Gene profiles of peripheral white blood cells as potential predictors of pregnancy in embryo-recipient heifers

BACKGROUND: The bovine endometrium undergoes cellular, molecular, and functional changes to support embryo survival and development to term. These changes involve a finely coordinated series of events at both local and systemic levels. We postulated that circulating white blood cells (WBCs) may provide valuable biomarkers to predict pregnancy success in heifers when embryo transfer is used sampled from heifers during a preconception cycle and a conception cycle before embryo transfer. WBCs were isolated using PAXgene Blood RNA tubes collected from heifers on Days 7 and 14 of a preconception estrous cycle (PCD7 and PCD14) and, after a rest cycle, on Day 7 of the subsequent estrous cycle (1ETD7) just before embryo transfer. Circulating progesterone and oestradiol-17b were assayed and pregnancy was confirmed by either the delivery of a healthy calf or uterine flushing and conceptus collection on Day 18 post-estrus. Using a custom bovine gene expression microarray representing 19,479 unique transcript, comparison of transcriptomes between heifers classified as non-pregnant or pregnant revealed 1,240, 896 and 1,023 differentially expressed genes (DEGs) at PCD7, PCD14 and 1ETD7 respectively. Our bioinformatics analyses revealed that pregnancy failure after embryo transfer was associated with upstream regulators, biological functions, canonical pathways and gene networks related to inflammation, immunity, apoptosis and cell death regulation, cell proliferation, membrane compounds, lipid metabolism, oxygen transport and ions transport. The heifers classified as non-pregnant showed significant increased transcripts levels of PTGR1 at the three time points (PCD7, PCD14 and 1ETD7), AIF1 at PCD14, FNDC3B, IL15 and SERPINE1 at 1ETD7. Our findings highlight the potential of peripheral WBCs as a non-invasive source of biomarkers for predicting pregnancy outcomes, offering promising insights for improving pregnancy success when reproductive biotechnologies are used in mammalian females. Transcriptional profiling was conducted using a custom 8x60K bovine gene expression microarray, designed based on annotated bovine Ensembl transcripts (http://www.ensembl.org/index.html, genome assembly UMD3.1) represented 19,479 transcripts, previously as described (Raliou, Dembele et al. 2019). RNA labeling was done with total RNA isolated from blood collected from 7 females classified as non-pregnant and 4 pregnant at PCD7 and PCD14. Due to blood sampling issues, at 1ETD7, microarray hybridization was carried out on total RNA from WBCs collected from 5 non-pregnant and 3 pregnant animals (Table 1). RNA labeling followed the Agilent “One-Color Microarray-Based Gene Expression Analysis” protocol (Raliou, Dembele et al. 2019). The microarray data were submitted to the GEO database (accession number GSE185325). Differentially expressed genes (DEGs) were identified using the fold change rank ordering (FCROS) method (Dembele and Kastner 2014) and t-test with the TREAT option of the LIMMA method (Smyth 2004), comparing non-pregnant with pregnant females. p-values were adjusted using the Benjamini & Hochberg method (Benjamini 1995), and a threshold was applied to select significant DEGs.