Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
A QTL affecting daily feed intake maps to Chromosome 2 in pigs Ross D. Houston,1,2* Chris S. Haley,3 Alan L. Archibald,3 Kellie A. Rance1 1 Aberdeen Centre for Energy Regulation and Obesity (ACERO), Energy Balance and Obesity Division, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, Scotland, UK 2 Division of Agriculture and Forestry, School of Biological Sciences, University of Aberdeen, AB24 4FA, Scotland, UK 3 Department of Genetics and Genomics, Roslin Institute (Edinburgh), Roslin, Midlothian EH25 9PS, Scotland, UK Received: 29 November 2004 / Accepted: 3 March 2005 Abstract Our understanding of the molecular genetic basis of several key performance traits in pigs has been significantly advanced through the quantitative trait loci (QTL) mapping approach. However, in contrast to growth and fatness traits, the genetic basis of feed intake traits has rarely been investigated through QTL mapping. Since feed intake is an important component of efficient pig production, the identification of QTL affecting feed intake may lead to the identification of genetic markers that can be used in selection programs. In this study a QTL analysis for feed intake, feeding behavior, and growth traits was performed in an F2 population derived from a cross between Chinese Meishan and European Large White pigs. A QTL with a significant effect on daily feed intake (DFI) was identified on Sus scrofa Chromosome 2 (SSC2). A number of suggestive QTL with effects on daily gain, feed conversion, and feeding behavior traits were also located. The significant QTL lies close to a previously identified mutation in the insulin-like growth factor 2 gene (IGF2) that affects carcass composition traits, although the IGF2 mutation is not segregating in the populations analyzed in the current study. Therefore, a distinct causal variant may exist on the P arm of SSC2 with an effect on feed intake. Introduction The advent of porcine molecular genetic marker maps and advances in statistical methodology have *Present address: Division of Genetics and Genomics, Roslin Institute (Edinburgh), Roslin, Midlothian EH25 9PS, Scotland, UK Correspondence to: Ross D. Houston; E-mail: ross.houston@bbsrc. ac.uk 464 facilitated the mapping of quantitative trait loci (QTL) for economically important traits in livestock. The genetic basis of performance traits, such as growth and fatness in pigs, has been well studied through QTL mapping in experimental crosses between various phenotypically divergent founder breeds such as Chinese Meishan or Wild Boar and European Large White (reviewed in Bidanel and Rothschild 2002). These QTL mapping studies have led to the identification of several genomic regions with significant effects on these performance traits. The identified QTL can be seen as a starting point for identifying the underlying causal polymorphisms, or closely linked markers, which can be used in marker-assisted selection programs to improve performance (Georges 1999). While feed intake is clearly an economically important trait in pigs and accounts for a significant proportion of the costs involved in pig production (Webb 1998), the molecular genetic basis of feed intake traits has received less attention than that of other performance traits. This may be partly because of the difficulty of measuring and analyzing accurate feed intake data in a group-housed system compared with traits such as daily gain and backfat (Hyun et al. 1997). However, with the development of electronic feed intake equipment tailored for use in such environments (Hyun et al. 1997), the opportunity exists to obtain accurate measures of feed intake, feed efficiency, and feeding behavior and, therefore, to investigate the molecular genetic basis of these traits. The relative difficulty in measuring accurate feed intake and feed efficiency indicates that marker-assisted selection may have advantages over traditional selection strategies for these traits, provided markers associated with suitably large and consistent phenotypic effects can be identified. The QTL mapping approach is often the starting point for locating such markers. DOI: 10.1007/s00335-004-4026-0 Volume 16, 464–470 (2005) Springer Science+Business Media, Inc. 2005 R.D. HOUSTON ET AL.: DAILY FEED INTAKE QTL IN PIGS A number of candidate genes have been associated with feed intake, growth, and fatness in pigs, including leptin (LEP), leptin receptor (LEPR), and melanocortin-4 receptor (MC4R) (reviewed in Roehe et al. 2003). It is interesting to examine whether any