Integrative Transcriptomic and Network Pharmacology Analysis Suggests Potential Mechanisms of Propofol in Early-Onset Preeclampsia
Source: PubMed Central Open Access, NCBI / U.S. National Library of Medicine
Purpose To investigate the mechanism of action of propofol in early-onset preeclampsia (EOPE). Patients and Methods Integrated transcriptomics and network approaches were employed to screen disease targets, drug targets, and differentially expressed genes from thetraining set (7 EOPE, 5 controls) andvalidation set (8 EOPE, 8 controls). Signature genes were further selected using machine learning algorithms (LASSO, SVM, Boruta). Immune infiltration levels were evaluated via ssGSEA and their correlations with signature genes were analyzed. Molecular docking validation of binding affinity was performed using CB-DOCK2. The expression patterns of signature genes in EOPE patients were ultimately confirmed using RT-qPCR in a clinical cohort of 36 EOPE patients and 36 healthy pregnant women. Results Nineteen potential intersection targets of propofol intervention in the EOPE were identified. Machine learning analysis further pinpointed three signature genes (,, and). Expression validation and ROC curve analysis confirmed their significant dysregulation, with area under the curve (AUC) values of 1.000, 1.000, and 0.886 in the training set and 0.891, 0.844, and 0.875 in the validation set, respectively, indicating good diagnostic performance. Immunoinfiltration analysis revealed significant alterations in 12 immune cell subsets in EOPE samples, which were closely associated with 3 signature genes. Molecular docking showed negative binding energies (from −5.0 to −5.8 kcal/mol), suggesti
Abstract
Purpose To investigate the mechanism of action of propofol in early-onset preeclampsia (EOPE). Patients and Methods Integrated transcriptomics and network approaches were employed to screen disease targets, drug targets, and differentially expressed genes from thetraining set (7 EOPE, 5 controls) andvalidation set (8 EOPE, 8 controls). Signature genes were further selected using machine learning algorithms (LASSO, SVM, Boruta). Immune infiltration levels were evaluated via ssGSEA and their correlations with signature genes were analyzed. Molecular docking validation of binding affinity was performed using CB-DOCK2. The expression patterns of signature genes in EOPE patients were ultimately confirmed using RT-qPCR in a clinical cohort of 36 EOPE patients and 36 healthy pregnant women. Results Nineteen potential intersection targets of propofol intervention in the EOPE were identified. Machine learning analysis further pinpointed three signature genes (,, and). Expression validation and ROC curve analysis confirmed their significant dysregulation, with area under the curve (AUC) values of 1.000, 1.000, and 0.886 in the training set and 0.891, 0.844, and 0.875 in the validation set, respectively, indicating good diagnostic performance. Immunoinfiltration analysis revealed significant alterations in 12 immune cell subsets in EOPE samples, which were closely associated with 3 signature genes. Molecular docking showed negative binding energies (from −5.0 to −5.8 kcal/mol), suggesting that propofol can spontaneously bind to all core targets with weak‑to‑moderate affinity, and the binding energy with CD68 was the lowest (–5.8 kcal/mol). RT-qPCR validation in the clinical cohort confirmed that the expression patterns of the three signature genes were consistent with the bioinformatics predictions. Conclusion This study identifies,, andas potential targets through which propofol may modulate EOPE-related pathways. Further mechanistic and in vivo validation studies are required.