Integrative Analysis of Whole-Genome Bisulfite Sequencing and RNA-Seq in Skeletal Muscle of Xin’anjiang Water Buffalo

Simple Summary

Xin’anjiang water buffalo (XAJB), an indigenous swamp buffalo (Bubalus bubalis) in China, is mainly distributed in the Xin’anjiang River basin in Anhui Province. In this study, we used whole-genome bisulfite sequencing (WGBS) and RNA sequencing (RNA-seq) to explore the DNA methylation profiles and gene-expression patterns in skeletal muscle in the female (BF) and male (BM) groups of XAJB. The results showed that the most methylation changes occurred in CG context, and the distinct methylation landscapes were associated with the transcriptional signatures in the two cohorts. Due to the limited sample-size in this study, further research is needed to fully understand the association of methylation patterns with transcriptome profiles in the skeletal muscle of XAJB.

Abstract

Xin’anjiang water buffalo (XAJB) is crucial for meat production and agricultural activities in Anhui Province of China. To generate hypotheses regarding how DNA methylation might correlate with transcriptional differences in skeletal muscle, WGBS and RNA-seq were performed on three BF and three BM adult XAJB. The results revealed 31,333 differentially methylated cytosines (DMCs), 1961 differentially methylated regions (DMRs), and 230 differentially expressed genes (DEGs) in skeletal muscle between the two groups. The qRT-PCR results of ten DEGs (COL1A1, THBS1, SLITRK4, VIPR2, IGFBP6, WIF1, MMP16, LMOD3, NPR3 and MYLK4) enriched in protein digestion and absorption pathway, PPAR signaling pathway, ECM-receptor interaction pathway or PI3K-Akt signaling pathway were consistent with the RNA-seq results. Most methylation changes occurred in CG context, and sixteen genes were predicted as dual differential genes in both methylation and transcriptome. Moreover, CG methylation showed a significant negative correlation with gene expression within the 2 kb upstream region (rho = −0.42, p < 0.001). Given the limited number of animals examined, additional investigations with expanded cohorts are essential to verify the association between the methylome and transcriptome signatures underlying skeletal muscle in XAJB.

1. Introduction

In China, the swamp buffalo, characterized by 48 chromosomes (Bubalus bubalis carabensis, 2n = 48) [1], is an important source of meat, a traditional draft animal, and a vital asset for agricultural activities [2,3]. The Xin’anjiang water buffalo (XAJB) is a representative native swamp buffalo breed distributed in the Xin’anjiang River basin of Anhui Province. Its coat color changes distinctly with age: calves and young animals are ochre yellow, turning brown at 3–4 years and black by 5–6 years. Adult animals often display a crescent-shaped white band on the lower neck and chest. As a local characteristic breed identified in the 3rd National Census of Livestock and Poultry Genetic Resources, the XAJB possesses unique genetic diversity. It exhibits excellent traits, including strong resistance to local diseases and parasites, and high-tolerance to low-quality and coarse-fiber diets, making it valuable genetic-material for livestock breeding. Additionally, the XAJB has superior meat-quality and unique flavor, and serves as an important driver of local economic development.

DNA methylation, a key epigenetic modification in eukaryotes, is involved in numerous biological processes such as gene-expression regulation, genome stability, embryo development, cell growth and apoptosis, disease occurrence, and tumor progression [4,5]. Skeletal muscle, the largest metabolic organ in animals, directly affects livestock market-value through its fiber composition and extracellular matrix structure, which determine carcass yield, tenderness, and intramuscular fat content. Myogenic regulatory factors, including myoblast determination protein 1 (MyoD) and myogenic factor 5 (Myf5), crucially regulate skeletal muscle lineage specification and development. Carrio et al. [6] and Brunk et al. [7] found that demethylation of enhancers is essential for MyoD and Myf5 activation during skeletal myogenesis and muscle–stem cell differentiation. Zhang et al. [8] showed that DNA methylation affects intramuscular fat (IMF) deposition by regulating genes like COL6A1 in intramuscular adipocyte differentiation models. Ran et al. [9] revealed that the genes related to skeletal muscle (TBX1, MEF2D, SPEG, CFL2, and TWF2) were strongly correlated with the methylation-mediated expression switch. DNA demethylation is therefore a key driver of myogenesis.

