Inerolysin, a cholesterol-dependent cytolysin produced by Lactobacillus iners

Abstract

Lactobacillus iners is a common constituent of the human vaginal microbiota. This species was only recently characterized due to its fastidious growth requirements and has been hypothesized to play a role in the pathogenesis of bacterial vaginosis. Here we present the identification and molecular characterization of a protein toxin produced by L. iners. The L. iners genome encodes an open reading frame with significant primary sequence similarity to intermedilysin (ILY; 69.2% similarity) and vaginolysin (VLY; 68.4% similarity), the cholesterol-dependent cytolysins from Streptococcus intermedius and Gardnerella vaginalis, respectively. Clinical isolates of L. iners produce this protein, inerolysin (INY), during growth in vitro, as assessed by Western analysis. INY is a pore-forming toxin that is activated by reducing agents and inhibited by excess cholesterol. It is active across a pH range of 4.5 to 6.0 but is inactive at pH 7.4. At sublytic concentrations, INY activates p38 mitogen-activated protein kinase and allows entry of fluorescent phalloidin into the cytoplasm of epithelial cells. Unlike VLY and ILY, which are human specific, INY is active against cells from a broad range of species. INY represents a new target for studies directed at understanding the role of L. iners in states of health and disease at the vaginal mucosal surface.

FIG. 1.

The genome of L. iners contains an ORF encoding a putative CDC. (A) Detection of the gene for INY in genomic DNA from several strains of L. iners. Numbers are strain identifiers, and MW represents the molecular weight marker. E. coli DNA was included as a negative control. (B) Unrooted phylogeny of CDCs. The topology shown was obtained independently by the MP, ML, and NJ techniques. The only topological differences noted were within the INY clade, where the sequences are so similar that there is little support and scarce phylogenetic signal to separate one INY lineage from another. The MP tree was 4,530 steps long. The ML tree had a score of −log L of 17,974.860993. Numbers on branches are bootstrap percentages for the MP and ML techniques (MP/ML). All bootstrap percentages are 100/100 unless otherwise indicated. Branch lengths are based on the ML tree.

FIG. 2.

Production of rINY and detection of INY in supernatants of L. iners. (A) Purification of an approximately 55-kDa protein (rINY) from E. coli. Lane MW contained protein molecular mass standards. (B) INY and PLY (10 μg/lane) were detected by an anti-PLY monoclonal antibody which recognizes the undecapeptide region at the C terminus of the protein. (C) Western blotting of concentrated supernatants from growth of several distinct isolates of L. iners (numbers are strain identifiers). INY was detected by an anti-PLY monoclonal antibody. Concentrated E. coli supernatant was included as a negative control. The values to the left of the panels are molecular masses in kilodaltons.

FIG. 3.

Non-species-restricted hemolytic activity of INY. Lysis of murine, human, and ovine erythrocytes by (A) INY, (B) ILY, and (C) PLY. (D) Polyclonal anti-VLY antibody inhibits the activity of INY (1.5 μg/ml) on ovine erythrocytes. *, P < 0.05; **, P < 0.001.

FIG. 4.

Cholesterol-dependent hemolytic activity of INY. (A) INY (1.25 μg/ml) was incubated with ovine erythrocytes in the presence of the indicated concentrations of cholesterol or a vehicle control. P values for this and subsequent figures: *, P < 0.05; **, P < 0.001; ***, P < 0.0001. NS, no statistically significant difference. (B) Kinetic assay of ovine erythrocyte lysis by rINY (1.25 μg/ml) in the presence of the indicated concentrations (μg/ml) of cholesterol or a vehicle control. Hemolysis was measured by determining OD₇₀₀ every 3 min for 1 h.

FIG. 5.

INY is a thiol-activated cytolysin. (A) INY (200 ng/ml) was incubated in the presence of various concentrations of DTT or a vehicle control for 5 min at room temperature. This toxin was then used in endpoint (A) and kinetic (B) hemolysis assays.

FIG. 6.

pH-dependent activity of INY. (A) INY, ILY, and VLY (125 ng/ml) were incubated with human erythrocytes in an endpoint hemolysis assay under various pH conditions at 37°C. (B) INY and LLO (125 ng/ml) were preincubated at the indicated temperature for 20 min before use in an erythrocyte lysis assay at the indicated temperature. The hydrophobicities of LLO (C) and INY (D) were determined over time with the fluorescent probe ANS. A 1 μM concentration of each toxin was injected into buffer containing 50 μM ANS 100 s after data acquisition started.

FIG. 7.

INY induces epithelial cell lysis at high concentrations and activates proinflammatory signaling at low concentrations. HeLa (A) or COS-7 (B) cells were incubated with increasing concentrations of INY, and epithelial cell lysis was measured by LDH release. Values were normalized to 100% lysis using Triton X-100. HeLa (C) or COS-7 (D) cells were incubated with medium alone (negative) or INY (3 μg/ml) for 30, 60, or 90 min. Cells were lysed and probed with phospho-p38 antibody (top) and total p38 antibody (bottom). HeLa cells were incubated with PLY as a positive control. (E) HeLa cells were incubated with fluorescently labeled phalloidin (final concentration of 3.3 nM) in the presence or absence of INY (final concentration of 1.8 μg/ml). Internalized phalloidin was detected by fluorescence microscopy. Scale bars represent 50 μm.