cm2 compared to 3.01, 4.44, and4.68 log CFU/cm2 attached to cut surfaces of cabbage, Iceberg lettuceand Romaine lettuce, respectively (Fig. 2). Salmonella attachedpreferentially (at higher numbers) to damaged surfaces of all produce.Initial Salmonella populations (0h) on cut surfaces were 0.3 to1 log CFU/cm2 higher than the populations on intact surface. However,the difference between initial Salmonella attachment (0 h) to intactand cut surfaces for specific produce types was not statistically sig-nificant. For all Salmonella serovars, initial attachment to intact and cutRomaine lettuce was significantly greater than attachment to intact orcut cabbage surface. Further, initial attachment of serovar Tennessee(4.28 log CFU/cm2) was significantly greater than the attachment ofNewport (3.39CFU/cm2) and Negev (3.18CFU/cm2) serovars on intactRomaine lettuce. The number of attached cells on intact and cut surfacesincreased with time for all produce types regardless of Salmonellaserovars. Cut cabbage had populations of Salmonella 1.4 log CFU/cm2

greater than on intact cabbage after 1 h of attachment. For Newport andTennessee serovars, number of attached cells on cut cabbage surfaces at1, 4, and24 hwere significantlyhigher than thenumberof attached cellson intact cabbage at corresponding times. Salmonella Tennessee attach-ment to intact and cut Romaine lettuce after 4 h (4.47 log CFU/cm2 and5.30 log CFU/cm2) was significantly higher than those populations onintact and damaged cabbage (3.15 log CFU/cm2 and 4.23 log CFU/cm2),respectively. After 24 h, Salmonella populations attached to cut cabbagesurface were significantly higher than the population on intact surfacefor all serovars except Negev. Salmonella attachment after 24 h rangedfrom 5.10 to 5.84 log CFU/cm2, 5.29 to 6.17 log CFU/cm2, and 5.66 to6.25 log CFU/cm2 fordamaged surfaces of cabbage, Iceberg andRomainelettuce, respectively.

3.3. Attachment strength of Salmonella serovars

In general, the attachment strength (SR) of Salmonella to cabbagewas significantly lower than that to Iceberg and Romaine lettuce(Fig. 3). The SR values of serovars calculated for intact and damagedsurfaces of all three produce at 0 h varied from 0.08 to 0.26. Initial SRvalues were similar (PN0.05) irrespective of produce or type of sur-face (cut, intact) for Braenderup, Negev, and Newport serovars. Theinitial SR value for serovar Tennessee on intact Romaine lettuce (0.25)

plates. Biofilm formation is measured by optical density at 600nm (OD600).

Fig. 2. Attachment of Salmonella enterica serovars to intact and damaged (a) cabbage (b) Iceberg and (c) Romaine lettuce surfaces over time at 10 °C. Cabbage coupons weresubmerged in suspensions containing 6 log CFU/ml Salmonella. Error bars are the standard deviations from three replications.

44 J. Patel, M. Sharma / International Journal of Food Microbiology 139 (2010) 41–47

was significantly higher than the SR value on intact cabbage (0.05). Ingeneral, although attachment strength increased with time, the dif-ference was not significant between 4 and 24 h. The SR of all Salmonella

serovars on intact Romaine lettuce was significantly higher than theSR on intact or damaged cabbage at 4 and 24 h. Within produce types,there were no differences (PN0.05) in attachment strengths of serovars

Fig. 3. Attachment strength (SR) of Salmonella enterica serovars to intact and damaged (a) cabbage (b) Iceberg and (c) Romaine lettuce surfaces over time at 10 °C. The SR values werecalculated as the ratio of strongly attached cells/(strongly+loosely attached cells).

45J. Patel, M. Sharma / International Journal of Food Microbiology 139 (2010) 41–47

to intact and cut cabbage surfaces at 1, 4 or 24 h sampling times. Similarobservations were found with Iceberg lettuce in most cases. However,the attachment strengths of all serovars except Negev to intact Romainelettuce was significantly higher than that to cut Romaine lettuce at 4 h.Likewise, the attachment strengths of Braenderup, Negev, and Tennes-see serovars to intact Romaine lettuce were significantly higher thantheir attachment strengths to cut Romaine lettuce at 24 h.

