Integrative Single-Cell RNA Sequencing and Machine Learning Reveals Candidate Plasma Protein-Associated Gene Signatures for Osteoporosis: A Preliminary Exploratory in Silico Study.

PURPOSE: Plasma proteins are associated with the onset and progression of osteoporosis (OP) and can serve as indicators for this disease. This study aimed to comprehensively analyze and identify plasma protein-associated biomarkers in OP, and clarify their potential molecular mechanisms.

METHODS: Data related to OP and plasma protein-associated genes were retrieved from public databases. Biomarkers were screened and validated via high-dimensional Weighted Gene Co-Expression Network Analysis (hdWGCNA), differential expression analysis, machine learning algorithms, and expression profiling. Additionally, nomogram construction, enrichment analysis, and immune infiltration analysis were performed to further investigate the regulatory roles of biomarkers. Molecular docking was conducted to computationally predict potential drug-target interactions, which require experimental verification. Single-cell RNA sequencing (scRNA-seq) data from one OP patient were used for exploratory examination of biomarker expression patterns in key cell types, providing cellular context for the bulk transcriptomic findings. Finally, reverse transcription quantitative polymerase chain reaction (RT-qPCR) validated biomarker expression in lymphocytes from 6 OP patients and 6 controls, while ELISA measured serum CTSD activity in 10 independent samples per group.

RESULTS: Two biomarkers (JUN and CTSD) were identified as significantly linked to plasma protein in OP. The nomogram developed based on these biomarkers showed potential diagnostic utility that warrants further evaluation. The biomarkers were significantly enriched in the ribosome, proteasome, lysosome, and spliceosome pathways. Notably, JUN showed strong positive correlations with activated dendritic cells and mast cells. Molecular docking further demonstrated that the biomarkers could bind favorably with streptozocin and bruceantin. Moreover, JUN and CTSD exhibited differential expression changes during the differentiation of bone marrow mesenchymal stem cells. Finally, we validated the expression of CTSD and JUN in lymphocytes from OP and control groups, finding that CTSD expression was increased in the OP group compared with the control group.

CONCLUSION: In summary, we identified JUN and CTSD as transcriptome-derived plasma protein-associated gene biomarkers in OP. These findings may provide preliminary insights that could inform future diagnostic and therapeutic investigations, although further validation in larger cohorts and direct proteomic studies are warranted.