Uncovering Hidden Mechanisms of Different Prescriptions Treatment for Osteoporosis

Osteoporosis (OP) is a systemic disease susceptible to fracture due to the decline of bone mineral density and bone mass, the destruction of bone tissue microstructure, and increased bone fragility. At present, the treatments of OP mainly include bisphosphonates, hormone therapy, and RANKL antibody therapy. However, these treatments have observable side effects and cannot fundamentally improve bone metabolism. Currently, the prescription of herbal medicine and their derived proprietary Chinese medicines are playing increasingly important roles in the treatment of OP due to their significant curative effects and few side effects. Among these prescriptions, Gushukang Granules (GSK), Xianling Gubao Capsules (XLGB), and Er-xian Decoction (EXD) are widely employed at the clinic on therapy of OP, which also is in line with the compatibility principle of "different treatments for the same disease" in herbal medicine. However, at present, the functional interpretation of "different treatments for the same disease" in herbal medicine still lacks systematic quantitative research, especially on the detection of key component groups and mechanisms. To solve this problem, we designed a new bioinformatics model based on random walk, optimized programming, and information gain to analyze the components and targets to figure out the Functional Response Motifs (FRMs) of different prescriptions for the therapy of OP. The distribution of high relevance score, the number of reported evidence, and coverage of enriched pathways were performed to verify the precision and reliability of FRMs. At the same time, the information gain and target influence of each component was calculated, and the key component groups in all FRMs of each prescription were screened to speculate the potential action mode of different prescriptions on the same disease. Results show that the relevance score and the number of reported evidence of high reliable genes in FRMs were higher than those of the pathogenic genes of OP. Furthermore, the gene enrichment pathways in FRMs could cover 79.6, 81, and 79.5% of the gene enrichment pathways in the component-target (C-T) network. Functional pathway enrichment analysis showed that GSK, XLGB, and EXD all treat OP through osteoclast differentiation (hsa04380), calcium signaling pathway (hsa04020), MAPK signaling pathway (hsa04010), and PI3K-Akt signaling pathway (hsa04151). Combined with experiments, the key component groups and the mechanism of "different treatments for the same disease" in the three prescriptions and proprietary Chinese medicines were verified. This study provides methodological references for the optimization and mechanism speculation of Chinese medicine prescriptions and proprietary Chinese medicines.