Listeria seeligeri Isolates from Food Processing Environments Form Two Phylogenetic Lineages

Abstract

Seven different actA subtypes forming two phylogenetic lineages could be distinguished by sequencing the actA gene of Listeria seeligeri isolates from different habitats. Isolates of the two lineages differ in hemolytic as well as phospholipase activities and in the arrangement of the virulence gene cluster. The presence of a serine protease gene resembling orf2110 of L. monocytogenes in some isolates further supports the hypothesis that L. seeligeri is subject to ongoing adaptation to changing environments.

The genus Listeria comprises the species L. monocytogenes, L. ivanovii, L. seeligeri, L. innocua, L. welshimeri, L. grayi, and L. marthii (4, 7, 17). Of these only, L. monocytogenes (15) and L. ivanovii (1, 18) are considered as pathogens. The pathogenicity is closely associated with a virulence gene cluster, although other genes like those coding for internalines are implicated in pathogenesis too (16). Like L. monocytogenes and L. ivanovii, L. seeligeri also carries a virulence gene cluster. Although genes of the virulence cluster are presumably not expressed in a correct or functional fashion (3, 20), L. seeligeri has been associated with human infection in one case so far (14). Recent studies reveal exceptions from the traditional classification into pathogenic and nonpathogenic Listeria species. L. innocua strains carrying a virulence gene cluster similar to that of L. monocytogenes have been isolated (6, 22). Moreover, nonhemolytic L. seeligeri isolates lacking the major part of the virulence gene cluster have been described (21).

L. seeligeri variants.

In this study, a total of 2,492 samples from different habitats of the food chain were investigated with respect to the occurrence of members of the genus Listeria. Of 715 isolates obtained from animals (organs and feces), butcher shops (drains and intermediate products), or humans (feces), a total of 44 could be affiliated with L. seeligeri (Table 1). The L. seeligeri isolates were characterized by sequencing the actA gene, which had been used previously for subtyping of L. monocytogenes (2, 10, 23, 24). Sequencing of 253 bp of the L. seeligeri actA gene with the primers actaf and actar of Volokhov et al. (21) disclosed seven different ActA types (Fig. 1) grouped in two phylogenetic lineages (Fig. 2). While two independent isolates originating from locations not interconnected and 50 km apart are grouped in lineage I, the other 42 L. seeligeri isolates from this study are grouped in lineage II. The two lineages can be distinguished also in the case of other genes of the virulence gene cluster (e.g., the seeligeriolysin gene hly) and genes not affiliated with the virulence gene cluster (e.g., the 16S rRNA coding genes, the iap gene, and the putative protease gene orf2110). Isolates of lineage I exhibit higher hemolytic and phospholipase activities on selective agars than those of lineage II (Fig. 3). ActA sequences of lineage II exhibit a high variability on the protein level (from 79.0% to 99.5% identity), resulting in six different actA subtypes (Fig. 1). Notable differences are an insert of seven additional amino acids in subtypes 1 and 2, a leucine-proline repeat in subtypes 5 to 7, and a stop codon in the actA gene in subtype 4.

TABLE 1.

Occurrence of Listeria spp. in organs of animals, feces of humans and animals, and butcher shops^a in Upper Franconia, Germany

Origin (n)	% of samples positive for:
	Listeria spp.	L. monocytogenes	L. innocua	L. welshimeri	L. seeligeri
Animal organs (63)	11.1	6.3	1.6	0.0	3.2
Animal feces (31)	9.7	9.7	0.0	0.0	0.0
Drain (522)	41.8	27.2	14.0	2.5	3.4
Sausage meat (789)	59.1	30.4	23.8	10.4	3.0
Human feces (1,087)	1.9	1.1	0.7	0.4	0.0

^aDrains and sausage meat.

FIG. 1.

Hypothetical protein sequences of the seven identified ActA subtypes of L. seeligeri (Ls). The protein sequences correspond to the hypothetical translation products of the 253-bp actA gene fragments (Fig. 2). For every subtype, an example is given. Alignments were done with the program BioEdit. The asterisk within the type 4 sequence denominates a stop codon. Gaps are indicated by the “∼” symbol.

FIG. 2.

Dendrogram showing phylogenetic relationship among L. seeligeri (Ls) isolates based on partial actA sequences (253 bp) from L. seeligeri isolates obtained in Upper Franconia, Germany, and reference strains L. seeligeri ATCC 35967 (AY878348) and L. seeligeri NRRL (AY510074). The tree was constructed by the minimum evolution method in the MEGA version 4 package. The bootstrap values presented at corresponding branches were evaluated from 1,000 replications. The phylogenetic lineages of the strains are symbolized with Roman numerals and the corresponding actA subtypes with Arabic numerals. The bar indicates 1.0% sequence divergence. Strains marked with an asterisk carried a gene similar to the orf2110 gene of L. monocytogenes F2365 (AE017262).

FIG. 3.

Phosphatidylinositol-specific phospholipase activity on ALOA agar plates (Merck, Darmstadt, Germany). For analysis of the phospholipase activity, 10 μl of a overnight culture in brain heart infusion (BHI) liquid medium of L. monocytogenes 1 (A), L. seeligeri 90 (B), and L. seeligeri 187 (C) was dropped onto ALOA agar plates and incubated at 37°C for 48 h.

Virulence clusters of L. seeligeri lineages I and II.

