Clonal Clustering Using 10-Gene Multilocus Sequence Typing Reveals an Association Between Genotype and Listeria monocytogenes Maximum Growth Rate in Defined Medium

Abstract

We used a 10-gene (10G) multilocus sequence typing scheme to investigate the diversity and phylogenetic distribution of 124 Listeria monocytogenes strains across major lineages, major serotypes, and seven epidemic clones that have been previously associated with outbreaks. The 124 isolates proved to be diverse, with a total of 81 sequence types (10G-STs) belonging to 13 clonal complexes (CCs), where all STs of the same CC differ from one another in up to 3 of the 10 alleles (named as 10G-triple-locus-variant-clonal-complexes [10G-TLV-CCs]). Phenotypic characterization for 105 of the 124 strains showed that L. monocytogenes had variable maximum growth rate (μ_max) in a defined medium at 16°C, and classification by lineage or serotype was not able to reflect the genetic basis for the difference of this phenotype. Among the six major 10G-TLV-CCs, 10G-TLV-CC4 that included lineage I strains had significantly lower μ_max (Tukey honestly significant difference adjusted [adj.] p < 0.05) compared to 10G-TLV-CC1 and 10G-TLV-CC3 that both comprised lineage II strains, indicating a distinct difference in growth of these L. monocytogenes isolates under nutrient-limited conditions among some of the CCs. However, the other three (10G-TLV-CC2, 6, and 10) of the six major 10G-TLV-CCs containing either lineage I or lineage II strains did not show significantly different μ_max compared to the others (adj. p < 0.05). Our findings highlighted the importance of using molecular typing methods that can be used in evolutionary analyses as a framework for further understanding the phenotypic characteristics of subgroups of L. monocytogenes.

Introduction

L isteria monocytogenes is an opportunistic foodborne pathogen that poses a particular risk for pregnant women, neonates, elderly people, and other immunologically susceptible persons. It is ubiquitous in the environment and almost exclusively transmitted by food (Schlech et al., 1983; Farber and Peterkin, 1991). L. monocytogenes isolates have been differentiated into four genetic lineages (I– IV) (Ward et al., 2008; Orsi et al., 2011) and 13 known serotypes (McLauchlin, 1990). Despite having similar prevalence in food and environment samples, lineage I and II strains are more likely to be associated with human hosts and nonhuman (e.g., animal, food, and environment) sources, respectively (Haase et al., 2014), whereas lineages III and IV are mainly from animal hosts (Orsi et al., 2011). Serotypes 4b, 1/2a, and 1/2b are most commonly associated with human illnesses (McLauchlin, 1990). Phenotypic characterization of L. monocytogenes isolates has shown apparent differences among lineages and serotypes in virulence (Wiedmann et al., 1997; Buncic et al., 2001) and growth under various conditions (De Jesus and Whiting, 2003; Bergholz et al., 2010; Tang et al., 2013). However, with the aim of developing control strategies that prevent L. monocytogenes outbreaks, further study is needed to investigate (1) whether particular genetic subtypes of L. monocytogenes are more persistent in environment or food, potentially increasing human exposure to these strains; and (2) which phenotypic characteristics, if any, are linked to L. monocytogenes that are more likely to be associated with outbreaks.

Multivirulence-locus sequence typing (MvLST) (Zhang et al., 2004; Chen et al., 2007) together with pulsed-field gel electrophoresis (Buchrieser et al., 1993), have been used to define epidemic clones (ECs) of L. monocytogenes, where isolates that are genetically related but implicated in temporally and geographically distinct outbreaks are classified in one EC (Kathariou et al., 2003; Chen et al., 2007; Cheng et al., 2008; Lomonaco et al., 2013). Multilocus sequence typing (MLST) (Meinersmann et al., 2004), which targets seven housekeeping genes (7G-MLST), has been used to define the clonal complexes (CCs) of L. monocytogenes (Feil, 2004; Ragon et al., 2008). Cantinelli et al. (2013) recently confirmed the correspondence between ECs and CCs in terms of phylogenetic clustering and discriminatory power. In this study, we used 10-gene MLST (10G-MLST) (den Bakker et al., 2010) to investigate the phylogenetic diversity of L. monocytogenes strains. Further, we compared the available classification results based on these previously described typing schemes (CCs and ECs identified by using 7G-MLST and MvLST scheme, respectively) to the one we used (10G-MLST scheme), to understand the correlations among them and enable cross-referencing between the phylogenetic classification results of these schemes.

The ability to persist at low temperatures and in nutrient-limited food processing environments elevates the risk of food contamination and transmission of this pathogen through foods to humans (Lappi et al., 2004; Swaminathan et al., 2007), thus possibly representing one of the important phenotypes to identify the L. monocytogenes isolates with greater potential to spread through the food supply. Since Lineages III and VI of L. monocytogenes are rarely isolated from human-related cases or from food, we focused on Lineages I and II, and compared the maximum growth rates (μ_max) of L. monocytogenes isolates growing in a chemically defined medium (DM) at 16°C by lineage (I and II) and 10G-MLST defined CCs (where all STs of the same CC differ from one another in 3 or fewer of the 10 alleles). This will facilitate better understanding of the correlation between phenotype and genotype of L. monocytogenes.

Materials and Methods

Isolates

For measurement of μ_max in DM, a group of 105 isolates was assembled from our collection (the Cornell University Food Safety Laboratory (CUFSL]) (Supplementary Table S1; Supplementary Data are available online at www.liebertpub.com/fpd) to represent isolates from a wide range of sources (Tables 1 and 2) and diverse genetic backgrounds based on ribotype data (Supplementary Table S1). For 10G-MLST analysis, we (1) sequenced the MLST genes for 89 of the 105 isolates, and (2) extracted MLST gene sequences for the remaining 16 isolates from the public sequence repositories. Second, in addition to these 105 isolates, 19 isolates with full genome sequence data available were included in the 10G-MLST analysis by extracting MLST gene sequences from the public sequence repositories (Supplementary Table S1). The whole isolate set included (1) 24 human and food isolates from 15 outbreaks, (2) 64 human and food isolates from nonoutbreak cases, (3) 12 animal isolates, and (4) 18 environmental isolates, for a total of 124 L. monocytogenes isolates representing Lineages I, II, III, and IV, as well as 7 of the ECs.

