Rifaximin alters intestinal bacteria and prevents stress-induced gut inflammation and visceral hyperalgesia in rats

Abstract

Background & aims: Rifaximin is used to treat patients with functional gastrointestinal disorders, but little is known about its therapeutic mechanism. We propose that rifaximin modulates the ileal bacterial community, reduces subclinical inflammation of the intestinal mucosa, and improves gut barrier function to reduce visceral hypersensitivity.

Methods: We induced visceral hyperalgesia in rats, via chronic water avoidance or repeat restraint stressors, and investigated whether rifaximin altered the gut microbiota, prevented intestinal inflammation, and improved gut barrier function. Quantitative polymerase chain reaction (PCR) and 454 pyrosequencing were used to analyze bacterial 16S ribosomal RNA in ileal contents from the rats. Reverse transcription, immunoblot, and histologic analyses were used to evaluate levels of cytokines, the tight junction protein occludin, and mucosal inflammation, respectively. Intestinal permeability and rectal sensitivity were measured.

Results: Water avoidance and repeat restraint stress each led to visceral hyperalgesia, accompanied by mucosal inflammation and impaired mucosal barrier function. Oral rifaximin altered the composition of bacterial communities in the ileum (Lactobacillus species became the most abundant) and prevented mucosal inflammation, impairment to intestinal barrier function, and visceral hyperalgesia in response to chronic stress. Neomycin also changed the composition of the ileal bacterial community (Proteobacteria became the most abundant species). Neomycin did not prevent intestinal inflammation or induction of visceral hyperalgesia induced by water avoidance stress.

Conclusions: Rifaximin alters the bacterial population in the ileum of rats, leading to a relative abundance of Lactobacillus. These changes prevent intestinal abnormalities and visceral hyperalgesia in response to chronic psychological stress.

Keywords: Antibiotic; CRD; EMG; Gut Flora; IBS; IFN-gamma; IL; Intestine; Irritable Bowel Syndrome; PCR; PXR; RF; TNF-α; VMR; WAS; colorectal distention; electromyographic; interferon-gamma; interleukin; irritable bowel syndrome; polymerase chain reaction; pregnane X receptor; rRNA; ribosomal RNA; rifaximin; tumor necrosis factor−α; visceromotor response; water avoidance stress.

Figure 1

Effect of chronic stress and antibiotics on VMR to CRD. (A) EMG amplitude expressed as mean change from baseline in rats subjected to chronic WAS or sham WAS. (B) EMG amplitude expressed as mean change from baseline after chronic treatment with vehicle (water) or rifaximin (150 mg/kg, twice daily, oral gavage) in rats previously subjected to chronic WAS or sham WAS. (C) EMG amplitude expressed as mean change from baseline after chronic treatment with vehicle or neomycin (150 mg/kg, twice daily, oral gavage) in rats previously subjected to chronic WAS or sham WAS. n = 8–9. *P < .05, compared with the sham WAS, or sham WAS + water gavage controls; ^#P < .05, compared with WAS + water gavage rats. (D) EMG amplitude expressed as mean change from baseline after chronic treatment with vehicle (water) or rifaximin (150 mg/kg, twice daily, oral gavage) in rats previously subjected to repeat RS or sham RS. n = 6. *P < .05, compared with sham RS + water gavage controls; ^#P < .05, compared with RS + water gavage rats.

Figure 2

Comparison of bacterial community composition and diversity, and total bacterial load in the ileal contents of rats previously subjected to WAS or sham WAS. (A) Phylotypes classified to the level of family in order of rank abundance as a percentage of the total community. (B) Diversity (Shannon diversity index and Shannon evenness) of bacterial communities. (C) Realtime qPCR analysis of total bacterial 16s rRNA gene copies (total bacterial load). n = 9. For all figures, *significantly different than sham WAS (P < .05).

