GDF15 Suppresses Lymphoproliferation and Humoral Autoimmunity in a Murine Model of Systemic Lupus Erythematosus

Abstract

Growth and differentiation factor 15 (GDF15), a divergent member of the transforming growth factor-β superfamily, has been associated with acute and chronic inflammatory conditions including autoimmune disease, i.e., type I diabetes and rheumatoid arthritis. Still, its role in systemic autoimmune disease remains elusive. Thus, we studied GDF15-deficient animals in Fas-receptor intact (C57BL/6) or deficient (C57BL/6lpr/lpr) backgrounds. Further, lupus nephritis (LN) microdissected kidney biopsy specimens were analyzed to assess the involvement of GDF15 in human disease. GDF15-deficiency in lupus-prone mice promoted lymphoproliferation, T-, B- and plasma cell-expansion, a type I interferon signature, and increased serum levels of anti-DNA autoantibodies. Accelerated systemic inflammation was found in association with a relatively mild renal phenotype. Splenocytes of phenotypically overall-normal Gdf15-/- C57BL/6 and lupus-prone C57BL/6lpr/lpr mice displayed increased in vitro lymphoproliferative responses or interferon-dependent transcription factor induction in response to the toll-like-receptor (TLR)-9 ligand CpG, or the TLR-7 ligand Imiquimod, respectively. In human LN, GDF15 expression was downregulated whereas type I interferon expression was upregulated in glomerular- and tubular-compartments versus living donor controls. These findings demonstrate that GDF15 regulates lupus-like autoimmunity by suppressing lymphocyte-proliferation and -activation. Further, the data indicate a negative regulatory role for GDF15 on TLR-7 and -9 driven type I interferon signaling in effector cells of the innate immune system.

Keywords: Autoantibodies; Autoimmunity; Growth and differentiation factor 15; Inflammation; Lupus nephritis; Macrophages; Toll-like-receptor.

Fig. 1

GDF15 suppresses systemic lymphoproliferative disease in 6-month-old lupus-prone C57BL/6^lpr/lpr mice. a 6-month-old C57BL/6 (n = 20), Gdf15−/− (n = 16), C57BL/6^lpr/lpr (n = 49) and Gdf15−/−^lpr/lpr (n = 30) female mice showed normal body weights. At 6 months of age female GDF15-deficient C57BL/6^lpr/lpr mice showed increased cervical, axillar, and mesenteric lymph node- (b) as well as spleen-weights (c) compared with C57BL/6^lpr/lpr or GDF15-deficient or intact C57BL/6, respectively (n = 8–41 as indicated in a figure). Data are presented as box-whiskers-plots, **p < 0.01; ***p < 0.001. d C57BL/6 (n = 8) and C57BL/6^lpr/lpr (n = 11) female mice were bled at age of 6 months to determine the serum levels of GDF15 protein. e Spleen sections of 6-month-old C57BL/6^lpr/lpr (n = 12) and Gdf15−/−^lpr/lpr (n = 12) female mice were stained with Ki67 proliferation marker (brown) and quantified (dot plot) using image software. RNA was isolated from spleens of 6 months old female Gdf15−/−^lpr/lpr (n = 6) and control C57BL/6^lpr/lpr (n = 6) mice for real-time PCR analysis. Data are expressed as means of the ratio of Ki67 versus that of GAPDH mRNA. Data are presented as means ± SD, **p < 0.01 versus control mice. f, g RNA was isolated from Gdf15−/−^lpr/lpr (n = 6) and control C57BL/6^lpr/lpr (n = 6) mice. Real-time PCR analysis represents DNAse-I and Caspase-3 relative expression; **p < 0.01 versus control mice. h PI+ splenocytes were quantified by flow cytometry (**p < 0.01). Cell-free DNA was quantified in serum by real-time PCR (n = 6 per group) (i) and picogreen assay (j). Data are presented as means ± SD (j), ***p < 0.001 versus control mice (n = 6–26 per group as indicated in a figure).

