CDC confirmed identification of L. monocytogenes using the AccuProbe LISTERIA MONOCYTOGENES Culture Identification Test (Gen-Probe, San Diego, CA, USA) and by FDA according to the FDA Bacteriological Analytical Manual (www.fda.gov/Food/ScienceResearch/LaboratoryMethods/BacteriologicalAnalyticalManualBAM/default.htm). Isolates were serotyped by using commercial antisera (Denka Seiken, Tokyo, Japan) and analyzed by PFGE (9) (Table; Figure 1). The Technical Appendix shows the relative distribution of the 4 PFGE profiles among clinical, food, or environmental samples.

Isolates were grown overnight in tryptic soy broth with yeast extract at 37°C, and DNA was extracted by using the Ultra Clean Microbial DNA Isolation Kit (Mo Bio Laboratories, Solana Beach, CA, USA) for isolates from CDC and USDA and the Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA) for isolates from FDA. Sequence types (STs) identified by using MLST were assigned as described (6) on the basis of whole genome sequence data (C. Tarr, Y. Chen, unpub. data) and compared with those publicly available (www.pasteur.fr/mlst). MVLST data were obtained as described (3) or extracted from whole genome sequences (Y. Chen, unpub. data). Sequences were compared with those on the MVLST database available in the laboratory of S.K. (3,4) and analyzed by using MEGA5.0 (10). New virulence types (VTs) were assigned to USDA isolates: VT60 (isolates 239, 441, 442, 458, 541, 565, 577); VT68 (350, 470); VT69 (247); VT70 (502); VT71 (450); VT72 (342), and VT73 (267). comK prophage JFs were sequenced as described (5). Prophage types (PTs) were assigned by comparing JF sequences with those available from previous reports (4,5). comK prophage JF sequences were submitted to GenBank for isolate L2676 (accession nos. JQ407079 and JQ407080) and 3 USDA isolates (accession nos. JQ750615–JQ750618).

Isolates L2624, LIS0075, and LIS0078 (PFGE profile 1) belonged to the globally disseminated ST5 (6) and had the same VT (VT63) as 5 other 1/2b isolates in the database: isolates 10-0810 and 10-0811, from an imitation crabmeat–borne outbreak in Canada during 1996 (4,11); and isolates 98-0041, 233, and 466 (Table; Figure 2). Because VT63 isolates were associated with multiple outbreaks, they should be considered part of a novel EC (ECVI). ECVI isolates from cantaloupe and USDA isolate 233 showed no amplification of comK prophage JFs (Table). PT11/11 was identified during the 1996 imitation crabmeat–associated outbreak in Canada (4) and in USDA isolate 466 (Table). Further research is needed to determine why comK PTs were identical during different years and in different geographic locations and food processing plants.

Isolate L2625 (VT74, PFGE profile 2) from cantaloupe differed by 1 single nucleotide polymorphism in inlC from 3 other serotype 1/2a VT61 isolates (10-4758, 10-4754, and 06-6956) associated with the 2002 cheese-associated listeriosis outbreak in Canada (4,12) (Table; Figure 2). L2625 was assigned to ST29, an infrequent sequence type in the Institut Pasteur MLST database that differs from the ST (ST405) assigned to the isolates from cheese in the 2002 outbreak in Canada. No amplification of comK prophage JFs was observed, consistent with the PTs in the 2002 cheese-associated outbreak in Canada (4). Given this evidence, isolate L2625 does not represent a novel EC but should be considered a novel outbreak strain.

Isolates L2626 and LIS0077 (PFGE profile 3, ST7) and L2676, LIS0072, and LIS0087 (PFGE profile 4, ST561) from cantaloupe samples shared the same VT (VT56) as isolates 10-0813 and 10-0812 associated with a listeriosis outbreak related to whipping cream during 2000 in Canada (4,12) and isolates 06-6909, BL0047, 261, and 498 (Table; Figure 2). These Listeria isolates from cantaloupe displayed 2 highly similar PFGE profiles and STs, and the same serotype, ApaI PFGE pattern, and VT (Table; Figure 1). Isolates L2626 and LIS0077 showed no amplification of comK prophage JFs, which was also consistent with the upstream PT in the outbreak associated with whipping cream in Canada (Table). The JF sequences in isolates L2676, LIS0072, and LIS0087 were identical to those in USDA isolate 261 (Table). These isolates matched those from the whipping cream–associated outbreak in Canada in terms of VT56 and downstream PT (PT13) (Table). However, the upstream JF could not be amplified in the strain identified in whipping cream (4), possibly because of extensive recombination within the comK prophage (13), especially in the upstream JF (5). These STs and VTs were also found in clinical isolates over extended periods (6). Therefore, by definition (2,3), these isolates also represent a novel EC (ECVII).

Different clones, particularly ECVI and ECVII, might have cocolonized niches or harborage sites within the cantaloupe processing facility, possibly explaining the multiple strains associated with this outbreak. Serotype 4b L. monocytogenes strains, of the same genetic lineage as serotype 1/2b strains, reportedly survived and grew substantially better in mixed-serotype biofilms containing a specific strain of serotype 1/2a (14). Although a biofilm was not detected in the cantaloupe facility, because the facility had already been extensively cleaned and sanitized before FDA sampling, further research is needed to determine the potential for these strains to cocolonize with biofilms.

Six of the 7 currently identified ECs were found at some point in 1 or both of the US chicken processing plants included in the study (Figure 2). Listeriosis cases and outbreaks have been associated with consumption of undercooked raw chicken and ready-to-eat poultry products (2,4). Additional research is needed to determine whether poultry or poultry processing plants could be responsible for the global dissemination of ECs of L. monocytogenes.

The molecular epidemiology of L. monocytogenes strains involved in the 2011 multistate cantaloupe-associated outbreak was greatly enhanced by the use of subtyping markers with different levels of epidemiologic resolution. Particularly, MVLST enabled the detection of 1 novel 1/2a outbreak strain and 2 novel ECs of L. monocytogenes. In contrast to focusing on isolates from a single outbreak (15), our findings demonstrate that to detect new ECs it is important to analyze isolates from many sources around the world.

Number of isolates of Listeria monocytogenes encountered in clinical and food or environment samples collected by the Centers for Disease Control and Prevention during a 2011 L. monocytogenes outbreak related to cantaloupe.