Functionally decoupled cis-regulatory signals at PCOLCE2 link fibroblast-lineage collagen maturation to pelvic organ prolapse
Background. Pelvic organ prolapse (POP) is a common failure of fibrocollagenous pelvic support. Genome-wide association studies (GWAS) report a sub-genome-wide suggestive association at PCOLCE2 (regional minimum p = 2.5×10⁻⁷), which encodes a secreted regulator of procollagen C-terminal processing, but they do not establish which cell type, which molecular species or which regulatory signal carries the risk. Methods. This was a pre-specified study of a five-gene collagen-assembly panel and of one GWAS-suggestive locus, not an unbiased genome-wide screen, and it used public summary-level and single-cell data only; no functional assay was performed. The analysis combined cis-Mendelian randomisation (MR) with fibroblast expression quantitative trait loci (eQTL) instruments, colocalisation, the regional causal-inference method CAUSE, stratified fine-mapping across four molecular phenotypes, plasma protein QTL (pQTL) deconvolution, Steiger filtering, a fifteen-endpoint specificity panel, single-cell and pseudo-bulk transcriptomics, and exploratory regulatory annotation. Results. Conditional on the sub-genome-wide index signal set out under Background, genetic evidence supports a triangulated statistical causal inference for a fibroblast-lineage cis-regulatory effect at PCOLCE2 that raises POP risk (OR 1.282 per SD, 95% CI 1.112–1.477), with specificity of this cis-regulatory effect to pelvic-floor endpoints, colocalisation (PP4 = 0.954) and a causal rather than shared-confounding architecture (CAUSE ΔELPD = −7.99 ± 1.64). Within a single haplotype block the locus resolves into two functionally decoupled, statistically inferred cis-regulatory profiles (r² = 0.066); the blood-restricted profile controls circulating PCPE-2 but is unrelated to POP, and every variant that raises plasma PCPE-2 falls significantly below the POP effect predicted by a circulating-dosage model. The fibroblast signal also accounts for the locus's bone-density association, which we read as parallel pleiotropy rather than skeletal mediation. PCOLCE2 was expressed selectively in pelvic-floor fibroblasts, and exploratory, under-powered pseudo-bulk analysis (16 versus 5) placed an observed disease-associated expression increment within ECM-remodelling fibroblasts; there the collagen-assembly module was positively but non-significantly displaced (Hedges' g = +0.45, label-shuffling p = 0.38) and only the inhibitor module reached significance, so assembly-biased matrix decay remains a working mechanistic hypothesis. Conclusions. Conditional on independent POP GWAS replication — the index signal is sub-genome-wide significant and every inference here depends on it — the disease-relevant signal at this locus is a fibroblast-restricted cis-regulatory effect on PCOLCE2 rather than circulating protein dosage. More generally, the locus shows that apparent gene-level pleiotropy at matrix genes can resolve into functionally decoupled, tissue-restricted cis-regulatory profiles within a single haplotype block — a principle that changes how such GWAS signals should be read and followed up.
Authors
- Suting LI
- Songming He (ORCID: https://orcid.org/0000-0002-1048-2256)
- 李秉枢
- Li Hong (ORCID: https://orcid.org/0000-0002-8064-1961)
- Hongming Li
- Jinhua Tian
- Jie Min
- Zhi Wang
Institutions
- Renmin Hospital of Wuhan University (CN)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-19
- DOI
- https://doi.org/10.21203/rs.3.rs-11780387/v1
- Primary Topic
- Genetic Associations and Epidemiology
- Type
- preprint