The existence of horns in cattle is a safety concern for both handlers and animals, thus their physical removal via disbudding or dehorning must be done. This, however, raises animal welfare concerns. A naturally occurring hornless phenotype in cattle, referred to as the polled phenotype, has been observed, though its biological mechanism is still unknown. Four variants have been associated with the polled phenotype in cattle, two of which are dominant alleles at the POLLED locus (PF and PC) identified in Friesian dairy and Celtic beef breeds (e.g. Angus), respectively. In addition, a long intergenic non-coding RNA (lincRNA#1) has been found to be upregulated in the horn bud region of polled fetuses compared to horned fetuses, indicating a potential role in the development of horns. It has been demonstrated that introgression of the PC allele, a 212 bp duplication that replaces 10 bp, into a pp horned cell line followed by somatic cell nuclear transfer (SCNT) results in the polled phenotype, however SCNT is inefficient, and is not practical for a production setting. Direct cytoplasmic injection (CPI) of gene editing reagents into one-cell zygotes would be a better alternative; however, this method first needs to be optimized to avoid potential off-target effects, reduce mosaicism, and increase biallelic editing efficiencies. In these studies, we created a CPI protocol to optimize CRISPR-Cas9 genome editing efficiencies in bovine embryos to investigate the underlying mechanism behind the polled phenotype in cattle. In the first experiment, gRNAs were designed targeting the POLLED locus (as well as two other loci for separate experiments) and mutation rates in embryos injected 18 hours post insemination (hpi) with Cas9 protein (84.2%) were significantly higher and mosaicism rates were lower (94.2%) than those injected with Cas9 mRNA (68.5%; 100%, respectively), with off-target effects not cause for concern. After optimization of guide RNA (gRNA) design and discovering the most efficient type of Cas9 (mRNA or protein), we created a dual gRNA approach to induce deletions in the embryo genome and aimed to address two hypotheses: (1) the absence of the 10 bp sequence observed in the p allele at the POLLED locus, but absent in the PC allele, is sufficient to result in the polled phenotype, and (2) the absence of lincRNA#1 expression will result in horn development irrespective of the alleles present at the POLLED locus. To test the first hypothesis, we optimized the dual gRNA approach to delete a 133 bp region containing the 10 bp, testing timing of gRNA and Cas9 protein microinjection either 6, 8 or 18 hpi as well as comparing in vitro transcribed (IVT) verses synthetic gRNAs. We found that embryos injected 6 hpi had significantly higher rates of deletion (53%) in comparison to those injected 8 (12%) and 18 hpi (7%), and synthetic gRNAs performed significantly better than IVT gRNAs, with 84% deletion rates seen compared to 53%, respectively. Embryo transfers were performed, and fetuses were collected between three to five months of gestation. A total of seven fetuses were collected, two containing biallelic deletions; however, all exhibited horn bud development, indicating that the 10 bp deletion alone is insufficient to cause the polled phenotype. To address the second hypothesis, we utilized our optimized dual gRNA protocol to create a large (~3.7 kb) deletion in bovine embryos to knockout lincRNA#1. An 80% biallelic lincRNA#1 knockout rate was achieved; however, the genotypically polled fetuses still presented as polled, indicating lincRNA#1 does not play a role in inducing the polled phenotype. Overall, we optimized an efficient CPI system in embryos with high editing efficiency and no off-target effects and further optimized it to create a dual guide approach to create deletions in the bovine embryo genome. Utilizing the dual guide method resulted in high rates of deletion with lower rates of mosaicism than previously seen, making it a promising approach to create genome edited livestock via CPI.
