T70. A COMPUTATIONAL TRANSCRIPTOMICS APPROACH TO DRUG REPURPOSING IDENTIFIES NEW THERAPEUTIC CANDIDATES FOR TREATMENT OF CHRONIC PAIN

• Published in: European neuropsychopharmacology Vol. 75; p. S199
• Main Authors: Cote, Alanna, Johnston, Keira J.A., Seah, Carina, Young, Hannah, Charney, Alexander W., Huckins, Laura
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.10.2023
• ISSN: 0924-977X; 1873-7862
• DOI: 10.1016/j.euroneuro.2023.08.354

Chronic pain is defined as pain persisting for greater than three months, or beyond the time of expected healing. Chronic pain is a common condition, with recent incidence estimates exceeding that of well-known conditions like diabetes, depression, and hypertension. Despite its high prevalence and distressing physical and psychological symptoms, research is lacking in appropriate long-term treatment for chronic pain, and chronic pain persists at high rates even with intervention. Recent GWAS indicate that chronic pain can be studied as a distinct neuropsychiatric illness with genetic risk. We used a validated functional genomics approach to drug repurposing, called signature mapping, to identify drug candidates for chronic pain. In a signature mapping analysis the transcriptomic effects of disease and drug perturbations are compared, and drugs with opposite effects on gene expression as the disease are nominated as therapeutic candidates. This phenotype-driven, whole transcriptome approach to drug discovery can yield additional insights beyond the single receptor- or molecule-level investigation in traditional drug development pipelines. We applied three chronic pain disease expression signatures: 1) presence of chronic pain, 2) severity of chronic pain, and 3) resolution of acute pain. Transcriptomic signatures for presence and severity of chronic pain were obtained by summary-based transcriptome-wide association studies (TWAS) of GWAS for chronic pain phenotypes. We performed TWAS using S-PrediXcan across 17 blood, skeletal muscle, and central nervous system tissue models built from the Genotype-Tissue Expression Project Project, Depression Genes and Networks, the CommonMind Consortium, and PsychENCODE reference datasets. Expression signatures for resolution of acute pain were taken from a differential expression study of acute low back pain over a three-month period. Our hypothesis was that drugs with similar effects on expression to the resolution of acute pain may be protective against chronic pain development. For our drug gene expression signatures, we used signatures downloaded from the CMap LINCS 2020 Resource, including 89,878 signatures across 7,668 compounds and 156 cell types. We performed the signature mapping analysis using an ensemble connectivity score combining five popular individual scores: the weighted connectivity score and four eXtreme scores (Xsum, XCor, XSpe, and XCos). The signature mapping analysis nominated 1,354 putative drug candidates across all three disease signatures (resolution of acute pain: 927, severity of chronic pain: 544, presence of chronic pain: 408). 308 of 1,354 candidate compounds are prescribable in the United States. This study nominates drug candidates across many RxNorm-defined classifications, including anti-diabetic medications, antithrombotic agents, analgesics, diuretics, sex hormones and modulators of the genital system, psychoanaleptics, psycholeptics, and antineoplastic agents. Drug repurposing can be a fast and cost-effective alternative to drug discovery for the pharmaceutical treatment of complex diseases. Interim findings highlight new genetically-informed avenues for drug treatment of chronic pain conditions. We are currently working to validate these drug candidates using the Mount Sinai electronic health record system, testing for relationships between exposure to each candidate medication and chronic pain prevalence or pain symptom severity.