Table 1.

Lineage I Isolates Used for 10-Gene Multilocus Sequence Typing (10G-MLST) Analysis

Strain ID ^a	Source	10G-ST ^b	10G-SLV-CC ^c	10G-TLV-CC ^d	7G-ST ^e	7G-CC ^f	μ_max ^g (log[OD₆₀₀/mL]/d)
FSL J1-0220	1979, human, Boston vegetable outbreak	2	2	2	NA^h	NA	0.047
FSL J1-0225	1983, human, Massachusetts milk outbreak	9	2	2	290	2	0.058
FSL R2-0583	1983, human, Massachusetts milk outbreak	9	2	2	NA	2	0.039
FSL R2-0578	1983, human, Massachusetts milk outbreak	38	2	2	NA	2	0.05
FSL J1-0020	1987, human, Philadelphia multiple foods outbreak	9	2	2	NA	2	0.052
FSL J1-0129	1988–1989, human, UK pate outbreak	2	2	2	NA	2	0.046
FSL R2-0585	1988–1990, food, UK pate outbreak	2	2	2	NA	2	0.021
FSL R2-0589	1988–1990, food, UK pate outbreak	36	2	2	NA	2	0.041
FSL J1-0116	1988–1990, human, UK pate outbreak	2	2	2	2	2	0.062
HPB2262	1997, Italian gastroenteritis outbreak	2	2	2	2	2	NA
FSL F2-0091	1999, human, sporadic	2	2	2	NA	NA	0.055
FSL M2-0042	1999, human, sporadic	37	2	2	NA	NA	0.041
FSL F2-0656	2001, human, sporadic	2	2	2	NA	NA	0.041
FSL F2-0661	2001, human, sporadic	2	2	2	NA	NA	0.041
FSL S4-0848	2002, environment (sidewalk floor)	39	2	2	2	2	0.034
ATCC 19117	Animal (sheep), sporadic	2	2	2	2	2	NA
SLCC 2755	1967, animal (chinchilla)	47	5	4	66	3	NA^h
FSL R2-0502	1994, food, Illinois chocolate milk outbreak	1	5	4	3	3	0.04
FSL R2-0598	1994, human, Illinois chocolate milk outbreak	1	5	4	NA	3	0.06
FSL N1-0017	1998, food	21	5	4	3	3	0.033
FSL M1-0006	1998, human, sporadic	19	Sⁱ	4	NA	NA	0.012
FSL F2-0369	2000, food (ready-to-eat pasta salad)	1	5	4	NA	NA	0.043
FSL R2-0154	2001, food (smoked seafood)	1	5	4	NA	3	0.055
FSL L3-0051	2002, food (ready-to-eat salmon)	1	5	4	NA	3	0.034
FSL F6-0386	2007, food (smoked salmon)	20	5	4	NA	NA	0.049
FSL J1-0049	Human sporadic	1	5	4	3	3	0.028
FSL J1-0194	1997, human, sporadic	30	3	5	88	S	0.036
FSL F2-0521	2000, food (smoked fish salad)	31	3	5	NA	NA	0.048
FSL R8-5459	2010, environment (floor)	32	3	5	NA	NA	0.05
FSL N1-0260	Human	33	3	5	NA	NA	0.031
FSL R2-0557	Human, sporadic	34	3	5	NA	NA	0.033
FSL F6-0366	1998–1999, food (meat), US hot dog outbreak	4	6	6	6	6	0.034
FSL N1-0225	1998–1999, human, US hot dog outbreak	29	6	6	6	6	0.05
FSL N1-0061	1998, food (salmon brine)	27	6	6	NA	NA	0.053
FSL E1-0201	2001, environment (soil)	4	6	6	NA	NA	0.053
FSL N3-0780	2002, animal (bovine feces)	4	6	6	6	6	0.055
FSL N3-0692	2002, environment (soil)	4	6	6	6	6	0.054
FSL R2-0763	2002, human, Northeastern states sliced deli meat outbreak	4	6	6	6	6	0.044
FSL F3-0950	2005, human, sporadic	4	6	6	NA	6	0.049
FSL F6-0095	2005, human, sporadic	28	6	6	NA	NA	0.055
FSL F2-0699	2002, human, sporadic	14	7	7	NA	NA	0.049
FSL F3-0757	2002, human, sporadic	14	7	7	NA	NA	0.028
FSL H5-0804	2006, environment (deli floor drain)	59	7	7	NA	NA	0.052
FSL J1-0175	Environment (water)	58	7	7	87	S	0.052
FSL J1-0108	1981, human, Halifax coleslaw outbreak	8	8	8	1	1	0.059
FSL J1-0126	1983–1987, human, Switzerland cheese outbreak	8	8	8	1	1	0.033
FSL J1-0110	1985, food, Mexican-style cheese outbreak	8	8	8	1	1	0.052
FSL C1-0122	1998, human, sporadic	22	8	8	1	1	0.043
SLCC 2378	Poultry	45	8	8	73	1	NA
FSL J2-0045	1992, animal (sheep), clinical	25	10	9	NA	NA	0.044
FSL F2-0018	1999, human, sporadic	24	S	9	NA	NA	0.047
FSL F2-0366	2000, human, sporadic	26	10	9	NA	NA	0.042
FSL S4-0643	2002, environment (bench)	11	9	12	NA	4	0.052
L312	Food (cheese)	11	9	12	4	4	NA
07PF0776	Human (myocardial abscess), sporadic	42	9	12	4	4	NA
FSL J1-0169	1996, human, sporadic	7	S	Sⁱ	5	5	0.051
FSL E1-0041	2000, animal (sheep), clinical	41	S	S	NA	NA	0.051
FSL F2-0493	2000, food (raw chunk beef patties)	35	S	S	NA	NA	0.047
FSL R2-0501	2000, human, North Carolina cheese outbreak	40	S	S	558	S	0.043
FSL R2-0182	2001, food (smoked seafood)	7	S	S	NA	5	0.061
FSL F2-0693	2001, human, sporadic	60	S	S	NA	NA	0.036
FSL S4-0440	2002, environment (sidewalk floor)	23	S	S	379	NA	0.049
FSL S10-630	2010, environment (produce farm)	43	S	S	NA	NA	0.048
SLCC 2540	1956, human, sporadic	46	S	S	617	3	NA

Isolates are listed in numerical order of the “Lineage,” and then in “10G-TLV-CC.”