QTL identified for feed intake or growth traits coincide with the regions of the genome to which these genes map. Also, since adiposity stores are a key variable in regulating the feed intake of an animal (Woods and Seeley 2000), it is worthwhile assessing whether any of the previously identified major QTL affecting fatness are associated with a pleiotropic effect on feed intake. Furthermore, identification of genomic regions associated with feed intake may be of value to human obesity research through the use of comparative maps. Therefore, the objective of this study was to analyze electronically recorded feed intake data for a Meishan · Large White F2 population in a genome scan to identify QTL with effects on feed intake, growth, and feeding behavior traits. Materials and methods Animals. The animals used in this study came from an F2 population derived from a cross between Meishan and Large White pigs at the Roslin Institute. The subset of animals that had feed intake data recorded came from two separate batches—one born in 1995 and the other born in 1996. For both of these populations, reciprocal F1 crosses were produced from unrelated Meishan and Large White F0 animals. The F0 Meishan pigs were derived from the importation of 11 males and 21 females from the Jiadan county pedigree on the Lou Tang research farm in China in 1987 (Haley et al. 1992). The F0 Large White pigs were from a British control population, derived from a broad sample of genotypes in 1982 (Walling et al. 1998). For the animals used in the current study, a total of 9 F0 males and 18 F0 females were used to produce 4 F1 males and 58 F1 females, which were intercrossed to produce 88 F2 males and 67 F2 females (total n = 155) Performance testing. The F2 population was performance tested using single-spaced FIRE feeders (Feed Intake Recording Equipment; Osborne Europe Ltd, Newcastle-Upon-Tyne). The feeder animals were approximately three months old (or between 30 and 35 kg) at the start of the trial and were tested until they reached approximately 80 kg, although some animals remained in the feeders for longer. During the trial animals were group-housed in pens with solid floors and bedded with straw, with each pen containing 13 single-sex animals and one feeder 465 station, to which the animals were given 24-hour access. They were fed a dry pelleted proprietary ration containing 18% crude protein, 5.09% oil, 6.38% fiber 5.78% ash, and 1.05% lysine ad libitum. Each animal was assigned an individual identity number (ID) and a tag number, which were used by the electronic recording system for animal identification. Data summary. Every time an animal visited the feeder, the ID and weight of the animal, the total feed intake during the visit, and the time spent in the feeder were electronically recorded and used to give daily measurements of median weight (of all measurements taken during the day), total feed intake, number of visits to the feeder, and the total time in the feeder for each animal. Through individual scrutiny of each animalÕs data records, the suitability of each animal for inclusion in the analysis was determined based on the quality of the growth curve and the quantity of measurement errors. The daily data records for the included animals were summarized to give measurements of average daily gain (ADG, g), daily feed intake (DFI, g), feed conversion ratio (FCR), number of visits to the feeder per day (NVD), time in feeder per day (DFT, min), average feed per visit (AFV, g), and average feeding rate (AFR, g/sec). These traits were calculated for the early (35–55 kg), late (55–80 kg), and entire (35– 80 kg) test periods. Map construction. Genotype data for 114 markers across the 18 autosomes and the sex chromosome for the QTL pedigree were available on the ResSpecies database hosted at the Roslin Institute (www.resspecies.org). Marker order and sex-averaged marker positions were estimated using the BUILD option in Cri-Map 2.4 (Green et al. 1990). QTL Analysis. The fixed effects of litter, gender, batch (year of birth), and feeder number were assessed using a restricted maximum likelihood (REML) approach in Genstat 6.1 (Genstat Committee 2002) to determine which effects were explaining a significant proportion of the phenotypic variance in the feeder traits. Because of the relatively small sample size (n = 155) and the large number of full-sib families in the F2 population (n = 58), it was decided not to include litter as a fixed effect in the QTL analyses because of the substantial loss of degrees of freedom. Feeders were