Genomic and transcriptomic analyses revealed potential candidate genes for egg production traits in domesticated ducks

Abstract

The process of domestication in ducks has led to distinct morphological, behavioral, and physiological traits that differentiate from their wild ancestors. Among these traits, reproductive performance, particularly egg production, has been specifically enhanced. In this study, in order to understand the molecular mechanisms underlying the great differences of egg production between domesticated ducks and their wild ancestors, we integrated genomic and transcriptomic analyses to identify candidate genes and pathways associated with egg production. Selective sweep analyses based on Di revealed genomic regions under selection in multiple egg-type duck breeds. A total of 987 candidate genes were identified, with 11 genes shared across breeds, suggesting common selection pressures. Transcriptomic analyses between egg-type ducks and mallard revealed differentially expressed genes involved in hormone regulation, lipid transport, and follicular development in the liver and gonads. Multiple candidate genes (such as FN1, PDE10A, SEMA5B, UBE2E3) and enriched pathways (such as ECM-receptor interaction, Wnt signaling, steroid hormone biosynthesis) were found to be potentially linked to egg production traits. Notably, only FN1 and SEMA5B were identified by both genomic and transcriptomic analyses. Our study provide additional insights into the genetic basis of egg production in ducks.

Introduction

It is well known that egg production is an important economic trait in poultry breeding. Through long-term domestication and artificial selection, domesticated ducks have undergone great improvement in reproduction performance, particularly in egg production, compared to their wild ancestors (Bao et al., 2020; Duggan et al., 2015; Miller, 1977). However, egg production is a complex polygenic trait (Bello et al., 2022), and the genetic basis of these improvements is still not fully understood. Therefore, it is important to understand the genetic mechanism underlying these differences in egg production between domesticated ducks and their wild ancestors.

Comparative genomic and population genomic analyses of domesticated animals and their wild ancestors provide evidence for selection signatures and candidate genes underlying domestication traits (Wang et al., 2014). Rubin et al. (2010) identified a shared selective sweep among domesticated chicken populations with similar traits, involving TSHR gene, which is likely associated with the loss of seasonal reproduction during domestication. In ducks, recent studies have shown that comparisons between the genomes of egg-type and meat-type ducks and that of mallards have revealed distinct genomic selection during domestication (Gu et al., 2020; Zhang et al., 2018; Zhou et al., 2018).

With the development of high-throughput sequencing technology and the reduction of cost, sequencing technology has been widely used in animal genetic breeding research. Genome sequencing and transcriptome sequencing have been used to analyze candidate genes of egg production in poultry (Bello et al., 2021; Huang et al., 2024; Ouyang et al., 2025; Sun et al., 2021; Tao et al., 2017; Zhao et al., 2019). This study aims to use genomic and transcriptomic data to identify candidate genes associated with duck egg production traits. These results will contribute to a deeper understanding of the potential molecular mechanisms affecting egg production.

Materials and methods

Sample and genomic sequencing

A total of 212 duck samples were used in this study, including 28 breeds. The duck samples were obtained by our lab. We also downloaded 82 sets of sequencing data from NCBI. The detailed information can be shown in Table S1. To investigate selection signals related to egg production, we focused on five high-laying duck breeds as representative populations, based on their performance records in the Chinese Poultry Genetic Resources. These included Jinding ducks (n = 8), Shaoxing ducks (n = 8), Shanma ducks (n = 8), Putian black ducks (n = 8), and Wendeng black ducks (n = 8). Di was then calculated using these breeds to identify putative regions under selection.

Genomic variant calling and auality control

The raw reads were processed using fastp (v0.20.0) (Chen et al., 2018) with default parameters. The filtered reads were aligned with the duck reference genome (cau_duck1.0) using Burrows-Wheeler Aligner (BWA) (v0.7.15) (Li and Durbin, 2009). Duplicate reads were removed and aligned data were sorted using PICARD (http://broadinstitute.github.io/picard/). Genome Analysis Toolkit (GATK v4.1.8.1) (McKenna et al., 2010) was used to perform local realignment and base quality recalibration of reads by the "FixMateInformation", "BaseRecalibrator", and "ApplyBQSR" parameters. SNPs were called using the “Haplotypecaller” parameter. Variants were filtered using the ‘VariantFiltration’ parameter with the following standards: " QUAL < 30.0 ||QD < 5.0 || FS > 60.0 || MQ < 40.0 || MQRankSum < −12.5 || ReadPosRankSum < −8.0″. A total of 17,622,141 variants were retained for the following analysis.

