Influence of sire fertility on the metabolism of in vitro produced embryos

Bull field fertility varies greatly and is largely influenced by differences in early embryo survival rather than fertilization failure alone. We, therefore, investigated whether sire fertility status is reflected in the metabolomic profile of embryo-conditioned culture media. Six Holstein-Friesian bulls were selected from a population of 840 sires and classified using an adjusted sire fertility index. Mean (± SD) values were 4.27 ± 0.61 % for highly fertile (HF; n=3) and -13.1 ± 1.85 % for sub-fertile (SF, n=3) bulls. Cumulus-oocyte complexes were fertilized using frozen-thawed semen from individual bulls. On Day 7 post-fertilization, Grade 1 blastocysts (pools of n=15 per bull; n=3 replicates) were transferred to fresh culture medium for 24 h. On Day 8, embryo-conditioned media from both fertility treatments and unconditioned media were collected, snap-frozen, and subjected to high-throughput untargeted metabolomic profiling. A total of 149 metabolites were detected. Unsupervised analysis revealed substantial overlap between HF and SF sire derived embryo-conditioned media, while both differed from unconditioned media. The relative abundance of 30 metabolites was altered by embryo presence, irrespectively of sire fertility. Direct comparison of HF vs. SF sire derived embryo-conditioned media identified only three differentially abundant metabolites. However, multivariate machine learning analysis identified a panel of eight N-acetylated amino acids which discriminated embryos derived using semen from bulls of divergent fertility status with high accuracy (AUROC=0.903). In conclusion, sire fertility status imparts a subtle but detectable metabolic imprint on the early embryo which can be exploited to non-invasively discriminate embryos according to potential developmental competence.