Investigation of a QTL region for loin eye area and fatness on pig 1

Laura Grapes, Max F. Rothschild

Department of Animal Science and Center for Integrated Animal Genomics, 2255 Kildee Hall, Iowa State University, Ames, Iowa 50011, USA

Received: 22 December 2005 / Accepted: 14 March 2006

quantitative trait loci (QTL) that impact traits of Abstract economic importance. However, very few of the Previously, quantitative trait loci (QTL) for tenth-rib and/or mutations underlying the observed backfat (TENTHRIB) and loin eye area (LEA) were QTL have been identified (Kim et al. 2000; Van Laere identified on pig Chromosome 1 (SSC 1) near micro- et al. 2003). In fact, few genes have been identified as satellite S0008 from a three-generation Berkshire · significant positional candidates for observed QTL Yorkshire cross (BY). This work attempted to refine effects. Ciobanu et al. (2004) showed that mutations these QTL positions and identify genes associated in the calpastatin , which is located in a region with these QTL. Genotypes of BY (n = 555) were suggestive for average instron force, tenderness, determined by PCR-RFLP or PCR tests for 13 poly- juiciness, and chewiness QTL, were significantly morphisms identified in BY F0 individuals for candi- associated with these traits. Clearly, further work is date genes, BAC end sequences, and genomic clones. necessary to identify the genes underlying the Using least-squares regression interval mapping, the results from the numerous porcine genome scans. LEA QTL was estimated at S0008; the TENTHRIB Malek et al. (2001a) reported genome-wise sig- QTL position was shifted approximately 1 cM nificant QTL for loin eye area (LEA) and tenth-rib downstream from S0008. Of the genes/sequences backfat (TENTHRIB) on pig Chromosome 1 (SSC 1) mapped in the QTL region, CL349415 was signifi- near microsatellite S0008. These QTL were identi- cantly associated with TENTHRIB (p = 0.02) and fied in a three-generation Berkshire · Yorkshire solute carrier family 2, member 12 (SLC2A12) was breed cross (BY). Alleles originating from the Berk- significantly associated with LEA (p = 0.02). These shire breed interestingly were shown to be cryptic results suggest that the gene(s) responsible for the and increase LEA and decrease TENTHRIB; these LEA and TENTHRIB QTL effects are tightly linked to QTL explained 4.21% and 4.78% of the phenotypic S0008 or that the high informativeness of S0008 rel- variance for LEA and TENTHRIB, respectively, in ative to surrounding markers is influencing the QTL BY F2 individuals. In the same chromosomal region, position estimates. In addition, janus kinase 2 (JAK2) a chromosome-wise significant QTL for average was mapped to a suggestive LEA QTL region and backfat (AVBFAT) and suggestive QTL for marbling showed association with LEA (p = 0.009), fatness, score (MARB), total lipid percentage (TOTLIPPR), color, and pH traits in BY. lumbar fat (LUMBAR), and cholesterol (CHOLES) were identified in BY (Malek et al. 2001a,b). Thom- sen et al. (2004) performed further QTL analyses in the same BY population after adding 33 microsatel- lite markers to the overall genome map, 3 of which Introduction were located on SSC 1. The LEA and TENTHRIB Several whole-genome scans have been conducted in QTL identified by Malek et al. (2001a) were con- a variety of pig breed crosses (Malek et al. 2001a,b; firmed, but the AVBFAT QTL dropped just below Milan et al. 2002; Ovilo et al. 2002; Varona et al. significance (Thomsen et al. 2004). These results 2002) to identify chromosomal regions containing indicate that there are gene(s) located near S0008 that have a large impact on LEA and fatness traits in Correspondence to: Max F. Rothschild; E-mail: mfrothsc@ Berkshire and/or Yorkshire pigs. Hence, the objec- iastate.edu tive of this work was to fine map these QTL regions

DOI: 10.1007/s00335-005-0188-7 Volume 17, 657 668 (2006) Ó Springer Science+Business Media, Inc. 2006 657  À  658 L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1 as an initial step toward identifying the underlying 2001) and the gene encoding it (RRAGD) maps to HSA genes. 6q15. The nuclear receptor 2 E1 gene (NR2E1) maps to HSA 6q21 and was selected as a representative gene from this human region to determine the comparative Materials and methods map position in pig. Connective tissue growth factor Resource population. A three-generation resource functions in insulin-like growth factor binding population was created and managed as described by and stimulation of cell proliferation (reviewed in Malek et al. (2001a). Briefly, two purebred Berkshire Brigstock 1999) and the gene (CTGF) maps to HSA 6q grandsires were crossed with nine purebred Yorkshire 23.2. Solute carrier family 2, member 12 gene granddams, producing nine litters from which 8 F1 (SLC2A12) maps to HSA 6q 23.2, is expressed in adi- sires and 26 F1 dams were selected. These F1 individ- pose tissue, and has glucose transport properties (re- uals were crossed to produce 525 F2 individuals. viewed in Joost and Thorens 2001). Ectonucleotide pyrophosphatase 1 gene (ENPP1) also maps to HSA Phenotypic data collection. Phenotypic values 6q23.2, about 140 kb upstream of CTGF, and muta- for loin eye area, tenth-rib backfat, average backfat, tions in the gene have been significantly associated lumbar backfat, last rib backfat, marbling score, total with insulin resistance and obesity (Pizzuti et al. lipid percent, and cholesterol were collected on pigs 1999; Meyre et al. 2005). 8 as described in Malek et al. (2001a,b). family, member A1 gene (ALDH8A1) maps to HSA 6q23.3 and plays a role in the pathway of 9-cis-retinoic Genetic marker development. Microsatellite acid biosynthesis (Lin and Napoli 2000). In addition, marker genotypes for SSC 1 described in Malek et al. the janus kinase 2 gene (JAK2) was selected for study (2001a) and Thomsen et al. (2004) were available. To because of its role in muscle cell differentiation and increase marker density in the region significant for proliferation and its known location on SSC 1 (Wang LEA and TENTHRIB QTL, several candidate genes, et al. 2004). BAC end sequences, and genomic clones were se- After candidate genes CTGF and ENPP1 were lected for polymorphism discovery and genotyping. mapped in the BY flanking S0008, pig BAC end se- quence bE1I10T7 was identified as homologous to Candidate gene, BAC end sequence, and clone human sequence between CTGF and ENPP1 using the selection. Candidate genes were selected based upon Porcine BES Search tool available from The Wellcome comparative human pig chromosomal information Trust Sanger Institute (http://www.sanger.ac.uk/ and known functionalÀ information. The microsatel- cgi-bin/Projects/S_scrofa/BESsearch.cgi) and selected lite marker S0008, nearest to the most significant for marker development. Two porcine genomic positions for LEA and TENTHRIB QTL, had been clones, CL349415 and CL353852, were also selected physically mapped to SSC 1p22-23 (Robic et al. for marker development based on their comparative 1996). This region is syntenic to the q arm of human human map positions (Rogatcheva et al. 2006). For (HSA 6; Fronicke et al. 1996). How- NR2E1, no annotated pig sequence was available. In- ever, the synteny between the p arm of SSC 1 and the stead, one BAC end sequence, 199F3SP6, identified as q arm of HSA 6 is not well defined, and gene order in homologous to the human gene position for NR2E1 this region of SSC 1p has been inverted compared to using the Porcine BES Search tool, was selected. HSA 6q. Therefore, candidate genes from HSA 6q12 q23 were selected for marker development and Polymorphism identification. For all genes and subsequentÀ mapping in the pig to determine overall sequences, PCR primers were designed from avail- gene order on SSC 1p and increase marker density in able human or pig sequence using the default set- the QTL regions. tings of Primer3 web-based software (http:// The Ph.D. finger protein 3 gene (PHF3) maps to frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) HSA 6q12 and was chosen to determine whether to amplify genomic DNA from the 11 F0 animals of genes from this human region may map to the pig the BY population. Accession numbers of human QTL region of interest. The potassium channel gene and pig sequences from which primers were de- (KCNQ5) maps to HSA 6q13 and is involved in volt- signed, the sequences of those primers, and the age-gated potassium transfer (Lerche et al. 2000). The resulting fragment sizes are given in Table 1. Se- gene encoding phosphoglucomutase 3 (PGM3) maps quences from the F0 individuals were compared for to HSA 6q14.2 and is involved in carbohydrate each gene/clone/BAC end sequence using Sequen- metabolism (Hopkinson and Harris 1968). Ras-related cher v3.0 software (Gene Codes Corporation, Ann GTP binding D is a G protein that acts in numerous Arbor, MI) to identify polymorphisms. Most poly- cell processes and signaling pathways (Sekiguchi et al. morphisms discovered were single nucleotide poly- L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1 659

