Abstract: In this study, we searched for the mutations in Myostatin (MSTN) gene among the parental generation of our F2 population, consisted of Japanese Black and Limousin breeds, as the positional candidate gene of the QTL for several beef quality and quantity traits detected at the centromeric end of BTA2. Twenty single nucleotide polymorphisms were detected in MSTN gene and the one of them was nonsynonymous mutation, which cause amino acid substitution of phenylalanine to leucine at position 94 (F94L). The significant genotypic effects of F94L mutation for 5 carcass grading traits (carcass grade, M. longissimus area, BMS, luster and firmness), 3 meat quality traits (crude fat, crude protein and moisture of the M. longissimus) and 8 computer image analysis traits (M. longissimus area, M. longissimus lean area, Ratio of Fat Area (RFA) to M. longissimus area, RFA to M. semispinalis capitis area, M. semispinalis lean area, RFA to M. semispinalis area, M. trapezius lean area and RFA to M. trapezius area) were determined. Seventy five Japanese Black and 147 Holstein bull samples were then genotyped for assessing the frequency of F94L in the commercial population in Japan, but there observed no variation and fixed to phenylalanine homozygote in both 2 breeds. It was strongly suggested that F94L mutation of MSTN gene have an effect for beef quality and quantity traits, however, it seems almost impossible to apply the F94L information to the current breeding system of the Japanese Black breed.

INTRODUCTION

The Quantitative Trait Loci (QTL) effects on 18 meat quality and quantity traits have been detected in the same centromeric region of BTA2 in our F2 population from a cross between Japanese Black sires and Limousin dams (Abe et al., 2008). Among the 18 traits, 6 were related to muscle mass and 8 were related to fat quality or fat accumulation. F2 animals that inherited the Limousin allele had greater muscle area and conversely, animals that inherited the Japanese Black allele gained higher values in fat-related traits. The highest F-value was observed in the lean area of M. longissimus derived from Computer Image Analysis (CIA) and a QTL effect for this lean area trait was also detected in other muscle parts (M. semispinalis and M. trapezius). Hence, this QTL seemed to play a significant role in muscle mass.

There are several reports on the detection of beef quality or quantity traits in the centromeric region of BTA2 (Casas et al., 1998, 2001; Alexander et al., 2007; Sellick et al., 2007). This region is also known as the muscular hypertrophy locus, where the Myostatin (MSTN) gene was isolated as the causative gene of the double-muscling phenotype in cattle (Grobet et al., 1997; Kambadur et al., 1997; McPherron and Lee, 1997b). MSTN is a Transforming Growth Factor-β (TGF-β) family member that plays an essential role in regulating muscle growth (McPherron et al., 1997a). Several distinct mutations have been identified that explain the increased muscling in cattle (Grobet et al., 1998; Dunner et al., 2003). This study fosuc on the detection of DNA polymorphisms in the MSTN gene of the F2 parental generation. A detected mutation was then genotyped in the F2 population and its relationship to the observed traits was assessed. Sire bulls of Japanese Black and Holstein breeds were also genotyped to estimate its genotypic frequency in the commercial population in Japan.

MATERIALS AND METHODS

Animals and phenotypes: The F2 population used in this study is described by Abe et al. (2008). Five carcass grade traits (carcass grade, M. longissimus area, BMS, luster and firmness), 5 meat quality traits by physiochemical analysis (crude fat, crude protein, moisture and C16:1 and C18:2 contents in intramuscular fat of M. longissimus) and 8 Computer Image Analysis (CIA) traits (M. longissimus area, M. longissimus lean area, Ratio of Fat Area (RFA) to M. longissimus area, RFA to M. semispinalis capitis area, M. semispinalis lean area, RFA to M. semispinalis area, M. trapezius lean area and RFA to M. trapezius area) of 186 F2 individuals were measured to assess their relationship with the detected mutation in the MSTN gene. All these traits had a detected QTL in the centromeric region of BTA2.

Table 1:	Oligonucleotide primers used in this study
aUTR: Untranslated Region

The details of the measurement methodologies and values are also described by Abe et al. (2008).

DNA extraction: To detect mutations in the MSTN gene, the genomic DNA of the parental generation (2 Japanese Black sires (JB-A and JB-B) and 2 Limousin dams (L-A and L-B)) of the F2 family were used. In addition, 1 Japanese Black and 1 Holstein bull samples (JB-50 and Hol-1) were used to detect mutations. This Japanese Black bull (JB-50) did not have a strong blood relationship with JB-A or JB-B. Genomic DNA was extracted from frozen semen from the 2 Japanese Black sires and from the whole blood of the 2 Limousin dams. Genomic DNA of JB-50 and Hol-1 was also extracted from frozen semen.

The DNAs of F2 individuals were extracted from whole blood samples. Seventy five Japanese Black and 147 Holstein bull samples were genotyped for the F94L polymorphism in the MSTN gene, for assessing the genotypic frequency. DNA was extracted from all the frozen semen stored in our stations.

