Porcine single nucleotide polymorphisms and their functional effect: an update

Objective

To aid in the development of a comprehensive list of functional variants in the swine genome, single nucleotide polymorphisms (SNP) were identified from whole genome sequence of 240 pigs. Interim data from 72 animals in this study was published in 2017. This communication extends our previous work not only by utilizing genomic sequence from additional animals, but also by the use of the newly released Sscrofa 11.1 reference genome.

Results

A total of 26,850,263 high confidence SNP were identified, including 19,015,267 reported in our previously published results. Variation was detected in the coding sequence or untranslated regions (UTR) of 78% of the genes in the porcine genome: 1729 loss-of-function variants were predicted in 1162 genes, 12,686 genes contained 64,232 nonsynonymous variants, 250,403 variants were present in UTR of 15,739 genes, and 15,284 genes contained 90,939 synonymous variants. In total, approximately 316,000 SNP were classified as being of high to moderate impact (i.e. loss-of-function, nonsynonymous, or regulatory). These high to moderate impact SNP will be the focus of future genome-wide association studies.

Electronic supplementary material

The online version of this article (10.1186/s13104-018-3973-6) contains supplementary material, which is available to authorized users.

A survey of single nucleotide polymorphisms identified from whole-genome sequencing and their functional effect in the porcine genome. Anim Genet 2017;48:404-411. [PMID: 28485055 DOI: 10.1111/age.12557]

Genetic variants detected from sequence have been used to successfully identify causal variants and map complex traits in several organisms. High and moderate impact variants, those expected to alter or disrupt the protein coded by a gene and those that regulate protein production, likely have a more significant effect on phenotypic variation than do other types of genetic variants. Hence, a comprehensive list of these functional variants would be of considerable interest in swine genomic studies, particularly those targeting fertility and production traits. Whole-genome sequence was obtained from 72 of the founders of an intensely phenotyped experimental swine herd at the U.S. Meat Animal Research Center (USMARC). These animals included all 24 of the founding boars (12 Duroc and 12 Landrace) and 48 Yorkshire-Landrace composite sows. Sequence reads were mapped to the Sscrofa10.2 genome build, resulting in a mean of 6.1 fold (×) coverage per genome. A total of 22 342 915 high confidence SNPs were identified from the sequenced genomes. These included 21 million previously reported SNPs and 79% of the 62 163 SNPs on the PorcineSNP60 BeadChip assay. Variation was detected in the coding sequence or untranslated regions (UTRs) of 87.8% of the genes in the porcine genome: loss-of-function variants were predicted in 504 genes, 10 202 genes contained nonsynonymous variants, 10 773 had variation in UTRs and 13 010 genes contained synonymous variants. Approximately 139 000 SNPs were classified as loss-of-function, nonsynonymous or regulatory, which suggests that over 99% of the variation detected in our pigs could potentially be ignored, allowing us to focus on a much smaller number of functional SNPs during future analyses.

Litter-of-origin trait effects on gilt development

The preweaning litter environment of gilts can affect subsequent development. In a recent experiment designed to test the effects of diet on gilt development, litter-of-origin traits including individual birth weights, immunocrits (a measure of colostrum intake), sow parity, number weaned, and individual weaning weights were collected for approximately 1,200 gilts that were progeny of approximately 300 sows. Subsequently, BW, LM area, and backfat were measured at 100 d of age and at 28-d intervals until slaughter (260 d of age). From 160 d of age to slaughter, gilts were observed daily for estrus. At slaughter, the reproductive tract and 1 mammary gland were recovered. The reproductive tract was classified as cyclic or prepubertal; the number of corpora lutea was counted. Uterine horn lengths and ovarian dimensions were measured. Uterus and ovary samples from every 10th gilt were prepared for histological evaluation of uterine gland development and follicle counts, respectively. Mammary gland tissue protein and fat were assayed. Day of the estrous cycle at slaughter was calculated using the first day of the most recent standing estrus (d 0) recorded previous to slaughter. Each gilt development trait was analyzed for association with each litter-of-origin trait, after adjusting for dietary treatment effects. Uterine length, ovarian dimensions, mammary gland protein and fat, and uterine gland development were also adjusted for day of the estrous cycle at slaughter. All litter-of-origin traits were associated ( < 0.05) with growth traits. Top-down (backward elimination) multiple regression analysis indicated that BW and LM accretion in gilts was positively associated with immunocrit ( < 0.01), birth weight ( < 0.01), preweaning growth rate ( < 0.01), and parity ( < 0.01). Backfat accretion was positively associated with preweaning growth rate ( < 0.01), number weaned ( < 0.05), and parity ( < 0.05). Age at puberty was associated with birth weight (positive; < 0.01) and preweaning growth rate (negative; < 0.01). Total uterine length was positively associated with only birth weights ( < 0.05). Mammary gland protein was negatively associated with preweaning growth ( < 0.01). Mammary gland fat was positively associated with birth weight and number of piglets weaned ( > 0.05). These results indicate that colostrum consumption, birth weights, preweaning growth rate, number weaned, and parity are associated with gilt development traits during later life.

