APPLIED AND ENVIRONMENTAL MICROBIOLOGY, May 2011, p. 3279–3292 Vol. 77, No. 100099-2240/11/$12.00 doi:10.1128/AEM.00546-11Copyright © 2011, American Society for Microbiology. All Rights Reserved.

comK Prophage Junction Fragments as Markers for Listeria monocytogenesGenotypes Unique to Individual Meat and Poultry Processing Plants

and a Model for Rapid Niche-Specific Adaptation, BiofilmFormation, and Persistence�†

Bindhu Verghese,1§ Mei Lok,1§‡ Jia Wen,1 Valentina Alessandria,2 Yi Chen,3Sophia Kathariou,4 and Stephen Knabel1*

Department of Food Science, The Pennsylvania State University, University Park, Pennsylvania 168021; Department of Exploitation andProtection of Agricultural and Forest Resources, Agricultural Microbiology and Food Technology Sector, Faculty of Agriculture,University of Turin, Turin, Italy2; Microbial Methods Development Branch, Division of Microbiology, Office of Regulatory Science,

Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, College Park, Maryland 207403; andDepartment of Food Science, North Carolina State University, Raleigh, North Carolina 276954

Received 9 March 2011/Accepted 14 March 2011

Different strains of Listeria monocytogenes are well known to persist in individual food processing plants andto contaminate foods for many years; however, the specific genotypic and phenotypic mechanisms responsiblefor persistence of these unique strains remain largely unknown. Based on sequences in comK prophage junctionfragments, different strains of epidemic clones (ECs), which included ECII, ECIII, and ECV, were identifiedand shown to be specific to individual meat and poultry processing plants. The comK prophage-containingstrains showed significantly higher cell densities after incubation at 30°C for 48 h on meat and poultryfood-conditioning films than did strains lacking the comK prophage (P < 0.05). Overall, the type of strain, thetype of conditioning film, and the interaction between the two were all highly significant (P < 0.001).Recombination analysis indicated that the comK prophage junction fragments in these strains had evolved dueto extensive recombination. Based on the results of the present study, we propose a novel model in which theconcept of defective comK prophage was replaced with the rapid adaptation island (RAI). Genes within the RAIwere recharacterized as “adaptons,” as these genes may allow L. monocytogenes to rapidly adapt to differentfood processing facilities and foods. If confirmed, the model presented would help explain Listeria’s rapid nicheadaptation, biofilm formation, persistence, and subsequent transmission to foods. Also, comK prophagejunction fragment sequences may permit accurate tracking of persistent strains back to and within individualfood processing operations and thus allow the design of more effective intervention strategies to reducecontamination and enhance food safety.

Listeria monocytogenes is a unique food-borne pathogen thatcauses listeriosis, which ranges from febrile gastroenteritis tomore severe life-threatening invasive diseases, especially forimmunocompromised individuals (94). It is widely distributedin many wild and domestic animals and various natural envi-ronments and is resistant to a wide variety of environmentalstresses (33). L. monocytogenes is considered a model organismfor the study of host-pathogen interactions, especially as amodel for intracellular pathogens of humans (47). However, L.monocytogenes may also be an excellent model for pathogen-environment interactions, because it is well known to cyclebetween being a pathogen in many wild and domestic animalsand a saprophyte in diverse environments, including thosefound in various types of food processing facilities (41). How-

ever, while much is known about the pathogenic lifestyle of L.monocytogenes, much less is known about its saprophytic life-style, including the genetic and phenotypic factors affecting itscolonization and persistence in food processing and retail en-vironments and subsequent transmission to ready-to-eat(RTE) foods. This has resulted in numerous costly recalls,disease cases, and outbreaks, which are often associated withhigh mortality (94). As a result, a zero-tolerance policy for L.monocytogenes currently exists for RTE meat and poultry prod-ucts manufactured in the United States.

Specific resident strains of L. monocytogenes have beenfound to persist for months to more than a decade (70, 78, 95,105) in various types of food processing plants manufacturingmeat products (5, 38, 40, 90, 105), poultry products (4–6, 32,78, 85), dairy products (4, 5, 37, 48, 53, 101), seafood products(4, 5, 73, 86, 95), vegetables (5), and sandwiches (8) and also inretail environments (88). These persistent strains are geneti-cally distinct from transient strains that are isolated sporadi-cally (4, 5). Persistent strains of L. monocytogenes show greateradherence to food-contact surfaces than do nonpersistentstrains (66, 74) and as a result contaminate foods more fre-quently than do sporadic strains (5, 48). Autio et al. (5) foundthat some pulsotypes of L. monocytogenes were repeatedlyfound in the same food product from the same producer while

* Corresponding author. Mailing address: 437 Food Science Build-ing, Department of Food Science, The Pennsylvania State University,University Park, PA 16802. Phone: (814) 863-1372. Fax: (814) 863-6132. E-mail: sjk9@psu.edu.

§ Contributed equally.‡ Present address: Strategic Diagnostics Inc., 128 Sandy Drive, New-

ark, DE 19713.† Supplemental material for this article may be found at http://aem

.asm.org/.� Published ahead of print on 25 March 2011.

3279

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

other pulsotypes were repeatedly found in products from mul-tiple producers. While many reports have clarified the roles ofnumerous virulence genes and their coordinated expression incausing listeriosis, the genotypic and phenotypic mechanismsresponsible for the persistence of different strains of L. mono-cytogenes in individual food processing and retail facilities re-main to be elucidated.

Most bacteria, including L. monocytogenes, grow on surfacesand form biofilms (12, 93). Cells of L. monocytogenes withinbiofilms are more resistant to biocides (102), which increasesthe risk of food contamination. Initial adherence is critical forbiofilm formation and depends on the physiochemical proper-ties of the environmental surfaces as well as the biofilm-form-ing potential of the bacterial cells (18, 83, 102). L. monocyto-genes can adhere to abiotic surfaces such as stainless steel,glass, plastic, polymers, and rubber that are present in the foodprocessing environment (44, 102, 107). These bacteria alsoadhere to biotic surfaces such as other microorganisms andplant and animal tissues (44). Strains of L. monocytogenes showspecificity in forming biofilms. For example, serotype 1/2astrains have been reported to adhere faster to surfaces and toproduce larger biofilms than do other serotypes (35, 54). How-ever, L. monocytogenes strains that are low biofilm producershave been reported to form dense biofilms in the presence ofa strong biofilm producer (80). Various studies have demon-strated that specific genes in L. monocytogenes are required fordifferent phases of biofilm formation. In L. monocytogenes, agr(accessory gene regulator) and ami (autolysin-adhesion gene)mutants were shown to be defective in the initial attachment toa surface (57, 84, 91). Also, genomic studies exploring geneexpression during growth of L. monocytogenes on a surfacedemonstrated that the DegU orphan response regulator wasimportant for biofilm formation (43, 57).

Conditioning films are well known to enhance the adherenceof various microorganisms to numerous surfaces in naturalenvironments (29, 39) and the surfaces of teeth (27, 56) andimplanted medical devices (24). Bowden and Li (11) stressedthe importance of nutritional influences of conditioning filmson biofilm development. They indicated that conditioning filmsmay act as important nutrient sources and that competition forsuch nutrients may serve as a strong selective force in theevolution of strains that form biofilms and colonize varioussurfaces. Many authors have speculated that biofilm formationis the reason for the persistence of specific strains in foodprocessing plants. However, other explanations for this phe-nomenon have also been suggested, including continual rein-troduction of the same strain over extended periods of time,random primary colonization, ability to grow and survive at lowtemperatures, competitiveness for nutrients, ability to mount astringent response and undergo physiological adaptation tonutrient deprivation, resistance to sanitizers and/or heavy met-als and antibiotics, ability to interact with other microorgan-isms to form a stable ecosystem within biofilms, or some com-bination of the above (70, 78, 83, 105).

In order to implement more effective intervention strategiesto prevent persistent strains of L. monocytogenes from contam-inating RTE foods, we must first identify their reservoirs andunderstand how they are transmitted from these reservoirs tofoods. To accomplish this, it is critical to develop an accuratemolecular subtyping method by selecting the right molecular

subtyping marker(s) and applying it to a relevant collectionof strains (36, 94). Multi-virulence-locus sequence typing(MVLST) of L. monocytogenes very accurately identified anddifferentiated the four epidemic clones (ECs) of L. monocyto-genes (ECI, -II, -III, and -IV), outbreak clones that did notbelong to a specific epidemic clone, and nonoutbreak strains(21, 65). However, MVLST was not able to differentiate out-break clones within epidemic clones of L. monocytogenes (21,65). To overcome this limitation, Chen and Knabel (20) tar-geted regions within the comK and PSA prophages of L. mono-cytogenes and designed PCR-based approaches that accuratelydifferentiated different outbreak clones within epidemicclones. Orsi et al. (78) subsequently performed whole-genomesequencing on lineage II serotype 1/2a ECIII strains of L.monocytogenes from a 1988 sporadic case and the 2000 out-break linked to RTE poultry products from the same process-ing plant in Texas. Their findings supported the conclusion ofChen and Knabel (20) that prophages in L. monocytogenes areexcellent markers for differentiating outbreak clones withinepidemic clones. The results of Orsi et al. (78) also revealedextensive recombination throughout the comK prophage inECIII, which they suggested was due to recombination be-tween the comK prophage in the ECIII strain in lineage II andthe comK prophage in a serotype 1/2b strain in lineage I.

Most species of bacteria, including L. monocytogenes (49,62), contain prophage DNA which can interact with infectingphage (49, 50). As a result, phage-prophage interactions arethought to be a dynamic force acting on bacterial populationsand to account for a major portion of bacterial evolution thatis occurring via lateral gene transfer (13, 15, 49, 50, 52). There-fore, temperate phages which integrate their genomes intohost genomes could be a major source of novel genes thatenhance the fitness of the bacterial host (34, 46, 49, 52, 103).Wang et al. (103) demonstrated that cryptic (defective) pro-phages in Escherichia coli K-12 enhance its resistance to os-motic, oxidative, and acid stresses and increase growth andbiofilm formation. Prophages, including those in L. monocyto-genes, constitute the major differences between genomes ofclosely related strains of bacteria (13). This is especially thecase for epidemic clones and outbreak clones of L. monocyto-genes, which are highly clonal in terms of their backbone ge-nome (22, 72, 78, 81, 82). Extensive mosaicism is typical ofprophages in many prokaryotes (71), including the comK pro-phage of L. monocytogenes (78), which is rendered inactive(defective) by gene deletions following insertion into the comKgene (16).