4. Discussion

The attachment and biofilm forming abilities provide Salmonellanumerous avenues to contaminate fresh and fresh-cut produce. Ourstudy evaluated the biofilm formation of five Salmonella serovars andits correlation to the attachment to intact and cut surfaces of lettuceand cabbage. In our study, biofilm formation by serovars Tennessee

46 J. Patel, M. Sharma / International Journal of Food Microbiology 139 (2010) 41–47

and Thompson in TSB and diluted TSB (1:10) was significantly greaterthan the other three serovars (Braenderup, Negev and Newport). S.Tennessee displayed the most biofilm formation in vitro and alsoattached toRomaine lettuce at higher populations than other Salmonellaserovars. Stepanovic et al. (2004) hypothesized that stronger biofilmformation in diluted TSB could be due to the induction of biofilm understarvation stress. Our results are in agreement with Kroupitski et al(2009) who reported increased biofilm formation in diluted TSB butnot in diluted LB medium. E. coli O157:H7 cells grown in TSB werehydrophilic in nature and attached better to lettuce surfaces than thosegrown in nutrient broth (Hassan and Frank, 2004). In our study, thedifference in Salmonella biofilm formation when grown in TSB or LBcould be due to the compositional difference of these media.

The initial attachment strength of the Tennessee serovar was alsosignificantly stronger to intact Romaine lettuce than to intact cabbage.Thesefindings indicate that in vitro biofilm formationmay be correlatedto attachment to some produce surfaces, even though in vitro biofilmformation and produce attachment studies were carried out at differenttemperatures. Our results are in agreement with others who reportedcorrelation of microtiter plate assay with biofilm formation for Listeriamonocytogenes (Djordjevic et al., 2002). Otherworkers have shown thatSalmonella Enteritidis strains lacking genes involved in producingcellulose (bcsA) and capsule assembly genes (yihO) attached to alfalfasprouts in lower numbers than wild-type strains (Barak et al., 2007).These same workers showed that strains lacking the bcsA could notattach to surfaces in high numbers in vitro but could still attach toproduce surfaces in numbers similar to the wild-type strains. Differ-ences in our results presented here and thosefindingsmay be due to theuse of different Salmonella serovars, produce types used, and method-ology for evaluating attachment to produce types.

In our study, Salmonella serovars attached rapidly to both intactand damaged surfaces of produce. A significant number of cells wereattached to these surfaces within 5 min (0 h). Populations of Salmonellaattached to produce types increased with time. Some of these increasescould be attributed to the proliferation of attached cells on producesurfaces. Along with an increase in populations of strongly attachedSalmonella, the attachment strength also increased in all three commod-ities over time. Rapid attachment of Salmonella to lettuce and cabbage isconsistentwith the reportsof other investigators. Listeriamonocytogenesattached to intact and cut lettuce surfaceswithin 5 minwhen incubatedat 10, 22 or 37 °C (Ells and Hansen, 2006). Almost immediate attach-ment of Salmonella cells after introduction has been demonstrated totomatoes (Iturriaga et al., 2003). Salmonella attached to green peppersurfaces within 30s of inoculation (Liao and Cooke, 2001).

The attachment strength of Salmonella serovars was significantlyhigher onRomaine lettuceover intact cabbage. Salmonella cells can use acombination of cellulose, thin aggregative fimbriae, and O-antigen asfactors to attach to produce surfaces (Barak et al., 2007). Other authorshave pointed to a role for non-hemagglutinating pili, fibrillae andflagella in bacterial attachment to produce surfaces (Silagyi et al., 2009).A recent study by Kroupitski et al (2009) reported comparable attach-ment of Salmonella Typhimurium to intact and cut lettuce when incu-bated at 4 °C for 18 h, and other authors have stated that Salmonella didnot differ in levels of attachment to cut and intact lettuce surfaces(Takeuchi et al, 2000).

The presence of extracellular substances can influence the surfacehydrophobicity and charge of bacterial cells (Ryu et al., 2004). Bacteriawith greater hydrophobic characteristics may attach to the cuticle, ahydrophobic layer on plant surfaces composed of fatty acids, poly-saccharides and waxes (Solomon and Sharma, 2009). The surface ofintact produce that is covered by a hydrophobicwaxy cuticlemay allowhydrophobic Salmonella cells to attach to the waxy cuticle. However,breaks in the cuticle exposes hydrophilic structures from withinallowing intimate contact between bacterial cells and the leaf, andmay release previously unavailable nutrients to enteric bacteria,makingthen good sites for colonization.

