‘Double-muscling’ and pelvic tilt phenomena in rabbits with the cystine-knot motif deficiency of myostatin on exon 3

Gene mutations at different gene sites will produce totally different phenotypes or biological functions in gene-edited animals. An allelic series of mutations in the myostatin (MSTN) gene can cause the ‘double-muscling’ phenotype. Although there have been many studies performed on MSTN-mutant animals, there have been few studies that have investigated the cystine-knot motif in exon 3 of MSTN in rabbits. In the current study, CRISPR/Cas9 sgRNA anchored exon 3 of a rabbit’s MSTN was used to disrupt the cystine-knot motif to change the MSTN construction and cause a loss of its function. Eleven MSTN-KO founder rabbits were generated, and all of them contained biallelic modifications. Various mutational MSTN amino acid sequences of the 11 founder rabbits were modeled to the tertiary structure using the SWISS-MODEL, and the results showed that the structure of the cystine-knot motif of each protein in the founder rabbits differed from the wild-type (WT). The MSTN-KO rabbits displayed an obvious ‘double-muscling’ phenomena, with a 20−30% increase in body weight compared with WT rabbits. In the MSTN-KO rabbits, all of the MSTN−/− rabbits showed teeth dislocation and tongue enlargement, and the percentage of rabbits having pelvic tilt was 0% in MSTN+/+, 0% in MSTN+/−, 77.78% in female MSTN−/− rabbits, and 37.50% in male MSTN−/− rabbits. The biomechanical mechanism of pelvic tilt and teeth dislocation in the MSTN-KO rabbits requires further investigation. These newly generated MSTN-KO rabbits will serve as an important animal model, not only for studying skeletal muscle development, but also for biomedical studies in pelvic tilt correction and craniofacial research.


Introduction
Myostatin (MSTN) can negatively regulate the growth and development of skeletal muscle. It is a member of the transforming growth factor β (TGF-β) superfamily of growth and differentiation factors [1,2]. MSTN functions to inhibit excessive growth of muscle in order to maintain an overall balance of this tissue relative to adipose tissue [3]. Natural myostatin null animals, including cattle [4], dogs [5], humans [6], and sheep [7], shows a dramatic and widespread increase in skeletal muscle mass; namely, a 'double-muscling' phenomenon.
Since its discovery in mice [8], MSTN has been extensively investigated in other animals. A large number of variants have been identified in cattle, most of which are silent or neutral. Two of the variants in the third exon strongly affect the 'double-muscling' phenotype in Belgian Blue cattle [9]. An 11-bp deletion at the bovine myostatin nucleotide 821 [nt821(del11)] in the coding region results in a frameshift mutation. Also, in Piedmontese cattle [4], a G to A transition at nucleotide 938 results in the substitution of cysteine by tyrosine (this mutation is referred to as C313Y). It has been shown that an 11-bp deletion [nt821(del11)] in the myostatin gene results in a truncation of the bioactive carboxyterminal domain of the protein [9,10]. In addition, the C313Y mutation predicts the replacement of cysteine at amino acid 313 with tyrosine. This cysteine is the fifth in a series of nine whose appearance and spacing is extremely conserved among TGF-β family members [2,4]. These two mutations remove the portion of the myostatin protein that is most highly conserved among TGF-β family members. The rabbit myostatin gene consists of three exons and two introns. The third exon of the rabbit myostatin gene codes for mature processed myostatin, which is the active region of the molecule and shares extensive homology with other TGF-β members [2,11]. Similar to other TGF-β family members, three kinds of proteins are translated by myostatin: a precursor protein consisting of an N-terminal signal sequence; a propeptide domain; and a growth factor domain that contains the characteristic cystine-knot motif (C272-C282, C281-C340, C309-C372, and C313-C374). C339 forms the intermolecular dimerization disulfide bond [12,13]. The C313Y mutation removes the C313-containing disulfide bond, an integral part of the characteristic TGF-β cystine-knot structural motif. Some research has suggested that the cystine-knot motif, which is essential for myostatin structural stability and covalent dimerization, is required for receptor-mediated signaling by myostatin growth factor in vivo, with its absence leading to a double-muscled phenotype ('double-muscling') [14]. Hence, it is feasible to prepare a rabbit model of myostatin using the knockout cystine-knot motif.
Previous studies have reported that MSTN naturally mutant cattle generally show a 20−40% increase in muscle weights due to muscle hypertrophy or hyperplasia relative to wild-type (WT) cattle [15]. This is consistent with a report that found that individual muscle fibers of homozygous MSTN-KO mice (MSTN −/− mice) produced by the homozygous deletion of the C-terminal region of the myostatin gene in embryonic stem cells were double or triple in mass compared with those of their heterozygous and WT littermates [2]. Some research has reported that the masseter and temporalis muscles of MSTN −/− mice were over 50% larger in mass than WT controls due to larger cross-sections of the muscle fibers and increased numbers of muscle cells [2,16]. Therefore, MSTN −/− mice were used to explore the effects of increased muscle mass on sagittal suture complexity. The mice were also used to indicate that a consistent amount of force needs to be applied over a much greater area of the craniofacial skeleton in the myostatin knockout, and this should, in turn, lead to altered biomechanical stress and bony morphology [3,16]. An MSTN −/− mouse has significantly larger temporalis, masseter, and medial and lateral pterygoid muscles than WT controls [17]. These characteristics, coupled with morphologic findings of the skull in MSTN −/− mice (e.g., shorter crania, longer, and rounder mandibles), provide further evidence of an altered craniofacial loading environment due to MSTN deficiency [3,[16][17][18][19][20]. Therefore, these MSTN-KO mice have been used as a model for studying muscle-bone interactions, such as to investigate the role of increased mandibular elevators and bite forces on temporomandibular joint morphology. In the past 5 years, many MSTN-KO animals (e.g., pigs, goats, sheep, and rabbits) with 'double-muscling' have been generated due to the emergence of the CRISPR/Cas9 gene editing technique as a newly versatile genome engineering tool [21][22][23][24]. However, the morphology and the muscle-bone interactions of these MSTN-KO animals have rarely been investigated.
The aim of the present study is to generate MSTN-KO rabbits using the CRISPR/Cas9 system anchored to the cystine-knot motif by microinjecting Cas9 mRNA and sgRNA into fertilized rabbit eggs. The MSTN-KO rabbits can be used as a basis to create new breeds of rabbit to improve muscle yield, with the ultimate goal of approval for human consumption. In addition, an important animal model is proposed for use in further studies that investigate muscle-bone interactions.

