1 2 DR THAISE MELO (Orcid ID : 0000-0003-0983-4602) 3 DR MARINA R. S. FORTES (Orcid ID : 0000-0002-7254-1960) 4 5 6 Article type : Original Article 7 8 9 Short Running Title: Across-breed QTL validation for sexual precocity in tropical cattle 10 Title: ACROSS-BREED VALIDATION STUDY CONFIRMS AND IDENTIFIES 11 NEW LOCI ASSOCIATED WITH SEXUAL PRECOCITY IN BRAHMAN AND 12 NELLORE CATTLE1 13 Thaise Pinto de Melo*, Marina Rufino Salinas Fortes†‡, Ben Hayes‡, Lucia Galvão de 14 Albuquerque*§, Roberto Carvalheiro*§ 15 16 *Department of Animal Science, School of Agricultural and Veterinarian Sciences, 17 FCAV/ UNESP - Sao Paulo State University, Jaboticabal, Sao Paulo, 14884-900, 18 Brazil. 19 †The University of Queensland, School of Chemistry and Molecular Biosciences, St 20 Lucia, Queensland 4072, Australia. 21 ‡The University of Queensland, Queensland Alliance for Agriculture and Food 22 Innovation, St Lucia, Queensland 4072, Australia. 23 §National Council for Scientific and Technological Development (CNPq), Brasília, 24 Distrito Federal, Brazil. 25 Corresponding author: Roberto Carvalheiro, School of Agricultural and Veterinarian 26 Sciences, FCAV/ UNESP - Sao Paulo State University, Jaboticabal, Brazil. Email: 27 [email protected] Manuscript 28 This is the author manuscript accepted for publication and has undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/JBG.12429 This article is protected by copyright. All rights reserved 29 ABSTRACT: The aim of this study was to identify candidate regions associated with 30 sexual precocity in Bos indicus. Nellore and Brahman were set as validation and 31 discovery populations, respectively. SNP selected in Brahman to validate in Nellore 32 were from: gene regions affecting reproductive traits (G1), and significant SNP (P ≤ 10- 33 3) from a meta-analysis (G2). In the validation population early pregnancy (EP) and 34 scrotal circumference (SC) were evaluated. To perform GWAS in validation population 35 we used regression and Bayes C. SNP with P ≤ 10-3 in regression and Bayes Factor ≥ 3 36 in Bayes C were deemed significant. Significant SNP (for EP or SC) or SNP in their ± 37 250 Kb vicinity region, which were in at least one discovery set (G1 or G2) were 38 considered validated. SNP identified in both G1 and G2 were considered candidate. For 39 EP 145 SNP were validated in G1 and 41 in G2, for SC these numbers were 14 and 2. 40 For EP 21 candidate SNP were detected (G1 and G2). For SC no candidate SNP were 41 identified. Validated SNP and their vicinity region were located close to QTL or genes 42 related to reproductive traits and were enriched in gene ontology terms related with 43 reproductive success. These are therefore, strong candidate regions for sexual precocity 44 in Nellore and Brahman. 45 Keywords: Bos indicus, discovery population, reproductive traits, SNP validation, 46 tropical beef cattle 47 INTRODUCTION 48 49 Across-breed validation studies are commonly used to validate quantitative trait 50 loci (QTL) for several traits. In Genome-wide association studies (GWAS) those 51 significant markers under an empirical P-value for the same or correlated traits in 52 different breeds are likely tagging QTL that segregate across-breeds. These QTLs may 53 harbor some important genes affecting both populations. Karlsson et al. (2007) used this 54 approach to validate single nucleotide polymorphisms (SNP) markers in different dog 55 breeds. They notice that this strategy was highly efficient to fine-mapping across 56 breeds. 57 The probability of finding common QTL for correlated traits in breeds that share 58 common ancestry is expected to be higher than in breeds with very distinct genetic Author Manuscript 59 origin. This is because breeds that do not share a recent common ancestor are more 60 genetically distinct from each other. Genetic differences increase with distance to 61 common ancestors. Distinct genetic origin affects the linkage disequilibrium pattern 62 between SNPs at long-ranges, and by consequence it affects QTL mapping (Goddard & This article is protected by copyright. All rights reserved 63 Hayes, 2009). As Nellore and Brahman are both Bos indicus breeds, and Brahman was 64 originally developed by three base breeds, Gir, Guzerat and Nellore (Briggs & Briggs, 65 1980), the likelihood of both breeds sharing QTL controlling correlated traits is higher 66 than in unrelated breeds. 67 Several strategies have been used to conduct across-breed validation studies. 68 Pryce et al. (2010) used two dairy breeds to validate QTL for milk production and 69 fertility traits. They distributed Holstein bulls in a discovery population and younger 70 Holstein bulls and Jersey bulls in a validation population. SNP that were detected as 71 significant at an empirical threshold P-value in discovery and validation populations 72 were considered validated. Also validating fertility traits in dairy cattle, Höglund et al. 73 (2014) used three breeds to validate genomic associations. They used one breed as 74 discovery population and the other two breeds as validation populations. They argued 75 that using two populations simultaneously to validate significant associations is a 76 powerful strategy to decrease the risk of false positive association. 