Evidence for New Alleles in the Protein Kinase Adenosine Monophosphate-Activated ₃-Subunit Gene Associated With Low Glycogen Content in Pig Skeletal Muscle and Improved Meat Quality

Corresponding author: Max Rothschild, Department of Animal Science, Iowa State University, Ames, IA 50011., mfrothsc{at}iastate.edu (E-mail)

Communicating editor: C. HALEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Several quantitative trait loci (QTL) affecting muscle glycogen content and related traits were mapped to pig chromosome 15 using a three-generation intercross between Berkshire x Yorkshire pigs. On the basis of the QTL location the PRKAG3 (protein kinase, AMP-activated, ₃-subunit) gene was considered to be a good candidate for the observed effects. Differences in the PRKAG3 gene sequences of the founder animals of the intercross were analyzed. The RN^- mutation previously reported was not present in the cross but three missense substitutions and a polymorphic short interspersed element (SINE) were identified. To confirm the hypothesis that at least one of these mutations was associated with differences in meat quality, >1800 animals from several unrelated commercial lines were genotyped for the candidate substitutions and an association study was performed. The results demonstrate the presence of new economically important alleles of the PRKAG3 gene affecting the glycogen content in the muscle and the resulting meat quality. Haplotype analysis was shown to resolve the effects of PRKAG3 more clearly than analysis of individual polymorphisms. Because of their prevalence in the more common commercial breeds, the potential implications for the pig industry and consumers are considerably greater than the original discovery of the RN^- mutation. Furthermore, these results illustrate that additional alleles of genes involved in major mutations may play a significant role in quantitative trait variation.

THE recent discovery (MILAN et al. 2000 ) of a nonconserved substitution in the PRKAG3 gene has explained the dominant mutation (denoted RN^-) that accounted for large differences in meat quality and processing yield in the Hampshire pig breed (MONIN and SELLIER 1985 ; LEROY et al. 1990 ). This substitution (R200Q) in the PRKAG3 gene caused a 70% increase in glycogen in muscle in RN^- homozygous and heterozygous animals that then resulted in the observed lower muscle pH 24 hr after slaughter, in reduced water-holding capacity in the muscle, and in much lower yield of a cured cooked ham product. The 200Q allele is associated with all RN^- animals and was present in a very high percentage of Hampshire pigs but not in pigs with an rn⁺ phenotype or in other breeds (MILAN et al. 2000 and this study).

Mammalian adenosine monophosphate (AMP)-activated protein kinase (AMPK) plays a key role in regulating energy homeostasis in eukaryotes (HARDIE et al. 1998 ). It consists of a catalytic subunit () and two regulatory subunits (ß and ). Two isoforms have been identified for both the - and ß-subunit and there are three isoforms reported for the -subunit in several mammals (GAO et al. 1996 ; STAPLETON et al. 1996 , STAPLETON et al. 1997 ; MILAN et al. 2000 ). The ₃-peptide, encoded by the PRKAG3 gene, is one of three options for the regulatory subunit of AMPK. When eukaryotic cells are subjected to environmental or nutritional stress factors and the AMP/ATP ratio rises significantly, then the "AMPK cascade" is induced, initiating measures to conserve energy (THORNTON et al. 1998 ) and to induce the ATP synthetic pathways (HARDIE et al. 1998 ).

The identification of quantitative trait loci (QTL) for meat quality traits in the region of the PRKAG3 gene in an rn⁺ resource population (MALEK et al. 2001 ) suggested that new allelic variation in this gene may be responsible for the observed effects. In this article we report the presence of new economically important alleles of the PRKAG3 gene affecting the glycogen content in muscle and in general the meat quality traits of pigs that include ultimate pH and color measures and that are correlated with water-holding capacity, drip loss, tenderness, and cooking loss (SELLIER 1998 ). Initial estimates of the allelic and haplotype effects and frequencies suggest that these alleles may have significant economic potential for the pig industry and ultimately for consumers in terms of improved pork quality.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Pedigree, linkage, and QTL mapping:
We have generated an intercross between Berkshire and Yorkshire (B x Y) pig breeds, yielding 525 F₂ offspring, and used this pedigree to map QTL for meat quality (MALEK et al. 2001 ) using an interval mapping method (HALEY et al. 1994 ). In this cross, the Berkshire breed was chosen because it is regarded as having very good meat quality, particularly in terms of pH, color, water-holding capacity, and tenderness. The PRKAG3 gene was mapped to the B x Y family linkage map using the CRI-MAP (version 2.4) mapping program (GREEN et al. 1990 ). The interval mapping method (HALEY et al. 1994 ), including the PRKAG3 site information, was used to map the QTL for meat quality for pig chromosome 15 (SSC 15; Fig 1). The QTL effects were estimated and represent the average Berkshire allelic effect compared to the average Yorkshire allelic effect.

View larger version (40K): In this window In a new window Download PPT slide   Figure 1. F-ratio curves for evidence of QTL associated with meat quality for SSC 15. The x-axis indicates the relative position on the linkage map. The y-axis represents the F-ratio. Arrows on the x-axis indicate the position where a marker was present. (- - -) 5% chromosome-wise significance; (—) 5% genome-wise significance; (· · ·) 1% genome-wise significance. (a) Average glycogen, average lactate, and average glycolytic potential traits. (b) pH traits.