CRISPR/Cas9 expression plasmid construction and in vitro transcription
To produce gene-modified New Zealand white rabbits with double-muscling, two sites that are associated with the cystine-knot motif in the MSTN rabbit gene were selected as target genes ( Figure 1A,B). The sgRNA-1 anchors the MSTN gene site that encodes C372, and the sgRNA-2 anchors the MSTN gene site that encodes C339 and C340.

Generation of MSTN-KO rabbits using the CRISPR/Cas9 system
In the present study, a total of 48 injected zygotes were transferred into the oviducts of three surrogate rabbits. After full-term gestation, the three surrogate mothers successfully gave birth to 12 live pups (Table 1).
Genomic DNA was extracted from the ear tissue of the newborn pups. The 12 individuals used for genotyping were numbered M1−M12. The T-cloning and PCR-sequence results showed that the MSTN mutation was detected in 11 pups, and the indels ranged from 1 to 221 bp ( Figures 1C and 2A,B). The theoretical amino acid sequences of the  MSTN knockout alleles are shown in Figure 2C,D. In addition, fragment deletions between the two sgRNAs targeting sites were frequently observed in the present study. Here, only one rabbit (M10♀) harbored mutations in exon 3 of the MSTN gene that did not cause a frameshift mutation. Eleven rabbits (M1−M11) deleted C372 and C374, and four rabbits (M8−M11) deleted C339 and C340 in two alleles, C281-C340. M7♀ and M10♀ were homozygous, and both of them lost C372 and C374 in the cystine-knot motif. In addition, M10♀ also deleted C339 and C340. As shown in Figure 2E,F, the structure of each protein differs from the WT for the cysteine delete (like C339, C340, C372, and C374). Five rabbits (M1♀, M2♂, M3♀, M5♀, and M7♀) deleted C272, C281, C282, C339, and C340. In the cysteines mentioned above, four cysteines formed two intramolecular disulfide bonds (C272−C282, C281−C340), and C339 was involved in intermolecular disulfide bond formation between two MSTN monomers. The typical phenotype of double-muscling was observed in F0 MSTN-KO rabbits at four months of age, and they were compared with their WT counterparts, as shown in Figure 3A,B. To investigate whether gene mutations abolished the MSTN protein translation, Western blotting and immunofluorescence staining were performed. As shown in Figure 4, the MSTN protein was significantly decreased in MSTN −/− and MSTN +/− rabbits compared with MSTN +/+ counterparts.