10G-ST, sequence type (ST) identified by using 10G-MLST scheme.

10G-SLV-CC, single-locus-variant-clonal complex identified by using 10G-MLST scheme and eBURST algorithm, where all STs in the clonal complex (CC) differ from one another in 1 of the 10 alleles.

10G-TLV-CC, triple-locus-variant-clonal complex identified by using 10G-MLST scheme and eBURST algorithm, where all STs in the same CC differ from one another in no more than 3 of the 10 alleles.

7G-ST, ST identified by using 7-gene multilocus sequence typing (7G-MLST) scheme.

7G-CC, CC identified by using 7G-MLST scheme, where all ST in the same CC differ from one another in one of the seven alleles.

μ_max, the growth parameter, maximum growth rate, of L. monocytogenes isolates growing in defined medium (DM) at 16°C.

NA, not available.

S, singleton, (1) for 10G-SLV-CCs and 10G-TLV-CCs, singletons were defined as STs differing from all of the other STs in more than 1 or more than 3 of the 10 alleles, respectively; (2) for 7G-CCs, singletons were defined as STs differing from all of the other STs in more than 1 of the 7 alleles.

Table 2.

Lineage II Isolates Used for 10-Gene Multilocus Sequence Typing (10G-MLST) Analysis

Strain ID ^a	Source	10G-ST ^b	10G-SLV-CC ^c	10G-TLV-CC ^d	7G-ST ^e	7G-CC ^f	μ_max ^g (log[OD₆₀₀/mL]/d)
FSL R2-0559	1998, food, sporadic	5	1	1	11	11	0.054
FSL J1-0101	1998, human, sporadic	5	1	1	86	11	0.04
FSL C1-0111	1998, human, sporadic	77	S^h	1	NAⁱ	NA	0.037
FSL F2-0237	1999, food (smoked salmon)	18	S	1	NA	NA	0.045
FSL F2-0141	1999, human, sporadic	5	1	1	NA	NA	0.055
FSL F2-0048	1999, human, sporadic	70	1	1	NA	NA	0.063
FSL R2-0011	2000, food (ready-to-eat deli salad)	16	1	1	NA	NA	0.057
FSL F6-0154	2000, food, U.S. turkey deli meat outbreak	5	1	1	86	11	0.053
FSL F2-0516	2000, human, sporadic	16	1	1	NA	NA	0.02
FSL F2-0405	2000, human, sporadic	71	1	1	NA	NA	0.054
FSL R2-0499	2000, human, U.S. turkey deli meat outbreak	76	1	1	11	11	0.059
FSL T1-0073	2001, food (raw Norwegian)	81	S	1	NA	NA	0.065
FSL R2-0487	2001, food (ready-to-eat bagged salad)	74	1	1	NA	NA	0.059
FSL L3-0123	2002, environment (floor drain)	72	1	1	NA	NA	0.061
FSL F3-0566	2003, human, sporadic	73	1	1	NA	NA	0.066
FSL F3-0995	2006, human, sporadic	18	S	1	NA	NA	0.055
FSL R8-0879	2008, environment (floor drain)	75	S	1	NA	NA	0.061
FSL X1-0001	1987, human, sporadic	69	4	3	85	7	0.047
FSL F2-0194	1999, human, sporadic	68	S	3	7	7	0.063
FSL L4-0100	2002, environment (fish-processing floor mat)	49	4	3	NA	NA	0.059
FSL L4-0096	2002, environment (fish-processing plant drain)	50	4	3	NA	NA	0.059
FSL F3-0744	2002, human, sporadic	3	4	3	NA	NA	0.058
FSL F3-0631	2003, human, sporadic	3	4	3	NA	NA	0.048
FSL F6-0084	2004, human, sporadic	3	4	3	NA	NA	0.055
FSL R6-0896	2007, environment (floor drain)	3	4	3	NA	7	0.064
SLCC 5850	1924, animal (rabbit)	3	4	3	12	7	NA
FSL F2-0539	1924, animal (rabbit)	6	11	10	35	9	0.044
SLCC 2372	1935, human, sporadic	13	11	10	122	9	NA
SLCC 2479	1966, human, sporadic	6	11	10	9	9	NA
FSL J1-0022	Human, sporadic	13	11	10	NA	9	0.052
FSL G2-0003	Human, reference strain	61	S	10	210	9	0.034
FSL J1-0125	Human, sporadic	6	11	10	NA	9	0.039
FSL R2-0561	Human, sporadic	6	11	10	9	9	0.053
FSL R2-0089	2000, food (ready-to-eat deli salad)	62	12	13	NA	NA	0.047
FSL F2-0515	2000, food (ready-to-eat turkey deli meat)	15	12	13	NA	NA	0.039
FSL L4-0151	2002, environment (raw salmon room floor drain)	15	12	13	NA	NA	0.044
SLCC 7179	1986, food (cheese)	51	S	S	91	NA	NA
FSL J2-0003	1993, animal (bovine), clinical	79	S	S	89	NA	0.048
FSL J2-0054	1993, animal (sheep), clinical	48	S	S	412	S	0.061
Finland1998	1998, Finland butter outbreak	10	S	S	155	NA	NA
FSL C1-0051	1998, human, sporadic	10	S	S	NA	NA	0.054
FSL C1-0115	1998, human, sporadic	80	S	S	370	S	0.047
FSL F2-0032	1999, food (smoked whitefish)	66	S	S	NA	NA	0.054
FSL F2-0216	1999, human, epidemic	10	S	S	NA	NA	0.054
FSL F2-0039	1999, human, sporadic	63	S	S	NA	NA	0.049
FSL E1-0154	2001, environment (soil)	65	S	S	NA	NA	0.048
FSL R2-0493	2001, food (ready-to-eat deli salad)	64	S	S	NA	NA	0.018
FSL L4-0396	2002, environment (finished product area drain)	67	S	S	NA	NA	0.048
FSL N3-0165	2002, environment (soil)	78	S	S	90	90	0.057
08_5578	2008, Canadian deli meat outbreak	17	S	S	292	8	NA
08_9523	2008, Canadian deli meat outbreak	17	S	S	120	8	NA