Tissue samples and DNA/RNA isolation:
Blood samples and phenotypes were collected and recorded on the F₀, F₁, and F₂ animals from the intercross family (MALEK et al. 2001 ) together with blood samples and muscle tissue from the ham and loin area of several F₃ animals. We also obtained a large collection of blood samples from five different commercial lines of pigs (Landrace, Large White, Duroc, Duroc Synthetic, and Berkshire). Genomic DNA was isolated from whole blood by standard salting-out procedures and total RNA was extracted from ham and loin muscle tissue using the TRIzol reagent method according to the manufacturer's instructions (GIBCO/BRL, Rockville, MD).

PCR, reverse transcription-PCR, rapid amplification of cDNA ends, and polymorphism discovery:
On the basis of the PRKAG3 pig gene sequence available in GenBank (no. AF214521), we designed primers to amplify the entire coding regions of the PRKAG3 gene. The PCR reactions were performed using 12.5 ng of porcine genomic DNA, 1.5 mM MgCl₂, 0.125 mM dNTP, 0.3 µM of each primer and 0.35 units Taq DNA polymerase (Promega, Madison, WI), and PCR buffer (10 mM Tris-HCl, 50 mM KCl, and 0.1% TritonX-100) in a 10-µl final volume. The reverse transcription of total RNA (3.5 µg) was performed by random hexanucleotide priming and Superscript II (GIBCO/BRL) according to the manufacturer's protocol (primers: Set A, forward 5' ATGAGCTTCCTAGAGCAAGGAG 3' and reverse 5' CAGGTCTCAATCTTATGTTCTTC 3'; set B, forward 5' CGTCCGAGCGGCACCTTTGT 3' and reverse 5' AAGGTTCCAAGGTTCTCAGGC 3'). 5' rapid amplification of cDNA ends (RACE) experiments were performed using the FirstChoice RLM-RACE kit (Ambion, Austin, TX) according to the manufacturer's instructions, followed by sequencing of the PCR products (gene-specific primers: outer 5' CCCACGAAGCTCTGCTTCTT 3' and inner 5' TCCTTGCTCTAGGAAGCTCAT 3'). The amplicons were sequenced using dye terminators (PE Applied Biosystems, Foster City, CA) on an ABI 377 automated sequencer. We used Sequencer software (Gene Codes, version 4.0.5, Ann Arbor, MI) to assemble the sequences and to identify polymorphisms.

Genotyping and PCR-restriction fragment length polymorphism analysis:
The region flanking each analyzed missense mutation was amplified using the same pair of primers for the T30N and G52S substitutions (forward 5' ATGAGCTTCCTAGAGCAAGGAG 3' and reverse 5' GGCTGCATGATGTTATGTGCCT 3') and a different pair for I199V (forward 5' GGAGCAAATGTGCAGACAAG 3' and reverse 5' CCCACGAAGCTCTGCTTCTT 3'). After digestion with BsaHI (for I199V), HphI (for G52S), and StyI (for T30N) restriction enzymes, the digested PCR products were separated on 4% NuSieve agarose (FMC, Rockland, ME) gels and stained with ethidium bromide. For the short interspersed element (SINE) polymorphism, PCR amplification (primers: forward 5' GAAACTCTTCTCCCCACAGAC 3' and reverse 5' GGCTGCATGATGTTATGTGCCT 3') was followed by separation of the products on a 1% agarose (AMRESCO, Solon, OH) gel. After genotyping for these polymorphisms, all the animals with haplotype 2 (Table 6) were also genotyped for the R200Q substitution to increase the chance of finding the RN^- or 200Q allele (see MILAN et al. 2000 ). Two homozygotes for the 200Q allele and four carriers were found and these were removed from further statistical analyses so that the RN^- mutation did not affect our analysis of the other substitutions. For the R200Q substitution we used the same primers as for the I199V mutation and the digestion was performed with the BsrBI restriction enzyme. As a final check, a random sample of 100 animals with different haplotypes was also scored for the R200Q substitution, but none of the animals carried the 200Q allele.

Phenotypic trait measurement:
Phenotypic measures for the B x Y family were made using typical industry techniques (MALEK et al. 2001 ) and included pH, color, and glycolytic potential. For the pigs from five commercial lines, data were collected at a commercial packing plant and individual meat color (loin and ham reflectance—lower values preferred) and individual loin and ham pH 24 hr after harvest (higher values preferred) were obtained. For the packing plant data no measures of glycogen or glycolytic potential were obtained. The measures of color and pH phenotypic traits are common industry measures of meat quality that are indirectly correlated with glycogen and glycolytic potential.

Statistical analysis:
Berkshire x Yorkshire F₂ population analysis: Associations between the PRKAG3 I199V substitution and glycogen, lactate, glycolytic potential, and meat quality traits in the B x Y F₂ population were tested using general linear model procedures (SAS procedure GLM; SAS Institute, Cary, NC) with a model that included dam, slaughter date, sex, and I199V genotype. Least-squares (LS) means for all three genotypes were obtained for the I199V substitution.

Commercial lines analyses: The associations between the PRKAG3 polymorphisms and meat quality traits were tested using mixed-model procedures (SAS procedure MIXED; SAS Institute) with a model that always included sire as a random effect and slaughter date and marker genotype(s) as fixed effects. Line was added as a fixed effect for across-line analyses. Sex and farm were not included because all traits were measured on females only and no more than one farm was represented on each slaughter date. While males were not used in this portion of the analysis, our results in the B x Y suggest no sex-by-genotype effect. A full model including a separate genotype effect for each of the three substitution sites was fitted across the five commercial lines. Nonsignificant genotype effects were removed by backward elimination (P to remove >0.10) to identify which substitutions were associated with effects on the meat quality traits.

