Inflammation and Immunity Pathways Regulate Genetic Susceptibility to Diabetic Nephropathy

Abstract

Diabetic nephropathy (DN) is a leading cause of end-stage renal disease worldwide, but its molecular pathogenesis is not well defined, and there are no specific treatments. In humans, there is a strong genetic component determining susceptibility to DN. However, specific genes controlling DN susceptibility in humans have not been identified. In this study, we describe a mouse model combining type 1 diabetes with activation of the renin-angiotensin system (RAS), which develops robust kidney disease with features resembling human DN: heavy albuminuria, hypertension, and glomerulosclerosis. Additionally, there is a powerful effect of genetic background regulating susceptibility to nephropathy; the 129 strain is susceptible to kidney disease, whereas the C57BL/6 strain is resistant. To examine the molecular basis of this differential susceptibility, we analyzed the glomerular transcriptome of young mice early in the course of their disease. We find dramatic differences in regulation of immune and inflammatory pathways, with upregulation of proinflammatory pathways in the susceptible (129) strain and coordinate downregulation in the resistant (C57BL/6) strain. Many of these pathways are also upregulated in rat models and in humans with DN. Our studies suggest that genes controlling inflammatory responses, triggered by hyperglycemia and RAS activation, may be critical early determinants of susceptibility to DN.

Figure 1

Fasting blood glucose, urinary AE, and renal pathology in mice. A: Compared with normal levels in WT mice, fasting blood glucose was significantly elevated in all of the Akita mouse groups during the period from 2 to 6 months of age, and there were no significant differences in blood glucose values between the various Akita mouse lines at these time points. B: Twenty-four–hour urinary AE was significantly higher in 24-week-old 129/SvEv strain (white bars) RenTg or Akita-RenTg mice compared with corresponding lines at 24 weeks on the C57BL/6 strain (black bars). Data are expressed as mean ± SEM. *P < 0.05; **P < 0.001 compared with C57BL/6; ‡P < 0.001 compared to 129/SvEv WT. C–F: Photomicrographs of periodic acid-Schiff–stained sections of mouse kidneys. Representative low- (C and E) and high-magnification (D and F) images of periodic acid-Schiff–stained kidney sections from 24-week-old 129/SvEv mice with RenTg alone (C and D) or 24-week-old double-heterozygous 129/SvEv Akita-RenTg mice (E and F). 129/SvEv RenTg mice have minimal pathological abnormalities, whereas 129/SvEv Akita-RenTg mice exhibit marked mesangial expansion and frank glomerular sclerosis with focal areas of interstitial inflammation.

Figure 2

Differential expression of cytokine and cytokine receptor interaction pathways. Mean-average plots comparing mean expression levels and fold change in gene expression in 129 (A) and B6 (B) studies. Gray dots represent all genes analyzed on the microarray, and the black crosses refer to genes in the cytokine–cytokine receptor interaction pathway that significantly contribute pathway enrichment. The x-axis represents the average gene expression over the groups, and y-axis represents the fold change (expressed as log ratio).

Figure 3

Changes in inflammation-related pathway gene expression from IPA. Tree map representation of the top-level inflammatory response biofunction in 129 and C57BL/6 strains of the Akita-RenTg and WT mice. The visualization is a hierarchical heat map in which the major boxes represent top-level biofunctions. Within each box, each individual rectangle is a subfunction related to the top-level function. The size of a rectangle is correlated with increasing overlap significance among the gene members of the biofunction and the query genes. Intensity-shaded orange and blue colors represent activated and inhibited states of the subfunctions, respectively, as determined by their z scores (darker shades indicate higher absolute z scores). Nonsignificantly enriched biofunctions are indicated in gray. APC, antigen-presenting cell; CNS, central nervous system.

Figure 4

Increased chemokine expression in mice with DN. Expression of CXCR4 (A) and CCL5 (B) were significantly higher in kidneys of 129– versus C57BL/6–Akita-RenTg mice. *P = 0.002 vs. C57BL/6–Akita-RenTg; **P < 0.0001 vs. C57BL/6–Akita-RenTg.

Figure 5

TNF-based, probable causal mechanistic networks in 129/SvEv, C57BL/B6, and human transcriptomic profiles. A: The predicted activation or inhibition state of a regulatory network consisting of TNF and its interacting partners are shown for 129/SvEv (129) and C57BL/6 (B6). Molecules are shaped according to function (ellipse, transcription factor; rectangle, cytokine; and triangle, kinase). Solid lines indicate a direct relationship (no intermediates) and dashed lines indicate an indirect relationship between interacting molecules. Activating relations are depicted by arrows, and inhibitory effects are depicted by a line with a bar. The color-coding of the regulators and their interactions are as follows: orange, predicted activation; blue, predicted inhibition; gray, effect not predicted; and yellow, effect inconsistent with predicted. Reflecting the broader analyses, these proinflammatory networks are upregulated on the susceptible 129 background and downregulated on the C57BL/6 background. B: Upregulation of the TNF mechanistic network is also observed in transcriptomic profiles from Nephromine data, similar to the pattern seen in the susceptible 129 mouse strain background.

Figure 6

Increased TNF-α protein levels in kidneys of mice with DN. TNF-α levels were measured in kidney lysates from 129– and C57BL/6–Akita-RenTg mice. A: Representative Western blot image from kidney lysates immunoprobed with anti–TNF-α and anti-GAPDH antibodies. Lanes 1–3 are samples from C57BL/6 mice, and lanes 4–6 are from 129 mice. B: TNF-α levels were more than threefold higher in the 129–Akita-RenTg group compared with C57BL/6–Akita-RenTg mice. *P = 0.03.

Figure 7

Interspecies comparisons of gene and pathway changes in comparable phenotypes. Comparison of direction of fold change for differentially expressed genes identified in common between the rat study vs. 129/SvEv (A) and rat vs. C57BL/6 (B). Black bars indicate results from the rat study, and white bars represent results in mice.