Analysis of genetic structure and identification of important genes associated with muscle growth in Fujian Muscovy duck

Abstract

Fujian Muscovy duck is a well-known meat waterfowl in Fujian Province due to its high meat production, superior breeding potential, and strong resistance. To fully explore the genetic characteristics of these advantages, Fujian black Muscovy duck and white Muscovy duck were used for whole-genome re-sequencing and transcriptome analyses. Population structure analysis showed significant differentiation between the two feather strains. Runs of homozygosity analysis indicated a stronger artificial influence on the black-feathered strain, with ROH island genes notably enriched in muscle tissue-related terms and pathways. Selective sweep and transcriptome analysis revealed a significant enrichment of genes linked to muscle tissue and muscle fiber-related terms and pathways. Key candidate genes identified, such as MEF2C, MYOZ2, and METTL21C, are believed to play crucial roles in meat production in Fujian Muscovy duck. This study offers a new perspective on improving meat production in Fujian Muscovy duck, which can benefit breeding strategies and production management.

Introduction

China has the world's largest meat consumer market. Between 2017 and 2021, total meat consumption increased. Pork consumption decreased from 63 % to 59 %, while poultry consumption increased from 22 % to 26 %. In Fujian, the market share of poultry, beef, and lamb continues to rise due to factors such as African swine fever and environmental changes. The development of animal husbandry in Fujian Province is influenced by climate and topography, resulting in advantages and disadvantages. Large livestock, such as cattle and sheep, meet challenges such as heat stress, while fodder cultivation is significantly limited, leading to high costs for large-scale expansion (Mathieu et al., 2020; Mishra, 2021). One optional approach to fill the meat supply gap in Fujian is breeding waterfowl.

Muscovy duck (Cairina moschata) is native to tropical regions of Central and South America. It has a long history of breeding in China dating back to 1729, mainly as a meat duck in the livestock industry. Fujian Muscovy duck can be divided into black-feathered strain and white-feathered strain. The black-feathered strain has black feathers with a greenish sheen, with only a few white feathers on the primary or secondary wings. The adult male muscovy ducks weigh between 3.3-4.2 kg, and the female muscovy ducks weigh between 1.9-2.5 kg. The white-feathered strain has pure white feathers, with the adult male muscovy ducks weighing between 3.0-4.1 kg and the female muscovy ducks weighing between 1.8-2.2 kg. Both strains exhibit excellent meat characteristics. The breeding performance of Fujian Muscovy duck has great potential. Crossbreeding with local duck varieties in Fujian, such as Jinding duck or Putian Black duck, produces offspring with excellent meat performance and nutritional value.

There has been extensive research on selection signals in poultry genomes due to the annual decrease in high-throughput sequencing costs. For instance, through de novo sequencing of the muscovy duck genome, genes that show strong adaptive evolutionary signatures related to immunity in muscovy ducks compared to domestic ducks have been identified (Jiang et al., 2021). Additionally, whole-genome resequencing (WGRS) identified genes associated with muscle contraction in the Muscovy duck (Gu et al., 2020b). WGRS integrated with RNA-Seq identified several genes related to muscle growth in Beijing ducks, including MEF2A and TNNT2 (Wang et al., 2017). Genome-wide association study revealed that EDNRB2 and XBP1 expression cause variation in black spot size on the body surface of domestic ducks (Xi et al., 2021). Gene chips were used to conduct linkage analysis of differently colored chickens, identifying WNT3 and GH as genes that control plumage color in chickens (Sun et al., 2015).

To improve production efficiency in poultry, exploring methods to increase muscle growth rates is a logical approach (Keel and Lindholm-Perry, 2022). However, research on Muscovy duck muscle growth remains insufficient. This study used WGRS and RNA-Seq methods to identify the genetic relationship between the two feather-color strains of Fujian Muscovy ducks. Population structure analysis was conducted to screen genes related to meat growth in Fujian Muscovy ducks through selection clearance analysis and RNA-Seq analysis. Potential selection markers in the Muscovy duck genome were revealed through runs of homozygosity (ROH) analysis. The study contributes data to research on poultry muscle growth and provides a basis for the development and utilization of Fujian Muscovy duck resources.