Studies revealed that the variations in gene expression were negatively correlated with DNA methylation levels. Huang et al. [10] found that highly expressed genes exhibit a negative correlation between methylation and expression at multiple developmental stages of muscle-related tissues in Qinchuan cattle, indicating that differential gene methylation may partially contribute to growth difference between fetal and adult stages. Using MethylRAD-seq and RNA-seq, Chen et al. [11] demonstrated that DNA methylation status is associated with longissimus muscle-quality traits in Yunling, Wenshan, and Simmental cattle. M. Baik et al. [12] showed that DNA methylation regulates tissue-specific expression of adipogenic and lipogenic genes in the IMF and muscle portions of the longissimus muscle in Korean cattle. Overall, DNA methylation may regulate adipogenesis and lipogenesis, thereby influencing IMF deposition in the longissimus dorsi muscle of beef cattle.

XAJB is an important livestock breed used both as a draft animal and for meat production. Previous studies by Li et al. [13] have demonstrated associations between DNA methylation and gene expression across various genomic regions in the longissimus dorsi of Duroc pigs. Nevertheless, the transcriptomic and methylomic landscapes of skeletal muscle in XAJB remain poorly characterized. To explore potential associations between DNA methylation patterns and transcriptional profiles in skeletal muscle, we performed WGBS and RNA-seq on XAJB skeletal muscle to generate comprehensive methylome and transcriptome profiles. qRT-PCR was subsequently used to validate the RNA-seq data, followed by an integrated analysis to clarify DNA methylation-gene expression correlations.

2. Materials and Methods

2.1. Animal Materials and Sample Collection

Six 18-month-old XAJB (3 males: 326.9 ± 31.1 kg, and 3 females: 317.1 ± 37.8 kg) were randomly selected from the conservation farm in Xiuning County, Anhui Province, which has a subtropical monsoon climate at 300–600 m above sea level, a mean annual temperature of 22–23 °C, relative humidity of 85–90%, and annual precipitation of 900–1700 mm. All animals were fed a total mixed ration (TMR) comprising 20% concentrate and 80% roughage (dry straw and corn silage), with water provided ad libitum, and were housed for 12 months under identical feeding and management conditions. When male and female buffaloes were slaughtered at 30 months of age, their body weights reached 407.5 ± 38.2 kg and 394.3 ± 26.1 kg, respectively, and longissimus dorsi muscle samples were immediately collected, frozen in liquid nitrogen, and stored at −80 °C for subsequent analyses. Notably, the present study used a small sample size (n = 6, 3 males and 3 females) due to the protected status of XAJB, which represents a limitation that should be considered when interpreting the study results. A photograph of the XAJB is shown in Figure 1.

2.2. DNA Extraction, Library Construction and Bisulfite Sequencing (BS-Seq)

Six individuals from the F (n = 3) and M (n = 3) groups, named F1, F2, F3, M1, M2 and M3, were selected for bisulfite sequencing. Genomic DNA was extracted from longissimus dorsi muscle tissues using the traditional phenol–chloroform method. The quality and concentration of DNA were assessed by 0.8% agarose-gel electrophoresis and a NanoDrop^TM 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). DNA libraries for bisulfite sequencing were prepared as follows: genomic DNA was fragmented into 100–300 bp by sonication (Covaris, Woburn, MA, USA) and purified using the MiniElute PCR Purification Kit (QIAGEN, Hilden, Germany). The fragmented DNA was then enzymatically converted using the EpiArt^® DNA Enzymatic Methylation Kit (Vazyme, Nanjing, China). After end-repair and adenine (A) overhang addition, the fragments were ligated to methylated sequencing adapters. The converted DNA was amplified by PCR and sequenced on an Illumina HiSeq^TM 2500 platform (Gene Denovo Biotechnology Co., Guangzhou, China), yielding a mean depth of 30× with 90% of sites covered by ≥5 reads.

To obtain high-quality clean reads, raw data was filtered to remove reads containing more than 10% unknown nucleotides (N) or more than 40% low-quality bases (Q-value ≤ 20). Subsequently, paired-end reads mapping to the Escherichia coli genome (GCF_000005845.2) were removed using Bowtie2 (v2.4.5) [14]. The remaining clean reads were aligned to the reference genome GWHAAJZ00000000 (https://ngdc.cncb.ac.cn/) (accessed on 15 December 2024) using BSMAP (v2.90) [15], and cytosine methylation was identified as described by Lister et al. [16]. Finally, cytosine methylation information was extracted for subsequent analysis.

