Co-expression analysis of placental genes in the search for key signaling pathways and biomarkers of the great obstetrical syndromes

Abstract

Objective. To study the molecular mechanisms responsible for the development of diseases grouped within the great obstetrical syndromes (GOS) at the level of the transcriptome of human maternal placenta.Material and Methods. We gathered the results of genome-wide transcriptome studies of the human placental tissue using Gene Expression Omnibus (GEO) data repository for the following phenotypes: physiological pregnancy, preeclampsia (PE), premature birth, and intrauterine growth restriction (IUGR). Eleven data sets were selected and supplemented with our experimental data; a total of 481 samples of human placental tissue were included in the integrative analysis. Bioinformatic data processing and statistical analyses were performed in the R v3.6.1 software environment using the Bioconductor packages. The pooled dataset was used to search for common molecular targets for GOS via weighted gene co-expression network analysis (WGCNA). The functional annotation of genes and the resulting clusters was carried out with the DAVID database; protein-protein interaction network was built using the STRING software; and the hub genes for the network were identified using the MCC analysis with plugin cytoHubba in Cytoscape software 3.7.2.Results. We obtained a table of expression levels for 15,167 genes in 246 samples. Hierarchical clustering of this network allowed to find 55 modules of co-expressed genes in the group with PE, 109 modules in the group with PB, 75 modules in patients with IUGR, and 56 modules in the control group. The preservation analysis of co-expressed modules for the studied phenotypes suggested the presence of a common cluster comprising eight genes specific only for patients with PE and IUGR, as well as the module of 23 co-expressed genes typical only for patients with PB and IUGR. Protein-protein interaction network was built for these gene sets, and the SOD1, TXNRD1, and UBB genes were the central nodes in the network. Based on network topology evaluation with cytoHubba, six hub genes (rank ˂ 5) were identified as follows: SOD1, TKT, TXNRD1, GCLM, GOT1, and ACO1.Conclusion. The obtained results allowed to identify promising genetic markers for preeclampsia, intrauterine growth restriction, and miscarriage. Moreover, the study also made it possible to identify the most important overlapping molecular mechanisms of these diseases occurring in the placental tissue.