Materials and methods

Samples collection

The study involved 30 300-day-old Fujian black muscovy ducks and 30 300-day-old Fujian white muscovy ducks, which were raised on Zhangzhou Changlong Agriculture and Animal Husbandry Co., Ltd. (Zhangzhou, Fujian, China). Blood samples were collected from the wing veins of those 60 muscovy ducks. Breast muscle samples were collected after beheading at the neck under deep anesthesia. The experimental animals were selected to cover almost all the existing genetic relationships of the species.

DNA extraction and WGRS

DNA was extracted from blood samples using the TIANamp Blood DNA Kit (Tiangen Biotech Co., Ltd., Beijing, China). The concentration and integrity of the DNA were checked by BioTek Epoch2 (Agilent, CA, USA) and agarose gel electrophoresis at a concentration of 1.5 %. Yingzi Gene Technology Co., Ltd (Wuhan, China) prepared DNA libraries for WGRS, which were sequenced on a DNBSEQ-T7 platform with 150 bp paired-end reads to obtain raw reads.

Variants calling

The raw reads underwent processing using fastp (v0.23.1) (Chen et al., 2018). The following standards were applied: (1) removal of reads with more than 40 % of low-quality bases (Phred quality score ≤20), (2) removal of reads less than 30 bases, (3) removal of barcode adapters, and (4) removal of reads with >10 % unidentified nucleotides (N). Subsequently, module Mark Duplicate of Picard Toolkits (https://broadinstitute.github.io/picard/, v2.18.24) was used to eliminate PCR-duplicate reads. Modules bwa and driver of sentieon were used to align the clean reads to the muscovy duck reference genome (GWHBJBF00000000), followed by sorting, and modules Haplotyper and GVCFtyper were used for single nucleotide polymorphisms (SNPs) detection at the population genomic level (Freed et al., 2017). ANNOVAR was used to functionally annotate all genomic variations (Yang and Wang, 2015). PLINK (v2.0) (–geno 0.1), Picard Toolkits (–MIN_GQ 20), and bcftools (v1.5) (MAC > 0) were used for quality control of the obtained SNPs to obtain high-quality biallelic SNPs (Chang et al., 2015; Danecek et al., 2021).

Population structure analysis

The black-feathered ducks and white-feathered ducks were divided into black group and white group respectively. ADMIXTURE was used to individually calculate the genome variation of each sample with kinship (K) set from 2 to 5, which observed Hardy-Weinberg equilibrium (Alexander et al., 2009). A Bayesian model-based on clustering algorithms was then used to compute the likelihood that the variation originated from the Kth generation, thereby inferring population structure. High-quality biallelic SNPs from section 2.4 were processed using PLINK (v2.0) to calculate genetic distances between pairs of samples (Chang et al., 2015). An evolutionary distance matrix was constructed using the distances between the two groups. Sample clustering analysis was then performed using MEGA11 by iteratively merging the two nodes with the smallest distance, constructing a phylogenetic neighbor-joining (NJ) tree to infer genetic relationships between populations (Tamura et al., 2021). After the evolutionary tree was constructed, R package ggtree was used for visualization (Yu, 2020). Module smartPCA of EIGENSOFT was used to conduct principal component analysis (PCA), and the significance of eigenvalues was tested using Tracy-Widom (Price et al., 2006). Principal component plots were generated via R package ggplot2 (Wickham, 2016). PopLDdecay was employed to analyze the results of the linkage disequilibrium decay, with the parameter "-MaxDist 1000 -MAF 0.05″ utilized (Zhang et al., 2019). The LD decays were subsequently plotted using the R package ggplot2 (Wickham, 2016).

ROH analysis

Based on previous studies, the following parameters were used to calculate ROH via PLINK (v2.0): (1) a minimum of 30 SNPs to constitute an ROH, (2) at least 1 SNP per 100 Kb within each ROH, (3) a minimum length of 250 Kb for each ROH, (4) allowing at most one possible heterozygous genotype and two missing genotypes within each ROH, and (5) a minimum distance of 1000 Kb between consecutive homozygous and heterozygous SNPs (Chang et al., 2015; Gao et al., 2023). To calculate an individual's genomic inbreeding coefficient (F_ROH), divide the total length of all detected ROHs by the total length of the autosomal genome covered by SNPs. ROH islands are defined as regions where the SNP density within each ROH accounts for the top 1 % of all ROH regions, following previous research. The R package rMVP used to visualize the distribution of ROH across the whole genome of two populations (Yin et al., 2021). The KOBAS were utilized for GO and KEGG enrichment analysis of genes in ROH islands, and the results were visualized using the R package ggplot2 (Wickham, 2016; Bu et al., 2021).

