Pityriasis rubra pilaris transcriptomics implicate IL-17 signaling and correlate with response to IL-17A inhibition

Pityriasis rubra pilaris (PRP) is a rare inflammatory skin disorder associated with significant patient morbidity. To further characterize the molecular landscape of PRP, we conducted a transcriptomic analysis of epidermis and dermis collected from patients with moderate-to-severe PRP compared to matched healthy controls, on no systemic therapy and then following 24 weeks of treatment with the IL17A inhibitor, ixekizumab. Patients with PRP not only experienced clinical improvements, but also correction of their cutaneous transcriptome dysregulation after therapy. Moreover, those who has >50% improvement in PASI score after treatment demonstrated distinct transcriptomic profiles when compared to those who failed to respond, despite the lack of a discernable phenotypic difference. Pathway enrichment was strongest among DEGs belonging to inflammatory processes for patients with PRP – in particular, enrichment of the Th17 immune response was seen both prior to and following treatment with IL17A inhibition. Positive regulation of natural killer cell chemotaxis was also seen pre- and post-treatment, however, the pathway’s odds of enrichment decreased following treatment in both epidermis and dermis. Additionally, negative regulation of IL-12 production was highly enriched following therapy, suggesting decreased activity of the Th1 axis. Several individual genes of interest similarly displayed marked differential expression by response status. In parallel with their overall normalization in gene expression, responders had a greater normalization toward healthy control skin than non-responders in the Th17 family genes IL17C, IL23A, CCL20, as well as IL36A and IL36F (implicated in pustular psoriasis), Phospholipase A2 group IVD, and the antimicrobial protein S100A7, all of which remained elevated in non-responders. Intriguingly, responders had higher baseline levels of IL17A and IL17F than non-responders. Whereas dermal IL17A/F levels improved in responders, epidermal expression remained high despite the clinical response to IL17A inhibition. In contrast, non-responder IL17A/F epidermal levels normalized and dermal levels were significantly suppressed compared to healthy controls, despite the lack of clinical improvement. This suggests alternative pathways such as specific targeting of IL17C or IL36 may be preferred in some subsets of PRP. The study cohort consisted of 11 adults with moderate-to-severe PRP (as defined by a Psoriasis Area and Severity Index [PASI] ≥ 10) undergoing 24 weeks of IL17A inhibitor therapy and 14 matched healthy controls. All studies were approved by the Oregon Health & Science University Institutional Review Board (#18031), and subjects gave written informed consent. Lesional PRP and matched healthy control punch biopsy specimens were separated into epidermis and dermis, and RNA was extracted, as previously described (Haynes et al., 2020). Short read sequencing assays were performed by the OHSU Massively Parallel Sequencing Shared Resource and postprocessed with standard pair-end trimming via Trimmomatic-0.36 and alignment to the human genome version GRCh38.89 using STAR (v2.5.3a). A counts matrix was generated using FeatureCounts (Subread package v2.0.1) and gencode v38 alignment. Differential expression analysis was conducted with DESeq2 (v1.34.0) (Love et al., 2014) and results filtered by adjusted p value < 0.05 and absolute log2 fold change (log2FC) > 2. Gene expression was compared between healthy controls and PRP patients, with delimitations by tissue depth (epidermis vs. dermis) as well as time on therapy (prior to receiving therapy [pre] vs. week-24 of therapy [post]) and response to therapy (≥50% improvement in PASI [responder] vs. <50% improvement [non-responder]) for PRP patients. Two models were generated, one combining all PRP patients, and another separating responders from non-responders. Pathway enrichment analysis was conducted using Enrichr (adjusted p value < 0.05) (Xie et al., 2021) separated by genes that were increased or decreased in abundance; Gene Ontology (GO) biological process terms (Ashburner et al., 2000) were selected for publication, sorted by the top 3 largest odds ratios per condition. A literature review was also performed to identify individual differentially expressed genes (DEGs) with biological relevance to PRP.