A Mutation in PRKAG3 Associated with Excess Glycogen Content in Pig Skeletal Muscle

See allHide authors and affiliations

Science  19 May 2000:
Vol. 288, Issue 5469, pp. 1248-1251
DOI: 10.1126/science.288.5469.1248


A high proportion of purebred Hampshire pigs carries the dominant RN mutation, which causes high glycogen content in skeletal muscle. The mutation has beneficial effects on meat content but detrimental effects on processing yield. Here, it is shown that the mutation is a nonconservative substitution (R200Q) in the PRKAG3 gene, which encodes a muscle-specific isoform of the regulatory γ subunit of adenosine monophosphate–activated protein kinase (AMPK). Loss-of-function mutations in the homologous gene in yeast (SNF4) cause defects in glucose metabolism, including glycogen storage. Further analysis of the PRKAG3 signaling pathway may provide insights into muscle physiology as well as the pathogenesis of noninsulin-dependent diabetes mellitus in humans, a metabolic disorder associated with impaired glycogen synthesis.

The presence of a dominant mutation (denoted RN ) in Hampshire pigs with large effects on meat quality and processing yield was first recognized by segregation analysis of phenotypic data (1). Meat from RN pigs has a low ultimate pH (measured 24 hours after slaughter), a reduced water-holding capacity, and gives a reduced yield of cured cooked ham (2, 3). These effects are due to a ∼ 70% increase in muscle glycogen content in RN(RN/rn+ orRN/RN ) animals. No pathological effects of theRN mutation have been reported, and it does not cause a glycogen storage disease. Because theRN allele has been found only in Hampshire pigs, it is likely that the mutation arose in this breed and has increased in frequency due to its favorable effects on growth rate and meat content in the carcass (3). TheRN allele is of considerable economic significance in the pig breeding industry, and most breeding companies would like to eliminate the mutation because of its negative effects on processing yield.

To identify the RN mutation, which resides on pig chromosome 15 (4–6), we screened a porcine Bacterial Artificial Chromosome (BAC) library (7) and constructed a 2.5 megabase pair (Mbp) contig of the RN region (Fig. 1C). The BAC clones were used to develop new genetic markers in the form of microsatellites (MS) and single nucleotide polymorphisms (SNPs). The markers were used to construct a high-resolution linkage map based on 1019 informative meioses (8) (Fig. 1A). We could excludeRN from the region proximal to SLC11A1 and distal to SNP S1010. A porcine radiation hybrid panel was exploited for high-resolution mapping of genetic markers and coding sequences (9) (Fig. 1B). The corresponding region on human chromosome 2q and mouse chromosome 1 did not contain any obvious candidate genes for RN. Linkage disequilibrium analysis indicated complete association between RN and marker alleles at S1006 and S1007 (Fig. 1D). These marker alleles most likely define the haplotype in which theRN mutation arose. The two markers are present on the overlapping BAC clones 127G6 and 134C9, suggesting that RN may reside on the same clone or on one of the neighboring clones.

Figure 1

Genetic and physical map of theRN region. (A) Linkage map with distances in centiMorgan (cM). (B) Radiation hybrid map with distances in centiRay (cR6500); the order of framework markers is shown and the position of additional loci is indicated. (C) Selected clones from the ∼2.5 Mbp BAC contig on pig chromosome 15, showing the location of genes and genetic markers. (D) Linkage disequilibrium estimated with DISMULT (39) using a random sample of 91 Swedish Hampshire pigs.

A shotgun library of the BAC clone 127G6 was constructed and more than 1000 individual sequences were determined and assembled into contigs (10). BLAST (11) searches of the National Center for Biotechnology Information (NCBI) nucleotide database (12) yielded three convincing matches of coding sequences. Two of these were matched to human cDNA sequences or genes (KIAA0173 and CYP27A1) but did not appear to be plausible candidate genes for RN. The third coding sequence in BAC 127G6 showed significant sequence similarity to AMP-activated protein kinase (AMPK) γ subunits, including Snf4 in yeast. AMPK has a key role in regulating energy metabolism in eukaryotic cells and is homologous to the SNF1 kinase in yeast (13,14). AMPK (SNF1) is composed of three subunits (the analogous designations in yeast are given in parentheses): the catalytic α chain (Snf1) and the two regulatory subunits β (Sip1, Sip2, and Gal83) and γ (Snf4). AMPK is activated by an increase in the ratio of AMP to adenosine triphosphate (AMP:ATP). Activated AMPK turns on ATP-producing pathways and inhibits ATP-consuming pathways. AMPK can also inactivate glycogen synthase, the key regulatory enzyme of glycogen synthesis, by phosphorylation (13). Several isoforms of the three different AMPK subunits are present in mammals. In humans, PRKAA1 and PRKAA2 encode α subunits, PRKAB1 and PRKAB2 encode β subunits, and PRKAG1 and PRKAG2 encode γ subunits (15). The AMPK γ chain gene's localization in the region showing maximum linkage disequilibrium and the putative function of its protein product made it a strong positional candidate gene for RN.