2.3. Genome-Wide Methylation Level Analysis

The methylation level was calculated as the percentage of methylated cytosines across the whole genome, each chromosome, and different genomic regions, for each sequence context (CpG, CHG, and CHH). Based on average methylation levels in 100 bp windows, we plotted the methylation profiles over the gene body (or transposable element) and its flanking 2 kb regions to evaluate context-specific patterns [17].

Differential DNA methylation between the two groups was detected with methylKit (v1.34.0) using Pearson’s χ² test [18]. To identify differentially methylated cytosines (DMCs), we applied sequence-context-specific thresholds: CG and CHG: |Δmethylation| ≥ 0.25 and q ≤ 0.05; CHH: |Δmethylation| ≥ 0.15 and q ≤ 0.05; all cytosines (C): |Δmethylation| ≥ 0.20 and q ≤ 0.05. For differentially methylated regions (DMRs), windows were required to contain at least the minimum number of cytosines listed below, with the same |Δmethylation| and q-value cut-offs: CG: ≥5 CpGs per window, CHG: ≥ 5 CHG sites per window, CHH: ≥ 15 CHH sites per window, all C: ≥20 cytosines per window.

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were subsequently performed on genes associated with DMCs/DMRs.

2.4. RNA Extraction, Library Construction, and Transcriptome Analysis

Total RNA was extracted using the TRIzol reagent kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. RNA purity was assessed by electrophoresis on RNase-free agarose gels, and RNA integrity was evaluated with an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA). Libraries were sequenced on the Illumina HiSeq 4000 platform (Gene Denovo Biotechnology Co., Guangzhou, China).

For data processing, fastp (v0.23.2) was used to filter raw reads, and sequences containing adapters, >10% unknown bases (N), or quality scores < 20 were discarded [19]. Clean reads were aligned to the reference genome GWHAAJZ00000000 using HISAT2 (v2.2.1) [20]. Transcript assembly and abundance estimation were performed with StringTie (v1.3.4) [19,20], and read counts per gene were obtained with featureCounts (v2.0.3) [21,22]. Gene-expression levels were quantified as fragments per kilobase of transcript per million mapped reads (FPKM). Differential expression was analyzed with DESeq2 (v1.46.0) [23], and genes with p < 0.05, FDR < 0.05 and |log2FC| > log2(1.5) were considered significantly differentially expressed. GO and KEGG enrichment analyses of the DEGs were performed using the Wallenius non-central hyper-geometric distribution implemented in the GOseq R package (v1.54.0) [24] and KOBAS software (v3.0.3) [25], respectively. GO terms or KEGG pathways with p < 0.05 were regarded as significantly enriched.

2.5. DEGs Verification Using qRT-PCR

After evaluated, the total RNA was reverse transcribed using HiScript^® IV 1st Strand cDNA Synthesis Kit (Vazyme Biotech Co., Ltd., Nanjing, China) according to the manufacturer’s instruction. For COL1A1 (collagen alpha-1(I) chain), SLITRK4 (SLIT and NTRK like family member 4), MMP16 (matrix metalloproteinase-16), LMOD3 (leiomodin-3), VIPR2 (vasoactive intestinal peptide receptor 2), IGFBP6 (insulin like growth factor binding protein 6), THBS1 (thrombospondin 1), WIF1 (Wnt inhibitory factor 1), VIPR2 (vasoactive intestinal peptide receptor 2), MYLK4 (myosin light chain kinase family member 4), NPR3 (natriuretic peptide receptor 3) and GAPDH (glyceraldehyde-3-phosphate dehydrogenase) target gene, real-time PCR was carried out on an ABI 7500 (Applied Biosystems, Foster City, CA, USA) platform with the Taq Pro Universal SYBR qPCR Master Mix (Vazyme Biotech Co., Ltd., Nanjing, China). The primer sequences (listed in

Table 1. Primer of DEGs for expression-profile analysis.

