Rapid Genotyping of MSTN Gene Polymorphism Using High-resolution Melting for Association Study in Rabbits

Animals, growth traits and carcass traits collection

This study was conducted in a total of 222 commercial meat rabbits including 88 Ira rabbits, 86 Champagne rabbits and 48 Tianfu black rabbits. The nutritional levels and feeding management were described previously (Zhang et al., 2011), with CP = 16%, DE = 10.8 MJ/kg. The carcasses were dissected according to the norms of the World Rabbit Science Association, and the data of growth traits and carcass traits were collected. The growth traits included 35-d-weight, 70-d-weight, and 84-d-weight; the carcass traits included eviscerated weight, semi-eviscerated weight, eviscerated slaughter rate, and semi-eviscerated slaughter rate (Blasco et al., 1993).

Genomic DNA isolation and mutation screening

Ear tissues were collected for genomic DNA extraction using AxyPrep Genomic DNA Miniprep Kit (Axygen, USA) and the DNA was stored at −20°C. A polymerase chain reaction (PCR) primer pair (Forward 5’-GAC TTGTGACAGACAGGGTTT-3’ and Reverse 5’-AATCT TTATGGCTTCTATTCTTG-3’) was newly designed in this study to amplify the 5’regulatory region according to the reference sequence of rabbit MSTN (GenBank accession number GU938461.1) by Primer 5.0, the PCR product size was 437 bp. The PCR reaction was performed with the following condition: one denaturation cycle at 94°C for 5 min; 35 cycles of 94°C for 30 s, 56°C for 30 s and 72°C for 40 s; and ended with an extension cycle at 72°C for 10 min. The 30 μl reaction volume included 15 μl 2×Taq PCR MasterMix (TIANGEN, Beijing, China), 3 μl DNA template (20 ng/μl), 9.6 μl ddH₂O, 1.2 μl of each primer (10 pmol/μl). PCR products from 16 rabbits were randomly selected to scan the genetic polymorphism in the 5’regulatory region using direct sequencing method by Shanghai Invitrogen Biotechnology Co., Ltd (Shanghai, P. R. China).

Genotyping using HRM

A primer pair (Forward 5’-TCATAGACCCTGACGAC ACT-3’ and Reverse 5’-CTTTTCAGTAATGGCAAGC-3’) was newly designed to amplify the small fragment containing the SNP revealed by direct sequencing, and the product size was 141 bp. The 10 μl reaction volume included 5 μl 1×SsoFast EvaGreen supermix (Bio-Rad, USA), 1 μl DNA template (20 ng/μl), 3.2 μl ddH₂O, 0.4 μl of each primer (10 pmol/μl), and was pipetted into each well on the 96-well PCR plate. The cycling condition was: one denaturation cycle at 98°C for 3 min; 40 cycles of 5 s at 98°C and 25 s at 57°C; following the step of 95°C for 10 s; a melting step ramped from 65°C to 95°C rising by 0.2°C each step with date acquisition for 10 s. All samples were amplified in duplicate, each plate included DNA from sequencing-verified wild-type, heterozygote, homozygote and no template control (NTC). The PCR reaction was performed on Bio-Rad CFX96 real-time PCR detection system (Bio-Rad, Inc., Hercules, CA, USA). The amplicon was subsequently subjected to HRM analysis (Martino et al., 2010).

Data analysis

The DNA sequences were assembled and analyzed using the DNAstar program (DNAS Inc, Madison, WI, USA). HRM curve data were analyzed using the Precision Melt Analysis™ software (Martino et al., 2010). Allele frequencies, genotype frequencies, heterozygosity (He), effective number of alleles (Ne) and Polymorphic Information Content (PIC) were estimated based on the study of previously described (Bindu et al., 2012). The effects of genotype on the traits were analyzed by the general linear model (GLM) procedure of SAS 9.2 program according to the following statistical model:

Where Y_ijk is studied traits, μ is the overall mean of observations, G_i is the fixed genotype effect of the MSTN mutation site, S_j is gender effect, B_k is breed effect, and e_ijk is the residual error.

