Rifaximin ameliorates cirrhotic portal hypertension through suppression of microbiome-derived deoxycholic acid

Microbiome-derived deoxycholic acid (DCA) elevates serum 5-hydroxytryptamine (5-HT), a mediator of portal hypertension (PH). Rifaximin, a non-absorbable antibiotic, is known to reduce DCA levels. We aimed to elucidate the role of DCA in cirrhotic PH and evaluate the therapeutic potential of rifaximin.

PH was induced in mice by thioacetamide (TAA) injection or bile duct ligation (BDL). Mice were treated with antibiotics (ABX) or rifaximin, with or without exogenous DCA supplementation. A cohort of 51 patients with cirrhosis and 19 healthy controls was analyzed to validate correlations among DCA, 5-HT, and hepatic venous pressure gradient (HVPG). Mice with tissue-specific knockout of gut epithelial Tph1 (Tph1^VKO), vascular smooth muscle cell Htr1a (Htr1a^ΔVSMC), or Kcnj9 (Kcnj9^ΔVSMC) were used for mechanistic studies.

Fecal DCA positively correlated with portal pressure (PP) in TAA-induced PH mice (r = 0.631, p <0.001) and with HVPG in patients (r = 0.5874, p <0.001). ABX treatment reduced fecal DCA, serum 5-HT, and PP in TAA- or BDL-induced PH mice. Exogenous DCA reversed the ABX-induced reductions in serum 5-HT and PP, an effect abolished in Tph1^VKO mice. GIRK3 (encoded by Kcnj9) was upregulated in portal veins from PH mice and patients. VSMC-specific Kcnj9 deletion attenuated PH and prevented 5-HT-induced PP elevation. Mechanistically, 5-HT triggered portal vein smooth muscle cell contraction via the HTR1A-GIRK3-Ca²⁺-MLC2 pathway. Rifaximin alleviated PH by reducing DCA in wild-type mice but showed no additional PP reduction in Tph1^VKO, Htr1a^ΔVSMC, or Kcnj9^ΔVSMC mice.

Microbiome-derived DCA exacerbates PH by enhancing TPH1-dependent 5-HT biosynthesis, which activates portal vein smooth muscle cell contraction via HTR1A-GIRK3 signaling. Rifaximin alleviates cirrhotic PH by reducing DCA levels, highlighting a potential therapeutic strategy for clinical PH management.

Portal hypertension is a key driver of cirrhosis-related complications, yet current therapies exhibit suboptimal efficacy. Herein, we elucidate the role of the gut microbial metabolite deoxycholic acid in the pathophysiology of cirrhotic portal hypertension through the TPH1-5-HT/HTR1A-GIRK3 axis and provide preclinical evidence that rifaximin ameliorates cirrhotic portal hypertension by reducing deoxycholic acid. These insights may open a new avenue for the clinical management of cirrhotic portal hypertension.