Genetic variants of the growth differentiation factor 8 affect body conformation traits in Chinese Dabieshan cattle

Objective The growth differentiation factor 8 (GDF8) gene plays a key role in bone formation, resorption, and skeletal muscle development in mammals. Here, we studied the genetic variants of GDF8 and their contribution to body conformation traits in Chinese Dabieshan cattle. Methods Single nucleotide polymorphisms (SNPs) were identified in the bovine GDF8 gene by DNA sequencing. Phylogenetic analysis, motif analysis, and genetic diversity analysis were conducted using bioinformatics software. Association analysis between five SNPs, haplotype combinations, and body conformation traits was conducted in 380 individuals. Results The GDF8 was highly conserved in seven species, and the GDF8 sequence of cattle was most similar to the sequences of sheep and goat based on the phylogenetic analysis. The motif analysis showed that there were 12 significant motifs in GDF8. Genetic diversity analysis indicated that the polymorphism information content of the five studied SNPs was within 0.25 to 0.5. Haplotype analysis revealed a total of 12 different haplotypes and those with a frequency of <0.05 were excluded. Linkage disequilibrium analysis showed a strong linkage (r2>0.330) between the following SNPs: g.5070C>A, g.5076T>C, and g.5148A>C. Association analysis indicated these five SNPs were associated with some of the body conformation traits (p<0.05), and the animals with haplotype combination H1H1 (-GGGG CCTTAA-) had greater wither height, hip height, heart girth, abdominal girth, and pin bone width than the other (p<0.05) Dabieshan cattle. Conclusion Overall, our results indicate that the genetic variants of GDF8 affected the body conformation traits of Chinese Dabieshan cattle, and the GDF8 gene could make a strong candidate gene in Dabieshan cattle breeding programs.


INTRODUCTION
Dabieshan cattle are a local breed, mainly raised in the central plains of China [1]. They are usually tawny, but individuals can also have black coloration. Dabieshan cattle are one of the most important local breeds in China because of the quality of their meat and adaptability. Given their longer generation times, high costs, and the increased labor they require, the growth of Dabieshan cattle by traditional breeding is slow. Body conformation traits, such as body length (BL), wither height (WH), hip height (HH), and heart girth (HG) are markedly lower in Dabieshan cattle compared with other exotic commercial cattle breeds and urgently need to be improved.
The growth and differentiation factor 8 (GDF8) gene, also known as myostatin, was first cloned in the mouse muscle cDNA library in 1997 [2]. Thomas et al [3] indicated that GDF8 plays an important role in muscle differentiation and growth by inhibiting the formation and differentiation of muscle cells and impeding the growth of skeletal mus cles. In Belgian Blue and Piedmontese cattle breeds, natural mutations, or deletions in GDF8 gene have been shown to promote the proliferation and hypertrophy of muscle cells and fibers, which resulted in a 20% enhanced muscle mass [4]. In Shaanbei White Cashmere goats, a 5bp indel in GDF8 was shown to affect body conformation traits, such as body height, HH, and chest width index [5]. Yang [6] showed that single nucleotide polymorphisms (SNPs) in GDF8 affected the body height, BL, HG, and hucklebone width in Nanyang, Qinchuan, and Jiaxian red cattle. These findings suggest that GDF8 could make an excellent candidate gene for body con formation traits in livestock.
The objectives of this study were to investigate the SNPs within GDF8, and their associations with body conforma tion traits in Chinese Dabieshan cattle. The result of this study could be used to improve the design of breeding plans.

MATERIALS AND METHODS
All procedures involving animals in this study were approved by the Animal Care and Use Committee of the Anhui Acad emy of Agricultural Sciences (approval number A19CS08).

Animals, genotypes, and phenotypes
The body conformation traits of 380 Chinese Dabieshan cattle managed in the Species Resources Protection Farm in Anhui, China, were studied. The animals were not preg nant, and all individuals were separated by more than three generations. Cattle were fed a diet based on the Nutrient Requirements of beef cattle (8th, NRC, 2016), which includ ed 25% concentrate and 75% roughage (corn silage and dry straw) on a total mixed ration basis, along with abun dant water.
Ear marginal tissues were collected from these animals. Genomic DNA was extracted using the TIANamp Genomic DNA Kit (TIANGEN, Beijing, China) and measured by a spectrophotometer. Genomic DNA was then diluted to 50 ng/μL for polymerase chain reaction (PCR) analysis. We also collected phenotypic data following the methods of Wang et al [10] and Yang et al [11]. Traits measured in this study were BL, WH, HH, HG, abdominal girth (AGR), hip width (HW), and pin bone width (PBW).