Genome-wide selection signature analysis

Di introduced by Akey was calculated to detect selective sweeps using a sliding-window approach (10-kb windows with 5-kb steps) (Akey et al., 2010). For each population, Di was calculated by comparing allele frequencies between the target population and all other populations combined as a reference background. Windows within the top 1% of Di (empirical 99th percentile) were considered candidate selective sweep regions. The genomic regions consistently detected across all five high-laying breeds were regarded as candidate selection signatures associated with egg production. Candidate regions were annotated using BioMart in Ensembl (http://www.ensembl.org/biomart/martview/).

Transcriptomic analysis

Fertilized eggs were collected from Jinding ducks (JD) and mallards (MD) and incubated under standard conditions. At embryonic day 25, liver and left gonad tissues were collected from female embryos of JD (high egg-producing domestic breed) and MD (control group with lower egg-laying performance). For each tissue, four to five biological replicates were included per group. All samples were stored in RNAlater (Qiagen) until processing. Total RNA was extracted using TRIzol reagent (Ambion, Austin, TX). Libraries were sequenced with paired-end 150 bp reads on the Illumina HiSeq 4000 platform. The transcriptomic data have been deposited in NCBI under accession number PRJNA1419553.

The raw data were filtered to remove reads aligned to the barcode adapter using fastp (v0.20.0). The cleaned data was mapped to the genome (CAU_duck1.0) using HISAT2 (v 2.1.1) (Kim et al., 2019). The reads were assembled the gene expression levels were estimated using StringTie (v.2.1.4) (Pertea et al., 2015). The reads count files were merged with tissue, and expression levels were performed using DESeq2 (Love et al., 2014). Genes that met the absolute value of expression level change greater than 1 and adjusted P less than 0.05 were considered as differentially expressed genes. The related functions of DEGs were analyzed using KOBAS3.0 (Bu et al., 2021).

Integration of genomic and transcriptomic data

Considering the different biological roles of liver and gonad tissues in egg production, differential expression analyses were conducted separately for the two tissues. DEGs from both tissues were merged to generate a unified DEG set representing transcriptional changes in at least one key reproductive organ. This merged DEG set was subsequently intersected with genes located within Di-based selective sweep regions to identify overlapping genes.

Results

Selection signature analysis

Selection analysis was performed based on Di. With the threshold set to the top 1%, a total of 7,112 windows were identified, within which 924 genes were annotated. The top 1% windows within each breed were considered putative signatures of selection. A total of 1,618 windows with high di values were identified to be shared by 2 or more breeds. Specifically, 1,089 were shared between 2 groups, 316 were shared among 3 groups, and 173 was shared by 4 groups (Fig. 1 and Table S2). There were 40 Di windows shared by all five breeds, encompassing 11 candidate genes, including TBXAS1, PPIL4, ARID1B, SLC25A21, UBE2E3, SLC49A4, SEMA5B, SLC4A10, IFIH1, ZNF385B, FN1 (Fig. 2). The shared selection regions among breeds likely contain genes that affected egg production. Sliding-window analyses of pairwise FST in this gene region exhibit differentiation in high-egg-production populations relative to mallards, suggesting potential selection (Figure S1).

Fig. 1.

Genome-wide Di for the five breeds.

Fig. 2.

Venn diagrams illustrate the shared genes in the five breeds.

Transcriptome analysis reveals differentially expressed genes associated with egg development

Left gonad and liver samples from embryonic were collected for transcriptome sequencing and analysis of the functional factors involved in egg production. After quality control, 1.71 billion clean reads were obtained from all samples, and an average of 42,751,095 clean reads per sample was obtained. Using a fold change threshold of ≥1 and a p-value < 0.05, 990 DEGs were identified between JD and MD left gonad, with 675 down regulated and 315 up regulated (Fig. 3). Pathway enrichment analysis identified the pathways significantly enriched by the DEGs. KEGG identified 5 pathways, such as MAPK signaling pathway, Neuroactive ligand-receptor interaction, Calcium signaling pathway, Regulation of actin cytoskeleton, ErbB signaling pathway.

