Optimized Cas9:sgRNA delivery efficiently generates biallelic MSTN knockout sheep without affecting meat quality

CRISPR/Cas9-based genome-editing systems have been used to efficiently engineer livestock species with precise genetic alterations intended for biomedical and agricultural applications. Previously, we have successfully generated gene-edited sheep and goats via one-cell-stage embryonic microinjection of a Cas9 mRNA and single-guide RNAs (sgRNAs) mixture. However, most gene-edited animals produced using this approach were heterozygotes. Additionally, non-homozygous gene-editing outcomes may not fully generate the desired phenotype in an efficient manner. We report the optimization of a Cas9 mRNA-sgRNA delivery system to efficiently generate homozygous myostatin (MSTN) knockout sheep for improved growth and meat production. Firstly, an sgRNA selection software (sgRNAcas9) was used to preliminarily screen for highly efficient sgRNAs. Ten sgRNAs targeting the MSTN gene were selected and validated in vitro using sheep fibroblast cells. Four out of ten sgRNAs (two in exon 1 and two in exon 2) showed a targeting efficiency > 50%. To determine the optimal CRISPR/Cas9 microinjection concentration, four levels of Cas9 mRNA and three levels of sgRNAs in mixtures were injected into sheep embryos. Microinjection of 100 ng/μL Cas9 mRNA and 200 ng/μL sgRNAs resulted in the most improved targeting efficiency. Additionally, using both the highly efficient sgRNAs and the optimal microinjection concentration, MSTN-knockout sheep were generated with approximately 50% targeting efficiency, reaching a homozygous knockout efficiency of 25%. Growth rate and meat quality of MSTN-edited lambs were also investigated. MSTN-knockout lambs exhibited increased body weight and average daily gain. Moreover, pH, drip loss, intramuscular fat, crude protein, and shear force of gluteal muscles of MSTN-knockout lambs did not show changes compared to the wild-type lambs. This study highlights the importance of in vitro evaluation for the optimization of sgRNAs and microinjection dosage of gene editing reagents. This approach enabled efficient engineering of homozygous knockout sheep. Additionally, this study confirms that MSTN-knockout lambs does not negatively impact meat quality, thus supporting the adoption of gene editing as tool to improve productivity of farm animals.