The Food Safety and Inspection Service (FSIS) is responsi-ble for monitoring the safety of RTE meat and poultry prod-ucts manufactured in the United States. As part of this respon-sibility, FSIS conducts risk-based L. monocytogenes testingprograms in processing plants producing postlethality-exposedRTE meat and poultry products. As a result of its L. monocy-togenes sampling program, FSIS has generated more than 500L. monocytogenes isolates from different RTE meat processingfacilities throughout the United States (104). Small fractions(�4%) of these isolates were ECII based on pulsed-field gelelectrophoresis (PFGE) profiles and multilocus genotyping(MLGT) (104); however, these methods could not prove con-clusively that these isolates were ECII nor determine whetherthey represented different genotypes of ECII. Eifert et al. (32)

3280 VERGHESE ET AL. APPL. ENVIRON. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

also isolated L. monocytogenes strains that had ECII markersand PFGE patterns similar to those of the 1998 and 2002 L.monocytogenes outbreak clones (32) from two different turkeyprocessing plants from nonadjoining states in the UnitedStates. L. monocytogenes serotype 1/2a epidemic clones, ECIII(76) and ECV (2008 Canadian outbreak; S. Knabel, A. Re-imer, B. Verghese, M. Lok, J. Zergler, J. Farber, F. Pagotto,M. Graham, C. A. Nadon, Canadian Public Health LaboratoryNetwork (CPHLN), and M. W. Gilmour, unpublished data),also appear to persist within RTE meat and poultry processingplants. Therefore, the purposes of the present study were (i) toconfirm that the putative ECII isolates were ECII; (ii) to de-termine the genotypes of persistent strains of ECII, ECIII, andECV by amplifying and sequencing comK prophage junctionfragments; (iii) to assess whether recombination was occurringin these junction fragments; (iv) to determine the capacities ofthe different comK-prophage-containing genotypes to attach,grow, and form biofilms on RTE meat- and poultry-condition-ing films, compared to those of strains lacking the comK pro-phage; and (v) to develop a genotypic/phenotypic model thatmight explain how sequence variations in the comK prophagemight drive rapid niche-specific adaptation, biofilm formation,persistence, and transmission of L. monocytogenes.

MATERIALS AND METHODS

Bacterial isolates and DNA extraction. Isolate identification numbers, sources,and dates of isolation are given in Table S1 in the supplemental material. Tenisolates were obtained from the Listeria collection at CDC, and 18 isolates wereobtained from FSIS of the U.S. Department of Agriculture (USDA). For the 10isolates obtained from CDC, six were associated with two outbreaks of listeriosisinvolving ECII and four isolates were ECIII isolates from the same plant inTexas (two isolates associated with one sporadic case in 1988 and two isolatesassociated with one outbreak of listeriosis in 2000). The 18 isolates obtained fromFSIS were isolated by FSIS at different times from different meat processingfacilities as part of their L. monocytogenes testing programs. FSIS had deter-mined that these isolates had PFGE and MLGT subtypes similar to those ofpreviously characterized ECII isolates (104). Bacterial isolates were streaked ontryptic soy agar with 0.6% yeast extract (TSAYE) (BD, Franklin Lakes, NJ)with incubation at 35°C for 24 h. For each strain, one colony on the plate wasinoculated into 10 ml of tryptic soy broth with yeast extract (TSBYE) andthen incubated at 35°C overnight. Cultures grown overnight were adjusted toan optical density of 0.2 at 650 nm, which is equivalent to approximately 107

CFU/ml. For all isolates, bacterial genomic DNAs were extracted using anUltraClean microbial DNA extraction kit (Mo Bio Laboratories, SolanaBeach, CA) and stored at �20°C before use. Similar methods were used forgenomic DNA extraction from 10 isolates from the work of Eifert et al. (32)in S. Kathariou’s lab.

ECII PCR. PCR amplifications were performed using PCR master mix (QiagenInc., Valencia, CA) and ECII-specific primers (see Table S2 in the supplementalmaterial) with a Mastercycler gradient (Eppendorf Scientific Corp., Hamburg,Germany). Amplification conditions were 94°C for 15 min followed by 31 cyclesof 94°C for 30 s, 59°C for 1.5 min, and 72°C for 1.5 min and a final extension at72°C for 10 min.

MVLST. Intragenic regions of the six virulence genes (prfA, inlB, inlC, dal,clpP, and lisR) were amplified as previously described by Chen et al. (21).

Outbreak clone PCR. PCR amplification of the hypervariable comK prophageregion in the 1998 and 2002 ECII outbreak strains (LMOh7858_2426) wasperformed as previously described by Chen and Knabel (20).

Amplifications of upstream and downstream junction fragments. Amplifica-tion of the upstream and downstream junction fragments within comK wasperformed using methods modified from the work of Loessner et al. (64). Thesemodifications included redesigned primers (see Table S2 in the supplementalmaterial) and different cycling conditions. A single PCR program was used forboth upstream and downstream junction fragment amplifications (initial dena-turation at 95°C for 15 min followed by 40 cycles of 94°C for 30 s, 63°C for 1.5min, and 72°C for 1.5 min and a final extension at 72°C for 10 min).

DNA sequencing. All PCR products were purified using ExoSAP-IT (USBCorp., Cleveland, OH) prior to sequencing. DNA sequencing was performed atthe Pennsylvania State University Genomics Core Facility. The sequences of twoclinical isolates from the 2008 Canada outbreak (ECV) were downloaded fromNCBI GenBank.

Sequence analysis. Sequence analyses were performed using DNAStar soft-ware (v8.1) (DNAStar Inc.). Multiple sequence alignments were performedusing MEGA (version 4.0.2) (97). The unweighted pair group method witharithmetic mean (UPGMA) implemented in MEGA was used to construct clus-ter diagrams using the upstream and downstream junction fragment sequences ofL. monocytogenes isolates. One thousand random tree replications were per-formed, and bootstrap values were assigned to the nodes of the cluster diagram(see Fig. 2).

Recombination analysis. The extent of recombination within the upstream anddownstream junction fragments of L. monocytogenes was determined using theRecombination Detection Program (RDP), GENECONV, maximum chi-square,Chimera, and Sister Scan recombination detection methods implemented inRDP v.3.38 (68). Since no single program provided optimal performance underall conditions, a stringent criterion was applied, such that any event for detectionof positive recombination breakpoints had to be supported by five or moremethods with P values of �10�5.

Detection of spontaneous induction of the comK prophage. Spontaneous in-duction of the comK prophage was detected by attP and attB PCR. Amplifica-tions of the attP site within comK phage (Fig. 1A) and the attB attachment sitewithin comK in the L. monocytogenes genome (Fig. 1B) were performed usingmethods modified from the work of Loessner et al. (64). The modificationsincluded redesigned primers (see Table S2 in the supplemental material) andcycling conditions as described above for amplification of the upstream anddownstream junction fragments.

Detection of int gene in L. monocytogenes genomic DNA and comK phage DNAby PCR. The phage integrase gene (int) in L. monocytogenes genomic DNA andcomK phage DNA was amplified by PCR in strains F2365 (ECI), H7858 (ECII),N3-031 (ECIII), and 1001::A118 using primers described in Table S2 in thesupplemental material. ECI strain F2365 was isolated from Mexican-style softcheese during the 1985 California outbreak. The ECII and ECIII strains aredescribed in Table 4. L. monocytogenes strain 1001::A118 is an artificially lyso-genized serotype 1/2a strain. The cells were grown to early log phase (107

CFU/ml) at 35°C in tryptose broth (TB) supplemented with CaCl2 to a finalconcentration of 10 mM. Cells were centrifuged at 12,000 rpm for 1 min andresuspended in 5 ml of fresh TB. After overnight incubation in TB at 35°C,samples were centrifuged at 14,000 rpm for 10 min and the supernatant wasfiltered through a 0.2-�m sterile membrane filter (Thermo Scientific, Waltham,MA). Genomic DNA extraction was performed as described above. Phage DNAextraction was conducted as described by Asadulghani et al. (3). Briefly, phageDNA in the phage head was detected by PCR amplification of int after DNase(Rockland, Gilbertsville, PA), RNase (EMD, Gibbstown, NJ), and proteinase K(Rockland, Gilbertsville, PA) treatments. PCR amplification conditions were94°C for 15 min followed by 30 cycles of 94°C for 30 s, 62°C for 1.5 min, and 72°Cfor 1.5 min and a final extension at 72°C for 10 min. Phage DNA amplificationwas repeated at least three times for each isolate, and int was confirmed bysequencing. The int primer sequences for detection of phage DNA are shown inTable S2 in the supplemental material.

Plaque assay. To determine whether the prophages in strains H7858 (ECII)and N3-031 (ECIIII) were active or defective, a plaque assay was performed onthese strains. All media used in the plaque assay were supplemented with CaCl2to a final concentration of 10 mM. Cells were grown at 35°C in TB to early logphase (107 CFU/ml), centrifuged at 12,000 rpm for 1 min, and then resuspendedin 1.8 ml of fresh TB. After overnight incubation in TB at 35°C, samples werecentrifuged at 14,000 rpm for 10 min and the supernatant was filtered through a0.2-�m sterile membrane filter (Thermo Scientific, Waltham, MA). One hundredmicroliters of filtrate was mixed with 100 �l of exponential-phase cells of theserotype 1/2a indicator strain 1001 (provided by Martin Loessner, ETH, Zurich,Switzerland) or 4b indicator strain ATCC 23074; the mixture was added to 5 mlof molten TSAYE containing 0.75% agar at 50°C and then poured onto solidifiedTSAYE containing 1.5% agar. Plaque formation was observed after overnightincubation at 30°C. The plaque assay was also performed on strains F2365 (ECI)and 1001::A118. Each plaque assay was replicated twice.