Increased body weights in the MSTN-KO rabbits
Rabbits experience an exuberant growth period from 2 to 3 months, so the body weights of the MSTN-KO rabbits and their WT counterparts were recorded from 9 to 14 weeks. It was found that the average body weights of male MSTN-KO rabbits were significantly higher than the male WT rabbits from 9 to 14 weeks of age ( Figure 3C). The average body weights of female MSTN-KO rabbits were significantly higher than the female WT rabbits from 11 to 14 weeks of age ( Figure 3D). Significant differences (*P<0.05) from 10 to 14 weeks after birth were found between the MSTN-KO male rabbits and the WT male rabbits. Significant differences (*P<0.05) were found between the MSTN-KO female rabbits and the WT female rabbits from 13 to 14 weeks after birth. At 14 weeks, the average body weights of the female (2318.40 + − 183.15 g) and male (2579.33 + − 67.99 g) MSTN-KO rabbits increased by 17.91 and

Heritability and breeding of the MSTN-KO rabbits
The male founders (M2, M6, and M8) were mated with the female WT rabbits, and the female founders (M1, M3, M5, and M10) were mated with the male WT rabbits. Genomic DNA was extracted from the ear tissues of the F1 rabbits. PCR amplification and TA-cloning sequencing analysis showed that these F1 rabbits were monoallelic MSTN-KO rabbits. Then the same genotype of the F1 male and female rabbits was mated for homozygous mutants. The breeding results are shown in Supplementary Table S1.

Off-target assay
A total of ten potential off-target (OT) sites (OTs, five for sgRNA1 and five for sgRNA2) were successfully amplified using PCR and subjected to Sanger sequencing. No overlapping peaks were detected near the OTs.

Histological analysis
The MSTN-KO rabbits exhibited the double-muscled phenotype (Figure 3A), and a histological examination of the gluteus maximus and tongue showed muscle fiber hypertrophy relative to the fibers of the WT rabbit ( Figure 5A). In Figure 5B

Analysis of hematological and biochemical parameters
The hematological and biochemical parameters from 6-month-old MSTN +/+ , MSTN +/− , and MSTN −/− rabbits were analyzed. In Figure 6, the levels of serum creatinine (CREA, 325.97 + − 18.92 μmol/l), serum creatine kinase (CK, CREA is a product of muscle tissue metabolism in the body. Serum CK is related to the total amount of muscle. This means that the greater the muscle capacity, the higher the serum CK activity. Increased serum CREA and CK levels in the MSTN-KO rabbits should correlate with increased muscle mass. The serum CREA content reported for MSTN-KO Meishan pigs was also higher than that in WT pigs [25]. Interestingly, serum LIP was greater in the MSTN −/− rabbits than that in the MSTN +/+ rabbits at 6 months of age. Further experiments are required to investigate this phenomenon in MSTN −/− rabbits. There were no differences in the hematology characteristics and other biochemical parameters. The results are listed in Supplementary Tables S2 and S3.