Statistical analysis
At the five SNP sites, the genotypic and allelic frequencies were counted from the genotype data. The HardyWeinberg equilibrium (HWE) of alleles was examined via a χ 2 test, which was performed using the POPGENE 3.2 software package.  [13,14]. D' represents the different degrees of linkage, and r 2 >0.33 indicates a strong linkage between SNP sites [15]. The mean of body conformation traits was calculated us ing the Bonferroni method in the SPSS 24.0 (IBM Company, NY, USA) software package. The association between SNP markers, haplotype combinations, and body conformation traits was analyzed using a general linear model with Bon ferroni method as follows: y ijk = u+g i +a j +s k +e ijk [16] Where y ijk is the phenotypic observation, u is the popula tion mean, g i is the fixed effect of animal genotype, a j is the fixed effect of age, s k is the random effect of sire, and e ijk is the residual error. Finally, a Bonferroni correction was per formed to determine the pvalue. The data were expressed as the mean±standard error, and differences were considered significant at p<0.05.

Species homology, phylogenetic tree, and motif analysis
Our sequence analysis revealed that the cDNA in GDF8 consisted of an 1,128bp open reading frame, flanked by 133bp 5′UTR and 1,476bp 3′UTR sequences. The coding sequence regions encoded a polypeptide of 375 amino acid residues with a molecular mass of 25.74 kD and an isoelec tric point of 9.58. Table 2 shows that each of the 5′donor and 3′acceptor splice sites conformed to the GTAG rule. Figure 1 shows the multiple sequence alignment in seven species (cattle, human, rat, goat, mouse, sheep, and pig). Simi lar structures including precursor protein containing signal peptide, TGFb_propeptide, and transforming growth factor beta (TGFB) were often observed among species. On the other hand, the phylogenic analysis of seven species revealed that cattle were closest to goat and sheep for GDF8 sequence, whereas the pig, human, rat, and mouse were relatively distant from the cattle branch. (Figure 2). A total of 12 significant motifs were identified in the related groups based on analysis of the supersecondary structure of GDF8 ( Figure 3).

Sequence variants and genetic diversity
Five SNPs were detected in GDF8. Figure 4 shows the aga rose gel electrophoresis (1.5%) of PCR amplified products using F1R1 and F2R2 primers; Figure 5 shows the sequenced map of five SNPs in bovine GDF8. In exon 1, g.244C>G and g.400G>A were detected; in exon 3, g.5070C>A and g.5076T>C were detected; in the 3′UTR, g.5148A>C was detected. Except for the SNP in the 3′UTR, the SNPs in exon1 and exon3 were synonymous mutations (g.244C>G: glycine; g.400G>A: glutamicacid; g.5070C>A: isoleucine; g.5076T>C: tyrosine). Table 3 shows the genotype frequency, allele frequency, HWE, and population diversity parameters of five SNPs in GDF8. Three genotypes were observed in all five SNPs. The diversity parameters indicated that the H e values ranged from 0.425 to 0.472; the range of N e values was from 1.738 to 1.892, and the PIC values ranged from 0.335 to 0.360. These data indicated that Dabieshan cattle have a medium level of genetic diversity at these five SNP sites. According to the χ 2 statistic, the genotypic frequencies of the g.244C>G, g.400G>A, and g.5148A>C mutations were in Hardy-Weinberg equilibrium. The χ 2 statistic for g.5070C>A and g.5076T>C indicated that they were in HardyWeinberg disequilibrium, which may be a consequence of artificial selection and the diversity of breed ing methods employed. Table 4 presents the results of the LD analysis between SNP markers. We categorized r 2 value of 0.33 or higher as strong LD. The r 2 values between g.5070C>A, g.5076T>C and g.5148A>C in GDF8 were greater than 0.330, indicating that the three SNPs had strong LD in Dabieshan cattle. The mean of the r 2 values between adjacent SNPs was 0.329, and the mean of D′ values between adjacent SNPs was 0.772.

