https://orcid.org/0000-0001-7509-5030
Figures
Abstract
Litter size is one of the most important economic traits in sheep. Identification of gene variants that are associated with the prolificacy rate is an important step in breeding program success and profitability of the farm. So, to identify genetic mechanisms underlying the variation in litter size in Iranian Baluchi sheep, a two-step genome-wide association study (GWAS) was performed. GWAS was conducted using genotype data from 91 Baluchi sheep. Estimated breeding values (EBVs) for litter size calculated for 3848 ewes and then used as the response variable. Besides, a pathway analysis using GO and KEGG databases were applied as a complementary approach. A total of three single nucleotide polymorphisms (SNPs) associated with litter size were identified, one each on OAR2, OAR10, and OAR25. The SNP on OAR2 is located within a novel putative candidate gene, Neurotrophic receptor tyrosine kinase 2. This gene product works as a receptor which is essential for follicular assembly, early follicular growth, and oocyte survival. The SNP on OAR25 is located within RAB4A which is involved in blood vessel formation and proliferation through angiogenesis. The SNP on OAR10 was not associated with any gene in the 1Mb span. Moreover, gene-set analysis using the KEGG database identified several pathways, such as Ovarian steroidogenesis, Steroid hormone biosynthesis, Calcium signaling pathway, and Chemokine signaling. Also, pathway analysis using the GO database revealed several functional terms, such as cellular carbohydrate metabolic, biological adhesion, cell adhesion, cell junction, and cell-cell adherens junction, among others. This is the first study that reports the NTRK2 gene affecting litter size in sheep and our study of this gene functions showed that this gene could be a good candidate for further analysis.
Citation: Esmaeili-Fard SM, Gholizadeh M, Hafezian SH, Abdollahi-Arpanahi R (2021) Genome-wide association study and pathway analysis identify NTRK2 as a novel candidate gene for litter size in sheep. PLoS ONE 16(1): e0244408. https://doi.org/10.1371/journal.pone.0244408
Editor: Rajakumar Anbazhagan, National Institute of Child Health and Human Development (NICHD), NIH, UNITED STATES
Received: August 12, 2020; Accepted: December 8, 2020; Published: January 22, 2021
Copyright: © 2021 Esmaeili-Fard et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The datasets supporting the conclusions of the manuscript are available in the Figshare repository under the following DOI number: https://doi.org/10.6084/m9.figshare.12515189.v1
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Flock profitability is greatly affected by ewe reproductive efficiency. The outcome of reproductive traits in sheep is quite different from one breed to another and also within breeds. Identification of ewes with higher prolificacy rates is an important step in sheep breeding success and profitability of the farm [1]. In sheep, understanding the genetic basis for high prolificacy is valuable for genetic improvement and management and provides basic knowledge that may facilitate modification of fertility and prolificacy in the future [2]. The heritability of the reproductive traits varies from low to moderate and, therefore, response to selection under classic phenotypic breeding programs is expected to be low. As an alternative, identifying major gene variants that affect different parts of the reproductive process, such as ovulation rate and litter size could be very helpful and used in marker-assisted selection (MAS) context [3]. Characteristics of a major gene affecting prolificacy in a population include high variation in ovulation rate and litter size, combined with high repeatability. Reproduction is a complex process, and traits such as ovulation rate and litter size are genetically affected by many minor genes and also some major genes, called fecundity (Fec) genes [4]. To date, a number of this fecundity genes have been identified in sheep, including GDF9 [5], BMP15 [6], BMPR1B [7], B4GALNT2 [8], FecX2W [9], and FecD [10] located on ovine chromosomes 5, X, 6, 11, X, and an unknown autosome, respectively. So far, several studies using different methods have been performed to identify gene variants affecting litter size in Iranian sheep breeds, but nearly none of the effective variants of these known genes have been identified yet [1, 11–13]. However as some of the identified variants might be population-specific, further scanning of the genome is needed to identify additional strong candidate genes in the population of interest.
Baluchi sheep are one of the Iranian fat-tailed breeds and comprise approximately 30% of the country’s total sheep inventory. This breed is well-adapted to the arid subtropical environments of eastern Iran and can play an important role in supplying the need for meat due to its higher twinning rate than other Iranian breeds and acceptable growth rates [14, 15]. To date, only one genome-wide association study on the litter size in Baluchi sheep was conducted. Gholizadeh et al. [1] using a GLM model and birth type records as the response variable, performed a whole-genome scan for each parity separately and identified two significant single nucleotide polymorphisms (SNPs), rs407696726 and rs412433416, on OAR10 and OAR15. The SNP on OAR15 is located within the DYNC2H1 gene, but SNP on OAR10 was not associated with any gene in a 1 Mb span. However, in the present study, we used a two-step GWAS accompanied by a gene-set analysis (GSA) as a complementary approach to identify genes and pathways affecting litter size in Iranian Baluchi sheep. Here, the estimated breeding values (EBVs) were used as the response variable.
Complex traits are controlled by several genes and therefore, association studies identify only the most significant SNPs (due to using stringent threshold) and their neighboring genes that signify only a small portion of the genetic variation [16]. So, several poorly associated SNPs are always ignored and the large majority of the genetic variants contributing to the trait remains hidden [17, 18]. An alternative strategy to tackle the aforesaid problem is to move up the analysis from the SNP to the gene and gene-set levels [19]. gene set or pathway-based analysis has been proposed as a complementary approach to investigate complex traits from a genetic and biological perspective and work based on evaluating modules of functionally related genes, rather than focusing only on the most significant markers [16, 20]. In this approach, a set of related genes, such as genes in a specific pathway, that harbor significant SNPs previously detected in GWAS (although with a less stringent threshold), is tested for over-representation in a specific pathway [21].
Failure to identify known genes affecting litter size in Baluchi breeds as well as other Iranian sheep breeds reinforces the possibility that another mechanism may be involved in the litter size variation in Iranian sheep. So, to unravel the genomic architecture underlying the litter size in Baluchi sheep, a GWA analysis using EBVs as a response variable was performed. Gene-set analysis was used as a complementary approach for GWAS.
Materials and methods
Phenotypic and genotypic data
All the required information for this study was provided by the Department of Animal Science of Sari Agriculture Science and Natural Resource University (SANRU), Iran [1]. The phenotype dataset consisted of 3,848 birth records from 1,506 ewes that were collected from 2004 to 2012 at Abbas Abad Baluchi sheep Breeding Station, Iran. Descriptive statistics of the litter size trait is presented in Table 1. The pedigree file encompassed 4,727 animals with 178 sires, 1,509 dams, and 818 founders.
