Introduction

Myostatin (MSTN, also known as growth differentiation factor 8) is a member of the transforming growth factor-β (TGF-β) superfamily (Sharma et al. 1999), which is well recognized as a negative regulator of skeletal muscle mass in mammalian species (Elliott et al. 2012). Mutations in MSTN acquired during early development may cause both hypertrophy and hyperplasia in adult skeletal muscle fibers (McPherron et al. 1997; Hamrick et al. 2000; Gentry et al. 2011). In addition to muscle, the expression of MSTN occurred in many tissues, e.g., liver, spleen, kidney, lung, and placenta etc. (Garikipati et al. 2006; Mitchell et al. 2006; Arai and Nishiyama 2007; Peiris et al. 2010; Jiao et al. 2011; Zheng et al. 2011), which suggests that the myostatin may have additional functions other than myogenesis (Argiles et al. 2012). In chicken, MSTN is expressed ubiquitously and most strongly in the testis and ovary (Kubota et al. 2007), and myostatin seems to have some important roles in the development. The deficiency of myostatin induced by the treatment using myostatin-specific small-interfering RNA delays cell differentiation and causes significant alterations in the cellular morphology of chicken embryonic myotubes (Sato et al. 2006).

The MSTN gene is highly polymorphic (Stratil and Kopecny 1999; Hill et al. 2010; Sjakste et al. 2011; Han et al. 2012; Sanchez-Ramos et al. 2012; Zhu et al. 2012), and many polymorphisms are breed specific. Seven single nucleotide polymorphisms (SNPs) in MSTN were identified in broiler and/or layer chicken lines (Baron et al. 2002), and four SNPs were identified in the MSTN of Bian chick (Zhang et al. 2011a; Zhang et al. 2012). Further studies on the growth traits show that the SNPs in chicken MSTN may affect the abdominal fat weight and percentage, breast muscle weigh and percentage, birth weight, and adult weight (Gu et al. 2004; Zhang et al. 2011a, b, 2012). At the same time, in many species, polymorphisms in the non-coding region of MSTN can also cause similar effects comparing with the phenotypes of the mutations in the MSTN coding region (Clop et al. 2006; Zhang et al. 2012), especially those in the enhancer box (E-box) motif (CANNTG), which is considered as one of the critical regulatory components in muscle gene expressions (Spiller et al. 2002; Salerno et al. 2004; Du et al. 2007; Funkenstein et al. 2009). Moreover, E-box has been shown to participate in the regulation of many myogenic factors (Dechesne et al. 1994; Jane et al. 2002; Heidt et al. 2007), e.g., the E-box promoter elements are suggested to induce the expression of cathepsin B gene during human myoblast differentiation (Jane et al. 2002).

The effects of MSTN E-box motif mutants on the expression of myostatin are well recognized in many animal models, including sheep, fish, mouse, and bovine (Spiller et al. 2002; Salerno et al. 2004; Du et al. 2007; Funkenstein et al. 2009), while the number and position of E-boxes show great varieties among different species. In the bovine MSTN upstream region, ten E-boxes locate in three clusters (Spiller et al. 2002), and seven E-boxes have been identified in murine MSTN promoter region, among which the E-box 5 (E5) plays a role in regulating the promoter activity in murine (Salerno et al. 2004). However, little is known about the distribution and function of E-boxes in chicken MSTN promoter region. Some SNPs in the MSTN upstream region have been described in chicken (Zhang et al. 2012), while none of them is in the E-box motif.

The Wenshang Luhua chicken breed (Gallus gallus) originally lived in a few ranges in Shandong province of China, mainly in Wenshang county, and it has been recognized as a domestic bird with good economic traits. As a result of competition by exotic breeds, Wenshang Luhua chicken is now an endangered breed. In this study, we sought to identify the E-box motifs in Wenshang Luhua chicken MSTN promoter region, determine the SNPs in the E-boxes and investigate the potential effects of these SNPs on the MSTN expressions.