Detection of DNA polymorphisms in the MSTN gene: To design primers for the detection of DNA polymorphisms in the MSTN gene, 3 nucleotide sequences deposited in the GenBank database (AJ438578 for the promoter region of MSTN, AB076403 for the MSTN gene and the partial sequence of AAFC03052712.1 (nt20353-nt20681) for the 3' untranslated region of MSTN) were used. The nucleotide position numbers in this report are relative to the nucleotide sequence of AB076403. The primers designed are shwon in Table 1. Initially, Polymerase Chain Reaction (PCR) amplification with genomic DNAs of 4 parents and 2 bulls (Japanese Black and Holstein) was performed using the primers listed in Table 1 (Mseq 01-27), in a 15 μL reaction mixture containing 20 ng of cDNA, 1.67 mM MgCl₂, 6.25 pmol of each primer, 0.2 mM deoxynucleotides (dNTPs) and 0.375 U Taq DNA polymerase (ABgene, Epsom, UK). Amplifications were carried out under the following conditions: 5 min at 94°C; 30 cycles of 30 sec at 94°C, 30 sec at the annealing temperature and 30 sec at 72°C and a final extension of 7 min at 72°C. The annealing temperature was 60°C for all primers. After amplification, DNA cycle sequencing was performed with approximately 20 ng amplified product by using a BigDye® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster city, CA) in an ABI PRISM® 3130x L genetic analyzer, using both forward and reverse primers.

Genotyping of F94L mutation by PCR-restriction fragment length polymorphism: In order to perform PCR-Restriction Fragment Length Polymorphism (RFLP) detection, a primer pair to amplify the fragment containing the MSTN mutations g. 415 C>A was designed (Table 1, F94L_TaqI). The PCR conditions were the same as those described earlier. The amplified 274 bp fragment was subsequently digested with TaqI to detect the g 415 C>A genotype. The restriction enzyme digestions were performed using 5 μL of PCR product mixed with 5 units of the appropriate restriction enzyme, followed by incubation at 65°C for 4 h. All the animals used in this study were genotyped using this method.

Statistical analysis: The effects of the F94L genotype on traits observed in F2 were assessed for significance using a one way analysis of variance with the FASN genotype as a treatment following the SAS GLM procedure (SAS Inst., Inc. Cary, NC). To test the pairwise differences between the effects of genotype, Tukey’s studentized range (HSD) test was also carried out.

RESULTS AND DISCUSSION

Twenty Single Nucleotide Polymorphisms (SNPs) were identified in the 6 investigated individuals (Table 2). Among these mutations, 2 SNPs were located in exon 1 (g 400 A>G and g 415 C>A). The SNP g 415 C>A was reflected in the codon at position 94 of the MSTN protein, TTC and TTA, which code for phenylalanine and leucine, respectively. This nonsynonymous mutation clearly differed between the 2 parental breeds of F2 and was C/C in Japanese Black and A/A in Limousin. The genotypes of F94L in JB-50 and Hol-1 were identical to JB-A and JB-B.

Next, F2 individuals were genotyped for the polymorphism of F94L in the MSTN gene by PCR-RFLP (Fig. 1) and the genotypic effects on 5 carcass grade, 5 meat quality and 8 computer image analysis traits were examined (Table 3-5). Significant effects were determined in all assessed traits, except for the C16:1 and C18:2 percentage of intramuscular fat, involved in 5 meat quality grade traits (Table 4). Overall, individuals that inherited the A allele at F94L showed higher values in muscle-related traits (crude protein (%), lean area of each of M. longissimus, M. semispinalis and M. trapezius (cm²)). Especially, the mean value of the M. longissimus lean area differed by almost 10 cm² between the two kinds of individuals homozygous in F94L. This trait marked the highest F-value in QTL analysis and the alleles that originated from the Limousin parents had an enormous effect on muscle mass (Abe et al., 2008). The observed effect of the A allele at F94L was the same as the QTL analysis results.

Table 2:	Detection of MSTN mutations in the 6 investigated individuals
aNumbers of nucleotide position were based upon AB076403 of GenBank database. bUTR: Untranslated Region

Table 3:	Carcass grade traits of MSTN genotype F94L mutation
Mean±SE: Mean values with different superscript letters in the same row differ significantly; a-b, a-c, a-d, a-ep<0.001; d-cp<0.01; b-cp<0.05; e-cp<0.1

On the other hand, individuals that inherited the C allele at F94L showed higher values in fat-related traits (BMS, crude fat (%), RFA to M. longissimus, M. semispinalis capitis, M. semispinalis and M. trapezius (%)) and its allelic effects were also the same as the QTL analysis results described in previous report (Abe et al., 2008).

