Introduction
For the last years, meat quality traits have increasingly attracted more attention in pig breeding because selection for high growth rate and lean meat deposition resulted in reduction of meat quality such as the increasing incidence of PSE (pale, soft and exudative) muscle and decreasing of intramuscular fat (IMF) content (Zuo et al., Reference Zuo, Xiong, Su, Deng, Zheng and Jiang2003). Use of marker-assisted selection (MAS) is especially interesting for meat quality traits, because improvement of meat quality is difficult using conventional selection methods since most traits of interest can only be measured after slaughter and, therefore, only information on relatives can be used for selection (Malek et al., Reference Malek, Dekkers, Lee, Bass and Rothchild2001). The successful application of MAS in the animal population will depend on the identification of major genes or tightly linked markers. The study of associations of candidate genes is a step for the knowledge of the genetic basis of productive traits, and compared with other genomic approaches (quantitative trait loci (QTL) detection) is potentially more easily and efficiently implemented in breeding programmes (Ovilo et al., Reference Ovilo, Fernandez, Rodriguez, Nieto and Silio2006).
The skeletal muscle glycogen synthase (GYS1) gene is a key enzyme of non-oxidative pathway of glucose metabolism that has been reported to be related to insulin resistance in non-insulin-dependent diabetic (NIDDM) patients (Rissanen et al., Reference Rissanen, Pihlajamaki, Heikkinen, Kekalainen, Mykkanen, Kuusisto, Kolle and Laakso1997; Shimomura et al., Reference Shimomura, Sanke, Ueda, Hanabusa, Sakagashira and Nanjo1997; Orho-Melander et al., Reference Orho-Melander, Almgren, Kanninen, Forsblom and Groop1999; Fenger et al., Reference Fenger, Poulsen, Beck-Nielsen and Vaag2000; Motoyama et al., Reference Motoyama, Emoto, Tahara, Komatsu, Shoji, Inaba and Nishizawa2003). The A/G mutation in intron 14 of the porcine GYS1 gene has been discovered, but has no effect on the content of glycogen in skeletal muscle (Te Pas et al., Reference Te Pas, Leenhouwers, Knol, Booij, Priem and Van der Lende2003). Preoxisomal acyl-CoA oxidase 1 (ACOX1) catalyses the beta-oxidation of very long-chain fatty acids, and thus plays an essential role in fatty acid degradation. Fan et al. Reference Fan, Pan, Chu, Lee, Kluckman, Usuda, Singh, Yeldandi, Rao, Maeda and Reddy(1996) found that homozygous ACOX1-null mice were viable, but growth retarded and infertile. Expression of the ACOX1 gene was significantly increased in male mice fed a high-fat diet, compared with a low-fat diet (Kim et al., Reference Kim, Sohn, Ahn, Lee, Lee and Lee2004). The porcine GYS1 and ACOX1 gene was physically mapped to SSC6 and SSC12, respectively (Cirera et al., Reference Cirera, Jorgensen, Sawera, Raudsepp, Chowdhary and Fredholm2003; Fontanesi et al., Reference Fontanesi, Davoli, Nanni Costa, Scotti and Russo2003), where significant QTL affecting meat quality traits have been reported (De Koning et al., Reference De Koning, Janss, Rattink, Van Oers, De Vries, Groenen, Van der Poel, De Groot, Brascamp and Van Arendonk1999; Clop et al., Reference Clop, Ovilo, Perez-Enciso, Cercos, Tomas, Fernandez, Coll, Folch, Barragan, Diaz, Oliver, Varona, Silio, Sanchez and Noguera2003). Therefore, those two genes have been suggested as the promising candidate genes for meat quality traits, given their roles in the glycogen synthesis and metabolism of fatty acids.
Compared with the Chinese fat-type Meishan pigs, the meat-type Large White pigs have a higher growth rate, higher feed to body weight conversion ratio and higher carcass lean percentage, but have lower IMF and inferior eating quality from the Chinese perspective. In order to seek for the major gene or marker underlying the important economic traits such as meat quality traits, a three generation resource family was established by Large White boars and Meishan dams in our lab. The objective of the current study was to detect FokI restriction endonuclease variants in the seventh intron of the GYS1 gene and PstI restriction endonuclease variants in the ninth intron of the ACOX1 gene in the Large White × Meishan pig resource family, and to determine whether these were associated with variation in meat quality traits, and thus could contribute to breeding programmes.
