ObjectiveThe relationship between epilepsy and glioma has long been widely recognized, but the mechanisms of interaction remain unclear. This study aimed to investigate the shared genetic signature and treatment strategies between epilepsy and glioma. MethodsWe subjected hippocampal tissue samples from patients with epilepsy and glioma to transcriptomic analysis to identify differential genes and associated pathways, respectively. Weight gene co-expression network (WGCNA) analysis was performed to identify conserved modules in epilepsy and glioma and to obtain differentially expressed conserved genes. Prognostic and diagnostic models were built using lasso regression. We also focused on building transcription factor-gene interaction networks and assessing the proportion of immune invading cells in epilepsy patients. Finally, drug compounds were inferred using a drug signature database (DSigDB) based on core targets. ResultsWe discovered 88 differently conserved genes, most of which are involved in synaptic signaling and calcium ion pathways. We used lasso regression model to reduce 88 characteristic genes, and finally screened out 14 genes (EIF4A2, CEP170B, SNPH, EPHA4, KLK7, GNG3, MYOP, ANKRD29, RASD2, PRRT3, EFR3A, SGIP1, RAB6B, CNNM1) as the features of glioma prognosis model whose ROC curve is 0.9. Then, we developed a diagnosis model for epilepsy patients using 8 genes (PRRT3, RASD2, MYPOP, CNNM1, ANKRD29, GNG3, SGIP1, KLK7) with area under ROC curve (AUC) values near 1. According to the ssGSEA method, we observed an increase in activated B cells, eosinophils, follicular helper T cells and type 2T helper cells, and a decrease in monocytes in patients with epilepsy. Notably, the great majority of these immune cells showed a negative correlation with hub genes. To reveal the transcriptional-level regulation mechanism, we also built a TF-gene network. In addition, we discovered that patients with glioma-related epilepsy may benefit more from gabapentin and pregabalin. ConclusionThis study reveals the modular conserved phenotypes of epilepsy and glioma and constructs effective diagnostic and prognostic markers. It provides new biological targets and ideas for the early diagnosis and effective treatment of epilepsy.
Read full abstract