gender-specific and, therefore, fitting feeder number in the analyses accounts for variance resulting from gender. Since the same feeder may have functioned differently in the different years, a composite vari- 466 able of feeder number · batch was included as a fixed effect in the QTL analysis. The covariates of age at 35 kg or age at 55 kg were also included in the model, depending on which growth period was under assessment. The halothane gene was not segregating in the Large White founder animals (Cameron et al. 1988). To account for genetic effects from other chromosomes, QTL detected at the suggestive level were included as cofactors in any further analysis of that particular trait. The web-mounted QTL Express software (http:// qtl.cap.ed.ac.uk; Seaton et al. 2002) was used to calculate the information content across the genome and the proportion of animals with missing genotype information. The statistical principles of Haley et al. (1994) were then applied to perform the QTL analysis in which the genotype information is used to calculate the probabilities of F2 individuals having a particular putative QTL genotype, based conditionally on marker genotypes and accounting for the parental origin of the alleles. The probabilities are then used in a least-squares framework to investigate the genetic model underlying the trait (Knott et al. 1998). To test for evidence of an imprinting effect associated with the QTL, an imprinting term was included as an addition to the additive and dominance model. The imprinting model was compared with a model assuming no QTL (F statistic, 3 degrees of freedom) and the best-single-QTL model (F statistic, 1 degree of freedom). Thresholds. Suggestive and significant thresholds were calculated following the guidelines of Lander and Kruglyak (1995) using a permutation test (Churchill and Doerge 1994) with the map and genotype data described above and randomly created phenotype data with a mean of 0 and a standard deviation of 1 (these data were used because of the large number of traits and growth periods analyzed). Five thousand permutations were performed in the QTL Express program, which calculates the F ratio that is equivalent to the 0.05-level that chromosomewide thresholds automatically. Following the principles of Knott et al. (1998), the suggestive thresholds were determined from the 0.05 chromosome-wide level for each chromosome separately. For the genome-wide significance threshold, the permutation process was repeated using the actual phenotype data for the trait that gave the highest F-ratio in the QTL analysis (DFI 2). The thresholds for determining the genome-wide significance of the imprinting model were calculated as described in Knott et al. (1998). For significant QTL an approximation of the confidence interval was calculated by the 1-LOD drop support method (Lander and Botstein 1989). R.D. HOUSTON ET AL.: DAILY FEED INTAKE QTL IN PIGS Results Linkage map. In order to build a linkage map based on the genotype data for the Meishan · Large White F2 population, the Cri-Map BUILD option was used. The map details are presented in Fig.1. The gender averaged map length was 2131 cM, which gives an average distance between markers of 18.7 cM. The information content was generally maintained above 0.5 throughout the genome. The animals used in the current study were a subset of those used in Lee et al. (2003), and the estimated gender-averaged order and positions for the markers common to both studies were similar. The marker order was consistent with other published linkage maps (Archibald et al. 1995; Rohrer et al. 1996), while estimated marker positions were similar. Thresholds. A permutation analysis in QTL Express was used to calculate the genome-wide significant and suggestive thresholds. The genomewide significance threshold was calculated to be 8.38 at the 5% level, while the suggestive threshold varied between 4.19 and 5.42, depending on the chromosome under analysis. The 5% genome-wide threshold calculated using permutation of the trait DFI 2 was calculated to be 8.50. These thresholds are comparable to previously published studies (e.g., Knott et al. 1998; Walling et al. 1998; de Koning et al. 1999). The genome-wide significance threshold for the imprinting model was estimated to be 6.5, while the suggestive thresholds for SSC2 was estimated to be 4.2. The F ratio corresponding to the nominal significance level when comparing the imprinting model to the best single-QTL model was 3.9. QTL analysis. A genome-wide scan for QTL affecting the calculated feed intake, feeding