Galacto-oligosaccharides attenuate renal injury with microbiota modification

Abstract

Tubulointerstitial injury is central to the progression of end-stage renal disease. Recent studies have revealed that one of the most investigated uremic toxins, indoxyl sulfate (IS), caused tubulointerstitial injury through oxidative stress and endoplasmic reticulum (ER) stress. Because indole, the precursor of IS, is synthesized from dietary tryptophan by the gut microbiota, we hypothesized that the intervention targeting the gut microbiota in kidney disease with galacto-oligosaccharides (GOS) would attenuate renal injury. After 2 weeks of GOS administration for 5/6 nephrectomized (Nx) or sham-operated (Sham) rats, cecal indole and serum IS were measured, renal injury was evaluated, and the effects of GOS on the gut microbiota were examined using pyrosequencing methods. Cecal indole and serum IS were significantly decreased and renal injury was improved with decreased infiltrating macrophages in GOS-treated Nx rats. The expression levels of ER stress markers and apoptosis were significantly increased in the Nx rats and decreased with GOS. The microbiota analysis indicated that GOS significantly increased three bacterial families and decreased five families in the Nx rats. In addition, the analysis also revealed that the bacterial family Clostridiaceae was significantly increased in the Nx rats compared with the Sham rats and decreased with GOS. Taken altogether, our data show that GOS decreased cecal indole and serum IS, attenuated renal injury, and modified the gut microbiota in the Nx rats, and that the gut microbiota were altered in kidney disease. GOS could be a novel therapeutic agent to protect against renal injury.

Keywords: Endoplasmic reticulum stress; galacto‐oligosaccharides; gut microbiota; indoxyl sulfate; kidney disease.

Figure 1.

Blood pressure and body weight. The (A) systolic blood pressure was significantly higher in the Con Nx and GOS Nx rats than in the corresponding Sham groups, but no significant differences were found between the Con Nx and GOS Nx rats or between the Con Sham and GOS Sham rats. No significant changes in (B) body weight were found among the four groups.

Figure 2.

Concentrations of (A) cecal indole and (B) serum indoxyl sulfate. (A) GOS significantly decreased the cecal indole concentrations both in the Sham and Nx rats (amount per 1 g of the cecal contents). (B) GOS significantly decreased the serum indoxyl sulfate concentrations in GOS Nx rats compared with Con Nx rats at 3 and 4 week. *Difference with P < 0.05.

Figure 3.

Concentrations of blood urea nitrogen (BUN) and serum creatinine (sCre). (A) BUN and ( B) sCre were significantly higher in the Con Nx and GOS Nx rats than in the corresponding Sham groups, but no significant changes were found between the Con Nx and GOS Nx rats or between the Con Sham and GOS Sham rats.

Figure 4.

Semiquantitative analysis of tubulointerstitial injury and immunohistochemistry of macrophages. The injury worsened in (C) Con Nx rats compared with (A) Con Sham rats and ameliorated in (D) GOS Nx rats compared with (B) Con Nx rats. Semiquantitative analysis of (E) tubulointerstitial injury revealed the significant deterioration in Con Nx than Con Sham and the significant amelioration in GOS Nx rats compared with CON Nx rats. Infiltrating macrophages were increased in (H) Con Nx rats compared with (F) Con Sham rats and decreased in (I) GOS Nx rats compared with Con Nx rats. Quantitative analysis of (J) infiltrating macrophages revealed a significant increase in Con Nx rats compared with Con Sham rats and a significant decrease in GOS Nx rats compared with Con Nx rats. Original magnification: 200×; Scale bars: 50 µm. *Difference with adjusted P < 0.05.

Figure 5.

Expressions of CCAAT/enhancer‐binding protein homologous protein (CHOP). (A) CHOP mRNA expression was examined with quantitative PCR. The expression was significantly upregulated in Con Nx rats compared with Con Sham rats and significantly downregulated in GOS Nx rats compared with Con Nx rats. Immunohistochemistry revealed that CHOP‐positive cells were increased in (D) Con Nx rats compared with (B) Con Sham rats and decreased in (E) GOS Nx rats compared with Con Nx rats. Quantitative analysis of CHOP‐positive cells (F) revealed a significant increase in Con Nx rats compared with Con Sham rats and a significant decrease in GOS Nx rats compared with Con Nx rats. Original magnification: 200×; Scale bars: 50 µm. *Difference with adjusted P < 0.05.

Figure 6.

Expressions of glucose‐regulated protein (GRP) 78. Immunohistochemistry of GRP78 revealed that the GRP78‐positive area was increased in Con Nx rats (C) compared with Con Sham rats (A) and decreased in GOS Nx rats (D) compared with Con Nx rats. Quantitative analysis of the GRP78‐positive area (E) revealed a significant increase in Con Nx rats compared with Con Sham rats and a significant decrease in GOS Nx rats compared with Con Nx rats. Western blotting of GRP78 (F, G) showed the same statistically significant changes. Original magnification: 200×; Scale bars: 50 µm. *Difference with adjusted P < 0.05.

Figure 7.

Terminal deoxynucleotidyl transferase‐mediated dUTP nick end labeling (TUNEL) assay and expression of cleaved caspase‐3. TUNEL‐positive cells were increased in Con Nx rats (C) compared with Con Sham rats (A) and decreased in GOS Nx rats (D) compared with Con Nx rats. Quantitative analysis of TUNEL‐positive cells (E) revealed a significant increase in Con Nx rats compared with Con Sham rats and a significant decrease in GOS Nx rats compared with Con Nx. Immunohistochemistry of cleaved caspase‐3 showed that cleaved caspase‐3‐positive cells were increased in Con Nx rats (H) compared with Con Sham rats (F) and decreased in GOS Nx rats (I) compared with Con Nx rats. Quantitative analysis of cleaved caspase‐3‐positive cells (J) revealed a significant increase in Con Nx rats compared with Con Sham rats and a significant decrease in GOS Nx rats compared with Con Nx rats. Western blotting of cleaved caspase‐3 (K, L) showed the same statistically significant changes. Original magnification: 400×; Scale bars: 30 µm. *Difference with adjusted P < 0.05.

Figure 8.

Proportions (%) of the bacterial families in the cecum analyzed with pyrosequencing methods (A–C) and concentrations of short‐chain fatty acids (SCFAs) in the cecum (D). The proportion of “Clostridiaceae” (A) was significantly increased in Con Nx rats compared with Con Sham rats and was significantly decreased in GOS Nx rats compared with Con Nx rats. The proportions of “Bifidobacteriaceae,” “Clostridiales; Incertae Sedis XIV,” and “Porphyromonadaceae” (B) were significantly increased in GOS Nx rats compared with Con Nx rats. The proportions of “Ruminococcaceae,” “Peptostreptococcaceae,” “Clostridiaceae,” “Streptococcaceae,” “Veillonellaceae,” and “Clostridiales; Incertae Sedis XIII” (C) were significantly decreased in GOS Nx rats compared with Con Nx rats. Six SCFA concentrations were examined (D), and succinic acid was significantly increased in GOS Nx rats compared with Con Nx rats. *Difference with adjusted P < 0.05.

Figure 9.

The alterations in the gut microbiota and the amelioration of renal injury. In 5/6 Nx rats without GOS (A), Clostridiaceae was increased and serum IS was elevated resulting in more severe renal injury with more severe ER stress and apoptosis. In 5/6 Nx rats with GOS (B), the microbiota were altered and serum IS was decreased with lower indole concentration resulting in less mild renal injury with less ER stress and apoptosis.