Genome-wide association studies for sperm traits in Assaf sheep breed

Genome-wide association studies for sperm traits in Assaf sheep breed M. Serrano a, ⁎,M. Ramón b , J.H. Calvo c,f , M.Á. Jiménez a , F. Freire d , J.M. Vázquez d ,J.J. Arranz e a Departamento de Mejora Genética Animal, INIA, 28040 Madrid, Spain b IRIAF-CERSYRA, Valdepeñas 13300, Ciudad Real, Spain c Unidad de Tecnología en Producción Animal, CITA, 59059 Zaragoza, Spain d OVIGEN, Granja Florencia s/n, Ctra. Villalazán-Peleagonzalo, 49800 Toro, Zamora, Spain e Departamento de Producción Animal, Universidad de León, 24007 León, Spain f ARAID, 50004 Zaragoza, Spain abstractarticle info Article history: Received 6 May 2020 Received in revised form 20 August 2020 Accepted 21 August 2020 Available online xxxx Keywords: Association study Pseudo-phenotypes Rams Single-nucleotide polymorphisms genotypes Sperm parameters Sperm quality traits routinely collected by artificial insemination (AI) center for rams progeny test are related with the capacity to produce sperm doses for AI and, in more or less grade, with males' fertility. Low-quality ejaculates are unuseful to perform AI sperm doses, which suppose high economic loses for the AI center. Moreover, sperm quality traits have low heritability values which make traditional genetic selection little efficient to its improvement. In this work, a genome-wide association study (GWAS) was conducted by using sperm quality traits data and 50 K Affymetrix customchipgenotypesof 429 rams of Assafbreed from OVIGENAI centre. Furthermore, 47 of these rams were also genotyped with the Illumina HD Ovine BeadChip, and therefore HD genotypes were imputed for all rams with phenotype data. Previous to the GWAS, a linear regression model was fitted including sperm traits as dependent variables; the flock of origin, date of sperm collection, and jump number as fixed effects; rams age at collection in months as covariate; and ram permanent effect as random. Pseudo-phenotypes obtained from this model were used as input for GWAS. Associations at the chromosome-wise level (FDR 10%) of 76 single-nucleotide polymorphisms (SNPs) in 4 chromosomes for ejaculate concentration (CON), 20 SNPs in 3 chromosomes for ejaculate volume (VOL), 32 SNPs in 1 chromosome for ejaculate number of spermatozoa (SPZ), and 23 SNPs for spermatozoa mass motility (MOT) in 17 chromosomes were found. Only SNPs associated with MOT overcame the genome-wide significance level. Some candidate genes for sperm traits variability were SLC9C1 (OAR1), TSN (OAR2), and FUT10 (OAR26) for MOT;. DOCK2,CPLANE1,SPEF2, and RAI14 (OAR16) for CON; SCAPER and PSMA4 (OAR18) for VOL; and PARM1 and LOC101110593 (OAR6) for SPZ. SNPs associated with sperm traits were not found to be correlated with milk production genetic variation; however, the high frequencies of some SNPs with negativeeffectoverspermtraits found in animals at the top milk yieldestimatedbreeding values (EBVs) ranking would allow to exert some selective presure to improve rams sperm performances. Effects and frequencies of some of the SNPs detected over sperm quality traits make these variants good candidates to be used in marker-assisted selection to improve sperm characteristics of Assaf rams and AI center efficiency to produce sperm doses. © 2020 The Authors. Published by Elsevier Inc. on behalf of The Animal Consortium. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Implications Finding out the genetic basis of sperm quality traits in the ovine specie is a very important issue since their moderate to low heritability makes traditional selection methods little efficient to its improvement. Sperm traits are highly related to artificial insemination centers efficiency to perform rams progeny test in sheep dairy genetic programs. The genome mutations associated with the variability of sperm traits here detected could allow to perform a marker-assisted selection to improve rams sperm characteristics, and therefore, their reproductive ability and capacity to produce sperm doses. Introduction Artificial insemination (AI) is an essential reproductive tool in genetic breeding programs of dairy ruminants. The use of this technique enables progeny tests to predict estimated breeding values (EBVs)of males, contributes to connect flocks, and is the best strategy to disseminate the genetic improvement achieved by the genetic program using elite rams. In sheep insemination centers, the fertilizing capacity of the spermatozoa and the suitability of the semen for producing AI doses are assessed using three parameters: the volume of the ejaculate Animal xxx (2021) xxx ⁎Corresponding author. E-mail address: [email protected] (M. Serrano). ANIMAL-100065; No of Pages 9 https://doi.org/10.1016/j.animal.2020.100065 1751-7311/© 2020 The Authors. Published by Elsevier Inc. on behalf of The Animal Consortium. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Contents lists available at ScienceDirect Animal The international journal of animal biosciences Please cite this article as: M. Serrano, M. Ramón, J.H. Calvo, et al., Genome-wide association studies for sperm traits in Assaf sheep breed, Animal, https://doi.org/10.1016/j.animal.2020.100065 (VOL), the concentration of spermatozoa (CON), and the spermatozoa mass motility (MOT) measured on a subjective scale. VOL and CON phenotypes are related to the effective number of doses that can be elaborated from one ejaculate, while sperm motility has been described as a good indicator of sperm fertilizing ability (Gadea, 2005;Broekhuijse et al., 2012). However, a direct relationship between these sperm parameters and ram's fertility is not evident. The moderate heritability estimates found for sperm characteristics in some sheep breeds (David et al., 2007; Pelayo et al., 2019)jointo the fact that these traits are only collected in a very limited number of males (only rams from AI centers) makes conventional selection ineffective to improve them. Furthermore, the genetic correlation between dairy traits, milk yield and fat and protein contents, the main selection objective in most dairy sheep breeds, and sperm quality traits is unknown. If this correlation was negative, selection of elite rams for milk traits may lead to an impairment of ram's spermatic quality. Current genomic tools, such as medium and high-density genotyping single-nucleotide polymorphism (SNP) chips, could help to identify genomic regionsassociated with such low heritability traits, enablingto conduct marker assisted selection of sperm quality traits in livestock. In last years, some works dealing with genome-wide association studies (GWAS) of sperm quality traits have been conducted in cattle (Suchocki and Szyda, 2014; Hering et al., 2014; Fonseca et al., 2018; Yin et al., 2019;Qin et al., 2017), pigs (Diniz et al., 2014; Marques et al., 2018), and goats (Wang et al., 2020). All