Brief Summary of Minutes of Annual Meeting

Members of the NC1004 Multistate Project (Genetic and Functional Genomic Approaches to Improve Production and Quality of Pork) met on January 14, 2006 in conjunction with the NRSP-8 swine sub-committee and the Plant and Animal Genome Meetings in San Diego, CA. This was the fourth meeting of the committee and covered the period January 1 through December 31, 2005 Efforts have continued at several stations to pursue detection of quantitative trait loci (QTL), with efforts at Michigan and Iowa focused primarily on meat quality traits. Iowa has continued their analyses of the Berkshire x Yorkshire population and joint analyses of the ISU Berkshire x Yorkshire and the Univ. of Illinois Berkshire x Duroc populations. Joint data from two F2 crosses between commercial breeds (Berkshire-Yorkshire at Iowa State University (ISU), and Berkshire-Duroc at the University of Illinois (UOI)) were used to detect QTL with a unique mode of genomic imprinting, polar overdominance (POD), in which all genotypes have equal effect, except one of the heterozygotes (Qq or qQ). All animals were genotyped for 39 (ISU) and 32 (UOI) markers on chromosomes (SSC) 2, 6, 13, and 18. Twenty-six growth, carcass composition, and meat quality traits were analyzed, and tests to differentiate POD QTL from Mendelian, paternal or maternal expression QTL were applied. Two POD QTL, for 24-hr Loin pH (49 cM) and lipid % (151 cM) were detected on SSC2 at a 5% genome-wise (GW) level in the ISU population. An additional two QTL for firmness and juiciness were detected at a 5% chromosome-wise (CW) level on SSC2 and SSC13, respectively. Three POD QTL were detected at the GW level in UOI population, for 24-hr Loin Hunter (89 cM in SSC2), last-rib and average back fat (72 cM, 115 cM in SSC6), and four POD QTL at the CW level, for last-rib and tenth-rib back fat in SSC6, and color score and 24-hr Loin Hunter in SSC18. Joint analyses revealed 3 POD QTL at the GW level for last-rib back fat, lipid %, and 24-hr Loin pH in SSC6, which were detected at the GW level in ISU or UOI. An additional 9 POD QTL were detected at the CW level, of which 5 were not detected in either ISU or UOI. Although false positive results cannot be excluded, their study suggests evidence of POD QTL in pigs and advantages of combining data to increase power for QTL detection. Mode of expression of detected QTL requires verification in other studies. Iowa is also using QTL regions identified to investigate biological and positional candidate genes that are associated with growth, backfat, appetite, and meat quality in commercial pigs. Genes are identified, polymorphisms detected and mapped, and association analyses are conducted. In addition, reproduction and reproductive longevity candidate genes are being identified, mapped and association studies conducted. While Michigan report results from their Duroc x Pietrain reference population. A total of 55 QTL for 22 of the 29 measured growth traits were found to be significant at the 5% chromosome-wise level. Of these 55 QTL, 16 were significant at the 1% chromosome-wise, 11 at the 5% genome-wise, and 10 at the 1% genome-wise significance thresholds. A total of 33 QTL for 15 of the 16 animal random regression terms were found to be significant at the 5% chromosome-wise level. Of these 33 QTL, 11 were significant at the 1% chromosome-wise, 2 at the 5% genome-wise, and 2 at the 1% genome-wise significance thresholds. Putative QTL were discovered for 10th rib and last rib backfat on SSC 6, body composition traits on SSC 9, backfat and lipid composition traits on SSC 11, 10th rib backfat and total body fat tissue on SSC 12, and linear regression of body weight, longissimus muscle area, and 10th rib backfat on SSC 18. A total of 94 QTL for 35 of the 38 carcass and meat quality traits analyzed were found to be significant at the 5% chromosome-wise level. Of these 94 QTL, 43 were significant at the 1% chromosome-wise, 27 at the 5% genome-wise, and 14 at the 1% genome-wise significance thresholds. Putative QTL were discovered for 45 min pH and pH decline on SSC 3, marbling score and carcass backfat on SSC 6, carcass length and number of ribs on SSC 7, marbling score on SSC 12, and color measurements and tenderness score on SSC 15. These results will facilitate fine mapping efforts to identify genes controlling growth, composition and meat quality traits that can be incorporated into marker-assisted selection programs to accelerate genetic improvement in pig populations. The Nebraska station furthered their search for QTL affecting reproduction and applied new statistical models to existing data. Data from the F2 resource population were reanalyzed with models that allowed composite interval mapping and search for chromosomal regions with imprinted effects. Based on these findings, a simulation program was developed to simulate the pedigree structure and selection background from Generation 0 to 23 in the Nebraska