Background
Genome editing is a well-established technique for the modification of genomes of living organisms. The application of genome editing in farm animals is promising for agricultural and biomedicine industries [1,2]. The clustered regularly interspaced short palindromic repeats/CRISPR-associated protein 9 (CRISPR/ Cas9) technology [3,4] has been widely used to induce genome modification, including gene knockout, gene knockin, and single nucleotide substitutions, on a wide range of organisms. CRISPR-based genome editing tools rapidly evolved from the canonical CRISPR/Cas9 system to the more recent CRISPR-associated transposases [5,6], base editors [7,8], and prime editors [9,10]. Although the wide application of CRISPR-based systems, optimization of the delivery methods and parameters of these systems based on target species is still needed. Gene knockout is a favorable approach to disrupt the function of genes that negatively regulate desirable economically important traits in farm animals. However, most gene-edited farm animals generated with zygote microinjection were heterozygous knockouts. Thus, in order to generate authentic homozygous gene-edited animals in a time-saving and cost-effective manner, it is necessary to optimize the gene knockout approach.
At present, the main methods used for genetic modification of animals are somatic cell nuclear transfer (SCNT) of gene-edited cells and embryonic microinjection. SCNT can be used to generate homozygous gene-edited animals, but this approach presents several technical challenges due to high embryonic lethality [11,12]. On the other hand, microinjection provides a technically less challenging approach for efficient genome modification, being successfully applied to several mammalian species to target crucial genes, such as the myostatin (MSTN) gene to promote muscle mass gain [13][14][15][16][17]. However, animals with MSTN knockout using microinjection usually harbor heterozygous and/ or chimeric gene-editing products [18][19][20].
The MSTN gene is a member of the transforming growth factor-beta (TGF-β) superfamily, negatively regulating skeletal muscle tissue production [21][22][23]. MSTN affects the growth and development of muscle tissue by regulating the proliferation of myoblasts [24,25]. Inactivation of the MSTN gene was shown to promote proliferation of myocytes and muscle fiber hypertrophy [26][27][28][29][30][31][32][33][34]. Therefore, the MSTN gene can be considered a genome editing target for exploring the MSTN signaling pathway and production of geneedited animals with improved muscle mass gain and growth rates. Moreover, inducing mutations in both MSTN alleles (i.e., in a homozygous manner) may enable a more efficient gene disruption, hence generating animals with more desirable muscle mass phenotype.
Although MSTN-knockout sheep models have been generated previously using microinjection of CRISPR/ Cas9 reagents [14,[18][19][20]35], generation of homozygous MSTN-knockout individuals was relatively inefficient, which might be attributed to ineffective sgRNA design and the use of undetermined microinjection concentrations of CRISPR/Cas9 reagents. In most agricultural animal studies, delivery of CRISPR/Cas9 reagents has been performed based on methods established in mice [36]; thus, this approach requires optimization to ensure suitability for agricultural animal studies.
Therefore, in this study, parameters of sgRNAs design and concentration of CRISPR/Cas9 microinjection were optimized. Homozygous MSTN-knockout sheep were efficiently generated, and meat quality of MSTN-knockout lambs was assessed. This study constitutes a practical reference for generating homozygous knockout farm animals based on the optimization of CRISPR/Cas9 reagents. It was also confirmed that MSTN knockout in farm animals increases muscle mass without affecting meat quality.

Analysis of off-target mutations in gene-edited animals
To evaluate the off-target effects potentially induced by CRISPR/Cas9 system, 10 off-target sites (OT1-OT10) were selected using Cas-OFFinder [38] (see Additional file 1: Table S3). Nucleotide sequences around predicted off-target sites were amplified by PCR from genomic DNA of eight gene-edited founders and evaluated using Sanger sequencing. No off-target editing occurred in gene-edited founders (see Additional file 2: Fig. S2), highlighting the accuracy of the CRISPR/Cas9 system when optimized sgRNAs are used.

Phenotype assessment of gene-edited animals
The MSTN gene is a negative regulator of muscle growth and development, and loss of MSTN causes overdevelopment of muscles in animals, which contributes to the generation of the desirable double-muscling phenotype. To evaluate the expression of MSTN in muscle tissues of gene-edited sheep, five lambs [mutated (Mut): #44 and #48; wild-type (WT): #06, #24, and #38] were selected (Fig. 3a). Expression of the MSTN gene in gene-knockout lambs was significantly lower than that in WT lambs, which is consistent with previous reports [45,46].
Then, effects of genetic modification on muscle development of MSTN-knockout lambs were assessed using hematoxylin and eosin (H&E) staining. Previous studies highlighted that the loss of MSTN function causes an increase in muscle mass, resulting from the combination of hyperplasia and hypertrophy [31,33,47]. It is indicated that the increased muscle mass in constitutive MSTN knockout mice is primarily due to myofiber hypertrophy [47], while in cattle with a naturally occurring MSTN mutation is primarily due to hyperplasia [31,33]. In the current study, the diameter of myofibers of MSTN-knockout lambs (e.g., #48) was notably larger than that of WT lambs (Fig. 3e). These results are consistent with previous gene editing studies conducted on cattle, pigs, goats, and rabbits [17,40,48,49], which showed that disruption of the MSTN gene leads to the desirable increased muscle mass phenotype in animals, thus providing a novel way for increased meat production.