Material and methods
Animals and data collection
Two F2 populations used in the association analysis were derived from the intercross of Large White and Meishan pigs. One population was formed by 140 F2, 28 F1 and 10 grandparent animals in 16 full-sib families, and the other population consisted of 194 F2, 26 F1 and 11 grandparent animals in 21 full-sib families. They were fed twice daily with diets formulated according to age under a standardised feeding regimen and free access to water. The average live weight at slaughter was 87.0 ± 7.07 kg. The F2 pigs were slaughtered at 2000 and 2003 following a common protocol (Xiong and Deng, Reference Xiong and Deng1999). Meat quality traits including pH of m. longissimus dorsi at post slaughter 45 min (pHLD), pH of m. biceps femoris at post slaughter 45 min (pHBF), pH of m. semipinalis capitis at post slaughter 45 min (pHSC), drip loss rate (DLR, %), water-holding capacity (WHC, %), meat colour value of m. longissimus dorsi (MCV1), meat colour value of m. biceps femoris (MCV2), meat marbling score of m. longissimus dorsi (MMS1), meat marbling score of m. biceps femoris (MMS2), IMF (%) and water moisture (WM, %) were measured according to the method of Xiong and Deng Reference Xiong and Deng(1999). Genomic DNA was prepared from blood samples using a standard phenol: chloroform extraction method.
PCR-RFLP genotyping
The DNA from the F2 pigs was used as template to perform PCR. According to the obtained cDNA sequences (Genbank accession number AY870324) and the exon/intron organisation of human GYS1 gene, PCR primers (forward: 5′- TAT GAG TTC TCC AAC AAG GGG -3′; reverse: 5′- GAT GAA GAA AGC AAC CAC TGT C -3′) were designed to amplify porcine GYS1 gene. According to the obtained cDNA sequences (Genbank accession number DQ842227) and the exon/intron organisation of human ACOX1 gene, PCR primers (forward: 5′- GGA AAT GAA CCC GAC CAG TA -3′; reverse: 5′- TGC GTC TCA GGA AGC AGT AAG -3′) were designed to amplify porcine ACOX1 gene. The reaction mixtures comprised 25 ng porcine genomic DNA as template, 0.25 μmol/l of each primer, 0.25 μmol/l of each dNTP, 1 × PCR buffer and 1 U Taq DNA polymerase (Biostar Internation, Toronto, Canada). The PCR amplifications were performed in 20 μl on a GeneAmp PCR system 9600 (Perkin Elmer, Foster City, CA, USA) with the following cycling parameters: 95°C initial denaturation for 4 min, 35 cycles of 95°C denaturation for 45 s, 60°C (GYS1) or 61°C (ACOX1) annealing for 45 s, and 72°C extension for 45 s. A final extension was performed at 72°C for 10 min. 8 μl of PCR amplifications obtained with above primer pairs were digested with 10 U FokI and PstI restriction enzyme (TaKaRa, Dalian, China), electrophoresed on 1.5% (GYS1) or 2.5% (ACOX1) agarose gel in 1 × TAE buffer and stained with 0.5 μg/ml ethidium bromide.
Analysis
The effects of single genotypes on the traits studied were analysed by the least-squares method as applied in the general linear model (GLM) procedure of Statistical Analysis Systems Institute (SAS; 2000) according to the following statistical model:
where Tijkl is the observed values of a given trait, μ is the overall mean, Si is effect of sex (i = 1 for male or 2 for female), Yj is the effect of year (j = 1 for year 2000 or 2 for year 2003), Gk is the effect of genotype (k = AA, AB and BB), Fl is the effect of family (l=37), bijkl is the regression coefficient of the slaughter age for meat quality traits, Xijkl is the slaughter age, and eijkl is the random residual. Both additive and dominance effects were estimated using the REG procedure of SAS version 8.0, where the contrast coefficients for the additive effect were denoted as −1, 0 and 1 for AA, AB and BB, respectively, and the contrast coefficients for the dominance effect were denoted as 1, −1 and 1 for AA, AB and BB, respectively (Liu, Reference Liu1998).
In order to determine whether the significant associations of GYS1 gene were due to the marker or due to other co-inherited blocks, the genetic mapping were performed by CRI-MAP software version 2.4 (Green et al., Reference Green, Falls and Crooks1990) using genotypes for FokI PCR-RFLP and four microsatellite markers (SW1302,SW1473,S0121 and SW322) information available on SSC6 (Zuo et al., Reference Zuo, Xiong, Deng, Su, Wang, Lei, Li, Jiang and Zheng2005; Zhang et al., Reference Zhang, Xiong, Zuo, Lei, Jiang, Li, Zheng and Li2007). Least-square regression interval mapping as described by Haley et al. Reference Haley, Knott and Elsen(1994) was used for QTL detection. QTL analysis was carried out on the Internet (http://qtl.cap.ed.ac.uk). For the meat quality traits, sex, year and full-sib family were included as fixed effects with the slaughter date as a covariate.