Implementing meta-analysis from genome-wide association studies for pork quality traits

Pork quality plays an important role in the meat processing industry. Thus, different methodologies have been implemented to elucidate the genetic architecture of traits affecting meat quality. One of the most common and widely used approaches is to perform genome-wide association (GWA) studies. However, a limitation of many GWA in animal breeding is the limited power due to small sample sizes in animal populations. One alternative is to implement a meta-analysis of GWA (MA-GWA) combining results from independent association studies. The objective of this study was to identify significant genomic regions associated with meat quality traits by performing MA-GWA for 8 different traits in 3 independent pig populations. Results from MA-GWA were used to search for genes possibly associated with the set of evaluated traits. Data from 3 pig data sets (U.S. Meat Animal Research Center, commercial, and Michigan State University Pig Resource Population) were used. A MA was implemented by combining -scores derived for each SNP in every population and then weighting them using the inverse of estimated variance of SNP effects. A search for annotated genes retrieved genes previously reported as candidates for shear force (calpain-1 catalytic subunit [] and calpastatin []), as well as for ultimate pH, purge loss, and cook loss (protein kinase, AMP-activated, γ 3 noncatalytic subunit []). In addition, novel candidate genes were identified for intramuscular fat and cook loss (acyl-CoA synthetase family member 3 mitochondrial []) and for the objective measure of muscle redness, CIE a* (glycogen synthase 1, muscle [] and ferritin, light polypeptide []). Thus, implementation of MA-GWA allowed integration of results for economically relevant traits and identified novel genes to be tested as candidates for meat quality traits in pig populations.

Plasma concentrations of acyl-ghrelin are associated with average daily gain and feeding behavior in grow-finish pigs

The objectives of this study were to determine the effect of sex, sire line, and litter size on concentrations of acyl-ghrelin and total ghrelin in plasma of grow-finish pigs and to understand the relationship of plasma concentrations of ghrelin with feeding behavior, average daily gain (ADG), and back fat in grow-finish swine. Yorkshire-Landrace crossbred dams were inseminated with semen from Yorkshire, Landrace, or Duroc sires. Within 24 h of birth, pigs were cross-fostered into litter sizes of normal (N; >12 pigs/litter) or small (S; ≤ 9 pigs/litter). At 8 wk of age, pigs (n = 240) were blocked by sire breed, sex, and litter size and assigned to pens (n = 6) containing commercial feeders modified with a system to monitor feeding behavior. Total time eating, number of daily meals, and duration of meals were recorded for each individual pig. Body weight was recorded every 4 wk. Back fat and loin eye area were recorded at the conclusion of the 12-wk feeding study. A blood sample was collected at week 7 of the study to quantify concentrations of acyl- and total ghrelin in plasma. Pigs from small litters weighed more (P < 0.05) and tended (P = 0.07) to be fatter than pigs from normal litters. Postnatal litter size did not affect ADG, feeding behavior, or concentrations of ghrelin in plasma during the grow-finish phase. Barrows spent more time eating (P < 0.001) than gilts, but the number of meals and concentrations of ghrelin did not differ with sex of the pig. Pigs from Duroc and Yorkshire sires had lesser (P < 0.0001) concentrations of acyl-ghrelin than pigs from Landrace sires, but plasma concentrations of total ghrelin were not affected by sire breed. Concentrations of acyl-ghrelin were positively correlated with the number of meals and negatively correlated with meal length and ADG (P < 0.05). A larger number of short-duration meals may indicate that pigs with greater concentrations of acyl-ghrelin consumed less total feed, which likely explains why they were leaner and grew more slowly. Acyl-ghrelin is involved in regulating feeding behavior in pigs, and measuring acyl-ghrelin is important when trying to understand the role of this hormone in swine physiology.