Chemical composition and meat quality analysis of gluteal muscles
Chemical analyses were then performed to assess meat quality parameters of MSTN-knockout and WT animals. As shown in Table 2, no differences in pH value, contents of total crude protein, fat, moisture, and shear force of gluteal muscle tissue obtained from MSTN-knockout founders and their WT counterparts. Therefore, these results indicate that the editing of the MSTN gene in lambs can promote muscle growth without affecting main meat quality parameters and these findings are in line with previous gene editing studies conducted on pigs [50][51][52]. Additionally, naturally occurring mutations in the sheep MSTN gene highlighted similar observations [53][54][55].

Conclusions
In this study, the CRISPR/Cas9 system was optimized for gene editing of MSTN in sheep by in vitro selection of highly-efficient sgRNAs. Moreover, the optimal microinjection concentration to efficiently generate biallelic gene knockout animals was determined. Additionally, homozygous MSTN-knockout sheep were shown to possess the desirable increased body mass phenotype without affecting meat quality. Therefore, the optimized gene editing system described herein can be potentially applied to enhance desirable traits in food animals.

Animals
All animals were used in the experiments raised at the Ningxia Tianyuan Sheep Farm, Hongsibu, Ningxia Autonomous Region, China. Water and standard feed were supplied ad libitum for both Mut founders and their WT counterparts. Animals were treated according to the Guidelines of Northwest A&F University for the Care and Use of Laboratory Animals, China.

In vitro transcription of sgRNAs and Cas9 mRNA
Oligonucleotides were synthesized and annealed to form double-stranded oligos to construct sgRNAs vectors for in vitro transcription (see Additional file 1: Table S6). Double-stranded oligos were sub-cloned into the pUC57-T7-gRNA vector as previously described [56]. Clones that contained the desired sequences were identified by Sanger sequencing and amplified in culture medium. Plasmids were then obtained using the plasmid extraction kit (AP-MN-P-250G; Axygen, Union City, CA, USA). sgRNAs were in vitro transcribed using the MEGAshortscript Kit (AM1354; Ambion, Austin, TX, USA) and purified using the MEGAClear Kit (AM1908; Ambion, USA). Linearized Cas9 in vitro transcription vector (Addgene; No. 44758) was used as template to produce Cas9 mRNAs as previously described [14].

Determining the optimal concentration of Cas9 mRNA:sgRNAs at the embryonic level
Four high-efficiency sgRNAs (sgRNA 1-4 , sgRNA 1-5 , sgRNA 2-1 , and sgRNA 2-2 ) were selected for further validation experiments in sheep embryos (Fig. 2a). In order to determine the optimum microinjection concentration, 50 healthy ewes (~ 3-5 years old) with normal estrous cycles were selected as donors for zygote collection in October. The procedure of superovulation of donors was carried out as previously described [40]. Briefly, an EAZI-BREED ™ controlled internal drug release (CIDR) devise for sheep and goats containing 300 mg of progesterone was inserted into the vagina of donor ewes for 12 days. Superovulation was performed 60 h prior to the removal of the CIDR device using a total of 300 units of FSH ( (Fig. 1c).

Generation of MSTN homozygous knockout sheep
Embryos were obtained by surgical oviduct flushing from 33 females previously subjected to superovulation. Collected embryos were cytoplasmically coinjected with a mixture of 100 ng/μL of Cas9 mRNA and 200 ng/μL of sgRNAs (as revealed by the optimal result of Group #8 at the embryonic level) using the Eppendorf FemtoJet system. The following parameters were used: microinjection pressure, 45 kPa; compensatory pressure, 7 kPa; and time, 0.1 s. Microinjection was performed in an Olympus ON3 micromanipulation system [39]. Microinjected embryos were cultured in Quinn's Advantage Cleavage Medium (Sage) for 24 h and subsequently transferred into surrogates as previously reported [19]. Pregnancy was confirmed by observing estrus behaviors in surrogates at each ovulation cycle. After approximately 150 days of pregnancy, 16 newborn lambs were genotyped. Full care and monitoring were given to the lambs after delivery.