In our work presented here, there may be several possibilities forwhy the attachment strength of Salmonella to intact and lettucesurface may increase over 24 h. Salmonella cells infiltrated niches onthe leaf surface, like areas around stomata and in cell margins wheremicrocolonies and biofilms of epiphytic bacteria colonize (Warneret al., 2008). Other workers have demonstrated that Salmonella maysurvive for longer durations when they are associated with aggregatesof other bacteria on the leaf surface (Monier and Lindow, 2005).Infiltration and association with aggregates may prevent Salmonellafrom being dislodged from the foliar surface during agitation, andmayhave resulted in increased attachment strength over time. Theconsistent lower attachment of Salmonella to cabbage surface maybe a result of a less damaged cuticle providing fewer niches and sitesof colonization for Salmonella cells.

Our study shows that attachment of Salmonella to lettuce andcabbage surfaces can differ based on specific properties of the serovarsand commodities being evaluated. In this limited instance, the for-mation of biofilm by S. Tennesseewas a good indicator of its attachmentstrength to lettuce surfaces. These studies also reveal that Salmonellaserovars more strongly attach to lettuce than to cabbage, regardless ofthe condition (cut, intact) of the produce surface. More studies areneeded to investigate the interactions between produce surfaces andother strains of Salmonella serovars to devise effective interventions andstrategies to decrease contamination of leafy greens.

Acknowledgements

The authors thank Katherine Hopkins and Ernie Paroczay fortechnical assistance, and Dr. Bryan Vinyard for statistical analysis.

References

Anonymous, 2009. National Salmonella Typhimurium Outbreak Linked to Lettuce?Available at: http://www.foodpoisonjournal.com/2009/09/articles/foodborne-ill-ness-outbreaks/national-salmonella-typhimurium-outbreak-linked-to-lettuce/2009Accessed September 23, 2009.

Aruscavage, D., Lee, K., Miller, S., LeJeune, J.T., 2006. Interactions affecting the proliferationand control of human pathogens on edible plants. Journal of Food Science 71,R89–R99.

Barak, J.D., Whitehand, L.C., Charkowski, A.O., 2002. Differences in attachment ofSalmonella enterica serovars and Escherichia coliO157:H7 to alfalfa sprouts. Appliedand Environmental Microbiology 68, 4758–4763.

Barak, J.D., Jahn, C.E., Gibson, D.L., Charkowski, A.O., 2007. The role of cellulose and O-antigen capsule in the colonization of plants by Salmonella enterica. MolecularPlant–Microbe Interactions 20, 1083–1091.

Berger, C.N., Shaw, R.K., Brown, D.J., Mather, H., Clare, S., Dougan, G., Pallen, M.J.,Frankel, G., 2009. Interaction of Salmonella enterica with basil and other saladleaves. ISME Journal 3, 261–265.

Beuchat, L.R., 2002. Ecological factors influencing survival and growth of humanpathogens on raw fruits and vegetables. Microbes and Infection 4, 413–423.

Boyer, R.R., Sumner, S.S., Williams, R.C., Pierson, M.D., Popham, D.L., Kniel, K.E., 2007.Influence of curli expression by Escherichia coli O157:H7 on the cell's overallhydrophobicity, charge, and ability to attach to lettuce. Journal of Food Protection70, 1339–1345.

CDC, 2006. Ongoing multistate outbreak of Escherichia coli serotype O157:H7 infectionsassociated with consumption of fresh spinach—United States. Morbidity andMortality Weekly Report 55, 1045–1046.

Dickson, J.S., Koohmaraie, M., 1989. Cell surface charge characteristics and theirrelationship to bacterial attachment to meat surfaces. Applied and EnvironmentalMicrobiology 55, 832–836.

Djordjevic, D., Wiedmann, M., McLandsborough, L.A., 2002. Microtiter plate assay forassessment of Listeria monocytogenes biofilm formation. Applied and Environmen-tal Microbiology 68, 2950–2958.

Ells, T.C., Hansen, L.T., 2006. Strain and growth temperature influence Listeria spp.attachment to intact and cut cabbage. International Journal of Food Microbiology111, 34–42.