of them were performed by using medium-density SNP genotyping arrays and a variable number of genotyped animals. Some putative candidate genes were associated with more or less degree with sperm mass motility, volume, concentration, and number of spermatozoa. In goats (Wang et al., 2020), a strong association of the DSCAML1 gene with sperm motility and density was found. However, to date, no studies dealing with genome-wide association of sperm traits in sheep have been conducted. The main objective of this work is to identify SNPs and genomic regions associated with sperm quality traits by using genotypes from the 50 K Affiymetrix custom chip and the Illumina HD Ovine BeadChip. Material and methods Data Sperm quality traits were obtained from 27 635 ejaculates of 429 matured rams present in OVIGEN AI center, recorded between 2006 and 2018. Ejaculates were obtained after natural ejaculation in an artificial vagina. For a given ram, 1 or 2 successive ejaculates were collected over a 2–5 min period and evaluated immediately after collection. Threetraits were assessed for each ejaculate: VOL (ml), which was measured using a graduated collection tube, CON (spermatozoa × 10 6 /ml), which was determined using a standard spectrophotometer, and MOT, which was assessed for undiluted semen under a microscope. MOT was scored subjectively, based on wave motion on a continuous scale from 0 (no motion) to 5 (frequent rapid and vigorous waves) according to the original method described by Evans and WMC (1987). A fourth derived trait, the number of spermatozoa (SPZ) was computed as the product of VOL and CON and also included in the analyses. Data regarding these measurements are shown in Table 1. Statistics Phenotypes used in GWA studies were obtained from the sperm quality data described above. For that, in a first step, we performed a mixed linear regression analysis that allowed us to obtain a pseudophenotype for the GWAS of the second step. The four sperm traits, VOL, CON, SPZ, and MOT described above, evaluated in a total 27 635 ejaculates belonging to 429 matured rams, were considered as dependent variables in the analyses. A mixed linear regression analysis, including the ram flock of origin (51 levels), the date of sperm collection (1806 levels), and the ejaculate number (2 levels, first or second jump) as fixed effects; the age of the ram at sperm collection in months as a covariate; and the ram permanent environmental effect as random (429 levels), was fitted. Male's individual estimates adjusted by factors included in the model were obtained and used as pseudo-phenotypes (inputs) in the GWAS. BLUPF90 suite (Misztal et al., 2002) was used to run the mixed linear regression analyses. Once pseudo-phenotypes were obtained, a GWAS was carried out in a second step. From the 429 phenotyped rams, only 342 have 50 K SNP Affymetrix microarray genotypes (47 702 SNPs) in the current genotyped Assaf population. From these 342 males genotyped with the 50 K microarray, 47, were also genotyped with the HD Illumina OvineBeadChip (604 317 SNPs). Medium density 50 K and HD genotyping platforms have in common 43 511 SNPs. Imputation of 50 K to HD genotypes was conducted with BEAGLE4.0 software (Browning and Browning, 2007). Only genotypes with an imputation probability higher or equal to 95% were retained. After imputation and filtering by SNPs call rate (0.9), individual call rate (0.9), and MAF (0.001), a total of 540 411 SNPs distributed on the 26 ovine autosomes and X chromosome from 342 rams were included in subsequent analyses. GWAS was conducted with the mixed linear model based association analysis (MLMA) of the genome-wide complex trait analysis software (Yang et al., 2011) for the whole genome and including a genetic relationship matrixto control for therandom effects of genetic similarity and excluding the chromosome on which the candidate SNP is located (leavingone-chromosome-out LOCO) applying the following model: yj¼uþSNPiþgjþej where y j is the pseudo-phenotype for the sperm trait analyzed for the genotyped animal j;uis the overall mean; SNP is the effect of the i SNP (assumed as a covariate coded as 0, 1, or 2, respectively, to Table 1 Basic statistics of sperm traits data collected in the first and second ejaculates from Assaf rams. Trait N Mean SD Min Max CV Volume (ml) 27 635 1.23 0.57 0.1 8 0.46 Volume 1st ejaculate (ml) 21 643 1.31 0.57 0.1 5 0.43 Volume 2nd ejaculate (ml) 5 887 0.91 0.40 0.9 5 0.44 Concentration (×10 6 spz.ml −1 ) 26 414 3 978.64 1 328.49 120 9 701 0.33 Concentration 1st ejaculate (×10 6 spz.ml −1 ) 20 858 4 088.13 1 358.45 120 9 636 0.33 Concentration 2nd ejaculate (×10 6 spz.ml −1 ) 5 453 3 569.43 1 119.33 120 9 701 0.31 Number of spermatozoa (×10 6 ) 26 411 5 191.88 3 165.30 109 37 968 0.61 Number of spermatozoa 1st ejaculate (×10 6 ) 20 857 5 657.60 3 261.03 109 25 170 0.58 Number of spermatozoa 2nd ejaculate (×10 6 ) 5 451 3 433.42 1 950.33 109 37 968 0.57 Motility 25 699 4.82 0.58 1 5 0.12 Motility 1st ejaculate 20 339 4.80 0.62 1 5 0.13 Motility 2nd ejaculate 5 360 4.89 0.42 1 5 0.09 N: number of records; Min: minimum value; Max: maximum value; Spz = spermatozoa number. M. Serrano, M. Ramón, J.H. Calvo et al. Animal xxx (2021) xxx 2 genotypes aa, Aa, and AA); g j is the random additive genetic effect; and e j is the residual error. Chromosome-wise significance association was assessed using a false discovery rate (FDR multi-test correction threshold of 10%). The choice of a threshold value of 10% is justified because this work is mainly exploratory and a first attempt to find variants associated with spermatic characters in sheep, so we decided to be more conservative. Visualization of the association results was performed in Manhattan plots and quantile-quantile plots using R software (R core Team, 2019). Variant Effect Predictor from Ensembl (www.emsembl.org)was used to annotate significant SNPs detected at the genome and chromosome levels. The positional candidate genes were identified in the 250Kb region on both sides of the significant SNPs according to the sheep genome assembly Oar_v3.1 and based on the Ensembl release 81. For functional protein interaction networks, the STRING software was used (Szklarczyk et al., 2019;https://string-db.org/). To examine the frequencies of the SNPs associated with sperm traits detected in the GWAS analysis in an increased number of Assaf animals, we used the whole population currently genotyped in this breed,which consisted of 1503 (1077 males and 426 females) genotyped for the 50 K custom platform and imputed to the HD ovine BeadChip (the 1077 males included the 342 phenotyped rams used in the GWAS analysis). Milk yield EBVs of the 10% top and 10% bottom ranking of these 1503 animals for the significant SNPs associated with sperm traits were studied in order to decipher theputative relationship between milk