Listeria monocytogenes is a facultative intracellular pathogen that causes a disease known as listeriosis in humans as well as in a variety of animal species.1,2 In humans, listeriosis may be characterized by invasive systemic infections, such as meningitis, encephalitis, septicemia, and late-term spontaneous abortions, that occur primarily in susceptible individuals.1 In the United States, an estimated 2,500 cases of clinical listeriosis in humans result in nearly 500 deaths annually,3 making L monocytogenes an important public health concern. Nearly all (99%) cases of listeriosis in humans are thought to result from consumption of contaminated foods.3
Listeriosis in cattle is manifest in 2 major forms including encephalitis and septicemia, which is often complicated by a fetal infection resulting in spontaneous abortion.2 The encephalitic form of listeriosis develops when L monocytogenes infects the nervous system, and this form of disease is typically characterized by facial paralysis, excessive salivation, depressed appetite, and fever.2Listeria monocytogenes can also cause septicemia, and this manifestation of listeriosis can lead to spontaneous abortions in pregnant animals.4Listeria monocytogenes is most commonly acquired through oral transmission in ruminants and is often linked to consumption of contaminated silage.2 Once L monocytogenes infects the brain, it is able to replicate efficiently for an extended period in this protected niche.5 Specifically, results of 1 study6 suggest that nervous tissue serves as an important niche for uninhibited proliferation of L monocytogenes during infection of a host because the number of L monocytogenes in gerbils decreased in the liver and spleen but not the brain 5 days after infection.
Genes encoded on the Listeria pathogenicity island 1 (ie, hly, plcA, plcB, actA, mpl, and prfA)7 as well as the internalin genes (ie, inlA, inlB, inlC, inlC2, inlD, inlE, inlF, inlG, and inlH) are thought to be important for L monocytogenes virulence.8 The gene actA encodes ActA, which initiates polymerization of host-cell actin and facilitates the intracellular spread of L monocytogenes from host cell to host cell.9 Internalin A interacts with epithelial cells that express an appropriate E-cadherin isotype, and there is evidence that internalin B interacts with other host-cell surface proteins and plays a role in infection of hepatic cells.10 Little is known about the remaining internalin genes, but their organization along the chromosome has been determined.11,12
Many studies have been completed to characterize L monocytogenes beyond the species level. Various molecular subtyping methods permit classification of L monocytogenes isolates into 3 genetic lineages (lineages I, II, and III).13Listeria monocytogenes isolates in lineage I appear to be overrepresented among clinical cases of listeriosis in humans, whereas lineage II contains strains linked to disease in humans and animals as well as strains that have been predominantly isolated from nonhost environments.14–17Listeria monocytogenes isolates belonging to lineages I and II appear to be equally represented among clinical isolates from animals.15Listeria monocytogenes lineage III represents a third, rare lineage, and isolates belonging to this lineage have mainly been collected from cases of listeriosis in animals.15
Listeria monocytogenes infections have been linked to a broad range of clinical signs within mammalian hosts, ranging from a healthy fecal carrier state and noninvasive disease to invasive systemic infections. Another Listeria species, Listeria ivanovii, primarily causes abortions and stillbirths in sheep.2 This example of tissue specificity within pathogenic Listeria leads us to hypothesize that L monocytogenes may also have tissue specificity in mammalian hosts. Whereas a number of authors have suggested that selection may favor pathogen subtypes that have evolved towards reduced virulence, evidence suggests that pathogens may evolve to become more or less virulent depending on selective forces.18 In particular, pathogens that are transmitted by mechanisms other than direct host-to-host transmission, such as food or waterborne transmission, may benefit from causing severe disease that may facilitate generation of high pathogen numbers for subsequent dispersal (eg, Vibrio cholerae and Bacillus anthracis).19 The purpose of the study reported here was to determine whether specific L monocytogenes strains, as determined by genetic characteristics and virulence phenotypes, were associated with distinct clinical manifestations of listeriosis in cattle and thus may potentially have tissue specificity.
