Abstract
The ability to survive and proliferate in acidic environments is a prerequisite for the infection of Listeria monocytogenes. The glutamate decarboxylase (GAD) system is responsible for acid resistance, and three GAD homologs have been identified in L. monocytogenes: gadD1, gadD2, and gadD3. To examine whether GAD genes are specific to lineage, serovar, or certain subpopulation, we performed a systematic investigation on the prevalence of GAD genes in 164 L. monocytogenes. In contrast to gadD2 and gadD3 conserved in all L. monocytogenes strains, gadD1 was identified in 36.6% (60/164) of L. monocytogenes strains, including all serovar 1/2c and 68.5% (37/54) of serovar 1/2a strains, as well as a small fraction of serovar 1/2b (3.4%, 1/29) and lineage III (13.8%, 4/29) strains. All serovar 4b and lineage IV strains lacked this gene. According to the ascB-dapE structure, L. monocytogenes strains were classified into four subpopulations, carrying inlC2DE, inlGC2DE, inlGHE, or no internalin cluster, respectively. All L. monocytogenes strains with inlGC2DE or inlGHE pattern harbored gadD1, whereas those bearing inlC2DE or no internalin cluster between ascB and dapE lacked gadD1. In addition, other five non-monocytogenes Listeria species lacking ascB-dapE internalin cluster were gadD1-negative. Overall, the presence of gadD1 is not fully dependent on lineages or serovars but correlates with ascB-dapE internalin profiles, suggesting gadD1 might have co-evolved with the ascB-dapE internalin cluster in the primitive L. monocytogenes before divergence of serovars.
Introduction
L
Acid represents one of the most frequently encountered stresses for L. monocytogenes (Ryan et al., 2008; Chen et al., 2011a). The glutamate decarboxylase (GAD) system has been associated with acid resistance in many bacteria. Three GAD homologs, located in three distinct loci, have been identified in L. monocytogenes: gadD1 (lmo0447), gadD2 (lmo2363), and gadD3 (lmo2434) in the strain EGD-e (Glaser et al., 2001). L. monocytogenes is the only bacterium that we know possesses three gad genes (Cotter et al., 2005). While GadD2 is primarily responsible for surviving severe acid challenge (Cotter et al., 2001), GadD1 plays a major role in growth in mildly acidic conditions (Cotter et al., 2005), and also contributes to the tolerance to bile (Begley et al., 2002) and to lantibiotic nisin (Begley et al., 2010). Previous studies observed that gadD1 was absent in a large number of non-serovar 1/2 L. monocytogenes strains. To examine whether GAD genes are specific to lineage, serovar, or certain subpopulation, we performed a systematic investigation of the prevalence of gadD1, gadD2, and gadD3 in L. monocytogenes strains.
Methods
Bacterial strains
A total of 164 L. monocytogenes strains were examined, including 58 lineage I (29 serovar 1/2b, 29 serovar 4b), 72 lineage II (54 serovar 1/2a, 18 serovar 1/2c), 29 lineage III (26 serovar 4a, one serovar 4b, two serovar 4c), and five lineage IV (four serovar 4a, one serovar 4b) strains (Table 1). Also analyzed were six L. innocua strains belonging to four evolutionary subgroups (Chen et al., 2010a), as well as three L. welshimeri, two L. ivanovii, two L. seeligeri, and two L. grayi strains.
Serovar identification was performed by a multiplex PCR with primers from ORF2819, ORF2110, lmo0737, and lmo1118 (Doumith et al., 2004).
DNA manipulations
Genomic DNA was extracted as described previously (Chen et al., 2009), and oligonucleotide primers were synthesized by Invitrogen Biotechnology (Shanghai, China) (Table 2). Polymerase chain reaction (PCR) was conducted using a thermal cycler (MJ Research Inc., Boston, MA), with annealing temperatures depending on primer pairs (Table 2), and the duration of extension depending on the expected length of amplicon (1 kb per min). For DNA sequencing analysis, PCR fragments were purified with the AxyPrep DNA Gel Extraction Kit (Axygen Inc., Union City, CA) and their sequences determined by dideoxy method on ABI-PRISM 377 DNA sequencer.
PCR using primer pair F1/R1 was expected to produce a 4,000-bp fragment from strains harboring inlGC2DE, a 2,241-bp fragment from strains harboring inlGHE, and no fragment from those harboring inlC2DE or being empty between ascB and dapE, while PCR with primer pair F1/R2 only yielded a 2,241-bp fragment from strains harboring inlGC2DE. For strains being consistently negative using primers F1/R1 and F1/R2, PCR employing primer pair F2/R1 was performed and expected to obtain 1,782-bp fragment from strains harboring inlC2DE (Chen et al., 2009).
Due to considerable nucleotide identities (65–75%) between gadD1, gadD2, and gadD3, three pairs of primers were designed based on the distinguishable regions through sequence comparison.
F, forward; R, reverse.
Determination of the organization of ascB-dapE internalin cluster
Two sets of PCR were conducted using one common upstream-primer (F1, targeting inlG) and two downstream-primers (R1, targeting inlE; R2, targeting inlD) as previously reported (Table 2) (Chen et al., 2009). For strains being consistently negative in the above two PCR sets, another PCR using primers F2 (targeting inlC2) and R1 was performed (Table 2), which only obtained a 1,782-bp fragment from strains harboring inlC2DE (Fig. 1).