lines. A QTL for litter size, the one with the largest effect identified in the genome scan, with frequency of .5 for each of two alleles was incorporated at a known position on one chromosome and SNP were inserted at 1 cM intervals along this chromosome. The objective was to determine the number of SNP to include in haplotypes to maximize the efficiency of fine mapping QTL of the magnitude found in the NE selection lines and with a similar population structure. Simulated data were compared with actual selection responses and parameter estimates to validate the simulation model. Fine mapping of QTL identified in previous scans was accomplished by using phenotypic and genotypic data from pigs of several generations from the lines described above. Phenotypic data were collected for birth weight (BWT, n = 1422), weaning weight (WWT, n = 1311), age at puberty (AP, n = 669), ovulation rate (OR, n = 797), number of fully formed pigs (FF, n = 841), number of pigs born alive (BA, n = 841), number of mummified pigs (MUM, n = 841), number of nipples (NN, n = 1434), splayleg incidence (n = 458), and number of stillborn pigs (SB, n = 841). Age at puberty was recorded in gilts of Lines I and C1 from Generation 2 through Generation 15 and in gilts of Lines IOL, COL, and C2 through Generation 16. Ovulation rate was recorded in gilts of Lines I and C1 through Generation 11, and in gilts of Lines IOL, COL, and C2 through Generation 16. Number of fully formed pigs, MUM, SB, and BA were recorded in gilts within 24 h of parturition each generation. Four regions on chromosomes 6, 11, 12, and 13 were chosen to fine map and validate QTL because these regions had previously shown to harbor QTL in the F2 resource population (Cassady et al., 2001; Holl et al., 2004). Monsanto Choice Genetics developed SNP panels to genotype samples. A total of 1167 animals were genotyped for 118 single nucleotide polymorphism (SNP) markers. The Iowa, Nebraska, North Carolina, and Ohio stations continue to create and maintain selection lines of pigs. At Iowa after four generations of selection for intramuscular fat (IMF), the average EBV for select line pigs is 1.03% greater than for control line pigs. Of the pigs harvested in generation four, line LS means for tenth rib backfat and loin muscle area were 18.58 mm and 42.94 cm2 in the control line, and 21.62 mm and 39.22 cm2 in the select line (P < 0.05), respectively. Analysis of STAGES data evaluated on all 810 pigs in generation four revealed no significant difference between lines for days to 114 kg, however, compositional differences similar to those found in the pigs harvested were noted. Results through generation four indicate that selection for IMF has resulted in slightly more tenth-rib backfat and less LMA, while having no significant effect on growth performance. Chemical analysis of a loin sample from pigs harvested revealed a significant phenotypic response in IMF (3.04% in the control line vs. 3.97% in the select line) similar to the difference between lines for IMF EBV. Line LS means for pigs harvested in generation four for 24 h Hunter L and Minolta were 45.77 and 21.99 in the control line, and 49.79 and 24.67 in the select line (P < 0.05), respectively. Subjective measures of marbling were significantly different between lines (2.25 in the control line vs. 3.00 in the select line); however, subjective measures of color and firmness revealed no significant difference. Other meat quality characteristics such as Instron tenderness, pH, and percent cooking loss, as well as sensory panel evaluations of juiciness, tenderness, chewiness, flavor, and off-flavor were not significantly different after four generations of selection for IMF. Selection on IMF EBV has yielded correlated responses in terms of slightly lighter and less desirable objective measures of color; however, it has had no effect on other objective measures of meat quality. At Nebraska selection for litter size and its component traits ovulation rate, embryonic survival, and uterine capacity was initiated 1981. Five lines were developed including 1) Line I: Index selection line, selected 11 generations for ovulation rate and embryonic survival, and an additional 13 generations for increased litter size; 2) C1, randomly selected control line contemporary to I, 3) Line IOL; derived from Line I at Generation 8 and subsequently selected 8 generations for ovulation rate and uterine capacity followed by 7 generations of litter size selection; 4) Line COL: derived from Line C1 at Generation 8 and subsequently selected 8 generations for ovulation rate and uterine capacity followed by 7 generations of litter size selection; and 5) Line C2: a line contemporary with Lines IOL and COL, derived from Line C1 at Generation 8 and subsequently selected randomly. At Ohio data were collected from the initial random mated population of Landrace and Berkshire purebred swine were used to estimate genetic (co)variance parameters for use in establishing