Materials and Methods

Cloning MSTN and its upstream region.

The chickens (breed name: Wenshang Luhua, G. gallus) used in this study were bought from Poultry Institute, Shandong Academy of Agricultural Science, China. All chickens were housed, fed, and sacrificed as per the local standards for animal welfare issues. This study was approved by the Animal Welfare Committee of Shandong University (ECAESDUSM 20123010). The genomic DNA was isolated from the leg of 22 hatched 11-d chickens using the TIANamp Genomic DNA kit (Tiangen, Beijing, China). The full encoded MSTN-Wlc gene (MSTN gene in Wenshang Luhua chicken) flanking regions of each exon was amplified and cloned as described previously (Gu et al. 2004). The 2.3-kb wild-type MSTN upstream promoter fragment was amplified by PCR performed in a 50-μl reaction mixture containing 1 μg of genomic DNA, with the following PCR parameters: 33 cycles of (94°C for 40 s, 55°C for 40 s, and 72°C 5 min). The primers were Upper-F and Upper-R listed in Table 1. The PCR products were sequenced by Boya Biotechnology Company (Shanghai, China).

Table 1. Primers used in this study
Full size table

Bioinformatical analysis.

The sequence of MSTN-Wlc was compared with the database at the National Center for Biotechnology Information (NCBI) by using BLAST. The promoter sequence of MSTN-Wlc was analyzed with Mat Inspector (Cartharius et al. 2005) to predict potential E-box, myocyte enhancer factors 2 (MEF2) factor, glucocorticoid receptor (GRE), muscle-specific Mt-binding site (MTBF) and progesterone response element (PRE). All the currently available promoter sequences of vertebrates MTSN were retrieved from NCBI database to conduct the phylogenetic analyses. The nucleoside sequences of the promoters were completely aligned by ClustalX (Chenna et al. 2003) and phylogenetic reconstruction of the promoter sequences was carried out by using distance/neighbor joining (NJ) programs in MEGA v.3.1 software (Kumar et al. 2004). The interior branch length supports were from 1,000 replicates.

Construction of MSTNP W-GFP: the green fluorescent protein (GFP) expression vectors driven by the upstream promoter sequences of MSTN.

The 2.3-kb MSTN promoter fragment was cloned into the T vector system (Promega, Madison, WI) to generate TMSTNP, which was subsequently served as the template for generating other MSTN constructs. The PCR parameters were 30 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 2 min 30 s. The primers were T-F and T-R (Table 1). The GFP gene was amplified from Cam35S-GFP vector (Invitrogen, Carlsbad, CA) using the PCR (30 cycles of 94°C for 30 s, 58°C for 30 s, and 72°C for 1 min ) with the GFP-F and GFP-R primers (Table 1). The PCR was performed to obtain MSTN promoter-GFP fusion fragment using MSTN promoter fragment and GFP gene as templates, and T-F and GFP-R as primers (Table 1). The PCR parameters were 30 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 3 min 30 s. The vector pcDNA3.1 (−) (V795-20, Invitrogen) was digested with MluI and BamHI to remove cytomegalovirus (CMV) promoter, and the MSTN promoter-GFP fusion fragment with the same digestion was subsequently inserted in the vector to construct MSTNPW-GFP expression vector.

Construction of vectors MSTNP M4-GFP and MSTNP M3,4-GFP.

Vectors MSTNPM4GFP and MSTNPM3,4 were constructed with site-specific mutagenesis using the vector MSTNPW-GFP as the template. To construct vector MSTNPM4-GFP containing full-length MSTN promoter with a point mutation in E4-box (CACACG, the sequence of wild-type E4-box is CACATG), PCR was performed in a 50-μl reaction mixture containing E4-F and E4-R as primers (Table 1). The PCR parameters were 30 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 4 min 30 s. The expression vector carrying full-length MSTN promoter (8.5 kb) with both mutated E3 (GAAATG, the sequence of wild-type E3-box is CAAATG) and E4-boxes (MSTNPM3,4-GFP) was constructed using PCR with primers E4-F and E3-R (Table 1). All the vectors were sequenced to confirm the presence of correct sequences.