Table 1. Pig/human sequences used to design initial PCR primers, primer sequences, and resulting fragment sizes for genes and sequences examined on SSC 1 Gene/Sequence (Accession No. / TIGR gene index) Primer sequences Fragment size (bp) PHF3 exon 16 (NM_015153) F: 5¢ GCCTCCGAATGTCTTTAACC 3¢ 237 R: 5¢ GATTTCCATTTCCAGGACCTC 3¢ KCNQ5 exon 14 (TC145264) F: 5¢ AGCTCCGCACAGAACAGC 3¢ 756 R: 5¢ CAAGCTGAGGGCGTCTGTAT 3¢ PGM3 exons 5-6 (BF189324) F: 5¢ TGACTATGGCTTGTTGACGC 3¢ 1267 R: 5¢ TCCCTTTGGTCCCATCATTA 3¢ RRAGD exons 3-4 (TC166509) F: 5¢ GACCAGGGCCTACAAAGTGA 3¢ 1087 (expected)a R: 5¢ AGCAAATTCTCCCAGAGTGG 3¢ BAC end sequence 199F3SP6 F: 5¢ TAGAGCCCTAGAGCCAGCAGG 3¢ 725 R: 5¢ CATCAGAGAGCACCAAACTTCTT 3¢ CTGF exons 4-5 (U70060) F: 5¢ TTACCGACTGGAAGACACGTT 3¢ 836 R: 5¢ GTTGTAATGGCAGGCACAAG 3¢ ENPP1 exons 8-9 (TC213901) F: 5¢ ACCCTGAGTGGTACAAAGGA 3¢ 783 R: 5¢ CATCTGATCCTGGCCAAAAA 3¢ SLC2A12 exon 5 (TC173292) F: 5¢ GAACTATGTGAAAAACAACATT 3¢ 581 R: 5¢ TCCTAACCAGAACCCTGTCC 3¢ ALDH8A1 exon 7 (BM499254) F: 5¢ CATGCTTCCCACGGTGATA 3¢ 252 R: 5¢ GCAAGGTTCAGCTCTCTGATG 3¢ BAC end sequence bE1I10T7 F: 5¢ AATTTTTCCACACTACCCTGCT 3¢ 875 R: 5¢ ATTTCAATGGAAGAGTGTTGCTC 3¢ CL349415 F: 5¢ CCCAAAAGAGATGTATGTATAAAAGGA 3¢ 544 R: 5¢ TCACTCACATCAGCAAGATGG 3¢ CL353852 F: 5¢ TCTTGTGGGATCTGACACCT 3¢ 586 R: 5¢ ACAGTGTGGGTGGCATGTAA 3¢ JAK2 exons 13-14 (NM_214113) F: 5¢ TGGGCCATGCATTTTCTAGT 3¢ 1229 R: 5¢ GGAAGACAGGAAATCCTCCC 3¢ aFragment size of the sequence amplified from RRAGD exons 3-4 is given as an expected size (based on human intron size) because the complete porcine sequence of this fragment was not obtained due to sequencing difficulties across intron 3. morphisms (SNPs); however, a 9-bp insertion/dele- 199F3SP6, a PCR-based test was designed to distin- tion was found for BAC end sequence 199F3SP6. One guish genotypes (Table 2). A restriction enzyme that polymorphism from each gene/sequence was se- could recognize the selected SNP for each gene/se- lected to be genotyped in the BY population. For quence was identified using NEBCutter v2.0 (http:// genes PHF3, KCNQ5, PGM3, RRAGD, CTGF, tools.neb.com/NEBcutter2/index.php). All digestions ENPP1, SLC2A12, ALDH8A1, and JAK2, sequences were performed according to the manufacturer’s obtained from the BY F0 individuals were submitted instructions (New England Biolabs, Ipswich, MA). to GenBank. Accession numbers for those submitted Digestion/PCR products were loaded on 4% NuSieve sequences are used in Tables 2, 3, and 4. agarose gels (Cambrex Corporation, East Rutherford, NJ) stained with ethidium bromide, electrophoresed, Genotyping conditions. For each selected SNP, a and viewed by UV light to observe genotypes. The PCR restriction fragment length polymorphism identified polymorphisms, enzymes, and resulting (PCR-RFLP) test was designed for genotyping the BY fragment sizes for all digestions of PCR products are population. For genes KCNQ5, PGM3, RRAGD, and given in Table 3. Fragment sizes for the PCR-based JAK2, primers had to be redesigned to conduct PCR- test of 199F3SP6 genotypes are also given in Table 3. RFLP tests. All PCR primers for these tests were de- Allele frequencies for all polymorphisms identified in signed from amplified sequences (see Table 2 for Gen- the BY F0 individuals are given in Table 4. Bank accession numbers) and by using the default settings of Primer3 web-based software (http://frodo. Statistical analysis. Genotypes from BY were wi.mit.edu/cgi-bin/primer3/primer3_www.cgi). For all used to position each gene/sequence on the SSC 1 other genes and sequences, PCR-RFLP tests were car- linkage map using the ‘‘all’’ option of CRI-MAP v2.4 ried out using the original primers designed to amplify software (Green et al. 1990) to obtain the best map the gene/sequence (Table 1). Primer sequences, PCR order and the ‘‘fixed’’ option to obtain map distances. reaction and cycling conditions, and resulting frag- To refine LEA and fatness QTL positions on SSC ment sizes are given in Table 2 for all PCR-RFLP tests. 1, least-squares regression interval mapping was For the 9-bp indel found in BAC end sequence performed using QTLExpress software (http:// Table 2. PCR primer sequences, reaction and cycling conditions, and fragment sizes for PCR and PCR-RFLP tests for SSC 1 660 Gene/Sequence Fragment PCR reaction (GenBank accession No.) Primer sequences size (bp) conditionsa PCR cycling conditionsb PHF3 exon 16 F: 5¢ GCCTCCGAATGTCTTTAACC 3¢ 237 Standard reaction Standard cycling conditions; (DQ322274) R: 5¢ GATTTCCATTTCCAGGACCTC 3¢ conditions 51°C annealing temp KCNQ5 exon 14 F: 5¢ ATTTAATATTACCTGACCAAGGACCGGTCT 3¢ 399 Standard reaction 94°C 2 min; 30 cycles of (DQ322246) R: 5¢ CAAGCTGAGGGCGTCTGTAT 3¢ conditions 94°C 30 sec, 53°C 30 sec, 72°C 45 sec; 72°C 7 min PGM3 intron 5 exon 6 F: 5¢ TGCTTTATGCTTCACTTTTAC 3¢ 203 Standard reaction Standard cycling conditions; (DQ322247) À R: 5¢ TCCCTTTGGTCCCATCATTA 3¢ conditions 60°C annealing temp RRAGD intron 3-exon 4 F: 5¢ GAAGGGGACTAGAATGTGTGC 3¢ 224 Standard reaction Standard cycling conditions; (DQ322249) R: 5¢ AGCAAATTCTCCCAGAGTGG 3¢ conditions 54°C annealing temp BAC end sequence F: 5¢ TGGCCATTAAGAGGAAAATTACA 3¢ 173 Standard reaction Standard cycling conditions; 199F3SP6 R: 5´ CATCAGAGAGCACCAAACTTCTT 3¢ conditions 60.5°C annealing temp .Gae n ..Rtshl:F Rothschild: M.F. and Grapes L. CTGF exons 4-5 F: 5¢ TTACCGACTGGAAGACACGTT 3´ 836 Standard reaction 94°C 2 min; 35 cycles of (DQ322250) R: 5¢ GTTGTAATGGCAGGCACAAG 3¢ conditions 94°C 30 sec, 58°C 30 sec, 72°C 1 min; 72°C 7 min ENPP1 exons 8-9 F: 5¢ ACCCTGAGTGGTACAAAGGA 3¢ 783 Standard reaction 94°C 2 min; 35 cycles of (DQ322251) R: 5¢ CATCTGATCCTGGCCAAAAA 3¢ conditions 94°C 30 sec, 60°C 30 sec, 72°C 1 min; 72°C 7 min SLC2A12 exon 5 F: 5¢ GAACTATGTGAAAAACAACATT 3¢ 581 Standard reaction Standard cycling conditions; (DQ322252) R: 5¢ TCCTAACCAGAACCCTGTCC 3¢ conditions 57°C annealing temp ALDH8A1 exon 7 F: 5¢ CATGCTTCCCACGGTGATA 3¢ 252 Standard reaction Standard cycling conditions; (DQ322253) R: 5¢ GCAAGGTTCAGCTCTCTGATG 3¢ conditions 60°C annealing temp BAC end sequence F: 5¢ AATTTTTCCACACTACCCTGCT 3¢ 875 Standard reaction Standard cycling conditions; bE1I10T7 R: 5¢ ATTTCAATGGAAGAGTGTTGCTC 3¢ conditions 54°C annealing temp Genomic clone F: 5¢ CCCAAAAGAGATGTATGTATAAAAGGA 3¢ 544 Standard reaction Standard cycling conditions; INE