within and across breed heritability and genetic correlation estimates for use in subsequent long-term genetic selection programs for improved pork quality in the swine species. Data from 1,065 purebred Landrace and 203 purebred Berkshire barrow and gilt progeny were included in the estimation process. Heritability estimates were as follows: Days to 113.4 kg (0.22), Backfat (0.62), Loin Muscle Area (0.57), Minolta Color (0.36), Visual Marbling Score (0.26), Intramuscular Fat Percentage (0.35), Warner-Bratzler Shear Force (0.35), and 24 Hour pH (0.25). Notable genetic correlations were Backfat and pH (r = 0.17), Minolta Color and pH (r = -0.50), Marbling Score and Warner-Bratzler Shear (r = -0.40), Intramuscular Fat Percentage and Warner-Bratzler Shear (r = -0.45). Analyses also showed significant breed effects for most production, carcass, and pork quality traits. In general, Berkshire pigs grew slower, having more tenth rib backfat and smaller tenth rib loin muscle area. In addition, loins from Berkshire pigs were darker in visual and instrumental color, with more visual and chemical marbling content, less purge loss and much more tender measures of instrumental tenderness. The data provided to date are being used to initiate a long-term selection project to improve pork quality in the Landrace breed. North Carolina evaluated the relationships between indirect measures of behavior and performance. The objective of this research was to test the phenotypic relationships among the backtest, resident-intruder test scores (RIS), growth, LM area, and backfat in pigs. Little is known about relationships among measures of pig behavior and economically important traits. However, it may be expected that a pigs behavior affects its performance and the performance of its pen mates. The backtest and resident-intruder test were each done twice on pigs (n = 145) from 20 litters. During the backtest a pig was gently restrained in a supine position for 60 s. Number of bouts of struggling by the pig and total time spent struggling were recorded. Cumulative number of escape attempts (BTS) and cumulative time spent struggling (TTS) during both backtests were analyzed. The resident intruder tests, done with 30 to 50 d old pigs, measures the tendency for aggressive behavior toward other pigs. A solid divider was placed down the center of the pen to separate the resident pig from its penmates. An intruder pig of the same sex and smaller size was placed into the pen with the resident pig. When an attack initiated by the resident pig occurred pigs were immediately separated and the test was terminated (RIS = 1). If after 5 minutes no attack occurred, the test was terminated (RIS = 0). The BTS and TTS were correlated (r = 0.77; P < 0.05) and RIS and BTS were uncorrelated (r = 0.1; P = 0.29). Dam effects influenced BTS, TTS, and RIS (P < 0.03). The BTS and TTS affected ADG in the nursery (P < 0.04) and BTS also affected backfat (P < 0.07) and LM area (P < 0.06). The RIS affected ADG in the nursery, Kg of acceptable standardized fat-free lean per day, total Kg of acceptable standardized fat-free lean and backfat (P < 0.05) and tended to affect days to 110 Kg and LM area (P < 0.1). Piglets with RIS = 2, showing more aggression, had fewer days to 110 Kg, greater lean gain, and more Kg of fat-free lean than pigs with RIS = 1 or 0 (P < 0.05). Increased BTS and RIS were associated with increased kilograms of lean. In conclusion, phenotypic associations do exist among measures of behavior and performance traits, and increased lean gain was associated with increased aggression. Groups at Iowa, Nebraska, and Minnesota have also developed bioinformatic tools. Minnesota continues their work on a relational database for conducting comparative and functional genomics research. MANGOdb contains an exhaustive laboratory information management system (LIMS) for storing data about such diverse topics as animal pedigrees and treatment, tissue and nucleic acid isolation, library production and sequencing, contig building and annotation, genetic and physical mapping, and macro and microarray annotation and hybridization. Current work on the database is focused on the storage and analysis of microarray-based gene expression data and is incorporating the front end and tables from the Stanford Microarray Database, with annotation from a customized version of SOURCEdb. While Iowa and Nebraska have developed computer modeling and simulation programs to develop, optimize, and evaluate strategies for marker-assisted selection. Groups at Nebraska and BARC continue to collaborate on project related to pig disease. At Nebraska two replications of a PRRSV challenge experiment with a total of 400 pigs were conducted. Replication 1 occurred during summer 2002, and Replication 2 in winter of 2003. In each replicate a total of 100 PRRSV-negative, SEW pigs of the NE Index line (Line I) and 100 pigs of a commercial Duroc-Hampshire (DH) line were used. Line I is an inbred