52 Determining changes in blood transcriptome during pregnancy in the cow

During pregnancy, the maternal immune system experiences profound changes to tolerate fetal growth, triggering an adaptation response in circulating blood cells. These alterations can be estimated through the peripheral blood transcriptome, composed of blood cells and cell-free fetal RNA in women. In species with superficial placentation, such as cattle, previous studies have shown that the blood transcriptome also differs at certain time points, but how these adjustments occur across gestation is still unknown. Therefore, this study aimed to apply a bioinformatic approach to determine changes in gene expression during cow gestation using publicly available RNA sequencing (RNA-seq) data from multiple studies. Data integration was done on five different studies (GSE146041, GSE103628, GSE210665, GSE230294, GSE179946), covering gestational ages from 0, 21, 42, 55, and 105 days, through R packages. Batch effects were removed with ComBat-seq (sva), and normalization was done using RUVg (RUVSeq), which uses negative control genes to remove unwanted variation. For this, non-differentially expressed genes between pregnant and nonpregnant samples were used. A principal component analysis demonstrated that the studies were now separated based on gestational stage rather than batch (each study). Next, a time course gene expression analysis was done using maSigPro. Genes were fitted using a quartic regression model, and differentially expressed genes were found with a Benjamini-Hochberg false discovery rate (FDR) correction (Q < 0.05). Hereafter, stepwise regression was used to retain the most significant genes, which were grouped into six clusters according to their similar expression profiles through hierarchical clustering. Functional analysis of the genes within each cluster was done using the DAVID database to determine enriched ontological terms (FDR < 0.05). The clusters contained 203, 82, 154, 31, 42, and 35 genes. Cluster 2 (n = 82) contained genes that showed an elevated expression at Day 21 (embryonic period) compared with Day 0 and all the other gestational ages, and this cluster was involved in immunological pathways related to interferon response. Genes in Cluster 6 (n = 35) exhibited a clear pattern of increased expression during the fetal period (Days 42, 55, and 105) compared with Days 21 and 0. This gene cluster was strongly enriched for oxidative phosphorylation (P = 5.9e-22), a pathway found to be involved with a maternal blood gene signature positively correlated with fetal weight at Day 42 (Rabaglino et al. 2023 Biol. Reprod. 109, 749–758). In conclusion, this study shows the potential of using publicly available RNA-seq data to determine molecular signatures in the blood that underlie the complexity of pregnancy in the cow. Integrating transcriptomic data is challenging due to inherent variations among studies. While these differences cannot be fully excluded, we designed the bioinformatics pipeline to minimize those batches and retain the biological information from each data set. Future investigations and validations can pinpoint blood changes related to fetal growth or fetal characteristics that could be used to track fetal development or even predict calf health.