Fig. 3.

Transcriptome analysis of JD and MD. a, The volcano plot of DEGs between JD and MD in ovary; b, KEGG enrichment analysis of DEGs in ovary; c, The volcano plot of DEGs between JD and MD in liver; b, KEGG enrichment analysis of DEGs in liver.

In the transcriptome analysis of liver, 858 DEGs were identified between JD and MD liver, with 316 down regulated and 542 up regulated (Fig. 3). Pathway enrichment analysis identified the pathways significantly enriched by the DEGs. KEGG identified a total of 34 pathways, such as ECM-receptor interaction, Neuroactive ligand-receptor interaction, Focal adhesion, Notch signaling pathway, AGE-RAGE signaling pathway in diabetic complications, Melanogenesis, Butanoate metabolism, Wnt signaling pathway, Fatty acid biosynthesis, Steroid hormone biosynthesis, Intestinal immune network for IgA production, Ferroptosis, Insulin signaling pathway, Fatty acid metabolism, Hedgehog signaling pathway.

Candidate gene analysis

To further study the possible regulatory mechanisms of these candidate genes in ducks, an integrated analysis was conducted of transcriptome and genomic analysis to identify overlapped genes. No overlapping DEGs were detected between liver and gonad tissues. Therefore, DEGs from both tissues were merged to generate a unified DEG set, representing genes showing expression changes in at least one key organ related to egg production. This merged DEG set was subsequently intersected with genes located in Di-based selective sweep regions to identify overlapped genes. There were 18 genes were identified, which shown different expression patterns and high Di values in 2 or more breeds (Table 1). And two genes (FN1 and SEMA5B) were identified in all five high-laying breeds. These genes were considered candidate genes potentially associated with egg production traits in ducks.

Table 1.

Overlapped genes identified by di statistic and transcriptomic analyses.

Tissue	Gene	Expressions status	Column
left gonad	EGFR	up	pt|jd
left gonad	ENSAPLG00000016422	up	wd|sx
left gonad	TRAF3	up	wd|sx
left gonad	ENSAPLG00000022017	down	wd|sm
left gonad	ENSAPLG00000017302	up	wd|jd
left gonad	STRN	up	sx|sm
left gonad	MREG	down	sx|sm
left gonad	ERBB4	up	pt|sx|jd
left gonad	SEMA5B	up	pt|wd|sx|sm|jd
liver	GALNTL6	down	pt|sx
liver	ENSAPLG00000014415	down	wd|sx
liver	MLPH	up	wd|sx
liver	ADAMTS20	up	sx|sm
liver	SMYD3	down	sx|jd
liver	BRSK2	up	sx|jd
liver	TRPM8	down	wd|sx|sm
liver	PDE10A	up	sx|sm|jd
liver	FN1	up	pt|wd|sx|sm|jd

Discussion

Reproductive performance is an important economic characteristic in poultry breeding (Bao et al., 2020). The laying performances of domesticated duck breeds differ considerably from those of mallards, providing a large space for researching and utilization. As a key reproductive organ in poultry, the ovary plays a crucial role in folliculogenesis and ovulation through its development and physiological activity (Xiang et al., 2023). As the primary organ for energy and fatty acid metabolism (Pearce, 1971), the liver plays a crucial role during egg production in poultry by synthesizing yolk proteins and lipids and transporting them to the ovary (Johnson A.L., 2015). To explore the genetic basis and early regulatory mechanisms underlying the differences in egg production between domestic ducks and mallards, we integrated genomic data with transcriptomic data of the left gonad and liver at embryonic day 25. This study aimed to identify candidate genes and pathways potentially associated with egg production traits in domestic ducks.