behavior, and growth traits was performed. The abbreviations and descriptive statistics for the measured traits are presented in Table 1. The only QTL that reached genome-wide significance was located on SSC2 at 28 cM, and it affects daily feed intake during the 55–80 kg growth period (Table 2, Fig. 2). There was also suggestive evidence of an effect on daily feed intake during the entire growth period with a similar additive effect, but the best-estimated position was 3 cM (Table 2, Fig. 2). There was suggestive evidence for QTL that affect average daily gain (Chromosomes 8, 11, and 17), daily feed intake (Chromosomes 11, 13, and 17), feed conversion ratio (chromosomes 11, 12, and 14), daily feeding time (Chromosome 6), average feed per visit (chromosomes 11 and 15), and average feeding rate (chromosomes 3 and 14) (Table 2). R.D. HOUSTON ET AL.: DAILY FEED INTAKE QTL IN PIGS 467 Fig. 1. Names of the markers used in the study, with their calculated order and linkage map positions (given in centimorgans). While the best-estimated position on SSC2 of the significant QTL affecting DFI was 28 cM, the test statistics were high over much of the chromosome (Fig. 2). A calculation of the 1-LOD drop support confidence intervals suggests that the QTL was located between 1 and 63 cM. The QTL that affect DFI (55–80 kg and 35–80 kg) were additive in nature with no evidence of a dominance effect. The additive effects associated with the QTL were large at 244 g (±58) and 179 g (±46) for the 55–80 kg and 35–80 kg growth periods, respectively (Table 2). The positive estimation of the additive effect indicates that the feed intake–increasing allele originates from the Large White line. There was no evidence that including an imprinting term significantly improved the model compared with the additive and dominance model. Therefore, it can be concluded that there is no parent-of-origin effect associated with the QTL. Discussion To locate and characterize QTL with effects on feed intake, feeding behavior, and growth traits, a genome-wide QTL analysis was performed in a Meishan · Large White F2 population which had been per- formance-tested with electronic feeding equipment. Potentially, the most interesting QTL was located on SSC2 and had an additive effect on daily feed intake during the late test period, which was significant at the genome-wide level, and an additive effect on daily feed intake during the entire test period, which reached the suggestive level. A number of other QTL that reached the suggestive threshold were located, including QTL affecting average daily gain, feed conversion ratio, and various feeding behavior traits. The appropriate significance thresholds for determining suggestive and significant QTL were determined by permutation tests in the current study. While these account for the testing of multiple positions across the genome, they do not account for the testing of multiple traits. For example, with 21 different traits analyzed, one expects approximately 21 QTL to surpass the suggestive threshold by chance alone. Therefore, as with many QTL genome scans, it is important to be alert to the possibility that a number of the putative QTL detected may be false-positive results. Nonetheless, it is important to note these putative QTL since comparisons across multiple studies and populations can provide much stronger evidence for a QTL 468 R.D. HOUSTON ET AL.: DAILY FEED INTAKE QTL IN PIGS Table 1. Summary of the feeder traits analyzed in each growth period (means and standard deviations) 35–55 kg (1) 55–80 kg (2) 35–80 kg (3) Trait Abbrev. Mean SD Mean SD Mean SD Average daily gain (g) Daily feed intake (g) Feed conversion ratio Number of visits per day Time in feeder per day (min) Average Feed per visit (g) Feed rate (g/min) ADG DFI FCR NVD DFT AFV AFR 619 1795 2.95 6.87 58.3 287 31.6 122 275 0.45 2.24 12.5 93 6.66 635 2207 3.52 6.54 55.6 370 40.5 123 397 0.53 2.25 14.8 122 9.18 620 2005 3.26 6.66 56.6 329 36.2 105 306 0.41 2.08 13.2 104 7.44 SD = standard deviation composition traits has been discovered (Van Laere et al. 2003). The causal polymorphism is a guanineto-adenine polymorphism in intron 3 of the imprinted IGF2 gene (IGF2-intron3-G3072A) and has been shown to affect gene expression through the disruption of a transcription factor binding site (Van Laere et al. 2003). IGF2-intron3-G3072A lies close to the best-estimated positions of the daily feed intake QTL identified in the current study, which raises the possibility that the mutation has a pleiotropic effect on DFI. The animals analyzed in the