CL349415 R: 5¢ TCACTCACATCAGCAAGATGG 3¢ conditions 60.5°C annealing temp M

Genomic clone F: 5¢ TCTTGTGGGATCTGACACCT 3¢ 586 Standard reaction Standard cycling conditions; APPING CL353852 R: 5¢ ACAGTGTGGGTGGCATGTAA 3¢ conditions 54°C annealing temp JAK2 exon 13-intron 13 (DQ322254) F: 5¢ TGGGCCATGCATTTTCTAGT 3¢ 867 Standard reaction 94°C 2 min; 35 cycles of R: 5¢ TTAGAAGTATATTAAACAATAA conditions 94°C 30 sec, 56°C 30 sec, QTL TCTATACCCGTGA 3¢ 72°C 1 min; 72°C 7 min FOR aStandard reaction: 1X GoTaq PCR Buffer (Promega, Madison, WI); 2 pmol dNTPs; 2.5 pmol each primer; 5U GoTaq polymerase (Promega). b Standard cycling conditions: 94°C for 2 min; 35 cycles of 94°C for 30 sec, annealing temp for 30 sec, 72°C for 30 sec; 72°C for 5 min. L OIN E YE A E AND REA F TESON ATNESS S 1 SSC L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1 661

Table 3. Polymorphisms identified, restriction enzymes, and resulting fragment sizes for PCR and PCR-RFLP tests for SSC 1 Gene/Sequence (GenBank accession No.) Polymorphisma Restriction enzyme Fragment sizes (bp) PHF3 exon 16 (DQ322274) T119C AvaII T allele 153, 69, 15 C allele À 101, 69, 52, 15 KCNQ5 exon 14 (DQ322246) C388T BsaI T allele À 253, 146 C allele À 221, 146, 32 PGM3 intron 5 exon 6 (DQ322247) C1016G ApoI C allele À 203 À G allele À 171, 32 RRAGD intron 3-exon 4 (DQ322249) C90T BsrI C allele À 224 T allele À 173, 51 BAC end sequence 199F3SP6 TGTTAGGTG (+) / ()) TGTTAGGTGÀ (+) allele 173 ()) allele 162 À CTGF exons 4-5 (DQ322250) C668T BsgI T allele À713, 123 C allele À 653, 123, 60 ENPP1 exons 8-9 (DQ322251) T649C NlaIII T allele À 635, 148 C allele À 504, 148, 131 SLC2A12 exon 5 (DQ322252) G233T AccI G allele À 581 T allele À 347, 234 ALDH8A1 exon 7 (DQ322253) G31A FokI A allele À 252 G allele À 208, 44 BAC end sequence bE1I10T7 T378C NlaIII C allele À 842 T allele À 463, 79, 34 Genomic clone CL349415 G378A TaqI A allele À 545 G allele À 376, 169 Genomic clone CL353852 G388A NlaIII G allele À 477, 96, 13 A allele À 390, 96, 87, 13 JAK2 exon 13-intron 13 (DQ322254) G831T NmuCI T allele À 498, 369 G allele À 462, 369, 36 À aPolymorphism location is given as the base position within the sequence listed in the adjacent column. The more frequent allele in the Berkshire sires is listed first. qtl.cap.ed.ac.uk/; Seaton et al. 2002). Regression compare results for all traits and limit computa- models for each trait were those used by Malek et al. tional requirements, chromosome-wise significance (2001a) and included sex and year-season as fixed thresholds were derived based on five traits (loin eye effects and live weight as a covariate. Chromosome- area, tenth rib backfat, last rib backfat, cholesterol, wise significance thresholds for the resulting F and marbling), similar to Malek et al. (2001a). statistic of a single trait were obtained using 5000 Genotypes of each gene/sequence were tested for random permutations of the data in QTLExpress. To association with LEA and the fatness traits of

a Table 4. Allele frequencies in BY F0 individuals for polymorphisms on SSC 1 selected for genotyping

Allele frequencies BY F0 individuals Gene/Sequence (GenBank accession No.) Polymorphismb B sires Y dams PHF3 exon 16 (DQ322274) T119C T 0.5; C 0.5 T 0.83; C 0.17 KCNQ5 exon 14 (DQ322246) C388T T 0.25; C 0.75 T 0.67; C 0.33 PGM3 intron 5 exon 6 (DQ322247) C1016G C 1.0; G 0.0 C 0.67; G 0.33 RRAGD intronÀ 3-exon 4 (DQ322249) C90T C 0.0; T 1.0 C 0.52; T 0.48 BAC end sequence 199F3SP6 TGTTAGGTG (+) / ()) TGTTAGGTG 1.0; )0.0 TGTTAGGTG 0.08; )0.92 CTGF exons 4-5 (DQ322250) C668T T 0.0; C 1.0 T 0.22; C 0.78 ENPP1 exons 8-9 (DQ322251) T649C T 0.75; C 0.25 T 0.72; C 0.28 SLC2A12 exon 5 (DQ322252) G233T G 0.75; T 0.25 G 1.0; T 0.0 ALDH8A1 exon 7 (DQ322253) G31A A 1.0; G 0.0 A 0.17; G 0.83 BAC end sequence bE1I10T7 T378C T 1.0; C 0.0 T 0.79; C 0.21 Genomic clone CL349415 G378A A 0.0; G 1.0 A 0.17; G 0.83 Genomic clone CL353852 G388A G 0.75; A 0.25 G 0.63; A 0.37 JAK2 exon 13-intron 13 (DQ322254) G831T T 0.5; G 0.5 T 0.0; G 1.0 aBY: Three-generation Berkshire · Yorkshire population. bPolymorphism location is given as the base position within the sequence listed in the adjacent column. The more frequent allele in the Berkshire sires is listed first. 662 L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1

interest in BY F2 individuals using the MIXED pro- cedure of SAS (SAS Institute, Cary, NC). The model fit for all traits included the fixed effects of sex, slaughter date, and genotype, the random effect of dam, and live weight as a covariate. Differences be- tween genotypic least-squares means estimated from this model for each trait were tested for significance using pair-wise t-tests under the ‘‘pdiff’’ option in the SAS software.

Results Linkage mapping results. The new linkage map of SSC 1 in BY contains 25 candidate genes and mi- crosatellite markers and is 146 cM long (Fig. 1). This linkage map is in agreement with that of Thomsen et al. (2004), which contained only 10 markers, al- though it is approximately 20 cM longer. Genotypes were checked for errors as a source of map expansion but none could be found. The length of the most comprehensive, publicly available linkage map of SSC 1 from USDA Meat Animal Research Center is 153 cM, indicating that the BY map of SSC 1 is within the expected range. While gene order on SSC 1p was inverted compared to HSA 6q in a broad sense, the order of PHF3, KCNQ5, and PGM3 was similar to that on HSA 6q (Fig. 1). Relative distances between genes differed in pig compared to human. The most interesting example involves ENPP1 and CTGF, which are separated by less than 60 kb in humans but were found to be nearly 8 cM apart in the porcine linkage map. However, the position of BAC end sequence bE1I10T7 could not be distin- guished from CTGF on the BY linkage map, and bE1I10T7 has greater than 80% identity with human sequence spanning the region between ENPP1 and CTGF. Eight candidate genes/genomic clones/BAC end sequences (ALDH8A1, SLC2A12, CTGF, ENPP1, PHF3, CL353852, CL349415, BAC end bE1I10T7) were mapped to the region originally shown in Malek et al. (2001a) to contain QTL for LEA and TENTHRIB. In addition, JAK2 was mapped to a region suggestive for a LEA QTL (Malek et al. Fig. 1. Linkage map of SSC 1 in three-generation Berkshire · 2001a). Yorkshire population. For newly mapped genes and se- quences, the comparative human map position is listed in QTL mapping results. Position estimates for parentheses. multiple QTL on SSC 1 were determined using the new linkage map and genotypic data. However, few of the QTL positions were markedly changed rela- location of S0008 (32.9 cM), with F values of 8.66 and tive to Malek et al. (2001a). The positions of the 4.20, respectively (Figs. 2 and 3). The position of the LEA, TENTHRIB, and AVBFAT QTL were originally TENTHRIB QTL was estimated at 34 cM with an F 29 cM, 1.6 cM downstream from microsatellite value of 8.24 (Fig. 2). The AVBFAT QTL no longer S0008, with F values of 10.34, 11.32, and 6.79, reached the 5% chromosome-wise significance respectively. The new positions of the LEA and threshold with the new markers incorporated AVBFAT QTL were estimated at 33 cM, nearly at the (Fig. 3). However, the previously suggestive MARB L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1 663