Identification of SNP in rabbit MSTN

In previous studies, SNPs in rabbit MSTN was focused on the exon, intron and 3’-untranslated region. Fontanesi et al. (2008) sequenced the entire MSTN in four different breeds of rabbit and only detected one SNP in intron 2 (c.747+34C>T). Then Fontanesi et al. (2011) re-sequenced the MSTN including the three coding exons in 14 rabbits with different conformations and muscle mass, and identified another three SNPs of MSTN in exon 1 (c.108C>T), exon 2 (c.713T>A) and 3’-untranslated region (c.*194A>G). Kurkute et al. (2011) detected all three exons of MSTN in three rabbit breeds and found two mutations in exon 1 (c.1A>G) and exon 3 (c.799G>A).

In the present study, a T/C variant at −125 bp (relative to ATG start codon) of rabbit MSTN was identified by direct sequencing method (Figure 1). The T/C variant is a newly discovered variant in 5’regulatory region of MSTN. In silico analysis revealed that the allele C increased a transcription factor binding site of AP-2 and decreased a transcription factor binding site of CAAT (http://tfbind.hgc.jp/). The CAAT box is a motif present in the conserved sequence upstream of the start points in eukaryotic transcription units, which is recognized by a large group of transcription factors. AP-2 transcription factors are localized predominantly in the nucleus where they bind to target sequences and regulate transcription of target genes (Dawid et al., 2005). The change of these transcription factor binding sites might affect RNA transcription, which should be further investigated.

Figure 1.

The T/C variant at −125 bp of rabbit MSTN was detected by direct sequencing.

Genotyping of SNP in rabbit MSTN by HRM

HRM was performed on sequencing-verified samples in order to determine the Tm and characterize of the melting curve profiles of different genotypes. In this study, each genotype produced unique melt curves (amplitude and/or shape of the curves) and could be discriminated directly (Figure 2). The genotyping of SNP in MSTN by HRM was done as following: first, the curves were auto grouped by the software based on the melt transition, and then the observer confirmed the difference between Tm and melt curves. There was no difference in the classification by the observer compared to that grouped by the software.

Figure 2.

Three genotypes of the rabbit MSTN were grouped by HRM analysis.

The homozygous CC (79.4°C) melted at a higher temperature than the homozygous TT (79.0°C) due to the higher stability of the G ≡ C pairing compared to A = T. The difference in temperature could unambiguously differentiate the two homozygous sequences. The heterozygotes were identified based on the changes in the shapes of the melt curves as compared to homozygotes. The presence of both forms of homozygotes together with heteroduplexes results in a change in the melting pattern characterized by a flatter curve shape/melting transition (Raghavendra et al., 2011), as observed in the normalized melting curves (Figure 2).

Validation of sensitivity and specificity of genotyping by HRM

We selected 47 samples which were on the same plate to be sequenced for verifying the sensitivity and specificity of genotyping by HRM. Sequencing was performed by Shanghai Invitrogen Biotechnology Co., Ltd (Shanghai, P. R. China). The sensitivity and specificity of HRM were 0.96 (45 true positive/(45 true positive+2 false negative)) and 0.98 (92 true negative/(92 true negative+2 false positive)), respectively. Recent studies have also reported the sensitivity of HRM close to 100% except in the situation where amplicons have a high GC content (Garritano et al., 2009; Laurie et al., 2009).