A total of 91 of 1,506 ewes were genotyped by Illumina 50K SNP panel for 54,241 markers. A similar number of single- and twin-bearing ewes with minimum pedigree relationships to one another were selected for genotyping. These ewes had a total of 435 repeated records across six years.
Quality control of SNP genotypes was performed in GenABEL [22] R [23] package. The SNPs which had minor allele frequency (MAF) less than 0.01, genotype call rate less than 93%, Hardy–Weinberg equilibrium deviations less than 10−6 p-value cutoff, and SNPs without known genomic location were removed. Also, ewes with a genotyping call rate of less than 95%, were excluded from the dataset.
Genome-wide association study
For GWA analysis, a two-step approach was used. In the first step, the estimated breeding values of all ewes for the litter size trait were calculated using the repeatability BLUP (Best linear unbiased prediction) model. To do this, a primary GLM model was run in R software to identify significant fixed effects on litter size. Then the following model was used for EBV calculation:
where, y is a vector of birth type (1 and 2); b is a vector of fixed effects including dam age (2–6 years), birth year (2006–2012), and parities (1–5); u is the vector of random additive direct genetic effects, pe is the vector of permanent environmental effects, and e is the vector of random residual effects. The matrices X, Z, and W are the incidence matrices relating animal records to fixed and random effects, respectively. In this model, the random effects have multivariate Gaussian (co)variance,
Where A is the pedigree-based relationship matrix; I is the identity matrix; n is the number of ewes in the dataset (n = 1,506) and N is the total number of records (N = 3,848). AIREMLF90 and BLUPF90 were used for variance components estimation and EBVs calculation, respectively [24, 25]. In the second step, EBVs as the response variable and SNP genotypes as the fixed effects fitted in a GLM model as follows,
where XSNP is the incidence matrix relating EBVs to SNP genotypes and βSNP is the regression coefficient. The analysis was performed using the GenABEL [22] package in the R environment. In this step to account for population stratification, a principal component analysis (PCA) was performed and the first six principal components (based on cumulative variance) were incorporated in the model as covariates. However, after using PCs as covariates, partial inflation was corrected using the genomic control (GC) method, and all p-values were presented without any inflation (λ = 1). For multiple testing corrections, the simpleM [26] approach was used. This method works based on the effective number of independent markers (tests). Based on the average number of independent tests on each chromosome and the 0.05 p-value cutoff, we determined a chromosome-wide threshold (0.05/495 = 1.01*10−4). Briefly, The SimpleM first computes the eigenvalues from the pair-wise SNP correlation matrix created with composite LD from the SNP dataset and then infers the effective number of independent tests (Meff_G) through principle component analysis. Once Meff_G is estimated, a standard Bonferroni correction is applied to control for the multiple testing. The Manhattan plot was drawn by CMplot (https://github.com/YinLiLin/R-CMplot) R package.
Two well-known databases including BioMart-Ensembl (www.ensembl.org/biomart) and UCSC Genome Browser (http://genome.ucsc.edu) were used along with the Ovis aries reference genome assembly (Oar_v3.1) to identify genes.
Genotypes mean comparison
To investigate the genotype effect of significant SNPs on litter size, differences between each genotype means were analyzed by Tukey’s multiple pairwise-comparison test using R. The p-values less than 0.05 (p < 0.05) were considered statistically significant. The ggplot2 [27] R package was used to visualize the results.
Gene-set analysis
Gene-set analysis is usually done in 3 steps [17]. In the first step, an arbitrary threshold of p-value ≤ 0.05 was used to determine significant SNPs [17, 28]. The biomaRt [29, 30] R package and the Oar_v3.1 ovine reference genome assembly were used to assign SNPs to genes if they were located within the genes or 15 kb upstream or downstream of a gene. The size of the flanking region was based on the finding that most SNPs that affect the expression of genes are located within 15 kb of the gene [31]. Genes with at least one significant SNP were considered as significantly associated genes. In the second step, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases were used to define the functional gene-sets. And finally, the significant association of a particular gene-set with the litter size was calculated using Fisher’s exact test based on the hypergeometric test [17]. The p-value of g significant genes in the gene-set was computed by,
where s is the total number of significant genes associated with the litter size, m is the total number of analyzed genes, and k is the total number of genes in the gene-set under consideration [17]. Both GO and KEGG enrichment analyses were performed using the R package clusterProfiler [32]. Gene-sets with more than 5 and less than 500 genes were tested. Categories with a p-value less than or equal to 0.05 (P ≤ 0.05) were considered significant sets. The ggplot2 [27] R package was used to visualize the GO and KEGG analysis results as bar plots.
Results
Estimates of genetic parameters
Estimates of variance components, heritability (h2), repeatability (r), and range of estimated breeding values in genotyped ewes are shown in Table 2. The estimate of heritability of the litter size trait was 0.08 which was in the range reported by previous authors [33–35] and as expected, it was low.
Genome-wide association analysis
In this study, a two-step approach was applied to perform a GWA analysis to identify gene variants affecting litter size in Baluchi sheep. We used EBVs as the response variable for GWAS which can increase the power to some extent as we have a better estimate of the actual genetic variance. EBVs can largely compensate for the limited number of genotypes to get reasonable estimates. After quality control, 84 ewes and 45342 SNPs were retained for the GWA analysis. We applied the simpleM method for multiple testing correction. This approach can take into account the dependence among the SNPs that are in linkage disequilibrium (LD) and consequently a less stringent threshold can be applied. Although false discovery tests are important for controlling false positives, if there are numerous SNPs in strong LD, they could cause substantial loss of statistical power and increase risks of missing true associations (false negatives) [36]. In addition, we used both the kinship matrix (K) and PC levels (Q) for considering family relatedness and population structure. Previous studies [37–39] have debated that Bonferroni correction for marker effects using both Q and K could result in over-correcting and the need to use a lower significance level of the p-value. The results of GWA analysis are shown in the Manhattan plot in Fig 1.
X-axis: SNPs positions on chromosomes, Y-axis: -Log10 p-value. The red dashed line indicates the threshold for chromosome-wide (P<1.01*10−4) statistical significance based on the simpleM approach. All p-values were presented without any inflation (ʎ = 1).
