168 An intronic variant that activates cryptic splicing of the adenylate kinase 9 gene causes extreme subfertility in bulls

A significant range in field fertility exists among bulls used in artificial insemination (AI). One extreme example is a Holstein Friesian bull released in the Irish market in Spring 2020 who, despite passing stringent quality control checks, had a phenotypic pregnancy rate of 18% from 2,976 AIs. Subsequent sperm evaluation using computer-assisted sperm analysis and flow cytometry did not reveal any difference in sperm motility, kinematics, viability, acrosome integrity, membrane fluidity, mitochondria membrane potential, superoxide production, DNA fragmentation, or protamine content relative to a control high fertile bull (same throughout the experiment). However, impaired caffeine-induced hyperactivation was observed. Whole-genome sequencing from semen and subsequent reference-guided variant detection revealed an intronic variant in adenylate kinase 9 (AK9) as a candidate causal variant. Sequencing of RNA extracted from testicular tissue revealed that most of the AK9 transcripts were affected by impaired splicing, which leads to a premature termination codon and a severely truncated protein. The aim was to fully characterise the subfertility phenotype associated with the identified mutation causing extreme subfertility. Data were analysed by chi-squared analysis (proportions) and Student’s t-test (count data). In vitro fertilisation (IVF) revealed compromised cleavage rate (41.6 vs 91.2%), cell cycle kinetics (%, 5- to 8-cell embryos at 48 h post-insemination [hpi] 6.4 vs 51.8%) and blastocyst yield (7.8 vs 36.8%) compared with a control fertile bull (P < 0.05). To assess fertilisation rate, presumptive embryos were fixed at 17 hpi (n = 552) to assess pronuclear formation and at 48 hpi (n = 166) to assess embryo cell number and number of sperm bound to the zona pellucida (ZP). At 17 hpi, the majority (95.9%) of oocytes were unfertilised and arrested in metaphase II of meiosis in the subfertile bull while 74.3% of oocytes in the control had two pronuclei (P < 0.05). At 48 hpi, % > 5-cell embryos was higher in the control (67 vs 0.07%; P < 0.05). Combining the data from both time points, mean (± s.e.m) number of bound sperm (1.4 ± 0.1 vs 0.2 ± 0.04) and % structures with at least one sperm bound (63.8 vs 21%) were higher in the control compared with the subfertile bull, respectively (P < 0.05). Finally, to assess the ability of sperm to reach the site of fertilisation and bind with the oocyte in vivo, a superovulation model was employed. Heifers were superstimulated with 1,500 IU of equine chorionic gonadotrophin and inseminated with frozen-thawed semen from the subfertile bull or a control bull (12 heifers/bull). The reproductive tract of each heifer was flushed postmortem on Day 3.5 to 4 (Day 0 = time of expected ovulation). Cleavage rate (78.2 vs 4.6%; P < 0.001), % > 5-cell embryos (70.8 vs 0.03%; P < 0.05), and mean number of bound sperm (1.4 ± 0.4 vs 0) were higher for the control versus the subfertile bull, respectively. In conclusion, subfertility due to cryptic splicing in AK9 impairs fertility through altered sperm binding and penetration of the bovine ZP.

This work was funded by Science Foundation Ireland (16/IA/4474).