Isolates are listed in numerical order of the “Lineage,” and then in “10G-TLV-CC.”

10G-ST, sequence type (ST) identified by using 10G-MLST scheme.

10G-SLV-CC, single-locus-variant-clonal complex identified by using 10G-MLST scheme and eBURST algorithm, where all STs in the clonal complex (CC) differ from one another in 1 of the 10 alleles.

10G-TLV-CC, triple-locus-variant-clonal complex identified by using 10G-MLST scheme and eBURST algorithm, where all STs in the same CC differ from one another in no more than 3 of the 10 alleles.

7G-ST, ST identified by using 7-gene multilocus sequence typing (7G-MLST) scheme.

7G-CC, CC identified by using 7G-MLST scheme, where all ST in the same CC differ from one another in 1 of the 7 alleles.

μ_max, the growth parameter, maximum growth rate, of Listeria monocytogenes isolates growing in defined medium (DM) at 16°C.

NA, not available.

MLST

All of the 124 isolates were characterized using a previously described 10G-MLST scheme (den Bakker et al., 2010), which includes DNA sequences of partial open reading frames for 10 genes: ldh, lmo0490, prs, sigB, polC, rarA, lmo1555, pbpA, addB, and lmo2763 (den Bakker et al., 2010). This method was selected as it was designed to be used for the majority of Listeria species, not just monocytogenes, and has the potential to be more discriminatory with the inclusion of 10 genes rather than 7. The 10G-MLST data for the isolates from our collection (CUFSL, Supplementary Table S1) can be found on Food Microbe Tracker (http://www.foodmicrobetracker.com). Sequence types based on 10G-MLST (10G-STs) (Tables 1 and 2) were determined by using DnaSP v.5 (Librado and Rozas, 2009). For each 10G-MLST locus, an allele number was given to each distinct sequence variant, and a distinct 10G-ST number was attributed to each distinct combination of alleles at the 10 genes. 10G-ST numbers were initially based on highest frequency for the 10G-STs, and were subsequently increased arbitrarily.

The eBURST V3 (Feil et al., 2004) algorithm was used for the classification of single-locus-variant-clonal complexes (10G-SLV-CCs) and triple-locus-variant-clonal complexes (10G-TLV-CCs). All 10G-STs assigned to the same 10G-SLV-CC and 10G-TLV-CC share identical alleles at 9 and 7 of 10 loci, respectively. Singletons were accordingly defined as 10G-STs having at least two and four allelic mismatches with all other 10G-STs (Supplementary Fig. S1, Tables 1 and 2) for 10G-SLV-CCs or 10G-TLV-CCs, respectively. The central 10G-ST of a clonal complex (CC) was defined as the ST that differed from the largest number of other STs at only a single locus or at three loci for 10G-SLV-CCs or 10G-TV-CCs, respectively. The goeBURST 1.2.1 (Francisco et al., 2009) algorithm was used to visualize the relationship of the 10G-STs by generating a full minimum spinning tree with PHYLOViZ (www.phyloviz.net) (Fig. 1).

FIG. 1.

Ten major 10-gene-triple-locus-variant-clonal-complexes (10G-TLV-CCs) and the minimum spanning tree generated by goeBURST for all 10G-sequence types (STs). 10G-TLV-CCs were identified by using 10G-MLST scheme and eBURST algorithm, where all 10G-STs of the same 10G-TLV-CC differ from one another in no more than 3 of the 10 alleles. Ten major 10G-TLV-CCs (defined as the 10G-TLV-CCs containing more than 2 10G-STs) are depicted by red (Lineage I) and green (Lineage II) circles. Seven epidemic clones and the 7G-CCs are designated next to their corresponding 10G-TLV-CCs and specifically pointed to the group founder 10G-STs. The identified group founder 10G-STs are represented by light-green elements, and the other 10G-STs are represented by blue. The size of each element reflects the number of isolates within a 10G-ST. Locus variant levels (levels 1–10) are shown on links between 10G-STs.

For 60 of the 124 isolates, CC (defined as 7G-CC in present study) were assigned to each strain based on sequence types (ST) (defined as 7G-ST in present study) (Tables 1 and 2) presented in the Pasteur MLST database (http://www.pasteur.fr/recherche/genopole/PF8/mlst/Lmono.html) and in the studies of Ragon et al. (2008) and Cantinelli et al. (2013). All 7G-STs assigned to the same 7G-CC share identical alleles at six of seven loci, and singletons were accordingly defined as 7G-STs having at least two allelic mismatches with all other 7G-STs (Tables 1 and 2). This scheme includes DNA sequences of partial open reading frames for seven genes including abcZ, bglA, cat, dapE, dat, ldh, and lhkA (Ragon et al., 2008). There is one gene (ldh) in common between the 7G-MLST and 10G-MLST scheme.