Selective sweep analysis

Hypothetical selective regions were identified on the sample genomes and calculated F_ST using the R package PopGenome, following established study (Pfeifer et al., 2014; Gu et al., 2020a; Liu et al., 2021). Specifically, each chromosome was segmented into 250 Kb-long windows. The starting position of each window was determined by shifting 10 Kb forward from the preceding window's start, and the initial window on each chromosome was positioned at the chromosome's first base. Following computation, regions with F_ST values in the top 5 % and exceeding 0.25 were selected, and genes in regions were extracted. The distribution of F_ST values across the entire genome was visualized using the R package rMVP (Yin et al., 2021). Subsequently, the KOBAS were utilized for GO and KEGG enrichment analysis of genes in regions, and the results were visualized using the R package ggplot2 (Wickham, 2016; Bu et al., 2021).

Preparation and observation of muscle slices

The weight of breast muscle samples was measured first. Five samples were randomly selected from each group, immersed in a 4 % paraformaldehyde solution at a 20-fold volume, fixed for 24 h, and subsequently cross-sectioned. Two random locations were selected from each sample for sectioning. Following the completion of slicing, fields with satisfactory integrity were randomly chosen and photographed at 40x magnification. Employing the ImageJ (v2.14.0) Fiji distribution, the area occupied by intact muscle cells within each field of view was observed and quantified (Schindelin et al., 2012; Rueden et al., 2017). The data were analyzed and visualized by Graphpad Prism 8, and results are presented as means ± standard error (SEM).

RNA-Seq analysis

For the 10 samples selected in previous section, three samples in each group were randomly selected for RNA extraction before 4 % paraformaldehyde solution treatment. The total RNA was extracted from six samples using Trizol reagent (AGBio Tech Co., Ltd., Hunan, China) and SteadyPure RNA Extraction Kit (AGBio Tech Co., Ltd., Hunan, China). Yingzi Gene Technology Co., Ltd (Wuhan, China) prepared the cDNA libraries necessary for RNA-Seq. The DNBSEQ-T7 platform was used for sequencing with 150bp paired-end reads length. Quality control was performed to obtain clean reads. The clean reads were aligned to the reference genome (GWHBJBF00000000) using Hisat2, and gene annotation was conducted (Kim et al., 2019). Transcript expression levels were assessed using StringTie, which led to the identification of differential expression genes (DEGs) and prediction of novel transcripts (Shumate et al., 2022). The six samples were divided into two groups: the first group comprised three white ducks, and the second group included three black ducks. DEGs were filtered using DESeq2 with criteria set at |log2fc| > 1 and FDR < 0.05 (Love et al., 2014). Enrichment analysis for GO and KEGG was performed using the KOBAS (Bu et al., 2021). The iRegulon module of Cytoscape (v3.10.1) was used to conduct transcription factor analysis on the obtained DEGs with default parameter settings, and the motifs and tracks database of transcription factor analysis was constructed by Stein Aerts Lab (http://iregulon.aertslab.org/collections.html#motifcolldownload) (Shannon et al., 2003; Janky et al., 2014).

Quantitative real-time PCR

The total RNA extracted in previous section was also utilized for qRT-PCR. Subsequently, reverse transcription into cDNA was performed using the Evo M-MLV RT Mix Kit (AGBio Tech Co., Ltd., Hunan, China). A total volume of 20 μL was prepared for qRT-PCR using 10 μL of 2× SYBR Green Pro Taq HS Premix (ROX Plus) (AGBio Tech Co., Ltd., Hunan, China), 0.5 μL of cDNA, 0.4 μL each of sense and antisense primers with concentration of 10 μM, and 8.7 μL of water DEPC-treated. The reactions were conducted on the QuantStudio™ 5 Real-Time PCR System (Thermo Fisher Scientific, Waltham, USA) to accurately quantify cDNA expression levels. The amplification conditions consisted of an initial denaturation at 95 °C for 30 s, followed by 40 cycles of denaturation at 95°C for 5 s, annealing at 60 °C for 30 s. Each sample was prepared in triplicate to ensure mutual validation. Expression levels of the target gene were normalized to β-actin as a housekeeping gene. The primers used are listed in Table S6. Data analysis was conducted using the Livak method, and results are presented as means ± standard error (SEM).