Effect of MSTN mutation on the teeth and pelvis in MSTN-KO rabbits
To study if there were any effects of the targeted MSTN mutation on the rabbits, the body structure and physiological activity of the MSTN-KO rabbits were closely observed. It was found that the MSTN −/− rabbits showed a greater mass of the masseter muscles compared with the WT and MSTN +/− rabbits ( Figure 7A). In addition, teeth dislocation and pelvic tilt were present in the MSTN −/− rabbits at 3

Discussion
In the present study, the MSTN-KO rabbits showed a 'double-muscling' phenomenon and a dramatic increase in body weight. At the end of 14 weeks, the MSTN-KO rabbits exhibited an approximately 30% increase in body weight compared with WT rabbits that might be attributed to the MSTN gene mutation factor, irrelevant to age and sex. These results are consistent with the MSTN-KO mice produced by the homozygous deletion of the C-terminal region of the myostatin gene [2,8,26], indicating that disruption of cystine-knot motif can increase muscle mass. This also shows that the structure of the cystine-knot motif plays an important role in myostatin activity. An intricate cysteine-knot motif contains nine cysteine residues (C272, C281, C282, C309, C313, C339, C340, C372, and C374) in the C-terminal region of mammalian myostatin protein [27]. The nine cysteine residues form disulfide bonds that are crucial for the formation of three-dimensional protein structures and could influence the function of other motifs (such as ligand binding or signaling) in the same protein [28,29]. Disruption of the cystine-knot and decreased covalent dimerization likely results in a less compact structure that is then unsuitable for type I and type II Ser/Thr kinase transmembrane receptors activation [14,30]. An analysis of the C313A and C374A mutations confirmed that it is the disruption of the cystine-knot by removal of the C313-C374 disulfide bond, and not the introduction of the tyrosine, that is the dominant factor in the decreased stability and reduced disulfide-linked dimerization of C313Y myostatin [14]. This means that removing the disulfide bonds destroys the cystine-knot motifs. Various mutational MSTN genes of the 11 founder rabbits were modeled using the SWISS-MODEL, and the cystine-knot motifs were disrupted, leading to disulfide bond loss and great changes in the tertiary structures. Furthermore, as this research progresses and more offspring are produced, this MSTN-KO model can be used to incorporate more reliable parameters and may offer insights into the mechanisms that control signaling in skeletal muscle.
In the present study, all of the homozygous mutant rabbits showed pelvic tilt, teeth dislocation, and tongue enlargement. This OT phenomenon was not found in the MSTN − MSTN-KO rabbits [24,31] and MSTN-KO pigs [21]. Previous studies have demonstrated that an enlarged tongue is primarily attributed to muscle fiber hypertrophy [24]. However, the teeth dislocation and pelvic tilt phenomena have rarely been reported in other MSTN −/− animals. Myostatin is expressed only in muscle and not in normal bone or other connective tissue [8]. It has long been recognized that increased muscle mass and strength are associated with a resultant increase in bone mass [32][33][34]. Studies have shown that MSTN has an effect on the bone mass and bone volume fraction in mice [35]. Byron et al. [36] and Vecchione et al. [37] employed a hypermuscular mouse model (mice lacking myostatin) to explore the effects of increased muscle mass. They reported that adult MSTN-KO mice were generally more brachycephalic, had a smaller cranial vault and maxillary lengths, and exhibited significantly different mandibular shapes, with the mandible longer and shortened in the vertical dimension compared with WT controls. Byron et al. [16] further revealed that increasing masticatory muscle mass and bite force would increase sagittal suture waveform complexity based on research of MSTN-KO mice. Recently, a study on MSTN-KO Meishan pigs found that 20% MSTN-KO Meishan pigs had one extra thoracic vertebra, and the researchers demonstrated that this phenomenon was caused by the MSTN mutation [25]. To our knowledge, the effect of MSTN deficiency on rabbit teeth and the pelvis has rarely been reported in other MSTN-KO animals. Statistically significant differences have been found in joint proportions and bone mineral densities between normal and MSTN-KO mice [20]. Moreover, a recent study has shown that mice homozygous for the myostatin mutation had increased muscle mass and craniofacial dysmorphology in adulthood [3]. Significant correlations have also been noted between masseter muscle weight and mandibular body length [37]. In this study, the teeth dislocation phenomenon may have been due to masticatory hypermuscularity that resulted in significantly altered craniofacial morphology, likely due to altered biomechanical stress. Rabbits excel at sprinting and jumping due to short forelimbs and long hind legs. In MSTN −/− rabbits, the elevated hindquarter muscle mass and physiological cross-section enhanced jumping forces compared with similar-sized MSTN +/− rabbits. Following alterations in the local biomechanical conditions, the elevated hindquarter muscle loads led to an alteration in the morphology of the pelvis in MSTN −/− rabbits. These findings demonstrate that muscle function plays an important role in bone morphology. It was found in this study that MSTN +/− rabbits displayed mild phenotypes and mild abnormalities. This result is consistent with studies of both MSTN +/− mice and MSTN +/− cattle, and the abnormalities were likely caused by haploinsufficiency [4]. In these allelic mutation rabbits, it was found that the incidence of pelvic tilt in female MSTN −/− rabbits (77.78%) was significantly higher than that in male MSTN −/− rabbits (37.50%). The conclusion drawn was that this may be due to the difference in the pelvic structures of males and females. In mammals, the pelvis is significantly larger in females than in males to allow babies to pass through pelvis when they are born. However, it was still unclear if these craniofacial and pelvic tilt abnormalities were related to the MSTN knockout condition during the pre-or neonatal period or were simply a result of postnatal compensatory changes. The effect of the molecular mechanism of the loss of myostatin function on the craniofacial and pelvic tilt abnormalities in rabbits needs further investigation.
The MSTN-KO mouse has typically been used as a new animal model for the study of the relationship among craniofacial morphology, hypermuscularity, and increased bite force since its muscles have been demonstrated to be significantly larger [16,38]. In a clinical setting, non-specific chronic low back pain (cLBP) is associated with pelvic tilt [39]. It can be concluded that the MSTN-KO rabbits will be a better model for pelvic tilt correction and craniofacial research due to their observable pelvis and skulls, appropriate body, longer life cycles, and easy management due to low feeding costs.

