Identification and validation of a novel candidate gene regulating net meat weight in Simmental beef cattle based on imputed next‐generation sequencing

Abstract Objectives Genome‐wide association studies (GWAS) represent a powerful approach to detecting candidate genes for economically important traits in livestock. Our aim was to identify promising candidate muscle development genes that affect net meat weight (NMW) and validate these candidate genes in cattle. Materials and methods Using a next‐generation sequencing (NGS) dataset, we applied ~ 12 million imputed single nucleotide polymorphisms (SNPs) from 1,252 Simmental cattle to detect genes influencing net meat yield by way of a linear mixed model method. Haplotype and linkage disequilibrium (LD) blocks were employed to augment support for identified genes. To investigate the role of MTPN in bovine muscle development, we isolated myoblasts from the longissimus dorsi of a bovine foetus and treated the cells during proliferation, differentiation and hypertrophy. Results We identified one SNP (rs100670823) that exceeded our stringent significance threshold (P = 8.58 × 10−8) for a putative NMW‐related quantitative trait locus (QTL). We identified a promising candidate gene, myotrophin (MTPN), in the region around this SNP Myotrophin had a stimulatory effect on six muscle‐related markers that regulate differentiation and myoblast fusion. During hypertrophy, myotrophin promoted myotube hypertrophy and increased myotube diameters. Cell viability assay and flow cytometry showed that myotrophin inhibited myoblast proliferation. Conclusions The present experiments showed that myotrophin increases differentiation and hypertrophy of skeletal muscle cells, while inhibiting their proliferation. Our examination of GWAS results with in vitro biological studies provides new information regarding the potential application of myotrophin to increase meat yields in cattle and helpful information for further studies.


| INTRODUC TI ON
Domesticated animals, especially cattle, are recognized for their economic importance in countries around the world. 1 Cattle breeds have undergone organized selection to enhance beef production 2 according to breeding agendas based on meat-related traits. 3 Beef breeders prefer rapid lean muscle mass growth to meet dramatically increasing consumer demands for lean meat. 4 Hence, muscle tissue growth characteristics are economically very important.
Skeletal muscle development in the foetal stage is crucial because there is no net increase in muscle fibre number after birth.
Foetal skeletal muscle development involves myogenesis, adipogenesis and fibrogenesis, which are all produced by mesenchymal stem cells. 5 Skeletal muscle growth is achieved by an increase in myofibre number (hyperplasia) and size (hypertrophy). Hypertrophy, generally defined as an expansion of myotube size, is the main determinant of skeletal muscle mass. 6  viding key information about differentiation in myogenesis. MyHC isoforms also represent a robust tool for characterizing muscle fibre types in skeletal muscle. 8 MYH1 and MYH4 are typically expressed in fast muscle fibres, 9 whereas MYH3 is expressed preferentially in slow muscle fibres. 10 Hence, identifying genes that can upregulate myoblast differentiation and hypertrophy can provide a benefit for beef cattle production.
The Simmental breed of cattle is one of the oldest and most widespread cattle breeds in almost every region of the world. This breed is typically bred for fast-growing performance and lean meat production under a proper feeding regimen. [11][12][13] Many studies have examined Simmental cattle in terms of economically important traits, such as growth, 14 carcass quality, 15 meat quality 16 and meat yield. 17 A better understanding of the genetic variation in meat quality and yield will help guide breeders to adopt methods that enable beef market demands to be efficiently addressed.
As meat production in the beef industry has grown considerably, breeders have become increasingly focused on pinpointing quantitative trait loci (QTLs) and candidate genes that may affect cattle growth and meat production traits. Genes associated with muscle characteristic and development are distributed over many chromosomes, with relevant QTLs identified on bovine chromosomes 2, 3, 4, 6, 20 and 29. [18][19][20][21][22][23] Genome-wide association studies (GWAS) employing high-density SNP panels represent a powerful approach to detecting regions of the genome and genetic variants that can explain variation in complex disorders and clinically important traits in humans 24-26 and domesticated animals. 19,[27][28][29] GWAS outputs are sensitive to several factors, including sample size and the number of variants influencing a target trait. 30 Next-generation sequencing (NGS) can be used to identify many more genetic variants than are used in association studies employing SNP arrays. Sharma et al used NGS to localize 18 putative variants related to Mendelian diseases in Hanwoo cattle, 31 and another 33 genes related to domestication. 32 This method provides a robust strategy with which to explore genes with important influences on complex traits. In Simmental beef cattle, NGS has revealed several genes that regulate the dimensions of the hind quarters, including SLC13A1, LMOD2, WASL, IQUB, NDUFA5, ASB15 and PLXNA4. 18 However, the high densities of SNPs in NGS datasets complicate quantification of marker effects. 18,33 Moreover, a shortcoming of GWAS is errant rejections of the null hypothesis, leading to many false positives. 34 Hence, GWAS results must be validated before such findings can be applied for selection of the traits of interest as well as for confirming gene function and verifying the biological significance of detected genes. 34,35 Net meat weight (NMW, carcass weight without bone) has become an important parameter in the meat industry because of its direct correlation with other economically important traits, such as live weight and carcass weight. 36 The objectives of this research were firstly to conduct a GWAS with an imputed NGS dataset aimed at detecting NMW candidate genes in Simmental beef cattle and, secondly, to validate the functions of identified genes appearing to affect both the differentiation and proliferation of myoblasts. This line of research is important for breeding programmes because it provides comprehensive knowledge and confirmation of the associations. Ultimately, our long-term aim is to provide validated information for useful candidate genes to help Simmental breeders select for breeding animals that will yield offspring with increasing meat yields for consumers.