Statistical and reproducibility

Statistical analysis was performed by Graphpad Prism 8 software. The section and qRT-PCR data were presented as means ± standard error (SEM). The two-tailed student's t-test was used to assess the significance of the comparison between two groups. P values < 0.05 and 0.01 or 0.001 were present as significant and extremely significant difference, respectively.

For the RNA-Seq data, the two-tailed student's t-test was first used to test the significance of Log2FC between two groups, and then the false discovery rate is calculated via Benjamini and Hochberg method. padj < 0.05 were present as significant difference.

RESULTS

WGRS of fujian muscovy duck

The average raw total bases per sample were 12.44 Gb. Reads with more than 40 % of low-quality bases, unidentified nucleotide (N) content >10 %, less than 30 bases, and barcode adapters were removed, followed by quality control to obtain clean reads. The analysis results (Table S1) showed that the Q20 quality range was between 91.61 % and 98.11 %, with a mean of 96.92 %, while Q30 ranged from 74.39 % to 94.1 %, with a mean of 90.57 %. After variant calling, a total of 947,742 SNPs were detected (Figure S1).

Population structure analysis

A combined approach of ADMIXTURE and Cross-Validation (CV) was used to determine the ancestry proportions between the two sample groups and conduct population structure analysis. The results of population structure analysis and CV error (Figure 1A, S2) showed that the black and white groups diverged at the second generation. At k = 3, the black group exhibited a more complex lineage compared to the white group.

Samples with high sequence similarity were merged, and the Kimura two-parameter model was used to resample 1000 times, generating reliable phylogenetic tree results with bootstrap values exceeding 80 % (Fig. 1B). The analysis indicated that the black group diverged into two branches, while the white group formed a single branch. Notably, individuals 20, 23, 24, 27, and 28 of the black group exhibited differences from other individuals within the same group. In the results structure, when k = 4, these five individuals showed different lineages from the other individuals, providing further evidence that the white group may be influenced by natural or artificial selection.

Fig. 1.

Population structure analysis of the black and white groups. a) Population structure of the two duck groups; b) Phylogenetic tree of 60 individuals of Fujian Muscovy ducks, with red representing the black group and blue representing the white group; c) PCA of 60 Fujian Muscovy ducks, with red representing the black group and blue representing the white group. The red and blue circles represent a 95 % confidence level.

Note: n = 30 per group.

The PCA results for both groups (Fig. 1C) showed that samples from each group formed two distinct clusters along the first component. However, the black group exhibited more excellent dispersion than the white group along the second component, which is consistent with the results of the population structure analysis.

The linkage disequilibrium decay analysis revealed that the white group exhibited a faster decay than the black group (Figure S3). Furthermore, the entire group of muscovy ducks in this study decayed faster than the individual plumage color subgroups. This suggests that the black group was subjected to a higher intensity of domestication than the white group, which is consistent with the results described above.

ROH analysis

In Fujian Muscovy ducks, we detected a total of 15,173 ROH, with their lengths and quantities illustrated in Fig. 2A, and distribution of F_ROH shown in Fig. 2B. Specifically, black group exhibited 8,202 ROH, totaling 7,462.09 Mb, with a median F_ROH of 0.2114, while the white group had 6,971 ROH, totaling 5,606.07 Mb, with a median F_ROH of 0.1676. These ROH characteristics suggest that the black group is more influenced by natural or artificial selection compared to the white group, which aligns with the results of our population structure analysis. Additionally, we calculated the individual lengths of ROH, revealing that the proportion of ROH lengths exceeding 1 Mb was more significant in the black group than in the white group (Table 1).

Fig. 2.

The genome-wide ROH scan. a) The X-axis represents the number of ROH per group, while the Y-axis represents the total length of ROH per individual. The color red represents the black group, and blue represents the white group. b) Boxplots illustrating the F_ROH for both groups, derived from F_ROH values. Outliers are denoted by points, with red indicating the black group and blue indicating the white group. c) Manhattan plot showing ROH distribution across the genomes of 60 Fujian Muscovy ducks. The red dotted line is 93.33, which represents the ROH threshold, with regions above the dashed line indicating ROH islands.

Note: n = 30 per group.

Table 1.