Materials and Methods
Bacterial isolates—Thirty-two L monocytogenes isolates from cases of listeriosis in cattle were obtaineda Isolates were collected from various farms between January 1, 1986, and December 31, 2000. All isolates thus represent unrelated cases of listeriosis. These isolates had been characterized by EcoRI ribotyping and actA allelic profiling and had been assigned to 1 of 3 genetic lineages, as described by Wiedmann et al.17 Isolates were maintained at −80°C in a 15% glycerol solution. Prior to characterization, isolates were streaked onto brain heart infusion agarb plates and grown for 18 hours at 37°C. A single isolated colony from each plate was transferred into brain heart infusionc broth and incubated at 37°C for 18 hours. Bacterial lysates for use as templates in PCR assays were prepared as previously described.20
Amplification via PCR assay—Amplifications of the full actA open reading frame (1,920 bp) and approximately 800 bp of the 3′ end of inlA were performed via PCR assay. Previously developed PCR assaysd were used to detect the presence or absence of select internalins (ie, inlE, inlC, inlG, inlF, inlC2, inlD, and inlH) to assign an internalin profile to each isolate. Polymerase chain reaction assays were performed by use of 1X PCR assay buffer, Thermus aquaticus DNA polymerase, MgCl2 at a final concentration of 1.5 mM, and deoxynucleotide triphosphatese at a final concentration between 50 and 100 μM. Primers used for PCR assays and DNA sequencing are depicted (Appendix 1). Polymerase chain reaction assay thermocycling conditions included an initial denaturation at 94° to 96°C for 2 to 10 minutes followed by 28 to 40 cycles of denaturation at 94° to 96°C for 1 minute, annealing at 45° to 58°C for 1 minute, and extension at 72°C for 1 to 2 minutes. A hot-start PCR assay was performed for amplification of all internalin gene fragments. Deoxyribonucleic acid polymerasef was used in a touchdown PCR assay, which involved a decrease in the annealing temperature to 0.5°C/cycle, to amplify the 5′ region of actA. Target amplicons were confirmed by use of 1.5% agarose gel electrophoresis followed by ethidium bromide staining and visualization with UV light.
Molecular serotyping—Molecular serotyping was performed by use of a multiplex PCR assay as described previously.21Listeria monocytogenes isolates were assigned to 1 of 5 multiplex PCR assay profiles that can be used to differentiate the major L monocytogenes serotypes. Specifically, profile 1 represents serotypes 1/2a and 3a; profile 2 represents serotypes 1/2b, 3b, and 7; profile 3 represents serotypes 1/2c and 3c; profile 4 represents serotypes 4b, 4d, and 4e; and profile 5 represents serotypes 4a and 4c isolates.
Purification and sequencing of DNA—Deoxyribonucleic acid was purified by use of 1 of 2 purification kits.g,h The DNA concentration of the purified products was determined via a fluorescent DNA quantification kiti by use of a microplate analyzerj or by comparing amplicon band intensities with a known DNA concentration of a standard DNA markerk via computer software.l The DNA sequencing was performed by 1 of 2 facilities.m,n Sequencing was performed by use of terminatoro chemistry and DNA polymerase,p and reactions were run on 1 of 2 DNA analyzers.q,r Sequences were proofreads and alignedt by use of computer software. Each isolate's actA allelic type was confirmed by examining sequences for the presence or absence of a 105-bp deletion, which indicates actA type 3 and actA type 4, respectively.17 The DNA sequences were deposited in a databaseu and are available for download.
Phylogenetic analysis—The appropriate model of DNA substitution was selectedv to explain the evolution of actA and inlA sequences.22 On the basis of the DNA substitution parameters selected,v maximum likelihood phylogenies were generated by use of phylogenetic analysis software.23 Equal weights for all sites and the tree-bisection-reconnection branch-swapping algorithm were used to perform heuristic searches. Lineage III sequences were used to root the trees, as they are distantly related to both lineage I and II isolates.17 A 100-replicate bootstrap analysis was performed to assign confidence measures for the observed tree topologies.
Analyses to detect the presence of positive selection—Inferences about selective pressure acting on the evolution of a gene can be made by calculating ω (ratio of the number of nonsynonymous substitutions per nonsynonymous site to the number of synonymous substitutions per synonymous site). If ω > 1, there is significant evidence that positive selection has occurred.24 A software program25,w was used to test for the presence of positively selected sites within actA and inlA sequences. A series of nested null (0, 1, and 7) and alternative models (3, 2, and 8, respectively) was compared by use of the LRT, as previously described.24 Comparison between models 0 and 3 tests for variation in selective pressure along a sequence. Models 1 and 2 are compared to test for positive selection by use of discrete ω values, whereas comparing models 7 and 8 provides a more robust test of positive selection by use of a discrete approximation of a continuous distribution of ω values. Models 2 and 8 allow a site class in which ω > 1, and if models 1 and 7 are rejected in favor of models 2 and 8, respectively, positive selection is determined to have acted on a given sequence. Models 2 and 8 also calculate the posterior probability that an amino acid site fits into a site class in which ω > 1, and sites that have a high posterior probability (P > 0.95) of belonging to this site class are considered to have evolved by positive selection. Models 2 and 8 were run twice, once with an initial ω estimate of 0.4 and again with an initial ω estimate of 4.0. The run with the best likelihood estimate was used in the LRT to compare nested models of heterogeneous codon substitution.
In vitro cytopathogenicity assays—Plaque assays were performed essentially as previously described14 by use of a mouse fibroblast L2 cell line.x Cells were maintained at 37°C in Dulbecco modified Eagle mediumy with 10% fetal bovine serum, streptomycin, and penicillin. The L2 cells were seeded into 6-well plates at a density of 3 × 105 cells/well and grown to confluency in Dulbecco modified Eagle medium with fetal bovine serum and without antimicrobials. Bacterial cultures were grown at 30°C for 18 hours to stationary phase without shaking and subsequently centrifuged and resuspended in PBS. Approximately 48 hours after seeding L2 cells, each L monocytogenes isolate was used to infect 2 wells containing L2 cell monolayers and the laboratory strain 10403S26 was included in each assay as an internal standard.14 Computer softwarez was used to measure the size of approximately 25 plaques for each isolate. Mean plaque sizes were standardized to 10403S, which was assigned a plaque size of 100%. Mean plaque sizes for each isolate were calculated from 2 independent plaque assays.