LS means for the three genotype classes were obtained within the commercial lines for each of the substitutions analyzed individually. No line-by-genotype interactions were found and therefore, to improve the reliability of the estimates of the allele effects, the data from five lines were pooled for an across-line analysis.

The combined effects of the three substitutions were estimated as haplotype substitution effects. Contrasts between haplotypes were estimated from a model including sire (random), slaughter day, and one variable for each haplotype with values -1, 0, and 1 corresponding to the animal having 0, 1, or 2 copies of the haplotype in question. The haplotype substitution effects were presented as deviations from the mean of the haplotypes and reflect the differences from the worst to the best haplotype. The number of animals used in association analyses varied on the basis of the trait measured and are listed in Table 3 Table 4 Table 5.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Marker development and linkage mapping:
Several significant QTL were detected on SSC15 (MALEK et al. 2001 ) in the region where the PRKAG3 gene was located (MILAN et al. 2000 ) between the markers SW1683 and SW1983 (Fig 1). These included QTL for average glycogen content and glycolytic potential that have been reported (MILAN et al. 2000 ) to be affected by the PRKAG3 200Q allele, as well as the traits 24-hr ham and loin pH and 24-hr loin Hunter L values (light reflectance). The favorable allele at this QTL, which, interestingly, has an additive effect (the RN^- mutation is dominant), was derived predominantly from the Berkshire breed (generally regarded as having very good meat quality), as expected (Table 1). The PRKAG3 gene was the unique candidate gene in this area, based on the recent development of the bacterial artificial chromosome contig in the porcine RN region (MILAN et al. 2000 ), the high degree of linkage order conservation of the porcine map in this area with the human transcript map (JEON et al. 2001 ), and the recently developed human genome map (LANDER et al. 2001 ). We first tested the founder animals, two Berkshire sires and nine Yorkshire dams, for the published RN^- substitution (R200Q). All the founder animals had the rn⁺ allele (200R). By sequencing the entire coding region of the PRKAG3 gene in B x Y family founders and in four F₃ individuals with extreme values for meat quality, we identified three missense mutations: the T30N and the I199V substitutions previously described (MILAN et al. 2000 ) and a new missense mutation (G52S). Another nonsynonymous substitution (P53L) found by MILAN et al. 2000 was not found to be segregating in the founders of the B x Y family where they were all 53P. Due to the lack of information on the 5' untranslated region, we used RACE to find the complete 5' flanking sequence and gene organization in that region. An intronic SINE polymorphism was discovered starting 79 bp upstream of the start codon but this was present in only three Yorkshire grandams. On the basis of the differences in allele frequency of each site between the founders of the intercross family, we considered the G52S and I199V substitutions as the most likely candidates for the meat quality QTL reported previously. Using the I199V substitution we mapped the PRKAG3 gene in the B x Y linkage map to a position below the broad peak(s) of the QTL for glycogen, lactate, and glycolytic potential and 24-hr pH (Fig 1). After adding the PRKAG3 I199V information, the map length and marker order on SSC 15 was the same as in MALEK et al. 2001 . Reanalysis of the QTL including PRKAG3 I199V (Fig 1) caused small changes in the F value and the location of the QTL peaks on SSC 15 (from 0 to 3 cM) when compared with the results of MALEK et al. 2001 .

F₂ association study:
Using an association analysis, we found significant effects of all three of the substitutions (T30N, G52S, and I199V) on average glycogen and lactate content and also on glycolytic potential in the F₂ B x Y population (data shown only for I199V substitution; Table 2). The most significant effects were revealed for I199V substitution for most of the traits analyzed, including glycogen and lactate content and glycolytic potential measures, but also for some of the meat quality traits associated with these measures. From the F₂ data, the 30T, 52G, and 199I alleles were favorable in terms of meat quality. Given the large expected linkage disequilibrium in the intercross, it was necessary to investigate and confirm the effects of these mutations in several outcross commercial lines of pigs to determine whether this gene is likely to be involved directly in the observed variation in meat quality.

Analysis of commercial populations:
The genotypic frequencies for the analyzed substitutions are presented in Table 3. For all three substitution sites, the Berkshire line had a higher frequency for the genotypes associated with low glycogen content (and higher meat quality) in skeletal muscle on the basis of the B x Y F₂ data. The other commercial populations have lower frequencies of the favorable alleles with this being particularly marked for the I199V substitution when compared with the Berkshire population.

The PRKAG3 mutations and their associations with meat quality were tested for each of the five commercial lines and also across all of the lines. Backward elimination of substitution sites, in the across-line analysis, kept I199V in the model for all six traits; G52S for ham pH, loin pH, loin Minolta L, and loin Minolta b; and T30N was kept for ham Minolta L, loin Minolta L, and ham Minolta b.

Because each of the substitutions showed distinct associations with at least three of the traits, the effects of each substitution were estimated independently. Least-squares estimates of the genotype means across lines (Table 4) and within lines (Table 5) showed significant effects between the analyzed substitutions and measures of meat quality, suggesting that several additional (new) rn⁺ alleles may exist.

The association study revealed that the largest effects across the lines (Table 4) and also within the lines (Table 5, data shown only for I199V) were obtained with the I199V substitution for all the traits analyzed. For this substitution the associations were highly significant (P < 0.0005) for all of the meat quality traits used in this study when analyzed across lines. Significant associations with at least one of the traits were revealed for the same substitution within each of the individual lines, with highly significant effects for ham Minolta b in Duroc and Duroc Synthetic and for loin pH in Duroc Synthetic. These two breeds (Duroc Synthetic and Duroc) have the best frequency distribution for association analysis with a sufficient number of animals for each genotype (Table 5). In the across-line analysis and most of the individual line results, the effects were in the same direction for all traits with allele 199I being the favorable allele for high meat quality.