Genomewide association analysis for average birth interval and stillbirth in swine

Reproductive efficiency has a great impact on the economic success of pork production. Stillborn pigs and average birth interval contribute to the number of pigs born alive in a litter. To better understand the underlying genetics of these traits, a genomewide association study was undertaken. Samples of DNA were collected and tested using the Illumina Porcine SNP60 BeadChip from 798 females farrowing over a 4-yr period (all first parity). Birth intervals and piglet birth status (stillborn or alive) were determined by videotaping each farrowing event. A total of 41,148 SNP were tested using the Bayes C option of GenSel (version 4.61) and 1-Mb windows. These 1-Mb windows explained proportions of 0.017, 0.002, 0.032, 0.029, and 0.030 of the total variation, respectively, for litter average birth interval after deletion of the last piglet born, last birth interval in the litter, number of stillborn piglets ignoring the last piglet born, number of stillborns in the last birth position, and percent stillborn ignoring the last piglet. Significant 1-Mb nonoverlapping SNP windows were identified by using a conservative approach requiring 1-Mb windows to have a genetic variance ≥1.0% of genomic variance and these were considered to be QTL. Quantitative trait loci were located for number of stillborn piglets ignoring the last piglet born (1 QTL), number of stillborns in the last birth position (1 QTL), and percent stillborn ignoring the last piglet (3 QTL). In addition, 2, 13, 3, and 6 suggestive 1-Mb nonoverlapping SNP windows were identified for litter average birth interval after deletion of the last piglet born, number of stillborn piglets ignoring the last piglet born, number of stillborns in the last birth position, and percent stillborn ignoring the last piglet, respectively. Possible candidate genes affecting both birth interval and stillbirth included () and (). Possible genes affecting only birth interval included (), and (), and those affecting only stillbirth included (), LOC100518697 (a nostrin-like gene), and (). The QTL and the suggestive 1-Mb nonoverlapping SNP windows may lead to genetic markers for marker assisted selection, marker assisted management, or genomic selection applications in commercial pig populations.

Meta-analysis of genome-wide association from genomic prediction models

Genome-wide association (GWA) studies based on GBLUP models are a common practice in animal breeding. However, effect sizes of GWA tests are small, requiring larger sample sizes to enhance power of detection of rare variants. Because of difficulties in increasing sample size in animal populations, one alternative is to implement a meta-analysis (MA), combining information and results from independent GWA studies. Although this methodology has been used widely in human genetics, implementation in animal breeding has been limited. Thus, we present methods to implement a MA of GWA, describing the proper approach to compute weights derived from multiple genomic evaluations based on animal-centric GBLUP models. Application to real datasets shows that MA increases power of detection of associations in comparison with population-level GWA, allowing for population structure and heterogeneity of variance components across populations to be accounted for. Another advantage of MA is that it does not require access to genotype data that is required for a joint analysis. Scripts related to the implementation of this approach, which consider the strength of association as well as the sign, are distributed and thus account for heterogeneity in association phase between QTL and SNPs. Thus, MA of GWA is an attractive alternative to summarizing results from multiple genomic studies, avoiding restrictions with genotype data sharing, definition of fixed effects and different scales of measurement of evaluated traits.

Genomewide association and identification of candidate genes for ovulation rate in swine

