Immunosuppression causes dynamic changes in expression QTLs in psoriatic skin

BackgroundPsoriasis is a chronic, systemic inflammatory condition, primarily affecting skin and joints. It is an immune-mediated disease driven by IL-17-producing T cells in the skin. However, epidermal cells such as keratinocytes are now also recognized as essential contributors to pathogenesis. We hypothesized that, under immunosuppressive treatment, skin inflammation alters the effect of genetic variants on gene expression in key pathways and cell types within the skin tissue. Understanding dermal and epidermal cell dynamics in psoriasis may help identify novel therapeutic targets. MethodsDuring the "Psoriasis Treatment with Abatacept and Ustekinumab: A Study of Efficacy" (PAUSE) trial, we obtained longitudinal lesional and non-lesional skin biopsies from participants at baseline, during treatment, and after withdrawal of study medication. Using whole-tissue genome-wide gene expression and genotype data, we mapped expression quantitative trait loci (eQTLs). Using transcriptional data, we defined local skin inflammation status using Skin Psoriatic Inflammation Transcriptional Score (SPITS) comparing lesional and non-lesional baseline skin biopsies. We identified interactions between eQTLs and SPITS status, as well as other features. We then examined the resulting eQTL genes (eGenes) and eQTL SNPs (eSNPs) to detect key pathways containing the eGenes whose regulation was altered by inflammation status. ResultsWe report 953 significant eQTLs at a Bonferroni p-value threshold (0.05/7,475,856). We assessed inflammation status in each biopsy based on SPITS and identified 116 eQTLs that are modified by changes in SPITS (FDR < 0.20); these eQTLs largely overlapped with those that were modified by an IL-17 pathway score, which is defined by genes induced in IL-17 stimulated keratinocytes (97/116=83.62%). In contrast, we detected few eQTL interactions with treatment (N = 0) and psoriasis area and severity index (PASI) score (N = 2 at FDR< 0.05; N = 5 at FDR < 0.20). Using psoriatic skin single-cell gene expression (84/116 found in the single cell dataset), we observed that most SPITS-interacting eGenes are skin cell-specific (62/84=73.81%) rather than immune cell-specific. We found the most significant SPITS interaction eQTL to be rs1491377616-LCE3C, and both the eSNP and eGene are associated with psoriasis genetic risk. Finally, we found SPITS-interacting eSNPs are enriched in NRF2 transcription factor binding motifs and may regulate four eGenes in the NRF2 pathway. ConclusionsThis is the first eQTL study of psoriatic skin in a clinical setting. The study highlights the potential to use biospecimens from clinical trials to discover in vivo eQTL interactions with therapeutically relevant environmental variables. We found that eQTLs are largely modified by local factors rather than systemic factors, such as medication usage or global psoriasis disease activity. We observed that the majority of eQTLs that are modulated by local skin inflammation in psoriasis are expressed in dermal and epidermal cell types and not immune cell types. Our results suggest that the rs1491377616 risk allele may be modulating LCE3C expression in keratinocytes to cause disease. These results suggest that the landscape of changing gene regulation in dermal and epidermal cell types may be an important component of psoriasis, and the dermal and epidermal genes may be therapeutic targets.