Fig. 3. F-statistic curves indicating QTL position estimates for three backfat measures (lumbar, LUMBAR; last rib, Fig. 2. F-statistic curves indicating QTL position estimates LASTRIB; average, AVBFAT) and total lipid percent for loin eye area (LEA) and tenth-rib backfat (TENTHRIB) (TOTLIPPR), cholesterol (CHOLES), and marbling score on SSC 1. The x axis indicates relative position on SSC 1 in (MARB) on SSC 1. The x axis indicates relative position on cM. The y axis indicates the F statistic. Arrows on the x SSC 1 in cM. The y axis indicates the F statistic. Arrows on axis indicate a marker position. The positions of micro- the x axis indicate a marker position. The position of mi- satellite marker S0008 and the janus kinase 2 gene (JAK2) crosatellite marker S0008 is included. The chromosome- are included. The chromosome-wise 5% and 1% signifi- wise 5% and 1% significance thresholds are indicated by a cance thresholds are indicated by a solid line and a dashed solid line and a dashed line, respectively. line, respectively.

QTL surpassed the 5% chromosome-wise signifi- (Table 5). For SLC2A12 and JAK2, the alleles asso- cance threshold with new position estimates at 29 ciated with greater loin eye area (Table 5) were and 47 cM with F values of 7.6 and 7.1, respectively unique to the Berkshire grandparents (Table 4). For (Fig. 3). Suggestive evidence for TOTLIPPR, LUM- the RRAGD C90T SNP, the C allele was unique to BAR, and CHOLES QTL persisted in the area near the Yorkshire grandparents. Interestingly, mean S0008 (Fig. 3). Also, a new, suggestive QTL for LEA LEA for the RRAGD TT and CC genotypes were was identified, spanning approximately 40 cM be- nearly identical, while the mean of the CT geno- tween microsatellite markers S0312 and SW974 type was significantly lower than either of the two (Fig. 2). homozygotes (Table 5). The LEA and TENTHRIB QTL previously span- Genomic clone CL349415 G378A and JAK2 ned 16 and 20.2 cM, respectively, in the BY. With the G831T genotypes were found to be significantly new marker data, the LEA QTL span remained 16 associated with TENTHRIB in BY F2s(p < 0.05, cM and the TENTHRIB QTL span was reduced to 18 Table 6). The T allele of JAK2 G831T SNP cM (Fig. 2). The QTL positions were not greatly re- was derived from Berkshire grandsires (Table 4) and fined likely because of the limited amount of was associated with decreased TENTHRIB. Clone recombination available in F2 individuals. CL349415 and ALDH8A1 G31A genotypes were significantly associated with AVBFAT and LUMBAR Association analysis results. Each polymor- (p < 0.05, Table 6). Alleles of the CL349415 G378A phism was tested for association with LEA and the SNP and ALDH8A1 G31A SNP, contributed only by fatness traits measured on the BY F2 individuals. For Yorkshire granddams (Table 4), were associated with the SLC2A12 G233T SNP, the TT genotype was not increased backfat for all three measures (Table 6). found in the BY F2 individuals and so was not in- For AVBFAT, genotypes from ENPP1 T649C SNP cluded in association analyses. For BAC end se- also showed significant association (p < 0.06) in F2 quence bE1I10T7 T378C SNP, the CC genotype was individuals (Table 6). excluded from analyses because only 11 F2 individ- Numerous genes and sequences showed signifi- uals had this genotype. The AA genotype for cant association with total lipid percent and mar- CL349415 G378A SNP was not included in the bling in the BY F2 individuals (Table 7). The alleles analyses because only 10 BY F2 animals had this associated with increased MARB for ALDH8A1 genotype. G31A, CTGF C668T, and BAC end sequence Significant associations (p < 0.05) with LEA bE1I10T7 T378C were inherited from Yorkshire were found for the following polymorphisms: granddams (Table 4). The alleles associated with SLC2A12 G233T, RRAGD C90T, and JAK2 G831T increased MARB and TOTLIPPR for RRAGD C90T 664 L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1

a b c 2 d Table 5. Association of SLC2A12, RRAGD, and JAK2 genotypes with loin eye area (cm ) in BY F2 individuals Gene Polymorphism Genotypic least-squares meanse (SE) and genotype F probf SLC2A12 G233T 35.19 (0.60) c 36.42 (0.67) d 0.018 GG GT RRAGD C90T 37.00 (0.77) a 35.74 (0.57) b c 36.70 (0.62) d 0.034 CC CT TT JAK2 G831T 37.61 (0.89) e 36.33 (0.60) c 35.27 (0.59) d f 0.009 TT TG GG aSLC2A12: solute carrier family 2, member A12. bRRAGD: ras-related GTP binding D. cJAK2: janus kinase 2. dBY: Three-generation Berkshire · Yorkshire population. eSignificance levels for pairwise t-tests of genotypic mean differences: a, b = p < 0.10; c, d = p < 0.05; e, f = p < 0.01. fProbability for the F-test of differences between genotypic means. and the BAC end sequence 199F3SP6 indel (Table 7) Discussion were also inherited from Yorkshire granddams (Table 4). In addition, the T alleles from SLC2A12 QTL position estimates. Although eight additional G233T and JAK2 G831T SNPs coming from Berk- markers were added to the region originally identi- shire grandsires (Table 4) were related to decreased fied as containing LEA and TENTHRIB QTL by MARB and, in the case of JAK2, TOTLIPPR as well Malek et al. (2001a), the QTL positions remained (Table 7). Similar to AVBFAT (Table 6), animals similar with little refinement in span. This is likely heterozygous for ENPP1 T649C SNP had lower to the result of multiple factors, including (1) lack of TOTLIPPR and MARB than either homozygote recombination events in the QTL area in F2 indi- group (Table 7). viduals; (2) Berkshire grandsires and Yorkshire Finally, genotypes from CL352852 G388A, granddams had unique genotypes for microsatellite CL349415 G378A, SLC2A12 G233T, and BAC end S0008; and (3) allele frequencies for markers closest sequence bE1I10T7 T378C SNPs were associated to microsatellite S0008 were not dramatically dif- with CHOLES (Table 8). The A allele of CL349415 ferent between the Berkshire and Yorkshire grand- G378A and C allele of bE1I10T7 T378C, derived parents (Table 4). Thus, the genotype differences from Yorkshire granddams (Table 4), were associated between breeds for S0008 likely have a strong impact with increased cholesterol (Table 8). Conversely, the on the LEA and TENTHRIB QTL position estimates T allele of SLC2A12 G233T, derived from Berkshire due to the reliance of least-squares regression inter- grandsires (Table 4), was associated with decreased val mapping on between-breed genetic differences. It cholesterol (Table 8). is also possible that the true causative genes/muta-