Three SNPs located on OAR2, OAR10, and OAR25 were identified as chromosome-wide significantly associated with litter size in Baluchi sheep (Table 3).
The Pairwise-comparisons of significant SNP genotypes are shown in Fig 2. The significant SNP, rs406187720, on OAR2, was located within the NTRK2 gene. For this SNP, EBV for A/A (n = 1) genotype was greater than the mean EBV for G/A (n = 18) and G/G (n = 64) genotypes and the difference between G/A and G/G was statistically significant (p < 0.05). Mean comparisons (Fig 2) and the regression coefficient of A allele and A/A genotype (Table 2) of this SNP showed an additive effect. The SNP on OAR10 was not near any gene in 1Mb span, but genotypes mean comparison of this SNP showed that the addition of an A allele could increase the EBV for ewes carrying that allele (Fig 2). Another significant SNP, rs430079982, was identified on OAR25 within ovine known gene RAB4A. For this SNP, the EBV of ewes with the genotype G/G (n = 44) was significantly greater than that of the ewes with the genotype G/A (n = 31) and A/A (n = 9) (p < 0.01 and p < 0.001, respectively) (Fig 2). As it is clear from regression coefficients in Table 2, this SNP has a negative effect on litter size.
Differences between EBV means of each genotype in each significant SNP were calculated. Tukey’s multiple pairwise-comparison test was used for mean comparisons between genotypes. For significant SNP on OAR2, mean comparison only performed between G/A and G/G genotypes. *: p < 0.05, **: p < 0.01, ***: p < 0.001, n.s: not significant.
Gene set analysis
The genes or genomic regions that were identified in GWAS explain only part of the genetic variation. To overcome this limitation, the GWAS was complemented with a gene set analysis using the GO and KEGG database to detect potential functional categories underlying the litter size. In total, from 45,342 SNPs tested in the GWAS, 23,462 SNPs were located within or 15 kb upstream or downstream of 15,815 genes in the Oar.v3.1 ovine genome assembly. About 1,235 genes, out of 15,815 genes contained at least one significant SNP (p-value ≤ 0.05) and defined as significantly associated genes with the litter size. GO terms and KEGG pathways with a nominal p-value ≤ 0.05 were reported as significant (Fig 3).
The numbers within the bars denote the number of significant genes inside GO/KEGG terms.
Two KEGG terms related to steroid hormones, including Ovarian steroidogenesis (oas04913) and Steroid hormone biosynthesis (oas00140) showed an overrepresentation of significant genes associated with litter size. Many KEGG terms related to signaling also were identified as significant terms including Calcium signaling pathway (oas04020), Sphingolipid signaling pathway (oas04071), Chemokine signaling pathway (oas04062), and Phospholipase D signaling pathway (oas04072). Also, we identified several GO terms associated with litter size. Pathways related to the adhesion process e.g. biological adhesion (GO:0022610) and cell adhesion (GO:0007155) were among the significant terms. The neuromuscular junction (GO:0031594), cell junction (GO:0030054), and cell-cell adherens junction (GO:0005913) were other GO terms that enriched with significant genes. Two significant GO terms including ion channel regulator activity (GO:0099106) and calcium channel regulator activity (GO:0005246) were related to ion transport and channel activity. The cell cycle G2/M phase transition (GO:0044839) was another GO term which overrepresented with significant genes.
Discussion
Genome-wide association analysis
In the present study, to identify genomic regions and candidate genes associated with the litter size in sheep, a whole-genome scan was performed. Also, to cover some limitations of GWAS, a gene set analysis was used as a complementary approach. We identified the Neurotrophic receptor tyrosine kinase 2 (NTRK2) gene on OAR2 as a candidate gene affecting litter size in sheep. As shown in Fig 2, in the NTRK2 gene, the addition of an A allele increases the average EBV in the A/G genotype, and the difference between the A/G and G/G genotypes was significant. Although the individual with A/A genotype has the greatest EBV than the other two genotypes, due to a single observation, the result needs to be confirmed by larger sample size. Of course, it should be noted that the EBVs are calculated not only by the individual records but also by using their relatives’ records and means that contain information from both the individuals and relatives of the individuals. However, the NTRK2 gene functions in oocyte development and survival, make the result sensible.
Neurotrophic receptor tyrosine kinase 2 (NTRK2) is a transmembrane receptor that binds to brain-derived neurotrophic factor (BDNF) and neurotropin-4/5 (NTF-4/5) with high affinity. The BDNF and NTF-4/5 are members of neurotrophins (NTs). It has long been thought that NTs and their receptors to be exclusively required for the development of the nervous system. But, findings suggest that they are vital for the control of ovarian functions and contribute to both the formation and development of follicles [40]. Expression of BDNF and NTRK2 during the preimplantation have been observed in the oviduct, trophectoderm cells, and uterus in mice [41]. Also, their expression in oviducts and uteri of nonpregnant sheep and high expression in uteri of pregnant sheep has been reported [42]. Studies using Ntrk2-null mice showed that this molecule work as a receptor which is essential for follicular assembly, structural integrity and, early follicular growth [43, 44]. It has been shown that primordial follicle formation is decreased in the absence of NTRK2 or its ligands, NTF4/5, and BDNF. This deficiency just takes place after the initiation of folliculogenesis in the absence of the NTRK2 receptor. The ligand-mediated activation of the NTRK2 receptor enhances the capacity of the infantile ovary to respond to FSH with the synthesis of cyclin D2 which is a cell cycle protein mediating the proliferative actions of FSH in the ovary [45]. Both the number of secondary follicles and FSH receptor (FSHR) expression are reduced in Ntrk2-null ovaries and suggested that NTRK2 facilitate subsequent follicle development by inducing the formation of functional FSH receptors [46].
It has been shown that the preovulatory surge of gonadotropins increases BDNF synthesis in granulosa cells which facilitate oocyte development by mediating NTRK2 receptors expressed in the oocyte. Consequently, this process supports the development of zygotes into preimplantation embryos by PI3K-AKT signaling pathway activation [41, 47]. Oocytes without the NTRK2 gene can’t respond to gonadotropins due to a lack of PI3K-AKT mediated signaling activation [48]. Notably, the PI3K-Akt signaling pathway (oas04151) was one of our identified KEGG pathways with 21 significant genes, but it was not significant overall. Recently, using RNA-seq analysis, the PI3K-Akt signaling pathway was reported as a significant KEGG term affecting prolificacy [49].