Primers	Primer Sequence (5′-3′)	Size (bp)	TM (°C)
COL1A1-F	CGGCGGCTACGACTTGA	225	60.0
COL1A1-R	GTTCTACACGGTGAGGCT
SLITRK4-F	AGGGTGCCTCCTCTTACA	202	60.0
SLITRK4-R	TAAAGTGGCTCAAGCTCC
MMP16-F	TTCCTCCACCAACAAGAC	234	60.0
MMP16-R	TCACCCTGTGGTTTCTCA
LMOD3-F	CGATGGGGAGATTGATG	151	60.0
LMOD3-R	CCCTGTTGGTGGCTTGT
VIPR2-F	AGATGTTGGCGAGACCG	343	60.0
VIPR2-R	AGAGCCACGACCAGTTC
IGFBP6-F	GGGTCTACACTCCCAACTG	225	60.0
IGFBP6-R	TGGAGATGGTGAGGAAGGG
THBS1-F	ATTGCCAAAGGAGGTGTCA	183	60.0
THBS1-R	GTAGGCGTGGCTGATGTAA
WIF1-F	GCGAGAGTGCTCATAGGAC	233	60.0
WIF1-R	TGGGTTGGCAGTTACAGGG
NPR3-F	CCAAGATGGGCGAGATGAT	159	60.0
NPR3-R	TGTAGGCGGACGTATGCAA
MYLK4-F	GGACCTGAAGCCTGAGAAC	175	60.0
MYLK4-F	AGTGGGGAAGGAGACAAAA
GAPDH-F	GGGTGTGAACCACGAGAAGT	233	60.0
GAPDH-R	TAGAAGCAGGGATGATATTC

) were designed using Primer Express 5, and the PCR protocol was as follows: denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 20 s and 60 °C for 20 s, then 72 °C for 30 s. Relative mRNA-expression levels were analyzed by the 2^−ΔΔCT method and normalized using GAPDH as an internal reference gene.

2.6. Correlation of DNA Methylation and Gene Expression

To determine the relationship of DNA methylation and gene-expression patterns in the skeletal muscle of XAJB, methylated genes were divided into three classes based on the methylation level: low-, medium-, and high methylation groups. Spearman’s correlation analysis was used to assess the statistical relationship between DNA methylome and transcriptome within gene bodies and their ±2 kb flanking regions; rho > 0 indicates a positive correlation, whereas rho < 0 indicates a negative correlation. Additionally, genes common to both DMR-related genes and DEGs were also predicted.

3. Results

3.1. Genome-Wide DNA Methylation Profiling in the Skeletal Muscle of XAJB

In order to ascertain the impact of DNA methylation on phenotypic diversity in skeletal muscle, WGBS was performed on the skeletal muscle tissue of XAJB. The bisulfite conversion efficiency was assessed using unmethylated λ-DNA spike-in controls and reached 99.21 ± 0.11%. The average depth profile and cumulative depth distribution across all samples fully satisfied the experimental requirements (Figure S1A,B), and the PCA chart revealed a distinct biological grouping in the BF and BM cohorts (Figure S1C). After data filtering, each sample yielded roughly 510 million high-quality clean reads. A total of 89.81% and 89.12% of the reads mapped to the reference genome (GWHAAJZ00000000) and were processed for further evaluation, respectively (Supplementary Tables S1 and S2). The average proportion of methylated cytosine residues was 4.73%, as detailed in Supplementary Table S3.

The global DNA methylation patterns revealed three distinct contexts, CG, CHH, and CHG, with ‘H’ representing A, C, or T. Among these contexts, CG sites exhibited the highest methylation levels in both the BF and BM groups, whereas CHG and CHH contexts displayed low methylation (Figure S2). Genomic methylation in female XAJB was characterized by percentages of 67.20%, 1.78%, and 1.82% for CG, CHG, and CHH contexts, respectively, and male buffalo showed similar patterns (67.73%, 1.75%, and 1.78%; Supplementary Table S3). Methylation rates of CG, CHG, and CHH were consistent across chromosome in the BF and BM cohorts, although the BF cohorts exhibited greater variation (Supplementary Table S4). The CG-methylation rate among chromosomes ranged from 60.18% to 72.63%, CHH from 1.17% to 1.85%, and CHG from 1.70% to 1.92%. Sequence characteristics of methylation sites (mCG, mCHH, mCHG) are presented in Figure S3A. Figure S3B shows the distribution of methylation levels; CG displayed a pronounced peak in the high methylation region (85–90%), whereas CHH and CHG were concentrated in the low-methylation region (0–5%). Different subregions of coding genes often exhibit distinct groupings. The average methylation levels of the gene body, exon, intron, CDS, 5′-UTR, 3′-UTR, upstream_2k, and downstream_2k regions are given in

Table 2. Statistical analysis of methylation levels of cytosine contexts across gene regions.