Epifluorescence microscopy. Epifluorescence microscopy was used to studythe effects of type of strain of L. monocytogenes, type of food-conditioning film,and their interaction on the density of L. monocytogenes cells on glass slides afterincubation at 30°C for 48 h. Seven strains of L. monocytogenes were used in thisstudy, including two strains that lacked the comK prophage (a lineage III strain,W1-111, and a lineage I ECI strain, F2365) and five strains that contained the

VOL. 77, 2011 comK PROPHAGE JUNCTION FRAGMENTS 3281

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

comK prophage (three ECII strains [J1703, H7858, and OB020790], an ECIIIstrain [N3-031], and an ECV strain [08-5923]). Sterile TSBYE containing no cellsserved as the negative control for strains. Five large, unsliced, unopened cookedRTE foods from a local supermarket deli were used to make the food-condi-tioning films. These foods included Brie soft cheese, pork and beef hot dogs(containing salt, potassium lactate, sodium phosphate, sodium diacetate, sodiumerythorbate, and sodium nitrite), fully cooked turkey breast (containing salt andsodium phosphate), pasteurized canned ham (containing salt, sodium phosphate,and sodium nitrite), and chicken breast (containing salt, potassium lactate, so-dium phosphate, and sodium diacetate). Briefly, all foods (except the cannedham) were steamed for 60 min to destroy all vegetative cells that might bepresent on the surface of the foods and then cooled in an ice-water bath. Tengrams of each food was then aseptically sampled from the center of the foodproducts and blended with 40 g of sterile water in a sterile blender jar using anOsterizer blender (Oster, Shelton, CT). Sixty microliters of the homogeneousfood slurry of each food was transferred into each compartment of an eight-compartment CultureSlide (Falcon; Becton Dickinson, Franklin Lakes, NJ) (seeFig. S1 in the supplemental material). Food slurries in CultureSlide compart-ments were then air dried at 35°C for 3 h to form food-conditioning films on thesurface of the glass CultureSlides. Sterile water containing no food served as thenegative control for food-conditioning film.

The protocol for preparation of inoculates and fluorescent microscopy wasadapted from the report by Kushwaha and Muriana (58). Briefly, all strains of L.monocytogenes were incubated in TSBYE at 35°C for 17 h and then diluted10�5-fold with sterile TSBYE. Two hundred microliters of diluted culture ofeach strain was then added to each compartment of the CultureSlide (except forthe no-L. monocytogenes control compartment), and then the CultureSlides wereincubated at 30°C for 48 h (see Fig. S1 in the supplemental material for exper-imental design). After incubation, each compartment was rinsed three timesusing 200 �l of Tris buffer (pH 7.4; 0.05 M) and then stained using 200 �l of5,6-carboxy-fluorescein diacetate (5,6-CFDA; Invitrogen, Carlsbad, CA) solutionwith incubation at 25°C for 15 min. After fluorescent staining, each compartmentwas rinsed three times with Tris buffer (pH 7.4; 0.05 M). The chambers of theCultureSlides were then removed, and the remaining slides were examined using

a BX51 fluorescence microscope (excitation wavelength, �490 nm; detectionwavelength, �510 nm) equipped with a DP20 camera (Olympus Optical, Tokyo,Japan). Pictures of each slide were taken at magnifications of �100 and 400�,and the cell density of each strain on each food-conditioning film was analyzedvisually and given cell density scores from 0 (no cells present) through 1 (verylow), 2 (low), 3 (moderate), and 4 (high) to 5 (very high) by three separateindividuals using visual scoring standards (see Fig. 3A). The experiment wasreplicated twice, and mean scores were calculated and analyzed statistically usinganalysis of variance (ANOVA) with Minitab software (version 16.0; Minitab,State College, PA). Pairwise comparisons were made by using Tukey’s leastsignificant difference test (� � 0.05).

Sequence analysis of comK prophage in 4b and 1/2a strains. Sequence analyseswere conducted on the comK prophage regions in the sequenced genomes of L.monocytogenes serotype 1/2a and 4b using BLAST, PSI_BLAST (1), and BLAST2 (98). Homology searches were performed with both DNA and protein se-quences with various levels of stringency to detect genes that have 100% se-quence identity within an outbreak clone and unique genes within a serotype.BLAST searches for phage genes were also carried out on comK prophagesequences of ECII, ECIII, and ECV against the NCBI GenBank database.

Nucleotide sequence accession numbers. Gene sequences were deposited intoGenBank under accession numbers JF791254 to JF791319.

RESULTS

ECII PCR. Positive amplifications of the ECII marker (19)indicated that all isolates from FSIS and the work of Eifert etal. (32) were ECII (Table 1).

MVLST. Isolates from FSIS and the work of Eifert et al. (32)had the same MVLST sequence type (virulence type 19[VT19]) as that of ECII (21). MVLST analysis of the two fullysequenced 2008 Canada outbreak isolates (ECV) revealed that

FIG. 1. Schematic diagram of phage A118 integration into and excision out of the L. monocytogenes chromosome at comK. (A) Phage A118with attP attachment site. Horizontal arrows indicate PCR priming sites for amplifying the fragment containing attP. (Modified from reference 64with permission of John Wiley & Sons.) (B) L. monocytogenes comK gene containing attB attachment site. Horizontal arrows indicate priming sitesfor amplifying comK with attB attachment site. (C) Lysogenized L. monocytogenes showing the locations of the forward and reverse primers(horizontal arrows) for PCR amplification of upstream and downstream comK prophage junction fragments. The location of gene locusLMOh7858_2426 in the lysogenized strain (20) is shown with an upward-pointing arrow. (D) Lysogenized L. monocytogenes showing the PCRtargets for the upstream and downstream comK prophage junction fragment PCR (RBS, ribosomal binding protein; NC, noncoding region; att,attachment site). The partial gene targets are underlined.

3282 VERGHESE ET AL. APPL. ENVIRON. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

they had an MVLST sequence type that was distinct from thatof ECIII (VT1) or any other strain in our collection, and thus,they were classified into their own unique virulence type(VT59) in serotype 1/2a in lineage II (Knabel et al., unpub-lished).

Outbreak clone PCR. Two primer pairs (2426_1998 and2426_2002) (see Table S2 in the supplemental material) tar-geting the prophage region LMOh7858_2426 were able to dif-ferentiate between the 1998-1999 and 2002 outbreak clones ofECII based on the sizes of the amplified products (20). Isolatesfrom plants F, G, J, and O had only the 1998-1999 ECIIoutbreak clone profile, the isolate from plant I had the 2002ECII outbreak clone profile, and isolates from plant N hadboth the 1998-1999 and 2002 ECII outbreak clone profiles(Table 1). However, isolates from plant H (see Fig. S2) andplants K, L, and M (data not shown) showed no amplificationwith either of the above primer pairs.

Amplification of upstream and downstream junction frag-ments. PCR experiments targeting the upstream and down-stream junction fragments were conducted to determinewhether or not the comK prophage was absent in isolates thatfailed to amplify using the above prophage PCR assay. Positiveamplifications of the downstream junction fragment were seenin all isolates tested, including isolates from plant H (see Fig.S3 in the supplemental material), which as mentioned abovehad previously shown no amplification with prophage primerpair 2426_1998 or 2426_2002. Amplification of the upstreamand downstream junction fragments was observed among allisolates analyzed in the present study using modified primerpairs. Sequencing results for ECII isolates from plants K andM (both from New York) and plant L (from New Jersey)indicated that the comK prophage was present. However, morethan one product was amplified using these junction fragmentprimer pairs (data not shown), even when cycling conditionswere modified and alternative primer pairs targeting upstreamand downstream junction fragments were employed. There-fore, these isolates were not included in the present study.

Cluster diagrams were constructed based on the upstream

and downstream junction fragment sequences (Fig. 2A and B).Concatenated junction fragment sequences gave the same treeresolution as did the downstream junction fragment, and soisolates with identical downstream junction fragment se-quences were assigned to the same prophage type (PT). Thepresent study identified nine prophage types (Fig. 2), one eachfor ECIII and ECV and seven PTs associated with ECII iso-lates from meat and poultry processing plants in the UnitedStates. Of the seven ECII PTs, five PTs were unique to indi-vidual processing plants and two were found in multiple plants(Fig. 2; Table 2). In the latter cases, PT2 was found in threedifferent processing plants and PT6 was found in two differentprocessing plants (Fig. 2A and B). Both upstream and down-stream junction fragments indicated that PT2 included isolatesfrom plant B in Pennsylvania, plant C in New Jersey, and plantI in Indiana. Interestingly, plants B and C both produced thesame type of food, RTE turkey deli meat. The upstream anddownstream junction fragments indicated that PT6 containedisolates from plant G in Pennsylvania and one isolate fromplant O (state information not available; Fig. 2). Similar to thesituation above with PT2 isolates, both plant O (S. Kathariou,unpublished data) and plant G (FSIS, unpublished data) pro-cessed the same type of food; however, in this case it was rawturkey. Single nucleotide polymorphisms (SNPs) within thejunction fragments that differentiated prophage types of L.monocytogenes were identified (Table 3; see also Table S3 inthe supplemental material).

Recombination analysis. Possible recombination events inthe upstream and downstream fragments were tested with theRDP v.3.38 program. Three putative recombination eventswere identified in the downstream junction fragment. Amongthem, only two putative recombination events (prophage typesPT5, PT6, and PT8) were supported by at least four methods(P � 10�3) (see Fig. S4 in the supplemental material). Fourputative recombination events were detected in the upstreamjunction fragment, and two of these recombination events inPT4 and PT6 were supported by five methods implemented inthe RDP program with high statistical significance (P � 10�5).