Results
Phenotype, genotype and allele frequencies
Phenotypic means, standard deviations (s.d.) and coefficients of variation (CV) for meat quality traits were given in Table 1. From this table, it was found that the CVs ranged from 1.04 to 31.80%. The traits such as DLR, meat colour value, IMF content had the higher CV, while the other traits showed relatively lower variation. The distribution of genotypic and allelic frequencies in the pig population is given in Table 2. Overall, the allele frequencies showed almost equal proportion of alleles for these two genes except the GYS1 allele frequency in population 2003. For the analysis of the combined genotypes, the expected frequencies of the genotypes and their combinations were calculated by simple allele counting. All of the 9 (32) theoretically possible combinations of two individual genotypes were observed and most of the combined genotypes found clearly followed the Hardy–Weinberg equilibrium (Table 3), showing that both genes, GYS1 and ACOX1, are independent which seems reasonable as they are located in different chromosomes, GYS1 in chromosome 6 and ACOX1 in chromosome 12.
Phenotypic means, standard deviation (s.d.) and coefficients of variation (CV) for meat quality traits
Distribution of genotypic and allelic frequencies in the resource population
Expected frequencies of combined genotypes and comparison of observed and expected numbers of animals
GYS1 gene effects
The results of the GLM analysis of association between the GYS1 gene and meat quality performance in pigs are summarised in Table 4. Differences among FokI genotypes were only significant for pH of m. semipinalis capitis. No differences were detected for other meat quality traits. The AB pigs had significantly higher pH of m. semipinalis capitis than AA pigs, but there was no significant difference as compared with BB. This locus seemed to be significantly dominant in action and the dominance effects were −0.012 ± 0.007 for pH of m. semipinalis capitis.
Statistical analysis of association between GYS1 FokI-RFLP genotypes with meat quality traits
a,bWithin a row, means marked with different superscript letters are significantly different (P < 0.05).
Linkage analysis showed that the FokI marker was significantly linked with the selected markers on SSC6. Two-point linkage analysis revealed linkage to microsatellite markers SW1302 (recombination fraction = 0.25; LOD = 7.02) and SW1473 (recombination fraction = 0.20; LOD = 9.76) on the sex-average linkage. The most probable order produced by Build option is as follows (Kosambi cM; sex-average values): SW1302–26.9 – GYS1–21.3 – SW1473–23.3 – S0121–27.8 – SW322. However, the QTL analysis showed that there was no significant QTL for meat quality traits.
ACOX1 gene effects
The results of the GLM analysis of association between the ACOX1 gene and traits in pigs were summarised in Table 6. Significant effects of ACOX1 alleles were found on meat colour value of m. longissimus dorsi (MCV1), meat colour value of m. biceps femoris (MCV2), meat marbling score of m. longissimus dorsi (MMS1), and meat marbling score of m. biceps femoris (MMS2). This locus seemed to be significantly over-dominant in action for meat colour value and meat marbling score, and pigs with genotype AB had significantly lower meat colour value, but higher meat marbling score as compared those with genotype BB. However, other important meat quality traits, such as water moisture and pH, were scarcely affected by ACOX1 alleles.
Discussion
The present work was based on the analysis of a F2 segregation population derived from the intercross of Chinese Meishan and Large White pigs. Due to the significant phenotypic difference between Large White and Meishan pigs, the F2 segregation population showed great variation in meat quality traits. The frequency of genotypes AA, AB and BB for ACOX1 gene conformed to 1 : 2 : 1 Mendelian segregation, because the A allele was fixed in the founder Large White pigs and B allele fixed in the founder Chinese Meishan pigs.