Reproductive efficiency has a great impact on the economic success of pork production. Ovulation rate is an early component of reproduction efficiency and contributes to the number of pigs born in a litter. To better understand the underlying genetics of ovulation rate, a genomewide association study was undertaken. Samples of DNA were collected and tested using the Illumina Porcine SNP60 BeadChip from 1,180 females with ovulation measurements ranging from never farrowed to measurements taken after parity 2. A total of 41,848 SNP were tested using the Bayes C option of GenSel. After the Bayes C analysis, SNP were assigned to sliding windows of 5 consecutive SNP by chromosome-position order beginning with the first 5 SNP on SSC1 and ending with the last 5 SNP on SSCX. The 5-SNP windows were analyzed using the Predict option of GenSel. From the Predict analysis, putative QTL were selected having no overlap with other 5-SNP window groups, no overlap across chromosomes, and the highest genetic variation. These putative QTL were submitted to statistical testing using the bootstrap option of GenSel. Of the putative QTL tested, 80 were found to be statistically significant (P < 0.01). Ten QTL were found on SSC1, 12 on SSC2, 4 on SSC3, 8 on SSC4, 3 on SSC5, 3 on SSC6, 3 on SSC7, 4 on SSC8, 2 on SSC9, 4 on SSC10, 1 on SSC12, 4 on SSC13, 2 on SSC14, 4 on SSC15, 4 on SSC16, 6 on SSC17, 4 on SSC18, and 1 on SSCX. Sixteen QTL were found to be statistically significant at the P < 0.001 level. Six additional QTL were significant at the P = 0.001 level. These 22 QTL accounted for 71.10% of the total genetic variance. The most compelling candidate genes in these regions include Estrogen receptor 1, growth differentiation factor 9, and inhibin βA. These QTL, when combined with information on genes found in the same regions, should provide useful information that could be used for marker assisted selection, marker assisted management, or genomic selection applications in commercial pig populations.

Genome-wide association study of swine farrowing traits. Part I: genetic and genomic parameter estimates

The primary objective of this study was to determine genetic and genomic parameters among swine (Sus scrofa) farrowing traits. Genetic parameters were obtained using MTDFREML. Genomic parameters were obtained using GENSEL. Genetic and residual variances obtained from MTDFREML were used as priors for the Bayes C analysis of GENSEL. Farrowing traits included total number born (TNB), number born alive (NBA), number born dead (NBD), number stillborn (NSB), number of mummies (MUM), litter birth weight (LBW), and average piglet birth weight (ABW). Statistically significant heritabilities included TNB (0.09, P = 0.048), NBA (0.09, P = 0.041), LBW (0.20, P = 0.002), and ABW (0.26, P < 0.0001). Statistically significant genetic correlations included TNB-NBA (0.97, P < 0.0001), TNB-LBW (0.74, P < 0.0001), NBA-LBW (0.56, P < 0.0017), NSB-LBW (0.87, P < 0.0395), and LBW-ABW (0.63, P < 0.0002). Genetic parameters are similar to others found in the literature. The proportion of phenotypic variance explained by genomic markers (GP) generated by GENSEL was TNB (0.04), NBA (0.06), NBD (0.00), NSB (0.01), MUM (0.00), LBW (0.11), and ABW (0.31). Limited information is available in the literature about genomic parameters. Only the GP estimate for NSB is significantly lower than what has been published. The GP estimate for ABW is greater than the estimate for heritability found in this study. Other traits with significant heritability had GP estimates half the value of heritability. This research indicates that significant genetic markers will be found for TNB, NBA, LBW, and ABW that will have either immediate use in industry or provide a roadmap to further research with fine mapping or sequencing of areas of significance. Furthermore, these results indicate that genomic selection implemented at an early age would have similar annual progress as traditional selection, and could be incorporated along with traditional selection procedures to improve genetic progress of litter traits.

Validation and fine mapping of a QTL for ovulation rate on swine chromosome 3

Ovulation rate (OR) is an important component of litter size, but mutation(s) in gene(s) underlying OR QTL have yet to be identified in pigs. Markers within an OR QTL on SSC3 were genotyped in three white composite lines selected for ten generations for increased OR or uterine capacity (UC), with one line being an unselected control. Numbers of corpora lutea (CL) and UC (number of fully formed fetuses) were collected at approximately 105 days of gestation, as well as ovary weight (OW), uterine length (UL) and uterine weight (UW) measurements at 160 d of age in generation 12 and 13 females from all three lines. Six microsatellites and ten single nucleotide polymorphisms (SNPs; 0-42 cM) were genotyped in pigs from all lines of generations 11 through 13. The allele frequencies of 24269.1, SW2429, 7907.2 and 7637.2 were different (P < 0.01) in the OR line compared to the control line. A significant (P < 0.05) association of CL with 24269.1 (additive effect 0.65 ± 0.32) was detected, and additive genotypic effects approached significance for markers at 28 through 35 cM (16963.2, 27514.1 and SWR1637). Haplotyping of 7637.2 and 16963.2 (31 through 32 cM) identified a significant additive association of haplotype 1 with CL (-0.62 ± 0.30). These markers were also associated with OW (24296.1 and SWR1637), UL (16963.2, 27514.1 and haplotypes of 7637.2/16963.2) and UW (haplotypes of 7637.2/16963.2). This study verifies an OR QTL on SSC3. However, based on the data, it was concluded that there may be two genes, at 13 through 18 cM and 28 through 35 cM, controlling OR on SSC3p.