Statistical analysis—All statistical analyses were performed by use of computer software.aa χ2 Tests of independence were performed to analyze associations between clinical manifestation of listeriosis (ie, encephalitis vs septicemia and fetal infection) and L monocytogenes genotypes (ie, genetic lineage, ribotype, internalin profile, and actA allelic type). χ2 Tests of independence were also used to assess the association among amino acid residues at positively selected sites in actA and inlA with L monocytogenes genetic lineage and clinical manifestation of listeriosis. When > 25% of the expected values in a given table were < 5, the Fisher exact test was used. Analysis of variance was used to describe the association among L monocytogenes virulence phenotypes, as determined by mean plaque size and categoric variables (ie, clinical manifestation of listeriosis, genetic lineage, ribotype, and actA allelic type). Statistical analyses were performed only for ribotypes that were observed at least 4 times in our sample set (ie, ribotypes DUP-1038B and DUP-1039C). Values of P ≤ 0.05 were considered significant.
Results
Internalin profile analyses for selected L monocytogenes strains—Polymerase chain reaction assay amplifications were performed to screen for the presence or absence of specific internalin genes (ie, inlE, inlC, inlG, inlF, inlC2, inlD, and inlH). On the basis of the presence or absence of these internalin genes, L monocytogenes isolates were assigned to 1 of 6 internalin profiles, which were previously defined by use of 120 L monocytogenes isolates from clinical cases of listeriosis in humans and animals as well as in foodsd (Figure 1; Appendix 2). All L monocytogenes lineage I isolates were classified as internalin profile I and carried inlE, inlC, inlC2, andinlD. Of the 17 lineage II isolates studied, 16 were internalin profile II, carrying inlE, inlC, inlG, inlF, inlC2, andinlD, whereas 1 isolate (FSL E1-003) was designated profile I. Lineage I isolates appear to have lost inlF and inlG. Lineage III isolates had greater diversity in their internalin profiles, compared with lineage I and II isolates, with 1 isolate falling into profile IV (ie, inlE, inlC, inlG, inlC2, and inlD) and another falling into profile VI (ie, inlE, inlG, inlC2, and inlD). Additionally, 2 lineage III isolates (FSL J2-067 and J2-069) could not be classified into any of the previously defined internalin profiles.d The FSL J2-067 isolate carried inlC2, inlD, andinlE (newly designated internalin profile VII), whereas FSL J2-069 carriedinlE, inlC, and inlF (newly designated internalin profile VIII). All isolates that were not classified into internalin profiles I and II were grouped into a category termed “other” for statistical analyses.
Results of agarose gel (1.5%) electrophoresis of results of PCR assay to amplify a fragment of inlG from isolates of Listeria monocytogenes obtained from cattle with clinical manifestations of listeriosis. Lane 1, standard DNA markerk; lane 2, lineage II isolate; lanes 3 and 4, lineage I isolates; lane 5, lineage II positive control; lane 6, negative control by using sterile deionized water as template.
Citation: American Journal of Veterinary Research 67, 4; 10.2460/ajvr.67.4.616
Associations among clinical manifestations of listeriosis and genetic characteristics of L monocytogenes—Analyses were performed to probe associations between clinical manifestation of listeriosis in cattle (ie, encephalitis vs septicemia and fetal infection) and L monocytogenes genotypes (ie, genetic lineage, ribotype, actA type, and internalin profile; Appendix 2). An overall significant (P < 0.05) association was observed between genetic lineage and clinical manifestation of listeriosis in cattle. Specifically, L monocytogenes isolates from cases of encephalitis were overrepresented (P = 0.05) among lineage I isolates (Table 1). On the other hand, isolates from cattle with both major clinical manifestations of listeriosis (ie, encephalitis vs septicemia and fetal infection) were common (P > 0.05) among lineage II. Isolate actA allelic types were also overall significantly (P = 0.004) associated with clinical manifestation of listeriosis. Isolates of L monocytogenes linked to cases of encephalitis were associated with actA type 3, whereas isolates causing septicemia and fetal infection were associated with type 4. Whereas an overall significant (P < 0.001) association between L monocytogenes ribotypes and clinical manifestation of listeriosis was observed, it was not possible to discern significant associations among individual ribotypes and clinical manifestations of listeriosis attributable to the small number of observations for each ribotype. However, all ribotype DUP-1038B isolates studied were obtained from cases of encephalitis. As internalin profiles were nearly exclusive to genetic lineage, associations between internalin profile and manifestation of listeriosis were essentially the same as those observed between lineage and form of disease.