Figure 3

Effect of chronic stress and antibiotics on inflammatory cytokine expression in ileal tissue and gut permeability. (A, B, C, D) Cytokine mRNA levels measured with real-time qRT-PCR. Data represent fold change in target mRNA levels relative to expression in control samples after normalization to GAPDH. (A) IL-17, IL-6, IL-10, TNF-α, IFN-γ, and IL-1β mRNA levels in rats previously subjected to chronic WAS or sham WAS. (B) IL-17, IL-6, and TNF-α mRNA levels after chronic treatment with vehicle (water) or rifaximin (150 mg/kg, twice daily, oral gavage) in rats previously subjected to chronic WAS or sham WAS. (C) IL-17, IL-6, TNF-α, IL-10, IFN-γ, and IL-1β mRNA levels in rats previously subjected to repeat RS or sham RS. Note: rifaximin (150 mg/kg twice daily) significantly reduced the increase in IL-17, IL-6 and TNF-α mRNA evoked by chronic stress. Rifaximin also normalized the reduced IL-10 mRNA levels observed in repeat RS rats. (D) IL-17, IL-6, and TNF-α mRNA levels after chronic treatment with vehicle or neomycin (150 mg/kg, twice daily, oral gavage) in rats previously subjected to chronic WAS or sham WAS. n = 8-9. *P < .05, compared with sham WAS or sham RS. P < .05, compared with WAS or RS + water gavage).

Figure 4

Effect of chronic treatment with rifaximin (150 mg/kg, twice daily, oral gavage) on total bacterial load and bacterial community composition in ileal contents. (A) Total bacterial 16s rRNA gene copies. n = 8. *P < .05, compared with sham WAS + water gavage controls; P < .05, compared with WAS + water gavage rats. (B) Variation in bacterial community composition in the terminal ileum at phylum and family levels. (C) Relative abundance of selected phylotypes, identified at family level. n = 9 in sham WAS and WAS groups, n = 3 in WAS + rifaximin group. * significantly different than sham WAS; significantly different than WAS (Kruskal–Wallis with Conover–Inman post hoc test for multiple comparisons, P < .05). (D) Bacterial community composition analyzed with NMDS plots using a θ_YC distance matrix of OTU-based data. Circles represent distinct bacterial communities identified in luminal contents of terminal ileum of individual rats. Circles that appear larger are closer in the axis not represented and circles that are smaller are further away in the axis not represented; some circles have been made transparent. For this NMDS plot, stress = 0.09, r² = 0.97.

Figure 5

Effect of chronic treatment with rifaximin (150 mg/kg, twice daily, oral gavage) on gut permeability and ileal tight junction protein expression. (A, B) Gut permeability measured by appearance of FITC-labeled dextran in plasma. n = 8. *P < .05, compared with sham WAS + water gavage controls or sham RS + water gavage controls; ^#P < .05, compared with WAS + water gavage rats or RS + water gavage rats. (C) Densitometric quantification of occludin mRNA normalized to GAPDH band intensity. n = 8. *P < .05, compared with sham WAS + water gavage controls; #P < .05, compared with WAS + water gavage rats. (D) Densitometric quantification of occludin protein normalized to GAPDH band intensity. n = 6. *P < .05, compared with sham WAS + water gavage controls; ^#P < .05, compared with WAS + water gavage rats. (E) Densitometric quantification of occludin mRNA normalized to GAPDH band intensity, n = 6. *P < .05, compared with sham RS + water gavage controls; ^#P < .05, compared with RS + water gavage rats. (F) Densitometric quantification of occludin protein normalized to GAPDH band intensity. n = 6. *P < .05, compared with sham RS + water gavage controls; ^#P < .05, compared with RS + water gavage.

Figure 6

Effect of chronic neomycin treatment (150 mg/kg, twice daily, oral gavage) on total bacterial load and bacterial community composition in ileal contents. (A) Total bacterial 16s rRNA gene copies. *P < .05, compared with sham WAS + water gavage controls; ^#P < .05, compared with WAS + water gavage rats. (B) Variation in bacterial community composition in the terminal ileum at phylum and family levels. (C) Relative abundance of selected phylotypes, identified at the family level. *significantly different than sham WAS; ^#significantly different than WAS (Kruskal–Wallis with Conover–Inman post hoc test for multiple comparisons, P < .05). (D) Bacterial community composition analyzed with NMDS plots using a θ_YC distance matrix of OTU-based data. For this NMDS plot, stress = 0.06, r² = 0.98. n = 4. N.D. = not detected.

Figure 7

Effect of chronic treatment with neomycin (150 mg/kg, twice daily, oral gavage) on gut permeability and ileal tight junction protein expression. (A) Gut permeability measured by appearance of FITC-labeled dextran in plasma. n = 8. (B) Representative gel images of semiquantitative RT-PCR of tight junction protein occludin mRNA in ileal tissue. (C) Densitometric quantification of occludin mRNA normalized to GAPDH band intensity. n = 6. (D) Representative immunoblots show occludin expression in ileal tissue. (E) Densitometric quantification of occludin protein normalized to GAPDH band intensity. n = 6. For all groups, *P < .05, compared with sham WAS + water gavage controls.