Background and objectives Upregulation of type I IFN response genes is described for several autoimmune diseases, including systemic lupus erythematosus (SLE), multiple sclerosis (MS), myositis (IIM) and rheumatoid arthritis (RA). While IFNα contributes to SLE pathology, IFNβ therapy is beneficial in MS, implying different immunoregulatory roles for these IFNs. This study aimed to investigate potential diversification of IFNα-and IFNβ-mediated response programs in autoimmune diseases.
Materials and methods Peripheral blood gene expression of 23 prototypical type I IFN response genes (IRGs) was determined in healthy controls (HCs, n = 54), SLE (n = 47), IFNβ-treated MS (n = 71), untreated MS (n = 160), IIM (n = 78) and RA patients (n = 76). An IFN score was calculated and patients with a type I IFN signature (IFN score >mean+2SD in HC) were selected for analysis.
Results We identified IFNα- and IFNβ-specific response programs by comparing IRG expression in SLE (IFNα-mediated) and IFNβ-treated MS patients. Statistical testing revealed increased expression of 5 IRGs in SLE (Gene cluster (GC-)A, p ≤ 0.006) and 13 IRGs in IFNβ-treated MS (Gene cluster (GC-)B, p ≤ 0.044). Concordantly, the GC-A/GC-B log-ratio was positive for all SLE patients and negative for virtually all IFNβ-treated MS patients.
Applying this information to other autoimmune diseases, IIM patients displayed positive GC-A/GC-B log-ratios, indicating predominant IFNα activity. The GC-A/GC-B log-ratios in RA were lower and approached zero in part of the patients, implying relative importance of both clusters. Remarkably, GC-A/GC-B log-ratios appeared most heterogeneous in IFNβ-naive MS; half of the patients displayed GC-A dominance, whereas others showed GC-B dominance or log-ratios near zero. We confirmed these results for SLE, RA and MS using two public microarray datasets containing 22 SLE patients, 112 RA patients, 58 IFNβ-treated MS patients and 62 untreated MS patients.
Exploring functional differences between GC-A and GC-B genes revealed enrichment of the ISGF3-binding response element ISRE and IRF8 binding sites only in GC-B, suggesting that IFNβ is more potent in activating these transcription factors than IFNα.
Conclusions Conclusively, we demonstrated that type I IFN signatures in autoimmune diseases appear less uniform than generally assumed. The diversification of the type I IFN response in autoimmune diseases suggests different pathogenic roles of the type I IFNs.