Isolates L. seeligeri 90 and L. seeligeri 187, representing lineage I and lineage II, respectively, exhibit an identical arrangement of the virulence gene cluster with respect to gene organization. Both lineages also show high homology in the housekeeping genes prs and ldh (100% at the protein level), while other virulence gene products vary with respect to their identity from 81.3% to 99.1% (Table 2). The most significant differences between the lineages are observed with the dplcB and orfY genes. While the dplcB gene of lineage II represents a truncated form of the plcB gene, as described previously (3), the dplcB gene of lineage I includes 534 additional nucleotides, resembling the full-length adjacent plcB gene. The hypothetical translation product, a preproenzyme, shows an identity of only 48.5% to the PlcB phospholipase of the same strain. It differs, however, in five out of nine zinc ion coordinating amino acids from PlcB. This would result in a nonfunctional enzyme, as an exchange of Trp1 to Leu1 would impair the binding of a zinc ion at site 1.

TABLE 2.

Similarities and identities between gene and protein sequences of L. seeligeri 90 and 187 and L. monocytogenes F2365

Lineages compared and comparison parameter	% similarity for comparisona
	prsb	orfC	orfD	prfA	orfE	plcA	hly	orfK	mpl	actA	dplcB	plcB	orfY	orfX	orfI	orfP	orfB	orfA	ldhb	iapb,c
L. seeligeri 90 and 187
Similarity (nt)	95.9	95.1	93.7	97.8	96.1	96.3	97.5	93.3	94.2	87.8	34.9	95.1	82.1	98.2	96.8	95.2	100.0	92.7	97.0	95.6
Identity (aa)	100.0	95.0	92.7	99.1	95.5	96.2	98.3	89.2	92.5	81.3	ND	93.5	80.3	98.2	95.3	96.6	100.0	93.2	100.0	95.3
L. seeligeri 187 and L. monocytogenes F2365
Similarity (nt)	85.4			75.9		65.9	75.7		68.5	47.0		68.3							85.7	39.6
Identity (aa)	95.0			73.8		63.4	81.1		64.3	21.4		62.3							97.0	39.8
L. seeligeri 90 and L. monocytogenes F2365
Similarity (nt)	87.9			76.4		65.3	75.7		69.7	46.4		68.3							85.7	39.4
Identity (aa)	95.0			73.8		63.7	81.5		65.2	20.8		61.6							97.0	40.2

^aNucleotide (nt) similarities and amino acid (aa) identities between gene and protein sequences of the two lineages of L. seeligeri isolates 90 (lineage I; GQ862951) and 187 (lineage II; GQ862952) and L. monocytogenes F2365 (AE017262). Alignments were done with the program BioEdit. ND, not determined.

^bPartial sequence.

^cThe iap gene encodes the invasion-associated protein P60.

The serine protease gene orf2110.

The orf2110 gene encodes a hypothetical serine protease carrying a V8-like Glu-specific endopeptidase domain (COG3591) (13) as well as three Chw (Clostridium hydrophobic tryptophan) domains (12). The orf2110 gene has been considered specific for L. monocytogenes serotypes 4b, 4d, and 4e (7). In this study, a gene showing high similarity to the orf2110 gene of L. monocytogenes (95.6% to lmof2365_1900 in AAT04669) (11) and to a lesser extent to the analogous gene in L. welshimeri (80.2% similarity; lwe1890 in CAK21308) (5) is described for the first time in L. seeligeri. orf2110 is always located between the genes encoding a hypothetical alkaline protease (lmof2365_1899 in L. monocytogenes F2365 and lwe1889 in L. welshimeri SLCC5334) and a phosphomutase family protein (lmof2365_1901 in L. monocytogenes F2365 and lwe1891 in L. welshimeri SLCC5334). The flanking genes are also present in strains without an orf2110 gene, like in L. innocua (e.g., lin1984 and lin1985 in AL592022 L. innocua strain Clip 11262) (8) and some L. seeligeri strains (Fig. 2). The orf2110 gene is present in the isolates of lineage I as well as in some isolates of lineage II, primarily in ActA types 4 and 5 (Fig. 2). It is noteworthy that the orf2110 genes of L. seeligeri resemble more closely the genes from L. monocytogenes, whereas the 16S and 23S rRNA coding genes as well as the iap gene are phylogenetically related to L. welshimeri (16).

Conclusions.

It has been suggested that L. seeligeri carries an ancestral form of the virulence gene cluster, which has been deleted from the chromosome of L. innocua and L. welshimeri (8) and some L. seeligeri isolates (21). This cluster would have been subsequently optimized by “pathoadaptive mutations” for the infection and survival in mammals during the evolution of L. monocytogenes and L. ivanovii (19). It has been hypothesized that the virulence gene cluster of L. seeligeri and its low expression are adequate for survival in bacteriovorous protozoae like Tetrahymena (9). However, our results suggest that L. seeligeri is subject to ongoing adaptation to changing environments. Accordingly, lineage II appears to be better adapted to natural environments, as indicated for example by the deletion in the dplcB gene, the stop codon within the actA gene, or the loss of the entire virulence gene cluster (21). On the other hand, lineage I might be better adapted to the mammalian hosts, as indicated by higher hydrolytic activities. The concomitant deletion in the orfY gene might also contribute to survival in mammalian hosts as this gene is lacking in L. monocytogenes too.