Distribution of Listeria monocytogenesmolecular subtypes in isolates from cattle with encephalitis and septicemia or fetal infection.
DNA sequencing and molecular phylogeny of virulence genes for isolates of L monocytogenes from cases of listeriosis in cattle—The actA alignment confirmed that all isolates previously designated as actA type 3 carried a 105-nucleotide deletion encoding for a proline-rich repeat region that was present in actA type 4 isolates. The actA phylogeny indicated that L monocytogenes isolates from cases of listeriosis in cattle can be divided into at least 2 major evolutionary lineages that correspond to genetic lineages assigned by EcoRI ribo-typing (Figure 2). Lineage I appeared to be a highly clonal clade, compared with the more genetically diverse lineage II clade. Results of molecular serotyping indicated that lineage I isolates represented molecular serotype profiles that contain serotype 1/2b (n = 5; serotype profile 2) and serotype 4b (6; serotype profile 4; Appendix 2).
Maximum likelihood phylogram inferred from actA sequence data. Taxa labels include isolate name, genetic lineage, clinical manifestation of listeriosis, and actAallele type. For example, J2-069. Encephalitis.4 represents isolate FSL J2-064, from a case of encephalitis that carries an actA type 4 allele. The phylogram was rooted by use of lineage III isolates (ie, J2-069, J2-067, J2-071, and J2-074) as an out-group. Maximum likelihood bootstrap support measures > 70% are indicated at the appropriate nodes. The scale bar represents the number of nucleotide changes per length of sequence.
Citation: American Journal of Veterinary Research 67, 4; 10.2460/ajvr.67.4.616
Partial inlA sequencing results indicated 3 different small in-frame deletions (9 nucleotides each) in 5 L monocytogenes isolates (ie, FSL E1-001, FSL E1-054, FSL J2-002, FSL J2-067, and FSL J2-069). Interestingly, FSL J2-067 and FSL J2-069 also had unique internalin profiles that had not been observed previously.d The phylogenetic tree based on partial inlA sequences also indicated clustering of isolates consistent with genetic lineage and, thus, internalin profile, with the exception of a lineage I isolate (E1-054), which clustered with lineage III isolates and a single lineage III isolate (J2-067), which clustered with lineage I isolates (Figure 3). As with actA, the molecular phylogeny ofL monocytogenes inferred from inlA indicated lineage I to be the most clonal clade, which appears to be more closely related to lineage III isolates than the more genetically diverse lineage II clade.
Maximum likelihood phylogenetic tree based on partial inlAsequence data. Taxa labels include the isolate name (eg, J2-064 represents isolate FSL J2-064), clinical manifestation of listeriosis (eg, fetal infection), and internalin profile (eg, I represents internalin profile I). The phylogram was rooted by use of lineage III isolates (ie, J2-069, J2-067, J2-071, and J2-074) as an outgroup. Nodes with bootstrap values > 70% are indicated at the appropriate nodes. The scale bar represents the number of nucleotide changes per length of sequence.
Citation: American Journal of Veterinary Research 67, 4; 10.2460/ajvr.67.4.616
Analysis for positive selection and distribution of amino acid residues at positively selected amino acid sites—The LRT was used to compare nested models of heterogeneous codon substitution to test for the presence of positive selection in actA and inlA. The LRT statistic calculated for model 0 versus model 3 was significant (P < 0.001) for actA and inlA, indicating that there is variation in selective pressure alongactA and the 3′ end of inlA (Table 2). By comparing models 1 and 2, we uncovered significant (P < 0.05) evidence to support that actA evolved by positive selection. However, for inlA, the null model (model 1) could not be rejected in favor of a model allowing positive selection (P > 0.05). Comparison of a more robust pair of nested models to detect positive selection (models 7 and 8) yielded a significant LRT for both actA (P < 0.01) and inlA (P < 0.05), which supports the hypothesis that both actA and the 3′ end of inlA have diversified by positive selection. Model 8 also identified amino acid sites deemed to have evolved under positive selection, (ω > 1) with a posterior probability > 0.95.22 Twenty-one actA and 12 inlA sites were found to have evolved by positive selection (Tables 3 and 4, respectively). Two actA sites (sites 112 and 449) and 2 inlA sites (sites 558 and 664) affected by positive selection had an overall significant (P < 0.05) association with clinical manifestation of listeriosis. Specifically, at actA site 112, aspartic acid was more common among isolates from cases of encephalitis, whereas glutamic acid was more common among isolates from cases of septicemia and fetal infection. At actA site 449, more isolates from cases of encephalitis had a threonine residue, whereas alanine was observed only in isolates from septicemia or fetal infection cases. At inlA site 558, asparagine was more common among isolates from cases of encephalitis and arginine was more prevalent in septicemia and fetal infection isolates. At inlA site 664, alanine was overrepresented among L monocytogenes isolates obtained from encephalitis cases. All positively selected actA sites had a significant (P < 0.05) overall association with genetic lineage, whereas 6 inlA positively selected sites had a significant (P < 0.05) overall association with genetic lineage.