Cell culture and transfection.

Human embryonic kidney (HEK) 293T cells were used in the transfection assays. The cells were plated at a density of 4 × 104 cells/well in 24-well cell culture plates and incubated for 18–24 h. The transfection was conducted using DNAFect transfection reagent (CoWin Bioscience, Menlo Park, CA) according to its standard protocol with minor modifications. Plasmid DNA (0.8 μg) was diluted in 25 μl serum-free medium high-glucose Dulbecco’s Modified Eagle Medium (DMEM, Gibco, Carlsbad, CA) to prepare the transduction complexes. The serum-free high-glucose DMEM was used as transfection medium and the incubation time for the cells with complexes was 4–8 h. After transfection, the old medium was aspirated and 500 μl fresh high-glucose DMEM supplied with 10% fetal bovine serum was added, the cells were cultured for additional 48 h, then collected. RNA was extracted from cells using RNAsimple total RNA extract kit (Tiangen) and stored at −20°C for further use.

Real-time PCR to detect the expression of GFP in vitro.

Real-time PCR was performed to detect the expression of GFP in the transfected cells. The RNA was reverse-transcribed to obtain the cDNA and the expression of GFP was monitored using GFP-specific primers (RT-GFP-F and -R, Table 1) and SYBR Green dye-based detection. The real-time PCR was carried out in a 20-μl reaction mixture comprised of 10 μL 2× SyBr Green Mix, 0.4 μL of each primer (10 μM), and 2 μL template with the following parameters: 95°C for 10 min, followed by 40 cycles of (95°C for 15 s and 60°C for 60 s). The Ct method was used to compare the relative levels of expression of GFP, using a calibration factor of 2(−ΔΔCt). The relative starting quantities were calculated by comparison with the sample derived from MSTNPw-GFP-transfected cells as the standard. All values were normalized to those for β-actin (ACTB) mRNA (ACTB-specific primers were purchased from CW Biotech, Beijing China), which served as the internal control. All of the samples were measured in triplicates.

Real-time PCR to detect the expression pattern of MSTN in vivo.

To explore the in vivo function of the E-boxes upstream of MSTN, we extracted RNA from the leg of 22 hatched 11-d chickens, and monitored the expression pattern of the MSTN using real-time PCR. The cDNA was obtained by reverse transcribing the total RNA as described above. The real-time PCR was carried out in 50 μl reaction mixtures comprised of 25 μL 2X Sybr Green Mix, 1 μL primer F (10 μM), 1 μL primer R (10 μM), and 1 μL template with the following parameters: 95°C for 10 min, followed by 40 cycles of (95°C for 15 s and 60°C for 60 s). The primers for MSTN expression were RT-MS-F and RT-MS-R, and the primers for β-actin as the internal control were RT-AC-F and RT-AC-R (Table 1). The data were analyzed as described above.

Results

MSTN gene and its upstream region.

The sequence of MSTN gene in Wenshang Luhua chicken (MSTN-Wlc) showed very high similarities with sequenced MSTN genes in other breed, e.g., 98–100% identities. Since we focused on the promoter region of MSTN-Wlc (NCBI access no. KC405577), all the currently available upstream sequences of MTSN genes from different species were retrieved from NCBI database to carry out the phylogenetic analyses. As shown in the phylogenetic tree (Fig. 1), three sequenced promoters from G. gallus, including the one from Whenshang Luhua breed, were very close-related within a phylogenetic distance of approximately 1%, while one sequence from G. gallus (NCBI access no. AF478391.1) extruded from the group. The analysis also showed that these 12 promoter sequences could be divided into four deeply branching groups, which suggested extensive sequence variations in the upstream region of MSTN from different species. After further exploring the promoter sequence of MSTN-Wlc, several E-boxes, MEF2, and GRE were identified (Fig. 2). Interestingly, no MTBFs or PRE were predicted in MSTN-Wlc upstream, which commonly existed in several mammals’ MSTN upstream (Du et al. 2007). This is consistent with the previous finding that the expression patterns of MSTN were different among chick, yak, and pig (Sundaresan et al. 2008; Jiao et al. 2011; Zheng et al. 2011).

