Since the breeding began, valuable individuals were obtained as a result of directed crosses, although only a small percentage of the offspring met the desired phenotypic expectations. The Brangus cattle breed serves as an example of directed breeding, as it is the result of crossing the Angus and Brahman breeds. This selective process, which is both time and money-consuming, can now be circumvented using gene editing technologies to obtain the desired genotype in a targeted manner.
In this sense, we proposed to increase muscle mass of Brangus cattle, consequently improving the productivity of these individuals. Furthermore, this approach could also lead to a reduction in the environmental footprint, as less feed would be required for higher quality meat36. To achieve this goal, we combined CRISPR/Cas9 editing of the MSTN gene with SCNT. Additionally, we used cell samples from prize-winning animals as a proof of concept that gene editing can be applied in breeding programs instead of traditional crossbreeding.
In order to obtain edited embryos, we employed SCNT rather than zygote microinjection for several reasons. First, the genetic background of the edited cells can be thoroughly evaluated, including sequencing of the targeted gene and possible off-targets37,38, before embryo reconstruction. Second, SCNT ensures that no mosaic embryos are obtained, as opposed to what happens when zygote microinjection is used2,39. Third, gene editing can be performed on a cell line obtained from an individual with proven genetic value. This represents a significant improvement in breeding programs, known as precision breeding.
It is important to consider that gene editing of cell cultures results in genetically diverse cell populations, and consequently edited embryos may also be genetically different. To address this, monoclonal cultures should be performed after cell editing. However, obtaining fibroblasts or MSC clonal cell lines can be challenging because primary cultures usually cease dividing after several passages40,41. Although anti-apoptotic components could be added to the cell culture in order to promote cell proliferation42,43, it would involve a significant increase in the number of cell divisions and passages, which has been shown to be detrimental for embryo development after SCNT44. Considering all of the above, the SCNT embryos produced in this work were obtained using the complete heterogenous pool of edited cells.
High editing efficiency was achieved both with the use of the donor plasmid and with the trac:crRNA/Cas9 RNP complex. Puromycin selection after donor plasmid transfection enabled the purification of the cell culture, with 96% editing rate obtained in BFF-E1 when gRNA2 was used. This means that in this case, almost all cells that incorporated the plasmid were indeed edited, but not when the other 3 gRNAs were used. Despite the potential for unintended exogenous DNA integration with the use of plasmids45, this is a cost-effective system that allowed the identification of the most efficient gRNA for performing subsequent experiments with the trac:crRNA/Cas9 RNP complex, which is a more expensive system.
When trac:crRNA2/Cas9 RNP complex was used in BFF-E2-male, MSC-E2-male and MSC-E2-fem cultures, BPE of 58.8%, 31% and 59% was obtained, respectively. A lower editing efficiency was expected compared to E1, because selection of the nucleofected cells is not possible. However, the use of the RNP complex is advantageous, especially when the primary goal is to obtain a live animal. First, the risk of exogenous DNA insertions (i.e., from the plasmid backbone), and thus generating transgenic animals, is eliminated. Previous studies have reported undesired insertions during genetic editing protocols using DNA vectors, whether ZFN46, TALEN45,47 or CRISPR/Cas948,49. Second, it is known that exogenous DNA delivery can activate cytosolic sensors, leading to a pro-inflammatory response and cyototoxicity42,50. Therefore, Cas9 RNP is better tolerated in primary cells, such as the ones used in this work. Finally, the Cas9 RNP complex has a reduced half-life compared to DNA vectors, reducing the possibility of undesirable off-target effects48,49,51.
BPE evaluated after Sanger sequencing of the edited cells was then validated by the MSTN gene sequence analysis of individual embryos reconstructed by SCNT using BFF-E1ed, BFF-E2-maleed, MSC-E2-male and MSC-E2-femed cell lines. To make this comparison, it must be taken into account that both alleles are represented in the Sanger sequencing results: BFF-E1ed cells had a 96% BPE in their cells and 100% (10/10) edited embryos; BFF-E2-maleed cells had a 58.8% BPE in their cells and 64% (16/25) edited embryos, with 58% (29/50 ) edited alleles; MSC-E2-maleed cells had a 31% BPE in their cells and 73.3% (11/15) edited embryos, with 43.3% (13/30 ) edited alleles; MSC-E2-femed cells had a 59% BPE in their cells and 66% (4/6) embryos edited, with 41.7% (5/12) edited alleles. In this last group, more embryos will be needed to draw more accurate conclusions.
Based on these results, we consider that BPE in cell culture is a reliable indicator. However, it is not possible to determine whether the proportion of edition obtained represents homozygous or heterozygous genotypes. For example, 50% BPE in cells could indicate either 100% of heterozygously edited cells or 50% of homozygously edited cells. This tool is valuable for determining CRISPR strategy efficiency, but it is not useful for estimating the genotype of the edited embryos that will be obtained.
The analysis of the edited embryos also revealed significant differences between those obtained by plasmid-based and RNP-based CRISPR. While 100% (10/10) of the analyzed embryos from E1 had bi-allelic editions, wild type and mono-allelic edited embryos were also obtained in E2 when trac:crRNA/Cas9 RNP complex was used. This might be due to the continuous expression of Cas9 and gRNA over several days when plasmids are incorporated into the cells51. As explained above, the longer the CRISPR system is active, the higher the chance of editing.
The embryos derived from MSC-E2-femed were selected for transfer to recipient cows because of the high BPE obtained in cells. The calf was born with both alleles of the MSTN gene edited, one with a T base insertion that generates a premature stop codon, and the other one with an in-frame deletion of 18 bp that results in the loss of 6 amino acids in the propeptide domain of the MSTN protein. Although in-frame deletions in exon 2 still disrupt the myostatin expression52 and can produce MSTN-KO phenotypes53, we were not able to evaluate the deleterious effect of the loss of these 6 amino acids. One possibility is that the phenotype of the MSTNKO/−6 calves is similar to that of heterozygous MSTN-KO animals, which also exhibit significantly greater weight and muscle mass than non-mutant animals at birth54,55. In this work, the weight of both edited calves was 5–11 kg higher than the control and 20–26 kg higher than the average weight of normal Brangus calves at birth.
In conclusion, in the present study MSTN-KO bovine cells and embryos were successfully obtained using both plasmid-based and RNP-based editing strategies. Moreover, two edited calves were obtained using RNP-based editing on MSCs. Therefore, in one generation we obtained a cloned animal that retained the genetic merit of its predecessor and incorporated a valuable commercial characteristic. The relatively high edition efficiency obtained after RNP complex nucleofection suggests that it is a valuable tool to be incorporated into breeding programs aiming to reduce breeding time and eliminate the risk of exogenous DNA insertions in the bovine genome.