Large White-Landrace population that has been closed for 24 generations and selected for increased litter size. It was expected to have low resistance to disease because of its relatively high inbreeding (~25%). Line DH is a non-inbred, terminal sire line that excels in growth and leanness and was expected to have greater disease resistance than Line I. Genetically diverse lines were used to maximize the opportunity to detect genetic resistance to PRRSV. Two pigs from each of 200 litters by 163 dams and 83 sires were sampled to ensure genetic diversity within lines to maximize the chance that genes for both resistance and susceptibility to PRRSV existed in the sample. Pigs were transferred from their farm of origin at approximately 25 d of age to the University of Nebraska Veterinary Biomedical Sciences animal research facility and randomly assigned within line and litter to isolated rooms for PRRSV challenge. Each replicate included four rooms with 25 pigs per room (12 or 13 from each population). Littermates were assigned to different rooms. Rooms were randomly assigned to treatments with two being control (no PRRSV challenge) and two rooms containing littermates to the pigs in the control rooms assigned to PRRSV challenge. Each uninfected littermate served as a control for its infected littermate. After a 5-d adaptation period, body temperature was recorded and blood was withdrawn from all pigs. Challenged pigs were infected with PRRSV RFLP-Iowa Strain, the standard virulent strain used by the VBMS virology lab of F. Osorio. Blood was withdrawn and body temperature recorded at 4, 7, and 14 d post-challenge. All pigs were sacrificed at d-14, and lung and lymph tissue collected and frozen. Blood, lymph tissue, and lung tissue were analyzed for presence of virus with a procedure that measures the ability of the pig to replicate the PRRSV virus. The presence of lesions in lungs and lymph nodes was characterized. Principal component analyses were used to identify pigs from each population with high and low phenotypic response to PRRSV. Traits included in analyses were weight gain, change in body temperature, viremia, lymph and spleen PRRSV titer, ELISA antibody ratio, and lung lesion score. Variation in the first principal component described overall phenotypic response to PRRSV. The left tail of the distribution contained pigs that had been infected with PRRSV, but showed very few symptoms of disease, whereas those in the right tail were infected and had severe cases of PRRS. Pigs in these tails were arbitrarily categorized as resistant or susceptible. Seven to eight pigs within each category within each population were identified. RNA was extracted from lung tissue from these pigs and their littermates and cDNA was created. Differences in expression of genes was evaluated with microarray analyses. In addition, a sample of lung of each pig was sent to Dr. Joan Lunney, USDA BARC, who evaluated differences in expression of 12 specific immune function genes. During 2005, RNA was extracted from lymph tissue of these same pigs and differences in gene expression between resistant and susceptible pigs was determined. BARC has focused on PRRS. They utilized real-time expression assays for panels of immune markers known to control vaccine and disease immunity (see database www.ba.ars.usda.gov/nrfl/nutri-immun-db/nrfl_query1new.html). Their infectious disease work has been aimed at determining effect or mechanisms which lead to innate resistance to infection or promote protective responses against infection. They have used real time gene expression assays to monitor immune gene activation during innate and adaptive [T helper 1 (Th1) and Th2] immune responses. They tested samples, collected as part of the national PRRS CAP grant, from swine infected with, or vaccinated for PRRS virus. Their data indicated that the slow and weak development of innate immunity as evidenced by interferon-alpha (IFNA), intereukin-1 (IL1), IL6, IL8 responses result in a weak interferon-gamma (IFNG) response, thus preventing protective responses and enabling PRRSV to persist in infected pigs. In summary, significant progress has been made toward accomplishing the outcomes and outputs defined for the NC1004 multi-state project. Numerous collaborative linkages have developed between the NC1004 member stations (examples include IA/IL, IA/BARC/OK, MI/OK, NE/BARC, NC/MARC, VT/MARC) and many of these partnerships are resulting in successful grant funding, which is allowing research efforts for the project to move forward. Shared resources and public availability of data has been a cornerstone of the project and these continue to be developed and distributed. Integration of quantitative and molecular information facilitates rapid discovery of genetic variation influencing pig health, production efficiency and pork quality. In addition, outreach activities and interaction with industry scientists and producers allow transfer of knowledge to the industry.