Figure 1.
figure 1

Phylogenetic analysis of upstream sequences of MSTN from 12 species/breeds. The bar indicated the evolutionary distance. The numbers on branch nodes were percentages of 1,000 sets of bootstrap supports.

Full size image
Figure 2.
figure 2

Sequence of the 2.3-kb region upstream of MSTN-Wlc. The sequences corresponding to the E-boxes are shown double-underlined and bold. The 13 E-boxes were denoted by E1 to E13, with E1 being the farthest upstream of the ATG start codon that is marked with dot beneath. The predicted MEF2 are labeled with shadow boxes and the GRE are in hollow boxes.

Full size image

Considering the important role in regulating the expression of MSTN (Spiller et al. 2002; Salerno et al. 2004; Du et al. 2007), the E-boxes were further analyzed. There were 13 E-boxes in the 2.3-kb upstream region of MSTN-Wlc, which were denoted by E1 to E13 according to their positions in the region. The E1-box was farthest from the coding sequence, and E13-box was nearest to the coding sequence (Fig. 2). Furthermore, many SNPs were revealed in the MSTN-Wlc promoter region (data not shown), but only two SNPs were in the identified E-boxes, i.e., E3- and E4-box. Polymorphisms were present in the first nucleotide of the E3-box and in the fifth nucleotide of E4-box (Table 2).

Table 2. Polymorphisms in the first nucleotide of the E3-box and in the fifth nucleotide of E4-box of MSTN-Wlc promoter
Full size table

The role of the MSTN-Wlc E-box polymorphism in regulating the expression pattern in vitro.

To further analyze the effects of the polymorphisms in the E-boxes of MSTN-Wlc, various upstream sequences were cloned into a GFP reporter construct, i.e., the vectors MSTNPW-GFP, MSTNPM4-GFP, and MSTNPM3,4-GFP, which carried the wild-type full-length upstream region, the upstream region with mutated E4-box, and the upstream region with both mutated E3- and E4-boxes, respectively. After transfection, the expression of GFP was monitored using real-time PCR. As shown in Fig. 3, the relative GFP expression level of the E4 mutant was about 55% of that of the wild-type, while the GFP expression level of the E3 and E4 mutants was 52% of that of the wild-type.

Figure 3.
figure 3

The effect of variations in the E-box sequences on the expression of GFP reporter gene in vitro. Expression vectors with the various E-box-containing sequences of MSTN-Wlc inserted upstream of the GFP gene were transfected into HEK 293T cells. The relative expression of GFP was measured by real-time PCR. On the x-coordinate of the plot, W represents the vector MSTNPW-GFP carrying the wild-type full-length MSTN-Wlc upstream sequence; E4 mutant represents vector MSTNPM4-GFP carrying full-length MSTN-Wlc upstream with the mutated E4-box (CACACG in the E4-box); E3 and E4 mutant represents vector MSTNPM3,4-GFP carrying full-length MSTN-Wlc upstream with the mutated E3- and E4-boxes (with both GAAATG in E3-box and CACACG in E4-box). Unpaired t test was employed for statistic analysis, *P < 0.05.

Full size image

Role of E-box polymorphisms in regulating the expression pattern of MSTN-Wlc in vivo.

The expression level of the MSTN-Wlc gene in the leg of 22 hatched 11-d chickens was analyzed by real-time PCR. The results showed that the polymorphisms in E3- and E4-boxes affected the expression level in vivo. For the two genome types of E4-box (TC and CC types), two sets of expression data showing the variation in MSTN-Wlc expression levels were obtained (Fig. 4). The results also indicated that chickens with the heterozygous genotype of E3-box and homozygous E4-box both had lower relative expression of MSTN-Wlc compared to that of wild-type (Fig. 4).