Genomic organization of internalin cluster between ascB and dapE in L. monocytogenes. The diversity of gene contents (in dark gray) is delimited by housekeeping genes ascB and dapE (in white). Three ORFs (in light gray) are identified between ascB and inlC2 in inlC2DE profile, encoding hypothetical protein and two homologs of ABC transporter respectively. Arrows indicate orientation of genes.
Detection of glutamate decarboxylase genes
The specificities of PCR targeting GAD genes were verified using L. monocytogenes strains F2365 (lineage I) (Nelson et al., 2004), EGD-e (lineage II) (Glaser et al., 2001), and M7 (lineage III) (Chen et al., 2011b), whose complete genome sequences were known, and using strains belonging to another 14 bacterial genera as positive and negative controls, respectively. Randomly selected amplicons were sequenced to further confirm the accuracy of PCR reactions.
Results and Discussion
All 57 lineage I strains contained inlC2DE cluster, except for one serovar 1/2b strain bearing inlGC2DE. Seventy-two lineage II strains showed three internalin patterns (Fig. 1), with serovar 1/2a strains carrying either inlC2DE (31.5%), inlGC2DE (64.8%), or inlGHE (3.7%), and 1/2c strains possessing inlGHE (83.3%) or inlGC2DE (16.7%). While the majority of lineage III (25/29, 86.2%) and all lineage IV (5/5, 100%) strains lacked internalin cluster in this locus, a small fraction of lineage III strains carried inlGC2DE (4/29, 13.8%). According to the ascB-dapE structure, L. monocytogenes strains could be classified into four subpopulations, with inlC2DE, inlGC2DE, inlGHE, or no-internalin-cluster pattern consisting of 74, 43, 17, and 30 strains, respectively (Table 1). In addition, another five non-monocytogenes Listeria species, which are nonpathogenic to humans, lacked internalin genes in this region.
We then examined the presence of three GAD homologs in these L. monocytogenes strains. In contrast to gadD2 and gadD3 conserved in all L. monocytogenes strains, gadD1 was identified in 60 L. monocytogenes strains (36.6%), including one 1/2b (1/29, 3.4%), 37 serovar 1/2a (37/54, 68.5%), 18 serovar 1/2c (100%), and four lineage III (4/29, 13.8%) strains. All 29 serovar 4b and five lineage IV strains lacked this gene. More strikingly, the presence of gadD1 was highly associated with inlGC2DE and inlGHE patterns (Table 1). Although the strain assembly harboring gadD1 was mainly composed of lineage II strains, occurrence of gadD1-positive serovar 1/2b and lineage III strains indicated that the presence of gadD1 was not fully dependent on lineages or serovars. Notably, the gadD1-positive serovar 1/2b strain appears to branch off from the lineage I main cluster in the cladogram based on nine genomic loci (Chen et al., 2010b), and four gadD1-positive lineage III strains constituted subpopulation IIIA-1 by polyphasic approaches (Zhao et al., 2011). In addition, another five non-monocytogenes Listeria species lacking ascB-dapE internalin cluster were consistently gadD1-negative. Thus, it is possible that gadD1 have co-evolved with the ascB-dapE internalin cluster. The same serovar displays distinct profiles of ascB-dapE region and GAD genes, suggesting that these loci might co-evolve in the common ancestor of current L. monocytogenes lineages before divergence of serovars.
The multigene internalin family is scattered in L. monocytogenes genomes, and play broad roles not merely in adhesion and invasion of host cells (Bierne et al., 2007; Milillo et al., 2009; Markkula et al., 2011). gadD1 forms part of the GAD system and is implicated in listerial tolerance to acid, bile, and bacteriocin (Begley et al., 2002, 2010; Cotter et al., 2005). Genetic diversity of the ascB-dapE internalin cluster and GAD locus might be a consequence of listerial adaption to various environments that present various stress conditions for bacterial growth and survival. The absence of gadD1 from serovar 4b strains, which are responsible for the majority of epidemic listeriosis (Swaminathan et al., 2007), indicates that its presence is not essential for the organism to bring about disease. However, the association between the absence of gadD1 and virulence remains unclear. One previous study indicated that the loss of lysine decarboxylase genes during evolution of Shigella species from Escherichia coli is pathoadaptive (Maurelli et al., 1998). It is intriguing to consider whether the absence of GAD from serovar 4b may, under certain circumstances, enhance pathogenic potential.
In conclusion, the presence of gadD1 correlates with ascB-dapE internalin profiles. Further studies should be carried out to examine the expression and regulation of these genes under different conditions, with the purpose of establishing their potential roles in listerial survival under environmental stresses and pathogenesis within the host.
Footnotes
Acknowledgments
We thank Drs. Martin Wiedmann (Cornell University) and John Bowman (University of Tasmania, Australia) for the kind gift of Listeria strains. We appreciate Drs. Dongyou Liu (Royal College of Pathologists of Australasia Quality Assurance Programs, Australia) and Lingli Jiang (University of Oslo, Norway) for helpful discussions on the subtyping of Listeria strains. This study was supported by grants from National Natural Science Foundation (contract 31101829).
Disclosure Statement
No competing financial interests exist.