Comparison of ROH length between two groups.

	Black	White	Black(%)	White(%)
0-1 Mb	6133	5518	74.77	79.16
1-2 Mb	1378	1030	16.8	14.78
2-3 Mb	424	270	5.17	3.87
>3 Mb	267	153	3.26	2.19
Total	8202	6971	100	100

To explore the impact of ROH on Fujian Muscovy ducks, we identified a total of 26 ROH islands, distributed on chromosomes 1-7, 9, 11, 12, 15, 16, and 27 (Fig. 2C, Table S2). Following the identification of ROH islands, annotation was performed on 9,478 SNPs located within these islands, resulting in 434 annotated genes (Table S2). Enrichment analysis revealed that these genes were collectively enriched in 2,266 GO terms and 68 KEGG pathways (Figure S4, Table S2). Notably, terms and pathways related to muscle structure growth, myofibril assembly, and pigment deposition were identified. For instance, CREB1 and RBM24 were enriched in muscle tissue development (GO:0060537), KIF13A was enriched in pigmentation (GO:0043473), and PPP1R12A and ROCK2 were enriched in Regulation of actin cytoskeleton (apla04810). Moreover, FZD5, FZD7, and ADCY1 were enriched in Melanogenesis (apla04916).

Selective sweep analysis

The F_ST method was used to filter the obtained 947,742 SNPs, resulting in 33,433 SNPs. Out of these, 14,979 SNPs were located in the coding region, including 665 SNPs in exon and 14,314 SNPs in intron. Annotation identified 837 annotated genes (Fig. 3, Table S3), which were then subjected to enrichment analysis for GO Terms and KEGG Pathways. The analysis (Figure S5, Table S3) revealed that these genes were enriched in 3,742 GO terms and 127 KEGG pathways. Terms and pathways associated with muscle cell proliferation, striated muscle tissue development, myofibril differentiation, and pigment deposition were identified. Specifically, MEF2C, IGF2, ACTC1, and CREB1 were enriched in muscle tissue development (GO:0060537), PPP1R12A and ROCK2 were enriched in Regulation of actin cytoskeleton (apla04810), while FZD5, FZD7, EDNRB, EDNRB2, WNT3, WNT9B, and CREB1 were enriched in Melanogenesis (apla04916).

Fig. 3.

Whole genome of Fujian Muscovy duck scanned by F_ST method. Each point represents a divided window, the red dotted line is 0.25, which represents F_ST threshold, and above the dotted line is the filtered windows.

Analysis of muscle slices

Fig. 4A represents the pectoral muscle cross-section for the black group, while Fig. 4B is for the white group. Fig. 4(C, and 4D) and Table S4 present the histological section and pectoral muscle weight statistics of the two groups. The cross-sectional area of muscle fibers in the black group is significantly larger than that in the white group (P < 0.01), and the pectoral muscle weight of the black group is extremely significantly larger than the white group (P < 0.001). Combined with the results of WGRS analysis above, we speculate that the observed difference in gene expression between the black-feathered and white-feathered populations of Fujian Muscovy ducks may contribute to this phenomenon.

Fig. 4.

Transverse sections of the pectoral muscles in Fujian Muscovy ducks. a) Pectoral muscle section of the white group. b) Pectoral muscle section of the black group. c) Statistical graph of the cross-sectional area of muscle fiber cells. d) Pectoral muscle weight of 60 individuals in this study. All data are presented as means ± standard error (SEM).

Note: n = 5 per group for a), b), and c), n = 30 per group for d). Muscle cell area was performed by ImageJ software. ** means P < 0.01, and *** means P < 0.001.

RNA-Seq analysis

After grouping six samples, we conducted RNA-Seq analysis using the white group as a control. 30 DEGs were identified, and the differential degree is shown in Fig. 5 and Table S5. Subsequently, we performed GO and KEGG enrichment analysis on these DEGs. This yielded a total of 812 enriched GO terms and 49 enriched KEGG pathways (Figure S6, Table S5). Significant associations were found between pathways related to muscle growth and the enriched analysis part of selective sweep analysis. For instance, KLF15, TRIM63, and MYOC were enriched in striated muscle adaptation (GO:0014888), while XIRP2 and SIK1 were enriched in muscle tissue development (GO:0060537). Additionally, MYOC enrichment was observed in myosin light chain binding (GO:0032027), PDGFB was enriched in Regulation of actin cytoskeleton (ko04810). Subsequent analysis of transcriptional regulation identified 211 potential upstream transcriptional regulatory factors that could regulate these DEGs. Of these, 15 genes overlapped with the results from the selective sweep analysis, as shown in Fig. 6 and Table S5.