| Genotype analyses and quality control
Illumina Bovine HD SNP Beadchip (770k) and genotype analyses were performed in Illumina Genome Studio (Illumina, SD, CA, USA). We conducted quality control by withdrawing animals with high Pi-Hat values (Showing duplication of sample). The following strict criteria for excluding SNPs and animals were applied: SNP call rate < 90%, minor allele frequency < 5%, Hardy-Weinberg equilibrium deviation P < 10 −6 , and > 10% animals missing genotype data.
These criteria were tested in PLINK v1.07 software. 37 Following exclusion of 94 animals, a final cohort of 1,252 cattle with 671,204 autosomal SNPs were included in subsequent analyses. were acquired, in which poor-quality reads with > 10% undiscovered bases, >10% mismatches or > 50% low-quality bases were excluded.

| Resequencing
We removed duplicates from PCR amplification readouts in the formation of our library. The samples had an average sequencing depth of ~ 20× (quantity of sequences for each base).

| Imputation of SNP
Sequences with a minor allele frequency > 0.05 were imputed. A total of 21,043,398 sequence variants from the 44 genetically sequenced animals were analysed in BEAGLE v4.1, 38 (default setting) with algorithms determined by population data to deduce genotypes for animals with missing information and haplotypes. For imputed sequence variants, genotypes were labelled as 0, 1 or 2 for homozygotes, heterozygotes and alternative homozygotes, respectively.
Ultimately, we retained 12,468,401 SNPs for chromosomes 1-29 from the RNA sequencing data applying the key criterion of imputation quality > 0.1. 39

| Statistical model
A general mixed linear model for NMW was developed according to the formula y = + Xb + m j b j + Zu + e, wherein y represents a phenotypic value and μ is a population mean. Two fixed effect variables based on single marker regression were applied, including b, representing noise related to fixed effects (gender, weight, birth year and fattening days), and b j , representing the effect of an SNP. The parameter m j represents the vector for the i th marker, u is the polygenic effect presumed with N (0, σ 2 K), and K corresponds to the kinship matrix. Although all SNPs on autosomal chromosomes were eligible for inclusion, those SNPs on the chromosomes where m j resided, were excluded. The σ 2 parameter represents additive genetic variance, X represents the incidence matrix by which phenotypic values relate to fixed effects, and Z represents the matrix by which phenotypic values relate to polygenic effects. Finally, the variable e represents random residual effects, presumed in the formula V(e) = I 2 e , wherein I is the identity matrix and 2 e is the residual variance.