Maximum growth rate in defined media

To simulate a nutrient-limited (e.g., food processing) environment for L. monocytogenes growth, we measured μ_max of 105 L. monocytogenes isolates using a DM specific to L. monocytogenes (Amezaga et al., 1995) at 16°C, with 25 mM glucose as the carbon source. For growth prior to inoculation in DM, we followed the approach previously described by Stasiewicz et al. (2010). In the DM, optical density (OD)₆₀₀ readings were taken from hour 0 to 3 as time point 0 (started from the first detectable OD₆₀₀ value of each sample), and at every 6 or 12 h after time point 0 until stationary phase (up to 96 h). The four-factor modified logistic growth model described by Baranyi and Roberts (1994) was used to calculate the growth parameter μ_max, (log[OD₆₀₀/mL]/day), for each strain. Regression was carried out by using the NLStools package (v 0.0-5) in R v2.6.2. For growth in DM, each strain was tested once; multiple strains per CC were considered replicates.

Analysis of growth rates

One-way analysis of variance (ANOVA) were used to determine whether μ_max was significantly different between lineages and among 10G-TLV-CCs for the 105 strains tested (Fig. 2), where the values of the μ_max served as the response. The linear model used for the ANOVA was: Y(μ_max) = lineage or 10G-TLV-CC + strain + E(error). Strain was treated as a random effect. Significant difference between a given pair of lineages or 10G-TLV-CCs was identified by the Student t-test and Tukey honestly significant difference (HSD) method (α = 0.05; JMP 7; SAS Institute, Inc., Cary, NC), respectively. Student t-test p-value (P) and Tukey HSD adjusted p-values (adj. P) are reported as significant when <0.05.

FIG. 2.

Boxplots of the distribution of Listeria monocytogenes growth parameter maximum growth rate (μ_max) for six major 10-gene-triple-locus-variant-clonal-complexes (10G-TLV-CCs) (defined as each has five or more isolates) in defined media at 16°C. 10G-TLV-CC1, 3, and 10 represent Lineage II strains, and 10G-TLV-CC2, and 4 and 6 represent Lineage I strains. The horizontal bar indicates the median for each treatment. Boxes represent the 25th to 75th percentile of the values; whiskers represent the 10th and 90th percentiles. Filled circles represent values outside the 10th to 90th percentiles. Boxplots with the same letter are not statistically different from each other (overall α = 0.05, Tukey honestly significant difference correction).

Results and Discussion

Ten-gene MLST allows for a high level of discrimination of L. monocytogenes

Based on 10G-MLST, the 124 isolates were found to represent a total of 81 10G-STs (Supplementary Fig. S1, Tables 1 and 2). Four 10G-STs (10G-ST1, 2, 3, and 4) predominated, with five or more isolates belonging to each of the 4 STs, and cumulatively account for 21% of isolates (26/124). Among them, 10G-ST1, 2, and 4 belong to lineage I, and 10G-ST3 belongs to lineage II. Overall, Lineage I (39 STs, 64 isolates) contains more STs than lineage II (34 STs, 51 isolates). All 10G-STs containing more than one isolate originated from more than one source, which is consistent with the finding of Haase et al. (2014). The 10G-STs, 10G-ST1, 2, and 3, all included outbreak related isolates and correspond to 7G-ST3, 7G-ST2, and 7G-ST6, respectively. Interestingly, despite that two isolates from the U.S. hot dog outbreak FSL F6-0366 (H7858) and FSL N1-0225 (H7550) were both assigned as 7G-ST6 previously, they were differentiated into two distinct 10G-STs (10G-ST4 and 10G-ST29). The 10G-MLST showed a higher discriminatory power (D = 0.9842) (Hunter, 1990) compared to 7G-MLST (D = 0.9774), indicating that, although with some exceptions, 10G-MLST is generally more discriminatory than 7G-MLST (Tables 1 and 2). This suggests that 10G-MLST is an effective approach for characterizing L. monocytogenes DNA sequence polymorphism with a higher resolution compared to the 7G-MLST scheme.

While 10G-MLST shows increased discrimination over 7G-MLST, most likely these methods will be replaced with whole genome sequencing (WGS) (Roetzer et al., 2013), which is being adopted by public health agencies for routine surveillance (Harris et al., 2013; den Bakker et al., 2014), including an MLST-based approach to analyze genome sequence data (wgMLST) (Jolley and Maiden, 2014). However, interpreting genetic variation will present new challenges to investigation using WGS or wgMLST, whereas the use of a smaller number of genes for clustering may still be useful and efficient to define genetically distinct groups with phenotypic differences. In addition, 7G- or 10G-MLST data will be easily comparable with genome-based data, since the necessary sequences can be easily extracted from WGS data.

eBURST clustered 124 isolates into 12 10G-SLV-CCs and 13 10G-TLV-CCs

We used the eBURST algorithm to identify 12 10G-SLV-CCs and 42 singletons from 81 10G-STs, where all 10G-STs belonging to the same 10G-SLV-CCs share identical alleles at 9 of the 10 loci. Two 10G-SLV-CCs, 10G-SLV-CC1 (12 isolates, 8 STs) and 10G-SLV-CC2 (16 isolates, 6 STs), predominated and accumulatively accounted for 23% isolates (28/124), and 17% of the 10G-STs (14/81); they belong to Lineage II and Lineage I, respectively. Overall, Lineage I (8 10G-SLV-CCs, 11 singletons) contains more 10G-SLV-CCs than Lineage II (4 10G-SLV-CCs, 22 singletons). A large proportion of the outbreak-associated isolates were classified into 10G-SLV-CC2, accounting for 42% of the outbreak-associated isolates (10/24). Overall, isolate classifications into 10G-SLV-CCs and 7G-CCs were consistent for the majority of the clonal complexes (Tables 1 and 2) with some exceptions; for instance, isolates FSL J1-0169 and FSL R2-0182 assigned as 7G-CC5 were differentiated by 10G-MLST into two singletons.