Predictive markers in calpastatin for tenderness in commercial pig populations

The identification of predictive DNA markers for pork quality would allow US pork producers and breeders to select genetically superior animals more quickly and efficiently for the production of consistent, high-quality meat. Genome scans have identified QTL for tenderness on SSC 2, which have been fine-mapped to the calpastatin locus. The objectives of this study were to identify the sequence variation in calpastatin that likely affects tenderness in commercial-level pig populations and to develop definitive DNA markers that are predictive of pork tenderness for use in marker-assisted selection programs. We resequenced the calpastatin regulatory and transcribed regions in pigs with divergently extreme shear force values to identify possible mutations that could affect tenderness. A total of 194 SNP were identified in this sequence, and 31 SNP were found in predicted transcription factor binding sites. We tested 131 polymorphisms in our research population and a subset (40) of these in samples of industry pigs for their association with objective measures of tenderness. We identified 4 SNP that were consistently associated with pork tenderness in all the populations studied, representing 2,826 pigs from 4 distinct populations. Gel shift assays were designed for these SNP and 12 other polymorphic sites. Six sites demonstrated a gel shift when probes were incubated with nuclear extract from muscle, heart, or testis. Four of these sites, a specificity protein 1 (Sp1) site around nucleotides 12978 and 12979, a potential thyrotroph embryonic factor (Tef) site at nucleotide 25587, an unknown site at nucleotide 48699, and myocyte enhancer factor-2 (Mef-2)/TATA sites with SNP at positions 49223 and 49228 were allele specific in binding nuclear proteins. The allele frequencies for the tender alleles were similar (0.11 to 0.36) in the 4 different commercial populations. These 4 SNP were not in complete linkage disequilibrium with each other and may independently affect calpastatin expression, tenderness, or both. These markers should be predictive of pork tenderness in industry populations.

Estimates of genetic parameters among scale activity scores, growth, and fatness in pigs

Genetic parameters for scale activity score (AS) were estimated from generations 5, 6, and 7 of a randomly selected, composite population composed of Duroc, Large White, and 2 sources of Landrace (n = 2,186). At approximately 156 d of age, pigs were weighed (BW) and ultrasound backfat measurements (BF1, BF2, and BF3) were done. While pigs were in the scale, an AS was assigned, which ranged from 1 (calm) to 5 (highly excited), where 58.1, 28.5, 8.9, 4.0, and 0.5% were scored as 1, 2, 3, 4, and 5, respectively. Statistical model effects were year-week of measurement, sex, covariates of age for AS and BW or BW for BF1, BF2, and BF3, and an animal direct genetic effect. A 5-trait linear mixed model was used. Estimated heritabilities were 0.23, 0.54, 0.56, 0.52, and 0.48 for AS, BW, BF1, BF2, and BF3, respectively. Estimated genetic correlations between AS and BW, AS and BF1, AS and BF2, and AS and BF3 were -0.38, -0.11, -0.12, and -0.16 respectively. Results indicated AS had a heritable genetic component and was genetically correlated with performance traits. Estimated genetic correlations between AS and backfat measurements adjusted to a common BW were negative, as was the genetic correlation of AS with BW. Therefore, selection for more docile animals would be expected to result in fatter, faster growing pigs.