Detection of spontaneous induction of the comK prophage.Positive amplifications with attB primer pairs were seen in all12 isolates tested (see Fig. S4 in the supplemental material;also data not shown). Positive amplifications with attP primerpairs were seen in nine of the isolates tested but not in isolatesOB070181, OB080398, and OB080567 (Fig. S4; also data notshown). Comparison of the attP and attB sequences fromstrains H7858 and N3-031 with the corresponding upstreamand downstream junction fragment sequences from these samestrains confirmed that the fragments amplified were the attPsite within phage A118 and the attB attachment site withincomK.

Detection of int in L. monocytogenes genomic DNA and comKphage DNA by PCR. Amplification of int in genomic DNA wasobserved in L. monocytogenes strains H7858 (ECII), N3-031(ECIII), and 1001::A118 but not in F2365 (ECI) (Table 4).Amplification of int in comK phage DNA was observed in L.monocytogenes strains N3-031 and 1001::A118 but not in F2365or H7858 (Table 4). Amplicons were sequenced to confirmthat int was amplified.

Plaque assays. Phage plaques were observed on plates con-taining filtrate from L. monocytogenes strain 1001::A118 but

TABLE 1. PCR amplification results for ECII PCR, outbreak clonePCR, and junction fragment PCR of L. monocytogenes

isolates from different processing plants

Processingplant

ECIIPCR

ECII outbreak clone PCR Junction fragment PCR

2426_1998primer pair

2426_2002primer pair Upstream Downstream

A � � � � �B � � � � �C � � � � �D � � � � �E � � � � �F � � � � �G � � � � �H � � � � �I � � � � �J � � � � �K � � � � �L � � � � �M � � � � �N � � � � �O � � � � �

VOL. 77, 2011 comK PROPHAGE JUNCTION FRAGMENTS 3283

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

were not observed with ECI strain F2365, ECII strain H7858,or ECIII strain N3-031 (Table 4).

Epifluorescence microscopy. The effects of strain, type ofRTE food-conditioning film, and their interaction on final celldensity were all highly significant (P � 0.001). Tukey’s multiplecomparison revealed that strains lacking the comK prophage(ECI and lineage III) showed significantly lower cell densitiesthan did 4b and 1/2a strains containing the comK propahge(P � 0.05). The two serotype 1/2a strains containing the comKprophage (ECIII and ECV) produced the highest cell densities(P � 0.05) and frequently formed biofilms on meat- and poul-try-conditioning films (Fig. 3A and B and data not shown). TheECV strain produced very high cell densities and mature bio-films when growing on conditioning films made from all fourRTE muscle foods (hot dogs, turkey, ham, and chicken). Com-pared to other strains, the ECV strain formed unique biofilms,which contained high numbers of cells embedded in web-likeextracellular polymeric substances (EPS) with empty spaceswithin the web-like network (Fig. 3A and B). In contrast, ECVshowed very low cell density on conditioning films made fromsoft cheese. Among all RTE foods tested, chicken breast pro-duced the highest average cell density across all strains (P �0.05), followed by ham, turkey, hot dog, and soft cheese (Table5). Food-conditioning films were always visible when cell den-sities were low or moderate; however, food-conditioning films

were never observed at very high cell densities where biofilmswere often seen (Fig. 3B). No biofilms were observed in theabsence of food-conditioning films (Fig. 3A), even though cellshad attached to the glass slides and nutrients were available inTSBYE.

Sequence analysis of comK prophage of 4b and 1/2a strains.Comparative genome analysis of L. monocytogenes revealed thatthe comK prophage from serotypes 1/2a and 4b have four openreading frames (ORFs) (LMOh7858_2410, LMOh7858_2411,LMOh7858_2421, and LMOh7858_2426) that show 100% se-quence identity within the 1988 sporadic and 2000 outbreakstrains of ECIII (77). comK prophage sequence comparison ofserotype 1/2a epidemic clones (ECs) (ECIII and ECV) with ge-nome-sequenced serotype 4b strain (ECII) revealed two ORFsunique to 1/2a ECs that also have 100% sequence identity withinECIII isolates (Fig. 4). Comparison of prophage protein se-quences of ECII, ECIII, and ECV against the sequences availablein NCBI GenBank revealed that while the majority of prophagegenes in ECII (serotype 4b) and ECV (serotype1/2a) are fromListeria phage A006, most prophage genes in ECIII (serotype1/2a) are from Listeria phage A118 (Fig. 5). In addition, comKprophage genes with homology to genes from Streptococcus andEnterococcus phages were also found in the comK prophage (datanot shown).

FIG. 2. Cluster diagrams based on upstream and downstream junction fragment sequences in L. monocytogenes isolates described in Table S1in the supplemental material; the plant information and prophage type are labeled in the figure. (A) Upstream junction fragment cluster diagram.(B) Downstream junction fragment cluster diagram. Nodes are labeled with bootstrap values. �, plant information is not available.

3284 VERGHESE ET AL. APPL. ENVIRON. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

DISCUSSION

The backbone genome of L. monocytogenes is highly clonal(22, 72, 81, 82), which is consistent with the fact that L. mono-cytogenes has the highest degree of purifying selection observedin bacteria (75). Within individual epidemic clones, the back-bone genomes are even more conserved (21, 25, 65). For ex-ample, Orsi et al. (78) used whole-genome sequence compar-ison to analyze two 1988 sporadic case isolates and two 2000outbreak isolates of ECIII that were associated with the sameprocessing plant over a 12-year period and were able to findonly 11 total SNPs in the backbone genome between thesestrains. In contrast, the �40-kb comK prophage had 1,274polymorphic sites that differentiated the 1988 isolates from the2000 isolates (78). This very large number of polymorphisms inthe comK prophage within apparently so short a period of time(12 years or less) was very likely due to extensive recombina-tion (78).

The strong congruence between the upstream and down-stream junction fragment sequences and their correlation withspecific processing plants in the present study strongly sug-gested that these prophage types were not evolving randomly.

Most meat and poultry processing plants in North Americahave been in existence for less than 100 years. Therefore, if thedifferent persistent prophage types identified in the presentstudy (Fig. 2) were due to rapid niche-specific adaptation inindividual plants, only the comK prophage and not the back-bone genome of L. monocytogenes would appear to provideenough sequence variation to account for this rapid adapta-tion. This is consistent with the concept of rapid evolution ofmicrobial ecotypes (55) being driven by niche-specific selection(31). While it is difficult to predict the genes responsible forthis ecological divergence, a DNA sequence-based approachwhich targets ecologically relevant markers could overcomethese challenges. In the present study, ecologically distinctprophage types could be identified by sequencing the comKprophage junction fragments (Fig. 2). The fact that some pro-phage types were found in isolates from multiple processingplants may be due to a common food source in these plants. Toachieve efficiency, many food processing plants run the sametype of RTE food over the same packaging lines 16 h per day,5 to 6 days per week, for months or years. This constant andprolonged exposure to the same or very closely related types ofRTE foods may result in strains/ecotypes of L. monocytogenesthat have evolved an increased ability to attach to, grow on,and form biofilms on these RTE food-conditioning films. Suchstrains would persist over other strains that are less fit (55).Initial bacterial cell attachment is critical for the formation ofbacterial biofilms (14, 79). In food processing plants, foodresidues accumulate in enclosed areas that are difficult to cleanand sanitize, e.g., slicers, etc., and thus represent harboragesites/reservoirs for L. monocytogenes. Because they containfood and water for extended periods of time, these harboragesites allow bacteria to become established and multiply (2, 23,99). In summary, a common food in a food processing plantwould likely produce a common food-conditioning film in har-borage sites throughout the plant (10), which might select forattachment, growth, and biofilm formation by isolates with acommon prophage type.

In order to test whether type of strain or type of food-conditioning film has an effect on cell attachment or biofilmformation, we utilized fluorescent microscopy to observe theeffect of different food-conditioning films on growth and bio-film formation by different prophage types of L. monocytogenes

TABLE 2. Virulence types, allelic profiles of upstream anddownstream junction fragments, and prophage types

(PTs) of isolates analyzed in this study

Processingplant

Identificationno.

MVLSTvirulence

type

Junction fragmentallelic profile Prophage

typeUpstream Downstream

A H7858 19 1 1 1A H7557 19 1 1 1B J1816 19 2 2 2B J1815 19 2 2 2C J1925 19 2 2 2

J1703 19 2 2 2D N3-031 1 3 3 3D J1-101 1 3 3 3D R2-603 1 3 3 3D R2-499 1 3 3 3E 08-5578 59 4 4 4E 08-5923 59 4 4 4F OB040119 19 5 5 5F OB050272 19 5 5 5F OB050273 19 5 5 5G OB050226 19 6 6 6G OB050347 19 6 6 6G OB050350 19 6 6 6G OB050351 19 6 6 6G OB050355 19 6 6 6G OB070122 19 6 6 6H OB020621 19 5 7 7H OB020790 19 5 7 7I OB030029 19 2 2 2J OB070181 19 7 8 8N 1493 19 7 9 9N 1495 19 7 9 9N 1496 19 7 9 9N 1498 19 7 9 9N 1503 19 7 9 9N 1506 19 7 9 9N 1513 19 7 9 9N 1514 19 7 9 9N 1516 19 7 9 9N 1117 19 6 6 6

TABLE 3. Summary of SNPs within different regions of theupstream and downstream junction fragmentsa

Region within junction fragmentNo. of SNPs

Total Nonsynonymous

Upstream junction fragmentcomK (RBS) 1 NAcomK 2 NANoncoding (prophage) 52 NAHP1 (prophage) 55 22HP2 (prophage) 14 7

Downstream junction fragmentInt (prophage) 65 13Noncoding (prophage) 9 NAcomK 3 NA

a NA, not applicable because these are noncoding regions/pseudoregions.RBS, ribosomal binding site.

