TLR-mediated albuminuria needs TNFα-mediated cooperativity between TLRs present in hematopoietic tissues and CD80 present on non-hematopoietic tissues in mice

Abstract

Transient albuminuria induced by pathogen-associated molecular patterns (PAMPs) in mice through engagement of Toll-like receptors (TLRs) is widely studied as a partial model for some forms of human nephrotic syndrome (NS). In addition to TLRs, CD80 has been shown to be essential for PAMP-mediated albuminuria. However, the mechanistic relationships between TLRs, CD80 and albuminuria remain unclear. Here, we show that albuminuria and CD80-uria induced in mice by many TLR ligands are dependent on the expression of TLRs and their downstream signalling intermediate MyD88 exclusively in hematopoietic cells and, conversely, on CD80 expression exclusively in non-hematopoietic cells. TNFα is crucial for TLR-mediated albuminuria and CD80-uria, and induces CD80 expression in cultured renal podocytes. IL-10 from hematopoietic cells ameliorates TNFα production, albuminuria and CD80-uria but does not prevent TNFα-mediated induction of podocyte CD80 expression. Chitohexaose, a small molecule originally of parasite origin, mediates TLR4-dependent anti-inflammatory responses, and blocks TLR-mediated albuminuria and CD80-uria through IL-10. Thus, TNFα is a prominent mediator of renal CD80 induction and resultant albuminuria in this model, and small molecules modulating TLR-mediated inflammatory activation might have contributory or adjunct therapeutic potential in some contexts of NS development.

Keywords: CD80; IL-10; Proteinuria; TLR; TNFα.

Fig. 1.

CD80^−/− mice are resistant to LPS- and poly(I:C)-mediated albuminuria. (A) Comparison of urinary albumin:creatinine ratios between B6 and CD80^−/− mice given PBS, LPS or poly(I:C) [number of mice per group as follows: B6+PBS=6, B6+LPS=10, B6+p(I:C)=3; CD80^−/−=5 in all groups]. (B) Comparison of urinary CD80 levels between B6 and CD80^−/− mice given PBS or p(I:C) (five mice per group). (D) Levels of albumin in urine of B6 mice given PBS or Pam3CSK4 (six mice per group). The results represent mean±s.e.m. ND, not detectable.

Fig. 2.

CD80 expression on podocytes is essential for albuminuria. (A) Comparison of urinary albumin:creatinine ratios in BM chimeric mice given LPS or PBS. The key identifies BM donors and recipients (five mice per group). (B) CD80 levels in BM chimeric mice given LPS or PBS (five mice per group). ‘B6’ in panels A and B indicates B6.SJL. (C) Relative CD80 mRNA expression in E11 podocytes exposed to serum from B6 mice that had been treated as indicated (n=6, data from three independent experiments). (D) Relative CD80 mRNA expression in E11 podocytes exposed to serum from B6 and CD80^−/− mice given LPS or PBS (n=6, data from three independent experiments). The results represent mean±s.e.m. ND, not detectable; ns, not significant.

Fig. 3.

TLR expression and signalling in non-hematopoietic cells is dispensable for induction of albuminuria. (A) Comparison of urinary albumin:creatinine ratios in C3H/OuJ and C3H/HeJ mice that had been treated with PBS, LPS or p(I:C) (five mice per group). (B) Comparison of urinary CD80 levels in C3H/OuJ and C3H/HeJ mice given LPS or PBS (eight mice per group). (C) Comparison of urinary albumin:creatinine ratios in BM chimeric mice in response to LPS (five mice per group). (D) CD80 levels in BM chimeric mice given LPS or PBS (five mice per group). (E) Comparison of urinary albumin:creatinine ratios in B6 and MyD88^−/− mice given LPS, p(I:C) or PBS (five mice per group). (F) Comparison of urinary CD80 levels in B6 and MyD88^−/− mice given LPS or PBS (eight mice per group). Keys identify BM donors and recipients. Data represents mean±s.e.m. ND, not detectable; ns, not significant.

Fig. 4.