Fig. 5.

RNA-Seq analysis of six Fujian Muscovy ducks. a) The volcano plot displays differentially expressed genes with red indicating upregulated genes, blue indicating downregulated genes, and gray indicating genes with no significant trend. b) A heatmap illustrates inter-group gene expression patterns post-clustering, where red indicating genes with relatively high expression levels, and blue indicating genes with relatively low expression levels.

Note: n = 3 per group.

Fig. 6.

Differential gene and potential upstream regulatory factor network diagram. The outer circle represents the upstream regulatory factors overlapping with F_ST screening. The inner circle represents the DEGs that are potentially regulated by the upstream transcriptional regulatory factors. The DEGs that were validated by qRT-PCR are shown in dark green.

Note: The motifs and tracks database of transcription factor analysis was constructed by Stein Aerts Lab.

Verification of DEGs by quantitative real-time PCR

To assess the reproducibility of RNA-Seq results, we selected 10 genes potentially associated with muscle growth for quantitative Real-Time PCR (qRT-PCR) validation. The qRT-PCR results (Fig. 7, Table S6) showed that KLF15, MYOC, MYOZ2, NR4A3, SDC4, SIK1, TRIM63, and XIRP2 were up-regulated in White compared to Black, while METTL21C and PDGFB were down-regulated, consistent with the RNA-Seq results (Fig. 5B).

Fig. 7.

Verification of gene expression level via qRT-PCR. The heat map below the column chart was calculated using TPM values. The red represents high abundance, and the blue indicates a low expression level. In qRT-PCR, β-actin was used as a housekeeping gene. The data in the Y-axis represent the relative expression level. All data are presented as means ± standard error (SEM).

Note: n = 3 per group. ns means P > 0.05, * means P < 0.05, ** means P < 0.01, and *** means P < 0.001.

Discussion

Understanding the formative process of animal breeds is crucial in livestock production. It helps in defining breeding objectives, devising rational breeding plans, ensuring the inheritance of superior genes during breeding, and can also aid in preventing and controlling potential genetic diseases (van Marle-Koster and Visser, 2018; Stock et al., 2020). Previous research on Muscovy ducks at the genomic level has primarily focused on exploring genes related to production efficiency (Bai et al., 2020; Xu et al., 2022). However, there is still a significant gap in the study of the genetic structure of local breeds. In this study, we analyzed the genetic structure of the black and white feather strains of Fujian Muscovy duck through WGRS analysis. This analysis contributes to a better understanding of the breeding process of Fujian Muscovy duck. During the ADMIXTURE analysis, the lineages of the two feather strains were distinctly separated when tracing back one generation. However, as the tracing continued, the lineage of the black feather ducks appeared to be more complex than that of the white feather ducks. According to the neighbor-joining (NJ) Tree, the white feather ducks occupied a single branch, while the black feather ducks were divided into two branches, possibly due to short-term artificial breeding (Stock et al., 2020). The PCA results indicate that the genetic similarity within the white feather duck population is higher than that of the black. This finding supports the accuracy of the NJ Tree. To the best of our knowledge, relatively recently did the selective breeding of black feather muscovy ducks begin in Fujian. Therefore, the population structure analysis indicates that the genomic similarity of the black feather strain of Fujian Muscovy ducks is lower than that of the white, which is consistent with our expectations.

ROH typically arises from the transmission of identical haplotypes from parents to offspring, providing information about population history and evolution. The detection of ROH allows for the estimation of genetic history and inbreeding levels of populations, which can aid in minimizing the effects of inbreeding (Gao et al., 2023). The length of ROH is positively correlated with the timing of inbreeding events (Peripolli et al., 2017). The present study found that the proportion of long ROH (>1 Mb) was higher in the black group than in the white group, suggesting that inbreeding events have had a greater impact on the black population. This could be attributed to the superior meat performance of black-feathered muscovy ducks compared to white-feathered ones, resulting in greater artificial selection pressure on the black group. This trend has also been observed in commercially produced chickens (Talebi et al., 2020). Furthermore, the F_ROH of the black group was higher than that of the white group, providing additional evidence to support the aforementioned conclusion. This finding is consistent with trends observed in previous studies on other animals (Forutan et al., 2018; Gao et al., 2023).

