Tissue-Specific Profiling of E3 Ubiquitin Ligases under Cold Stress: An Integrated in Silico Pipeline for PROTAC Target Discovery

Abstract

Background. Cells use ubiquitination i.e. tagging proteins with the small protein ubiquitin, to control which proteins survive and which get destroyed. This system is well studied for heat shock, low oxygen, and DNA damage, but its role during cold stress is far less clear. A recent review by Sheng and colleagues (2024) highlighted three gaps: we don’t know whether ubiquitination protects cells from extreme cold, we lack tissue-level maps of E3 ligases (the enzymes that pick which proteins get tagged), and only about a dozen E3 ligases can currently be used as anchors for PROTAC drugs. Methods. We built a six-phase in silico pipeline pulling data from public databases (GEO, UniProt, STRING, AlphaFold, ELM, ChEMBL, PROTAC-DB), with Welch differential expression across eight tissues and four cold time-points, tau-index tissue specificity, PPI network analysis, AlphaFold confidence checks, and a composite PROTAC priority score. Results. Starting from 31 reference human E3 ligases and 96 cold-exposure samples, the pipeline flagged 15 cold-responsive E3 candidates. Most were RING-type (10/15) and highly tissue-specific (mean τ = 0.85). HIF1A was the central network hub (betweenness = 0.36). Eight of ten top candidates had AlphaFold mean pLDDT ≥ 70. K48 chains which is the standard proteasome-targeting signal, accounted for 71% of 24 mapped ubiquitination sites. Composite PROTAC ranking placed MDM2 first (0.72), MARCH5 second (0.61, novel target), then PARKIN, SMURF1, and RNF8. All five passed productive ternary-complex geometry checks. Conclusions. MARCH5, RNF34, RNF183, and TRIM59 emerge as cold-relevant E3 ligases not yet exploited as PROTAC recruiters which are fresh handles for drugs targeting cold injury, organ preservation, and brown-fat metabolic disease.