Significant effects, but smaller when compared to the I199V, were revealed for the T30N substitution in five of the traits when analyzed across lines (Table 4). Within-line analyses of T30N revealed effects almost exclusively in Duroc and Duroc Synthetic populations (data not shown). In most of the situations, the effects were in the same direction, the allele 30T being associated with a better meat quality.

For the G52S substitution, significant (P < 0.05) effects were identified for only two of the traits (ham pH and loin Minolta L) in across-line analysis, and a different allele was identified as favorable for those traits. Within-line analysis revealed significant associations for loin Minolta color scores for just the Duroc Synthetic population (data not shown).

In the five commercial populations we tested, we found just four haplotypes (Table 6). The Berkshire population is the least polymorphic, having haplotype 3 (30T-52G-199I) at a high frequency (0.87). In Large White, haplotype 2 (30T-52S-199V) is the most frequent and haplotype 1 (30N-52G-199V) has the highest frequency in Landrace, Duroc, and Duroc Synthetic populations. Haplotype 4 (30T-52G-199V) has the lowest frequency in all the populations.

The haplotype substitution effects for each line and across lines were calculated as the deviation from the average of the four haplotypes (Fig 2). Across- and within-line analyses showed bigger differences between haplotypes for ham pH and color measurements than for traits of the loin. For ham pH, across- and within-line analyses showed haplotype 3 having the highest effect, which was significantly different from each of the other haplotypes in the across-line analysis (P < 0.0005) and from at least one other haplotype in each individual line analysis (P < 0.05). Haplotype 2 was the next best for most of the traits and lines with haplotypes 1 and 4 tending to be the worst with respect to meat quality. This hierarchy is not evident in the Berkshire population, where significant differences are seen only with haplotype 4, which has the lowest value corresponding to the across-line result. The nonsignificant results in Berkshire are likely to be due in part to the low level of polymorphism in this breed and the concomitant very low number of observations for haplotypes 1 and 4. The estimate for haplotype 4 in the Duroc Synthetic population appears to be different from that in the other lines [especially for ham pH, where it is significantly higher than haplotype 2 (P < 0.05) and haplotype 1 (P < 0.01)], but the frequency of haplotype 4 in this population was very low (0.07). The synthetic nature of this line (although its inception was six generations ago) also provides the opportunity for extended linkage disequilibrium to be present, increasing the chance for linked loci to contribute to the haplotype substitution effects.

View larger version (21K): In this window In a new window Download PPT slide   Figure 2. Haplotype substitution effects of PRKAG3 on pH and color scores in the ham and loin. Haplotype substitution effects are estimated across five lines (ALL) and within each line (, haplotype 1; , haplotype 2; , haplotype 3; , haplotype 4). Lines are based on Landrace (LR), Large White (LW), or Duroc (DU); a Duroc-based synthetic line (DS); and a Berkshire-based line (BE). A separate scale is used for the BE line. Estimates within a column that have the same superscript are significantly different at P < 0.0005 for the across-line estimates and at P < 0.005 for the within-line estimates.

The haplotype results for Minolta scores were in line with the pH results. Haplotype 3 was generally found to have the favorable effect (lower color scores). There are a few exceptions in the results from individual lines and these may be the result of sampling. The only significant deviation is with haplotype 2, which is associated with a lower Minolta b score in Berkshire (P < 0.05). In the across-line analysis haplotype 2 was second to haplotype 3 in most cases.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The results reported in this work provide important evidence in favor of the presence of new alleles of the PRKAG3 gene affecting meat quality traits. This conclusion is based on three points: (1) the known effect of PRKAG3 alleles rn⁺ and RN^- on meat quality; (2) observation of several QTL for related meat quality traits discovered on SSC15 in the region where PRKAG3 is located in the B x Y family (these QTL were discovered in the pig cross where the original R200Q substitution was not segregating); and (3) results presented here on the association between the PRKAG3 substitutions and glycogen and lactate content, glycolytic potential, and meat quality traits in the B x Y F₂ population and association with meat quality traits in several unrelated commercial pig lines.

Association analyses of the individual substitutions revealed that, of the three studied here, the I199V substitution showed the most significant and largest differences in meat quality traits. For example, B x Y F₂ analysis showed significant differences between the I199V genotype classes for glycolytic potential, but also in glycogen and lactate content (Table 2). Important effects were also revealed for most of the meat quality traits analyzed. Allele 199I was found to be associated with a lower level of glycogen, lactate, and glycolytic potential, with higher ham and loin pH and with better color scores. This marker was sufficiently informative in B x Y F₂ to provide good estimates of the allele effects.

In the analyses of the commercial populations, the I199V substitution is associated with significant differences in LS means between the homozygous classes up to 0.14 in the Landrace line and 0.10 across the lines for ham pH (Table 4 and Table 5). For one of the meat color measures, Ham Minolta L, significant LS means differences were found between homozygous genotypes up to 3.5 units of reflectance (in Landrace) and 2.0 across the lines. These effects are in the range of 0.5 to 1 phenotypic standard deviation. Important differences were also revealed for the other traits and breeds. Effects of this magnitude in traits important for overall meat quality are of great interest to the animal breeding industry.