It has been shown that kisspeptin and its receptor, KISS1R are vital for NTRK2 function, and oocytes lacking KISS1R do not respond to gonadotropins while NTRK2 expression is normal [50]. According to the model described by Dorfman et al. [50], before puberty, oocytes only express a truncated NTRK2 form (NTRK2.T1) receptors and KISS1R. The preovulatory surge of gonadotropins at puberty increases the production of BDNF and kisspeptin in granulosa cells which binds to their receptors (NTKR2.T1 and KISS1R) in oocytes and promote the development of full-length and active NTRK2 receptors (NTKR2.FL) that activate a PI3K-AKT-mediated pathway which ensures oocyte survival. In lack of full-length NTRK2 (NTRK2.FL) the follicles lose their integrity and the oocytes can’t survive [50]. This suggested model by Dorfman et al. [50] to explain the NTRK2 mechanism of action can be seen here (https://academic.oup.com/view-large/figure/62307720/zee9991476020008.tif). NTRK2.FL: Full-length NTRK2; NTRK2.T1: Truncated NTRK2.
To date, most reports of the NTRK2 gene have been limited to molecular studies, but in a recent study, using RNA-seq analysis, Xia et al. [51], tried to identify fecundity-related lncRNAs and mRNAs in Han sheep. Notably, they identified both 105,603,287 lncRNA and its target gene, NTRK2, as significantly upregulated genes associated with sheep prolificacy. In addition, they reported Ovarian steroidogenesis and Steroid hormone biosynthesis KEGG pathways as significant, which both of them were identified in this study. In another study, the NTRK2 was found to be differentially upregulated in the ovaries of Hu sheep in association with off-season reproduction [52]. Besides, it’s reported that a variant close to the NTRK2 gene are associated with birth weight in female twins in human [53]. Based on the results of this study and other studies mentioned above, the NTRK2 gene can be introduced as an important candidate for further studies related to sheep prolificacy.
The SNP on OAR10 was identified as significant but was not near any gene in the 1Mb span. Genotypes mean comparison showed that the addition of A allele could significantly increase the average of EBV in A/G and A/A genotypes in comparison with the G/G genotype. However, the difference between G/A and A/A genotype was not significant (Fig 2).
Another significant SNP was rs430079982 located within the RAB4A gene on OAR25. This gene encodes Ras-related protein Rab-4A in humans. RAB4A is a member of the Rab GTPase superfamily which has many roles in membrane trafficking. Also, this gene regulates receptor recycling from endocytic compartments to the plasma membrane [54, 55]. RAB4A was firstly identified in early endosomes in Chinese hamster ovary [56]. It has been reported that RAB4A has a role in endosome-to-plasma membrane recycling of vascular endothelial growth factor receptor 2 (VEGFR-2) which is vital for blood vessel formation, cell migration, intracellular signaling, and proliferation through angiogenesis [57]. Angiogenesis is linked with follicular development and is controlled independently within each follicle which potentially making the functioning of its vasculature critically important in determining its fate [58]. Establishment and remodeling of a complex vascular system enable the follicle and corpus luteum to receive the required supply of nutrients, oxygen, and hormonal support as well as facilitating the release of steroids [59]. It has been reported that the production and action of vascular endothelial growth factor A (VEGFA) are necessary for follicular growth, ovulation, and the development and function of the corpus luteum [60]. VEGF is a potent and specific stimulator of vascular endothelial cell proliferation acting through two tyrosine kinase receptors, VEGFR-1 and VEGFR-2 [58]. Administration of VEGFA, stimulate the development of secondary follicles in cows [60]. Inhibition of VEGF and its receptor VEGFR-2 can inhibit follicular development or prevent ovulation [58]. Genotypes mean comparison showed that the average EBV for G/G genotype was significantly different from G/A and A/A genotypes.
Gene set analysis
In this study, the GWAS was complemented with a gene-set analysis using KEGG and GO databases to detect potential functional categories underlying the litter size. A full list of significantly enriched pathways using GO and KEGG databases are presented in S1–S3 Tables. Notably, two KEGG pathways related to Steroid hormone including Ovarian steroidogenesis (oas04913) and Steroid hormone biosynthesis (oas00140) were significant. Recently, both pathways were identified in a transcriptome analysis in association with prolificacy in Han sheep [51]. Also, using protein profile analysis, both Ovarian steroidogenesis and Steroid hormone biosynthesis were reported as significant KEGG pathways affecting prolificacy [61]. In addition, they identified the protein digestion and absorption KEGG pathways as significant. Similarly, in this study, the Protein processing in endoplasmic reticulum (oas04141) KEGG term was identified as a pathway associated with protein metabolism. Both Ovarian steroidogenesis and protein digestion and absorption KEGG pathways have recently been reported in a GWAS study as pathways affecting prolificacy in sheep [62]. Many signaling KEGG pathways were identified and some of them were related to reproductive processes. The calcium signaling pathway as the most significant term was one of them. Calcium has an important function in the oocyte’s meiotic maturation. Recently, Hernández-Montiel et al. [63] using RNA-seq transcriptome analysis in prolific and non-prolific Mexican Pelibuey sheep breed, reported the Calcium signaling pathway as a significant KEGG term. In addition, Xu et al. [64] in a GWAS, identified calcium ion binding GO term as pathway affecting prolificacy in Chinese sheep. Notably, we identified calcium channel regulator activity (GO:0005246) as a significant GO term in our gene set analysis using the GO database.
The Chemokine signaling (oas04062) was another significant KEGG pathway. Different aspects of ovulation are very similar to the inflammatory process and a diverse set of leukocytes have been observed in different stages of follicle development and ovulation [65]. Phospholipase D signalling (oas04072) was another significant KEGG pathway in the present study. Banno et al. [66] suggested that Phospholipase D (PLD) has a role in growth factor regulation in at least two signaling pathways, including extracellular signal-regulated kinases (ERKs) and the PI3k-Akt pathway. In addition, PLD is involved in the inflammatory process. IL-8 as an inflammatory cytokine up-regulates the expression of PLD and enables PLD to facilitate the migration of immune cells [67]. Notably, we identified two functionally related KEGG pathways including Folate biosynthesis (oas00790) and Cysteine and methionine metabolism (oas00270). Folate (B9 vitamin) is essential for rapid cell growth and division, which usually occurs during follicle development and fetal growth. Folate deficiency can lead to homocysteine accumulation. Homocysteine originates from the methionine and is remethylated to methionine with folates acting as methyl donors [68]. Adverse impacts of homocysteine accumulation due to folate deficiency on female reproductive performance can include inflammatory cytokine production, low cell division [69], increase in apoptosis [70], and high oxidative stress [71] which affects many aspects of oocyte development. Women who take folic acid supplements have lower homocysteine concentrations in their follicular fluid, oocytes with better quality, and a high degree of mature oocytes [72]. It has been reported that the folate concentration in prolific breeds ewes is higher than from non-prolific breeds [73], but a more recent study reported that folic acid supplementation to prolific and nonprolific ewes increase folate concentration in red cells and plasma but hasn’t any effect on fertility or litter size [74].