Table 6. Association of genomic clone CL349415, JAK2,a ALDH8A1,b and ENPP1c genotypes with tenth-rib backfat (TENTHRIB, d cm), average backfat (AVBFAT, cm), and lumbar backfat (LUMBAR, cm) in BY F2 individuals Gene/Sequence Polymorphism Trait Genotypic least-squares meanse (SE) and genotype F probf CL349415 G378A TENTHRIB 3.21 (0.08) c 3.02 (0.06) d 0.022 AVBFAT 3.32 (0.08) c 3.18 (0.05) d 0.042 LUMBAR 3.61 (0.09) c 3.43 (0.06) d 0.036 AG GG JAK2 G831T TENTHRIB 2.91 (0.12) c 3.07 (0.07) a 3.18 (0.07) b d 0.049 TT TG GG ALDH8A1 G31A AVBFAT 3.21 (0.06) e 3.27 (0.06) c 3.41 (0.08) d f 0.019 LUMBAR 3.43 (0.07) c g 3.56 (0.07) c* d 3.74 (0.09) d* h 0.002 AA AG GG ENPP1 T649C AVBFAT 3.34 (0.06) c 3.20 (0.06) d 3.22 (0.11) 0.052 TT TC CC aJAK2: janus kinase 2. bALDH8A1: aldehyde dehydrogenase 8 A1. cENPP1: ectonucleotide pyrophosphatase 1. dBY: Three-generation Berkshire · Yorkshire population. eSignificance levels for pair-wise t-tests of genotypic mean differences: a, b = p < 0.10; c, d, and c*, d* = p < 0.05; e, f = p < 0.01; g, h = p < 0.001. fProbability for the F-test of differences between genotypic means. L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1 665

Table 7. Association of genomic clone CL353852, ALDH8A1,a SLC2A12,b CTGF,c BAC end sequence bE1I10T7, ENPP1,d PHF3,e f g h RRAGD, BAC end sequence 199F3SP6, and JAK2 genotypes with total lipid percent (TOTLIPPR) and marbling score (MARB) in BY F2 individuals Gene/Sequence Polymorphism Trait Genotypic least squares meansi (SE) and genotype F probj CL353852 G388A TOTLIPPR 3.06 (0.15) 2.92 (0.14) c 3.24 (0.18) d 0.065 GG GA AA ALDH8A1 G31A MARB 3.58 (0.07) a c 3.70 (0.07) b 3.81 (0.09) d 0.031 AA AG GG SLC2A12 G233T MARB 3.72 (0.07) c 3.56 (0.09) d 0.042 GG GT CTGF C668T MARB 4.06 (0.22) c 3.81 (0.11) e 3.52 (0.09) d f 0.003 TT TC CC bE1I10T7 T378C MARB 3.78 (0.09) c 3.60 (0.06) d 0.046 CT TT ENPP1 T649C TOTLIPPR 3.18 (0.15) c 2.89 (0.15) c* d 3.41 (0.24) d* 0.007 MARB 3.72 (0.07) a* 3.59 (0.08) a b* 3.83 (0.13) b 0.076 TT TC CC PHF3 T119C MARB 3.91 (0.11) a e 3.71 (0.10) b c 3.32 (0.16) d f 0.010 TT TC CC RRAGD C90T TOTLIPPR 3.13 (0.19) a 3.12 (0.14) c 2.84 (0.15) b d 0.035 MARB 3.72 (0.10) a 3.73 (0.06) c e 3.51 (0.07) b f 0.005 CC CT TT 199F3SP6 TGTTAGGTG (+) / ()) TOTLIPPR 2.97 (0.15) c c* 3.25 (0.16) d 3.46 (0.25) d* 0.013 MARB 3.59 (0.07) g 3.82 (0.07) h 3.84 (0.13) 0.001 (+)(+) (+)())())()) JAK2 G831T TOTLIPPR 2.87 (0.22) a 2.92 (0.15) e 3.27 (0.15) b f 0.014 MARB 3.42 (0.12) a e 3.65 (0.07) b 3.71 (0.07) f 0.079 TT TG GG aALDH8A1: aldehyde dehydrogenase 8 A1. bSLC2A12: solute carrier family 2, member A12. cCTGF: connective tissue growth factor. dENPP1: ectonucleotide pyrophosphatase 1. ePHF3: Ph.D. finger protein 3. fRRAGD: ras-related GTP binding D. gJAK2: janus kinase 2. hBY: Three-generation Berkshire · Yorkshire population. iSignificance levels for pairwise t-tests of genotypic mean differences: a, b and a*, b* = p < 0.10; c, d = p < 0.05; e, f = p < 0.01; g, h = p < 0.001. jProbability for the F test of differences between genotypic means. tions responsible for the observed LEA and S0008 and ENPP1 could not be identified, and addi- TENTHRIB QTL effects are tightly linked to S0008. tional information from markers in this region may Interestingly, no other LEA or TENTHRIB QTL has help to further refine the positions of the LEA and been detected in this region in other studies to date, TENTHRIB QTL in the BY. implying that these QTL may be unique to the The chromosome-wise significant AVBFAT QTL Berkshire breed. Unfortunately, markers between detected in Malek et al. (2001a) no longer reached

