Antibiotic-induced gut dysbiosis enhances pathogen-mediated coagulopathy via shifts in quorum sensing and microbial metabolites

Antibiotic-induced gut dysbiosis enhances pathogen-mediated coagulopathy via shifts in quorum sensing and microbial metabolites

Authors:
Jessica Cao, Rebecca Meltzer, Andrew Benjamin, John Alverdy, Robert Keskey

Body of Abstract:
Background: Critically ill trauma patients are exposed to selective pressures including antibiotics that disrupt the gut microbiota, favoring expansion of Proteobacteria including pathogens like Pseudomonas aeruginosa (Pa). Intestinal dysbiosis in this setting has been linked to adverse outcomes including bleeding complications. We recently demonstrated that gut-derived Pa can induce clinically significant coagulopathy via secreted exoproducts in humans and mice. However, the extent to which Pa virulence expression–and thus its ability to drive coagulopathy–is modulated by the gut environment following antibiotic exposure remains unknown. We hypothesize that antibiotic-induced changes in the intestinal milieu (i.e. higher levels of the interspecies quorum-sensing molecule autoinducer-2 [AI-2], depletion of gut-derived metabolites) increases Pa virulence and enhances its ability to induce coagulopathy.

Methods: Pa isolated from an ICU patient was grown in tryptic soy broth (TSB) and TSB supplemented with the following: cecal metabolites from antibiotic-decontaminated mice (AbxCec); supernatant of E. Coli that constitutively expresses AI-2 (Ec-AI2), supernatant of a control E. Coli strain (Ec-Ctrl); 100 mM acetate (Ace); 10 mM butyrate (But); 10 mM propionate (Prop); 1 mM kynurenic acid (Kyn); 1 uM indole-3-propionate (IPA); and 50 mM tryptophan (Trp). Supernatant from Pa grown in each of these conditions (Pa-Sup) was collected at stationary phase at a similar optical density, filtered, and added to citrate-anticoagulated whole blood from healthy volunteers. Coagulation was analyzed using TEG6S. 

Results: Growth in TSB supplemented with AbxCec increased Pa-induced coagulopathy in a dose-dependent manner. Compared to TSB Pa-Sup, AbxCec Pa-Sup prolonged R time when added to blood at 1:15 (54.18s vs 4.64s; p<0.001) and decreased maximum amplitude (MA) at 1:20 (18.74 vs 57.76; p<0.001) and 1:15 (2.1 vs 43.46; p<0.01). Ec-AI2 Pa-Sup decreased MA to a greater extent than Ec-Ctrl Pa-sup and TSB Pa-Sup at 1:10 (7.52 vs 13.92 vs 29.8; Ec-AI2 vs Ec-Ctrl vs TSB Pa-sups; p<0.05). Ec-AI2 Pa-Sup also prolonged R time to a greater extent (31.1s vs 18.6s vs 3.88s; NS), but this was not statistically significant. Growing Pa in the presence of Ace, Kyn, IPA, and Trp suppressed Pa-Sup’s ability to decrease MA when added to blood at 1:10 (18.53 vs 59.53 vs 43.23 vs 53.2 vs 52.6; TSB vs Ace vs Kyn vs IPA vs Trp Pa-Sups; p<0.05), while growing Pa with But and Prop did not have this effect. Conclusion: By shifting the balance of quorum-sensing signals (i.e. AI-2) and gut-derived metabolites (i.e. SCFAs, Trp derivatives), antibiotics create a metabolic niche that favors Pa predominance and potentiates its virulence expression in a way that amplifies pathogen-mediated coagulopathy. This underscores the concept that antibiotic-mediated perturbations of the gut microbiota can transform commensal-pathogen dynamics and ultimately hinder recovery in critically ill patients.