Jund is a determinant of macrophage activation and is associated with glomerulonephritis susceptibility

Abstract

Crescentic glomerulonephritis is an important cause of human kidney failure for which the underlying molecular basis is largely unknown. In previous studies, we mapped several susceptibility loci, Crgn1-Crgn7, for crescentic glomerulonephritis in the Wistar Kyoto (WKY) rat. Here we show by combined congenic, linkage and microarray studies that the activator protein-1 (AP-1) transcription factor JunD is a major determinant of macrophage activity and is associated with glomerulonephritis susceptibility. Introgression of Crgn2 from the nonsusceptible Lewis strain onto the WKY background leads to significant reductions in crescent formation, macrophage infiltration, Fc receptor-mediated macrophage activation and cytokine production. Haplotype analysis restricted the Crgn2 linkage interval to a 430-kb interval containing Jund, which is markedly overexpressed in WKY macrophages and glomeruli. Jund knockdown in rat and human primary macrophages led to significantly reduced macrophage activity and cytokine secretion, indicating conservation of JunD function in macrophage activation in rats and humans and suggesting in vivo inhibition of Jund as a possible new therapeutic strategy for diseases characterized by inflammation and macrophage activation.

Figure 1

Nephrotoxic nephritis (NTN) and macrophage phenotypes in parental (WKY, Lewis) and chromosome 16 congenic strains (WKY.LCrgn2, LEW.WCrgn2). For NTN phenotypes, rats were killed 10 d after injection of nephrotoxic serum. (a–d) Glomerular crescents (a), fibrin deposition (b), macrophage infiltration defined by ED-1–positive cells per glomerular cross-section (c) and proteinuria (d), measured using at least four rats per strain. (e) Macrophage activation, assessed by Fc receptor mediated phagocytosis and oxidization. WKY, LEW and WKY.LCrgn2 BMDMs (three rats per strain) were stimulated with Fc oxyBURST. LEW and WKY.LCrgn2 BMDMs showed significantly less activation than WKY at all time points (P < 0.001; error bars, s.e.m.). (f) Sandwich ELISA for secretion of MCP-1 in basal (unstimulated) and LPS (100 ng/ml)-stimulated BMDMs; secretion of IL-10 in LPS-stimulated (100 ng/ml) WKY, LEW and WKY.LCrgn2 BMDMs. *P < 0.05 and **P < 0.001 compared to WKY; error bars, s.e.m.

Figure 2

Jund expression and protein levels in the NTN-susceptible WKY rat. (a,b) Microarray (a) and QRT-PCR (b) analyses showed that Jund expression is higher in WKY than in LEW glomeruli at baseline (WKY_c, LEW_c) and in NTN-induced glomeruli (WKY_NTN, and LEW_NTN). At least three rats per strain and per condition (control or NTN) were used. Error bars, s.e.m. (c) Immunostaining for JunD protein in WKY and Lewis glomeruli after NTN induction (day 10), showing increased JunD in the WKY glomeruli. (d) Sequence analysis of the rat Jund promoter, showing a C/T polymorphism at −210 bp (asterisk) in the vicinity of an octamer binding motif (−219 to −212 bp). (e) Luciferase assay performed after transfecting COS7 cells with pGL3-Basic vector containing, 300 bp upstream of the transcription initiation site, either the WKY or LEW Jund promoters (pGL3-WKY and pGL3-LEW, respectively). Firefly luciferase activity, normalized to Renilla luciferase activity, is expressed relative to the activity of the empty pGL3-Basic vector. Error bars, s.e.m. for five different transfection experiments performed in replicates of six. (f) JunD binding to AP-1 site in BMDMs. Specific JunD binding to AP-1 consensus sequence nucleotides (5′-TGAGTCA-3′) was greater in WKY BMDM nuclear extracts than in LEW and WKY.LCrgn2 by TransAM assay. A₄₅₀, absorbance at 450 nm. *P < 0.05; error bars, s.e.m.

Figure 3

Combined gene expression and genetic mapping for NTN susceptibility in rat. Rat Jund promoter polymorphism between WKY and LEW was used to genotype 177 F₂ rats derived from WKY and LEW by PCR-based ARMS assay (see also Supplementary Fig. 4). (a) NTN phenotypes versus Jund genotype. *P < 0.05; **P < 0.01; *** P < 0.001, compared to C/C genotype. Error bars, s.e.m. (b) Macrophage infiltration determined by quantitative analysis of ED-1–positive cells in the glomeruli (including the Bowman's space) in WKY, LEW and four Kyoto-derived rat strains, 10 d after NTN induction. *P < 0.001 compared to LEW; NS, nonsignificant.

Figure 4

Chromosome 16 haplotype analysis. (a) Crgn2 genetic map. High and low macrophage infiltration denote, respectively, strains with macrophage infiltration that is or is not significantly greater than that in the LEW strain. Numbers denote classification of alleles: 1, LEW; 2, WKY; 3 and 4, other alleles. Genetic locations of other significantly differentially expressed genes in the overall Crgn2 congenic interval are also indicated. The boxed area corresponds to microsatellite markers (D16Rat78 and D16Rat82) mapping to the peak of linkage in Crgn2. Genetic distances are indicated in cM. (b) Fine haplotype mapping delineates a maximal 430-kb region of haplotypic identity segregating with macrophage infiltration. Physical location of the markers is indicated in Mbp. The Jund C/T promoter polymorphism showed the same haplotype as D16Got22 and Jund1.

Figure 5

Effect of modulating Jund expression on macrophage activity. (a–c) WKY BMDMs were cultured (n = 3 rats) and transfected either with Jund siRNA or control (nontargeting) siRNA. Jund mRNAs were measured 48 h after transfection by QRT-PCR (a) and macrophage activation was assessed by Fc oxyBURST assay (b) and by measuring Nos2 expression in basal (unstimulated) and LPS-stimulated (100 ng/ml) cells (c). WKY BMDMs with Jund knockdown showed significantly less Fc receptor–mediated activation at all time points than BMDMs transfected with control siRNA (P < 0.001); NS, nonsignificant. Error bars, s.e.m.

Figure 6

JunD expression and cellular activation in human primary macrophages. Human macrophages from five healthy donors were derived from elutriated monocytes. Five independent experiments were performed in which cells were incubated either with Jund siRNA (100 nM) or control (nontargeting) siRNA (100 nM) for 48 h and stimulated with LPS (10 ng/ml) for 24 h. (a) QRT-PCR for human Jund. (b) JunD and p38 protein blotting. (c) IL-10, TNF-α and IL-6 relative secretion were assessed by sandwich ELISA in basal (unstimulated) and 24 h LPS-stimulated (10 ng/ml) cell supernatants. Error bars, s.e.m.