Summary of analysis for variation in selective pressure and identification of positively selected amino acid sites in actA and inlA from L monocytogenesisolates from cattle with listeriosis.
Summary of distribution of amino acid residues by categoric variables for actAsites determined to have undergone positive selection.
Summary of distribution of amino acid residues by categorical variables for inlAsites determined to have undergone positive selection.
In vitro cytopathogenicity of isolates of L monocytogenes from cases of listeriosis in cattle—The mean plaque size formed by L monocytogenes isolates in mouse L2 cells was used to assess associations between the ability to spread intracellularly in mammalian host cells and L monocytogenes genetic characteristics, as well as clinical manifestation of listeriosis in cattle. Listeria monocytogenes isolates from cattle with clinical signs of listeriosis representing genetic lineage I formed significantly (P < 0.05) larger plaques and thus had an enhanced ability to spread intracellularly, compared with lineage II isolates (Table 5). Mean plaque sizes for L monocytogenes isolates carrying either actA type 3 or actA type 4 alleles were statistically similar (P > 0.05). Plaque sizes formed by L monocytogenes isolates from cases of encephalitis were typically larger than plaques formed by isolates responsible for cases of septicemia and fetal infection. Isolates representing ribotype DUP-1038B yielded the largest plaques and produced significantly (P < 0.05) larger plaques, compared with other L monocytogenes isolates. Whereas only a small number of lineage III isolates (n = 4) were represented in our study, providing insufficient sample size for statistical comparisons, lineage III isolates had a similar mean plaque size (115%), compared with lineage I isolates (116%).
Mean ± SD plaque sizes formed by L monocytogenes isolates from cattle with encephalitis or septicemia and fetal infection.
Discussion
Listeria monocytogenes isolates from cases of listeriosis in cattle were studied to investigate associations among pathogen factors and clinical manifestations of listeriosis (ie, encephalitis vs septicemia and fetal infection). Results of our study indicated that genetic characteristics of L monocytogenes are associated with distinct clinical manifestations of listeriosis in cattle; positive selection in 2 key virulence genes of L monocytogenes has contributed to the diversification of 2 major lineages that differ in their association with disease manifestation; and L monocytogenes lineage I isolates had a greater ability to spread intracellularly within infected host cells, compared with lineage II isolates, and are associated with an encephalitic manifestation of listeriosis rather than septicemia and fetal infections, further supporting that this lineage is particularly virulent. We conclude that L monocytogenes from clinical cases of listeriosis represents a highly clonal subpopulation (lineage I), which is associated with encephalitis in cattle, as well as a more genetically diverse lineage (lineage II), which is not associated with a specific manifestation of listeriosis.
Our analyses revealed a significant difference in the distribution of L monocytogenes isolates, causing cases of encephalitis versus septicemia and fetal infection among the 2 major genetic lineages within L monocytogenes. Specifically, L monocytogenes isolates from cases of encephalitis were overrepresented among genetic lineage I, and ribotype DUP-1038B, a lineage I subtype, was exclusively associated with cases of encephalitis, consistent with results of a previous study17 that also indicated that all DUP-1038B animal isolates characterized were obtained from cases of encephalitis. Ribotype DUP-1038B was also found to be associated with isolation from ruminant fecal samples as opposed to the farm environment.27 The combination of these findings suggests that DUP-1038B represents a host-associated subtype, which seems particularly likely to cause encephalitis in animals. Interestingly, this subtype is also a common cause of listeriosis in humans and outbreaks of listeriosis.14,15 Studies investigating prevalence and molecular subtyping have provided preliminary evidence that L monocytogenes subtypes are also associated with various forms of disease in humans and, possibly, tissue specificity in humans.28 Specifically, results of 1 study28 indicate that serotype 4b, which is typically classified into L monocytogenes genetic lineage I, was most commonly isolated from pregnant women and was thus suggested to be particularly virulent.28 Although we did not investigate associations among subtypes and CNS infections, this previous study further supports that lineage I isolates or possibly some subtypes within lineage I may have unique pathogenic potential in various hosts.
In our study, a significant overall association was also observed between manifestation of listeriosis and actA allelic types as well as internalin profiles. Cases of encephalitis were most frequently linked to isolates with internalin profile I (lacking inlF and inlG) and actA type 3 (lacking one actA PRR), whereas actA type 4 isolates were more commonly implicated in cases of septicemia and fetal infection, and internalin profile II isolates were associated with both forms of disease. Interestingly, lineage I isolates lack internalin genes (ie, inlF and inlG), which appear to play a less important role in the pathogenesis of listeriosis in at least some mammalian hosts as determined by in vitro characterization of inlF and inlGHE null mutants.11,12 It is thus tempting to speculate that gene-loss events (ie, for internalin genes) and deletion mutations (ie, in actA) may be associated with the virulence potential and tissue specificity patterns observed for lineage I and II isolates. Interestingly, enhanced virulence in subtypes with gene deletions has also been observed in other bacterial species. For example, a chromosomal deletion surrounding the cadA region in the Shigella genome was linked to greater pathogenicity, compared with strains that retained this region.29,30 Our results suggested that selective pressure for the maintenance or loss of L monocytogenes surface proteins or domains within surface proteins may play a role in the diversification of virulence phenotypes and, possibly, tissue tropism within this species.