| SNP propagation
Our NGS dataset included a massive quantity of high throughput markers, which can complicate SNP p-value determination by conventional methods. 18,33 We dealt with this problem by segregating chromosomes into segments: segment 1, chromosomes 1-10 with 5,830,727 SNPs; segment 2, chromosomes 11-20 with 4,063,690 SNPs; and segment 3, chromosomes 21-29 with 2,573,984 SNPs.
Accordingly, segment-specific p-value thresholds were used depending on the number of markers in each segment.

| Primary cell isolation, cell culture, MTPN treatments for differentiation and hypertrophy
The longissimus dorsi (300 mg) was removed from foetuses, washed with phosphate-buffered saline (PBS) and diced into small pieces.
The fragments were digested with Dulbecco's modified Eagle's medium (DMEM, Gibco, Grand Island, NY, USA) with 0.1% collagenase type IV (Sigma, MO, TX, USA) for 45 min on at shaker at 37°C. The medium mixture was filtered through the 40-μm diameter nylon meshes, then supplemented with growth medium (GM) consisting of DMEM with 10% foetal bovine serum (Gibco, Grand Island, NY, USA) and centrifuged. The cells were resuspended in GM and seeded in petri dishes. One day later following attachment of muscle cells, the GM was replaced to remove dead cells. Cells were subcultured with trypsin (Amresco, MO, TX, USA) when they reached 75% confluency. Subsequently, we cultured the cells in 12well plates with GM. When cells reached 100% confluence, GM was changed to differentiation medium (DM) consisting of DMEM supplemented with 5% horse serum, and this day was considered to be day 0.

| RNA isolation, reverse transcription and quantitative real-time PCR (qRT-PCR)
RNA was extracted from cells with TRIzol reagent (Invitrogen, Carlsbad, CA, USA). RNA concentration and purity were monitored with a NanoPhotometer N50 (Implen, MU, DE); contamination was determined by 1.5% agarose gel electrophoresis.

| Immunofluorescence
Myotubes were fixed with 3-4% paraformaldehyde in PBS, washed three times with cold PBS 43

| Statistical analysis
The data were subjected to one-way and two-way analyses of variance (ANOVAs) with Tukey's post hoc tests. For each experiment, at least three replicates were considered. Statistical analysis was executed in GraphPad Prism (version 6.0, GraphPad Software, San Diego, CA, USA) with a significance criterion of P < .05.

| Association analysis revealed MTPN as a potential candidate gene of the NMW trait
Of 9,621,765,847 raw reads obtained, our quality control processes retained 9,584,920,309 reads. The quality of the sequencing data was excellent (Q20 ≥ 94.84% and Q30 ≥ 88.71%). 46 In segment 1, we identified one significant SNP on bovine chromosome 4 (rs100670823, P = 3.2 × 10 -8 ), which we listed as a candidate putative QTL related to NMW (see Manhattan plot for Segment 1 in Figure 2A). The marker rs100670823 was located 44 kb upstream of the MTPN gene which was a very promising candidate gene for NMW trait influence. Haplotype studies illustrated that this SNP was placed in a 22-kb haplotype block within a high-LD 5-kb span, as shown in Figure 2B. Remarkably, the amino acid sequences of MTPN were found to have 100% identity between cattle and human ( Figure S1).