Dunner et al. (2003) had already detected this mutation in several European cattle breeds, including the Limousin breed. All Limousin individuals in their study had the same non-disruptive MSTN haplotype, but their body conformational phenotypes, measured by their own criterion, were comparatively similar to those of individuals with the disruptive MSTN haplotype, which show the double-muscling phenotype. On the other hand, Sellick et al. (2007) reported the effect of MSTN F94L on carcass yield in their experimental back-cross population between Jersey and Limousin breeds. Like our study, they also detected QTL for similar traits such as meat percentage, eye muscle area and silverside (meat block composed of M. gluteobiceps and M. semitendinosus) percentage on the centromeric end of BTA2 by using their back-cross population. They focused on the MSTN gene as a candidate and observed a significant effect of F94L on carcass yield traits. Taken together, these data strongly suggest that the MSTN F94L mutation has an effect on carcass quality in beef cattle by regulating muscle mass.

Fig. 1:	Genotyping of F94L polymorphisms in MSTN gene. PCR products were digested with restriction enzyme Taq I. Arrows indicate the size of the DNA fragments. These fragments were size fractionated on 2% agarose gels

Significant effects on several fat-related traits were also detected. Because the amount of fat in the carcasses of double-muscled cattle is significantly less than that observed in the carcasses of conventional cattle (Hanset et al., 1982; Hocquette et al., 1999; Thiessen et al., 1982), these results could be explained simply as a reflection of the enormous effect of MSTN F94L on muscle mass.

However, there were several reports on the involvement of MSTN in adipogenesis, both in vitro (Rebbapragada et al., 2003; Artaza et al., 2005) and in vivo (Feldman et al., 2006) and another possibility could be discussed that the MSTN F94L mutation might also have an effect on fat-related traits.

Table 4:	Meat quality traits of MSTN genotypes for F94L mutation
Mean±SE: Units are all percentage. Mean values with different superscript letters in the same row differ significantly: a-bp<0.001; c-dp<0.1

Table 5:	Computer image analysis traits of MSTN genotype for F94L mutation
Mean±SE; Mean values with different superscript letters in the same row differ significantly: a-b, a-c, b-dp<0.001; b-c, c-dp<0.01; RFA: Ratio of Fat Area

The detected MSTN F94L mutation was located in the propeptide region and not in the functional mature region. Like other TGF-β super family members, MSTN is produced as a precursor protein (McPherron et al., 1997a; Derynck et al., 1985). Proteolytic processing cleaves between the propeptide and the C-terminal domain and the mature protein is released (McPherron et al., 1997a; Thies et al., 2001; Lee and McPherron, 2001). Both unprocessed and mature MSTN forms disulfide-linked dimers (McPherron et al., 1997a). In vitro, MSTN binds noncovalently to its propeptide after proteolytic processing, thereby producing a biologically inactive complex (Thies et al., 2001; Lee and McPherron, 2001).

Furthermore, overexpression of MSTN propeptide leads to an increase in muscle mass in transgenic animals (Lee and Mcpherron, 2001; Yang et al., 2001). Substitution at the 94th amino acid of MSTN may lead to some functional defects in the propeptide, which block inactivation of the mature peptide. Hence, it is very interesting and important to clarify whether F94L really has a functional effect.

Finally, to assess the genotypic frequency of F94L in the commercial population in Japan, 75 Japanese Black and 147 Holstein bull samples were genotyped. Surprisingly, no individuals with the A allele were found.

All animals of both breeds used in this study were C/C at F94L. In Europe, F94L mutation had limited distribution in only a few breeds such as Limousin, Pirenaica, Aubrac and Devon (Dunner et al., 2003). At the end of the 19th century, the Meiji government imported several European breeds into Japan. These European breeds included the Devon breed, which was crossed with native cattle in Shimane, Yamaguchi and Kagoshima prefectures (Namikawa, 1992). Thus, there remains a possibility of the existence of the F94L A allele in the Japanese Black breed in Japan. In the early modern breeding history of the Japanese Black, however, individuals that expressed the posterior muscle mass phenotype, the so-called pig rump phenotype, were eliminated from populations.

It may be that individuals with the A allele at F94L were involved in the pig rump population and exposed to the selection. In that case, it would be difficult to find such individuals in the Japanese Black population.

On the other hand, the allelic frequencies of the Japanese Black breed revealed more polymorphism than that in the Limousin breed (Table 2). It could be inferred that there exists a certain kinds of MSTN haplotypes in Japanese Black population and it is interesting to investigate the haplotypic effects of MSTN gene on beef quality and quantity traits in this breed, which was reconstructed with detected SNPs in this study.

CONCLUSION

According to the results of this study, F94L mutation in MSTN gene has the strong effect on beef quality and quantity traits by regulating muscle mass, but it seems almost impossible to apply this information to the current breeding system of the Japanese Black breed because of its fixed genotype.

ACKNOWLEDGEMENT

We thank to the staff of Tokachi and Ohu station, National Livestock Breeding Center, for feeding and measuring various traits of F2 population.