Besides I199V, large effects were also estimated from single substitution analysis of T30N. However, only modest effects of the T30N substitution remained if I199V was also included in the analysis. Strong linkage disequilibrium between sites 30 and 199 is considered to be in large part responsible for the effects being detected for site 30. Small effects, which were mostly nonsignificant, were observed for the single-site analysis of G52S.

Haplotype analysis helped to dissect the effects of the nonsynonymous substitutions and provided additional evidence for an effect at position 199 as well as position 52. Haplotype 3, which is the only haplotype containing 199I, was the most favorable haplotype with respect to pH and meat color measurements. In most of the situations tested, haplotype 2, which is the only haplotype containing 52S, showed an intermediate value, especially for ham quality traits where the differences in effects were more significant and greater than in other traits. Values for haplotypes 1 and 4 are close together at the bottom of the range and in most cases not significantly different from each other.

The observation that the values of these two haplotypes (1 and 4) are relatively similar for most estimates leads us to the conclusion that the T30N substitution is making only a marginal contribution to meat quality variation. In across-line, Landrace, and Large White analyses, where the frequency of haplotype 4 is >0.10, we find a favorable effect of haplotype 1 on ham Minolta scores (this haplotype being associated with the 30N variant) when compared with haplotype 4. In the other populations, differences between these haplotype effects are poorly estimated due to very low frequency of haplotype 4.

The difference between haplotype 4 and haplotype 2 is only at the G52S site. The effects of haplotype 4 and 2 are significantly different for pH and Minolta L scores in both ham and loin in the across-line analysis and for several traits of the individual lines, most notably the Large White. Haplotype 2 (which contains 52S and encodes a serine) is favorable over haplotype 4 (which contains 52G and encodes a glycine). This is the opposite of what was found in the B x Y study where 52G was predicted to be the favorable allele. Strong linkage disequilibrium with the I199V site, due to the limited number of founders of the F₂, may have masked the true effect of the G52S substitution in this population. Interestingly, the individual analysis of G52S did not show any effect for most traits and lines. That analysis compares haplotype 2 with the other three combined. From Fig 2 it can be seen that a mixture of the other three haplotypes, depending on haplotype frequencies, can result in a mean value close to that of haplotype 2 so that a difference would not be detected when G52S is analyzed individually, which points out the value of haplotype-based analysis.

The 30T variant, present in haplotype 4, was found to be favorable for meat quality on the basis of single-site analysis, because it was associated with significant effects in the Duroc and Duroc Synthetic lines for most of the traits. In these two populations haplotype 3 has a moderate frequency (Table 6) and contains both the 30T and the favorable 199I variant. Thus, the 199I variant contributes to the higher effects for the 30T site variant due to linkage disequilibrium.

We conclude that the joint analyses of substitutions and the haplotype analyses demonstrate the presence of three nonsynonymous substitutions in the PRKAG3 gene with different size effects on meat quality measurements in pigs. This interesting model of "one gene-several polymorphisms-diverse phenotypes" is based on distinguishable additive effects of a complex phenotypic trait and can serve as a model for future studies with other traits. The presence of multiple alleles as a consequence of consecutive mutations in a gene under selection has also been proposed recently in pigs (JEON et al. 1999 ; NEZER et al. 1999 )

The I199V substitution is in a cystathionine ß-synthase (CBS) domain, a very conserved region in genes of this family (MILAN et al. 2000 ). The role of the CBS domain is still unclear but it has been suggested that it is involved in cytoplasmic targeting (PONTING 1997 ), protein-protein interaction (BATEMAN 1997 ), and/or regulation (BATEMAN 1997 ) of protein activity. There are four CBS domains in the PRKAG3 gene (MILAN et al. 2000 ) and the I199V substitution is located in the first and most conserved domain. Alignment between the CBS domain and the ₃-peptide obtained using Pfam software revealed that the preferred amino acid at this position is isoleucine (result not shown). Interestingly, in this study, allele 199I (coding isoleucine at the site 199) was found to be associated with better meat quality in commercial populations and the B x Y F₂ family and also with lower levels of glycogen, lactate, and glycolytic potential in the latter one.

MILAN et al. 2000 show that the 200Q variant (RN^-) is always found with 199V. However, 199V is found with both 200R and 200Q and 199I is always found with 200R. As only three nucleotides separate these substitution sites, the probability of recombination between them is extremely small. For this reason we can consider R200Q to be the most recent substitution, a hypothesis also supported by the presence of this mutation in only the Hampshire pig breed. Both of the haplotypes 199V-200R and 199I-200R could be ancestral, because each has been identified in most of the breeds analyzed to date (MILAN et al. 2000 ), including wild boar and several species of the suborder Suisformes (D. CIOBANU, G. PLASTOW and M. ROTHSCHILD, unpublished results).

The 199V-200R haplotype is associated with higher glycogen content and lower postmortem ham/loin pH when compared with the 199I-200R haplotype (B x Y F₂ data). The substitution at codon 199 presumably leads to an effect on glucose metabolism and therefore an increase in the muscle glycogen content. The third haplotype 199V-200Q confers the RN^- phenotype. The associated effect 199V-200Q on glycogen content is greater than the effect of other haplotypes and the 199V-200Q haplotype is dominant over the others. For these reasons we suggest that the RN^- phenotype could be a combined effect of the 199V-200Q haplotype rather than solely a result of the R200Q substitution. This effect could be caused by the modification of the CBS domain by these substitutions.