The topologies of phylogenetic trees inferred from full actA and partial inlA sequences observed in our study are consistent with results of other studies,16,31,32 indicating that L monocytogenes contains at least 2 major evolutionary lineages. Consistent with results of our study, results of other studies31,32 also indicate that L monocytogenes lineage I is highly clonal, whereas lineage II is more genetically diverse. It has been hypothesized that the clonal structure of lineage I may be explained by emigration of a small group of host-associated lineage II isolates through a population bottleneck, resulting in a highly clonal and possibly mammalian host-associated population that we designated as lineage I.16,31,32 Our findings further expand the hypothesis that L monocytogenes lineage I is a highly clonal, mammalian host-associated population and indicate that this lineage may be associated with specific clinical manifestation of listeriosis in cattle. The increased level of genetic diversity observed within lineage II isolates has been proposed to be attributed to a higher frequency of horizontal gene transfer events within this subpopulation.31,32 The increased level of genetic diversity within lineage II isolates may convey a greater ability of thisL monocytogenes subpopulation, compared with lineage I isolates, to be ecologically successful in a variety of host tissues and nonhost environments, as supported by the more common isolation of lineage II strains from non-host environments, compared with lineage I strains.14 Our results are thus also consistent with the hypothesis that lineage II represents a generalist lineage,32 which we found appears to be equally likely to cause various clinical manifestations of listeriosis in cattle.
We also identified several amino acid sites along actA and partial inlA sequences that have diversified via positive selection. Many codons found to have undergone positive selection encoded amino acid residues that were associated with genetic lineage and some were associated with clinical manifestation of listeriosis. Positive selection in actA and inlA may thus be an important mechanism for the diversification and fixation of lineages and clonal groups within L monocytogenes, including some groups that may be associated with specific disease manifestations in an infected host.
Interestingly, virulence differences and tissue specificity for strains within a given species have also been observed in other bacterial pathogens. For example, Melles et al33 found that disease-associated Staphylococcus aureus isolates are overrepresented in several phylogenetic clusters in this species. A specific clonal cluster was significantly associated with impetigo (a particular clinical manifestation of S aureus infection), and another strain was predominantly linked to bacteremia, indicating that S aureus strains differ in virulence and, potentially, in tissue specificity. Similarly, Streptococcus suis, a facultative swine pathogen linked to a healthy carrier state, as well as a range of invasive infections including pneumonia, arthritis, septicemia, and meningitis, has been found to be a genetically diverse species with a small number of sub-types linked to most cases of disease. Results of multilocus sequence typing indicated that 1 S suis sequence type is associated with isolation from animals with invasive disease and was characterized as highly virulent by use of a porcine infection model, whereas the other 2 common sequence types were predominantly isolated from lungs of pigs with respiratory infections and were thus described as less virulent.34 Our results are thus consistent with results of that study and the clonal theory of bacterial pathogens35 and suggested that host-specific; tissue-specific; and, particularly, virulent clonal groups can be defined in many pathogens of veterinary importance, including L monocytogenes.
Results of a mouse-cell plaque assay, evaluating cell-to-cell spread capabilities, indicated that lineage I L monocytogenes isolates formed significantly larger plaques than lineage II isolates, consistent with results of a number of other studies.14,17,36 This provides evidence that lineage I isolates from cattle form a subpopulation that not only is associated with encephalitis but also has enhanced ability to spread intracellularly, consistent with an increased virulence of these strains. Interestingly, isolates representing ribotype DUP-1038B, which was found only among cattle with encephalitis, formed larger plaques than other L monocytogenes isolates, consistent with an overall association between plaque size and ability to cause encephalitis. Although not significant, the association between isolates from cases of encephalitis and a greater ability to spread intracellularly in mammalian hosts cells may be biologically meaningful. It is tempting to speculate that an enhanced ability to spread intracellularly within an infected host may be associated with the ability to cause encephalitis.
Results of the study reported here provide further insight into the evolution of L monocytogenes and suggested that lineage I strains may have CNS tissue specificity in cattle. Gaining access to protected host niches, such as the CNS,5 may permit a pathogen to multiply to high numbers, kill the host, and be reintroduced into its primary reservoir (eg, soil and plants) at high levels. This process may be an evolutionary advantage for pathogens like L monocytogenes that infect new hosts via indirect routes (eg, feedborne transmission) rather than direct transmission from 1 host to another.18,19 This hypothesis is consistent with the experimental observation that L monocytogenes multiplies uninhibited in the CNS in animals that, at the same time, restricts its multiplication in the spleen and liver.6 We appreciate, however, that selection for highly efficient cell-to-cell spread and for specific actA allelic types, which coincidently could also confer increased animal virulence and CNS tissue specificity, may also occur in other and nonhost-associated niches, such as protozoan hosts.37 Although we thus conclude that the genetic and phenotypic data generated in our study support the conclusion that lineage I represents a subpopulation within L monocytogenes that is associated with a specific manifestation of listeriosis and possibly specific tissue preferences within the ruminant host, the selective pressures and mechanisms driving the evolution of virulence characteristics in the lineage remain to be elucidated further.