Systemic TNFα is responsible for albuminuria and CD80 upregulation. (A-C) TNFα (A), IL-6 (B) and IL-10 (C) levels in sera from B6 mice given PBS (CTRL) or LPS. Sera were collected at the times after treatment as indicated (four mice per group). (D) serum CD80 levels in B6 and CD80^−/− mice treated with LPS or PBS (five mice per group). (E) Serum TNFα levels 2 h after injection of LPS or PBS from C3H/HeJ and C3H/OuJ mice (five mice per group). (F) Serum TNFα levels 2 h after injection of LPS or PBS from MyD88^−/− and B6 mice (five mice per group). (G) Relative CD80 mRNA expression levels in E11 podocytes that had been exposed to serum from C3H/HeJ and C3H/OuJ mice given PBS or LPS (n=8, four independent experiments). (H) Serum TNFα levels 2 h after injection of 3 μg LPS or PBS from B6 and IL-10^−/− mice (five mice per group). (I) Relative CD80 mRNA expression levels in E11 podocytes that had been exposed to serum from B6 and IL-10^−/− mice given PBS or 3 μg LPS (n=6, three independent experiments). Data represents mean±s.e.m. ND, not detectable.

Fig. 5.

Characterization of the role of TNFα-mediated signalling in vivo and in vitro. (A) Relative CD80 mRNA expression in E11 podocytes in response to increasing concentrations of recombinant TNFα (rTNFα) (n=8, all groups, four independent experiments). (B) Comparison of urinary albumin:creatinine ratios in B6 mice given PBS, Etanercept (ETA) alone, LPS alone or in combination as shown (five mice per group). (C) Comparison of urinary CD80 levels in mice given PBS, ETA alone, LPS alone or in combination (number of mice in each group, PBS=5, ETA=5, LPS=7, ETA+LPS=7). (D) Relative CD80 mRNA expression levels in E11 podocytes with or without treatment with recombinant TNFα, in the presence or absence of inhibitors as shown (n=6, three independent experiments). Data represents mean±s.e.m. in, inhibitor; ND, not detectable; ns, not significant; UnT, untreated. (E) TNFα-mediated NFκB activation in nuclear extracts of cultured podocytes over time, determined with EMSA analysis using a DNA probe derived from the HIV LTR harbouring consensus κB sites (top panel), or from the CD80 promoter encompassing the −2969 to −2945 region (bottom panel). (F) The specificity of NFκB binding to the CD80 promoter sequence was examined in an oligonucleotide competition assay using 50-fold molar excess of unlabelled κB oligonucleotide or mutant oligonucleotide. *non-specific bands. Data are representative of three independent experiments. ssAb, antibodies used for supershift assay.

Fig. 6.

Chitohexaose blocks TLR-mediated TNFα induction leading to inhibition of CD80-uria and albuminuria. (A) Comparison of urinary albumin:creatinine ratios in B6 mice receiving PBS, chitohexaose (Chx), LPS or chitohexaose and LPS (five mice per group). (B) Comparison of urinary CD80 levels in B6 mice receiving PBS, LPS or chitohexaose and LPS (nine mice per group). (C,D) Serum TNFα (C) and IL-10 (D) levels in B6 mice given PBS alone, LPS alone or LPS and chitohexaose (four mice per group), 2 h post injection. (E) Relative CD80 mRNA expression in E11 podocytes treated with sera from B6 mice exposed to PBS (CTRL), chitohexaose alone, LPS alone or LPS and chitohexaose (n=8, four independent experiments). (F) Comparison of urinary albumin:creatinine ratios in B6 mice receiving PBS, p(I:C) or chitohexaose and p(I:C) (five mice per group). Data represent mean±s.e.m. ND, not detectable.

Fig. 7.

Chitohexaose blocks TLR-mediated TNFα induction through IL-10. (A) Comparison of urinary albumin:creatinine ratios in B6 and IL-10^−/− mice given 1 μg of LPS, 10 μg of p(I:C) or PBS (five mice per group). (B) Comparison of urinary CD80 levels in B6 and IL-10^−/− mice receiving 1 μg of LPS or PBS (six mice per group). (C) Comparison of urinary albumin:creatinine ratios in IL-10^−/− mice given PBS, 1 μg of LPS without or with 250 μg chitohexaose (five mice per group). (D) Comparison of urinary CD80 levels in IL-10^−/− mice given PBS, 1 μg of LPS without or with 250 μg of chitohexaose (five mice per group). (E) Relative CD80 mRNA expression levels in E11 podocytes that had been exposed to PBS (CTRL), TNFα alone (1 ng/ml), IL-10 alone (50 ng/ml) or in combination (n=6, three independent experiments). Data represent mean±s.e.m. ND, not detectable; ns, not significant.