The exact functions of the ß and regulatory subunits of the AMPK are still unclear. However, it is known that both are essential for kinase activity (HARDIE and CARLING 1997 ). In vitro experiments show that the ß-subunit may have an important role in the formation of the heterotrimeric structure of AMPK, as ß interacts with both the - and -subunits, which do not interact directly with each other (WOODS et al. 1996 ). Recent evidence suggests that the allosteric AMP-binding site may involve both the - and -subunits of the AMPK complex (CHEUNG et al. 2000 ). CHEUNG et al. 2000 proposed an elegant model in which, in the absence of AMP, the heterotrimeric complex may be predominantly inactive without interaction between the - and -subunits. In this situation, phosphorylation of the Thr¹⁷² site in the -subunit and interaction with substrates are blocked by the autoinhibitory region of the -subunit. In the active form of AMPK the interaction between the autoinhibitory region and one or more of the CBS domains prevents the autoinhibition, and AMP binds on both subunits to stabilize the assembly (CHEUNG et al. 2000 ). The alignment information, the proposed model of the regulation of the AMPK complex, and the presence of the R200Q site nearby support the hypothesis of a possible role of the I199V substitution in the activity of AMPK. Even though the molecular structure of the AMPK complex has not been resolved yet, we hypothesize that the amino acid change may also influence the structure and activity of the enzyme, resulting in the observed effect of the G52S substitution.

Although the ₃-subunit is highly expressed in skeletal muscle, AMPK activity appears to be associated more with ₁- and ₂-isoforms (CHEUNG et al. 2000 ). However, in a mechanism not yet understood, the R200Q substitution (or I199V-R200Q combination) in the PRKAG3 gene causes important differences in AMPK activity in Hampshire pigs (MILAN et al. 2000 ), which suggests that the ₃-isoform has an important role in glucose metabolism in skeletal muscle. Detailed functional studies of the different subunit combinations will be necessary to resolve the situation. The role of AMPK in glucose metabolism makes physiological sense on the basis of comparisons with the related SNF1 complex from yeast. Also, several studies show that AMPK participates in glycogen metabolism by inactivating glycogen synthase (POULTER et al. 1988 ; CARLING and HARDIE 1989 ; ZHANG et al. 1993 ), by activating the nitric oxide synthase (FRYER et al. 2000 ), and by increasing the translocation of the glucose transporter 4 to the plasma membranes (HAYASHI et al. 1998 ; BERGERON et al. 1999 ; HOLMES et al. 1999 ; KURTH-KRACZEK et al. 1999 ).

While the effects of the substitutions reported here on the measures of meat quality are of lesser magnitude than those of the dominant RN^- mutation, they are of importance both biologically and economically. In particular, these alleles are segregating in all of the commercial lines and breeds analyzed to date in contrast to the RN^- mutation, which is associated only with the Hampshire breed and has limited use in most pork production programs. The results reported here for PRKAG3 also suggest that geneticists should look for additional mutations with an economic impact in genes known to cause more drastic effects both within and between species. This notion is supported by reports of major effects associated with other genes outside the target species or breed, e.g., large effects of MC4R mutations in mice (HUSZAR et al. 1997 ), humans (YEO et al. 1998 ) and, to a lesser extent, in pigs (KIM et al. 2000 ).

The identification of novel genes with biochemical significance in animal species will also provide useful information for human biomedical targets. This knowledge is enhanced when new and interesting alleles are discovered. In the case of PRKAG3, it has been suggested (MILAN et al. 2000 ) that this gene and other AMPK-related genes in humans are interesting candidates for human type II diabetes on the basis of their function and QTL locations. For this reason, the effect of these new alleles may provide new insights about potential factors affecting glucose metabolism and should be considered in further investigations of this disease.

	ACKNOWLEDGMENTS

This work was financially supported by PIC International Group, the Iowa Agriculture and Home Economics Experimental Station, Ames (paper no. J-19159, project no. 3600), and the Hatch and State of Iowa funds.

Manuscript received June 12, 2001; Accepted for publication August 6, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BATEMAN, A., 1997  The structure of a domain common to archaebacteria and the homocystinuria disease protein. Trends Biochem. Sci. 22:12-13[Medline].

BERGERON, R., R. R. RUSSELL, III, L. H. YOUNG, J. M. A. REN, and M. MARCUCCI et al., 1999  Effect of AMPK activation on muscle glucose metabolism in conscious rats. Am. J. Physiol. 276:E938-E944.

CARLING, D. and D. G. HARDIE, 1989  The substrate and sequence specificity of the AMP-activated protein kinase. Phosphorylation of glycogen synthase and phosphorylase kinase. Biochim. Biophys. Acta 1012:81-86[Medline].

CHEUNG, P. C., I. P. SALT, S. P. DAVIES, D. G. HARDIE, and D. CARLING, 2000  Characterization of AMP-activated protein kinase gamma-subunit isoforms and their role in AMP binding. Biochem J. 346:659-669.

FRYER, L. G., E. HAJDUCH, F. RENCUREL, I. P. SALT, and H. S. HUNDAL et al., 2000  Activation of glucose transport by AMP-activated protein kinase via stimulation of nitric oxide synthase. Diabetes 49:1978-1985[Abstract].

GAO, G., S. FERNANDEZ, D. STAPLETON, A. S. AUSTER, and J. WIDMER et al., 1996  Non-catalytic beta- and gamma-subunit isoforms of the 5'-AMP-activated protein kinase. J. Biol. Chem. 271:8675-8681[Abstract/Free Full Text].

GREEN, P., K. FALLS and S. CROOKS, 1990 Documentation for CRIMAP, version 2.4., Washington University, School of Medicine, St. Louis.