Figure 4.
figure 4

The correlation between variations in the upstream sequences and the expression of MSTN-Wlc in vivo. The expressions of MSTN-Wlc gene in 22 Whenshang Luhua chickens were analyzed by real-time PCR and correlated with their E3- and E4-box genotypes as described in Materials and Methods. On the x-coordinate of the plot, W represents wild-type of E3- and E4-boxes; E3 +/− represents heterozygous for E3-box [(C/G)AAATG in E3-box, five chickens]; E4 +/− represents heterozygous for E4-box [CACA(T/C)G in E4-box, five chickens]; E4 −/− represents homozygous for E4-box (CACACG in E4-box, four chickens). The number of individuals with polymorphisms is presented in Table 2. Unpaired t test was employed for statistic analysis. *P < 0.05; **P < 0.01.

Full size image

Discussion

In this study, the upstream promoter region of MSTN in Wenshang Luhua chicken was analyzed by cloning and sequencing analysis. Thirteen E-boxes were identified in the upstream of MSTN-Wlc and the polymorphisms of E-boxes were explored for the first time. Subsequently, analysis of the expression of endogenous MSTN-Wlc in vivo and MSTN-GFP reporters in vitro were both conducted. The effect of the E-boxes on expression in vitro is not identical to their effects in vivo, which may due to the presence of other regulatory mechanisms of muscle tissues in vivo that are absent in cultured 293T cells, or the differences caused by an artificial reporter system. Nevertheless, both results indicated that the polymorphisms in E-box altered MSTN expression levels in chicken. Furthermore, chickens with the heterozygous genotype of the E3-boxes and homozygous E4-boxes had a significantly reduced expression of MSTN-Wlc. The heterozygosity of E4-box did not affect expression any further as the heterozygosity of E3-box did. Therefore, we suggest that in spite of that both E3- and E4-boxes could play roles in vivo, the E3-box might be the more important in regulating the expression of MSTN-Wlc versus E4-box. In addition, the GFP expression data also implies that both E3- and E4-boxes are involved in regulating the expression of MSTN in a cooperative manner.

The role of E-box motifs in the regulation of mysotatin expression has been studied in many animal models, and different E-boxes are proposed to be involved in the regulation of MSTN expression in different species (Spiller et al. 2002; Salerno et al. 2004; Du et al. 2007; Funkenstein et al. 2009). The highly polymorphic nature of the myostatin gene and differences in the lengths of the upstream regions revealed in many studies might count for the variation in the multiplicity of the E-boxes. Eight, seven, and ten E-boxes were identified in the 1.2-, 2.5-, and 1.6-kb upstream regions of the ovine, murine, and bovine MSTN genes, respectively (Spiller et al. 2002; Salerno et al. 2004; Du et al. 2007). Among these E-boxes, the E6 in the bovine, E5 in the murine, and E3 and E5 together with E7 in ovine were found to be involved in the regulating the expression of MSTN (Spiller et al. 2002; Salerno et al. 2004; Du et al. 2007). From the results of this study, in Wenshang Luhua chicken, the E3- and E4-boxes were extruded from 13 identified E-boxes due to their regulating activities.

Due to the effects of myostatin on muscle mass, growth and other traits, the variations in MSTN are of great interest to animal breeding. Knowledge of MSTN null alleles and hypomorphs has been utilized to improve the selection of beef cattle and sheep (Georges 2010). A better understanding of the polymorphisms in MTSN in chicken will benefit the establishment of marker-assisted selection strategies for improving the agronomical traits. At the same time, our observations need to be extended in larger sample sets, including other species of chicken. In the future study, the polymorphisms in the upstream region and MSTN expression might be correlated with some important agronomical traits, such as muscle mass, growth, and resistance to infections. Thus, it will help to address the applicability of the molecular and sequential data regarding myostatin in breeding programs in chicken.