Our gene set analysis using the GO database identified several terms associated with litter size. Two pathways related to carbohydrate metabolism including cellular carbohydrate metabolic process (GO:0044262) and carbohydrate derivative binding (GO:0097367) showed an overrepresentation of significant genes. Carbohydrate and lipid oxidation have a major impact on fecundity. Glucose and pyruvate are the main energy source for mouse ovarian follicles and oocyte, respectively. Using in vitro grown follicles, it has been reported that glucose uptake and lactate production by follicles are increased parallelly with development and are stimulated just before ovulation [75]. In another study ovarian proteome of hyper-prolificacy sheep was analyzed and more than 100 identified proteins were classified in carbohydrate transport and metabolism [76]. Also, Yang et al. [77] using transcriptome analysis, reported galactose metabolism and fructose and mannose metabolism as biological pathways affecting ovulation. Many GO terms related to cell communication such as biological adhesion (GO:0022610), cell adhesion (GO:0007155), cell junction (GO:0030054), and cell-cell adherens junction (GO:0005913) were identified. Given that the oocytes are surrounded by many cell layers, it makes sense that signaling and communication between the oocytes and the environment have a great impact on follicle growth and the ovulation process. Besides the other communication ways between the oocyte and surrounding cells, physical connections are essential for follicular growth and oocyte maturation. Cadherin family genes as part of the cell adhesion process, encode glycoproteins that are responsible for cell-cell communication through calcium ions [78]. The gap junction is one of the direct child terms of cell junction pathway which contains several proteins called connexin. These proteins form intercellular membrane channels that connect adjacent cells and allow for direct sharing of several metabolites such as glucose, amino acids, nucleotides, ions, and second messengers (e.g. cGMP) to support the oocyte development, metabolism, and hemostasis [79]. Many genomics, transcriptomics, and proteomics studies reported many pathways related to both cell adhesion and gap junction process as pathways which has vital roles in prolificacy [49, 62, 77, 80, 81].
It should be noted that, at this point, reported genes are just candidates. The identification of validated causal genetic variants that underlie reproduction traits is big challenges in livestock genetic research and methods such as whole-genome resequencing, high-throughput transcriptome sequencing, and functional studies are required to fully address the causal relationship between genetic variants and phenotypes [82, 83]. Therefore, although there are several reports about the NTRK2 gene functions in oocyte growth, development, and survival, its variant’s effect on the litter size needs to be confirmed with a larger sample size. Recently, however, its effect on sheep prolificacy has been reported using transcriptome analysis [51]. Finally, while this study does improve our understanding of an interesting but less characterized breed, it will still be useful to see if these results can be broadly applicable to other breeds as well.
Conclusions
In the present study, a two-step GWAS accompanied by a gene set analysis (GSA) as a complementary approach was utilized to identify genes and pathways affecting the litter size in Iranian Baluchi sheep. Two new genes including NTRK2 and RAB4A and one genomic region on OAR10 were identified as regions associated with the litter size in sheep. The NTRK2 gene product works as a receptor which is essential for follicular assembly, early follicular growth, and oocyte survival by activation of the PI3K-AKT signaling pathway. This is the first study that reports the NTRK2 gene affecting litter size in sheep. Another identified gene, RAB4A, is involved in intracellular signaling and proliferation through angiogenesis. Several KEGG pathways related to steroid hormone biosynthesis and signaling were identified. In addition, we identified many GO terms related to carbohydrate metabolism and cell communication via adhesion and junctions as significant terms. Finally, reported genes and pathways provide valuable information that can be helpful to a better understanding of the mechanisms underlying sheep litter size variation.
Supporting information
S1 Table. GO terms.
GO terms significantly (P ≤ 0.05) enriched using genes associated with litter size in Baluchi sheep.
https://doi.org/10.1371/journal.pone.0244408.s001
(XLSX)
S2 Table. KEGG pathways.
KEGG pathways significantly (P ≤ 0.05) enriched using genes associated with litter size in Baluchi sheep.
https://doi.org/10.1371/journal.pone.0244408.s002
(XLSX)
S3 Table. GWAS results.
GWAS results for litter size in Baluchi sheep.
https://doi.org/10.1371/journal.pone.0244408.s003
(XLSX)
Acknowledgments
SMEF thanks the Abureyhan Campus of the University of Tehran for hosting during the research.
References
- 1. Gholizadeh M, Rahimi-Mianji G, Nejati-Javaremi A, De Koning DJ, Jonas E. Genomewide association study to detect QTL for twinning rate in Baluchi sheep. J Genet. 2014 Aug;93(2):489–93. pmid:25189245
- 2. Kirkpatrick BW, Morris CA. A major gene for bovine ovulation rate. PLoS One. 2015;10(6):e0129025. pmid:26046917
- 3. Abdoli R, Zamani P, Deljou A, Rezvan H. Association of BMPR-1B and GDF9 genes polymorphisms and secondary protein structure changes with reproduction traits in Mehraban ewes. Gene. 2013;524(2):296–303. pmid:23583795
- 4. Drouilhet L, Lecerf F, Bodin L, Fabre S, Mulsant P. Fine mapping of the FecL locus influencing prolificacy in Lacaune sheep. Anim Genet. 2009;40(6):804–12. pmid:19466934
- 5. Hanrahan JP, Gregan SM, Mulsant P, Mullen M, Davis GH, Powell R, et al. Mutations in the genes for oocyte-derived growth factors GDF9 and BMP15 are associated with both increased ovulation rate and sterility in Cambridge and Belclare sheep (Ovis aries). Biol Reprod. 2004;70(4):900–9. pmid:14627550
- 6. Davis GH, McEwan JC, Fennessy PF, Dodds KG, Farquhar PA. Evidence for the presence of a major gene influencing ovulation rate on the X chromosome of sheep. Biol Reprod. 1991;44(4):620–4. pmid:2043732
- 7.