current study came from a larger Meishan · Large White resource population, in which all founders and F1 sires were within a particular region. A previous QTL analysis in pigs has examined daily feed intake, although no significant or suggestive QTL were located (Rohrer 2000), which may be partly because of the low number of animals with feed intake records in the analysis (n = 92). For this reason, further QTL studies for feeding traits in pigs are recommended, particularly to determine whether the finding of a significant QTL for daily feed intake on SSC2 can be replicated in other populations. The P arm of SSC2 has received much attention recently since the causal polymorphism underlying a QTL with a large effect on fatness and carcass Table 2. Details of all QTL detected at the suggestive or significant level, their estimated position on the chromosome, and additive and dominance effectsa SSC Traitb Pos.(cM)c F ratio Additive effecte(SE) Dominance effectf(SE) 2 2 3 3 6 7 7 7 8 11 11 11 11 12 12 12 13 14 14 14 15 15 17 17 Daily feed intake 2 Daily feed intake 3 Average feeding rate 2 Average feeding rate 3 Time in feeder per day 1 Average feeding rate 1 Average feeding rate 2 Average feeding rate 3 Average daily gain 2 Average daily gain 1 Daily feed intake 1 Feed conversion ratio 2 Average feed per visit 2 Feed conversion ratio 2 Feed conversion ratio 3 Average feeding rate 2 Daily feed intake 1 Average feeding rate 1 Average feeding rate 3 Feed conversion ratio 2 Average feed per visit 1 Average feed per visit 3 Average daily gain 3 Daily feed intake 3 28 3 49 46 48 150 150 150 43 76 76 30 0 4 0 73 0 81 81 77 71 61 16 21 9.3 7.9 8.2 5.2 6.5 6.1 5.7 6.8 6.7 5.1 5.2 7.8 5.7 5.6 7.1 6.4 6.7 5.9 7.1 5.4 6.9 4.8 4.8 4.9 244 179 5.72 3.65 73.5 0.28 2.78 0.55 )73.6 )40.3 )104 )0.30 )8.70 0.11 0.09 3.52 24.9 )1.95 )1.95 )0.22 )52.8 )54.1 19.6 28.4 )121 )39.4 2.68 1.29 )709 )3.86 )4.97 )5.77 43.4 )44.7 )46.4 0.30 )124 0.34 0.25 5.11 174 2.29 3.62 )0.08 27.6 30.7 )67.9 )272 a (58.0) (45.8) (1.54) (1.19) (126) (0.67) (1.23) (0.96) (20.6) (16.2) (33.6) (0.08) (22.7) (0.07) (0.05) (1.47) (37.7) (0.72) (1.02) (0.07) (15.4) (18.3) (14.2) (50.8) (89.9) (66.1) (2.64) (1.94) (199) (1.1) (2.06) (1.60) (48.8) (22.0) (45.5) (0.15) (36.6) (0.11) (0.07) (2.15) (58.0) (1.04) (1.48) (0.11) (28.7) (37.8) (25.6) (90.2) SSC = Sus crofa chromosome, cM = centimorgan, SE = standard error, Pos. = position. 1, 2, 3 refer to growth periods 35–55 kg, 55–80 kg, and 35–80 kg respectively. Sex-averaged map positions from the marker closest to the distal tip of the P arm. d Significance threshold F = 8.50, suggestive threshold F = 4.19–5.42 depending on chromosome under study. e Positive value indicates increasing allele originates from Large White, negative value indicates increasing allele originates from Meishan. f Positive value indicates heterozygotes have a higher mean trait value than the midparent. b c R.D. HOUSTON ET AL.: DAILY FEED INTAKE QTL IN 469 PIGS Acknowledgments The authors are indebted to the staff at Dryden Farm (Roslin) for care of the animals and data collection. They thank Rosalie Waldron, Heather Finlayson, Alison Downing, and Jen Anderson for technical assistance. This project was supported through the LINK SLP program, with funding from DEFRA, MLC, Cotswold Pig Development Company Ltd., JSR Healthbred Ltd., Newsham Hybrid Pigs Ltd., Rattlerow Farms, and PIC. References Fig. 2. A plot of the likelihood profile for daily feed intake across Chromosome 2. fixed for guanine at the mutation, except one heterozygous founder Large White and one heterozygous F1 sire (Van Laere et al. 2003). However, the F2 offspring derived from this heterozygous sire were not part of the subset of animals with recorded feed intake data, thus indicating that IGF2-intron3G3072A is not responsible for the feed intake effects shown in the current study. Therefore, a distinct causal variant on the P arm of Chromosome 2 is likely to have effects on feed intake in this Meishan · Large White population. Since the estimated confidence interval for the DFI QTL is very large, there are likely to be dozens of other positional candidate genes in this region, including the insulin gene (INS) which maps close to IGF2 (Van Laere et al. 2003), However, like IGF2, INS is paternally expressed (Van Laere et al. 2003) and there was no evidence for an imprinting effect in the current study. A number of other candidate genes map to the regions covered by the suggestive QTL identified. The prepro-orexin gene (HCRT) has been mapped to SSC 12 (Malek et al. 2000), close to