A β-TCP/PMMA composite bone cement promotes osteogenic differentiation of osteoporotic rBMSCs via Wnt/β-catenin...

Osteoporosis-induced bone defects represent a severe clinical challenge that requires high-performance bone repair biomaterials. To address the limited osteoinductivity, excessive curing heat, and poor degradability of conventional PMMA bone cement, we developed a 1:1 β-tricalcium phosphate (β-TCP)/polymethyl methacrylate (PMMA) composite and an integrated computation-experimental framework. For the first time, we systematically clarified the osteoinductive mechanism of this composite in osteoporotic bone repair via a multidisciplinary strategy combining machine learning, multi-omics analysis, and cellular validation, instead of focusing solely on material composition optimization. Characterization showed that the composite had a uniform porous structure, suitable mechanical properties, low curing temperature, high degradation rate, and sustained calcium ion release. The high-accuracy random forest-convolutional neural network (RF-CNN) model predicted osteogenic potential and identified calcium ion release and pore structure as dominant osteogenic factors. Bioinformatics analysis screened 326 differentially expressed genes (DEGs), mainly enriched in the Wnt/β-catenin pathway, with Runx2, Bmp2, and Sp7 as key hub genes. Subsequent in vitro experiments validated that the composite significantly promoted proliferation, osteogenic differentiation, and mineralization of osteoporotic rat bone marrow mesenchymal stem cells (rBMSCs), reduced apoptosis, and upregulated core osteogenic genes by activating this pathway. This study provides a promising biomaterial for osteoporotic bone repair and a novel computation-assisted paradigm for bone regenerative material development.