Sample	Region	C (%)	CG (%)	CHG (%)	CHH (%)
BF	Genebody	5.06	68.67	1.82	1.86
	Upstream_2k	3.84	26.61	1.78	1.79
	Downstream_2k	4.30	38.33	1.82	1.83
	5′_UTR	5.69	30.00	1.83	1.89
	3′_UTR	6.18	73.01	1.89	1.93
	Exon	6.25	49.36	1.85	1.85
	CDS	6.53	47.44	1.85	1.81
	Intron	4.96	72.28	1.81	1.87
BM	Genebody	5.11	69.02	1.78	1.82
	Upstream_2k	3.92	27.81	1.74	1.76
	Downstream_2k	4.46	41.24	1.78	1.78
	5′_UTR	5.58	28.27	1.78	1.83
	3′_UTR	6.26	73.02	1.84	1.87
	Exon	6.47	50.73	1.81	1.81
	CDS	6.97	50.35	1.82	1.77
	Intron	5.00	72.46	1.78	1.82

. The methylation levels of the BM and BF groups showed similar distribution patterns across all gene regions, with CG context exhibiting the highest methylation levels (ranging from 26.61% to 73.02%), while CHG and CHH contexts displayed relatively low methylation levels.

3.2. Identification of DMCs/DMRs and Functional Enrichment Analysis

By comparing the DNA methylation profiles of the BF and BM groups, extensive differences were uncovered. A total of 31,333 DMCs were identified, comprising 16,063 hyper-methylated and 15,270 hypo-methylated sites in BF vs. BM groups. Among the 18,509 CpG sites examined, 9637 were hyper-methylated and 8872 were hypo-methylated. Within the CHG context, 1557 sites showed an increase in methylation status, while 1463 sites exhibited a decrease in methylation status. In the CHH contexts, 4869 sites were identified with hyper-methylation and 4935 sites with hypo-methylation (Figure 2A). Furthermore, 1916 DMRs were detected in the CG context, one in the CHG context and 44 in the CHH context (Figure 2B). The heatmap of the top 100 DMR in CG are shown in Figure S1D.

To further understand the functional enrichment of genes affected by DMCs/DMRs, GO and KEGG pathway analyses were performed. GO annotations for genes associated with CG DMCs are presented in Supplementary Table S5, and the top 20 KEGG pathways for CG DMCs/DMRs-associated genes are displayed in Figure 2C,D and Supplementary Table S6. These genes are primarily involved in skeletal-muscle development. For instance, focal adhesion mediates muscle-cell attachment, migration, and differentiation; the phosphatidylinositol signaling system regulates cell growth, differentiation, and apoptosis; inositol phosphate metabolism modulates intracellular signaling; ECM-receptor interaction is required for muscle-cell adhesion, migration, and differentiation; and the Hippo signaling pathway is involved in the regulation of organ size and tissue homeostasis.

3.3. Identification of DEGs and Functional Enrichment Analysis

Comparative RNA-seq was performed to depicted the skeletal-muscle gene expression profiles in XAJB. The expression-level distribution plot and PCA chart demonstrated the robust reliability of the RNA-seq data (Figure S1E,F). The BF vs. BM comparison yielded 230 DEGs (172 upregulated and 52 downregulated genes) (Figure 3A and Supplementary Table S7). The GO enrichment of DEGs significantly enriched in terms are shown in Supplementary Table S8, and the top 20 enriched KEGG pathways are depicted in Figure 3B and Supplementary Table S9. Figure 3C shows the fold change in the selected DEGs determined by qPCR and RNA-seq (mean ± SD). Ten DEGs enriched in the protein digestion and absorption pathway, PPAR signaling pathway, ECM-receptor interaction pathway, or PI3K-Akt signaling pathway were selected for qRT-PCR. The mRNA of COL1A1, THBS1, SLITRK4, VIPR2, IGFBP6, WIF1, MMP16 was upregulated, and that of LMOD3, NPR3 and MYLK4 was downregulated in the qRT-PCR results, which is consistent with the RNA-seq results, indicating that the RNA-seq results were reliable.