Bindon BM, Piper LR. Booroola (F) gene: major gene affecting ovine ovarian function. In: Genetic Engineering of Animals. Springer; 1986. p. 67–93.
- 8. Drouilhet L, Mansanet C, Sarry J, Tabet K, Bardou P, Woloszyn F, et al. The highly prolific phenotype of Lacaune sheep is associated with an ectopic expression of the B4GALNT2 gene within the ovary. PLoS Genet. 2013;9(9):e1003809. pmid:24086150
- 9. Davis GH, Dodds KG, Wheeler R, Jay NP. Evidence that an imprinted gene on the X chromosome increases ovulation rate in sheep. Biol Reprod. 2001;64(1):216–21. pmid:11133677
- 10. Juengel JL, O’Connell AR, French MC, Proctor LE, Wheeler R, Farquhar PA, et al. Identification of a line of sheep carrying a putative autosomal gene increasing ovulation rate in sheep that does not appear to interact with mutations in the transforming growth factor beta superfamily. Biol Reprod. 2011;85(1):113–20. pmid:21415136
- 11. Moradband F, Rahimi G, Gholizadeh M. Association of polymorphisms in fecundity genes of GDF9, BMP15 and BMP15-1B with litter size in Iranian Baluchi sheep. Asian-Australasian J Anim Sci. 2011;24(9):1179–83.
- 12. Khodabakhshzadeh R, Mohammadabadi MR, Shahrebabak HM, Koshkoieh AE. Identification of available mutations in the first-half (from 5′ end) of exon 2 of GDF9 gene in crossbred sheep from crossing of Romanov and Lori-Bakhtiari breeds. Anim Prod Res. 2015;4(4).
- 13. Abdoli R, Mirhoseini SZ, Ghavi Hossein-Zadeh N, Zamani P, Gondro C. Genome-wide association study to identify genomic regions affecting prolificacy in Lori-Bakhtiari sheep. Anim Genet. 2018 Oct;49(5):488–91. pmid:30079564
- 14. Abbasi MA, Abdollahi-Arpanahi R, Maghsoudi A, Torshizi RV, Nejati-Javaremi A. Evaluation of models for estimation of genetic parameters and maternal effects for early growth traits of Iranian Baluchi sheep. Small Rumin Res. 2012 May;104(1–3):62–9.
- 15. Tahmoorespur M, Sheikhloo M. Pedigree analysis of the closed nucleus of Iranian Baluchi sheep. Small Rumin Res. 2011;99(1):1–6.
- 16. Peng G, Luo L, Siu H, Zhu Y, Hu P, Hong S, et al. Gene and pathway-based second-wave analysis of genome-wide association studies. Eur J Hum Genet. 2010;18(1):111. pmid:19584899
- 17. Han Y, Peñagaricano F. Unravelling the genomic architecture of bull fertility in Holstein cattle. BMC Genet. 2016;17(1):143. pmid:27842509
- 18. Wang K, Li M, Bucan M. Pathway-based approaches for analysis of genomewide association studies. Am J Hum Genet. 2007;81(6):1278–83. pmid:17966091
- 19. Dadousis C, Pegolo S, Rosa GJM, Gianola D, Bittante G, Cecchinato A. Pathway-based genome-wide association analysis of milk coagulation properties, curd firmness, cheese yield, and curd nutrient recovery in dairy cattle. J Dairy Sci. 2017;100(2):1223–31. pmid:27988128
- 20. Chasman DI. On the utility of gene set methods in genomewide association studies of quantitative traits. Genet Epidemiol Off Publ Int Genet Epidemiol Soc. 2008;32(7):658–68. pmid:18481796
- 21. Wang L, Jia P, Wolfinger RD, Chen X, Zhao Z. Gene set analysis of genome-wide association studies: methodological issues and perspectives. Genomics. 2011;98(1):1–8. pmid:21565265
- 22. Aulchenko YS, Ripke S, Isaacs A, Van Duijn CM. GenABEL: an R library for genome-wide association analysis. Bioinformatics. 2007;23(10):1294–6. pmid:17384015
- 23.
R Development Core Team. R Development Core Team. R: A Language and Environment for Statistical Computing, 2019. Available in< https//www r-project org. 2019;
- 24. Misztal I, Tsuruta S, Lourenco DAL, Masuda Y, Aguilar I, Legarra A, et al. Manual for BLUPF90 family of programs. 2018.
- 25.
Masuda Y. Introduction to BLUPF90 suite programs:Standard Edition. Univ Georg http//nce.ads.uga.edu/wiki/doku.php?id = documentation. 2019;
- 26. Gao X, Starmer J, Martin ER. A multiple testing correction method for genetic association studies using correlated single nucleotide polymorphisms. Genet Epidemiol Off Publ Int Genet Epidemiol Soc. 2008;32(4):361–9. pmid:18271029
- 27.
Wickham H. ggplot2: elegant graphics for data analysis. Springer; 2016.
- 28. Abdalla EA, Peñagaricano F, Byrem TM, Weigel KA, Rosa GJM. Genome‐wide association mapping and pathway analysis of leukosis incidence in a US Holstein cattle population. Anim Genet. 2016;47(4):395–407. pmid:27090879
- 29. Durinck S, Spellman PT, Birney E, Huber W. Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat Protoc. 2009;4(8):1184. pmid:19617889
- 30. Durinck S, Moreau Y, Kasprzyk A, Davis S, De Moor B, Brazma A, et al. BioMart and Bioconductor: a powerful link between biological databases and microarray data analysis. Bioinformatics. 2005;21(16):3439–40. pmid:16082012
- 31. Pickrell JK, Marioni JC, Pai AA, Degner JF, Engelhardt BE, Nkadori E, et al. Understanding mechanisms underlying human gene expression variation with RNA sequencing. Nature. 2010;464(7289):768–72. pmid:20220758
- 32. Yu G, Wang L-G, Han Y, He Q-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omi a J Integr Biol. 2012;16(5):284–7.
- 33. Yazdi MH, Johansson K, Gates P, Näsholm A, Jorjani H, Liljedahl L-E. Bayesian analysis of birth weight and litter size in Baluchi sheep using Gibbs sampling. J Anim Sci. 1999;77(3):533–40. pmid:10229348
- 34. Jafaroghli M, Safari A, Shadparvar AA, Ghavi Hossein-Zadeh N. Genetic Analysis of Ewe Productivity Traits in Baluchi Sheep. Iran J Appl Anim Sci. 2019;9(4):651–7.