ABBREVIATIONS
| LRT | Likelihood ratio test |
| FSL | Food safety laboratory |
Courtesy of the New York State Diagnostic Laboratory, Cornell University, Ithaca, NY.
Brain heart infusion agar, Difco, Detroit, Mich.
Brain heart infusion broth, Difco, Detroit, Mich.
Jia YKK, Nightingale KJ, Boor JE, et al, Department of Food Science, Cornell University, Ithaca, NY: Unpublished data, 2005.
Deoxynucleotide triphosphates, Promega, Madison, Wis.
AmpliTaq Gold, Applied Biosystems, Foster City, Calif.
Qiavac PCR purification kit, Qiagen Inc, Valenica, Calif.
Qiaquick PCR purification kit, Qiagen Inc, Valenica, Calif.
Bio-Rad fluorescent DNA quantification kit, Bio-Rad Laboratories, Hercules, Calif.
Packard Fusion Universal microplate analyzer, PerkinElmer, Shelton, Conn.
pGEM, Promega, Madison, Wis.
LabImage, Kapelan, Halle, Germany.
Macrogen Inc, Seoul, Korea.
Biotechnology Resource Center, Cornell University, Ithaca, NY.
Big Dye Terminator Cycle sequencing kit, Applied Biosystems, Foster City, Calif.
AmpliTaq-FS, Applied Biosystems, Foster City, Calif.
ABI 3730xl, Applied Biosystems, Foster City, Calif.
ABI 3700, Applied Biosystems, Foster City, Calif.
Seqman, DNAStar software, Lasergene 6, Madison, Wis.
Megalign, DNAStar software, Lasergene 6, Madison, Wis.
PathogenTracker, version 2.0, Cornell Food Safety Laboratory, Cornell University, Ithaca, NY. Available at: www.pathogentracker.net. Apr 19, 2005.
MODELTEST, David Posada, University of Vigo, Pontavedra, Spain. Available at: darwin.uvigo.es/software/modeltest.html. Accessed Jan 1, 2005.
Phylogenetic analysis by maximum likelihood, version 3.13, David Swofford, Sinaeur Associates Inc, Sunderland, Mass. Available at: paup.csit.fsu.edu/order.html. Accesessed Jan 1, 2005.
Courtesy of Dr. Helene Marquis, Department of Microbiology and Immunology, College of Veterinary Medicine, Cornell University, Ithaca, NY.
Dulbecco modified Eagle medium, Gibco, Grand Island, NY.
Sigmascan Pro, version 5.0, Statistical Solutions, Saugus, Mass.
SAS, SAS Institute, Cary, NC.
References
- 2↑
Wesley IV. Listeriosis in animals.. In: Ryser ET, Marth EH, eds. Listeria listeriosis and food safety. 2nd ed. New York: Marcel Decker Inc, 1999; 39–73.
- 3↑
Mead PS, Slutsker L, Dietz V, et al. Food-related illness and death in the United States. Emerg Infect Dis 1999; 5: 607–625.
- 5↑
Vazquez-Boland JA, Kuhn M, Berche P, et al. Listeria pathogenesis and molecular virulence determinants. J Clin Microbiol 2001; 14: 584–640.
- 6↑
Blanot S, Muffat Jolly M, Vilde F, et al. A gerbil model for rhombencephalitis due to Listeria monocytogenes. Microb Pathog 1997; 23: 39–48.
- 7↑
Kreft J, Vazquez-Boland JA. Regulation of virulence in Listeria. Int J Med Microbiol 2001; 291: 145–157.
- 8↑
Ireton K, Cossart P. Host-pathogen interactions during entry and actin-based movement of Listeria monocytogenes. Annu Rev Genet 1997; 31: 113–138.
- 9↑
Dussurget O, Pizarro-Cerda J, Cossart P. Molecular determinants of Listeria monocytogenes. Annu Rev Microbiol 2004; 58: 587–610.
- 10↑
Lecuit M. Understanding how Listeria monocytogenes targets and crosses host barriers. Clin Microbiol Infect 2005; 11: 430–436.
- 11
Dramsi S, Dehouse P, Lebrun M, et al. Identification of four new members of the Internalin multigene family of Listeria monocytogenes EGD. Infect Immun 1997; 65: 1615–1625.
- 12
Raffelsbauer D, Bubert A, Engelbrecht F, et al. The gene cluster inlC2DE of Listeria monocytogenes contains additional new Internalin genes and is important for virulence in mice. Mol Gen Genet 1998; 260: 144–158.
- 13↑
Wiedmann M. Molecular subtyping methods for Listeria monocytogenes. J AOAC Int 2002; 85: 524–531.