3.4. Correlation Analysis

To determine whether the genomic DNA methylation level might be associated with global gene-expression in XAJB, transcriptomic and methylome data were integrated. Genes were assigned to four expression classes based on the expression, including a non-expressed group (RPKM (reads per kilobase per million reads mapped) ≤ 1), a low-expressed group (1 < RPKM ≤ 10), a medium-expressed group (10 < RPKM ≤ 100), and a high-expressed group (RPKM > 100). As shown in Figure 4A, the non-expressed group displayed the highest CG methylation in the 2 kb upstream region, and the methylation in this interval declined progressively toward the transcription start site (TSS) across all groups. The high-expressed group exhibited the lowest methylation levels within the gene body and 2 kb downstream region.

Analogously, genes were classified into none, low-, moderate-, and high methylation classes. Figure 4B illustrates the relationship between the methylation level and gene expression. In the 2 kb upstream region, high- and moderate CG-methylation groups contained the largest proportion of low-expressed genes. Conversely, the high CG-methylation group showed the highest prevalence of highly expressed genes within the gene body and 2 kb downstream regions. For CHG and CHH contexts, the unmethylated group exhibited the greatest proportion of low-expressed genes across the gene body.

Figure 4C depicts the overlapping distribution of CG/CHH DMR-related genes and DEGs. The 1916 DMRs in the CG context were associated with 548 genes, and 44 DMRs in the CHH context were associated with 18 genes. Only three genes were identified in both CG DMR-related and CHH DMR-related gene sets, whereas fifteen genes overlapped between CG DMR-related genes and DEGs, and merely one gene was shared by CHH DMR-related genes and DEGs. Notably, no genes were common to all three groups (Supplementary Table S10).

The association between DMRs location and gene-expression levels in the CG context is illustrated in Figure 4D. In the 2 kb downstream region of the TTS, expression-level changes associated with DOWN_DMRs (downstream differentially methylated regions) were slightly greater than those of UP_DMRs (upstream differentially methylated regions). Within the gene body region, expression-level changes for both DOWN_DMRs and UP_DMRs were similar. In the 2 kb upstream region, DOWN_DMRs exhibited marginally lower expression-level changes than UP_DMRs. Overall, across all examined regions, the expression-level changes for DOWN_DMRs and UP_DMRs were broadly comparable.

To elucidate the relationship between DNA methylation and gene expression, Spearman’s correlation coefficients were calculated. A significant negative correlation between CG methylation and gene expression was observed within the 2 kb upstream region (rho = −0.42, p < 0.001). In contrast, CHG and CHH methylation levels showed a positive association with gene expression within the gene body and the 2 kb downstream region. The results are illustrated in Figure 5 and Supplementary Table S11.

4. Discussion

Water buffaloes (Bubalus bubalis) are an essential livestock species in Asia, contributing significantly to agricultural, transport and food security [26]. Our comprehensive analysis reveals that methylation patterns were associated with a transcriptome profile in the skeletal muscle of XAJB, providing a possibility for exploring the meat-quality in this local Chinese breed.

In the present study, WGBS revealed distinct DNA-methylation patterns between BF and BM groups in the skeletal muscle of XAJB. The average proportion of methylated cytosine residues was 4.73%, and the global DNA methylation occurred in three contexts: CG, CHG, and CHH. CG sites exhibited the highest methylation levels in both groups, whereas CHG and CHH contexts showed comparatively low methylation. Comparable patterns have been reported in other mammals, indicating conservation of DNA-methylation mechanisms across taxa [27]. Within the XAJB genome, CG methylation was concentrated in the 3′-UTR and gene body of both BF and BM groups. Methylation in the 3′-UTR may influence mRNA stability and translational efficiency, while gene-body methylation can modulate transcriptional elongation and splicing [28].