Table 8. Association of genomic clones CL352852 and CL349415, SLC2A12,a and BAC end sequence bE1I10T7 genotypes b with cholesterol in BY F2 individuals Gene/Sequence Polymorphism Genotypic least-squares meansc (SE) and genotype F probd CL353852 G388A 58.30 (0.70) a 58.39 (0.61) c 56.25 (0.92) b d 0.079 GG GA AA CL349415 G378A 59.43 (0.92) a 57.56 (0.50) b 0.051 AG GG SLC2A12 G233T 59.44 (0.57) c 57.54 (0.76) d 0.025 GG GT bE1I10T7 T378C 59.62 (0.88) c 57.57 (0.51) d 0.027 CT TT aSLC2A12: solute carrier family 2, member A12. bBY: Three-generation Berkshire · Yorkshire population. cSignificance levels for pairwise t-tests of genotypic mean differences: a, b = p < 0.10; c, d = p < 0.05. dProbability for the F-test of differences between genotypic means. 666 L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1 significance with the addition of new markers, microsatellite genotypes were tested for association similar to results from Thomsen et al. (2004). How- with LEA and fatness traits. The recoded genotypes ever, a sharp peak in the F-statistic curve for AVB- for S0008 of the two Berkshire grandsires were 23 FAT remained at 34 cM, near the position of S0008 and 23 (freq. allele 2 = 0.50; freq. allele 3 = 0.50). The (Fig. 3), similar to the position of the AVBFAT QTL recoded genotypes of the nine Yorkshire granddams identified in Malek et al. (2001a). Again, the small were six 11, two 12, and one 22 (freq. allele 1 = 0.78; between-breed genotypic differences of markers freq. allele 2 = 0.22). Results showed that S0008 closest to microsatellite S0008 may have impacted genotypes were significantly associated with every detection of the AVBFAT QTL. Another suggestive fatness trait tested except CHOLES (p < 0.1). How- QTL for AVBFAT was identified in a European wild ever, S0008 genotypes were not associated with LEA. boar · Large White (Yorkshire) cross near microsat- In fact, with S0008 removed from the QTL analysis, ellite SW1514 (Knott et al. 1998), which is telomeric the F-statistic curve for LEA barely reached the 5% to SW1515 on SSC 1p, the marker at 0 cM on the BY chromosome-wise significance threshold in the linkage map (Fig. 1). same region and peaked at the same position at With the addition of markers between micro- ENPP1. This supports the hypothesis that the caus- satellites S0312 and SW974, a suggestive QTL for ative mutation responsible for the observed QTL LEA emerged that was not observed in either Malek effect is tightly linked to S0008 or that the between- et al. (2001a) or Thomsen et al. (2004) (Fig. 2). Al- breed differences in S0008 F0 genotypes have a strong though the F statistic does not reach chromosomal impact on detection of the LEA QTL. significance in this region, three new pronounced Although the F statistic for LEA does not reach peaks were observed (Fig. 2). Evidence of QTL significance in the region between S0312 and SW974, affecting loin muscle length, area, and weight has there is a pronounced peak at the same position as been observed on SSC 1q in other studies (Milan et JAK2 at 93 cM. The genotypes for JAK2 G831T SNP al. 2002; Rohrer and Keele 1998; Sato et al. 2003; Su were significantly associated with LEA (p < 0.009), et al. 2004), but their positions were telomeric to the with alleles originating from the Berkshire breed suggestive LEA QTL found in BY. associated with increased LEA. However, this SNP is The F-statistic curve for MARB (Fig. 3) was sig- located in intron 13 of the JAK2 gene, and at this time nificantly changed compared with that of Malek no functional differences are known to exist between et al. (2001b), in which the F statistic near S0008 the alleles. It is possible that this SNP is in linkage barely crossed the 5% chromosome-wise significance disequilibrium with a causative mutation in JAK2 or threshold. There now seem to be three to four MARB a gene nearby that is responsible for the observed QTL on SSC 1, with two of them estimated at 29 and suggestive LEA QTL. 47 cM. The F values at these two positions are comparable to those at the new LEA and TENTHRIB Breed of origin allelic effects. As shown in QTL positions (Fig. 2), which previously reached Malek et al. (2001a), alleles originating from the genome-wise significance (Malek et al. 2001a). Berkshire breed were shown to increase LEA and decrease TENTHRIB for the QTL detected in the Candidate gene positions and trait associa- region of S0008. These would be considered cryptic tions. Candidate genes SLC2A12, CTGF, and ENPP1, alleles because the Berkshire breed was chosen as it along with genomic clones CL353852, CL349415, and was expected to be fatter. Similar results were ob- BAC end sequence bE1I10T7 were positioned in the served in this study for genes mapped to the LEA and region significant for LEA and TENTHRIB QTL. Of fatness traits’ QTL regions. For SLC2A12 G233T those, SLC2A12 showed significant association with SNP, the T allele was unique to the Berkshire LEA, and CL349415, closest to S0008, showed signif- grandparents and was associated with increased LEA icant association with TENTHRIB. In the region of and decreased marbling and cholesterol. For many of the significant MARB QTL at 29 cM and the sugges- the other genes/sequences mapped to this region tive AVBFAT, CHOLES, TOTLIPPR, and LUMBAR (ALDH8A1, CTGF, BAC end sequence bE1I10T7, QTL, candidate genes ALDH8A1, SLC2A12, CTGF, genomic clone CL349415), alleles unique to the ENPP1, and PHF3 were mapped, along with genomic Yorkshire breed were consistently associated with clones CL353852, CL349415, and BAC end sequence increased tenth-rib and average