Fig. 2

GDF15 selectively controls the activation of T lymphocytes. a We presented the expression of relevant genes from spleens of 6-month-old C57BL/6, Gdf15−/−, C57BL/6^lpr/lpr, and Gdf15−/−^lpr/lpr female mice (n = 5 per genotype) in form of a heatmap. The separation of autoimmunity patterns was based on average linkage hierarchical clustering. Dark pink colors on the heatmaps represent low expression values, whitish colors represent average values, and the green represents high values. The row Z-scores were used to ensure that the expression patterns are not overwhelmed by the expression values. Several genes that displayed different regulation between C57BL/6^lpr/lpr (n = 6) and Gdf15−/−^lpr/lpr (n = 7) genotypes in the preliminary experiment were selected for real-time RT-qPCR validation. b The percentage of (activated) T cells was assessed from spleens of 6-months old females by flow cytometry in Gdf15−/−^lpr/lpr (n = 6) and control C57BL/6^lpr/lpr (n = 6) mice. After dead cell exclusion (PI+) cells, CD3+CD4+CD8-(CD44+) and CD3+CD4−CD8+(CD44+) were gated according to online supplementary Figure 4. c The frequency of T regulatory cells in spleens was reduced in Gdf15−/−^lpr/lpr (n = 6) compared to control (n = 6) mice. d, e The expression of Prdm1 and April in spleens from 6-month-old Gdf15−/−^lpr/lpr (n = 6) and control C57BL/6^lpr/lpr (n = 6) female mice. f Reports frequencies of spleen B220+ CD19+ MHCII+ B-cells (% of alive cells after PI dead cell exclusion as depicted in the online suppl. Fig. 4) in Gdf15−/−^lpr/lpr (n = 6) versus C57BL/6^lpr/lpr (n = 6) mice. g RNA was isolated from spleens of Gdf15−/−^lpr/lpr (n = 6) and control C57BL/6^lpr/lpr (n = 6) mice for real-time PCR analysis of Flt3L. h After exclusion of dead cells (PI) percentage of CD86+ CD11c+ F4/80- dendritic cells and CD86+ F4/80+ CD11c-macrophages were assessed from spleens of 6-month-old Gdf15−/−^lpr/lpr (n = 6) and control C57BL/6^lpr/lpr (n = 6) female mice by flow cytometry. Data represent means ± SD; *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 3

GDF15 suppresses type I interferon signature and autoantibody production in lupus-prone mice. a RNA was isolated from spleens of C57BL/6 (n = 5), Gdf15−/− (n = 6), Gdf15−/−^lpr/lpr (n = 7), and control C57BL/6^lpr/lpr (n = 6) female mice for real-time PCR analysis of pro-inflammatory and interferon signature genes Ifit1 and Ifit3. b RNA was isolated from spleens of Gdf15−/−^lpr/lpr (n = 7) and control C57BL/6^lpr/lpr (n = 6) female type I and type III interferon signature genes. Data are presented as means ± SD, *p < 0.05; **p < 0.01 versus control mice. c C57BL/6 (n = 11–21), Gdf15−/− (n = 10–15), Gdf15−/−^lpr/lpr (n = 10–30) and control C57BL/6^lpr/lpr (n = 10–17) female mice were bled at month 6 to determine serum levels of IL-6, IL-12, IL-10, IFN-γ, MCP-1, and TNF-α; **p < 0.01; ***p < 0.001. d Furthermore, the levels of IFN- α were determined in serum (n = 6–14 per genotype). Data represent means ± SD; **p < 0.01. e Percentage of B220+ CD11c+ CD137+ spleen − pDCs were quantified in 6-month-old female mice (n = 6 per genotype) by flow cytometry. The graph presents the mean ± SD; *p < 0.05. f Spleen sections from C57BL/6 (n = 8), Gdf15−/− (n = 8), Gdf15−/−^lpr/lpr (n = 8), and control C57BL/6^lpr/lpr (n = 8) female mice were stained with antibodies against B220 and BLIMP1. The positive signal was quantified by PhotoShop Software (positive area per HPF). Data are presented as means ± SD, *p < 0.05; **p < 0.01 versus control mice. g Mature kappa-light chain+ CD138+ plasma cells were quantified in spleens of 6-month-old female mice (n = 9 per genotype) by flow cytometry. The histogram presents the mean ± SD, **p < 0.01 versus control mice. h, i C57BL/6 (n = 5–20), Gdf15−/− (n = 6–12), Gdf15−/−^lpr/lpr (n = 8–19) and control C57BL/6^lpr/lpr (n = 8–32) female mice were bled at month 6 to determine serum levels of immunoglobulins as well as (i) anti-dsDNA, anti-Smith, and anti-histones autoantibodies by ELISA. The lupus strain of mice with the severe phenotype (MRL^lpr/lpr; n = 4–5) was used as a positive control. The (n) number of animals is for every genotype per assay is indicated in a figure. The specificity of anti-dsDNA antibodies was confirmed by Critidiae luciliae assay (data not shown). Data show means ± SD; *p < 0.05; **p < 0.01. HPF, high power field.