By comparing the DNA methylation profiles of the BF and BM groups, 31,333 DMCs and 1916, 1, and 44 DMRs in the CG, CHG, and CHH contexts were detected, respectively. These DMC/DMRs-associated genes were reported to be involved in various biological processes that directly or indirectly participate in the regulation of meat-quality. For instance, the cAMP signaling pathway modulates muscle contraction and relaxation by altering intracellular calcium levels and protein phosphorylation [29], while the axon guidance pathway contributes to muscle innervation, which is essential for muscle development and function [30].

In the transcriptome analysis, we observed significant differences in gene expression profiles between BF and BM in XAJB. A total of 230 DEGs were identified, among which 172 DEGs were upregulated and 58 were downregulated in the BF vs. BM cohorts. The DEGs were enriched in multiple GO and pathways related to meat-quality development. Collagen is a major component of the muscle extracellular matrix. Genes encoding collagen proteins (COL1A1, COL1A2, COL3A1, and COL5A1) can affect meat-quality traits such as firmness, chewiness, and tenderness by determining the structure and mechanical properties of muscle connective-tissue [31]. The protein digestion and absorption pathway supplies amino acids for muscle protein synthesis, thereby promoting muscle growth and increasing meat yield [32]. The PPAR signaling pathway reduces muscle protein degradation by inhibiting the NF-κB pathway while activating the IGF-1 signaling pathway to promote protein synthesis, thereby increasing muscle mass and maintaining meat tenderness [33].

The PI3K-Akt signaling pathway activates downstream mTOR signaling to stimulate protein synthesis and cell proliferation, and modulates lipid metabolism to influence meat flavor and texture [34]. The ECM–receptor interaction pathway maintains muscle-cell structure and function, affecting meat toughness and elasticity, and its disruption can impair muscle development and reduce the meat quality [35]. It should be noted that these KEGG pathway enrichment results are based on a limited set of 230 DEGs identified in the present study, and thus the findings require further validation with larger sample sizes and more comprehensive transcriptomic analyses.

Integrated transcriptome-methylome analysis revealed that highly expressed genes showed relatively low methylation in both the gene body and the 2 kb region upstream of the TSS, whereas lowly expressed genes exhibited higher methylation. This inverse relationship is in accordance with previous reports across multiple species [33,36]. Furthermore, sixteen genes were predicated as dual differential genes displaying concurrent methylation and expression changes. Rnf122, CSRNP3, ME3, COG3, NPR3, and LMOD3 were higher expressed, and TMEM45A, PTK2B, HEYL, VAV2, MMP16, ADAMTS12, SLITRK4, IL1RAPL2, AFF2, and COL1A1 were lower expressed in BF cohorts. These overlapping genes are candidate genes potentially associated with the skeletal-muscle and meat-quality formation of XAJB, and their function requires further experimental validation. Luo et al. [37] indicated that AI has been widely applied in animal science. By integrating multi-dimensional data such as genomic- and epigenomic information, it is possible to accurately predict the meat quality-related traits of livestock, such as IMF content and tenderness. This suggests that in the future, AI technology can be combined to further verify the function of the overlapping genes in this study, providing a more efficient molecular-marker screening scheme for breed improvement.

Previous studies in cattle have also reported sexual dimorphism in muscle development and meat quality [38]. Wang et al. [39] discovered that 192 conserved transcriptional regulatory units (regulons) exist in mouse and human muscle-stem cells (MuSCs). This suggests that the differences in DNA methylation and gene expression in the skeletal muscles of males and females may be partially attributed to the sex-specific regulation of conserved regulatory factors, such as testosterone in males and estrogen in females. Engman et al. [40] also indicated that divergences in serum sex steroids regulate the hypothalamic-pituitary axis, modulate growth hormone (GH) release, and consequently reshape the specific skeletal-muscle profile of males versus females. Owing to the absence of data on serum testosterone, estradiol, and relevant verification experiments, the expanded sample size, dual-luciferase reporter assays and immunofluorescence assay should be employed in the future.

5. Conclusions

In summary, we revealed an association between the DNA methylome and the transcriptome landscape in the skeletal muscle of XAJB, which provides an idea for studying the meat quality in XAJB. Due to the limited sample size in this study, further studies underlying functional and experimental validation are required to confirm the biological relevance of the findings.