Yi-qi-yang-yin decoction ameliorates diabetic retinopathy: New and comprehensive evidence from network pharmacology, machine learning, molecular docking and molecular biology experiment

Yi-Qi-Yang-Yin Decoction (YQYY), a traditional Chinese medicine (TCM) formula, has been used to treat diabetic retinopathy (DR), yet its precise mechanisms of action remain poorly understood. In this study, two distinct diabetic models, namely spontaneous type 2 diabetic db/db mice and streptozotocin (STZ)-induced type 1 diabetic rats, were employed to assess the efficacy of YQYY in ameliorating DR. Ultra-high performance liquid chromatography-quadrupole time-of-flight mass spectrometry (UPLC-QTOF-MS) was employed to identify the chemical composition of YQYY in mouse serum. Next, the possible targets and key pathways of YQYY for the management of DR were predicted by integrating network pharmacology and weighted gene co-expression network analysis (WGCNA). Network-based indicators were then employed to evaluate the efficacy of the formulae on DR, and molecular docking, along with compound similarity analysis, was used to identify candidate drugs of YQYY for DR. Finally, molecular biology techniques were utilized to experimentally validate the identified targets. Experimental results from animal models showed that YQYY effectively improved hemoglobin A1C (HbA1C), reduced vessel branch points, and mitigated retinal tissue injury in both DR models. 17 herbal components were identified in the YQYY-containing serum by UPLC-QTOF-MS. Network pharmacology predicted 44 common targets of YQYY involved in the regulation of DR. These targets were found to mainly participate in inflammation-related signaling pathways, including the NF-κB signaling pathway, toll-like receptor signaling pathway, IL-17 signaling pathway, and TNF signaling pathway. By integrating the most relevant disease templates for DR with network pharmacology, we preliminarily identified two key functions of YQYY and their associated regulatory targets, which showed strong connections and correlations with the targets identified in the screened DR disease models. These results demonstrate the pivotal role of core targets, such as BAX, BCL2, MMP9, SIRT1, PPARγ, VCAM1, PTGS2, TNF-α, and RELA, in mediating the therapeutic effects of YQYY in managing DR. Network analysis of YQYY efficacy in DR revealed a significant correlation between the YQYY targets and DR-related genes. Furthermore, molecular docking and drug similarity comparisons suggested that kaempferol, formononetin, and caffeic acid show potential as therapeutic candidates for DR. Our investigation demonstrated the therapeutic efficacy of YQYY against DR, shedding light on novel perspectives regarding the active constituents and molecular pathways through which YQYY exerts its effects in managing DR.