VOL. 77, 2011 comK PROPHAGE JUNCTION FRAGMENTS 3285

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

and also strains of L. monocytogenes that lacked the comKprophage. The results demonstrated the critical role of food-conditioning films on attachment, growth, and biofilm forma-tion by these different strains of L. monocytogenes. All ECstrains, except ECII strain J1703, grew well on the respectivefoods from which they were originally isolated (see Table S1 inthe supplemental material). However, it is not quite clear whythe ECII strain J1703, which was originally isolated during the2002 turkey deli outbreak, failed to grow on turkey-condition-ing film in the present study. Research is under way in our lab

(the Knabel lab) to test whether this ECII strain is capable offorming biofilms on other brands of RTE turkey products. Alineage I ECI strain and a lineage III strain, both of whichlacked the comK prophage, showed significantly less growth onmeat- and poultry-conditioning films in the present study thandid ECII, ECIII, and ECV strains that contained the comKprophage (Fig. 3A; Table 5). This is consistent with a previousstudy (67) in which strain F2365 (ECI) was shown to be lesscapable of forming biofilms than were other 4b strains and also1/2a strains. The different ECII isolates also showed variability

TABLE 4. Detection of spontaneous induction of the comK prophage in L. monocytogenes strains F2365, H7858, and N3-031 using attP andattB PCR, comK phage int PCR, and plaque formation

Strain testeda Serotype

Detection ofspontaneous

excision

Amplification ofint in: Plaque

formationb Interpretation

attP attB Lysogen Phage

F2365 (ECI) 4b � � � � � comK prophage absent, no comK phage formedH7858 (ECII) 4b � � � � � comK prophage spontaneously induced, no or

undetectable levels of defective comK phageN3-031 (ECIII) 1/2a � � � � � comK prophage spontaneously induced,

defective comK phage1001::A118 1/2a � � � � �c comK prophage spontaneously induced, active

comK phage

a Epidemic clone designation is indicated within parentheses after strain tested.b Indicator strains tested were ATCC 23074 and 1001 for serotypes 4b and 1/2a, respectively.c Approximately 3.00 log10 PFU/ml of filtrate as described in Materials and Methods.

FIG. 3. (A) Epifluorescence photomicrographs showing different cell densities of L. monocytogenes on food-conditioning films. The number inthe upper right corner of each picture indicates the cell density score, with 0 indicating absence of cells (turkey, no cells added) and 1 indicatingvery small (glass only, no food-conditioning film), 2 indicating small (hot dog with strain J1703), 3 indicating moderate (hot dog with strainOB020790), 4 indicating large (chicken with strain J1703), and 5 indicating very large (turkey with strain 08-5923) amounts of cells observed onthe slides. Bars, 20 �m. (B) Epifluorescence photomicrographs showing the degradation of food-conditioning films and biofilm formation by theECV strain (08-5923). Red arrows indicate undegraded food-conditioning films, and yellow arrows indicate biofilm formation. Bars, 40 �m.

3286 VERGHESE ET AL. APPL. ENVIRON. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

in their attachment and growth on food-conditioning films,which might be due to their different comK prophage types.Similarly, Wang et al. (103) demonstrated that cryptic (defec-tive) prophages in E. coli K-12 increase growth and biofilmformation. However, it cannot be excluded that other (cur-rently unidentified) genetic differences may play a role in at-tachment, growth, and biofilm formation by both L. monocy-togenes and E. coli. Interestingly, while ECII, ECIII, andespecially ECV showed extensive growth and sometimes bio-film formation on muscle foods, they showed very little if anygrowth on soft cheese (Table 5). This is consistent with previ-ous reports that RTE meat residues increased attachment andbiofilm formation by L. monocytogenes (92), while milk pro-teins significantly reduced its attachment to various surfaces(106).

Many known bacterial cell surface-associated fitness deter-minants are located on genomic islands of prophage origin (42,61). Interestingly, sequence analysis of the comK prophagerevealed four open reading frames (ORFs) (Fig. 4) located inthe “black block” regions (regions of 100% sequence identityin the comK prophage) of the work of Orsi et al. (78). TheseORFs are present in all genome-sequenced 1/2a and 4b strainsof L. monocytogenes that contain the comK prophage. In ad-dition, 2 other ORFs, which are unique to genome-sequenced1/2a strains and not present in the phage A118 genome orECII, can also be found in these regions (Fig. 4). At present, itis unclear whether these genes or some other prophage genesare responsible for niche-specific adaption. We hypothesizethat comK prophage genes might have been modified to helpL. monocytogenes adapt to different food processing environ-ments, and so we find it appropriate to tentatively term thesegenes “adaptons.” This is very similar to E. coli probiotic strainNissle, where such determinants (similar to adaptons) are lo-cated on genomic islands that likely represent defective pro-phages and include adhesins thought to contribute to its sur-vival and successful colonization of the human gut (42).Putative adaptons in the comK prophage may encode similar

adhesins that mediate specific attachment to different food-conditioning films. While such adaptons would probably notfully explain attachment and biofilm formation by L. monocy-togenes, as many adhesins are chromosomally encoded (7, 51),they may help to explain the persistence of specific L. mono-cytogenes prophage types in individual food processing plantsthat manufacture the same food product over an extendedperiod of time (5, 70, 86) (Fig. 2B).

Given the evidence for extensive recombination within pro-phages in the present and other studies (3, 20, 78), we specu-late that transduction of the entire comK prophage via defec-tive comK phage particles might be mediating this rapid andextensive recombination. Phage-like elements that functionsolely for mediating gene exchange have been observed previ-ously in the case of gene transfer agents (GTAs) in alphapro-teobacteria (59) and Staphylococcus aureus pathogenicity is-land (SaPI) particles (87, 96). In the present study, the comKprophages in L. monocytogenes ECII and ECIII were shown tobe spontaneously inducible, as positive amplifications of attPand attB were seen in all lysogenized isolates analyzed (see Fig.S5 in the supplemental material) (data not shown). The resultsin Table 4 demonstrate that the comK phage was spontane-ously induced from ECIII strains N3-031 and 1001::A118 andthat the comK phage genome was packaged within phage par-ticles by these strains. These findings are consistent with thoseof Loessner et al. (63), who also revealed spontaneous induc-tion of the comK prophage and subsequent phage formation.However, the prophages in L. monocytogenes are fairly stable,as only a small fraction of the population is spontaneouslyinduced (63). Spontaneous induction and formation of defec-tive transducing phages have been seen in various other mi-croorganisms and are a mechanism by which various virulencegenes (3) and genomic islands (49, 52), including pathogenicityislands (87), are mobilized into other strains. A recent studywith E. coli K-12 showed that a spontaneously induced defec-tive prophage, “e14,” is critical for biofilm formation (103).

Plaque-forming bacteriophages of L. monocytogenes have

TABLE 5. Effects of type of strain, type of food-conditioning film, and their interaction on the cell density of L. monocytogenes on glassCultureSlides after incubation at 30°C for 48 h in TSBYEa

Strain PTb

Cell density of strain on FCFStrain means for

all FCFseNo-FCFcontrol Soft cheese Hot dog Turkey Ham Chicken

No-L. monocytogenes control 0 0 0 0 0 0 0 ALineage III (W1-111) NAc 1.0 1.0 1.0 1.5 2.5 1.0 1.4 BECI (F2365) NA 1.0 3.0 1.5 1.5 2.0 2.0 2.0 BCECII (J1703) 2 1.0 1.5 2.5 1.0 3.0 4.0 2.4 BCECII (H7858) 1 1.0 1.0 2.5 4.0 3.5 3.5 2.9 CDECII (OB020790) 7 1.0 3.0 3.0 3.0 2.5 4.0 3.1 CDEECIII (N3-031) 3 1.0 3.5 3.5 4.5 4.0 3.5 3.8 DEECV (08-5923) 4 1.0 1.0 5.0 5.0 4.5 5.0 4.1 EFCF means for all strainsd 1.0 a 2.0 ab 2.7 bc 2.9 bc 3.1 bc 3.3 c

a Data in the table are based on two replications of the experiment. The numbers in the table indicate cell densities of different strains present on differentfood-conditioning films (FCFs), with 0 indicating absence of cells and 1 indicating very small, 2 indicating small, 3 indicating moderate, 4 indicating large, and 5indicating very large amounts of cells observed on the slides. Fractional numbers are calculated based on the results of two replications.

b PT, prophage type.c NA, not applicable. The lineage III strain (W1-111) and ECI strain (F2365) do not contain the comK prophage.d Means of one FCF for all strains are calculated by averaging all the values in the column (except that for the no-L. monocytogenes control) corresponding to that

FCF; means in this row that do not share a lowercase letter are significantly different (P � 0.05).e Means of one strain for all FCFs are calculated by averaging all the values in the row (except that for the no-FCF control) corresponding to that strain; means in

the column that do not share an uppercase letter are significantly different (P � 0.05).

VOL. 77, 2011 comK PROPHAGE JUNCTION FRAGMENTS 3287

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

been reported to be serotype specific due to the different phagereceptors on the surface of L. monocytogenes strains of differ-ent serotypes (30, 100). However, finding genes from variousserotype-specific phages within defective comK prophages indifferent lineages of L. monocytogenes (Fig. 5) indicates thatdefective comK prophages recombine between phages/defec-tive phages that infect different lineages and serotypes. Inaddition, standard phage susceptibility tests are based onplaque formation; however, Chen and Novick (17) in theirstudies reported that several staphylococcal phages transducedL. monocytogenes but could not form plaques. Spontaneousinduction of defective comK phage from a donor cell of L.monocytogenes and subsequent transduction of the entire de-fective phage genome into a recipient cell that already containseither an infective or a defective comK prophage would likelyresult in frequent and large-scale homologous recombination.This phenomenon is known to produce rapid reshuffling ofgene modules and/or gene sequences within bacterial popula-tions (3, 13, 78) and is consistent with the hypothetical phage-mediated “life cycle” of genomic islands proposed by Dobrindtet al. (28) and the model proposed in this study (Fig. 6). Sucha mechanism would be fundamentally different from general-

ized or specialized transduction, where only a small fraction(�10�4) of the induced phage are actually transducing phages(62). The proposed mechanism would also be different thanthe lytic or lysogenic cycles of phage replication, because in-duction and recombination of defective comK phage would beoccurring for the purpose of rapid evolution of L. monocyto-genes in the absence of phage propagation and, if so, wouldlikely be under the control of the bacterial host (13, 50).