| MTPN promoted differentiation of myoblasts into myotubes
To investigate the effect of myotrophin on the differentiation of bovine foetus derived myoblasts, cells were treated with 0 ng/mL, 10 ng/mL, 50 ng/mL, 200 ng/mL and 1000 ng/mL MTPN during differentiation.
qRT-PCR experiment showed that 1000 ng/mL MTPN significantly up- control; Figure 3C), in the 200 ng/mL and 1000 ng/mL groups (both P < .0001). The numbers of nuclei in myotubes after MTPN treatment are shown in Figure 3D. These data show that MTPN promoted myogenic differentiation in a dose-dependent manner.

| MTPN promoted myotube hypertrophy
As shown in Figure 4A, qRT-PCR showed that MTPN treatment (1000 ng/mL) from day 3 onward increased expression of MyoD, MYH1 and MYH3 significantly (P < .001 for MyoD and MYH1 genes with almost twofold increase compared to control; P < .0001 for MYH3 with more than threefold increase compared to control).

Analysis of MyHC-immunolabelled myotubes showed that MTPN
(1000 ng/mL) increased myotube diameter significantly (P < .05, Figure 4B,C). These results indicate that MTPN can have a positive effect on myotube hypertrophy in cattle.

| Biological function validation is important in post-GWAS study
The application of NGS data in association studies has numerous advantages over low-density genome variant studies for candidate gene detection, including higher throughput sequencing, more specific applicability, and higher read quality for sequencing dataset. 47 There are a variety of validation methods that can be applied to GWAS results, such as high-power statistical analyses (ie broad replication), genetic filtering (ie resequencing, deep sequencing, fine mapping), statistical filtering (ie genetic modelling, multiple loci, replication with heterogeneity) and phenome mapping, 34 as well as studies using different populations 48 and studies with population substructure corrections. 49 Biological and functional validation, including molecular function, in vitro, and in vivo studies, is critical for supporting GWAS results 34 and guiding the planning of follow-up.
In the present study, we attempted to verify an identified candidate gene in an in vitro functional study with proliferation and myogenic differentiation assays demonstrated to be reliable in vitro tools with which to investigate skeletal muscle development. 50-52  It was shown that S-myotrophin had stimulating effect on protein synthesis; such an effect was not seen with myoblast proliferation previously. 59 Hayashi et al reported that MTPN played crucial anabolic impact on protein synthesis, without altering incorporation of H-leucine into proliferating myoblasts, demonstrating that their data reflects an effect on hypertrophy rather than proliferation. 58 Here, we used CCK-8 proliferation assays to examine MTPN effects on proliferation and ob- Activation of NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) has been shown to be involved in the hypertrophic response to MTPN in neonatal rat ventricular cardiomyocytes, implicating the protein kinase C-IκB kinase-NF-κB pathway in mediating the MTPN-induced hypertrophic response in cardiomyocytes. 60 Indeed, deficiency of classical NF-κB signalling members enhances myogenic differentiation and alters myotube homeostasis. 61 Lu et al reported that muscle-derived stem cells isolated from the p65 ± mice had enhanced proliferation and myogenic differentiation compared to those from wild-type littermates. 62 However, the exact mechanism by which MTPN regulates muscle development requires further investigation.

| Differentiation promotion
In our research, we identified MTPN as a promising candidate F I G U R E 5 MTPN decreased proliferation of skeletal muscle cells, revealed in CCK8 assays, across four time points and four concentrations. One-way and two-way ANOVAs with Tukey's post hoc testing; ****P < .0001, ***P < .001, **P < .01 and *P < .05. Means ± standard deviations of six independent experiments are shown

E TH I C A L A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
All procedures were performed in accordance with regulations set by China's Council on Animal Care, and the study protocol

This study was supported by Cattle Breeding Innovative Research
Team (ASTIPIAS03 and ZDXT2018006), the Program of National Beef Cattle and Yak Industrial Technology System (CARS-37) and Pingliang Science and Technology Planned Project.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data underlying this study have been uploaded to Dryad. The raw genotype data are accessible using the following https://doi. org/10.5061/dryad.4qc06. Additional data are available using the https://doi.org/10.5061/dryad.5v3k1ct.