A less stringent approach allowing three-locus variants between members identified 13 10G-TLV-CCs and 31 singletons among the 124 isolates (Tables 1 and 2). This approach allowed more isolates to be clustered into CCs, and reduced the number of singletons, facilitating a better understanding of the relationship between isolates. Among the 13 10G-TLV-CCs, 10 CCs that each contain more than 2 10G-STs were highlighted in Figure 1, and account for 69% isolates (85/124) and 60% of 10G-STs (49/81) (Tables 1 and 2). The 26 outbreak-associated isolates were distributed in 6 of the 13 distinct 10G-TLV-CCs, including 5 of the 10G-TLV-CCs that each represent 1 EC (Fig. 1, Supplementary Table S1). Among them, each of the ECs only reciprocally corresponds to one 10G-TLV-CCs, showing a consistency between the classification into ECs and 10G-TLV-CCs, which is similar to the findings of Cantinelli et al. (2013) that MLST-based CCs correspond to the EC classification.

Growth rates show considerable differences between lineages and between 10G-TLV-CCs

Our data showed that L. monocytogenes had highly variable μ_max across the 105 isolates in DM at 16°C, ranging from 0.012 to 0.066 log(OD₆₀₀/mL)/day (Tables 1 and 2). More specifically, a one-way ANOVA model with lineage effect (two levels: Lineages I and II) for the growth parameter μ_max showed a significant effect (p = 0.0059) of lineage on μ_max (Table 3). Lineage I isolates had significantly lower (p < 0.05) average μ_max (0.045 ± 0.010 log[OD₆₀₀/mL]/day) compared to Lineage II isolates [0.052 ± 0.011 log(OD₆₀₀/mL)/day] (Table 3). In contrast, Bergholz et al. (2010) demonstrated that Lineage I strains had a significantly higher growth rate than Lineage II strains in brain heart infusion broth (BHIB) with 6% salt at 37°C. Moreover, our previous study on L. monocytogenes growth on cold smoked salmon and in modified BHIB with 4.65% salt at 7°C showed that there was no significant difference between the average μ_max of Lineage I and Lineage II strains under either of the growth conditions (Tang et al., 2013). Possible explanations for these observed differences include the high salt content and the difference of the incubation temperatures, while the DM used in the present study simulated a nutrient-limited food-processing environment, usually with relatively low temperatures (about 16°C). Similarly to our finding, Bruhn et al. (2005) found that Lineage II strains outcompeted Lineage I strains in selective enrichment media. However, based on the findings of Haase et al. (2014) that environmental and food isolates are equally distributed between Lineage I and II, it is not likely that the ability of Lineage II strains to grow faster in nonhost environments (especially nutrient-limited food-processing environments) can increase the prevalence of all Lineage II strains in environmental and food samples. Our observation, combined with the findings of Haase et al. (2014), may suggest that it is more likely that just some of the subgroups of Lineage II strains can grow faster than Lineage I strains in such environments.

Table 3.

Lineage-Specific Effects and 10G-TLV-CC-Specific Effects on Maximum Growth Rate of Listeria monocytogenes Isolates Grown in Defined Medium at 16°C

Lineage (no. isolates)	μ_max ^a (log[OD₆₀₀/mL]/d)	10G-TLV-CC ^b (no. isolates)	μ_max (log[OD₆₀₀/mL]/d)
Lineage I (57)	0.045 ± 0.010^c A^d	CC2 (14)	0.046 ± 0.011 AB
		CC4 (9)	0.040 ± 0.015 B
		CC6 (9)	0.050 ± 0.007 AB
Lineage II (44)	0.052 ± 0.011 B	CC1 (17)	0.054 ± 0.012 A
		CC3 (8)	0.057 ± 0.006 A
		CC10 (5)	0.045 ± 0.009 AB

μ_max, the growth parameter, maximum growth rate, of L. monocytogenes isolates growing in defined media (DM) at 16°C.

10G-TLV-CC, triple-locus-variant-clonal complex identified by using 10-gene multilocus sequence typing (10G-MLST) scheme and eBURST algorithm, where all sequence types in the same clonal complex (CC) differ from one another in no more than 3 of the 10 alleles.

Results are summarized by means ± standard deviations for strains of each category; each strain was tested once.

Means within a given column with the same letter are not statistically different from each other (overall α = 0.05, Tukey's correction).

To further compare the growth rate of subgroups of Lineage I and Lineage II strains, a one-way ANOVA model with 10G-TLV-CC effect (six levels: 10G-TLV-CC1, 2, 3, 4, 6, 10, each having at least five isolates) for the growth parameter μ_max was conducted, and showed significant effect (p = 0.0087) of 10G-TLV-CCs on μ_max (Fig. 2, Table 3). The 10G-TLV-CC4 showed the lowest average μ_max (0.040 ± 0.015 log[OD_600/mL]/day) and was significantly lower (adj. p < 0.05) than 10G-TLV-CC1 (0.053 ± 0.011 log[OD₆₀₀/mL]/day) and 10G-TLV-CC3 (0.057 ± 0.006 log[OD₆₀₀/mL]/day) (Fig. 2, Table 3). The average μ_max of each of the 10G-TLV-CC2 (7G-CC2), 10G-TLV-CC 6 (7G-CC6) and 10G-TLV-CC10 (7G-CC9) was not significantly different (adj. p > 0.05) from the others (Fig. 2, Table 3). The 10G-TLV-CC4, representing partial Lineage I strains, corresponds to 7G-CC3. The 10G-TLV-C1 and 10G-TLV-C3, representing partial Lineage II strains, correspond to 7G-CC11 and 7G-CC7 (Supplementary Table S1). Interestingly, it has been reported that 7G-CC3 (10G-TLV-CC4) ranked among the four most common clones (7G-CC2, 7G-CC1, 7G-CC3, and 7G-CC9) in the world, and is one of the most highly prevalent 7G-CCs (7G-CC1, 7G-CC2, and 7G-CC3) in Lineage I (Chenal-Francisque et al., 2011). According to our result, it seems that the majority of the most prevalent 7G-CCs aforementioned do not have a significant higher growth rate in DM as compared to the less prevalent 7G-CCs.

