Blood-borne extracellular vesicles (i.e. exosomes and microvesicles) carrying microRNAs (miRNAs) could make excellent biomarkers of disease and different physiologic states, including pregnancy status. We tested the hypothesis that circulating extracellular vesicle-derived miRNAs might differentiate the pregnancy status of cows that had maintained pregnancy to Day 30 from non-pregnant cows or from those that exhibited embryonic mortality between Days 17–30 of gestation. Cows were randomly assigned for artificial insemination with fertile semen (n= 36) or dead semen (n=8; control group) on Day 0 (day of estrus). Blood was collected from all animals on Day 0 and on Days 17 and 24 after artificial insemination. Cows receiving live sperm were retrospectively classified as pregnant on Day 30 (n=17) or exhibiting embryonic mortality between Days 17–30 (n=19). Extracellular vesicles from Day-17 and -24 samples were isolated from serum using ultra-centrifugation, and their presence was confirmed by nanoparticle tracking and Western blot analyses (for CD81) prior to RNA extraction. MicroRNA sequencing was performed on pregnant, embryonic-mortality, and control cows (n=4 per day), for a total of 24 independent reactions. In total, 214 miRNAs were identified in serum, 40 of which were novel. Based on differential abundance parameters, we identified 32 differentially abundant loci, representing 27 differentially abundant mature miRNA. At Days 17 and 24, specific miRNAs (e.g. miR-25, -16b and -3596) were identified that differentiated the pregnancy status. In summary, we identified several circulating extracellular vesicles derived miRNAs that differ in abundance between embryonic mortality and pregnant cows.

No abstract

Somatic cell NT (SCNT) efficiency remains poor, preventing the technology from being regularly used in the agricultural industry. It is believed that faulty epigenetic reprogramming of SCNT embryos leads to the low overall success. A clear apoptotic signature is associated with inappropriate gene expression and epigenomic aberrancies in many experimental cell culture systems, and we hypothesised that an apoptosis biomarker could be used to effectively separate properly reprogrammed porcine SCNT embryos from those that are destined to fail due to incomplete reprogramming. Therefore, our objective was to evaluate global gene expression and DNA methylation patterns in high- and low-apoptosis individual embryos in an effort to characterise the extent of genomic reprogramming that had taken place. Porcine SCNT blastocysts on Day 6 of development were stained with a nontoxic, noninvasive caspase activity reporter, and the top and bottom 20% of detected caspase activity were classified as high and low apoptosis, respectively (3 replicate cloning sessions; n=13 embryos per group). Genomic DNA and total RNA were isolated from each individual blastocyst. The RNA sequencing libraries were prepared using the Ovation SoLo RNA-Seq system (NuGen, San Carlos, CA, USA). Reduced representation bisulfite sequencing libraries were prepared for DNA methylation analysis using a modification of the single-cell reduced representation bisulfite sequencing global DNA methylation analysis approach detailed by Guo et al. (2015 Nat. Protoc. 10, 645-59). The RNA sequencing analysis using EdgeR (https://bioconductor.org/packages/release/bioc/html/edgeR.html) revealed 175 total differentially expressed genes (fold change ≥1.5; false discovery rate ≤0.05) between the high- and low-apoptosis SCNT embryos. This list of differentially expressed genes was used to perform enrichment analysis to identify overrepresented Gene Ontology (GO) terms or Kyoto Encyclopedia of Genes and Genomes pathways (DAVID Ease version 6.8 (https://david.ncifcrf.gov/) against the Sus scrofa background genome). However, no significantly enriched GO terms or pathways were identified (false discovery rate P>0.05). Analysis of global DNA methylation patterns between high- and low-apoptosis SCNT embryos using MethylKit (Akalin et al. 2012Genome Biol. 13, R87) revealed 335 differentially methylated 100-bp regions with at least 25% difference in methylation (adjusted P ≤ 0.01). Gene transcription start sites associated with these regions were used for enrichment analysis; again, no significant enrichment of GO terms or Kyoto Encyclopedia of Genes and Genomes pathways was identified. Principal component analysis of CpG methylation showed the low-apoptosis embryos clustering more tightly than the high-apoptosis embryos, which were highly scattered. Ongoing comparisons of high- and low-apoptosis cloned embryos with naturally fertilized embryos produced invivo may provide more information about which embryos were properly reprogrammed. Although we are still pursuing a link between reprogramming and gene expression in high- and low-apoptosis embryos, we conclude that these data support a model of stochastic epigenetic reprogramming following SCNT and reinforce the necessity of identifying embryos most likely to be successful due to proper epigenetic reprogramming in order to increase SCNT efficiency.

The goal of this study was to evaluate global levels of a variety of histone modifications at different lysine (K) residues on Histone 3 (H3) within the chromatin of porcine germinal vesicle (GV)-stage oocytes that were aspirated from follicles of different sizes. We hypothesised that we would see evidence of a transition from open, transcriptionally active chromatin (in oocytes from smaller, growing follicles) to more closed, transcriptionally silent chromatin associated with fully grown oocytes (aspirated from large, preovulatory follicles). Cumulus-enclosed oocytes were aspirated from small (<3 mm) or large (>7 mm) follicles from abattoir-derived pig ovaries. Oocytes were denuded immediately after aspiration and then immunoprobed with antibodies specific for trimethylated (me3) H3K4, H3K9me3, and H3K27me3. Background-corrected nuclear fluorescence levels for each histone mark were collected from multiple oocytes from each of at least three experimental replicates (aspiration days). Data were subjected to one-way ANOVA with a Bonferroni multiple testing correction to determine whether there were differences in fluorescence intensities in the nuclei (germinal vesicles) of oocytes from small v. large follicles. Oocytes from large follicles displayed more intense nuclear staining for all 3 histone marks: average nuclear H3K4me3 intensity was 31.4% higher (P = 0.0004), H3K9me3 was 70.3% higher (P = 0.0218), and H3K27me3 was 32.0% higher (P = 0.0231) in oocytes from large follicles. An ancillary analysis of the data revealed no effect (P > 0.1) of pubertal status (i.e. whether small and large follicles were aspirated from pre- v. post-pubertal ovaries) on the intensity of nuclear fluorescence for any of the marks evaluated. In continuation, 3 oocytes from both follicle types were collected on each of 6 aspiration days (i.e. 18 individual oocytes from each follicle type), and the mRNA from these were used for an RT-qPCR experiment to detect the relative abundance of transcripts from 21 different genes coding for histone methyltransferase or demethylase enzymes in oocytes from large v. small follicles. Of the 21 genes tested, 5 genes (KDM4C, KDM4D, KDM5B, KDM5C, and SETD7) were not detectable in our individual oocyte samples, but transcripts from 6 of the 16 remaining genes (KDM6A, KMT2B, MLL3, SETD1B, SETDB1, and SUV39H2) were shown to be significantly more abundant in oocytes from large follicles (at least 2-fold greater abundance and P < 0.05). Although our expectation-that histone marks (and related transcripts) would consistently reflect a globally “repressive” chromatin configuration in oocytes from large follicles and a more “open” configuration in oocytes from small follicles-turned out to be untrue, the evidence suggests that the epigenetic constitution of oocytes from small follicles may indeed vary from that of oocytes from large, preovulatory follicles, and this phenomenon warrants further investigation.