(P121) DEFINING H2S EFFECTS USING SELECTIVE CYSTEINE UPTAKE MUTANT OF CROHN'S DISEASE DERIVED ADHERENT INVASIVE E. COLI

Background:  Mucus-digesting Ruminococcus gnavus (Rg), adherent-invasive E. coli (AIEC) and fecal hydrogen sulfide (H₂S) are increased in active Crohn’s disease (CD) vs. healthy subjects. We established a link between Rg and ileal CD AIEC strain LF82 by showing that dual colonization of ex-germ-free (GF) IL10^-/- mice potentiated colitis and cecal H₂S production. Rg precultured mucus increased LF82 growth and H₂S production, H₂S supported in vitro Rg survival in minimal medium, and H₂S enhanced LF82 adhesion and invasion both in vitro and in vivo (DDW 2023). Production of H₂S is selectively enhanced by the sulfur substrate L-cysteine, a mucus component that crosslinks mucus polymers. We hypothesize that cysteine is released by R. gnavus digestion of mucus to supply a substrate for AIEC H₂S synthesis. While H₂S is increased in active CD and UC and can interfere with epithelial mitochondrial energy production, evidence of its direct effect on inducing or potentiating colitis is lacking. Showing that H₂S directly drives colitis could guide new treatments or diagnostic tools targeting H₂S.

Hypothesis:  H₂S contributes to colitis, tested by colonizing gnotobiotic Il10^-/- mice with R. gnavus and wild type (WT) AIEC or L-cysteine transporter and desulfidase gene deletion (mutant) strains.

Methods:  We have generated an AIEC NRG857c mutant that is unable to take up cysteine. We measured H₂S production by each strain with 0.3 mM cysteine added to minimum media. Germ-free IL10KO mice were colonized with R. gnavus and either wild-type or mutant strains (R+WT, R+Mut) for 7 weeks and weekly fecal lipocalin-2 levels were measured by ELISA. Colorimetric H₂S assays were performed in freshly harvested cecal contents and bacterial culture.    

Results:  The NRG857c mutant produced less H₂S compared to WT in M9 minimal medium supplemented with L-cysteine (Fig.1A). Weekly fecal lipocalin-2 levels were not different between R+WT and R+Mut mice (Fig.1B). R+Mut induced less inflammation by cecal and distal colonic histology scores and produced less H₂S compared to R+WT (Fig.1C, D). However, 3 out of 8 R+Mut- colonized mice had high H₂S production. Some randomly chosen isolates from those samples showed spreadable colony morphology and high H₂S production compared to low H₂S producing mutant strains in aerobic and anaerobic conditions in vitro (Fig. 2).


Conclusions:  Less colitis in dual colonized mice with the cysteine transporter and desulfidase deletion NRG857c mutant and R. gnavus suggested that H₂S production by AIEC contributes to colitis. Variable in vivo results likely resulted from high H₂S producing variant mutant strains. We will repeat the experiment with the low H₂S variants to directly determine if H₂S causes colonic injury. Exploring genetic mechanisms of the variant strains may provide better understanding of AIEC H₂S production pathways.