Relationships among calpastatin single nucleotide polymorphisms, calpastatin expression and tenderness in pork longissimus

Genome scans in the pig have identified a region on chromosome 2 (SSC2) associated with tenderness. Calpastatin is a likely positional candidate gene in this region because of its inhibitory role in the calpain system that is involved in postmortem tenderization. Novel single nucleotide polymorphisms (SNP) in calpastatin were identified and used to genotype a population (n = 1042) of Duroc-Landrace-Yorkshire swine for association with longissimus lumborum slice shear force (SSF) measured at days 7 and 14 postmortem. Three genetic markers residing in the calpastatin gene were significantly associated with SSF (P < 0.0005). Haplotypes constructed from markers in the calpastatin gene were significantly associated with SSF (F-ratio = 3.93; P-value = 0.002). The levels of normalized mRNA expression of calpastatin in the longissimus lumborum of 162 animals also were evaluated by real-time RT-PCR and were associated with the genotype of the most significant marker for SSF (P < 0.02). This evidence suggests that the causative variation alters expression of calpastatin, thus affecting tenderness. In summary, these data provide evidence of several significant, publicly available SNP markers associated with SSF that may be useful to the swine industry for marker assisted selection of animals that have more tender meat.

Genetic variation in the mannosidase 2B2 gene and its association with ovulation rate in pigs

Ovulation rate is an important phenotypic trait that is a critical component of litter size in pigs. Despite being moderately heritable in pigs, selection for increased ovulation rate is difficult because it is difficult to measure and is a sex-limited trait. A QTL for ovulation rate residing on the p-terminal end of pig chromosome 8 has been detected in a Meishan-cross resource population. Comparative analysis of this region yielded a positional candidate gene, mannosidase 2B2 (MAN2B2), for this QTL. The entire coding region of MAN2B2 was resequenced in the Meishan and White Composite founder animals of the resource population to identify SNPs. Eleven polymorphisms that alter the protein product of MAN2B2 were discovered and tested for statistical associations with ovulation rate in three generations of the resource population. The polymorphism located at position 1574 of the mRNA (D28521:c.1574A>G) was the most significant polymorphism tested (P = 0.00005) where the additive effect of the c.1574A allele was estimated to be -0.89 ova. This polymorphism was determined to be more significantly associated with ovulation rate than the breed-specific analysis conducted during the line-cross QTL discovery. The c.1574A>G marker was not associated with ovulation rate in an occidental population. Therefore, either MAN2B2 has a unique epistatic interaction within the Meishan-cross population or the c.1574A>G SNP is in linkage disequilibrium with the actual causative genetic variation in the Meishan-cross population.

Single nucleotide polymorphisms for pig identification and parentage exclusion

Single nucleotide polymorphisms (SNPs) have become an important type of marker for commercial diagnostic and parentage genotyping applications as automated genotyping systems have been developed that yield accurate genotypes. Unfortunately, allele frequencies for public SNP markers in commercial pig populations have not been available. To fulfil this need, SNP markers previously mapped in the USMARC swine reference population were tested in a panel of 155 boars that were representative of US purebred Duroc, Hampshire, Landrace and Yorkshire populations. Multiplex assay groups of 5-7 SNP assays/group were designed and genotypes were determined using Sequenom's massarray system. Of 80 SNPs that were evaluated, 60 SNPs with minor allele frequencies >0.15 were selected for the final panel of markers. Overall identity power across breeds was 4.6 x 10(-23), but within-breed values ranged from 4.3 x 10(-14) (Hampshire) to 2.6 x 10(-22) (Yorkshire). Parentage exclusion probability with only one sampled parent was 0.9974 (all data) and ranged from 0.9594 (Hampshire) to 0.9963 (Yorkshire) within breeds. Sire exclusion probability when the dam's genotype was known was 0.99998 (all data) and ranged from 0.99868 (Hampshire) to 0.99997 (Yorkshire) within breeds. Power of exclusion was compared between the 60 SNP and 10 microsatellite markers. The parental exclusion probabilities for SNP and microsatellite marker panels were similar, but the SNP panel was much more sensitive for individual identification. This panel of SNP markers is theoretically sufficient for individual identification of any pig in the world and is publicly available.