- 35. Janssens S, Vandepitte W, Bodin L. Genetic parameters for litter size in sheep: natural versus hormone-induced oestrus. Genet Sel Evol. 2004;36(5):1–20. pmid:15339632
- 36. Kaler AS, Purcell LC. Estimation of a significance threshold for genome-wide association studies. BMC Genomics [Internet]. 2019;20(1):618. Available from: https://doi.org/10.1186/s12864-019-5992-7 pmid:31357925
- 37. Dundas IS, Nair RM, Verlin DC. First report of meiotic chromosome number and karyotype analysis of an accession of Trigonella balansae (Leguminosae). New Zeal J Agric Res. 2006;49(1):55–8.
- 38. Gao L, Turner MK, Chao S, Kolmer J, Anderson JA. Genome wide association study of seedling and adult plant leaf rust resistance in elite spring wheat breeding lines. PLoS One. 2016;11(2):e0148671. pmid:26849364
- 39. Kaler AS, Abdel-Haleem H, Fritschi FB, Gillman JD, Ray JD, Smith JR, et al. Genome-Wide Association Mapping of Dark Green color index using a Diverse panel of Soybean Accessions. Sci Rep. 2020;10(1):1–11. pmid:31913322
- 40.
Dissen GA, Garcia-Rudaz C, Ojeda SR. Role of neurotrophic factors in early ovarian development. In: Seminars in reproductive medicine. NIH Public Access; 2009. p. 24. https://doi.org/10.1055/s-0028-1108007 pmid:19197802
- 41. Kawamura K, Kawamura N, Fukuda J, Kumagai J, Hsueh AJW, Tanaka T. Regulation of preimplantation embryo development by brain-derived neurotrophic factor. Dev Biol. 2007;311(1):147–58. pmid:17880937
- 42. Mirshokraei P, Hassanpour H, Rahnama A, Foster WG. Gene expression of BDNF and its receptors, TrkB and p75 in the uterus and oviduct of pregnant and non-pregnant ewes. Res Vet Sci. 2013;95(1):164–8. pmid:23618676
- 43. Paredes A, Romero C, Dissen GA, DeChiara TM, Reichardt L, Cornea A, et al. TrkB receptors are required for follicular growth and oocyte survival in the mammalian ovary. Dev Biol. 2004;267(2):430–49. pmid:15013804
- 44. Spears N, Molinek MD, Robinson LLL, Fulton N, Cameron H, Shimoda K, et al. The role of neurotrophin receptors in female germ-cell survival in mouse and human. Development. 2003;130(22):5481–91. pmid:14507777
- 45. Sicinski P, Donaher JL, Geng Y, Parker SB, Gardner H, Park MY, et al. Cyclin D2 is an FSH-responsive gene involved in gonadal cell proliferation and oncogenesis. Nature. 1996;384(6608):470–4. pmid:8945475
- 46. Kerr B, Garcia-Rudaz C, Dorfman M, Paredes A, Ojeda SR. NTRK1 and NTRK2 receptors facilitate follicle assembly and early follicular development in the mouse ovary. Reproduction [Internet]. 2009/04/08. 2009 Jul;138(1):131–40. Available from: https://pubmed.ncbi.nlm.nih.gov/19357131 pmid:19357131
- 47. Kawamura K, Kawamura N, Mulders SM, Gelpke MDS, Hsueh AJW. Ovarian brain-derived neurotrophic factor (BDNF) promotes the development of oocytes into preimplantation embryos. Proc Natl Acad Sci. 2005;102(26):9206–11. pmid:15967989
- 48. Zheng W, Nagaraju G, Liu Z, Liu K. Functional roles of the phosphatidylinositol 3-kinases (PI3Ks) signaling in the mammalian ovary. Mol Cell Endocrinol. 2012;356(1–2):24–30. pmid:21684319
- 49. Hernández-Montiel W, Collí-Dula RC, Ramón-Ugalde JP, Martínez-Núñez MA, Zamora-Bustillos R. RNA-seq Transcriptome Analysis in Ovarian Tissue of Pelibuey Breed to Explore the Regulation of Prolificacy. Genes (Basel). 2019;10(5):358. pmid:31083386
- 50. Dorfman MD, Garcia-Rudaz C, Alderman Z, Kerr B, Lomniczi A, Dissen GA, et al. Loss of Ntrk2/Kiss1r signaling in oocytes causes premature ovarian failure. Endocrinology. 2014;155(8):3098–111. pmid:24877631
- 51. Xia Q, Li Q, Gan S, Guo X, Zhang X, Zhang J, et al. Exploring the roles of fecundity-related long non-coding RNAs and mRNAs in the adrenal glands of small-tailed Han Sheep. BMC Genet. 2020;21:1–11. pmid:31900126
- 52. Chen L, Liu K, Zhao Z, Blair HT, Zhang P, Li D, et al. Identification of sheep ovary genes potentially associated with off-season reproduction. J Genet Genomics. 2012;39(4):181–90. pmid:22546540
- 53. Metrustry SJ, Edwards MH, Medland SE, Holloway JW, Montgomery GW, Martin NG, et al. Variants close to NTRK2 gene are associated with birth weight in female twins. Twin Res Hum Genet. 2014;17(4):254–61. pmid:24950379
- 54. Stenmark H. Rab GTPases as coordinators of vesicle traffic. Nat Rev Mol cell Biol. 2009;10(8):513–25. pmid:19603039
- 55. Goueli BS, Powell MB, Finger EC, Pfeffer SR. TBC1D16 is a Rab4A GTPase activating protein that regulates receptor recycling and EGF receptor signaling. Proc Natl Acad Sci. 2012;109(39):15787–92. pmid:23019362
- 56. Van Der Sluijs P, Hull M, Zahraoui A, Tavitian A, Goud B, Mellman I. The small GTP-binding protein rab4 is associated with early endosomes. Proc Natl Acad Sci. 1991;88(14):6313–7. pmid:1906178
- 57. Jopling HM, Odell AF, Pellet-Many C, Latham AM, Frankel P, Sivaprasadarao A, et al. Endosome-to-plasma membrane recycling of VEGFR2 receptor tyrosine kinase regulates endothelial function and blood vessel formation. Cells. 2014;3(2):363–85. pmid:24785348