Direct sequencing is the gold-standard approach for detecting SNPs, but involves additional steps leading to a higher turnaround time. HRM is a closed tube, fast method following PCR amplification and eliminates the possibility of template contamination, where samples can be analyzed in a single run (Raghavendra et al., 2011). HRM does not have 100% sensitivity; however, mutation scanning is a trade-off between the speed, cost, and sensitivity. It is worth to emphasizing how rapid this technique is: a plate of 96 samples can be melted in 10 min (Laurie et al., 2009). SsoFast EvaGreen supermix was used in this study. EvaGreen dye is a fluorescent nucleic acid dye with spectral properties similar to SYBR Green I and fluorescein. Unlike SYBR Green I, EvaGreen dye exhibits very low PCR inhibition. Thus, it can be used at high concentrations to generate greater fluorescent signals and provide increased sensitivity. Consequently, HRM is a better candidate to detect sequence mutations on a large scale, as it is much quicker than direct sequencing.

Genotype and allele frequencies in 5’ regulatory region of rabbit MSTN

The genotype and allele frequencies in the 5’ regulatory region of MSTN in different rabbit breeds is shown in Table 1. Allele C was the predominant allele with frequencies of 0.56 in Iraq, 0.58 in Champagne and 0.76 in Tianfu rabbits, respectively. The frequencies of genotype TC for Iraq and Champagne rabbits were the highest (0.64 and 0.59, respectively), but the frequency of genotype CC was the highest in Tianfu rabbits (0.56). The polymorphic site showed high heterozygosity (He = 0.48) and a high effective number of alleles (Ne = 1.91), and the genetic diversity was reasonably informative (0.25<PIC<0.50) in the three rabbits breeds. The genetic diversity revealed higher genetic variation and a selective potential that could be expected to get more genetic progress. In previous studies, a C-T transition at position 34 of MSTN intron 2 has been reported in New Zealand White, Soviet Chinchilla, Belgian Hare and several other rabbit breeds (Fontanesi et al., 2008; Bindu et al., 2012). In the present study, we randomly selected three rabbit breeds that were different from previous studies. The results showed similar frequencies and genetic characteristics.

Table 1.

Genotype and allele frequencies in 5’regulatory region of rabbit MSTN

Association between MSTN genotypes and production traits in rabbits

Some studies have showed that mutations in MSTN can inactivate its expression and further affect muscle fiber development and cause dramatic muscularity and a “double muscling” phenomenon in cattle (Wiener et al., 2002; Esmailizadeh et al., 2008; Grisolia et al., 2009). In the present study, the association analysis (Table 2) showed that different genotypes had a significant effect on the 84-d-weight of rabbits, individuals with the genotype TC indicated a significant association with higher 84-d-weight than genotype CC (p = 0.047). But MSTN genotypes had no significant effect on other production traits. Similar findings have been reported previously. Fontanesi et al. (2011) indicated that MSTN is not an important source of variability for performance traits, at least in the rabbit population analyzed. Bindu et al. (2012) also showed that a single nucleotide polymorphism (C-T transition) at position 34 of intron 2 of rabbit MSTN was associated with higher body weight but the difference in body weight among different genetic groups was not statistically significant.

Table 2.

The association analysis between MSTN genotypes and production traits in rabbits

As previously mentioned, the T/C variant at −125 bp caused the change of transcription factor binding sites which might affect RNA transcription. Sadkowski et al. (2008) also showed that polymorphism in the 5’-flanking region of MSTN was associated with the level of mRNA in skeletal muscle of bulls. Lin et al. (2002) showed that MSTN expression levels of Asian pigs increased markedly at 3 moths and maintained a constant level thereafter. Sun et al. (2012) also demonstrated that MSTN expression levels in muscle of Hu sheep were lowest at two days old, and tended to increase with the increasing age before 60 days. For these reasons, we could assume that MSTN expression levels might be relatively low in the initial growth process. Therefore, the T/C variant at −125 bp of rabbit MSTN had no significant effect on 35-d-weight and 70-d-weight. As the rabbit grows, the expression levels should gradually increase and the inhibitory effect on muscle growth should also gradually become obvious. At 84-d, the rabbit growth reached a relatively stable period and the inhibitory effect on muscle growth of MSTN might be most obvious. That caused the individuals with the genotype TC to have a significantly higher 84-d-weight than genotype CC (p = 0.047).