High interferon response signatures in SLE patient leukocytes are associated with increased transposable element expression in gene introns and intergenic regions.

backgroundSystemic lupus erythematosus (SLE) is a spontaneous systemic auto-immune condition for which the inciting factors and genetic basis are generally unknown. Although heterogeneous in its manifestations and severity, SLE involves chronic inflammation along with sustained autoantibody production. The root causes and pathophysiology of the inflammation and breaches of tolerance are incompletely understood, but neutrophils are thought to be important elements of the pathophysiology. Type I interferons (IFN) in the bloodstream and an IFN-stimulated gene (ISG) signature in circulating leukocytes, including neutrophils, are common features in many patients. Earlier work has provided evidence of increased levels of transcripts derived from transposable elements (TEs) in peripheral blood cells of SLE patients. Using six leukocyte types, including neutrophils, we tested the correlation of TE expression with disease severity and explored the relationships between increased ISG and TE expression with attention to the genomic locations of the expressed TEs.

resultsWe reanalysed previously published data from neutrophils and other leukocytes of SLE patients sub-divided into ISG-high (termed IFNpos, n = 12) and ISG-low (termed IFNneg, n = 11) patients in the original study, examining RNA-seq data from B and T lymphocytes, conventional and plasmacytoid dendritic cells (DC), monocytes and PMN of IFNpos and IFNneg SLE patients compared to healthy controls. SLE patients pre-stratified as IFNneg showed no significant increase in TE expression. All IFNpos cell types had similar amounts of total TE-encoded RNA, but among the 6 cell types, PMN had the highest number of differentially expressed TEs and ISGs in IFNpos SLE patients compared to healthy controls. There was a strong correlation between expression of several specific TE families and disease activity assessed at the time of the visit. Although most upregulated TEs (∼ 80%) were present in introns of upregulated genes, a substantial portion (∼ 20%) were not. Of the upregulated genes linked to upregulated TEs, approximately 1/3 were not ISGs whereas ∼ 67% were ISGs. By mapping expressed TEs in ISGs, we found that high intronic TE expression correlated strongly with increased ISG expression as well as with splicing alterations in annotated exons flanking expressed TEs. Consistent with autonomous TE expression, upregulated TEs were also observed at intergenic sites distant from annotated genes, perhaps due to weakening of heterochromatin integrity.

conclusionsHere we describe a strong association between increased TE expression and an ISG-based gene network signature in multiple types of leukocytes centrally involved in SLE pathogenesis. The subset of SLE patients with higher ISG expression also showed high TE expression, compared to healthy donors and SLE patients with low ISG expression. Our data do not offer mechanistic insights into how TE expression might be linked to SLE pathogenesis or ISG expression. Rather, they emphasize the need for long-reads sequencing to understand the causes and consequences of high TE expression in SLE and other autoimmune/inflammatory disorders. Important questions for future study include whether TE expression in introns of ISGs and other genes is independently regulated or reflects exonization or partial intron retention, and how frequently TE expression correlates with splicing variations in adjacent exons. Although limited by short-read RNA-seq technology, our analyses support selective regulation of many TEs independent from the regulation of conventional genes, thereby suggesting (but not directly demonstrating) that an autonomous mechanism of TE de-repression as well as altered splicing patterns may each be factors in SLE pathogenesis. Important questions for future study include whether TE expression in introns of ISGs and other genes is independently regulated or reflects exonization or partial intron retention, and how frequently TE expression correlates with splicing variations in adjacent exons.