Detection of single nucleotide polymorphisms associated with ultrasonic backfat depth in a segregating Meishan x White Composite population

Multiple genomic scans have identified QTL for backfat deposition across the porcine genome. The objective of this study was to detect SNP and genomic regions associated with ultrasonic backfat. A total of 74 SNP across 5 chromosomes (SSC 1, 3, 7, 8, and 10) were selected based on their proximity to backfat QTL or to QTL for other traits of interest in the experimental population. Gilts were also genotyped for a SNP thought to influence backfat in the thyroxine-binding globulin gene (TBG) on SSC X. Genotypic data were collected on 298 gilts, divided between the F8 and F10 generations of the US Meat Animal Research Center Meishan resource population (composition, one-quarter Meishan). Backfat depths were recorded by ultrasound from 3 locations along the back at approximately 210 and 235 d of age in the F8 and F10 generations, respectively. Ultrasound measures were averaged for association analyses. Regressors for additive, dominant, and parent-of-origin effects of each SNP were calculated using genotypic probabilities computed by allelic peeling algorithms in GenoProb. The association model included the fixed effects of scan date and TBG genotype, the covariates of weight and SNP regressors, and random additive polygenic effects to account for genetic similarities between animals not explained by known genotypes. Variance components for polygenic effects and error were estimated using MTDFREML. Initially, each SNP was fitted (once with and once without parent-of-origin effects) separately due to potential multi-collinearity between regressions of closely linked markers. To form a final model, all significant SNP across chromosomes were included in a common model and were individually removed in successive iterations based on their significance. Across all analyses, TBG was significant, with an additive effect of approximately 1.2 to 1.6 mm of backfat. Three SNP on SSC3 remained in the final model even though few studies have identified QTL for backfat on this chromosome. Two of these SNP exhibited irregular parent-of-origin effects and may not have been detected in other genome scans. One significant SNP on SSC7 remained in the final, backward-selected model; the estimated effect of this marker was similar in magnitude and direction to previously identified QTL. This SNP can potentially be used to introgress the leaner Meishan allele into commercial swine populations.

A genome scan for loci affecting pork quality in a Duroc-Landrace F population

A genome scan was conducted on 370 F2 Duroc-Landrace pigs. Microsatellite markers (n = 182) were genotyped across the entire F2 population, all F1 parents and the paternal grandparents. Breed of origin of all chromosomal segments inherited in F2 progeny were predicted using GenoProb, where genotypic data, genetic maps and extended pedigrees were used as inputs. Statistical tests for quantitative trait loci (QTL) associations were conducted on 41 phenotypes with SAS using output from GenoProb for genotypic data. Fixed effects included sex and age at slaughter. For certain analyses carcass weight, RYR1 genotype and/or PRKAG3 genotype were also included as covariates. Subjective and objective measures of pork colour, marbling and tenderness were recorded, as well as measures of carcass fatness and muscularity. Test results were adjusted to a genome-wide level of significance. Five genomic regions presented significant evidence for QTL at chromosome 1 positions 6 cM (intramuscular fat) and 67 cM (Hunter L*), chromosome 2 position 62 cM (taste panel tenderness), chromosome 17 position 50 (loineye area and image analysis estimated loineye area) and X position 87 cM (carcass weight). Sixty-six suggestive associations were detected. Fourteen of these associations were within the regions with significant QTL on chromosomes 2, 17 and X, and the remaining 52 associations resided in 29 other regions on 13 different chromosomes of the porcine genome. The chromosome 2 region of 60-66 cM was associated with all measures of pork tenderness and the region on chromosome 17 (32-39 cM) was associated with both measures of intramuscular fat and loineye area. After verification, the QTL for marbling and tenderness should be useful in commercial production to improve pork quality as the population was developed from two of the three most utilized breeds of swine in the USA.

Rearranged gene order between pig and human in a quantitative trait loci region on SSC3

A quantitative trait locus (QTL) for ovulation rate on chromosome 3 that peaks at 36 cM has been identified in a Meishan-White composite resource population with an additive effect of 2.2 corpora lutea. As part of an effort to identify the responsible gene(s), typing of additional genes on the INRA-University of Minnesota porcine radiation hybrid (IMpRH) map of SSC3 and comparative analysis of gene order was conducted. We placed 52 known genes and expressed sequence tags, two BAC-end sequences and one microsatellite (SB42) on a framework map that fills gaps on previous RH maps. Data were analysed for two-point and multipoint linkage with the IMpRH mapping tool and were submitted to the IMpRH database (http://imprh.toulouse.inra.fr/). Gene order was confirmed for 42 loci residing in the QTL region (spanning c. 17 Mb of human sequence) by using the high-resolution IMpRH2 panel. Carthagène (http://www.inra.fr/internet/departments/MIA/T/CarthaGene) was used to estimate multipoint marker distance and order using all public markers on SSC3 in the IMpRH database and those typed in this study. For the high-resolution map, only data for markers typed in both panels were used. Comparative analysis of human and porcine maps identified conservation of gene order for SSC3q and multiple blocks of conserved segments for SSC3p, which included six distinct segments of HSA7 and two segments of HSA16. The results of this study allow significant refinement of the SSC3p region that contains an ovulation rate QTL.