Genotyping of generated founders
Peripheral venous blood samples of two-week-old lambs were collected and submitted to genomic DNA extraction. PCR amplification was conducted using primers listed in Additional file 1: Table S2, and obtained PCR products using the KOD-NEO-Plus enzyme (DR010A; TOYOBO, Japan) were submitted to Sanger sequencing.

Prediction of off-target sites
Potential off-target sites with maximum three mismatches were predicted using Cas-OFFinder online software [38]. Search for off-target sites was carried out as previously described [38]. Primers used in amplifications for off-target sites and Sanger sequencing are listed in Additional file 1: Table S7.

Targeted deep sequencing
Target genomic loci were amplified using KAPA HiFi Hot-Start PCR Kit (KK2501; KAPA Biosystems, Wilmington, MA, USA) for generating a deep sequencing library as previously reported [57]. PCR amplicons were sequenced as a pool using the Dual Index Sequencing with TruSeq HT Library Prep (Illumina, San Diego, CA, USA).

H&E staining
Samples of gluteus maximus muscle were obtained from MSTN-knockout (founder No. #48) and WT (#06) founders (180 days old) for tissue biopsies. The gluteal muscle tissue was immediately fixed with 4% paraformaldehyde at 4 °C overnight, then embedded into paraffin and sectioned. After cutting the samples into 3 μm slices, these slices were stained with H&E. Tissue sections were dewaxed, rehydrated, and stained using standard H&E protocols [19]. After staining, tissue sections were observed by microscopy and images were analyzed.  Table S8. Gene expression levels were calculated using the 2 -ΔΔCt method and normalized against housekeeping GAPDH gene. Each sample was run in triplicate.

Determination of meat quality
At the age of 6 months, we selected four MSTN-edited and four WT female lambs to measure meat quality. Meat quality analysis was performed as previously described [58]. Briefly, pH value was determined after 45 min of gluteal muscle tissue sample collection using a pH-STAT meter (SFK-Technology, Denmark). After 24 h postmortem, gluteal muscle tissue was sampled, as well as pH and shear force of referred samples were determined. Shear force was assessed using the Warner-Bratzler shear force (WBSF) approach [59]. Moreover, crude protein, intramuscular fat, and moisture content of gluteal muscle tissue samples were determined using recognized Association of Official Analytical Chemists (AOAC) methods [58]. Moisture content was determined by drying gluteal muscle tissue samples in an oven at 105 °C until constant weight was obtained. Total crude protein (N × 6.25) was determined using the Kjeldahl method [60]. Intramuscular fat content was determined by the Soxhlet extraction method [61]. Data were reported as g/100 g fresh muscle weight. The data are expressed as the mean ± SEM and analyzed using Student's t-test with significant differences considered at P < 0.05.
Additional file 1: Table S1. Average cleavage rate and mutation rate of experimental groups. Table S2. Primers used for genotyping and amplifying Cas9/sgRNA-targeted MSTN fragment. Table S3. List of predicted off-target sites. Table S4. Growth parameters of MSTN-knockout and wildtype sheep. Table S5. sgRNA sequences and target sites. Table S6. Oligonucleotides used for generating sgRNA-expressing vectors for in vitro transcription. Table S7. Primers used for genotyping and amplifying predicted off-target site fragments. Table S8. Primers used for measuring MSTN expression level by real-time quantitative PCR (RT-qPCR). Description: The file contains the sequences of on-and off-target sites, primers, and oligonucleotides, as well as other relevant data.
Additional file 2: Figure S1. Overlapping or discontinuous peaks in Sanger sequencing of DNA samples obtained from the eight founder animals. Figure S2. Detection of potential off-targeted sites in the eight founder animals by Sanger sequencing. Ten potential off-targeted sites (OT1-OT10) were predicted by Cas-OFFinder [38]. Description: The file contains the results of Sanger sequencing for on-and off-target loci in gene-edited animals.