production efficiency and the frequencies of these variants. Results The mixed-effects linear model of sperm traits All the fixed effects included in the model, the ram flock of origin, the date of sperm collection, the ejaculate number, and the age of ram at the collection showed a significant effect on sperm traits. Fig. 1 shows the estimates and IC95% from the mixed-effects linear model for the date of sperm collection (expressed as month of collection, in Fig. 1), ejaculate number and age of ram in months for the four sperm traits considered. Ejaculates showed larger VOL and SPZ (although CON decrease, in part due to higher VOL) during the reproductive season (from July to next January). A decrease in VOL, CON, and SPZ from the first to the second ejaculate was observed but showing the second ejaculate higher motility than the first one. Regarding the age of ram, although significant, the magnitude of the effect was always very close to 0. Genome-wide association study of sperm traits After the QC of genotype data from 342 rams, a total of 540 411 SNPs distributed on the 26 autosomes and X ovine chromosome were included in subsequent analyses. Supplementary Table S1 shows genome-wide complex trait analysis results for significant SNPs at the chromosome and genome-wide level for the sperm traits studied. Genome-wide 10% FDR threshold for MOT was set on a P-value of 2.4 × 10 −5 . Chromosome-wide 10% FDR thresholds corresponded to P-values ranging from 5.8 × 10 −6 to 1.2 × 10 −4 for MOT; from 2.3 × 10 −5 to 4.5 × 10 −5 for VOL; from 5.7 × 10 −6 to 4.8 × 10 −4 for CON; and from 2.3 × 10 −4 to 1.3 × 10 −4 for SPZ. Twenty-three SNPs distributed in 17 different chromosomes overcame the genome-wide significance level for MOT trait. For VOL, CON, and SPZ, significant associationswerefoundonlyatthechromosomelevelinchromosomes: OAR6 (11 SNPs), OAR18 (6 SNPs), and OAR23 (3 SNPs) for VOL; OAR10 (3 SNPs), OAR13 (1 SNP), OAR16 (71 SNPs), and OAR26 (1 SNP) for CON; and OAR6 (32 SNPs) for SPZ. MLMA and LOCO approaches yielded similar results for all traits analyzed, and only results for the LOCO approach are shown. Fig. 2 shows Manhattan and quantile-quantile plots of the GWAS analyses for the four sperm traits studied. quantile-quantile was plotted to explore that actual distribution of P_values obtained followed the expected distribution and no other effects, such population structure, were missed. Supplementary Table S2 shows allele frequencies and effects of significant SNPs in the 342 rams used for the GWAS and in the current Assaf genotyped population (1503 animals). For VOL, in most cases, the less frequent allele (A1) of the associated SNP showed a negative effect over the trait with allele substitution values (b) ranging between −0.07 and −0.15 and frequencies from 0.06 to 0.49. In some cases, the A1 allele had a low positive effect over VOL (0.07 to 0.09) with an intermediate allele frequencies (from 0.32 to 0.49) (Supplementary Table S2). In general, the A1 alleles showed a negative effect over CON trait with b values from −407 to −179 and frequencies ranging between 0.05 and 0.5. However, also positive effects of the A1 alleles on CON were observed (from 181 to 242) at A1 frequencies between 0.25 and 0.50 (Supplementary Table S2). In most cases, the A1 effect over SPZ trait was highly positive (from 426 and 760) or negative (from −451 to −700) with very similar frequencies (between 0.08 and 0.50) (Supplementary Table S2). Interestingly, the A1 allele exerts a negative substitution effect with magnitudes ranging between −1.62 and −0.24 and frequencies between 0.03 and 0.34, for all significant SNPs detected in the GWA study over the MOT trait (Supplementary Table S2). Supplementary Table S3 shows variant effect prediction results of significant SNPs associated with sperm traits. Table 2 shows putative causalgenes located in the 250Kb region on both sides of thesignificant SNPs for each trait, and the biological processes and molecular functions in which they are involved. Phenotypes, if exist, are also shown. For VOL trait, associations at the chromosome level for 20 SNPs located in OAR6, OAR18, and OAR 23 were found. Most variants were Fig. 1. Estimates (points) and IC95% (segments) from mixed-effects linear model for the date of sperm collection (expressed here as month of collection, Feb to Dec), ejaculate number (2nd ejaculate) and age of ram in months for the four sperm traits considered: sperm volume (VOL; ml, sperm concentration (CON; spermatozoa ×10 6 /ml, sperm motility (MOT; 0–5 scale), and total number of spermatozoa (SPZ; spermatozoa × 10 6 ). Level of significance is represented by asterisks (***: P≤0.001; **: P≤0.01; *: P≤0.05). M. Serrano, M. Ramón, J.H. Calvo et al. Animal xxx (2021) xxx 3 located in introns from genes such as ARHGEF38 (Rho Guanine Nucleotide Exchange Factor 38)andPCDH7 (Protocadherin 7)inOAR6andCABLES1 (Cdk5 Abl enzyme substrate 1) in OAR23. However, two missense variants (oar3_OAR18_30765484, and oar3_OAR18_30789799) with moderate effect (SIFT values 0.41 and 0.12, respectively) were found in the SCAPER (S-phase cyclin A-associated protein in the ER)geneinOAR18,withallele frequencies around 6% in both cases. For CON, 76 SNPs significant associated with the trait were found in OAR10, OAR13, OAR16, and OAR26, being the most significant those found in OAR16 (71 SNPs). Also in this case, most variants were located in introns, upstream and downstream regions of genes such as DLEU7 (Deleted in lymphocytic leukemia 7)inOAR10; OSMR (Oncostatin M receptor), RANBP3L (RAN binding protein 3 like), NADK2 (NAD kinase 2, mitochondrial), SKP2 (S-phase kinase associated protein 2), CAPSL (Calcyphosine like), IL7R (Interleukin 7 receptor), SPEF2 (Sperm flagellar protein 2), and RAI14 (Retinoic acid induced 14) in OAR16; and SORBS2 (Sorbin and SH3 domain containing 2)in OAR26. Two synonymous variants with low effect, oar3_OAR16_ 1971417 and oar3_OAR16_36979589, were detected in DOCK2 (Dedicator of cytokinesis 2)andCPLANE1 (Ciliogenesis and planar polarity effector 1) genes of OAR16 with allele frequencies of 13 and 5.6%, respectively. Fig. 2. Manhattan and Q-Q plots from genome-wide association study (GWAS) of sperm traits in Assaf sheep. Traits considered are ejaculate volume (ml; Figs. A and B); sperm concentration (sperm ×10 6 /ml; Figs. C and D); sperm motility (0–5 scale, Figs. E and F); and total number of spermatozoa (sperm × 10 6 ; Figs. G and H). Dashed line in Manhattan plots (A, C, E, and G) corresponds to average threshold value for a FDR of 10% evaluated at the chromosome level. M. Serrano, M. Ramón, J.H. Calvo et al. Animal xxx (2021) xxx 4 For the trait SPZ, 32 associated SNPs were located in introns, upstream and downstream regions of genes at OAR6. Some of the target genes were CCSER1 (coiled-coil serine rich protein 1), KCNIP4 (potassium