ROH islands represent genomic regions that may have undergone natural or artificial selection (Martin et al., 2023). In this study, 26 ROH islands were identified, consisting of genomic regions with a high concentration of SNPs. Our enrichment analysis revealed that genes located on ROH islands were predominantly involved in pathways related to muscle structure formation, muscle fiber composition, and myofibril binding. Some of these genes have been validated or proposed as candidate genes associated with muscle growth in other animals, including CREB1, MYF6, and RBM24. In studies on pigs and cattle, CREB1 has been identified as a regulator of myoblast proliferation and differentiation, and intramuscular fat content (Feng et al., 2022; Sun et al., 2022). MYF6 has been recognized as a regulator of skeletal muscle growth in chickens and ducks (Zhu et al., 2014; Li et al., 2022). RBM24, as an RNA-regulation motif, has been found in ducks to potentially participate in the differentiation of duck myoblasts and the development of skeletal muscle, showing a similar expression pattern to MEF2C and other myogenic regulatory factors (Sun et al., 2016). Additionally, genes located on the islands, such as KIF13A and FZD5, are enriched in pathways related to pigment deposition. KIF13A has been suggested as one of the factors inhibiting melanin production and is associated with pigment deposition in pigs (Delevoye et al., 2009; Du et al., 2022). FZD5 acts as a receptor for the WNT5A ligand, which has been confirmed to inhibit melanin production (Zhang et al., 2013). These genes may be involved in plumage differentiation in Fujian Muscovy ducks.

In livestock and poultry research, selective sweep analysis can efficiently and cost-effectively scan low polymorphic regions on the genome. This makes systematic study of selection markers possible, facilitating the rapid exploration of traits related to economic characteristics (Panigrahi et al., 2023). Selective sweep analysis identified candidate genes for muscle growth, including MEF2C, IGF2, ACTC1, and NFATC1, previously validated in other animals. Pectoral muscle samples were collected from six representative Fujian Muscovy ducks for RNA-Seq, revealing multiple genes related to muscle growth, such as METTL21C, MYOZ2, MYOC, PDGFB, TRIM63, SDC4, and XIRP2. The study explored the transcriptional regulation of factors such as MEF2C, NFATC1, and CREB1, identified in the selective sweep analysis as potential regulators of muscle growth genes. The obtained differentially expressed genes were analyzed to demonstrate the feasibility of integrating WGRS and RNA-Seq to identify candidate genes. Furthermore, melanogenesis (apla04916) is enriched with EDNRB, EDNRB2, WNT3, WNT9B, etc. Among these genes, EDNRB (Li et al., 2015), EDNRB2 (Nannan et al., 2024), and WNT Family (Yang et al., 2018) have been shown to be associated with plumage differentiation in birds. This indirectly confirms the accuracy of our WGRS results.

The investigation of duck MEF2C has established its crucial role in muscle development (Sun et al., 2013). Based on RNA-Seq analysis of Beijing ducks (Anas platyrhynchos), IGF2 was found to be closely linked to the regulation of muscle development (Ye et al., 2017; Hu et al., 2021). However, studies of IGF2 on muscovy ducks have primarily explored its association with egg production traits, with less research yet identifying its correlation with muscle growth traits. ACTC1 has been linked to muscle fibre hypertrophy in chickens (Li et al., 2020). In pigs, expression of ACTC1 is regulated by lincRNA, which facilitates intramuscular fat deposition (Dou et al., 2023). Research into porcine skeletal muscle satellite cells has shown that increased expression of NFATC1 effectively promotes its differentiation (Chen et al., 2017).

