EUVIMEDOpdag EUVIMED
Crossref

Skin eQTL –Cerebrospinal Fluid Metabolites–Keloid Axis: An Integrative Multi‐Omics Mediation Mendelian Randomization Study

Jie Xu, Mengqi Bai, Tongyun Liu iD

Chemical Biology & Drug Design · 2026-06 · 107 (6)

DOI: 10.1111/cbdd.70346 ↗

At a glance

Journal article in Chemical Biology & Drug Design, published online on 2026-06-23. Written by Jie Xu and 2 co-authors. Institution: Department of Dermatology, The First Affiliated Hospital of Kunming Medical University, Kunming, Yunnan, China. Major MEDLINE topics: Quantitative Trait Loci, Keloid, Skin. 63 references.

AbstractSubjectsBibliografiske oplysningerCiteRelated studies

Abstract (original language)

Keloids are fibroproliferative dermal disorders characterized by excessive extracellular matrix deposition and a strong tendency to recur, yet the mechanistic links between inherited genetic variation, molecular phenotypes and disease risk remain incompletely understood. We integrated GTEx v8 skin expression quantitative trait loci (eQTL) data, pooled GWAS statistics for 338 cerebrospinal fluid (CSF) metabolites, and two independent European keloid GWAS datasets within a three‐step Mendelian randomization (MR) framework to identify causal mediation pathways from skin gene expression through systemic metabolite levels to keloid susceptibility. Bulk RNA‐seq, single‐cell RNA‐seq and in vitro fibroblast and smooth muscle cell experiments were then used to validate the functional roles of the prioritized genes and metabolites. Four CSF metabolites—sphingomyelin, 1‐stearoyl‐2‐oleoyl‐glycerophosphocholine (SOPC), N‐acetylarginine and X‐23593—were causally associated with increased keloid risk (OR > 1) and replicated across both GWAS datasets. PARP14 emerged as a reproducible upstream risk gene (OR = 1.17, 95% CI 1.02–1.34) coordinating effects across all four mediators, whereas ALDH2 was identified as a SOPC‐mediated protective gene (OR < 1). Single‐cell profiling localized elevated PARP14 expression to smooth muscle cells and reduced ALDH2 expression to a disease‐expanded mesenchymal fibroblast subset, with both signals embedded within canonical pro‐fibrotic signaling programs (TGF‐β, Wnt, Hippo, PI3K–Akt/mTOR, ECM–receptor interaction and focal adhesion). In vitro, SOPC and sphingomyelin produced dose‐dependent fibroblast proliferation and induction of COL1A1, COL3A1, POSTN, and FN1, while modulation of PARP14 and ALDH2 by siRNA and lentiviral overexpression confirmed their roles in regulating proliferation and fibrotic marker expression. Together, these findings delineate a skin eQTL → CSF metabolite → cell‐state axis in keloid pathogenesis, nominating PARP14, ALDH2 and SOPC‐linked lipid remodeling as candidates for mechanistic study and therapeutic targeting.

Subjects

MeSH terms (MEDLINE)

Quantitative Trait LociKeloidSkinHumansMendelian Randomization AnalysisGenome-Wide Association StudyMultiomicsPolymorphism, Single NucleotideFibroblastsSphingomyelins

Author keywords

ALDH2Mendelian randomizationPARP14cerebrospinal fluid metabolomicseQTLfibrosiskeloidmediationsingle‐cell RNA‐seq

Substances

Sphingomyelins

Authors and institutions

  • Department of Dermatology, The First Affiliated Hospital of Kunming Medical University, Kunming, Yunnan, China.
  • Sixth United Lige Medical Cosmetology Hospital of Hangzhou, Hangzhou, Zhejiang, China.

ORCID profiles: Tongyun Liu

Cite

APA

Xu, J., Bai, M., & Liu, T. (2026). Skin eQTL –Cerebrospinal Fluid Metabolites–Keloid Axis: An Integrative Multi‐Omics Mediation Mendelian Randomization Study. Chemical Biology & Drug Design, 107(6). https://doi.org/10.1111/cbdd.70346

Vancouver

Xu J, Bai M, Liu T. Skin eQTL –Cerebrospinal Fluid Metabolites–Keloid Axis: An Integrative Multi‐Omics Mediation Mendelian Randomization Study. Chemical Biology & Drug Design. 2026;107(6). doi:10.1111/cbdd.70346

BibTeX
@article{Xu2026,
  author = {Xu, Jie and Bai, Mengqi and Liu, Tongyun},
  title = {Skin eQTL –Cerebrospinal Fluid Metabolites–Keloid Axis: An Integrative Multi‐Omics Mediation Mendelian Randomization Study},
  journal = {Chemical Biology & Drug Design},
  year = {2026},
  volume = {107},
  number = {6},
  doi = {10.1111/cbdd.70346}
}
Start for free →