voltage-gated channel interacting protein 4), GBA3 (cytosolic beta-glucosidase), STIM2 (stromal interaction molecule 2), and PARM1 (Prostate androgen-regulated mucin-like protein 1). Finally, 23 SNPs associated at the genome-wise level with MOT were located dispersed across the whole genomeatchromosomes 1, 2,3,4, 6, 7, 9, 10, 12, 13, 15, 17, 19, 20, 22, and 26. All SNPs were located in introns, 3′UTR, splice and downstream regions of genes such as SLC9C1 (Solute carrier family 9 member C1) in OAR1; TSN (Translin) in OAR2; GRM8 (Glutamate metabotropic receptor 8) in OAR4; BMP2K (BMP2 inducible kinase)andPAQR3 (Progestin and adipoQ receptor family member 3)inOAR6;NCALD (Neurocalcin delta) in OAR9; DIP2C (Disco interacting protein 2 homolog C)inOAR13;CMIP (c-Maf inducing protein) in OAR14; CUL9 (Cullin 9), CFB (Complement factor B) and C2 (Complement C2)in OAR20; and FUT10 (Fucosyltransferase 10) in OAR 26. Discussion Among the sperm traits collected routinely in AI centers, only mass sperm motility seems to be directly related with overall fertility (Colenbrander et al., 2003; Foote, 2003;Theau-Clément et al., 2011; David et al., 2015). However, sperm volume and concentrationare traits with some interest regarding AI centers efficiency to produce semen doses. Heritability of ovine sperm traits ranges from very low to moderate. For MOT heritability in Lacaune and Assaf breeds takes values of 0.07 and 0.03, respectively (David et al., 2007; Pelayo et al., 2019). Moderate estimates of 0.19 and 0.12 have been found for CON and VOL, respectively, in Assaf breed. Higher estimates of 0.27 and 0.18 for CON and VOL, respectively, were found in Lacaune breed (D avidetal., 2007; Pelayo et al., 2019). Estimates of genetic correlation among these traits are variable across sheep breeds. In Manchega and Assaf dairy breeds, estimates of genetic correlations between VOL and CON were −0.49 and −0.25, respectively; between VOL and MOT, −0.29 and −0.33, respectively; and between CON and MOT, 0.32 and 0.20, respectively (Pelayo et al., 2019). In addition, it has been described that intensive selection in milk production in cattle has been associated with an impairment of the fertilizing capacity (Berry et al., 2009). Therefore, the detection of genes associated with the phenotypes measured in AI centers is key to try to improve the reproductive properties of the sperm of AI rams. These features are more important in phenotypes such as MOT due to their low heritability and its relationships with the fertility of the seminal doses. In the study here conducted, some genomic regions have been associated with sperm quality traits, finding significant SNPs within or close to genes functionally related to reproductive traits in other species. Ejaculate concentration For CON, a strong signal has been detected at chromosome 16. The SNP oar3_OAR16_1971417 is a synonymous variant G > A located at Table 2 Putative causal genes detected in the genome-wide association study for sperm traits in Assaf rams. Trait Chr Gene symbol Gene name Biological process, molecular function, phenotype CON 16 DOCK2 Dedicator of cytokinesis 2 Membrane raft polarization, immune response, chemotaxis, actin cytoskeleton organization 16 CPLANE1 Ciliogenesis and planar polarity effector 1 Transmembrane protein, ubiquitous expression in testis (RPKM 2.7) 16 RANBP3L RAN binding protein 3 like Intracellular transport, benign prostatic hyperplasia 16 NADK2 NAD kinase 2, mitochondrial NAD metabolic process, regulation of systemic arterial blood pressure, calcium-mediated signaling 16 LMBRD2 LMBR1 domain containing 2 Integral component of membrane 16 CAPSL Calcyphosine like Calcium ion binding, high expression in testis (RPKM 3.9) 16 IL7R Interleukin 7 receptor Cell growth, homeostasis of number of cells, severe combined immunodeficiency 16 SPEF2 Sperm flagellar 2 Spermatogenesis, fertilization, sperm motility, immotile short-tail sperm 16 RAI14 Retinoic acid induced 14 Nucleoplasm, mitochondrion, actin cytoskeleton, spermatogenesis MOT 1 SLC9C1 Solute carrier family 9 member C1 Sperm motility, spermatogenesis, high expression in testis (RPKM 3.5) 2TSN Translin DNA-binding protein, very high expression in testis (RPKM 26.5) 4GRM8 Glutamate metabotropic receptor 8 Neuroactive ligand–receptor interaction, glutamatergic synapse 6 PAQR3 Progestin and adipoQ receptor family member 3 Negative regulation of protein phosphorylation, most expressed in testis (RPKM 20.7) 6BMP2K BMP2 Inducible kinase Skeletal development and patterning, protein phosphorylation 9NCALD Neurocalcin delta Neuronal calcium sensor, regulation of systemic arterial blood pressure 13 DIP2C Disco interacting protein 2 homolog C Plasma uric acid levels, AMP binding 14 CMIP c-Maf inducing protein T-cell signaling pathway, in utero embryonic development 20 CUL9 Cullin 9 Microtubule cytoskeleton organization, regulation of mitosis, most expressed in testis (RPKM 6.1) 20 CFB Complement factor B Complement activation, regulation of the immune reaction 20 C2 Complement C2factor B Complement activation, autoimmune diseases, response to nutrient 26 FUT10 Fucosyltransferase 10 Fertilization, protein folding, wound healing VOL 6 ARHGEF38 Rho guanine nucleotide exchange factor 38 Regulation of Rho protein signal transduction, most expressed in prostate (RPKM 3.4) 6TIGD2 Tigger transposable element derived 2 DNA binding 6PCDH7 Protocadherin 7 Cell–cell recognition and adhesion, calcium ion binding 18 SCAPER S-phase cyclin A associated protein in the ER Nucleic acid binding, zinc ion binding, most expressed in testis (RPKM 4.5) 18 PSMA4 Proteasome subunit alpha 4 Core alpha subunit of the 20S proteosome, ubiquitin-dependent protein catabolic process 23 CABLES1 Cdk5 and Abl enzyme substrate 1 Protein binding, regulation of cell cycle, Body Mass Index SPZ 6 CCSER1 Coiled-coil serine rich protein 1 DNA methylation, gastrointestinal microbiome, 6KCNIP4 Potassium voltage-gated channel interacting protein 4 Regulation of potassium ion transmembrane transport 6GBA3 Cytosolic beta-glucosidase Beta-glucoside catabolic process, protein stabilization 6STIM2 stromal interaction molecule 2 Cellular calcium ion homeostasis, activation of store-operated calcium channel activity 6OCIAD1 OCIA domain containing 1 Regulation of stem cell differentiation, protein binding, ovarian cancer 6HOPX HOP homeobox Chaperone-mediated protein assembly, negative regulation of cell differentiation, histone deacetylation 6LOC101110593 RE1-silencing transcription factor-like Negative regulation of cell proliferation, cellular response to electrical stimulus 6PARM1 Prostate androgen-regulated mucin-like protein 1 Positive regulation of telomerase activity Chr = chromosome; CON = sperm concentration; MOT = mass sperm motility; VOL = ejaculate volume; SPZ = number of spermatozoa. M. Serrano, M. Ramón, J.H. Calvo et al. Animal xxx (2021) xxx 5 exon 32 of the DOCK2 gene. Thisgeneencodes a large protein of 180 kDa involved in intracellular signaling networks and highly related to the immune response. DOCK2 deficiency has been related to placental abruption and prostate cancer. Furthermore, Liu et al. (2019) revealed that the methylation of DOCK2 in cattle was associated with a potential role in the regulation of male fertility, suggesting that sperm methylation patterns influence sperm fertility. Then, this SNP could be in linkage disequilibrium with other mutations in the promoter or regulation regions of the gene that could affect CpG islands. The SNP oar3_OAR16_36979589 (rs421824327) is a synonymous variant G > A located at exon 11 of the CPLANE1 (ciliogenesis and planar polarity effector 1) gene which is involved in severalbiological processes such as cilium assembly and protein localization to the ciliary transition zone. Mutations of this gene (also known as Jbts17) produce the Joubert syndrome, Meckel syndrome, and oral-facial-digital syndrome and cause ciliogenesis and ciliary trafficking defects resulting in a decreased of the SHH (Sonic Hedgehog Signaling Molecule) target gene expression (Toriyama et al., 2016). The sonic hedgehog signaling pathway is important in the murine epididymis for the development of sperm motility (Turner et al., 2006). The SNPs oar3_OAR16_38356302 C > A, oar3_OAR16_38429385 A > G, oar3_OAR16_38445552 C > A, and oar3_OAR16_38480186 A > G are intron variants located at the SPEF2 gene. SPEF2 plays an important role in spermatogenesis and flagellar assembly (Sironen et al., 2011) and is expressed in all ciliated cells and required for cilia function (Sironen et al., 2010). Fig. 3 (a) shows the functional protein association network of the SPEF2 gene. This network involves several genes related to sperm cells properties. In humans, variants of SPEF2 gene are associated with many sperm anomalies such as oligozoospermia, short sperm flagellum, and abnormal spermatogenesis and also with infertility (Liu et al., 2020). Probably, these intron variants are not causal mutations affecting CON but should be in linkage disequilibrium with other mutations in coding regions of the SPEF2 gene, not included in the genotyping platform. Four SNPs, oar3_OAR16_39293892 G > A, oar3_OAR16_39321712 G > A, oar3_OAR16_39334360 G > A, and AX-124358876 (rs410614862) T > C, located at introns 1 and 2 of the RAI14 (Retinoic acid induced 14, also named NORPEG) gene, were associated with the variability on CON. Rai14 is an actin binding protein which is expressed in the testis by both Sertoli and germ cells in the seminiferous epithelium. Rai14 regulatesspermatidpolarityandspermatid transportduringspermiogenesis (Qian et al., 2013) and is also involved in the reorganization of actin filaments in Sertoli cells during the epithelial cycle, participating in conferring spermatid cell adhesion in the testis (Qian et al., 2013b). Ejaculate volume Two missense variants in OAR18, oar3_OAR18_30765484 G > A and oar3_OAR18_30789799 A > G located in exons 24 and 23 of the SCAPER gene, respectively, produce and aminoacid change of Threonine to Isoleucine and Phenylalanineto Leucine, respectively, that are predicted as tolerated changes with SIFT values of 0.41 and 0.12, respectively. The SCAPER (S phase cyclin A-associated protein in the endoplasmic reticulum) gene encodes a transmembrane transport protein which is highly expressed in testicle (ProteomicsDB, https://www.proteomicsdb.org/). In humans, SCAPER has been associated with retinitis pigmentosa (Tatour et al., 2017) and also with hypogonadism (a decreased functionalityofthegonad)and abnormal testismorphology (Human Phenotype Ontology, https://www.genecards.org/cgi-bin/carddisp.pl?gene= SCAPER). Fig. 3 (b) shows the functional protein association network of the SCAPER gene. The SNPs oar3_OAR18_29981728 A > C (rs415094151) and oar3_OAR18_29985155 G > A are downstream gene variants of the PSMA4 (Proteasome 20S subunit alpha 4) gene located at OAR18. PSMA4 is a component of the 20S core proteasome complex involved in the proteolytic degradation of most intracellular proteins. The 20S proteasome mediates ubiquitin-independent protein degradation, which is required in several pathways including spermatogenesis (20S-PA200 complex). Many proteins such asmeiotic proteins, core histones, and unnecessary organelles are degraded during spermatogenesis. It has been demonstrated that the expression of PSMA8 (Proteasome 20S Subunit Alpha 8), which also be part of the 20S proteasome, in spermatocytes, is essential for mice male fertility (Zhang et al., 2019). However, female fertility does not require PSMA8 (Zhang et al., 2019). Although there are no data about PMSA4 involvement in males' fertility, it should share some of the properties described for PMSA8, because it's high expression in testis and role in the 20S proteasome. The TET2 (Tet methylcytosine dioxygenase 2) gene is located 332 bp downstream the oar3_OAR6_19723487 at OAR6 (ARHGEF38). The protein encoded by this gene is a methylcytosine dioxygenase that catalyzes the conversion of methylcytosine to 5-hydroxymethylcytosine and plays a key role in active DNA demethylation. Expression levels of this gene during human spermatogenesis are pivotal for male fertility (Ni et al., 2016). Authors showed that levels of mRNA isoforms of the TET genes in spermatozoa associated with sperm parameters and were significantly reduced in subfertile men. Ejaculate number of spermatozoa For the ejaculate number of SPZ all significant associated SNPs were located at OAR6. The SNP oar3_OAR6_89628167 G > A is an intron variant located at PARM1 (Prostate androgen-regulated mucin-like protein 1) gene. Parm1 is a highly glycosylated, mucin-like type 1 transmembrane protein involved in the positive regulation of telomerase activity, in the survival of prostate cells (Cornet et al., 2003), and contributes to ovulation and/or luteal function by acting as a regulator of progesterone metabolism (Park et al., 2013). In cattle, a SNP (rs111027720) located in thePARM1 gene has been associated with the development of cleaved embryos to the blastocyst stage in “in vitro”fertilization procedures Fig. 3. Functional protein association networks of SPEF2 (a) SCAPER (b) and SLC9C1 (c) genes. M. Serrano, M. Ramón, J.H. Calvo et al. Animal xxx (2021) xxx 6 (Cochran et al., 2013a). The same authors found an SNP in the PARM1 gene associated with some fertility traits such as heifers' conception rate, daughter pregnancy rate, and bull productive life in Holstein cattle (Cochran et al., 2013b). Three SNPs, oar3_OAR6_72193906 A > G, oar3_OAR6_72207209 G > A, and oar3_OAR6_72211624 A > G located at intron, upstream and downstream the LOC101110593 (RE1-silencing transcription factor-like) gene, were positively associated with SPZ. LOC101110593 is involved in whole