A compact five-gene immuno-fibrotic profile distinguishing systemic sclerosis subtypes with PXDN validated as a...

Systemic sclerosis (SSc) is a heterogeneous autoimmune disease characterized by a paucity of reliable biomarkers for accurate diagnosis and subtype stratification. The considerable clinical variability between diffuse and limited cutaneous subtypes underscores an urgent need for molecular tools that can dissect this heterogeneity and guide therapeutic strategies.

To address this, we employed a multi-step computational approach. Initially, Weighted Gene Co-expression Network Analysis (WGCNA) was applied to the GSE130955 dataset to identify disease-associated modules. This was followed by the application of three machine-learning algorithms-LASSO regression, random forest, and SVM-RFE-to the GSE181549 dataset for hub gene selection. Immune cell infiltration was estimated using CIBERSORT, and the expression of candidate genes was validated at single-cell resolution using the GSE138669 dataset. Experimental validation was performed for the top candidate, PXDN, assessing its protein expression in human fibroblasts, SSc patient skin, and a bleomycin-induced murine fibrosis model. Finally, molecular docking with AutoDock Vina was conducted to evaluate the binding affinity of small molecules to PXDN.

WGCNA identified a module strongly correlated with SSc status (r = 0.73, p < 2e-16), which was enriched in immune chemotaxis and extracellular matrix organization pathways. The convergence of the three machine-learning algorithms nominated a five-gene signature (CPXM1, ELN, GSTM5, PXDN, PDE7B), which demonstrated high diagnostic accuracy (AUC = 0.985, 95% CI: 0.965-1) and effectively distinguished diffuse from limited cutaneous SSc. Single-cell analysis confirmed predominant expression of these genes in fibroblast and macrophage populations within SSc lesional skin. Experimentally, PXDN was found to be upregulated by TGF-β1 in human fibroblasts and was significantly elevated in skin and lung tissues from SSc and IPF patients, as well as in the bleomycin-induced mouse model. Although molecular docking nominated Protokylol hydrochloride as a compound satisfying distal-cavity geometric criteria, this serves as a hypothesis-generating finding rather than a confirmed inhibitor.

Our integrative analysis identifies a concise immunofibrotic gene signature that robustly distinguishes SSc and its major subtypes. This signature highlights a potential nexus of immune-stromal interactions that may underlie disease heterogeneity, offering candidate biomarkers for molecular stratification and therapeutic targeting. Notably, PXDN emerges as a particularly promising target, warranting further experimental investigation to validate its functional role and therapeutic potential in SSc.