- 14↑
Gray MJ, Zadoks RN, Fortes ED, et al. Listeria monocytogenes isolates from foods form distinct but overlapping populations. Appl Environ Microbiol 2004; 70: 5833–5841.
- 15↑
Jeffers GT, Bruce JL, McDonough PL, et al. Comparative genetic characterization of Listeria monocytogenes isolates from human and animal listeriosis cases. Microbiology 2001; 147: 1095–1104.
- 16
Ward TJ, Gorski L, Borucki M, et al. Intraspecific phylogeny and lineage group identification based on the prfA virulence gene cluster of Listeria monocytogenes. J Bacteriol 2004; 186: 4994–5002.
- 17↑
Wiedmann M, Bruce JL, Keating C, et al. Ribotypes and virulence gene polymorphisms suggest three distinct Listeria monocytogenes lineages with difference in pathogenic potential. Infect Immun 1997; 65: 2707–2716.
- 19↑
Faruque SH, Chowdhury N, Kamruzzman M, et al. Genetic diversity and virulence potential of environmental Vibrio cholerae population in a cholera-endemic area. PNAS 2004; 101: 2123–2128.
- 20↑
Furrer B, Candrian U, Hoefelein C, et al. Detection and identification of Listeria monocytogenes in cooked sausage products and in milk by in vitro amplification of haemolysin gene fragments. J Appl Bacteriol 1991; 70: 372–379.
- 21↑
Doumith M, Buchrieser C, Glaser P, et al. Differentiation of the major Listeria monocytogenes serovars by multiplex PCR. J Clin Microbiol 2004; 42: 3819–3822.
- 22↑
Posada D, Crandall KA. MODELTEST: testing the model of DNA substitution. Bioinformatics 1998; 14: 817–818.
- 23↑
Swofford D. Phylogenetic analysis using parsimony (*and other methods). Sunderland, Mass: Sinauer Associates, 1998.
- 24↑
Yang Z, Nielsen R, Goldman N, et al. Codon-substitution models for heterogeneous selection pressure at amino acid sites. Genetics 2000; 155: 431–449.
- 25↑
Yang Z. PAML: a program package for phylogenetic analysis by maximum likelihood. Comput Appl Biosci 1997; 13: 555–556.
- 26↑
Bishop DK, Hinrichs DJ. Adoptive transfer of immunity to Listeria monocytogenes: the influence of in vitro stimulation on lymphocyte subset requirements. J Immunol 1987; 139: 2005–2009.
- 27↑
Nightingale KK, Schukken H, Nightingale CR, et al. Ecology and transmission of Listeria monocytogenes infecting ruminants and in the farm environment. Appl Environ Microbiol 2004; 70: 4458–4467.
- 28↑
McLauchlin J. Distribution of serovars of Listeria monocytogenes isolated from different categories of patients with listeriosis. Eur J Clin Microbiol Infect Dis 1990; 9: 210–213.
- 29
Day WA, Fernández RE, Maurelli AT. Pathoadaptive mutations that enhance virulence: genetic organization of the cadA regions of Shigella spp. Infect Immun 2001; 69: 7471–7480.
- 30
Maurelli AT, Fernández RE, Bloch CA, et al. “Black holes” and bacterial pathogenicity: a large genomic deletion that enchances the virulence of Shigella spp. and enteroinvasive Escherichia coli. Proc Natl Acad Sci U S A 1998; 95: 3943–3948.
- 31
Meinersmann RJ, Phillips RW, Wiedmann M, et al. 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.
- 32↑
Nightingale KK, Windham K, Wiedmann M. Evolution and molecular phylogeny of Listeria monocytogenes isolated from human and animal listeriosis cases and foods. J Bacteriol 2005; 187: 5537–5551.
- 33↑
Melles DC, Gorkink FJ, Boelens HA. Natural population dynamics and expansion of pathogenic clones of Staphylococcus aureus. J Clin Invest 2004; 114: 1732–1740.
- 34↑
King SJ, Leigh JA, Heath PF, et al. Development of a multilocus sequence typing scheme for the pig pathogen Streptococcus suis: identification of virulent clones and potential capsular serotype exchange. J Clin Microbiol 2002; 40: 3671–3680.
- 35↑
Musser JM. Molecular population genetic analysis of emerged bacterial pathogens: selected insights. Emerg Infect Dis 1996; 2: 1–17.
- 36
Norton DM, Scarlett JM, Horton K, et al. Characterization and pathogenic potential of Listeria monocytogenes isolates from the smoked fish industry. Appl Environ Microbiol 2001; 67: 646–653.
- 37↑
Ly TM, Muller HE. Ingested Listeria monocytogenes survive and multiply in protozoa. J Med Microbiol 1990; 33: 51–54.
Appendix 1
Description of isolates of Listeria monocytogenes from cattle with clinical manifestations of listeriosis and summary of subtype characteristics.
Appendix 2
Primers for actA sequencing, partial inlAsequencing and internalin profiling.