To validate the results of WGRS, we conducted RNA-Seq analysis between black and white feather Fujian Muscovy ducks and performed transcriptional regulation analysis of differentially expressed genes. In our study, MEF2C and CREB1 were not only identified as regulators of muscle growth but also predicted as potential regulatory factors of DEGs. Among these, MYOZ2 (Shum et al., 2012), MYOC (Judge et al., 2020), and XIRP2 (Hawke et al., 2007) are considered potential target genes of MEF2C. SDC4 is regarded as a muscle cell marker expressed in different states along with MEF2C, a phenomenon also observed in porcine muscle satellite cells (Jacob and David, 2020; Yi et al., 2024). MYOZ2 has been associated with inhibition of muscle growth in studies across various poultry, including ducks (Liu et al., 2018; Wei et al., 2022; Zhang et al., 2024). In feeding studies on ducks, TRIM63 has been suggested to increase the rate of protein degradation, thereby inhibiting muscle growth (Xia et al., 2022). Additionally, MYOC (Sharma and Grover, 2021) and XIRP2 (Scheffer et al., 2015) are associated with regulating the function of the muscle actin cytoskeleton, and XIRP2 has also been found to be associated with muscle actin repair (Wagner et al., 2023).

Furthermore, it was observed that the results of the transcriptional regulation analysis did not include METTL21C, which encodes a protein methyltransferase with protein-lysine N-methyltransferase activity, participating in protein modification by mainly mediating lysine dimethylation and trimethylation (Kernstock et al., 2012; Ke et al., 2023). Research has shown that overexpression of METTL21C in human muscle cells can inhibit protein degradation during muscle hypertrophy (Steinert et al., 2023). In genomic studies of cattle, sheep, and chickens, METTL21C has been identified as a key gene related to meat yield (Yang et al., 2019; Demir et al., 2023; Kong et al., 2023). In chickens, METTL21C eliminates the growth inhibitory effect of IGF2BP1 on chicken myoblasts by mediating lysine trimethylation modification of IGF2BP1 (Wang et al., 2023). Previous studies have found that METTL21C methylates VSP and HSPA8 (Zoabi et al., 2020; Chen et al., 2023), where VSP maintains muscle cell protein homeostasis (Kustermann et al., 2018), and HSPA8 inhibits necrotic apoptosis (Wu et al., 2023). The study found that genes MYOZ2 and TRIM63 had higher expression levels in white ducks compared to black ducks, while METTL21C was higher in black ducks than in white ducks, suggesting a potential role of protein methylation, which requires further experimental validation.

Conclusions

This study conducted WGRS on Fujian Muscovy ducks and performed RNA-Seq on breast muscle with different feather colors, comprehensively analyzing their population genetic structure and genomic differences. The results revealed significant genetic structural disparities between black-feathered and white-feathered Fujian Muscovy duck strains. MEF2C, MYOZ2, and METTL21C were identified as key genes associated with meat yield regulation in Fujian Muscovy ducks, with their regulatory mechanisms potentially linked to protein methylation. Overall, these findings enhance our understanding of the population structure and shed perspective on the genetic determinants of muscle growth in Fujian Muscovy ducks, offering valuable insights for further research into the molecular mechanisms governing muscle development in this population.

Data availability

The Fujian muscovy duck genome sequences have been deposited with links to BioProject accession number PRJNA1049164 in the NCBI BioProject database. The transcriptome data for RNA-seq is available with links to BioProject accession number PRJNA1049396.

Consent to publication

All authors have read the manuscript and approved for publication.

Author contributions

Ruiyi Lin contributed to the experimental design. Huihuang Li, Xinguo Bao, and Weimin Lin performed the data analysis of genome. Xiaobing Jiang, Jialing Qiu, and Lianjie Lai collected samples. Ruiyi Lin, Xiaobing Jiang, and Huihuang Li completed the manuscript writing. Fan Yang, Chengfu Pan and Weilong Lin participated in the writing instruction and revision of the manuscript. All authors have read and approved the manuscript.

Funding

The work was supported by National Natural Science Foundation of China (31702109), Special Fund for Science and Technology Innovation of Fujian Agriculture and Forestry University (KFb22064XA), Natural Science Foundation of Fujian Province (2017J01596 and 2023J01446), Fujian Province Young and Middle-Aged Teacher Education Research Project (JAT220059).

Consent for publication

Our manuscript contains no individual person's data in any form.

Ethics approval

All experiments and methods procedures concerning the muscovy ducks used in this study were performed in accordance with the ARRIVE guidelines. Ethical approval was granted by the Experimental Animal Care and Use Committee of Fujian Agriculture and Forestry University (FAFU2023-0012) according to the Regulations for the Administration of Affairs Concerning Experimental Animals (Ministry of Science and Technology, China, revised in July 2013).

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relation-ships that could be construed as a potential conflict of interest.