Integrating genomic structural equation modeling and experimental validation to unravel the genetic basis of male genital lichen sclerosus.

To investigate the genetic architecture of male genital lichen sclerosus (MGLSc) and to identify potential susceptibility loci, candidate genes, and biological pathways associated with disease pathogenesis by integrating genomic structural equation modeling (Genomic-SEM) with multi-omics analyses and experimental validation. Publicly available genome-wide association study (GWAS) summary statistics of MGLSc-related traits were integrated to construct a Genomic-SEM framework. Linkage disequilibrium score regression (LDSC) was used to estimate genetic correlations and evaluate model stability. Functional mapping and annotation were performed using FUMA, and novel loci were further screened through a GWAS subtraction strategy. Fine-mapping was conducted using SuSIE and FINEMAP to prioritize candidate causal variants. Transcriptome-wide association study (TWAS) and FOCUS were applied to identify candidate genes. MAGMA-based gene enrichment, partitioned heritability analysis, and polygenic risk score (PRS) analyses were further performed to characterize the biological relevance of associated loci. Finally, RT-qPCR was conducted in vitro to validate the expression of prioritized genes. The Genomic-SEM showed a good overall fit and generated an indirect GWAS framework comprising 2,451,318 SNPs for MGLSc. A total of 208 SNPs reached conventional genome-wide significance, and FUMA annotation identified 43 risk loci, 52 lead SNPs, and 14 candidate genes. Using the GWAS subtraction strategy, 13 novel SNPs were further identified, including rs715299 and rs10774625. Fine-mapping highlighted four high-confidence variants, namely rs3134608, rs3134952, rs3763307, and rs2076524, mainly clustered in the chromosome 6 major histocompatibility complex region. TWAS identified HLA-DPA1 as the most significant gene, and FOCUS further supported its likely causal role. MAGMA and enrichment analyses suggested that immune-related and regulatory regions contributed substantially to MGLSc heritability. PRS analysis demonstrated marked heterogeneity across chromosomes, with chromosome 6 showing the strongest SNP-level contribution. RT-qPCR confirmed that HLA-DPA1 expression was significantly decreased in MGLSc samples compared with controls (P < 0.0001), consistent with the bioinformatics prediction. In addition, several MAGMA-prioritized genes, including C4B, DDR1, BBS7, VARS1, PRRT1, PPT2, EGFL8, BTNL2, and EME1, were downregulated, whereas AGPAT1 and PBX2 were upregulated in MGLSc samples. This study provides a systematic view of the genetic basis of MGLSc by integrating Genomic-SEM, fine-mapping, transcriptomic prioritization, and experimental validation. Our findings indicate that MGLSc is influenced by a shared polygenic architecture enriched in immune-related loci, particularly within the HLA region. HLA-DPA1 emerged as a high-confidence susceptibility gene, and multiple novel loci and candidate genes were identified, offering new insights into the molecular mechanisms underlying MGLSc and potential targets for future mechanistic and translational studies.