The proposed model for high-frequency recombination driv-ing rapid niche-specific adaptation and persistence of L. mono-cytogenes in food processing plants (Fig. 6) is consistent withwhat is currently known about defective prophages and howthey have been used by various bacteria to rapidly adapt tonatural environments (45, 69). If this is also true for L. mono-cytogenes, we believe that defective comK prophages would bemore accurately described as rapid adaptation islands (RAI)(Fig. 6). This RAI concept is consistent with the recent formu-lation of a novel concept of interphage interactions in defectiveprophage communities in E. coli O157:H7 (3). The finding thatno comK phage DNA of ECII (H7858) was detected after PCR(Table 4) and the predominance of 1/2 phage genes in thecomK prophage of a serotype 4 strain (Fig. 5) suggest that

FIG. 4. Sequences of putative adaptons within comK prophages in epidemic clone strains of L. monocytogenes show 100% sequence identitywithin a processing plant over a 12-year period (plant D) but vary between processing plants (plants A and E). The putative adaptons ORFs HP1,HP2, gp15, gp13, and partial int (locus designations LMOh7858_2410, -2411, -2421, -2426, and -2475, respectively) are present in sequencedgenomes of 1/2a and 4b serotypes of L. monocytogenes that contain the comK prophage and show 100% sequence identity within 1988 and 2000ECIII isolates. The putative adaptons ORFs gp27 (LMOf6854_2338) and HP3 (LMOf6854_2375) (arrows shaded gray) indicate those comKprophage genes that are unique to serotype 1/2a strains but are not present in the ECII serotype 4b strain. comK, N-terminal comK fragment; HP1,hypothetical protein 1; HP2, hypothetical protein 2; gp27, phage gp27 protein; gp15, phage gp15 protein; gp13, major tail protein; HP3, hypotheticalprotein 3; int, integrase; comK, C-terminal comK fragment. Arrows point to the corresponding positions of these loci in the comK prophage inECII, ECIII, and ECV strains. Blocks (adaptons) with different shading indicate nonidentical sequences, and black blocks indicate 100% sequenceidentity within ECIII isolates.

3288 VERGHESE ET AL. APPL. ENVIRON. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

serotype 4b strains may act more like recipients than likedonors (Fig. 6). Serotype 1/2a strains in lineage II are known topredominate and persist in food processing plants (77), possi-bly due to their enhanced ability to attach, grow, and form

biofilms (9) (Table 5; Fig. 3B). Therefore, they are more likelyto serve as donors of comK genes to other strains of L. mono-cytogenes in lineages I and II, thus allowing them to also adaptto these environments.

FIG. 5. Comparison of comK prophage genes in 1/2a and 4b serotypes of Listeria monocytogenes with sequenced bacteriophage genomes.Regions that correspond to different bacteriophages are shaded.

FIG. 6. Proposed model for rapid niche-specific adaptation and persistence of L. monocytogenes. The cycle starts at the top with spontaneousinduction of a rapid adaptation island (RAI) in the donor cell, followed by RAI phage formation and transduction of donor RAI into a recipientcell, which also contains a defective RAI or infective lysogenized phage integrated into its chromosome. Recombination between donor RAI andrecipient RAI/comK prophage generates numerous RAI recombinants. Natural selection then acts on RAI recombinants to yield unique persistentprophage types that are adapted to individual processing plants or multiple plants manufacturing the same type of food product and thus producethe same type of food-conditioning film.

VOL. 77, 2011 comK PROPHAGE JUNCTION FRAGMENTS 3289

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

The results of the present study and the model presentedhelp explain many of the previously well-known, but not well-understood, observations/phenomena associated with persis-tent strains of L. monocytogenes found in food processingplants, such as that (i) persistent strains in individual plantshave unique genotypes/subtypes; (ii) persistent and prevalentstrains are found in harborage sites containing RTE food res-idues; (iii) serotype 1/2a strains predominate over 4b strains inprocessing plants and foods; (iv) serotype 4b strains of ECIIcontaining comK prophage are also known to be persistent infood processing plants; (v) ECI and lineage III strains of L.monocytogenes, which are of animal origin (6, 36, 60, 89) andlack the comK prophage (26; X. Deng, personal communica-tion), are typically isolated from raw processing environments(11), raw foods, or RTE foods that have been cross contami-nated (21, 24, 39) but not from postpasteurization environ-ments or most pasteurized RTE foods (36); and (vi) RTE meatand poultry products are more frequently contaminated withL. monocytogenes than are most other RTE foods and thus arein the highest risk category for listeriosis. While the results ofthe present study support the model presented, further re-search is obviously needed to test the hypothesis that the comKprophage is playing a key role in attachment, growth, andbiofilm formation on specific food-conditioning films. Specifi-cally, the comK prophage should be cured (deleted) from thoseECII, ECIII, and ECV strains that contain it and these iso-genic strains should be compared to wild-type (WT) strains ina food-conditioning film model system to test this hypothesis.These studies are under way in our (Knabel’s) laboratory.

The present study also confirms that biofilm formation ismore likely to occur in difficult-to-clean harborage sites thatcontain food and water over extended periods of time (23). Forthis reason, it is critical to understand and eliminate harboragesites in order to prevent L. monocytogenes contamination ofRTE foods. For example, the 2008 Canadian outbreak dem-onstrated that contaminated harborage sites deep within slic-ers likely provided long-standing reservoirs for L. monocyto-genes, which was thought to have contributed significantly tothis outbreak (2). Therefore, preventing L. monocytogenes con-tamination of RTE food products relies on detecting and elim-inating harborage sites by first purchasing equipment thatmeets sanitary equipment design standards and then imple-menting a regular deep-cleaning and sanitizing program toremove all food residues to prevent biofilm formation anddestroy all remaining L. monocytogenes bacteria.

Further research is needed to test the model proposed in thepresent study, especially RAI transduction and recombination,and the possible role(s) of putative adaptons in generatingpersistent prophage types of L. monocytogenes. If the model isconfirmed to be accurate, it would lead to a more fundamentalunderstanding of how L. monocytogenes rapidly adapts to andpersists in individual food processing plants. It could also helpidentify the routes by which specific prophage types of L.monocytogenes are transmitted between and within processingplants and retail operations. This information would help foodcompanies implement more effective intervention strategies toprevent persistent and dangerous strains of L. monocytogenesfrom colonizing food processing environments and thus pre-vent their transmission to RTE foods.

ACKNOWLEDGMENTS

We thank Peter Evans at FSIS, USDA, for identifying and providingPFGE profiles of putative ECII isolates in the FSIS culture collection;Martin Wiedmann at Cornell University for providing ECI and ECIIIstrains; Bala Swaminathan at CDC for providing ECII strains; andMartin Loessner at ETH, Zurich, Switzerland, for providing strains1001 and 1001::A118.

S. Knabel was supported by a United States Department of Agri-culture Special Grant on Milk Safety to the Pennsylvania State Uni-versity, and S. Kathariou was supported by a grant from the AmericanMeat Institute Foundation.

REFERENCES

1. Altschul, S. F., et al. 1997. Gapped BLAST and PSI-BLAST: a new gen-eration of protein database search programs. Nucleic Acids Res. 25:3389–3402.

2. Anonymous. 17 April 2009, posting date. Lessons learned: the CanadianFood Inspection Agency’s review of Est. 97B (Maple Leaf Consumer FoodsInc.). Canadian Food Inspection Agency, Ottawa, Ontario, Canada.

3. Asadulghani, M., et al. 2009. The defective prophage pool of Escherichiacoli O157: prophage-prophage interactions potentiate horizontal transfer ofvirulence determinants. PLoS Pathog. 5:e1000408.

4. Autio, T., R. Keto-Timonen, J. Lunden, J. Bjorkroth, and H. Korkeala.2003. Characterisation of persistent and sporadic Listeria monocytogenesstrains by pulsed-field gel electrophoresis (PFGE) and amplified fragmentlength polymorphism (AFLP). Syst. Appl. Microbiol. 26:539–545.

5. Autio, T., et al. 2002. Similar Listeria monocytogenes pulsotypes detected inseveral foods originating from different sources. Int. J. Food Microbiol.77:83–90.

6. Berrang, M. E., R. J. Meinersmann, J. F. Frank, D. P. Smith, and L. L.Genzlinger. 2005. Distribution of Listeria monocytogenes subtypes within apoultry further processing plant. J. Food Prot. 68:980–985.

7. Bierne, H., and P. Cossart. 2007. Listeria monocytogenes surface proteins:from genome predictions to function. Microbiol. Mol. Biol. Rev. 71:377–397.

8. Blatter, S., N. Giezendanner, R. Stephan, and C. Zweifel. 2010. Phenotypicand molecular typing of Listeria monocytogenes isolated from the processingenvironment and products of a sandwich-producing plant. Food Control21:1519–1523.

9. Borucki, M. K., J. D. Peppin, D. White, F. Loge, and D. R. Call. 2003.Variation in biofilm formation among strains of Listeria monocytogenes.Appl. Environ. Microbiol. 69:7336–7342.

10. Boulane-Petermann, L. 1996. Processes of bioadhesion on stainless steelsurfaces and cleanability: a review with special reference to the food indus-try. Biofouling 10:275–300.

11. Bowden, G. H., and Y. H. Li. 1997. Nutritional influences on biofilm devel-opment. Adv. Dent. Res. 11:81–99.

12. Brooks, J. D., and S. H. Flint. 2008. Biofilms in the food industry: problemsand potential solutions. Int. J. Food Sci. Technol. 43:2163–2176.

13. Brussow, H., C. Canchaya, and W. D. Hardt. 2004. Phages and the evolu-tion of bacterial pathogens: from genomic rearrangements to lysogenicconversion. Microbiol. Mol. Biol. Rev. 68:560–602.