We also assessed whether the source or serotype of isolates may influence μ_max in DM. However, a one-way ANOVA model with source effect (human, food, environment, and animal as four levels of this effect) or serotype (1/2a, 1/2b, and 4b as three levels of this effect) for the growth parameter μ_max did not show significant effect (p > 0.05) on μ_max. Although various studies have characterized L. monocytogenes isolates to probe for phenotypic differences between L. monocytogenes lineages or serotypes (Mereghetti et al., 2000; Buncic et al., 2001; Bergholz et al., 2010; Stasiewicz et al., 2010; Tang et al., 2013), our data suggest that it is necessary to compare between L. monocytogenes CCs, which are defined by sequence typing methods such as 10G-MLST, when studying the phenotypes of L. monocytogenes, as classification by lineage or serotype may not be able to reflect the genetic basis for phenotype difference between different phenotypes of L. monocytogenes clusters.

Taken together, our data showed that (1) L. monocytogenes isolates have diverse μ_max in DM at 16°C, and (2) the isolates assigned as 10G-TLV-CC4, which represented part of the Lineage I isolates and was predominated by serotype 1/2b, had lower average μ_max compared to two 10G-TLV-CCs that represented part of the Lineage II isolates and predominated by serotype 1/2a.

Conclusions

Our findings demonstrated that the 124 L. monocytogenes isolates are genetically diverse, and the outbreak-associated isolates were widespread through the diversity of L. monocytogenes. We also identified that, among the 105 isolates tested, there was considerable variation in maximum growth rate in a nutrient-limited growth condition. We found that certain genetic subgroups of Lineage II strains could grow faster than some genetic subgroups of Lineage I strains under this nutrient-limited condition, which may facilitate their persistence in nonhost environments, thus leading to a higher risk of food contamination and transmission of this pathogen through foods to humans. Our data highlighted the importance of using molecular typing methods that are amenable to evolutionary analysis for a further understanding of the phylogenetic distribution of L. monocytogenes isolates and for the phenotypic characterization of genetic subgroups of L. monocytogenes. Classification of L. monocytogenes based on DNA sequence polymorphism, in the long term, may facilitate the surveillance, detection, and control of L. monocytogenes strains that are more likely to cause outbreaks and sporadic listeriosis cases.

Footnotes

Acknowledgments

We thank Maureen Gunderson for assistance with media preparation. This work was supported by New York Sea Grant R/SHH-15, funded under award NA07OAR4170010 from the National Sea Grant College Program of the U.S. Department of Commerce's National Oceanic and Atmospheric Administration, to the Research Foundation of State University of New York, and by Agriculture and Food Research Initiative grant 2010-65201-20575 from the U.S. Department of Agriculture, National Institute of Food and Agriculture, Food Safety Program.

Disclosure Statement

No competing financial interests exist.

References

Amezaga

, Davidson

, McLaggan

, Verheul

, Abee

, Booth

. The role of peptide metabolism in the growth of Listeria monocytogenes ATCC 23074 at high osmolarity. Microbiology, 1995; 141:41–49.

Baranyi

, Roberts

. A dynamic approach to predicting bacterial growth in food. Int J Food Microbiol, 1994; 23:277–294.

Bergholz

, den Bakker

, Fortes

, Boor

, Wiedmann

. Salt stress phenotypes in Listeria monocytogenes vary by genetic lineage and temperature. Foodborne Pathog Dis, 2010; 7:1537–1549.

Bruhn

, Vogel

, Gram

. Bias in the Listeria monocytogenes enrichment procedure: Lineage 2 strains outcompete lineage 1 strains in University of Vermont selective enrichments. Appl Environ Microbiol, 2005; 71:961–967.

Buchrieser

, Brosch

, Catimel

, Rocourt

. Pulsed-field gel electrophoresis applied for comparing Listeria monocytogenes strains involved in outbreaks. Can J Microbiol, 1993; 39:395–401.

Buncic

, Avery

, Rocourt

, Dimitrijevic

. Can food-related environmental factors induce different behaviour in two key serovars, 4b and 1/2a, of Listeria monocytogenes?. Int J Food Microbiol, 2001; 65:201–212.

Cantinelli

, Chenal-Francisque

, Diancourt

, Frezal

, Leclercq

, Wirth

, Lecuit

, Brisse

. “Epidemic clones” of Listeria monocytogenes are widespread and ancient clonal groups. J Clin Microbiol, 2013; 51:3770–3779.

Chen

, Zhang

, Knabel

. Multi-virulence-locus sequence typing identifies single nucleotide polymorphisms which differentiate epidemic clones and outbreak strains of Listeria monocytogenes . J Clin Microbiol, 2007; 45:835–846.

Chenal-Francisque

, Lopez

, Cantinelli

, Caro

, Tran

, Leclercq

, Lecuit

, Brisse

. Worldwide distribution of major clones of Listeria monocytogenes . Emerg Infect Dis, 2011; 17:1110–1112.

10.

Cheng

, Siletzky

, Kathariou

. 11 Genomic divisions/lineages, epidemic clones, and population structure. In: Handbook of Listeria monocytogenes. Boca Raton, FL: CRC Press, 2008, p. 337.

11.

De Jesus

, Whiting

. Thermal inactivation, growth, and survival studies of Listeria monocytogenes strains belonging to three distinct genotypic lineages. J Food Prot, 2003; 66:1611–1617.

12.

den Bakker

, Allard

, Bopp

, Brown

, Fontana

, Iqbal

, Kinney

, Limberger

, Musser

, Shudt

. Rapid whole-genome sequencing for surveillance of Salmonella enterica serovar enteritidis. Emerg Infect Dis, 2014; 20:1306.

13.

den Bakker

, Bundrant

, Fortes

, Orsi

, Wiedmann

. A population genetics-based and phylogenetic approach to understanding the evolution of virulence in the genus Listeria . Appl Environ Microbiol, 2010; 76:6085–6100.