Fig. 4

GDF15 reduces kidney injury in 6-month-old lupus-prone mice. a Kidney function parameters (serum creatinine, proteinuria and albumin/creatinine ratio) were measured in all groups of 6-month-old female mice (n = 10–25; the nis for every genotype per assay is indicated in a figure). Data are presented as means ± SD; *p < 0.05. b Kidney sections from C57BL/6 (n = 6–7), Gdf15−/− (n = 6–7), Gdf15−/−^lpr/lpr (n = 14–24), and control C57BL/6^lpr/lpr (n = 14–17) female mice were stained with antibodies against Mac2 and CD45, as well as (c) complement component 9 (C9) (n = 6 and n = 12), total IgG (n = 10 and n = 12), and WT1 (n = 14 and n = 14). The pathological changes were quantified by cell counting or by using a semiquantitative score. Data are shown as box-whiskers-plots or means ± SD; **p < 0.01, ***p < 0.001. n = number of animals is indicated in the figure.

Fig. 5

GDF15 negatively regulates TLR-9 dependent inflammation and cell proliferation. a C57BL/6 (n = 7), Gdf15−/− (n = 8), Gdf15−/−^lpr/lpr (n = 10) and control C57BL/6^lpr/lpr (n = 7) female mice were bled at month 6 to determine serum levels of corticosterone and aldosterone. Data represent means ± SD. b Expression levels of Tlr7 and Tlr-9 in spleens of 6-month-old C57BL/6 (n = 5), Gdf15−/− (n = 5), Gdf15−/−^lpr/lpr (n = 6), and control C57BL/6^lpr/lpr (n = 6) females were quantified by real-time PCR. Data are shown as means ± SD of the ratio of the specific mRNA versus that of GAPDH mRNA, *p < 0.05. c The production of GDF15 was quantified in supernatants of BM derived macrophage isolated from C57BL/6 (n = 7) and C57BL/6^lpr/lpr (n = 12) mice by ELISA 24 h upon CpG 1668 (100 ng/mL) stimulation. Data represent means ± SD; ***p < 0.001. d The proliferative activity of spleen lymphocytes isolated from 8-weeks old mice, stimulated with 10% FCS (n = 4) or CpG (n = 8) was evaluated by MTT assay 72 h upon CpG 1668 (50 ng/mL). The graphs represent a representative experiment (from three independent experiments) performed in a single run; presented as mean ± SD; *p < 0.05. e Expression analysis of preselected interferon type I dependent genes in BM-derived macrophages isolated from all four strains of mice and stimulated with LPS (10 ng/mL), CpG 1668 (100 ng/mL), and Imiquimod (1 μg/mL); n = 5 per each group and genotype. The expression of selected genes was determined in BM-derived macrophages 4 h poststimulation. The graphs represent a representative experiment (from three independent experiments) performed in a single run. Data represent means ± SD; *p < 0.05; **p < 0.01, ***p < 0.001. f The differences in TNFα expression (n = 4 per genotype and stimulation) and IL12 and TNFα production upon CpG 1668 (100 ng/mL) stimulation (or costimulation with 20 ng/mL of IFNγ) in BM-derived macrophages from C57BL/6, Gdf15−/− as well as Gdf15−/−^lpr/lpr and control C57BL/6^lpr/lpr mice were evaluated with qPCR (4 h) and ELISA (24 h). The biological replicates of n = 5 per group were used. Data represent means ± SD; *p < 0.05; **p < 0.01. The graphs represent a representative experiment (from three independent experiments) performed in a single run.

Fig. 6

GDF15 expression in human LN. Gene expression analysis of GDF15, IFIT1, IFIT3, MKI67, and PRDM1 genes in glomerular (a) and tubular (b) compartment of manually microdissected kidney biopsies from patients with HTN (Glom: n = 15; Tub: n = 21), MCD (Glom: n = 14; Tub: n = 15), and LN (Glom: n = 32; Tub: n = 32). Pretransplantation kidney biopsies from LDs (Glom: n = 42, Tub: n = 42) were used as control renal tissue. Values are expressed as a log2-fold change compared to controls (LD). All represented genes are significantly changed (p < 0.05), except when indicated n.s. Spearman correlation matrix for each gene-gene correlation in glomeruli (c) and tubulointerstitium (d) of patients (n = 32) with LN and (n = 42) controls (LDs). Shown are the Spearman correlation coefficients. A p value below 0.05 was considered to be statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001). n.s., not significant.