Assignment of 204 genes localized on HSA17 to a porcine RH (IMpRH) map to generate a dense comparative map between pig and human/mouse

Bi- and uni-directional chromosome painting (ZOO-FISH) and gene mapping have revealed correspondences between human chromosome (HSA) 17 and porcine chromosome (SSC) 12 harboring economically important quantitative trait loci. In the present study, we have assigned 204 genes localized on HSA17 to SSC12 to generate a comprehensive comparative map between HSA17 and SSC12. Two hundred fifty-five primer pairs were designed using porcine sequences orthologous with human genes. Of the 255 primer pairs, 208 (81.6%) were used to assign the corresponding genes to porcine chromosomes using the INRA-Minnesota 7000-rad porcine x Chinese hamster whole genome radiation hybrid (IMpRH) panel. Two hundred three genes were integrated into the SSC12 IMpRH linkage maps; and one gene, PPARBP, was found to link to THRA1 located in SSC12 but not incorporated into the linkage maps. Three genes (GIT1, SLC25A11, and HT008) were suggested to link to SSC12 markers, and the remaining gene (RPL26) did not link to any genes/expressed sequence tags/markers registered, including those in the present study. A comparison of the gene orders among SSC12, HSA17, and mouse chromosome 11 indicates that intra-chromosomal rearrangements occurred frequently in this ancestral mammalian chromosome during speciation.

Genetic variation in sperm production

In boars, the primary determinant of daily sperm production is the number of Sertoli cells, which establishes testicular weight. The only breed comparison of foetal testicular development in boars contrasted two diverse breeds, White composite (WC, Landrace-Yorkshire) with Meishan, a Chinese breed that undergoes pubertal development at a young age and has small testicular size. During the prenatal period, the pattern of change in testicular development is similar in these two breeds with both having their greatest proportion of proliferating Sertoli cells at 90 days of gestation, and with WC boars possessing more Sertoli cells and greater mass of seminiferous tubules during the latter half of gestation. During the first month of life, Meishan boars accumulate Sertoli cells and mass of seminiferous tubules at a greater rate than WC boars, and Meishan boars undergo terminal differentiation of Sertoli cells at a younger age. Postpubertal boars, within each breed and crossbreds of the two breeds, with small testicular size have increased circulating concentrations of follicle-stimulating hormone. No direct breed comparisons of testicular development are apparent for postpubertal boars of other breeds. Accepting the limitations of data reported from different laboratories, Piau boars reach puberty at an older age and have a greater proportion of their testes occupied with seminiferous tubules than Meishan boars; both breeds have small testes. A gene or genes on the X chromosome code for small testicular size in Meishan crossbred boars; genetic determinants of testicular size and sperm production in other breeds remain to be identified.

Deciphering the pig genome to understand gamete production

The field of livestock genomics has made considerable advances in the past decade. In the area of pig reproduction, a number of genome scans have identified several genomic regions associated with variation in reproductive measures ranging from ovulation rate, litter size and testis size. Additionally, several candidate genes have been associated with variation in litter size. These studies primarily focused on developing genetic markers to facilitate selection decisions. To date, their results have made minor contributions to commercial pig performance and our knowledge on the inheritance of complex phenotypes. With the availability of additional resources for pig, as well as from human and mouse studies, future studies should be directed to identifying genetic variation that affects biological processes. To reach this goal, teams of diversely trained scientists need to be formed that include geneticists, physiologists, molecular biologists and bioinformaticists. A diversified team of scientists equipped with all of the available research tools (genomic sequence data, expression arrays, knowledge of gene product functions, etc.) and appropriate swine populations should be able to decode the genome's hidden secrets on how it controls reproductive processes.