backfat measures, as bE1I10T7. All of these showed significant association well as increased lumbar fat, marbling, and choles- with at least one of the five fatness traits, six of them terol. All of these genes/sequences map within a 5.5- with two or more traits. cM region and indicate that there is strong between- Because the BY F0 grandsires had unique geno- breed disequilibrium in the F2 population in this types for S0008 compared with the granddams, the region, i.e., there is a distinct Yorkshire haplotype L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1 667 and a distinct Berkshire haplotype in this region. Glenn, Marcos Ramos, Nguyet Thu Nguyen, and Future analyses of haplotypes for the eight SNPs J. D. Loy for their technical assistance. located in the LEA and fatness QTL region may help to further describe the between-breed differences. References The linkage analysis results indicate that additional recombination events in the population (i.e., addi- 1. Brigstock DR (1999) The connective tissue growth tional generations) are necessary to further refine the factor/cysteine-rich 61/nephroblastoma overexpressed positions of the LEA and fatness traits QTL and (CCN) family. Endocr Rev 20, 189 206 À identify the underlying causal mutation(s). 2. Ciobanu DC, Bastiaansen JW, Lonergan SM, Thomsen Interestingly, the T allele of JAK2 G831T SNP, H, Dekkers JC, et al. (2004) New alleles in calpastatin originating from the Berkshire breed in this popula- gene are associated with meat quality traits in pigs. J Anim Sci 82, 2829 2839 tion, was also consistently associated with increased À LEA and decreased TENTHRIB, MARB, and TOT- 3. Fronicke L, Chowdhary BP, Scherthan H, Gustavsson I (1996) A comparative map of the porcine and human LIPPR, even though the potential LEA QTL is the genomes demonstrates ZOO-FISH and gene mapping- only reasonable suggestive QTL observed at the based chromosomal homologies. Mamm Genome 7, position of JAK2. 285 290 4. GreenÀ P, Falls K, Crooks S (1990) Documentation for Conclusions CRIMAP, version 2.4 St. Louis, MO, Washington University School of MedicineGreen P, Falls K, Crooks A limited number of genes/sequences (8) were S (1990) Documentation for CRIMAP, version 2.4 (St. mapped to the original LEA and TENTHRIB QTL Louis, MO: Washington University School of Medi- regions identified by Malek et al. (2001a) near S0008 cine) in this study. However, several of the genes in this 5. Hopkinson DA, Harris H (1968) A third phosphoglu- comutase locus in man. Ann Hum Genet 31, 359 367 region were significantly associated with fatness À traits and one was associated with LEA. 6. Joost HG, Thorens B (2001) The extended GLUT- Alleles were identified uniquely from the York- family of sugar/polyol transport facilitators: nomen- clature, sequence characteristics, and potential func- shire breed for ALDH8A1, CTGF, BAC end sequence tion of its novel members (review). Mol Membr Biol bE1I10T7, and genomic clone CL349415 that were 18, 247 256 associated with increased fatness; these genes/se- 7. Kim KS,À Larsen N, Short T, Plastow G, Rothschild MF quences mapped to a region of multiple suggestive (2000) A missense variant of the porcine melanocortin- and significant QTL for fatness traits. These alleles 4 receptor (MC4R) gene is associated with fatness, should be tested in commercial Yorkshire/Large growth, and feed intake traits. Mamm Genome 11, White-based lines for informativeness, size of ef- 131 135 À fects, and potential use in marker-assisted selection 8. Knott SA, Marklund L, Haley CS, Andersson K, programs to make genetic improvements in leanness Davies W, et al. (1998) Multiple marker mapping of of maternal lines. quantitative trait loci in a cross between outbred The known role of JAK2 in muscle cell differen- wild boar and large white pigs. Genetics 149, 1069 1080 tiation and proliferation makes it an interesting can- 9. LercheÀ C, Scherer CR, Seebohm G, Derst C, Wei AD, didate gene for LEA. In the BY population, JAK2 et al. (2000) Molecular cloning and functional expres- mapped to a region suggestive for LEA QTL. In addi- sion of KCNQ5, a potassium channel subunit that may tion, an allele unique to the Berkshire grandsires was contribute to neuronal M-current diversity. J Biol identified and significantly associated with increased Chem 275, 22395 22400 LEA and decreased fatness. The existing G831T SNP 10. Lin M, Napoli JL (2000)À cDNA cloning and expression should be tested in commercial lines to determine if of a human aldehyde dehydrogenase (ALDH) active the T allele is present in other breeds, has similar ef- with 9-cis-retinal and identification of a rat ortholog, ALDH12. J Biol Chem 275, 40106 40112 fects on LEA and fatness as observed in the BY F2, and À is suitable for marker-assisted selection programs to 11. Malek M, Dekkers JC, Lee HK, Baas TJ, Rothschild MF increase lean percentage in sire lines. (2001a) A molecular genome scan analysis to identify chromosomal regions influencing economic traits in the pig. I. Growth and body composition. Mamm Acknowledgments Genome 12, 630 636 12. Malek M, DekkersÀ JC, Lee HK, Baas TJ, Prusa K, et al. Funding for this work was provided by Sygen Inter- (2001b) A molecular genome scan analysis to identify national and PIC group, the Iowa Agriculture and chromosomal regions influencing economic traits in Home Economics Experiment Station, and Hatch the pig. II. Meat and muscle composition. Mamm Genome 12, 637 645 and State of Iowa funds. The authors thank Kimberly À 668 L. Grapes and M.F. Rothschild: FINE MAPPING QTL FOR LOIN EYE AREA AND FATNESS ON SSC 1

13. Meyre D, Bouatia Naji1 N, Tounian A, Samson C, 19. Seaton G, Haley CS, Knott SA, Kearsey M, Visscher Lecoeur C, et al. (2005)À Variants of ENPP1 are associ- PM (2002) QTL Express: mapping quantitative trait ated with childhood and adult obesity and increase the loci in simple and complex pedigrees. Bioinformatics risk of glucose intolerance and type 2 diabetes. Nat 18, 339 340 Genet 37, 863 867 20. SekiguchiÀ T, Hirose E, Nakashima N, Ii M, Nishimoto 14. Milan D, BidanelÀ JP, Iannuccelli N, Riquet J, Amigues T (2001) Novel G proteins, Rag C and Rag D, interact Y, et al. (2002) Detection of quantitative trait loci for with GTP-binding proteins, Rag A and Rag B. J Biol carcass composition traits in pigs. Genet Sel Evol 34, Chem 276, 7246 7257 705 728 21. Thomsen H, LeeÀ HK, Rothschild MF, Malek M, 15. OviloÀ C, Clop A, Noguera JL, Oliver MA, Barragan C, Dekkers JC (2004) Characterization of quantitative et al. (2002) Quantitative trait locus mapping for meat trait loci for growth and meat quality in a cross be- quality traits in an Iberian · Landrace F2 pig popula- tween commercial breeds of swine. J Anim Sci 82, tion. J Anim Sci 80, 2801 2808 2213 2228 16. Pizzuti A, Frittitta L, ArgiolasÀ A, Baratta R, Goldfine 22. Van LaereÀ AS, Nguyen M, Braunschweig M, Nezer C, ID, et al. (1999) A polymorphism (K121Q) of the human Collette C, et al. (2003) A regulatory mutation in IGF2 glycoprotein PC-1 gene coding region is strongly asso- causes a major QTL effect on muscle growth in the ciated with insulin resistance. Diabetes 48, 1881 1884 pig. Nature 425, 832 836 17. Robic A, Riquet J, Yerle M, Milan D, Lahbib MansaisÀ 23. Varona L, Ovilo C, ClopÀ A, Noguera JL, Perez-Enciso Y, et al. (1996) Porcine linkage and cytogeneticÀ maps M, et al. (2002) QTL mapping for growth and carcass integrated by regional mapping of 100 microsatellites traits in an Iberian by Landrace pig intercross: addi- on somatic cell hybrid panel. Mamm Genome 7, tive, dominant and epistatic effects. Genet Res 80, 438 445 145 154 18. RogatchevaÀ MB, Meyers S, He W, Larkin DM, Marron 24. WangÀ H, Yu M, Liu B, Yerle M, Wang Y, et al. (2004) BM, et al. (2006) Piggy-BACing the I: Mapping of the porcine JAK2, JAK3 and TYK2 genes Constructing a porcine physical map through com- using somatic cell and radiation hybrid panels. Anim parative genomics. Anim Genet (in press) Genet 35, 258 259 À