HALEY, C. S., S. A. KNOTT, and J. M. ELSEN, 1994  Mapping quantitative trait loci in crosses between outbred lines using least squares. Genetics 136:1195-1207[Abstract].

HARDIE, D. G. and D. CARLING, 1997  The AMP-activated protein kinase—Fuel gauge of the mammalian cell? Eur. J. Biochem. 246:259-273[Medline].

HARDIE, D. G., D. CARLING, and M. CARLSON, 1998  The AMP-activated/SNF1 protein kinase subfamily: Metabolic sensors of the eukaryotic cell? Annu. Rev. Biochem. 67:821-855[Medline].

HAYASHI, T., M. F. HIRSHMAN, E. J. KURTH, W. W. WINDER, and L. J. GOODYEAR, 1998  Evidence for 5' AMP-activated protein kinase mediation of the effect of muscle contraction on glucose transport. Diabetes 47:1369-1373[Abstract].

HOLMES, B. F., E. J. KURTH-KRACZEK, and W. W. WINDER, 1999  Chronic activation of 5'-AMP-activated protein kinase increases GLUT-4, hexokinase, and glycogen in muscle. J. Appl. Physiol. 87:1990-1995[Abstract/Free Full Text].

HUSZAR, D., C. A. LYNCH, V. FAIRCHILD-HUNTRESS, J. H. DUNMORE, and Q. FANG et al., 1997  Targeted disruption of the melanocortin-4 receptor results in obesity in mice. Cell 88:131-141[Medline].

JEON, J. T., O. CARLBORG, A. TORNSTEN, E. GIUFFRA, and V. AMARGER et al., 1999  A paternally expresssed QTL affecting skeletal and cardiac muscle mass in pigs maps to the IGF2 locus. Nat. Genet. 21:157-158[Medline].

JEON, J. T., V. AMARGER, C. ROGEL-GAILLARD, A. ROBIC, and E. BONGCAM-RUDLOFF et al., 2001  Comparative analysis of a BAC contig of the porcine RN region and the human transcript map: implications for the cloning of trait loci. Genomics 72:297-303[Medline].

KIM, K. S., N. LARSEN, T. SHORT, G. PLASTOW, and M. F. ROTHSCHILD, 2000  A missense variant of the porcine melanocortin-4 receptor (MC4R) gene is associated with fatness, growth, and feed intake traits. Mamm. Genome 11:131-135[Medline].

KURTH-KRACZEK, E. J., M. F. HIRSHMAN, L. J. GOODYEAR, and W. W. WINDER, 1999  5' AMP-activated protein kinase activation causes GLUT4 translocation in skeletal muscle. Diabetes 48:1667-1671[Abstract].

LANDER, E. S., L. M. LINTON, B. BIRREN, C. NUSBAUM, and M. C. ZODY et al., 2001  Initial sequencing and analysis of the human genome. Nature 409:860-921[Medline].

LEROY, P., J. NAVEAU, J. M. ELSEN, and P. SELLIER, 1990  Evidence for a new major gene influencing meat quality in pigs. Genet. Res. 55:33-40[Medline].

MALEK, M., J. C. M. DEKKERS, H. K. LEE, T. J. BAAS, and K. PRUSA et al., 2001  A molecular genome scan analysis to identify chromosomal regions influencing economic traits in the pig. II. Meat and muscle composition. Mamm. Genome 12:637-645[Medline].

MILAN, D., J. T. JEON, C. LOOFT, V. AMARGER, and A. ROBIC et al., 2000  A mutation in PRKAG3 associated with excess glycogen content in pig skeletal muscle. Science 288:1248-1251[Abstract/Free Full Text].

MONIN, G. and P. SELLIER, 1985  Pork of low technological quality with a normal rate of muscle pH falls in the immediate postmortem period: the case of Hampshire breed. Meat Sci. 3:49-63.

NEZER, C., I. MOREAU, L. KARIM, B. BROUWERS, and W. COPPIETERS et al., 1999  An imprinted QTL with major effects on muscle mass and fat deposition maps to the IGF2 locus in pigs. Nat. Genet. 21:155-156[Medline].

PONTING, C. P., 1997  CBS domains in CIC chloride channels implicated in myotonia and nephrolithiasis (kidney stones). J. Mol. Med. 75:160-163[Medline].

POULTER, L., S. G. ANG, B. W. GIBSON, D. H. WILLIAMS, and C. F. HOLMES et al., 1988  Analysis of the in vivo phosphorylation state of rabbit skeletal muscle glycogen synthase by fast-atom-bombardment mass spectrometry. J. Biol. Chem. 175:497-510.

SELLIER, P., 1998 Genetics of meat and carcass traits, pp. 463–510 in The Genetics of the Pig, edited by M. F. ROTHSCHILD and A. RUVINSKY. CABI, Wallingford, United Kingdom.

STAPLETON, D., K. I. MITCHELHILL, G. GAO, J. WIDMER, and B. J. MICHELL et al., 1996  Mammalian AMP-activated protein kinase subfamily. J. Biol. Chem. 271:611-614[Abstract/Free Full Text].

STAPLETON, D., E. WOOLLATT, K. I. MITCHELHILL, J. K. NICHOLL, and C. S. FERNANDEZ et al., 1997  AMP-activated protein kinase isoenzyme family: subunit structure and chromosomal location. FEBS Lett. 409:452-456[Medline].

THORNTON, C., M. A. SNOWDEN, and D. CARLING, 1998  Identification of a novel AMP-activated protein kinase beta subunit isoform that is highly expressed in skeletal muscle. J. Biol. Chem. 273:12443-12450[Abstract/Free Full Text].