77 Genic regions are strong candidate regions to present QTL segregating across 78 related breeds because it is expected that the metabolic pathways in which these genes 79 are involved are conserved across breeds. Also, regions with pleiotropic effect across 80 related traits could result in higher number of true positive validated associations across 81 breeds, because genes that are controlling multiple traits in a breed might preserve 82 similar pattern of pleiotropic effect in another related breed (Saatchi et al., 2014). 83 The aim of this study was to validate in a Nellore population genomic regions 84 associated with sexual precocity that were reported for Brahman. We used as discovery 85 data two SNP sets pre-selected in a Brahman population, 1) from gene regions 86 previously reported as significant for reproductive traits, and 2) from significant 87 associations detected in a meta-analysis study of sexual precocity traits. 88 89 MATERIAL AND METHODS 90 Ethics Statement 91 All managements and procedures involving production, maintenance and use of 92 Nellore animals were certified and approved by the National Council of Animal Author Manuscript 93 Experimentation Control (CONCEA, 2008) and Use Committee at University of Sao 94 Paulo, Jaboticabal Campus (18.340/16). Regarding Brahman animals, Animal care and 95 Use committee approval was not required because the data is from existing databases 96 described in the following section. This article is protected by copyright. All rights reserved 97 98 Discovery population 99 The discovery population was composed by Brahman animals. Phenotypes were 100 provided by Cooperative Research Centre for Beef Genetic Technologies (Beef CRC). 101 Brahman phenotypes included the female traits age when the first corpus luteum (CL) 102 was observed (AGECL), first postpartum anoestrus interval (PPAI), ability to ovulate 103 prior to weaning the calf (PW), and the male traits scrotal circumference (SC) measured 104 at 12, 18 and 24 months of age (SC12, SC18, SC24). 105 The AGECL was defined as the number of days from the heifer birth to the first 106 CL detected. PPAI, measured in days, was calculated as the difference between the 107 calving date and the date of the first observed ovulation postpartum. PW, a binary trait 108 was defined as 0 for females that had success to ovulate before weaning her calf or 1 for 109 those females that failed. For all female traits ovarian ultrasounds were carried out to 110 verify the presence of CL that is an indicator of the ovulation, at every 4 to 6 weeks 111 after heifers achieved 200 Kg of weight. 112 Scrotal circumference was measured in cm, with a standard metal tape. For 113 females and males, contemporary groups (CG) were defined by the concatenation year 114 of birth and management group information (defined as cohorts). The age of young 115 bulls at recording was considered a covariate for SC. Details about animals, cohorts and 116 phenotypes are described in Johnston et al. (2009), Johnston et al. (2010), Burns et al. 117 (2013), Corbet et al. (2013) and Fortes et al. (2018). 118 Animals were genotyped with the Illumina BovineSNP50 V1 and V2. Genotypes 119 were imputed for high-density panel using Beagle software v.3.2 (Browning & 120 Browning, 2009) and a reference population of representative animals of the Beef CRC 121 population genotyped using the high-density Illumina Bovine HD Assay (Illumina, San 122 Diego, CA, USA), as described by Fortes et al. (2013a). Quality control excluded 123 samples with call rate < 98%, SNP in non-autosomal regions, with call rate < 85% and 124 minor allele frequency (MAF) < 0.02. The number of SNP after quality control was 125 625,041 for females and 612,992 for males. Details about genotypes and imputation are 126 described in Fortes et al. (2013a). Author Manuscript 127 128 Validation population 129 Data from Nellore animals were used as validation population. Phenotypic 130 information was obtained from Alliance Nellore dataset. The animals considered in this This article is protected by copyright. All rights reserved 131 study were born in eight farms distributed over Midwest, Southeast and Northeast of 132 Brazil. In general, two breeding seasons are applied during the year, where the females 133 are either artificially inseminated or naturally mated. The heifers are exposed in the 134 early breeding season at around 16 months of age. After 60 days of the early breeding 135 season, pregnancy is confirmed and those females that failed in conceiving in the first 136 breeding season had a second opportunity at around 2 years old.