Conclusion
In conclusion, MSTN-KO rabbits were generated with remarkably high efficiency by microinjecting Cas9 mRNA and sgRNA into pronuclear-stage embryos. The results of the present study suggest that Cas9 mRNA microinjection may provide a fast and efficient approach for gene knockout rabbit models. Although muscle overgrowth caused by homozygous mutations in myostatin was impressive, bone abnormalities caused by excessively growing muscles may be a health hazard for livestock. In addition, based on the abnormality results found in MSTN-KO rabbits, it is suggested that these rabbits would make an important animal model for craniofacial and pelvic research.

Animals and ethics statement
The New Zealand rabbits used in the present study were kept at the Animal Genetic Engineering Laboratory of Yangzhou University. They were fed twice a day and offered water limitlessly. Animal experiments and procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals (Ministry of Science and Technology of the People's Republic of China) and approved by the Animal Care and Use Committee of Yangzhou University, Yangzhou, China (license number: SYXK(Su)2017-0044).
To construct the recombinant vector for the preparation of sgRNA by in vitro transcription, a pair of complementary DNA oligos, shown in Supplementary Table S4, were annealed to be double-stranded. They were then subcloned into the linearized PYSY-T7-Cas9-T7-MSTN-gRNA cloning vector, which was provided by the Nanjing YSY Biotech Com., Ltd., Nanjing, China. The PCR product for the in vitro transcription of Cas9 mRNA was amplified using T7 primers (T7-1:5 -TGTAAAACGACGGCCAGT-3 and T7-2: 5 -TGGCACCGAGTCGGTGCTTT -3 ). The PCR products for the in vitro transcription of sgRNA were amplified using T7 primers (T7-3: 5 -ATTAACCCTCACTAAAGGGA-3 and T7-4: 5 -AAAAAAAGCACCGACTCGGTGCCAC-3 ). Cas9 mRNA and sgRNAs were transcribed in vitro using the ScriptMAX Thermo T7 Transcription Kit (TSK-101, Toyobo, Japan) and purified using a lithium chloride precipitation solution (AM9480, Ambion, U.S.A.) according to the manufacturer's instructions. The concentration and quality of the synthesized mRNA were determined using the spectrophotometer (OD-1000+, Nanjing Wuyi technology Com., Ltd., Nanjing, China) and agarose gel electrophoresis, respectively. Cas9 mRNA and sgRNAs were diluted into RNase-free water and stored at −80 • C for future use.