Identifying genomic selection regions that are shared by breeds can help pinpoint selection signatures most likely associated with egg production traits. Based on Di, a total of 987 genes were identified, among which 11 were shared across all five breeds (TBXAS1, PPIL4, ARID1B, SLC25A21, UBE2E3, SLC49A4, SEMA5B, SLC4A10, IFIH1, ZNF385B, FN1). Deng et al. (2023) reported that miR-143-3p directly binds to the 3′ untranslated region (UTR) of UBE2E3, resulting in the downregulation of UBE2E3 expression and promoting ovarian granulosa cell senescence. ARID1B is the subunit of the SWI/SNF chromatin complex (Sato et al., 2018). PPIL4 is essential for brain angiogenesis (Barak et al., 2021). Other genes such as IFIH1 and SLC4A10 are primarily known for immune (Najm et al., 2024) or metabolic functions (Parker et al., 2008).

Through RNA-Seq analysis of left gonad and liver tissues, some important DEGs related to egg production in ducks were identified. These genes were found to be enriched in pathways such as ECM-receptor interaction, Focal adhesion, Neuroactive ligand-receptor interaction, Calcium signaling pathway, MAPK signaling pathway, ErbB signaling pathway, Steroid hormone biosynthesis, Wnt signaling pathway. Several studys have reported the function in egg production of these signaling pathway (Ouyang et al., 2020; Bello et al., 2021; He et al., 2023).

Integration of Di and transcriptomic analyses in female ducks identified two crucial genes: FN1 and SEMA5B. FN1 plays an important role in the support of cell adhesion (Goossens et al., 2009). FN1 involved in ECM-receptor interaction (Nersisyan et al., 2021). In hens, the follicular wall of the ovary is mainly composed of extracellular matrix (ECM), which regulates cellular proliferation, differentiation, and functional activity (Huang et al., 2022, 2024). The ECM-receptor interaction pathway may mediate communication between granulosa cells and the follicular wall, affecting the development of follicles (Huang et al., 2024). Recent studies have identified the ECM-receptor interaction pathway as significantly enriched in ovarian transcriptomic analyses, indicating its involvement in reproductive performance (Bello et al., 2021; Chang et al., 2024; He et al., 2023; Huang et al., 2024). FN1 has been reported as a candidate gene involved in egg production based on hypothalamus and ovary transcriptome analyses in Muscovy ducks (Bello et al., 2021). This gene has also been implicated in early development in bovine preimplantation embryos, particularly in embryo compaction and blastocyst formation (Goossenset al., 2009).

Although direct evidence linking SEMA5B to egg production is limited, a recent study on the genomic characteristics and selective signals of Zhongshan duck suggested that SEMA5B may play a role in growth and development (Chen et al., 2023), indicating its broader involvement in economically important traits. Moreover, SEMA5B has also been identified as a candidate gene for cow fertility (Cai et al., 2019), further supporting its potential role in reproductive processes.

By joint analysis of genomic and transcriptomic analyses, FN1 and SEMA5B were identified as potential candidate genes for egg production in ducks. Several signaling pathways were significantly enriched and may play crucial roles in regulating egg production traits, such as ECM-receptor interaction, Focal adhesion, Neuroactive ligand-receptor interaction, Calcium signaling pathway, MAPK signaling pathway, ErbB signaling pathway, Steroid hormone biosynthesis, Wnt signaling pathway. These results provide theoretical data reference for revealing the genetic mechanism of egg production in ducks.

Ethics statement

The animal experimentation protocols were approved by the Animal Welfare Committee of China Agricultural University (permit XK622, Permit 046/2018), and the study was conducted in accordance with ethical guidelines for animal research.

CRediT authorship contribution statement

Xinye Zhang: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Conceptualization. Xiurong Zhao: Writing – review & editing, Software, Methodology, Data curation. Hongchang Gu: Software, Formal analysis, Data curation, Conceptualization. Tao Zhu: Software, Investigation, Formal analysis. Hongyan Sun: Software, Investigation. Yalan Zhang: Software, Methodology, Investigation. Gang Wang: Software, Resources, Investigation. Xufang Ren: Visualization, Investigation, Formal analysis. Anqi Chen: Visualization, Software, Resources. Xiaoyu Jiang: Visualization, Software. Lujiang Qu: Writing – review & editing, Methodology, Funding acquisition.

Disclosures

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.