Background Genome-wide association studies (GWAS) have identified single nucleotide polymorphisms (SNPs) that tag genomic sites with a high statistical association with a diagnosis of SLE. As most risk SNPs mark regulatory rather than coding regions of the genome, the functional impact of the risk loci on molecular mechanisms and pathways has not been clearly defined. In addition, how disease-associated loci differentially impact pathogenic mechanisms in females and males with SLE has not been explored. We analyzed RNA sequencing data relevant to previously identified risk loci to understand the molecular pathways impacted by genetic variation.

Methods Forty subjects (10 SLE females; 10 SLE males; 10 healthy females; 10 healthy males) were identified for study, with careful matching of SLE and healthy donor subjects for age and ethnicity. Next generation RNA sequencing was performed and unsupervised clustering of transcript expression displayed for genes previously identified in GWAS as near SNPs associated with a diagnosis of SLE. Identified pathways were also investigated using an expanded RNAseq dataset, including 30 patients with lupus nephritis, 29 patients with non-renal SLE and 3 healthy donors.

Results Four major clusters of gene transcripts were identified, along with multiple subclusters. The most striking transcript subcluster preferentially expressed in lupus males identified mechanisms involved in nucleic acid sensing and type I interferon production (IRF7, IFIH1, IKBKE, TLR7, JAK2, CXorf21/TASL, and SLC15A4). Another subcluster that favored males included T and B cell transcripts associated with generation of autoantibody producing B cells (ITGAX, IRF5, SH2B3, TET3, NOTCH4, ICAM4). The cluster with transcripts showing decreased expression in SLE included DEF6 and PHRF1, potentially contributing to impaired transcriptional regulation of lymphocyte function in patients. Genes located in the MHC (ATF6B, NOTCH4, MICB, TNXB, ITPR3) or an X chromosome risk locus (IRAK1, TMEM187, MECP2, NAA10, HCFC1) were distributed among different transcript subclusters, suggesting the broad impact of risk haplotypes at those loci on regulation of multiple disease-associated pathways. Of interest is the observation that PHRF1 and IRF7, adjacent genes identified by a risk SNP (rs4963128), distributed to distinct clusters, with PHRF1 decreased and IRF7 increased in SLE patients. The expanded SLE dataset was also analyzed to further characterize lupus risk-associated functional transcript clusters.

Conclusion Genes located near lupus risk SNPs are differentially expressed in patients with SLE and cluster based on functional relationships. Transcripts differentially expressed in male patients suggest important involvement of nucleic acid sensing pathways in their disease.