Haplotype analysis
Singlesite association analysis has shown that haplotypes can contribute greatly to phenotypic variation [17]. Table 5 shows the result of haplotype analysis in Dabieshan cattle. Hap 1 (CACTA) showed the highest frequency (32%), whereas Hap 5 (GGACC) showed the lowest frequency (5%) in our analysis.      Table 6 lists the association of SNPs with body conformation traits in Dabieshan cattle. For g.244C>G, animals with geno type GG had a significantly higher mean AGR compared with the other genotypes (p<0.05). For g.400G>A, the in fluence of GG genotype resulted in the highest mean HW compared with animals with genotype AA (p<0.05). For g.5070C>A site, animals carrying the AA genotype had significantly higher HH and AGR than those with the CC genotype (p<0.05). At the g.5076T>C locus, our study also found that animals with genotypes TT and CC had greater WH than animals with genotype TC (p<0.05), and animals with the genotype CC had greater HH than animals with other genotypes (p<0.05). At the g.5148A>C locus, animals with genotype CC exhibited greater HH than the animals    with genotype AA and AC (p<0.05). Table 7 shows the association of combined haplotypes with body conformation traits in Dabieshan cattle. The frequen cies of combined haplotypes <5.0% were not considered. Dabieshan cattle with H1H1 resulted in the highest means for WH, HH, HG, AGR, and PBW compared with animals with other combined haplotypes (p<0.05).

DISCUSSION
In China, Dabieshan cattle are generally considered as an economically important local breed that is extensively farmed and produces a high quality product [1]. GDF8 has been shown to possess complex biological functions, especially in muscular development [3,4], bone metabolism [18], and body development. Many studies have shown that SNPs in GDF8 are associated with body conformation traits in goat [19], yak [20], horse [21], and cattle [6]. In this study, multiple sequence alignment indicated that GDF8 was highly homologous in seven species, suggesting that it may possess critically important functions. The results of the phylogenic analysis in this study were consistent with Wu et al [22], which showed that the GDF8 sequence of cattle was most like sheep and goat. Three genotypes were identi fied for the five SNPs, and the PIC indicated that GDF8 showed an intermediate level of polymorphism (0.25<PIC<0.5) in Chinese Dabieshan cattle. The r 2 values above 1/3 (r 2 >0.33) indicate that LD was sufficiently strong for mapping [15]. The mean r 2 indicated that LD among the g.5070C>A, g.5076T>C, and g.5148A>C sites was strong, and this may be attributed to the lower recombination and higher genotypic variation at these sites [23].
It was worth noting that, mutations of g.244C>G, g.400G>A, g.5070C>A, and g.5076T>C were synonymous polymor phisms in Dabieshan cattle. Our results were consistent with the study of Huang et al [24], which found that a silent muta tion of the bovine sterol regulatory elementbinding protein1c (SREBP1c) gene was associated with cattle body weight. Our results were also consistent with those of Xu et al [25] who found that a 'silent' SNP (g. 4617 A>C) in PAX3 dramatically improves the WH and BL of Chinese Nanyang and Caoyuan cattle. As Hunt et al [26] reported, a silent mutation did not alter the protein primary structure; however, this mutation might affect protein folding, alter the function, and modify the cellular response to specific targets by affecting messenger RNA splicing, stability, and protein structure. The SNPs in the 3′ UTR may be within or in the vicinity of the microRNA (miRNA) binding site, which could impair the regulatory functions of the associated miRNA [27,28]. Bioinformatics analysis in the current study showed that g.5148A>C was upstream of the btamiR29abcd binding site (http://www. targetscan.org/vert/); this may alter the level of GDF8 gene expression and affect phenotypes [29,30]. These findings were consistent with studies of Guanzhong dairy goats [31].
The bovine GDF8 gene is localized on chromosome 2 and has three exons and two introns. In this study, we detected five SNPs in bovine GDF8 gene: g.244C>G, g.400G>A, g.5070C>A, g. 5076T>C, and g.5148A>C. Data analysis re vealed that all five SNPs were significantly associated with their body conformation traits, and animals with the com bined haplotype H1H1 (GGGGCCTTAA) had significantly improved body conformation traits than animals with other haplotypes. Hap1 (CACTA) might also be associated with enhanced body conformation traits in Chinese Dabieshan cattle. Hence, we suggest that the combined haplotype H1H1 (GGGGCCTTAA) could be used as a marker to aid the selection of desirable characteristics of Dabieshan cattle in breeding programs.

CONCLUSION
Dabieshan cattle are economically important local breed in China. In this study, five SNPs and their corresponding hap lotypes were identified in Chinese Dabieshan cattle, and the association analysis revealed that all five loci were signifi cantly associated with their body conformation traits. Our results indicated that the combined haplotype H1H1 (GG GGCCTTAA) could be used as a marker to improve the body conformation traits of Dabieshan cattle in breeding program. Given the complexity of cattle breeding, additional studies are needed to examine the functional effects of GDF8 functional effects on body conformation traits.