Embryo-derived proteins (PDI, CAPG, PSAT1) are highly conserved among placental mammals and alter the transcriptome of endometrial epithelial cells (bovine and human)

Pregnancy establishment in mammals requires a complex sequence of events, including bi-lateral embryo-maternal communication, leading up to implantation. This is the time when most pregnancy loss occurs in mammals (including humans and food production species) and dysregulation in embryo-maternal communication contributes to pregnancy loss. Embryo-derived factors modify the function of the endometrium for pregnancy success. We hypothesise that these previously unexplored conceptus-derived proteins may be involved in altering the function of the endometrium to facilitate early pregnancy events in mammals with different early pregnancy phenotypes. Here, we show that protein disulphide-isomerase (PDI) is a highly conserved protein among mammals, and provide evidence for a species-specific roles for PDI in endometrial function in mammals with different implantation strategies. We show how PDI alters the endometrial transcriptome in human and bovine in vitro in a species-specific manner, and using a microfluidic approach we demonstrate that it alters the secretome capability of the endometrium. We also provide evidence from in vitro assays using human-derived cells that MNS1, a transcript commonly downregulated in response to PDI in human and bovine endometrial epithelial cells, may be involved in the attachment phase of implantation. We propose that the trophoblast-derived protein PDI, is involved in supporting the modulation of the uterine luminal fluid secreted by the endometrium to support conceptus nourishment, and also in the process of embryo attachment to the uterine lumen for pregnancy success in mammals. Overall design: Recombinant protein production Recombinant bovine PDI (rbPDI), PSAT1 (rbPSAT1), and CAPG (rbCAPG) was produced at the Newcastle University Protein and Proteome Analysis centre using an E. coli pET3A expression vector system and subsequent purification steps. Recombinant ovine IFNT (roIFNT) was kindly gifted by Professor Fuller Bazer for this study, produced in P.pistoris to >80% purity. All recombinant proteins were eluted/purified into PBS and stored at -80?C prior to use. Primary bovine endometrial cell culture Primary bovine endometrial epithelial cells (bEECs) and primary bovine endometrial stromal cells (bESCs) were isolated by enzymatic digestion from late-luteal phase uterine tracts (n=3). bEECs and bESCs were maintained in complete bovine medium (RPMI 1640 [Gibco, Massachusetts, US], 10% charcoal-stripped FBS [PAA Cell Culture Company, UK], 1% antibiotic antimycotic solution [ABAM, Sigma-Aldrich, Missouri, USA]). bESCs at a concentration of 150,000 cells per well (n=5) and bEECs at a concentration of 300,000 cells per well (n=5) were plated into separate 6-well plates in 2 mL complete bovine medium. All cultures were maintained at 37?C/5% CO2 in a humidified incubator. At 70% confluence bEECs and bESCs were treated with one of the following for 24 hr: (1) Media control, (2) Vehicle control (PBS only), (3) 1000 ng/ml rbPDI, 4) 1000ng/mL rbPSAT1, 5) 1000ng/mL rbCAPG, or (6) 1000 ng/ml rbPDI/CAPG/PSAT1 combined with 100 ng/ml roIFNT. Human endometrial cell culture Ishikawa (ECACC 99040201, passage 10) immortalised human endometrial epithelial cells (hEECs) were plated in biological technical triplicate at 300,000 cells per well in a 6 well plate, in 2 mL complete human medium (DMEM/F12 [Gibco, Massachusetts, US], 10% FBS [PAA Cell Culture Company, UK], 1% Glutamine SStreptomycin PPenicillin [Gibco, Massachusetts, US]). The three replicates were from three different flasks of passage 10 cells, which had been split from a single flask of passage 9 Ishikawa cells. All cultures were maintained at 37 ?C/5% CO2 in a humidified incubator. At 70% confluence hEECs were treated with one of the following for 24 hours: (1) Media control, (2) Vehicle control (PBS only), (3) 1000ng/mL rbCAPG (4) 1000 ng/ml rbPDI, or (5) 1000 ng/ml rbPSAT1. RNA extraction and sequencing Following treatment cells were washed with PBS, lysed in 400µL mirVana lysis solution from mirVana RNA extraction kit (Invitrogen, St. Louis, US) and snap frozen. RNA was extracted from the cell pellets using the MirVana RNA extraction kit (Invitrogen, St. Louis, US) as per manufacturer's protocol. Genomic DNA removal was carried out using the DNA free kit (Thermo Fisher Scientific, Massachusetts, US) as per the manufacturer's protocol. RNA libraries were sequenced at the NGS Facility at the University of Leeds where they were prepared and sequenced as per their standard protocols using the Illumina TruSeq Stranded Total kit according to manufacturer's guidelines. The RNA libraries were sequenced using the Illumina NextSeq 500 machine (Illumina, California, USA) with a single end 75 bp length read. RNA sequencing data processing was conducted as follows. Briefly, statistical test for differential gene expression was conducted via DESeq2 with the cut-offs such as log2FoldChange >1 (or <-1) and padj <0.05. For principal component analysis (PCA) plotting of each group of samples, both protein-coding genes and long non-coding RNAs (lncRNAs) with RPKM value = 1 in at least one sample were used and subsequently log2(RPKM+1) transformation and a quantile normalisation were applied. Overrepresentation enrichment analysis of differentially expressed protein-coding gene sets was executed using WebGestalt for gene ontology terms and KEGG pathways. For gene ontology terms, biological process non-redundant datasets were chosen as functional database, and for both types of analyses, significance level was determined by FDR <0.05. Venn diagram analysis was performed using Venny.