WOODS, A., P. C. F. CHEUNG, F. C. SMITH, M. D. DAVISON, and J. SCOTT et al., 1996  Characterization of AMP-activated protein kinase beta and gamma subunits. Assembly of the heterotrimeric complex in vitro. J. Biol. Chem. 271:10282-10290[Abstract/Free Full Text].

YEO, G. S., I. S. FAROOQI, S. AMINIAN, D. J. HALSALL, and R. G. STANHOPE et al., 1998  A frameshift mutation in MC4R associated with dominantly inherited human obesity. Nat. Genet. 20:111-112[Medline].

ZHANG, W., A. A. DEPAOLI ROACH, and P. J. ROACH, 1993  Mechanisms of multisite phosphorylation and inactivation of rabbit muscle glycogen synthase. Arch. Biochem. Biophys. 304:219-225[Medline].

This article has been cited by other articles:

				M. E. Spurlock and N. K. Gabler The Development of Porcine Models of Obesity and the Metabolic Syndrome J. Nutr., February 1, 2008; 138(2): 397 - 402. [Abstract] [Full Text] [PDF]

				L. Miyamoto, T. Toyoda, T. Hayashi, S. Yonemitsu, M. Nakano, S. Tanaka, K. Ebihara, H. Masuzaki, K. Hosoda, Y. Ogawa, et al. Effect of acute activation of 5'-AMP-activated protein kinase on glycogen regulation in isolated rat skeletal muscle J Appl Physiol, March 1, 2007; 102(3): 1007 - 1013. [Abstract] [Full Text] [PDF]

				H. J. van Wijk, B. Dibbits, E. E. Baron, A. D. Brings, B. Harlizius, M. A. M. Groenen, E. F. Knol, and H. Bovenhuis Identification of quantitative trait loci for carcass composition and pork quality traits in a commercial finishing cross J Anim Sci, April 1, 2006; 84(4): 789 - 799. [Abstract] [Full Text] [PDF]

				E. C. Nilsson, Y. C. Long, S. Martinsson, S. Glund, P. Garcia-Roves, L. T. Svensson, L. Andersson, J. R. Zierath, and M. Mahlapuu Opposite Transcriptional Regulation in Skeletal Muscle of AMP-activated Protein Kinase {gamma}3 R225Q Transgenic Versus Knock-out Mice J. Biol. Chem., March 17, 2006; 281(11): 7244 - 7252. [Abstract] [Full Text] [PDF]

				P. L. Johnson, J. C. McEwan, K. G. Dodds, R. W. Purchas, and H. T. Blair Meat quality traits were unaffected by a quantitative trait locus affecting leg composition traits in Texel sheep J Anim Sci, December 1, 2005; 83(12): 2729 - 2735. [Abstract] [Full Text] [PDF]

				H. A. M. van der Steen, G. F. W. Prall, and G. S. Plastow Application of genomics to the pork industry J Anim Sci, June 1, 2005; 83(13_suppl): E1 - 8. [Abstract] [Full Text] [PDF]

				E. Hambrecht, J. J. Eissen, D. J. Newman, C. H. M. Smits, L. A. den Hartog, and M. W. A. Verstegen Negative effects of stress immediately before slaughter on pork quality are aggravated by suboptimal transport and lairage conditions J Anim Sci, February 1, 2005; 83(2): 440 - 448. [Abstract] [Full Text] [PDF]

				B. R. Barnes, S. Marklund, T. L. Steiler, M. Walter, G. Hjalm,, V. Amarger, M. Mahlapuu, Y. Leng, C. Johansson, D. Galuska, et al. The 5'-AMP-activated Protein Kinase {gamma}3 Isoform Has a Key Role in Carbohydrate and Lipid Metabolism in Glycolytic Skeletal Muscle J. Biol. Chem., September 10, 2004; 279(37): 38441 - 38447. [Abstract] [Full Text] [PDF]

				H. Thomsen, H. K. Lee, M. F. Rothschild, M. Malek, and J. C. M. Dekkers Characterization of quantitative trait loci for growth and meat quality in a cross between commercial breeds of swine J Anim Sci, August 1, 2004; 82(8): 2213 - 2228. [Abstract] [Full Text] [PDF]

				C. Kuhn, G. Thaller, A. Winter, O. R. P. Bininda-Emonds, B. Kaupe, G. Erhardt, J. Bennewitz, M. Schwerin, and R. Fries Evidence for Multiple Alleles at the DGAT1 Locus Better Explains a Quantitative Trait Locus With Major Effect on Milk Fat Content in Cattle Genetics, August 1, 2004; 167(4): 1873 - 1881. [Abstract] [Full Text] [PDF]

				H. Yu, N. Fujii, M. F. Hirshman, J. M. Pomerleau, and L. J. Goodyear Cloning and characterization of mouse 5'-AMP-activated protein kinase {gamma}3 subunit Am J Physiol Cell Physiol, February 1, 2004; 286(2): C283 - C292. [Abstract] [Full Text]

				H. A. Wiatrowski, B. J. W. van Denderen, C. D. Berkey, B. E. Kemp, D. Stapleton, and M. Carlson Mutations in the Gal83 Glycogen-Binding Domain Activate the Snf1/Gal83 Kinase Pathway by a Glycogen-Independent Mechanism Mol. Cell. Biol., January 1, 2004; 24(1): 352 - 361. [Abstract] [Full Text] [PDF]