the suggestive QTL for feed conversion ratio, and is a plausible candidate gene for feeding-related traits (Malek et al. 2000). Also, the mahogany (ATRN) gene maps to a region of SSC17 which approximately coincides with the suggestive QTL that affect daily gain and feed intake (Edeal et al. 2000). Further investigation of these and other positional candidate genes could lead to the identification of genes, and causal polymorphisms, underlying the QTL effects. This would increase the plausibility of including the QTL in a marker-assisted selection program to modulate these economically important performance traits. Furthermore, since the pig is implicated as an excellent model for human obesity (Roehe et al. 2003), the investigation of the causal gene may lead to novel information on the control of obesity-related traits in humans. 1. Archibald AL, Haley CS, Brown JF, Couperwhite S, McQueen HA, et al. (1995) The PiGMaP consortium linkage map of the pig (sus scrofa). Mamm Genome 6, 157–175 2. Bidanel JP, Rothschild M (2002) Current status of quantitative trait locus mapping in pigs. Pig News Inf 23, 39N–54N 3. Cameron ND, Curran MK, Thompson R (1988) Estimation of sire with feeding regime interaction in pigs. Anim Reprod 46, 87–95 4. Churchill GA, Doerge RW (1994) Empirical threshold values for quantitative trait mapping. Genetics 138, 963–971 5. de Koning DJ, Janss LL, Rattink AP, van Oers PA, de Vries BJ, et al. (1999) Detection of quantitative trait loci for backfat thickness and intramuscular fat content in pigs (Sus scrofd). Genetics 152, 1679–1690 6. Edeal B, Rumph JM, Mass R, Killinger K, Jerez N, et al. (2000) Rapid communication: Linkage mapping of the mahogany (attractin) locus in cattle and pigs. J Anim Sci 78, 2479–2480 7. Genstat Committee (2002) Genstat release 6.1 reference manual (Oxford: VSN International) 8. Georges M (1999) Towards marker assisted selection in livestock. Reprod Nutr Dev 39, 555–561 9. Green P, Falls K, Crooks S (1990) Documentation for crimap, version 2.4 (St. Louis: Washington University School of Medicine) 10. Haley CS, Agaro Ed , Ellis M (1992) Genetic components of growth and ultrasonic fat depth traits in Meishan and Large White pigs and their reciprocal crosses. Anim Prod 54, 105–115 11. Hyun Y, Ellis M, McKeith FK, Wilson ER (1997) Feed intake pattern of group-housed growing-finishing pigs monitored using a computerized feed intake recording system. J Anim Sci 75, 1443–1451 12. Knott SA, Marklund L, Haley CS, Andersson K, Davies W, et al. (1998) Multiple marker mapping of quantitative trait loci in a cross between outbred wild boar and Large White pigs. Genetics 149, 1069–1080 13. Lander ES, Botstein D (1989) Mapping Mendelian factors underlying quantitative traits using RFLP Linkage maps. Genetics 121, 185–199 14. Lander E, Kruglyak L (1995) Genetic dissection of complex traits: Guidelines for interpreting and reporting linkage results. Nat Genet 11, 241–247 470 15. Lee GJ, Archibald AL, Garth GB, Law AS, Nicholson D, et al. (2003) Detection of quantitative trait loci for locomotion and osteochondrosis-related traits in Large White · Meishan pigs. R Anim Sci 76, 155–165 16. Malek M, Marklund S, Dyer C, Matteri R, Rothschild R (2000) Linkage and physical mapping of the porcine prepro-orexin gene. Mamm Genome 11, 342–343 17. Roehe R, Plastow GS, Knap PW (2003) Quantitative and molecular genetic determination of protein and fat deposition. Homo 54, 119–131 18. Rohrer GA (2000) Identification of quantitative trait loci affecting birth characters and accumulation of backfat and weight in a meishan–white composite resource population. J Anim Sci 78, 2547–2553 19. Rohrer GA, Alexander LJ, Hu Z, Smith TP, Keele JW, et al. (1996) A comprehensive map of the porcine genome. Genome Res 6, 371–391 R.D. HOUSTON ET AL.: DAILY FEED INTAKE QTL IN PIGS 20. Seaton G, Haley CS, Knott SA, Kearsey M, Visscher PM (2002) QTL express: Mapping quantitative trait loci in simple and complex pedigrees. Bioinformatics 18, 339–340 21. Van Laere AS, Nguyen M, Braunschweig M, Nezer C, Collette C, et al. (2003) A regulatory mutation in igf2 causes a major QTL effect on muscle growth in the pig. Nature 425, 832–836 22. Walling GA, Archibald AL, Cattermole JA, Downing AC, Finlayson HA, et al. (1998) Mapping of quantitative trait loci on porcine chromosome 4. Anim Genet 29, 415–424 23. Webb AJ (1998) Objectives and strategies in pig improvement: An applied perspective. J Dairy Sci 81 Suppl 2, 36–46 24. Woods SC, Seeley RJ (2000) Adiposity signals and the control of energy homeostasis. Nutrition 16, 894–902