organism decreased fertility and embryo development reduced rate (Thakore-Shah et al., 2015). Spermatozoa mass motility For MOT trait, SNPs associations were found dispersed across the whole genome. The SNP oar3_OAR1_175397925 C > A at chromosome 1 is an intron 6 variant of the SLC9C1 (Solute carrier family 9 member C1, also named Sperm-Specific Na(+)/H(+) Exchanger and Sperm-NHE) gene. SLC9C1 is a member of the sodium-hydrogen exchanger (NHE) family and is required for male fertility and sperm motility (Wang et al., 2003). Sperm-specific sodium/hydrogen exchanger is involved in intracellular pH regulation of spermatozoa and required for sperm motility and fertility. Also has a role in sperm cell hyper activation, a critical step for sperm motility essential in the preparation of sperm for fertilization. SLC9C1 is an integral component of the plasma membrane as a channel transport of ions, glucose, other sugars, salts, organic acids, metal ions, and amine compounds. Fig. 3 (c) shows the functional protein association network of the SLC9C1 gene. The ADCY10 (Adenylate cyclase type 10, Epididymis Secretory Sperm Binding Protein Li 7a)gene,a predicted functional partner, interacts and is co-expressed with SLC9C1. ADCY10 (also named Testicular Soluble Adenylyl Cyclase SAC)playsacritical role in mammalian spermatogenesis by producing the cAMP which regulates cAMP-responsive nuclear factors indispensable for sperm maturation in the epididymis. Genetic approaches have demonstrated that Adcy10 is necessary for male fertility and specifically for sperm motility and capacitation (Xie et al., 2006). The variant oar3_OAR2_185913749 G > A is a 3′-UTR located at exon 6oftheTSN (Translin, also named Testis Brain-RNA Binding Protein) gene. The 3′-UTR region plays a crucial role in gene expression by influencing the localization, stability, export, and translation efficiency of an mRNA. The 3′-UTR can influence polyadenylation, translation efficiency, localization, and stability of the mRNA since it contains both binding sites for regulatory proteins as well as microRNAs. This gene is highly expressed in brain and testis and is involved in DNA damage repair and in mRNA transport (Cho et al., 2004; Wang et al., 2004). The protein Translin is involved in translational regulation during spermatogenesis, and binds to specific mRNAs in the testis, forming an RNP complex (Morales et al., 2002). Mice lacking the TB RBP gene (the TSN mouse orthologue) can sire offspring, but have reduced sperm production (Chennathukuzhi et al., 2003). The SNP oar3_OAR26_27800594 C > A SNP is an intron 2 variantof the FUT10 (Fucosyltransferase 10)gene.FUT10 encodes a fucosyltransferase protein with roles in protein folding, glycosylation, nervous system development, and fertilization. FUT10 is highly expressed in testis (6.67 TPM) and in cattle has been associated with the ejaculate volume (Qin et al., 2016). Despite the moderate positivecorrelation existing betweenCON and MOT in Assaf sheep breed (0.20), any common gene or genomic region was found in the GWAS for both traits. Relationship of candidate genes frequencies and estimated breeding values for milk yield Relationship between sperm traits studied and milk yield was explored. For that, 1503 animals with genotypes mentioned above were ranked based on their EBVs for milk yield (personal communication, Assaf sheep breeding program). Then, top and bottom 10% animals were chosen, and allele frequencies for the most significant SNPs found in this work were compared between both, top and bottom animals. Table 3 shows allele frequencies of the most significant SNPs found for sperm quality traits in the 10% top and 10% bottom ranking of EBVs for milk yield in 1503 Assaf animals, and thedifference between them in both groups (A1 EBVnegative-A1 EBVpositive). In general for all traits, frequencies of the A1 alleles are very similar in both groups of animals. However, two SNPs for CON in the RAI14 gene with negative effect over the trait and 2 SNPs for VOL in the PSMA4 gene with positive effect over the trait showed significant higher A1 frequencies in the group of animals with negative EBVs for milk yield. For the ejaculate number of SPZ the 3 SNPs located in the LOC101110593 gene, which exert a positive effect over the trait, have higher A1 Table 3 Allele frequencies of most significant single-nucleotide polymorphisms for sperm quality traits in the top and bottom 10% ranking for the milk yield estimated breeding values (EBV) of 1 503 Assaf animals genotyped with the 50 K Affymetrix custom chip and imputed for the HD Illumina BeadChip. EBV milk yield negative EBV milk yield positive Marker ID A1 > A2 Gene b Trait N freq A1 freq A2 N freq A1 freq A2 Diff oar3_OAR16_36979589 G > A CPLANE1 −407.02 CON 75 0.13 0.87 75 0.11 0.89 0.02 oar3_OAR16_1971417 G > A DOCK2 −269.76 CON 75 0.21 0.79 75 0.13 0.87 0.08 oar3_OAR16_39293892 G > A RAI14 −255.00 CON 75 0.34 0.66 75 0.29 0.71 0.05 oar3_OAR16_39321712 G > A RAI14 −207.40 CON 75 0.37 0.63 75 0.29 0.71 0.07 oar3_OAR16_39334360 G > A RAI14 −229.23 CON 75 0.41 0.59 75 0.26 0.74 0.15 AX-124358876 T > C RAI14 −224.85 CON 75 0.46 0.54 75 0.30 0.70 0.16 oar3_OAR16_38429385 A > G SPEF2 −290.53 CON 75 0.23 0.77 75 0.21 0.79 0.02 oar3_OAR16_38480186 A > G SPEF2 −350.15 CON 75 0.25 0.75 75 0.21 0.79 0.05 oar3_OAR16_38356302 C > A SPEF2 −206.52 CON 75 0.37 0.63 75 0.31 0.69 0.06 oar3_OAR16_38445552 C > A SPEF2 −188.11 CON 75 0.42 0.58 75 0.47 0.53 −0.05 AX-169032210 A > G CFB/C2 −0.63 MOT 75 0.06 0.94 75 0.06 0.94 0.00 oar3_OAR26_27800594 C > A FUT10 −0.73 MOT 75 0.18 0.82 75 0.22 0.78 −0.04 oar3_OAR1_175397925 C > A SLC9C1 −1.36 MOT 75 0.06 0.94 75 0.02 0.98 0.04 oar3_OAR2_185913749 G > A TSN −0.94 MOT 75 0.08 0.92 75 0.06 0.94 0.02 oar3_OAR6_72193906 A > G LOC101110593 659.10 SPZ 75 0.13 0.87 75 0.16 0.84 −0.03 oar3_OAR6_72211624 A > G LOC101110593 676.75 SPZ 75 0.06 0.94 75 0.13 0.87 −0.07 oar3_OAR6_72207209 G > A LOC101110593 647.38 SPZ 75 0.11 0.89 75 0.16 0.84 −0.05 oar3_OAR6_89628167 G > A PARM1 −538.34 SPZ 75 0.33 0.67 75 0.39 0.61 −0.06 oar3_OAR18_29981728 A > C PSMA4 0.07 VOL 75 0.51 0.49 75 0.39 0.61 0.13 oar3_OAR18_29985155 G > A PSMA4 0.08 VOL 75 0.56 0.44 75 0.43 0.57 0.13 oar3_OAR18_30789799 A > G SCAPER −0.15 VOL 75 0.10 0.90 75 0.04 0.96 0.06 oar3_OAR18_30765484 G > A SCAPER −0.15 VOL 75 0.11 0.89 75 0.07 0.93 0.04 A1 = less frequent allele; b = allele substitution effect; N = number of records; Diff = frequency A1 in the EBV negative group –frequency A1 in the EBV-positive group. Trait = sperm trait (CON: sperm concentration; MOT: mass sperm motility; SPZ: number of spermatozoa; VOL: ejaculate volume). M. Serrano, M. Ramón, J.H. Calvo et al. Animal xxx (2021) xxx 7 frequencies in the top EBVs group of animals, but also that located in the PARM1 gene with negative effect over SPZ. Despite the small number of animals