14.

Farber

, Peterkin

. Listeria monocytogenes, a food-borne pathogen. Microbiol Rev, 1991; 55:476.

15.

Feil

. Small change: Keeping pace with microevolution. Nat Rev Microbiol, 2004; 2:483–495.

16.

Feil

, Li

, Aanensen

, Hanage

, Spratt

. eBURST: Inferring patterns of evolutionary descent among clusters of related bacterial genotypes from multilocus sequence typing data. J Bacteriol, 2004; 186:1518–1530.

17.

Francisco

, Bugalho

, Ramirez

, Carriço

. Global optimal eBURST analysis of multilocus typing data using a graphic matroid approach. BMC Bioinformatics, 2009; 10:152.

18.

Haase

, Didelot

, Lecuit

, Korkeala

, Group

L.m.M.S.

, Achtman

. The ubiquitous nature of Listeria monocytogenes clones: A large-scale Multilocus Sequence Typing study. Environ Microbiol, 2014; 16:405–416.

19.

Harris

, Cartwright

, Török

, Holden

, Brown

, Ogilvy-Stuart

, Ellington

, Quail

, Bentley

, Parkhill

. Whole-genome sequencing for analysis of an outbreak of methicillin-resistant Staphylococcus aureus: A descriptive study. Lancet Infect Dis, 2013; 13:130–136.

20.

Hunter

. Reproducibility and indices of discriminatory power of microbial typing methods. J Clin Microbiol, 1990; 28:1903–1905.

21.

Jolley

, Maiden

. Using MLST to study bacterial variation: Prospects in the genomic era. Future Microbiol, 2014; 9:623–630.

22.

Lappi

, Thimothe

, Nightingale

, Gall

, Scott

, Wiedmann

. Longitudinal studies on Listeria in smoked fish plants: Impact of intervention strategies on contamination patterns. J Food Prot, 2004; 67:2500–2514.

23.

Kathariou

, Torrence

, Isaacson

. Foodborne outbreaks of listeriosis and epidemic-associated lineages of Listeria monocytogenes . In: Microbial Food Safety in Animal Agriculture. Torrence

and Isaacson

(eds.). Ames, IA: Iowa State University Press, 2003, pp. 243–256.

24.

Librado

, Rozas

. DnaSP v5: A software for comprehensive analysis of DNA polymorphism data. Bioinformatics, 2009; 25:1451–1452.

25.

Lomonaco

, Verghese

, Gerner-Smidt

, Tarr

, Gladney

, Joseph

, Katz

, Turnsek

, Frace

, Chen

. Novel epidemic clones of Listeria monocytogenes, United States, 2011. Emerg Infect Dis, 2013; 19:147.

26.

McLauchlin

. Distribution of serovars of Listeria monocytogenes isolated from different categories of patients with listeriosis. Eur J Clin Microbiol Infect Dis, 1990; 9:210–213.

27.

Meinersmann

, Phillips

, Wiedmann

, Berrang

. Multilocus sequence typing of Listeria monocytogenes by use of hypervariable genes reveals clonal and recombination histories of three lineages. Appl Environ Microbiol, 2004; 70:2193–2203.

28.

Mereghetti

, Quentin

, Marquet-Van Der Mee

, Audurier

. Low sensitivity of Listeria monocytogenes to quaternary ammonium compounds. Appl Environ Microbiol, 2000; 66:5083–5086.

29.

Orsi

, Bakker

HCd

, Wiedmann

. Listeria monocytogenes lineages: Genomics, evolution, ecology, and phenotypic characteristics. Int J Med Microbiol, 2011; 301:79–96.

30.

Ragon

, Wirth

, Hollandt

, Lavenir

, Lecuit

, Le Monnier

, Brisse

. A new perspective on Listeria monocytogenes evolution. PLoS Pathog, 2008; 4:e1000146.

31.

Roetzer

, Diel

, Kohl

, Rückert

, Nübel

, Blom

, Wirth

, Jaenicke

, Schuback

, Rüsch-Gerdes

. Whole genome sequencing versus traditional genotyping for investigation of a Mycobacterium tuberculosis outbreak: A longitudinal molecular epidemiological study. PLoS Med, 2013; 10:e1001387.

32.

Schlech

3rd , Lavigne

, Bortolussi

, Allen

, Haldane

, Wort

A.J

, Hightower

A.W

, Johnson

, King

, Nicholls

. Epidemic listeriosis—Evidence for transmission by food. New Engl J Med, 1983; 308:203–206.

33.

Stasiewicz

, Wiedmann

, Bergholz

. The combination of lactate and diacetate synergistically reduces cold growth in brain heart infusion broth across Listeria monocytogenes lineages. J Food Prot, 2010; 73:631–640.

34.

Swaminathan

, Cabanes

, Zhang

, Cossart

. Listeria monocytogenes In: Food Microbiology: Fundamentals and Frontiers. Doyle MPaB

, (ed.). Washington, DC: ASM Press, 2007, pp. 457–492.

35.

Tang

, Stasiewicz

, Wiedmann

, Boor

, Bergholz

. Efficacy of different antimicrobials on inhibition of Listeria monocytogenes growth in laboratory medium and on cold-smoked salmon. Int J Food Microbiol, 2013; 165:265–275.

36.

Ward

, Ducey

, Usgaard

, Dunn

, Bielawski

. Multilocus genotyping assays for single nucleotide polymorphism-based subtyping of Listeria monocytogenes isolates. Appl Environ Microbiol, 2008; 74:7629–7642.

37.

Wiedmann

, Bruce

, Keating

, Johnson

, McDonough

, Batt

. Ribotypes and virulence gene polymorphisms suggest three distinct Listeria monocytogenes lineages with differences in pathogenic potential. Infect Immun, 1997; 65:2707–2716.

38.

Zhang

, Jayarao

, Knabel

. Multi-virulence-locus sequence typing of Listeria monocytogenes . Appl Environ Microbiol, 2004; 70:913–920.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB