Integrative Single‐Cell and Spatial Transcriptomics Identifies CRIP1+ Prehypertrophic Chondrocytes as Inflammatory Hubs and a Valuable Diagnostic Biomarker for Osteoarthritis

Abstract

Osteoarthritis (OA) is a prevalent degenerative joint disease characterized by progressive cartilage destruction, yet the heterogeneity of transitional chondrocyte states and their contributions to pathogenesis remain incompletely understood. In this study, we constructed a single‐cell transcriptomic atlas of human knee cartilage from 16 OA samples and 2 non‐OA normal donors. High‐resolution sub‐clustering of the prehypertrophic chondrocytes (preHTC) compartment revealed eight transcriptionally distinct subtypes, among which CRIP1+ preHTC emerged as the most significantly expanded population in OA, particularly in weight‐bearing regions. Pseudotemporal trajectory inference and regulon analysis identified EGR3 as a candidate key regulator for CRIP1+ preHTC. High‐dimensional weighted gene coexpression network analysis (hdWGCNA) demonstrated matrix‐remodeling and proteostasis‐stress transcriptional programs were highly active in CRIP1+ preHTC. Cell‐cell communication analysis uncovered that CRIP1+ preHTC acts as a central signaling hub in OA, with markedly enhanced FN1 signaling pathway. Spatial transcriptomics confirmed that CRIP1+ preHTC co‐localizes with prefibroblasts (preFC) in pathological niches and interacts with other cells in the OA microenvironment. Finally, a machine‐learning‐based 10‐gene classifier (JUN, MCOLN3, RHOC, AKR7A2, HSPG2, AK4, B4GALT2, CLSTN1, LRRC41, and GADD45A) derived from CRIP1+ preHTC‐associated hub genes accurately discriminated OA from normal tissues in two independent bulk RNA‐seq cohorts. Collectively, these findings suggest that CRIP1+ preHTC may represent a pathogenic cell state in OA, provide a multi‐layered molecular framework that links a specific chondrocyte transitional subset to cartilage degradation, and offer a potentially useful gene signature for OA diagnosis.