Identification of a Cytopathogenic Toxin from Sneathia amnii

Abstract

Sneathia amnii is a poorly characterized emerging pathogen that has been implicated in amnionitis and urethritis. We found that S. amnii damages fetal membranes, and we identified and purified a cytotoxic exotoxin that lyses human red blood cells and damages cells from fetal membranes. The gene appears to be cotranscribed with a second gene that encodes a protein with identity to two-partner system transporters, suggesting that it is the "A," or secreted component of a type Vb system. The toxin is 1,881 amino acids with a molecular weight of approximately 200 kDa. It binds to red blood cell membranes and forms pores with a diameter of 2.0 to 3.0 nm, resulting in osmolysis. Because it appears to be the "A" or passenger component of a two-partner system, we propose to name this novel cytotoxin/hemolysin CptA for cytopathogenic toxin component A.IMPORTANCESneathia amnii is a very poorly characterized emerging pathogen that can affect pregnancy outcome and cause urethritis and other infections. To date, nothing is known about its virulence factors or pathogenesis. We have identified and isolated a cytotoxin, named CptA for cytopathogenic toxin, component A, that is produced by S. amnii CptA is capable of permeabilizing chorionic trophoblasts and lysing human red blood cells and, thus, may play a role in virulence. Except for small domains conserved among two-partner secretion system passenger proteins, the cytotoxin exhibits little amino acid sequence homology to known toxins. In this study, we demonstrate the pore-forming activity of this novel toxin.

Keywords: hemolysin; microbiome; pathogenesis; toxins.

FIG 1

S. amnii damages cells in fetal membranes. (A and B) Sections of human chorion from healthy, term vaginal deliveries were sterilized, and sections were exposed to S. amnii from the maternal side. (A) The tissues were stained with trypan blue, fixed, and sectioned. The maternal face of the chorion, on the top, is indicated with a yellow arrow, and the side normally facing the amnion (amnion is not present) is indicated with a blue arrow. Control membranes, which were not challenged with bacteria, did not contain any trypan blue-stained cells. S. amnii penetrated the membranes and permeabilized chorionic trophoblasts on both the maternal (exposed) side and the side distal to exposure as indicated by trypan blue uptake. (B) Membrane traversal was assessed by recovery of viable bacteria from medium on the fetal side of the chorion. E. coli traversed only one membrane section out of 18 sections from 3 placentas, whereas S. amnii traversed 16 out of 17 sections. (C and D) Primary human amniotic epithelial cells (HAEC) extracted from amnion (C) and JEG-3 chorionic trophoblasts (D) were exposed to S. amnii and stained with trypan blue. (E and F) Viability of the HAEC (E) and JEG-3 (F) was quantified by MTT assay. Percent viability was calculated by subtracting the OD₅₇₀ to OD₆₃₀ of methanol-killed cells (0% viability) from the OD₅₇₀ to OD₆₃₀ of the test values and dividing these values by that of the PBS control (100% viability). Student's t test was performed for PBS versus S. amnii for both HAEC and JEG-3; P < 0.0001.

FIG 2

S. amnii displays hemolytic activity. Bacteria were incubated with washed RBC at the MOI shown for 1 h. S. amnii lysed human and bovine RBC, but there was little lysis of horse, sheep, and rabbit RBC even at the highest MOI tested. Lysis was measured by the amount of liberated hemoglobin detected by OD₄₀₅ (RBC treated with 0.1% Tween 20 served as a positive control for 100% lysis).

FIG 3

Schematic of cptB and cptA gene locus and CptA protein. The cptB and cptA genes are proximal on the S. amnii Sn35 chromosome and are separated by only 9 nucleotides. The genes and their locus tags are shown. Full-length CptA is 1,881 amino acids. The purple box is the predicted Sec signal peptide (probability = 0.97 by SignalP 5.0). The predicted cleavage site is between A₂₃ and K₂₄. The orange box indicates a region with homology to the domain involved in the export of Fha through its autotransporter. The yellow box represents a repeat region (1,103 to 1,864) predicted by RADAR. The only cysteine in the gene (C_1,006) is indicated.

