The intracellular bacterium Listeria monocytogenes can invade several types of normally non-phagocytic cells. Entry into cultured epithelial cells requires the expression of inIA, the first gene of an operon, comprising two genes: inIA, which encodes internalin, an 800-amino-acid protein, and inIB, which encodes a 630-amino-acid protein. Several genes homologous to inIA are detected in the genome of L. monocytogenes; InIB is one of them. We have assessed the role of inIB in invasiveness of L. monocytogenes by constructing isogenic chromosomal deletion mutants in the inIAB locus. Our findings indicate that: i) inIB is required for entry of L. monocytogenes into hepatocytes, but not into intestinal epithelial cells; ii) inIB encodes a surface protein; iii) internalin plays a role for entry into some hepatocyte cell lines. These results provide the first insight into the cell tropism displayed by L. monocytogenes.

Human E-cadherin promotes entry of the bacterial pathogen Listeria monocytogenes into mammalian cells by interacting with internalin (InlA), a bacterial surface protein. Here we show that mouse E-cadherin, although very similar to human E-cadherin (85% identity), is not a receptor for internalin. By a series of domainswapping and mutagenesis experiments, we identify Pro16 of E-cadherin as a residue critical for specificity: a Pro→Glu substitution in human E-cadherin totally abrogates interaction, whereas a Glu→Pro substitution in mouse E-cadherin results in a complete gain of function. A correlation between cell permissivity and the nature of residue 16 in E-cadherins from several species is established. The location of this key specificity residue in a region of E-cadherin not involved in cellcell adhesion and the stringency of the interaction demonstrated here have important consequences not only for the understanding of internalin function but also for the choice of the animal model to be used to study human listeriosis: mouse, albeit previously widely used, and rat appear as inappropriate animal models to study all aspects of human listeriosis, as opposed to guinea-pig, which now stands as a small animal of choice for future in vivo studies.

SummaryStreptococcus agalactiae [group B streptococcus (GBS)] is the leading cause of neonatal pneumonia, sepsis and meningitis. An in silico genome analysis indicated that GBS strain NEM316 encodes five putative sortases, including the major class A sortase enzyme and four class C sortases. The genes encoding the class C sortases are tandemly arranged in two different loci, srtC1-C2 and srtC3-C4 , with a similar genetic organization and are thought to be involved in pilus biosynthesis. Each pair of sortase genes is flanked by LPXTG protein encoding genes, two upstream and one downstream, and a divergently transcribed regulatory gene located upstream from this locus. We demonstrated that strain NEM316 expresses only the srtC3-C4 locus, which encodes three surface proteins (Gbs1474, Gbs1477 and Gbs1478) that polymerize to form appendages resembling pili. Structural and functional analysis of this locus revealed that: (i) the transcriptional activator RogB is required for expression of the srtC3-C4 operon; (ii) Gbs1477, and either SrtC3 or SrtC4 are absolutely required for pilus biogenesis; and (iii) GBS NEM316 pili are composed of three surface proteins, Gbs1477, the bona fide pilin which is the major component, Gbs1474, a minor associated component, and Gbs1478, a pilus-associated adhesin. Surprisingly, pilus-like structures can be formed in the absence of the two minor components, i.e. the putative anchor Gbs1474 or the adhesin Gbs1478. Adherence assays showed that Gbs1478 confers adhesive capacity to the pilus. This study provides the first evidence that adhesive pili are also present in Grampositive pathogens.

SummaryListeria monocytogenes is an intracellular bacterial pathogen that expresses several surface proteins critical for the infectious process. Such proteins include InlA (internalin) and InlB, involved in bacterial entry into the host cell, and ActA, required for bacterially induced actin-based motility. Although the molecular mechanisms of attachment of InlA and ActA have been characterized, essentially nothing is known about how InlB is anchored to the bacterial surface. Using a genetic approach, we demonstrate that the last 232 amino acids of InlB are both necessary and sufficient for anchoring this protein to the bacterial surface. An InlB mutant protein deleted for the last 232 amino acids was secreted and not detected at the cell surface. A 'domain-swapping' strategy in which these 232 amino acids were used to replace the normal cell wall-anchoring domain of InlA resulted in a chimeric protein that was anchored to the cell surface and able to confer entry. Interestingly, surface association of InlB also occurred when InlB was added externally to bacteria, suggesting that association may be able to occur after secretion. This association was productive for invasion, as it conferred bacterial entry into host cells. The C-terminal anchoring region in InlB contains 80-amino-acid repeats beginning with the sequence GW that is also present in a newly identified surface-associated bacteriolysin of L. monocytogenes, called Ami. Addition of GW repeats to the C-terminal of InlB improves anchoring of the protein to the cell surface. These and other data suggest that such 'GW' repeats may constitute a novel motif for cell-surface anchoring in Listeria and other Gram-positive bacteria. This motif may have important consequences for the release of surface proteins involved in interactions with eukaryotic cells.