Zygote injection with Cas9/sgRNA
Healthy donors with regular estrus cycles were selected for zygote collection. Zygotes were collected using surgical oviduct flushing from donors after superovulation treatment and natural mating. In brief, donors were injected with follicle stimulating hormone (FSH) for estrus synchronization. A total of 60 IU FSH were administered by intramuscular injection in six dosages at 12-h intervals (the first dose was 15 IU and other doses decreased progressively to 5 IU). The animals were then injected with 100 IU of HCG 12 h after the FSH was administered. The females were then subsequently mated with a male rabbit.

DNA extraction, PCR, and sequencing
The genomic DNA of the sample was extracted from a small piece of ear biopsy using the phenol-chloroform extraction method. The target sites were PCR amplified using the primers (CAS9SG12-1 and CAS9SG12-2). The sequences of primers are shown in Supplementary Table S5. The PCR products were subjected to 1% agarose gel electrophoresis, and the map is shown in Figure 1C. To test the mutation patterns, the PCR products were purified using a PCR Purification kit (EP101-01, TransGen, China) and cloned into the pGEM-T vector (A1360, Promega, U.S.A.). A total of 50 positive T-clones were sequenced by the Beijing Genomics Institute. Lasergene (DNASTAR, Inc., U.S.A.) was used for the sequence analysis. The tertiary structure of the theoretical amino acid sequences was modeled using SWISS-MODEL (https://www.swissmodel.expasy.org/).

Western blotting and immunofluorescence analysis
Total protein was extracted from gluteus maximus tissue in MSTN +/+ , MSTN +/− , and MSTN −/− rabbits using radio-immunoprecipitation assay (RIPA) lysis buffer (P0013B, Beyotime, China) with protease inhibitors according to the manufacturer's instructions. The protein concentrations were determined using the bicinchoninic assay (BCA) method (CW0014S, CWBIO, China). A 40-μg protein sample was subjected to a 12% polyacrylamide Tris-glycine gel for 1 h. Separated proteins were then transferred to polyvinylidene difluoride (PVDF) membranes (F019531, Sangon, China). The membranes were then blocked for 2 h in a Tris buffered saline with Tween 20 (TBST) buffer containing 5% milk at room temperature. The membranes were subsequently incubated for 1.5 h at 37 • C with primary monoclonal antibodies against MSTN (sc-393335, Santa Cruz Biotechnology, U.S.A.) diluted to 1:100 with TBST buffer. The membranes were subsequently washed three times for five minutes with the TBST buffer and incubated with anti-mouse secondary antibodies (D110096, Sangon, China) for 1.5 h at 37 • C. The GAPDH antibody (CW0100M, CWBIO, China) was used as a reference control. The proteins were visualized with the ECL substrate solution (#170-5060, Bio-Rad, U.S.A.) using an imaging system (5200, Tanon, China).
Gluteus maximus tissue from MSTN +/+ , MSTN +/− , and MSTN −/− rabbits were collected and fixed with 4% paraformaldehyde, followed by embedding in paraffin according to routine protocol. Deparaffinized tissues were sectioned at 4-μm thicknesses and blocked in 5% bovine serum albumin (BSA). Sections were stained with a primary monoclonal antibody against MSTN (sc-393335, Santa Cruz Biotechnology, U.S.A.), and diluted to 1:50 with a blocking buffer by incubating at 4 • C overnight. After washing, sections were incubated with secondary antibodies

Increased body weights in MSTN-KO rabbits
In the preent study, the MSTN-KO founder rabbits and WT counterpart controls were bred and housed under the same conditions. Their feed was adjusted with respect to growth stages after weaning at 30 days of age. Their body weights were recorded daily from 9 to 14 weeks of age.

OT assay
To determine the site-specific cleavage of the CRISPR-Cas9 system in vivo, all possible sites of the entire rabbit genome with homology to the 23-bp sequence (sgRNA + PAM) were predicted using the CRISPR design tool (http://tools. genomeengineering.org). The mismatch parameter for the target sequence was set as 5 -NGG and was chosen as the protospacer adjacent motifs (PAM). The sites with 12-bp conserved proximal to PAM and total mismatches <4 and sites with total mismatches <5 were chosen as potential OT sites for subsequent testing. The selected potential OT sites were initially PCR-amplified using genomic DNA from MSTN-KO rabbits. The PCR products were then used for sequencing. The information of the OT sites and primer pairs used is listed in Supplementary Tables S6 and S7.