analyzed, it seems that these SNPs are not linked with genes involved in milk production. However, the high frequencies found for some SNPs with negative effect over sperm traits in the top milk yield EBVs group, would allow exerting some selection pressure over them to improve sperm characteristics in the high merit animals for milk production. Most SNPs that were significantly associated with some sperm trait belongs exclusively to the HD ovine Illumina BeadCheap. This fact indicates that high-density SNP platforms are more suitable to detect genes or genomic regions related to animal characteristics, at least in this sheep breed. In general, sheep breeds show a low linkage disequilibrium than other livestock species, such as cattle and pigs, among markers from low density genotyping chips, asthe 50K Affymetrix platform. Some of thegenes here detected in association with sperm quality traits could be used in marker assisted selection of rams to be used in AI centers for progeny test in dairy breeding programs. In summary, the GWA studies conducted in this work have revealed the existence of genomic regions and some putative causal genes associated with sperm quality traits. Among them, SPEF2 for ejaculate concentration, SCAPER and PSMA4 for ejaculate volume, PARM1 for ejaculate number of spermatozoa, and SLC9C1 and FUT10 for sperm mass motility seem to be good candidate genes to improve sperm quality traits in sheep. More studies increasing the number of genotyped and phenotyped animals will be necessary to validate the magnitude of the associations here detected and to establish the possibility of using these genes to improve rams’sperm traits. Supplementary materials Supplementary data to this article can be found online at https://doi. org/10.1016/j.animal.2020.100065. Ethics approval Not applicable. Data and model availability statement OVIGEN AI center is the owner of phenotypic data and genotypes from Assaf animals. None of the data were deposited in an officialrepository but available upon requests which must be accompanied by a description of what the data will be used for. Author ORCIDs Malena Serrano ORCID 0000–0002–1621-3102. Author contributions Malena Serrano: Conceptualization, formal analysis, investigation, resources, writing –original draft, writing –review & Editing. Manuel Ramón: Formal analysis, Writing –Original Draft.Jorge H Calvo: writing –review & editing, visualization. Mª Ángeles Jiménez: Data curation. Fernando Freire: Funding acquisition, resources. José Manuel Vázquez: Resources. Juan José Arranz: Writing –review & editing. Declaration of interest None. Acknowledgements We thanks ASSAF.E breeders association and OVIGEN to provide phenotype and genotype data of AI rams. Financial support statement This work was supported by the European Association for Innovation project 001958/G02 (FEADER), the contract INIA-GENOVIS Con19-043 and the National project RTA2013-00041-C02-01. References Berry, D.P., Friggens, N.C., Lucy, M., Roche,J.R., 2009. Milk production and fertility in cattle. Annual Review of Animal Biosciences 4, 269–290. Broekhuijse, M.L.W.J., Šoštarić, E., Feitsma, H., Gadella, B.M., 2012. Application of computer-assisted semen analysis to explain variations in pig fertility. Journal of Animal Science 90 (3), 779–789. https://doi.org/10.2527/jas.2011-4311 Epub 2011 Nov 7. Browning, S.R., Browning, B.L., 2007. Rapid and accurate haplotype phasing and missing data inference for whole genome association studies by use of localized haplotype clustering. American Journal of Human Genetics 81, 1084–1097. Chennathukuzhi, V., Stein, J.M., Abel, T., Donlon, S., Yang, S., Miller, J.P., Allman, D.M., Simmons, R.A., Hecht, N.B., 2003. Mice deficient for testis-brain RNA-binding protein exhibit a coordinate loss of TRAX, reduced fertility, altered gene expression in the brain and behavioural changes. Molecular and Cellular Biology 23, 6419–6434. Cho, Y.S., Chennathukuzhi, V.M., Handel, M.A., Eppig, J., Hecht, N.B., 2004. The relative levels of Translin-associated Factor X (TRAX) and testis brain RNA-binding protein determine their nucleocytoplasmic distribution in male germ cells. The Journal of Biological Chemistry 279, 31514–31523. Cochran, S.D., Cole, J.B., Null, D.J., Hansen, P.J., 2013a. Single nucleotide polymorphisms in candidate genes associated with fertilizing ability of sperm and subsequent embryonic development in cattle. Biology of Reproduction 89, 69. Cochran, S.D., Cole, J.B., Null, D.J., Hansen, P.J., 2013b. Discovery of single nucleotide polymorphisms in candidate genes associated with fertility and production traits in Holstein cattle. BMC Genetics 14, 49. Colenbrander, B., Gadella, B.M., Stout, T.A., 2003. The predictive value of semen analysis in the evaluation of stallion fertility. Reproduction in Domestic Animals 38, 305–311. R Core Team, 2019. R: A language and environment for statistical computing. R foundation for statistical computing. Austria. URL, Vienna https://www.R-project.org/. Cornet, A.M., Hanon, E., Reiter, E.R., Bruyninx, M., Nguyen, V.H., Hennuy, B.R., Hennen, G. P., Closset, J.L., 2003. Prostatic androgen repressed message-1 (PARM-1) may play a role in prostatic cell immortalisation. Prostate 56, 220–230. David, I., Druart, X., Lagrifoul, G., Manfredi, E., Robert-Granié, C., Bodin, 2007. Genetic and environmental effects on semen traits in Lacaune and Manech tête rousse AI rams. Genetics, Selection, Evolution 39, 405–419. David, I., Kohnke, P., Lagrifoul, G., Praud, O., Plouarboué, F., Degond, P., Druart, X., 2015. Mass sperm motility is associated with fertility in sheep. Animal Reproduction Science 161, 75–81. Diniz, D.B., Lopes, M.S., Broekhuijse, M.L., Lopes, P.S., Harliziu, B., Guimarães, S.E., Duijvesteijn, N., Knol, E.F., Silva, F.F., 2014. A genome-wide association study reveals a novel candidategene for sperm motility in pigs. Animal Reproduction Science 151, 201–207. Evans, G., WMC, Maxwell, 1987. Salamon’s artificial insemination of sheep and goats. Butterworths, Sydney, Australia. Fonseca, P.A.S., Dos Santos, F.C., Lam, S., Suárez-Vega, A., Miglior, F., Schenkel, F.S., Diniz, L. A.F., Id-Lahoucine, S., Carvalho, M.R.S., Cánovas, A., 2018. Genetic mechanisms underlying spermatic and testicular traits within and among cattle breeds: systematic review and prioritization of GWAS results. Journal of Animal Science 96, 4978–4999. Foote, R.H., 2003. Fertility estimation: A review of past experience and future prospects. Animal Reproduction Science 75, 119–139. Gadea, J., 2005. Sperm factors related to in vitro and in vivo porcine fertility. Theriogenology 63, 431–444. Hering, D.M., Oleński, K., Ruść, A., Kaminski, S., 2014. Genome-wide association study for semen volume and total number of sperm in Holstein-Friesian Bulls. Animal Reproduction Science 151, 126–130. Liu, S., Chen, S., Cai, W., Yin, H., Liu, A., Li, Y., Liu, G.E., Wang, Y., Yu, Y., Zhang, S., 2019. Divergence analyses of sperm DNA methylomes between monozygotic Twin AI Bulls. Epigenomes 3, 21. Liu, C., Lv, M., Xiaojin, H., Yong, Z., Amir, A.Y., Weiyu, L., Huan, W., Zine Eddine, K., Wangjie, L., Jingjing, Z., Qing, T., Shuyan, T., Yong Jun, Z., Yading, Z., Caihua, L., Shixiong, T., Zhiguo, Z., Li, J., Pierre, R., Feng, Z., Yunxia, C., 2020. Homozygous mutations in SPEF2 induce multiple morphological abnormalities of the sperm flagella and male infertility. Journal of Medical Genetics 57, 31–37. Marques, D.B.D., Bastiaansen, J.W.M., Broekhuijse, M.L.W.J., Lopes, M.S., Knol, E.F., Harlizius, B., Guimarães, S.E.F., Silva, F.F., Lopes, P.S., 2018. Weighted single-step GWAS and gene network analysis reveal new candidate genes for semen traits in pigs. Genetics, Selection, Evolution 50, 40. Misztal, I., Tsuruta, S., Strabel, T., Auvray, B., Druet, T., Lee, D.H., 2002. BLUPF90 and related programs (BGF90). 