Complement C5 activation promotes type 2 diabetic kidney disease via activating STAT3 pathway and disrupting the gut-kidney axis

Abstract

Diabetic kidney disease (DKD) is a severe DM complication. While complement C5 up-regulation and gut dysbiosis are found in T2DM, their roles in DKD are unclear. Here, we investigated the effect of C5 on the gut microbiota during DKD development. Renal C5a/C5a receptor (C5aR) expression changes were measured in T2DM patients and db/db mice. Db/db mice were treated with a C5aR antagonist (C5aRA), and renal function, gut microbiota and renal genome changes were analysed. The effects of C5a and short-chain fatty acids (SCFAs) on the signal transducer and activator of transcription 3 (STAT3) pathway were examined in vitro. C5a was up-regulated in glomerular endothelial cells (GECs) of T2DM patients and db/db mice. Although glucose and lipid metabolism were unchanged, C5aR blockade alleviated renal dysfunction, ECM deposition, macrophage infiltration and proinflammatory factor expression in db/db mice. C5aRA partly reversed the declines in gut microbiota diversity and abundance and gut SCFA levels in db/db mice. C5aRA down-regulated the expression of many immune response-related genes, such as STAT3, in db/db mouse kidneys. C5aRA and SCFAs suppressed C5a-induced STAT3 activation in human renal glomerular endothelial cells (HRGECs). Based on our results, C5 hyperactivation promotes DKD by activating STAT3 in GECs and impairing the gut-kidney axis, suggesting that this hyperactivation is a potential target for the treatment of DKD.

Keywords: SCFA; STAT3; complement C5; diabetic kidney disease; gut microbiota; inflammation.

FIGURE 1

Increased complement C5 activation in kidneys of DKD patients and db/db mice. (A) Double‐IF staining of C5a and CD31 in the kidneys of humans and mice (scale bar = 50 µm). (B) The fluorescent intensity of C5a in kidney detected by IF staining. (C) The concentration of C5a in serum and renal tissue of mice measured by ELISA. (D) Double‐IF staining of C5aR and CD31 in the kidneys of humans and mice (scale bar = 50 µm). (E) The protein level of C5a and C5aR in kidney of mice measured by Western blot. ^& P < .05, DKD patient vs Healthy. *P < .05, db/db or db/db + C5aRA group vs db/m; ^# P < .05, db/db + C5aRA group vs db/db group

FIGURE 2

Effect of C5aR blockade on general and renal function parameters in db/db mice. (A) Measurement of bodyweight (BW) and renal weight/bodyweight (RW/BW) in mice. (B) Measurement of UACR and UA in mice. (C) Measurement of glucose and lipids (CHOL, TG and LDL) in mice. (D) Representative images of H&E, PAS staining (scale bar = 25 µm) and TEM (scale bar = 2 µm) in renal tissues from mice. (E) Qualitative analysis of PAS‐positive area in renal sections from mice. (F) The gene expression of α‐SMA, TGF‐β, collagen I and FN‐1 in kidneys of db/db mice detected by RNAseq analysis. *P < .05, db/db or db/db + C5aRA group vs db/m; ^# P < .05, db/db + C5aRA group vs db/db group

FIGURE 3

Effect of C5aR blockade on oxidative stress and inflammation in db/db mice. (A) Measurement of H₂O₂ in serum and renal tissue of mice by commercial kit. (B) Measurement of IL‐6 and MCP‐1 in serum of mice by Luminex assay. (C) The gene expression of TLR2, MCP‐1 and F4/80 in kidneys of mice detected by RNAseq analysis. (D‐E) Measurement of 8‐OHdG, F4/80‐positive macrophages, IL‐1β and MCP‐1 expression in the kidneys of mice by IHC staining (scale bar = 50 µm). (F) Measurement of IL‐6 and MCP‐1 protein expression in the kidney of mice by Western blotting. *P < .05, db/db or db/db + C5aRA group vs db/m; ^# P < .05, db/db + C5aRA group vs db/db group

FIGURE 4

Effect of C5aR blockade on gut microbiota homeostasis in db/db mice. (A and C) Representative images of H&E and PAS staining of colons from mice (scale bar = 80 µm). (B and D) Quantification of the villi length (µm) and goblet‐positive cells detected by H&E and PAS staining. (E and F) Measurement of Claudin 1 expression in the small intestine of mice by IHC staining (scale bar = 40 µm). (G) The Shannon index of gut microbiota in mice detected by 16S rRNA sequencing. (H) Principal co‐ordinates analysis (PCoA) of gut microbiota in mice using 16S rRNA sequencing. (I and K) Measurement of the gut microbiota abundance of mice at the phylum level. (J and L) Measurement of the gut microbiota abundance of mice at the genus level. (M) GC‐MS analysis of SCFAs (acetate, butyrate and propionate) level in faecal samples of mice.*P < .05, db/db or db/db + C5aRA group vs db/m group; ^# P < .05, db/db + C5aRA group vs db/db group

FIGURE 5

Effect of C5aR blockade on renal transcriptome in db/db mice. (A) Scatter plot of renal gene expression between different groups of mice detected by RNAseq analysis. (B) Hierarchical clustering heatmap analysis of the changed genes among the three groups. (C) KEGG pathway enrichment analysis of the changed genes in kidneys of mice. (D) Gene network analysis of the changed genes in kidneys of mice. (E) The expression of the changed genes in kidneys of mice detected by RNAseq analysis. *P < .05, db/db or db/db + C5aRA group vs db/m group; ^# P < .05, db/db + C5aRA group vs db/db group

FIGURE 6

Effect of SCFAs on the C5a/C5aR‐STAT3 pathway in HRGECs. (A) Double‐IF staining of p‐STAT3 in HRGECs after different treatments as indicated (scale bar = 20 µm). (B) Quantification of the nucleus/cytoplasm ratio of p‐STAT3 in HRGECs. (C) Western blot and quantitative analysis of p‐STAT3 and STAT3 protein expression in HRGECs after different treatments as indicated (n = 3, *P < .05, control group vs other groups; ^# P < .05, C5a + SCFAs or C5a + C5aRA group vs C5a alone group). (D) Double‐IF staining of p‐STAT3 with CD31 or C5a in renal sections from mice (scale bar = 50 µm). (E) Western blotting analysis of p‐STAT3 and STAT3 protein in the renal tissues of mice. *P < .05, db/db or db/db + C5aRA group vs db/m group; ^# P < .05, db/db + C5aRA group vs db/db group

FIGURE 7

The potential mechanism of complement C5 activation‐induced renal injury in T2DM. In T2DM status, the up‐regulation of C5a activated the STAT3 pathway and thus promoted inflammatory response and further renal fibrosis in kidneys. Meanwhile, the elevated C5a caused gut microbiota dysbiosis (eg declined gut microbiota diversity) with decreased SCFAs production, which impaired the anti‐inflammatory roles of endogenous SCFAs. Conversely, C5aR blockade could attenuate the renal dysfunction through suppressing the STAT3 pathway and restoring the gut‐kidney axis