Preimplantation Genetic Testing for Aneuploidy Improves Live Birth Rates with In Vitro Produced Bovine Embryos: A Blind Retrospective Study

Abstract

Approximately one million in vitro produced (IVP) cattle embryos are transferred worldwide each year as a way to improve the rates of genetic gain. The most advanced programmes also apply genomic selection at the embryonic stage by SNP genotyping and the calculation of genomic estimated breeding values (GEBVs). However, a high proportion of cattle embryos fail to establish a pregnancy. Here, we demonstrate that further interrogation of the SNP data collected for GEBVs can effectively remove aneuploid embryos from the pool, improving live births per embryo transfer (ET). Using three preimplantation genetic testing for aneuploidy (PGT-A) approaches, we assessed 1713 cattle blastocysts in a blind, retrospective analysis. Our findings indicate aneuploid embryos have a 5.8% chance of establishing a pregnancy and a 5.0% chance of given rise to a live birth. This compares to 59.6% and 46.7% for euploid embryos (p < 0.0001). PGT-A improved overall pregnancy and live birth rates by 7.5% and 5.8%, respectively (p < 0.0001). More detailed analyses revealed donor, chromosome, stage, grade, and sex-specific rates of error. Notably, we discovered a significantly higher incidence of aneuploidy in XY embryos and, as in humans, detected a preponderance of maternal meiosis I errors. Our data strongly support the use of PGT-A in cattle IVP programmes.

Screening of in vitro-produced cattle embryos to assess incidence and characteristics of unbalanced chromosomal aberrations

“…The results indicate a 5% incidence rate of unbalanced chromosomal aberrations, taking 430 successfully genotyped stage 8 embryos into account, and even 8% when considering only the 268 samples properly classified (discarding the over-dispersed heterozygotes samples). Either way, the incidence rate of unbalanced chromosomal aberrations is somewhat higher than the 2% reported by Silvestri et al (2021) based on only 103 stage 8 to 9 embryos. Caution should be taken in comparing these incidences to human embryo studies, as in cattle the oocytes are matured in vitro, whereas that is not common practice in humans.…”

“…Although the number of unbalanced cases is limited, BTA 1, 6, 14, 15 were impaired multiple times ( Table 1 ). Silvestri et al (2021) showed that BTA 14 is highly susceptible to chromosomal issues, and also BTA 1, 6, and 15 were more prone to errors among others. The overall call rate of the ploidy issue cases ranged from 0.84 to 0.86, and of monosomy and trisomy cases from 0.87 to 0.97, indicating that overall call rates are not necessarily impaired ( Table 1 ).…”

“… Silvestri et al (2021) screened 1,713 embryo genotypes in a retrospect study, using a combination of log R ratio ( LRR ) and B allele frequency ( BAF ; Attiyeh et al, 2009 ), karyomapping ( Handyside et al, 2010 ; Turner et al, 2019 ), and Gabriel–Griffin plots ( Gabriel et al, 2011 ). They observed an incidence of chromosomal aberrations in 14% of the screened embryos, with the largest incidence in stage 5 embryos (24%), whereas the 103 stage 8 to 9 embryos showed only 2% incidence.…”

Selfish centromeres and the wastefulness of human reproduction

“…Limited post hatch compensation is possible in these species. By contrast, in all mammalian data (except inbred mice), there is an aneuploidy rate of approximately 1% per chromosome per embryo, this being seen in cows [ 10 , 75 ], pigs [ 76 ], hybrid mice [ 77 ], interstrain mouse crosses [ 15 ], outbred mouse lines [ 64 ], as well as humans [ 4 ]. The net percentage of aneuploid early embryos in a young female outbred mammal is thus approximately N%, where N is the haploid chromosome number ( S3 File ).…”

“…Similarly, some methods (e.g., FISH) report exceptionally high rates (>70%) of nondisjunction in eggs [ 2 , 3 ], consistent with very early stage (oocyte to zygote) mortality. While then some data suggest that large acrocentric chromosomes predominate as early embryonic aneuploids (e.g., in pigs [ 92 ]), and others suggest the opposite (in humans and cows) [ 4 , 10 ], inferences from such data are hard to draw.…”