7th world congress on genetics applied to livestock production 2002 Montpellier, France. Morales, C.R., Lefrancois, S., Chennathukuzhi, V., El-Alfy, M., Wu, X., Yang, J., Gerton, G.L., Hecht, N.B., 2002. A TB-RBP and Ter ATPase complex accompanies specific mRNAs from nuclei through the nuclear pores and into intercellular bridges in mouse male germ cells. Developmental Biology 246, 480–494. Ni, K., Dansranjavin, T., Rogenhofer, N., Oeztuerk, N., Deuker, J., Bergmann, M., Schuppe, H. C., Wagenlehner, F., Weidner, W., Steger, K., Schagdarsurengin, U., 2016. TET enzymes are successively expressed during human spermatogenesis and their expression level is pivotal for male fertility. Human Reproduction 31, 1411–1424. M. Serrano, M. Ramón, J.H. Calvo et al. Animal xxx (2021) xxx 8 Park, J.Y., Jang, H., Curry, T.E., Sakamoto, A., Jo, M., 2013. Prostate androgen-regulated mucin-like protein 1: a novel regulator of progesterone metabolism. Molecular Endocrinology 27, 1871–1886. Pelayo, R., Ramón, M., Granado-Tajada, I., Ugarte, E., Serrano, M., Gutiérrez-Gil, B., Arranz, J.J., 2019. Estimation of the genetic parameters for semen traits in spanish dairy sheep. Animals 9, 1147. Qian, X., Mruk, D.D., Cheng, Y., Cheng, C.Y., 2013. RAI14 (retinoic acid induced protein 14) is an F-actin regulator. Lesson from the testis. Spermatogenesis 3 e24824. Qian, X., Mruk, D.D., Cheng, C.Y., 2013b. Rai14 (Retinoic Acid Induced Protein 14) is involved in regulating f-actin dynamics at the ectoplasmic specialization in the rat testis. PLoS One 8 e60656. Qin, C., Yin, H., Zhang, X., Sun, D., Zhang, Q., Liu, J., Ding, X., Zhang, Y., Zhang, S., 2016. Genome-wide association study for semen traits of the bulls in Chinese Holstein.Animal Genetics 48, 80–84. Qin, C., Yin, H., Zhang, X., Sun, D., Zhang, Q., Liu, J., Ding, X., Zhang, Y., Zhang, S., 2017. Genome-wide association study for semen traits of the bulls in Chinese Holstein.Animal Genetics 48, 80–84. Sironen, A., Hansen, J., Thomsen, B., Andersson, M., Vilkki, J., Toppari, J., Kotaja, N., 2010. Expression of SPEF2 during mouse spermatogenesis and identification of IFT20 as an interacting protein. Biology of Reproduction 82, 580–590. Sironen, A., Kotaja, N., Mulhern, H., Wyatt, T.A., Sisson, J.H., Pavlik, J.A., Miiluniemi, M., Fleming, M.D., Lee, L., 2011. Loss of SPEF2 function in mice results in spermatogenesis defects and primary ciliary dyskinesia. Biology of Reproduction 85, 690–701. Suchocki, T., Szyda, J., 2014. Genome-wide association study for semen production traits in Holstein-Friesian Bulls. Journal of Dairy Science 98, 5774–5780. Szklarczyk, D., Gable, A.L., Lyon, D., Junge, A., Wyder, S., Huerta-Cepas, J., Simonovic, M., Doncheva, N.T., Morris, J.H., Bork, P., Jensen, L.J., von Mering, C., 2019. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Research 47, 607–613. Tatour, Y., Sanchez-Navarro, I., Chervinsky, E., Hakonarson, H., Gawi, H., Tahsin-Swafiri, S., Leibu, R., Lopez-Molina, M.I., Fernandez-Sanz, G., Ayuso, C., Ben-Yosef, T., 2017. Mutations in SCAPER cause autosomal recessive retinitis pigmentosa with intellectual disability. Journal of Medical Genetics 54, 698–704. Thakore-Shah, K., Koleilat, T., Jan, M., John, A., Pyle, A.D., 2015. REST/NRSF knockdown alters survival, lineage differentiation and signaling in human embryonic stem cells. PLoS One 10, e0145280. Theau-Clément, M., Ailloud, E., Sanchez, A., Duzert, R., Saleil, G., Brun, J.M., 2011. Relation entre les caractéristiques de la semence de lapin et sa fécondance. 14èmes Journées de la Recherche Cunicole. Le Mans, France, pp. 73–76. Toriyama, M., Lee, C., Taylor, S.P., Duran, I., Cohn, D.H., Bruel, A.L., Tabler, J.M., Drew, K., Kelly, M.R., Kim, S., Park, T.J., Braun, D.A., Pierquin, G., Biver, A., Wagner, K., Malfroot, A., Panigrahi, I., Franco, B., Al-Lami, H.A., Yeung, Y., Choi, Y.J., Duffourd, Y., Faivre, L., Rivière, J.B., Chen, J., Liu, K.J., Marcotte, E.M., Hildebrandt, F., ThauvinRobinet, C., Krakow, D., Jackson, P.K., Wallingford, J.B., 2016. The ciliopathy-associated CPLANE proteins direct basal body recruitment of intraflagellar transport machinery. Nature Genetics 48, 648–656. Turner, T.T., Bang, H.J., Attipoe, S.A., Johnston, D.S., Tomsig, J.L., 2006. Sonic hedgehog pathway inhibition alters epididymal function as assessed by the development of sperm motility. Journal of Andrology 27, 225–232. Wang, D., King, S., Quill, T., Doolittle, L.K., Garbers, D.L., 2003. A new sperm-specificNa+/ H+ exchanger required for sperm motility and fertility. Nature Cell Biology 5, 1117–1122. Wang, J., Boja, E.S., Oubrahim, H., Chock, P.B., 2004. Testis brain ribonucleic acid-binding protein/translin possesses both single-stranded and double-stranded ribonuclease activities. Biochemistry 43, 13424–13431. Wang, K., Kang, Z., Jiang, E., Yan, H., Zhu, H., Liu, J., Qu, L., Lan, X., Pan, C., 2020. Genetic effects of DSCAML1 identified in genome-wide association study revealing strong associations with litter size and semen quality in goat (Capra hircus). Theriogenology 146, 20–25. Xie, F., Garcia, M.A., Carlson, A.E., Schuh, S.M., Babcock, D.F., Jaiswal, B.S., Gossen, J.A., Esposito, G., van Duin, M., Conti, M., 2006. Soluble adenylyl cyclase (sAC) is indispensable for sperm function and fertilization. Developmental Biology 296 (353–36). Yang, J., Lee, S.H., Goddard, M.E., Visscher, P.M., 2011. GCTA: a tool for Genome-wide Complex Trait Analysis. Am J Hum Genet. 88 (1), 76–82. Yin, H., Zhou, C., Shi, S., Fang, L., Liu, J., Sun, D., Jiang, L., Zhang, S., 2019. Weighted singlestep genome-wide association study of semen traits in holstein bulls of China. Frontiers in Genetics 10, 1053. Zhang, Q., Ji, S.Y., Busayavalasa, K., Shao, J., 2019. YuC. Meiosis I progression in spermatogenesis requires a type of testis-specific 20S core proteasome. Nature Communications 10, 3387. M. Serrano, M. Ramón, J.H. Calvo et al. Animal xxx (2021) xxx 9