Type 1 interferon suppresses expression and glucocorticoid induction of glucocorticoid-induced leucine zipper (GILZ)

Abstract

SLE is a systemic multi-organ autoimmune condition associated with reduced life expectancy and quality of life. Glucocorticoids (GC) are heavily relied on for SLE treatment but are associated with detrimental metabolic effects. Type 1 interferons (IFN) are central to SLE pathogenesis and may confer GC insensitivity. Glucocorticoid-induced leucine zipper (GILZ) mediates many effects of GC relevant to SLE pathogenesis, but the effect of IFN on GC regulation of GILZ is unknown. We performed in vitro experiments using human PBMC to examine the effect of IFN on GILZ expression. JAK inhibitors tofacitinib and tosylate salt were used in vivo and in vitro respectively to investigate JAK-STAT pathway dependence of our observations. ChiP was performed to examine glucocorticoid receptor (GR) binding at the GILZ locus. Several public data sets were mined for correlating clinical data. High IFN was associated with suppressed GILZ and reduced GILZ relevant to GC exposure in a large SLE population. IFN directly reduced GILZ expression and suppressed the induction of GILZ by GC in vitro in human leukocytes. IFN actions on GILZ expression were dependent on the JAK1/Tyk2 pathway, as evidenced by loss of the inhibitory effect of IFN on GILZ in the presence of JAK inhibitors. Activation of this pathway led to reduced GR binding in key regulatory regions of the GILZ locus. IFN directly suppresses GILZ expression and GILZ upregulation by GC, indicating a potential mechanism for IFN-induced GC resistance. This work has important implications for the ongoing development of targeted GC-sparing therapeutics in SLE.

Keywords: GILZ; STAT1; autoimmunity; glucocorticoid; inflammation; interferon; systemic lupus erythematosus (SLE).

Figure 1

PBMC from SLE patients with high IFN score and higher disease activity express less GILZ. (A) IFN score in PBMC of healthy controls and SLE patients. (B) GILZ expression in PBMC of SLE patients with low (n=439) or high IFN (n=1317)score. (C) Correlation between IFN score (expressed as STD from mean healthy control expression level) and GILZ mRNA expression. (D–G) GILZ expression in PBMC of SLE patients, categorized by SLEDAI (D), C3 and C4 complement activity (E–F) and anti-dsDNA status (G). All data originated from GSE88884 (n=1,756 SLE patients, n=50 healthy controls). *p<0.05, ****p<0.0001. ns, Not significant.

Figure 2

Type I IFN inhibits GILZ expression in human and mouse. (A) Healthy PBMC were treated with 1000 IU IFNα for 1 or 3 hours (left panel), or with 100 IU or 1000 IU for 3 hours (right panel), after which GILZ expression was analysed by qPCR. n=3-5, pooled from at least three independent experiments. (B) GILZ expression in splenocytes from WT or IFNα-overexpressing mice (GSE123549, n=2). (C) GILZ expression in PBMC from MS patients, before and 4 or 24 hours after injection of 500 µg IFNβ (GSE138064, n=25-26). *p<0.05, **p<0.01, ***p<0.001. ns, Not significant.

Figure 3

Type I IFN inhibits DEX-induced GILZ. (A) Daily prednisolone dose in SLE patients with low or high IFN score (GSE88884, n=1756). (B) GILZ expression in PBMC of SLE patients stratified by daily prednisolone dose (GSE88884, n=1756). (C, D) Healthy PBMC were treated with IFNα, DEX or both, after which GILZ mRNA expression was analysed by qPCR (C), and GILZ protein level was analysed by flow cytometry in the same samples (D) GILZ protein level is represented as MFI normalized against the control MFI for each donor. n=9, pooled from three independent experiments. *p<0.05, **p<0.01, ****p<0.0001. #symbol indicates significance of difference from control (untreated) cells, # p<0.05, ## p<0.01, ### p <0.001, ####p<0.0001.

Figure 4

JAK1/Tyk2 signaling is involved in IFN-mediated inhibition of GILZ. (A) Kidneys from lupus-prone BWF1 mice treated with tofacitinib, DEX or both were analysed for GILZ expression by qPCR (n=9-15). (B–D) Healthy human PBMC were treated with the Jak1/Tyk2 inhibitor tosylate salt (TS) before exposure to IFNα, DEX or both. Expression of pSTAT1 and STAT1 was analysed by western blot (B), and RSAD2 (C) and GILZ (D) by qPCR. n=9-11, pooled from three independent experiments. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. ns, Not significant.

Figure 5

DEX-induced GR binding at a putative GILZ enhancer is reduced after IFNα pre-treatment. (A) DHS linkage predictions are mapped to indicate potential looping regulatory regions connected to the GILZ promoter (19). (B) ChIP-seq data of STAT1 binding in CD14+ monocytes treated with IFNγ (GSE99887), H3K27Ac histone acetylation and GR binding in THP1 cells treated with 100 nM DEX (GSE43036) and GR binding in SUP-B15 cells treated with 100 nM DEX (GSE107584). (C–F) ChIP for GR in L363 cells treated with DEX or IFNα and DEX. Enrichment was assessed in regions (C) A, (D) Z,(E) K and (F) L as indicated in figure B. n=3 from three independent experiments. ns, Not significant.