FIG 4

CptA antiserum blocks S. amnii-mediated hemolysis. S. amnii was preincubated with either preimmune serum (Pre) or CptA-specific antiserum (α-CptA) prior to addition to RBC. The bacteria were incubated with the RBC at an MOI of 10 bacteria/1 RBC for 1 h. CptA antiserum completely blocked the hemolytic activity. Statistical analysis was completed as follows: one-way ANOVA, F = 716.4, P < 0.001; Tukey’s HSD, P < 0.001 for PBS versus S. amnii, PBS versus S. amnii plus preimmune serum, S. amnii versus S. amnii plus anti-CptA, and S. amnii plus anti-CptA versus S. amnii plus preimmune serum. Tukey’s HSD was not significant for PBS versus S. amnii plus anti-CptA and S. amnii versus S. amnii plus preimmune serum.

FIG 5

Purified CptA and rCptA. Denaturing PAGE (A) and Western blot (B) with anti-CptA rabbit antiserum used as the primary antibody. (Left to right) BLUEstain protein ladder, 15 μg BME-EDTA extract from S. amnii, 0.1 μg CptA (∼0.5 pmol) purified by cation exchange chromatography. The antiserum recognized a number of peptides in the extract ranging in size from ∼72 to 250 kDa. Cation exchange chromatography enriched the 250-kDa peptide. (C) Denaturing PAGE analysis of (left to right) BLUEstain protein ladder, 0.1 μg (∼2.5 pmol) purified Nterm, and 0.5 μg (∼2.5 pmol) purified rCptA.

FIG 6

CptA antiserum blocks CptA-mediated hemolysis and cytotoxicity. Native CptA, CptA preincubated with preimmune serum (Pre), CptA preincubated with anti-CptA, recombinant CptA (rCptA), crude lysate from E. coli containing the Nterm expression construct, or purified Nterm was added to human RBC or JEG-3 cells. Percent hemolysis of human red blood cells measured by spectrophotometric quantification of hemoglobin release at 405 nm (A) or viability of JEG-3 chorionic trophoblasts assessed by MTT assay (B) and trypan blue staining (C). CptA antiserum completely blocked hemolytic and cytotoxic activity. Statistical analysis for hemolysis was as follows: one-way ANOVA, F = 712.7, P < 0.001; Tukey’s HSD, P < 0.001 for PBS versus CptA, PBS versus CptA plus preimmune serum, CptA versus CptA plus anti-CptA, CptA plus anti-CptA versus CptA plus preimmune serum, PBS versus rCptA, rCptA versus lysate, and rCptA versus Nterm. Hemolysis induced by rCptA was significantly higher than that induced by native CptA (P < 0.0025). Tukey’s HSD was not significant for PBS versus CptA plus anti-CptA, CptA versus CptA plus preimmune serum, PBS versus crude Nterm lysate, and PBS versus purified Nterm. Statistical analysis for JEG-3 viability was as follows: one-way ANOVA, F = 103.49; P < 0.001; Tukey’s HSD, P <0.001 for PBS versus CptA, PBS versus CptA plus preimmune serum, CptA versus CptA plus anti-CptA, CptA plus anti-CptA versus CptA plus preimmune serum, PBS versus rCptA, rCptA versus lysate, and rCptA versus Nterm. Tukey’s HSD was not significant for PBS versus CptA plus anti-CptA, CptA versus CptA plus preimmune serum, PBS versus crude Nterm lysate, PBS versus purified Nterm, and CptA versus rCptA.

FIG 7

CptA forms pores between 2.0 and 2.8 nm. Polyethylene glycols ranging in molecular weight from 500 to 4,000 g/mol were added to RBC treated with VLY or CptA. VLY-mediated hemolysis was not affected by the PEGs. CptA-mediated hemolysis was partially inhibited by PEG 1000 and nearly completely inhibited by PEG 2000, suggesting that CptA forms pores of defined diameter between these hydrodynamic diameters.