Ultimate pH of pork is the most commonly used trait to assess pork quality. A higher level of acidity within the muscle (lower pH) causes muscle protein to denature and lose the ability to hold water. Therefore, meat with higher pH will tend to have more desirable characteristics (Malek et al., Reference Malek, Dekkers, Lee, Bass and Rothchild2001). The pH of pork is correlated with glycogen content and glycolysis in postmortem muscle. GYS1 is a key enzyme of non-oxidative pathway of glucose metabolism and has been shown to strongly influence muscle glycogen content and glycolysis in skeletal muscle. This study demonstrates a significant genotype effect of GYS1 on pHSC, which is in good agreement with physiological functions of GYS1 gene. In order to discriminate between causal and neutral mutations in the F2 design, we made use of the information provided by the neutral genetic markers located in the adjacent region of this mutation and conducted the QTL analysis. The peak of QTL for pH was located around the mutation of GYS1 gene and F-ratio of QTL for pHSC was higher than that of pHLD and pHBF (Table 5). Therefore, the size of the GYS1 genotype effects on the different muscle pH in the association study may mainly depend on the linked QTL effects. However, no significant QTL for pHSC was detected on this region although significant association of this mutation with pHSC was found in the single marker association. It may be due to several reasons. (1) The F2 design has a great power to detect QTL provided by linkage disequilibrium, and also makes it difficult to discriminate between causal and neutral mutations. Therefore, a high percentage of false positives can be expected (Varona et al., Reference Varona, Gomez-Raya, Rauw and Noguera2005). (2) The number of animals we detected is not enough to demonstrate the true event. (3) The significant effects of GYS1 gene were balanced by another QTL or marker with negative effects, as the QTL region was broad. All these need further verification.
Estimated effects (mean ± s.e.) of QTL for pH on pig chromosome 6
Statistical analysis of association between ACOX1 PstI-RFLP genotypes with meat quality traits
a,bWithin a row, means marked with different superscript letters are significantly different (P < 0.05).
ACOX1 is the first and rate-limiting enzyme in the peroxisomal fatty acid beta-oxidation pathway, suggesting this gene may be a potential candidate gene for the traits related to fat metabolism. IMF content is a major determinant of meat quality. After the elimination of the halothane mutation, the next limiting factor for meat quality would be IMF (Webb, Reference Webb1998). The IMF can be measured by subjective and objective methods. Meat marbling score is one of subjective methods. The more the IMF content, the higher the meat marbling score. This study showed a significant effect of ACOX1 on meat marbling score. Although this SNP did not significantly contribute to the variation in IMF content in this population, the effect of this locus still approached the P < 0.05 statistical level (data not shown). As we have not performed the QTL analysis in this chromosome, we cannot determine whether the significant associations are due to the marker or due to other co-inherited blocks. However, the dominance effects were consistent within two F2 populations except for the effects on MMS2 (data not shown).
The potential gain of MAS would be in terms of reduced costs for sib testing after slaughter and reduction in sophisticated meat quality measurements as well as additional improvement of meat quality by early information from genetic markers (Ovilo et al., Reference Ovilo, Fernandez, Rodriguez, Nieto and Silio2006). At present, most of the market pigs are the hybrids among different specialised pig sire and dam lines. As for the dominant markers in the MAS programme, one of the alleles can be selected in sire line, and the other allele selected in the dam line, so the dominant effects can be realised in the hybrids. From this point of view, we can select the allele from Meishan pigs in the synthetic dam lines, and then the dam line can be crossbred with Large White pig sire lines. However, before the selection of these markers in the specialised dam lines, we should confirm the effects of those markers by comparing the meat quality traits of crossbred pigs carrying different genotypes, as the alternative allele was not completely fixed in the synthetic dam lines prior to the selection. In addition, we have to re-test for the effects in different crossbred populations, because it is likely that this polymorphism indirectly affects meat quality traits by being in linkage disequilibrium with another polymorphism that directly influences the quantitative traits analysed.
Conclusions
The results showed that the FokI variants of the GYS1 gene was significantly associated with pHSC in the association studies, but when taken into account the information provided by the neutral genetic markers on this chromosome, the peak of the QTL curve was located around this marker for pH, but it did not reach significance level. The effects of the PstI variants in the ninth intron of the ACOX1 gene polymorphism on marbling and IMF were consistent. These two loci seemed to be significantly dominant in action. Those dominant markers could be used in the MAS of crossbreeding of specialised pig sire and dam lines.
Acknowledgements
This study was supported financially by National Natural Science Foundation of P. R. China (30500358), the National ‘973’ programme of P. R. China (2006CB102102), the National High Technology Development Project and Natural Science Foundation of Hubei Province (2005ABA142). The authors gratefully acknowledge Dr Zhang JH for providing part of the microsatellite information. The authors also acknowledge the teachers and graduate students of the Key Laboratory of Swine Genetics and Breeding, Ministry of Agriculture, and the Swine Breeding Center of China for the collection of meat quality information.