14. Busscher, H. J., R. Bos, and H. C. van der Mei. 1995. Initial microbialadhesion is a determinant for the strength of biofilm formation. FEMSMicrobiol. Lett. 128:229–234.

15. Canchaya, C., C. Proux, G. Fournous, A. Bruttin, and H. Brussow. 2003.Prophage genomics. Microbiol. Mol. Biol. Rev. 67:238–276.

16. Chatterjee, S. S., et al. 2006. Intracellular gene expression profile of Listeriamonocytogenes. Infect. Immun. 74:1323–1338.

17. Chen, J., and R. P. Novick. 2009. Phage-mediated intergeneric transfer oftoxin genes. Science 323:139–141.

18. Chen, M.-Y. 2010. Towards real-time observation of conditioning film andearly biofilm formation under laminar flow conditions using a quartz crystalmicrobalance. Biochem. Eng. J. 53:121–130.

19. Chen, Y., and S. J. Knabel. 2007. Multiplex PCR for simultaneous detectionof bacteria of the genus Listeria, Listeria monocytogenes, and major sero-types and epidemic clones of L. monocytogenes. Appl. Environ. Microbiol.73:6299–6304.

20. Chen, Y., and S. J. Knabel. 2008. Prophages in Listeria monocytogenescontain single-nucleotide polymorphisms that differentiate outbreak cloneswithin epidemic clones. J. Clin. Microbiol. 46:1478–1484.

21. Chen, Y., W. Zhang, and S. J. Knabel. 2007. Multi-virulence-locus sequencetyping identifies single nucleotide polymorphisms which differentiate epi-demic clones and outbreak strains of Listeria monocytogenes. J. Clin. Mi-crobiol. 45:835–846.

22. Cheng, Y., R. M. Siletzky, and S. Kathariou. 2008. Genomic divisions/lineages, epidemic clones, and population structure, p. 337–358. In D. Liu(ed.), Handbook of Listeria monocytogenes. CRC Press, New York, NY.

23. Chmielewski, R. A. N., and J. F. Frank. 2003. Biofilm formation and controlin food processing facilities. Comp. Rev. Food Sci. Saf. 2:22–32.

3290 VERGHESE ET AL. APPL. ENVIRON. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

24. Costerton, J. W., L. Montanaro, and C. R. Arciola. 2005. Biofilm in implantinfections: its production and regulation. Int. J. Artif. Organs 28:1062–1068.

25. den Bakker, H. C., E. D. Fortes, and M. Wiedmann. 2010. Multilocussequence typing of outbreak-associated Listeria monocytogenes isolates toidentify epidemic clones. Foodborne Pathog. Dis. 7:257–265.

26. Deng, X., A. M. Phillippy, Z. Li, S. L. Salzberg, and W. Zhang. 2010.Probing the pan-genome of Listeria monocytogenes: new insights into in-traspecific niche expansion and genomic diversification. BMC Genomics11:500.

27. Diaz, P. I., et al. 2006. Molecular characterization of subject-specific oralmicroflora during initial colonization of enamel. Appl. Environ. Microbiol.72:2837–2848.

28. Dobrindt, U., B. Hochhut, U. Hentschel, and J. Hacker. 2004. Genomicislands in pathogenic and environmental microorganisms. Nat. Rev. Micro-biol. 2:414–424.

29. Donlan, R. M. 2002. Biofilms: microbial life on surfaces. Emerg. Infect. Dis.8:881–890.

30. Dorscht, J., et al. 2009. Comparative genome analysis of Listeria bacterio-phages reveals extensive mosaicism, programmed translational frameshift-ing, and a novel prophage insertion site. J. Bacteriol. 191:7206–7215.

31. Dunn, K. A., J. P. Bielawski, T. J. Ward, C. Urquhart, and H. Gu. 2009.Reconciling ecological and genomic divergence among lineages of Listeriaunder an “extended mosaic genome concept.” Mol. Biol. Evol. 26:2605–2615.

32. Eifert, J. D., et al. 2005. Molecular characterization of Listeria monocyto-genes of the serotype 4b complex (4b, 4d, 4e) from two turkey processingplants. Foodborne Pathog. Dis. 2:192–200.

33. Farber, J. M., and P. I. Peterkin. 1991. Listeria monocytogenes, a food-bornepathogen. Microbiol. Rev. 55:476–511.

34. Filee, J., P. Forterre, and J. Laurent. 2003. The role played by viruses in theevolution of their hosts: a view based on informational protein phylogenies.Res. Microbiol. 154:237–243.

35. Folsom, J. P., G. R. Siragusa, and J. F. Frank. 2006. Formation of biofilmat different nutrient levels by various genotypes of Listeria monocytogenes.J. Food Prot. 69:826–834.

36. Foxman, B., L. Zhang, J. S. Koopman, S. D. Manning, and C. F. Marrs.2005. Choosing an appropriate bacterial typing technique for epidemiologicstudies. Epidemiol. Perspect. Innov. 2:10.

37. Fretz, R., et al. 2010. Update: multinational listeriosis outbreak due to‘Quargel’, a sour milk curd cheese, caused by two different L. monocyto-genes serotype 1/2a strains, 2009-2010. Euro Surveill. 15(16):pii:19543.

38. Gilmour, M. W., et al. 2010. High-throughput genome sequencing of twoListeria monocytogenes clinical isolates during a large foodborne outbreak.BMC Genomics 11:120.

39. Goller, C. C., and T. Romeo. 2008. Environmental influences on biofilmdevelopment. Curr. Top. Microbiol. Immunol. 322:37–66.

40. Graves, L. M., et al. 2005. Microbiological aspects of the investigation thattraced the 1998 outbreak of listeriosis in the United States to contaminatedhot dogs and establishment of molecular subtyping-based surveillance forListeria monocytogenes in the PulseNet network. J. Clin. Microbiol. 43:2350–2355.

41. Gray, M. J., N. E. Freitag, and K. J. Boor. 2006. How the bacterial pathogenListeria monocytogenes mediates the switch from environmental Dr. Jekyllto pathogenic Mr. Hyde. Infect. Immun. 74:2505–2512.

42. Grozdanov, L., et al. 2004. Analysis of the genome structure of the non-pathogenic probiotic Escherichia coli strain Nissle 1917. J. Bacteriol. 186:5432–5441.

43. Gueriri, I., et al. 2008. The DegU orphan response regulator of Listeriamonocytogenes autorepresses its own synthesis and is required for bacterialmotility, virulence and biofilm formation. Microbiology 154:2251–2264.

44. Habimana, O., M. Meyrand, T. Meylheuc, S. Kulakauskas, and R. Brian-det. 2009. Genetic features of resident biofilms determine attachment ofListeria monocytogenes. Appl. Environ. Microbiol. 75:7814–7821.

45. Hacker, J., and E. Carniel. 2001. Ecological fitness, genomic islands andbacterial pathogenicity. A Darwinian view of the evolution of microbes.EMBO Rep. 2:376–381.

46. Hagens, S., and M. J. Loessner. 2007. Bacteriophages of Listeria, p. 265–279. In H. Goldfine and H. Shen (ed.), Listeria monocytogenes: pathogenesisand host response. Springer, New York, NY.

47. Hamon, M., H. Bierne, and P. Cossart. 2006. Listeria monocytogenes: amultifaceted model. Nat. Rev. Microbiol. 4:423–434.

48. Harvey, J., and A. Gilmour. 2001. Characterization of recurrent and spo-radic Listeria monocytogenes isolates from raw milk and nondairy foods bypulsed-field gel electrophoresis, monocin typing, plasmid profiling, andcadmium and antibiotic resistance determination. Appl. Environ. Micro-biol. 67:840–847.

49. Hatfull, G. F. 2008. Bacteriophage genomics. Curr. Opin. Microbiol. 11:447–453.

50. Hendrix, R. W. 2003. Bacteriophage genomics. Curr. Opin. Microbiol.6:506–511.

51. Jordan, S. J., et al. 2008. Listeria monocytogenes biofilm-associated protein

(bapL) may contribute to surface attachment of L. monocytogenes but isabsent from many field isolates. Appl. Environ. Microbiol. 74:5451–5456.

52. Juhas, M., et al. 2009. Genomic islands: tools of bacterial horizontal genetransfer and evolution. FEMS Microbiol. Rev. 33:376–393.

53. Kabuki, D. Y., A. Y. Kuaye, M. Wiedmann, and K. J. Boor. 2004. Molecularsubtyping and tracking of Listeria monocytogenes in Latin-style fresh-cheeseprocessing plants. J. Dairy Sci. 87:2803–2812.

54. Kalmokoff, M. L., et al. 2001. Adsorption, attachment and biofilm forma-tion among isolates of Listeria monocytogenes using model conditions.J. Appl. Microbiol. 91:725–734.

55. Koeppel, A., et al. 2008. Identifying the fundamental units of bacterialdiversity: a paradigm shift to incorporate ecology into bacterial systematics.Proc. Natl. Acad. Sci. U. S. A. 105:2504–2509.

56. Kolenbrander, P. E., N. S. Jakubovics, N. I. Chalmers, and R. J. Palmer, Jr.2007. Human oral multi-species biofilms: bacterial communities in healthand disease, p. 175–193. In S. Kjelleberg and M. Givskov (ed.), The biofilmmode of life: mechanisms and adaptation. Horizon Bioscience, Wymond-ham, Norfolk, United Kingdom.

57. Kumar, S., et al. 2009. A study on the effects of some laboratory-derivedgenetic mutations on biofilm formation by Listeria monocytogenes. World J.Microbiol. Biotechnol. 25:527–531.

58. Kushwaha, K., and P. M. Muriana. 2009. Adherence characteristics ofListeria strains isolated from three ready-to-eat meat processing plants. J.Food Prot. 72:2125–2131.

59. Lang, A. S., and J. T. Beatty. 2007. Importance of widespread gene transferagent genes in alpha-proteobacteria. Trends Microbiol. 15:54–62.