FDA, 2009. Guidance for Industry: Guide to Minimize Microbial Food Safety Hazards ofLeafy Greens; Draft GuidanceAvailable at: http://www.fda.gov/Food/Guidance-ComplianceRegulatoryInformation/GuidanceDocuments/ProduceandPlanPro-ducts/ucm174200.htm2009Accessed September 1, 2009.

Hassan, A.N., Frank, J.F., 2004. Attachment of Escherichia coli O157:H7 grown in trypticsoy broth and nutrient broth to apple and lettuce surfaces as related to cellhydrophobicity, surface charge, and capsule production. International Journal ofFood Microbiology 96, 103–109.

47J. Patel, M. Sharma / International Journal of Food Microbiology 139 (2010) 41–47

Herman, K., Ayers, T.L., Lynch, M., 2008. Foodborne disease outbreaks associated withleafy greens, 1973–2006. International Conference on Merging Infectious Diseases,March 16–19 Atlanta, GA.

Iturriaga, M.H., Escartin, E.F., Beuchat, L.R., Martinez-Peniche, R., 2003. Effect ofinoculum size, relative humidity, storage temperature, and ripening stage on theattachment of Salmonella Montevideo to tomatoes and tomatillos. Journal of FoodProtection 66, 1756–1761.

Kroupitski, Y., Pinto, R., Brandl, M.T., Belausov, E., Sela, S., 2009. Interactions of Salmonellaenterica with lettuce leaves. Journal of Applied Microbiology 106, 1876–1885.

Liao, C.-H., Cooke, P.H., 2001. Response to trisodium phosphate treatment of SalmonellaChester attached to fresh-cut green pepper slices. Canadian Journal of Micro-biolology 47, 25–32.

Monier, J.-M., Lindow, S.E., 2005. Aggregates of resident bacteria facilitate survival ofimmigrant bacteria on leaf surfaces. Microbial Ecology 49, 343–352.

Ryu, J.H., Kim, H., Beuchat, L.R., 2004. Attachment and biofilm formation by Escherichiacoli O157:H7 on stainless steel as influenced by exopolysaccharide production,nutrient availability, and temperature. Journal of Food Protection 67, 2123–2131.

Sewell, A.M., Farber, J.M., 2001. Foodborne outbreaks in Canada linked to produce.Journal of Food Protection 64, 1863–1877.

Silagyi, K., Kim, S.-H., Lo, Y.M., Wei, C-i, 2009. Production of biofilm and quorum sensingby Escherichia coli O157:H7 and its transfer from contact surfaces to meat, poultry,ready-to-eat deli, and produce products. Food Microbiology 26, 514–519.

Sivapalasingam, S., Friedman, C.R., Cohen, L., Tauxe, R.V., 2004. Fresh produce: agrowing cause of outbreaks of foodborne illness in the United States, 1973 through1997. Journal of Food Protection 67, 2342–2353.

Solomon, E.B., Sharma, M., 2009. Microbial attachment and limitations of decontam-ination methodologies. In: Sapers, G.M., Solomon, E.B., Matthews, K.R. (Eds.), TheProduce Contamination Problem: Causes and Solutions. Academic Press, New York,pp. 21–39.

Stepanovic, S., Cirkovic, I., Ranin, L., Svabic-Vlahovic, 2004. Biofilm formation bySalmonella spp. and Listeria monocytogenes on plastic surface. Letters in AppliedMicrobiology 38, 428–432.

Suslow, T.V., Oria, M.P., Beuchat, L.R., Garrett, E.H., Parish, M.E., Harris, L.J., Farber, J.N.,Busta, F.F., 2003. Production practices as risk factors inmicrobial food safety of freshand fresh-cut produce. Comprehensive Reviews in Food Science and Food Safety 2,38–77.

Takeuchi, K., Matute, C.M., Hassan, A.N., Frank, J.F., 2000. Comparison of the attachmentof Escherichia coli O157:H7, Listeria monocytogenes, Salmonella Typhimurium, andPseudomonas fluorescens to lettuce leaves. Journal of Food Protection 63,1433–1437.

Warner, J.C., Rothwell, S.D., Keevil, C.W., 2008. Use of episcopic differential interferencecontrast microscopy to identify bacterial biofilms on salad leaves and trackcolonization by Salmonella Thompson. Environmental Microbiology 10, 918–925.