- 58. Fraser HM. Regulation of the ovarian follicular vasculature. Reprod Biol Endocrinol [Internet]. 2006 Apr 12;4:18. Available from: https://pubmed.ncbi.nlm.nih.gov/16611363 pmid:16611363
- 59. Robinson RS, Woad KJ, Hammond AJ, Laird M, Hunter MG, Mann GE. Angiogenesis and vascular function in the ovary. Reproduction. 2009;138(6):869–81. pmid:19786399
- 60. Yang MY, Fortune JE. Vascular endothelial growth factor stimulates the primary to secondary follicle transition in bovine follicles in vitro. Mol Reprod Dev. 2007;74(9):1095–104. pmid:17290425
- 61. Tang J, Hu W, Chen S, Di R, Liu Q, Wang X, et al. The genetic mechanism of high prolificacy in small tail han sheep by comparative proteomics of ovaries in the follicular and luteal stages. J Proteomics. 2019;204:103394. pmid:31146049
- 62. Hernández-Montiel W, Martínez-Núñez MA, Ramón-Ugalde JP, Román-Ponce SI, Calderón-Chagoya R, Zamora-Bustillos R. Genome-Wide Association Study Reveals Candidate Genes for Litter Size Traits in Pelibuey Sheep. Animals. 2020;10(3):434. pmid:32143402
- 63. Garza Hernandez D, Mucha S, Banos G, Kaseja K, Moore K, Lambe N, et al. Analysis of single nucleotide polymorphisms variation associated with important economic and computed tomography measured traits in Texel sheep. Animal. 2018 May;12(5):915–22. pmid:29037276
- 64. Xu S-S, Gao L, Xie X-L, Ren Y-L, Shen Z-Q, Wang F, et al. Genome-wide association analyses highlight the potential for different genetic mechanisms for litter size among sheep breeds. Front Genet. 2018;9:118. pmid:29692799
- 65. Richards JS, Liu Z, Shimada M. Immune-like mechanisms in ovulation. Trends Endocrinol Metab. 2008;19(6):191–6. pmid:18407514
- 66. Banno Y, Takuwa Y, Yamada M, Takuwa N, Ohguchi K, Hara A, et al. Involvement of phospholipase D in insulin-like growth factor-I-induced activation of extracellular signal-regulated kinase, but not phosphoinositide 3-kinase or Akt, in Chinese hamster ovary cells. Biochem J. 2003;369(2):363–8. pmid:12385647
- 67. Kang DW, Choi K-Y. Functional regulation of phospholipase D expression in cancer and inflammation. J Biol Chem. 2014;289(33):22575–82. pmid:24990948
- 68. Lucock M. Folic acid: nutritional biochemistry, molecular biology, and role in disease processes. Mol Genet Metab. 2000;71(1–2):121–38. pmid:11001804
- 69. Gmyrek GB, Sozanski R, Jerzak M, Chrobak A, Wickiewicz D, Skupnik A, et al. Evaluation of monocyte chemotactic protein-1 levels in peripheral blood of infertile women with endometriosis. Eur J Obstet Gynecol Reprod Biol. 2005;122(2):199–205. pmid:15893866
- 70. Hussein MR. Apoptosis in the ovary: molecular mechanisms. Hum Reprod Update. 2005;11(2):162–78. pmid:15705959
- 71. Agarwal A, Gupta S, Sharma RK. Role of oxidative stress in female reproduction. Reprod Biol Endocrinol. 2005;3(1):28. pmid:16018814
- 72. Szymański W, Kazdepka-Ziemińska A. Effect of homocysteine concentration in follicular fluid on a degree of oocyte maturity. Ginekol Pol. 2003;74(10):1392–6. pmid:14669450
- 73. Girard CL, Castonguay F, Fahmy MH, Matte JJ. Serum and milk folates during the first two gestations and lactations in Romanov, Finnsheep, and Suffolk ewes. J Anim Sci. 1996;74(7):1711–5. pmid:8818819
- 74. Méthot H, Girard CL, Matte JJ, Castonguay FW. Effects of dietary supplements of folic acid on reproductive performance in ewes. Can J Anim Sci. 2008;88(3):489–97.
- 75. Harris SE, Adriaens I, Leese HJ, Gosden RG, Picton HM. Carbohydrate metabolism by murine ovarian follicles and oocytes grown in vitro. Reproduction. 2007;134(3):415–24. pmid:17709560
- 76. Miao X, Luo Q, Zhao H, Qin X. Ovarian proteomic study reveals the possible molecular mechanism for hyperprolificacy of Small Tail Han sheep. Sci Rep. 2016;6(1):1–10. pmid:28442746
- 77. Yang F, Wang M, Zhang B, Xiang W, Zhang K, Chu M, et al. Identification of new progestogen-associated networks in mammalian ovulation using bioinformatics. BMC Syst Biol. 2018;12(1):36. pmid:29615037
- 78. Angst BD, Marcozzi C, Magee AI. The cadherin superfamily: diversity in form and function. J Cell Sci. 2001;114(4):629–41. pmid:11171368
- 79. Winterhager E, Kidder GM. Gap junction connexins in female reproductive organs: implications for women’s reproductive health. Hum Reprod Update. 2015;21(3):340–52. pmid:25667189
- 80. Chen HY, Shen H, Jia B, Zhang YS, Wang XH, Zeng XC. Differential gene expression in ovaries of Qira black sheep and Hetian sheep using RNA-Seq technique. PLoS One. 2015;10(3). pmid:25790350
- 81. Pokharel K, Peippo J, Honkatukia M, Seppälä A, Rautiainen J, Ghanem N, et al. Integrated ovarian mRNA and miRNA transcriptome profiling characterizes the genetic basis of prolificacy traits in sheep (Ovis aries). BMC Genomics. 2018;19(1):104. pmid:29378514
- 82. Gao Y, Jiang J, Yang S, Hou Y, Liu GE, Zhang S, et al. CNV discovery for milk composition traits in dairy cattle using whole genome resequencing. BMC Genomics. 2017;18(1):265. pmid:28356085
- 83. Ma L, Li Z, Cai Y, Xu H, Yang R, Lan X. Genetic variants in fat‐and short‐tailed sheep from high‐throughput RNA‐sequencing data. Anim Genet. 2018;49(5):483–7. pmid:30069889