60. Linnan, M. J., et al. 1988. Epidemic listeriosis associated with Mexican-stylecheese. N. Engl. J. Med. 319:823–828.

61. Loeffler, J. M., and V. A. Fischetti. 2006. Lysogeny of Streptococcus pneu-moniae with MM1 phage: improved adherence and other phenotypicchanges. Infect. Immun. 74:4486–4495.

62. Loessner, M. J., and R. Calendar. 2006. The Listeria bacteriophages, p.593–601. In R. Calendar (ed.), The bacteriophages. Oxford University PressInc., New York, NY.

63. Loessner, M. J., S. Goeppl, and M. Busse. 1991. Comparative inducibility ofbacteriophage in naturally lysogenic and lysogenized strains of Listeria spp.by u.v. light and mitomycin C. Lett. Appl. Microbiol. 12:196–199.

64. Loessner, M. J., R. B. Inman, P. Lauer, and R. Calendar. 2000. Completenucleotide sequence, molecular analysis and genome structure of bacterio-phage A118 of Listeria monocytogenes: implications for phage evolution.Mol. Microbiol. 35:324–340.

65. Lomonaco, S., Y. Chen, and S. J. Knabel. 2008. Analysis of additionalvirulence genes and virulence gene regions in Listeria monocytogenes con-firms the epidemiologic relevance of multi-virulence-locus sequence typing.J. Food Prot. 71:2559–2566.

66. Lunden, J. M., M. K. Miettinen, T. J. Autio, and H. J. Korkeala. 2000.Persistent Listeria monocytogenes strains show enhanced adherence to foodcontact surface after short contact times. J. Food Prot. 63:1204–1207.

67. Marsh, E. J., H. Luo, and H. Wang. 2003. A three-tiered approach todifferentiate Listeria monocytogenes biofilm-forming abilities. FEMS Micro-biol. Lett. 228:203–210.

68. Martin, D., and E. Rybicki. 2000. RDP: detection of recombinationamongst aligned sequences. Bioinformatics 16:562–563.

69. McDaniel, L. D., et al. 2010. High frequency of horizontal gene transfer inthe oceans. Science 330:50.

70. Møretrø, T., and S. Langsrud. 2004. Listeria monocytogenes: biofilm forma-tion and persistence in food-processing environments. Biofilms 1:107–121.

71. Morgan, G. J., and W. B. Pitts. 2008. Evolution without species: the case ofmosaic bacteriophages. Br. J. Philos. Sci. 59:745–765.

72. Nelson, K. E., et al. 2004. Whole genome comparisons of serotype 4b and1/2a strains of the food-borne pathogen Listeria monocytogenes reveal newinsights into the core genome components of this species. Nucleic AcidsRes. 32:2386–2395.

73. Norton, D. M., et al. 2001. Molecular studies on the ecology of Listeriamonocytogenes in the smoked fish processing industry. Appl. Environ. Mi-crobiol. 67:198–205.

74. Norwood, D. E., and A. Gilmour. 1999. Adherence of Listeria monocyto-genes strains to stainless steel coupons. J. Appl. Microbiol. 86:576–582.

75. Novichkov, P. S., Y. I. Wolf, I. Dubchak, and E. V. Koonin. 2009. Trends inprokaryotic evolution revealed by comparison of closely related bacterialand archaeal genomes. J. Bacteriol. 191:65–73.

76. Olsen, S. J., et al. 2005. Multistate outbreak of Listeria monocytogenesinfection linked to delicatessen turkey meat. Clin. Infect. Dis. 40:962–967.

77. Orsi, R. H., H. C. Bakker, and M. Wiedmann. 2011. Listeria monocytogeneslineages: genomics, evolution, ecology, and phenotypic characteristics. Int.J. Med. Microbiol. 301:79–96.

78. Orsi, R. H., et al. 2008. Short-term genome evolution of Listeria monocy-togenes in a non-controlled environment. BMC Genomics 9:539.

79. Palmer, J., S. Flint, and J. Brooks. 2007. Bacterial cell attachment, thebeginning of a biofilm. J. Ind. Microbiol. Biotechnol. 34:577–588.

80. Pan, Y. W., F. Breidt, and S. Kathariou. 2009. Competition of Listeria

VOL. 77, 2011 comK PROPHAGE JUNCTION FRAGMENTS 3291

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.

monocytogenes serotype 1/2a and 4b strains in mixed-culture biofilms. Appl.Environ. Microbiol. 75:5846–5852.

81. Piffaretti, J. C., et al. 1989. Genetic characterization of clones of the bac-terium Listeria monocytogenes causing epidemic disease. Proc. Natl. Acad.Sci. U. S. A. 86:3818–3822.

82. Ragon, M., et al. 2008. A new perspective on Listeria monocytogenes evo-lution. PLoS Pathog. 4:e1000146.

83. Renier, S., M. Hebraud, and M. Desvaux. 18 November 2010. Molecularbiology of surface colonization by Listeria monocytogenes: an additionalfacet of an opportunistic gram-positive foodborne pathogen. Environ. Mi-crobiol. doi:10.1111/j.1462-2920.2010.02378.x.

84. Rieu, A., S. Weidmann, D. Garmyn, P. Piveteau, and J. Guzzo. 2007. Agrsystem of Listeria monocytogenes EGD-e: role in adherence and differentialexpression pattern. Appl. Environ. Microbiol. 73:6125–6133.

85. Rorvik, L. M., B. Aase, T. Alvestad, and D. A. Caugant. 2003. Molecularepidemiological survey of Listeria monocytogenes in broilers and poultryproducts. J. Appl. Microbiol. 94:633–640.

86. Rorvik, L. M., B. Aase, T. Alvestad, and D. A. Caugant. 2000. Molecularepidemiological survey of Listeria monocytogenes in seafoods and seafood-processing plants. Appl. Environ. Microbiol. 66:4779–4784.

87. Ruzin, A., J. Lindsay, and R. P. Novick. 2001. Molecular genetics ofSaPI1—a mobile pathogenicity island in Staphylococcus aureus. Mol. Mi-crobiol. 41:365–377.

88. Sauders, B. D., et al. 2004. Distribution of Listeria monocytogenes molecularsubtypes among human and food isolates from New York State showspersistence of human disease-associated Listeria monocytogenes strains inretail environments. J. Food Prot. 67:1417–1428.

89. Schlech, W. F., et al. 1983. Epidemic listeriosis—evidence for transmissionby food. N. Engl. J. Med. 308:203–206.

90. Senczek, D., R. Stephan, and F. Untermann. 2000. Pulsed-field gel electro-phoresis (PFGE) typing of Listeria strains isolated from a meat processingplant over a 2-year period. Int. J. Food Microbiol. 62:155–159.

91. Shi, X. M., and X. N. Zhu. 2009. Biofilm formation and food safety in foodindustries. Trends Food Sci. Technol. 20:407–413.

92. Somers, E. B., and A. C. L. Wong. 2004. Efficacy of two cleaning andsanitizing combinations on Listeria monocytogenes biofilms formed at lowtemperature on a variety of materials in the presence of ready-to-eat meatresidue. J. Food Prot. 67:2218–2229.

93. Stoodley, P., K. Sauer, D. G. Davies, and J. W. Costerton. 2002. Biofilms ascomplex differentiated communities. Annu. Rev. Microbiol. 56:187–209.

94. Swaminathan, B., and P. Gerner-Smidt. 2007. The epidemiology of humanlisteriosis. Microbes Infect. 9:1236–1243.

95. Takahashi, H., et al. 2009. Biofilm formation ability of Listeria monocyto-genes isolates from raw ready-to-eat seafood. J. Food Prot. 72:1476–1480.

96. Tallent, S. M., T. B. Langston, R. G. Moran, and G. E. Christie. 2007.Transducing particles of Staphylococcus aureus pathogenicity island SaPI1are comprised of helper phage-encoded proteins. J. Bacteriol. 189:7520–7524.

97. Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: MolecularEvolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol.Evol. 24:1596–1599.

98. Tatusova, T. A., and T. L. Madden. 1999. BLAST 2 Sequences, a new toolfor comparing protein and nucleotide sequences. FEMS Microbiol. Lett.174:247–250.

99. Tompkin, R. B. 2002. Control of Listeria monocytogenes in the food-pro-cessing environment. J. Food Prot. 65:709–725.

100. Tran, H. L., F. Fiedler, D. A. Hodgson, and S. Kathariou. 1999. Trans-poson-induced mutations in two loci of Listeria monocytogenes serotype 1/2aresult in phage resistance and lack of N-acetylglucosamine in the teichoicacid of the cell wall. Appl. Environ. Microbiol. 65:4793–4798.

101. Unnerstad, H., et al. 1996. Prolonged contamination of a dairy with Listeriamonocytogenes. Neth. Milk Dairy J. 50:493–499.

102. Van Houdt, R., and C. Michiels. 2010. Biofilm formation and the foodindustry, a focus on the bacterial outer surface. J. Appl. Microbiol. 109:1117–1131.

103. Wang, X., et al. 2010. Cryptic prophages help bacteria cope with adverseenvironments. Nat. Commun. 1:147.

104. Ward, T. J., et al. 2010. Molecular and phenotypic characterization ofListeria monocytogenes from FSIS surveillance of RTE foods and processingfacilities. J. Food Prot. 73:861–869.

105. Williams, S. K. 2011. Molecular ecology of Listeria monocytogenes andother Listeria species in small and very small ready-to-eat meat processingplants. J. Food Prot. 74:63–77.

106. Wong, A. C. 1998. Biofilms in food processing environments. J. Dairy Sci.81:2765–2770.

107. Zameer, F., S. Gopal, G. Krohne, and J. Kreft. 2010. Development of abiofilm model for Listeria monocytogenes EGD-e. World J. Microbiol. Bio-technol. 26:1143–1147.

3292 VERGHESE ET AL. APPL. ENVIRON. MICROBIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/a

em o

n 06

Feb

ruar

y 20

22 b

y 22

3.17

.85.

102.