The cDNA sequence of PRKAG3 was determined by reverse transcriptase–polymerase chain reaction (RT-PCR) and rapid amplification of cDNA ends (RACE) analysis with pig muscle mRNA from arn+/rn+ homozygote. A BLAST search revealed that this gene is distinct from mammalian PRKAG1 and PRKAG2 isoforms and orthologous to a human gene represented by the expressed sequence tag (EST) sequence AA178898 (GenBank). Most of the genomic sequence of this gene was recently released as part of a “working draft” sequence (GenBank AC009974) by the Genome Sequencing Center at the Washington University School of Medicine in St. Louis, Missouri. We suggest that this gene be denoted PRKAG3 because it is the third isoform of a mammalian AMPK γ subunit. The cDNA sequence of humanPRKAG3 was also determined by RT-PCR and 5′ RACE analysis with human skeletal muscle cDNA (16) (Fig. 2A). Conceptual translation of the cDNA and subsequent sequence alignment (17) revealed a protein with two regions: residues 1 to 159 show 65% sequence identity between pig and human PRKAG3, whereas residues 160 to 464 show as much as 97% identity. No known protein domains were detected in the former region, whereas the latter contains four cystathionine beta–synthase (CBS) domains shared with other AMPK γ sequences (18) (Fig. 2A).

Figure 2

Predicted protein sequence of PRKAG3, expression pattern, and mutation detection. (A) Protein sequence and alignment with other AMPKG/SNF4 sequences. The PRKAG3 mRNA sequences were determined by RT-PCR and RACE analysis except for the first two codons, which were predicted by GRAIL analysis of pig and human genomic sequences (37). The four CBS domains are indicated by a line above the codons, and theRN mutation at R200 is indicated by an arrow. PigG3, pig PRKAG3 (this study; AF214520); HumG3, human PRKAG3 (this study; AF214519); HumG1, Human PRKAG1 (U42412); HumG2, Human PRKAG2 (AJ249976); Dros, Drosophila (AF094764); Snf4, yeast (M30470); both the PRKAG2 and Drosophila sequences have longer NH2-terminal regions, but they do not show significant similarity to the NH2-terminal region of PRKAG3 and were not included. *, stop codon; , identity to master sequence; ., alignment gap. (B) Northern blot analysis of human mRNA using human PRKAG1, humanPRKAG2, and porcine PRKAG3 probes. H, heart; B, brain; Pl, placenta; L, lung; Li, liver; M, skeletal muscle; K, kidney; Pa, pancreas; S, spleen; Th, thymus; Pr, prostate; Te, testis; O, ovary; I, small intestine; C, colon (mucosal lining); PBL, peripheral blood leukocyte. Size markers are in kilobases and are on the left. (C) Pyrosequencing (40) of the reverse strand of nucleotides 595 to 599 showing the sequence (C/T)GGA(C/T). The presumed causative SNP at nucleotide 599, which corresponds to codon 200, is marked by arrows, and the linked SNP at nucleotide 595 is marked by stars. Pyrosequencing was performed with a Luc 96 instrument (Pyrosequencing AB, Uppsala, Sweden).

The candidate gene status of PRKAG3 was further strengthened by analysis of human multiple tissue Northern blots. WhereasPRKAG1 and PRKAG2 were widely expressed,PRKAG3 showed a distinct muscle-specific expression (19) (Fig. 2B). Consistent with this, ESTs representingPRKAG1 and PRKAG2 have been identified in various cDNA libraries, whereas a single PRKAG3 EST (GenBank entryAA178898) has been found in a muscle cDNA library. The muscle-specific expression is consistent with the fact that RN animals show high glycogen content in skeletal muscle but not in liver (2).

The entire PRKAG3 coding sequence was determined from onern+/rn+ and oneRN /RN homozygote by RT-PCR analysis. A total of seven nucleotide differences was found, four of which were nonsynonymous substitutions (Table 1). The screening of these seven SNPs with genomic DNA from additional rn+ and RN pigs of different breeds revealed five differentPRKAG3 alleles, but only the R200Q substitution was exclusively associated with RN (Fig. 2C). The nonconservative R200Q substitution occurs in CBS1, which is the most conserved region among AMPK γ chain isoforms, and R200 is conserved in mammalian and Drosophila AMPK γ isoforms (Fig. 2A). Interestingly, homozygosity for a nonconservative substitution (D444N) at the corresponding position in the regulatory domain of CBS causes loss of S-adenosylmethionine regulation and homocystinuria in humans (20). The 200Q allele was found in all RN animals but not in any rn+ animals from Hampshire or other breeds (21) (Table 2). This is consistent with the assumption thatRN originated in the Hampshire breed.

Table 1

Comparison of PRKAG3 sequences with associated rn + andRN alleles in parentheses for different pig populations (37). D, Duroc; H, Hampshire; L, Landrace; LW, Large White; M, Meishan; WB, Wild Boar. The numbering of codons differs slightly from the one in Fig. 2A, which is based on a multiple sequence alignment. Dash indicates identity to the top sequence.

View this table:
Table 2

Association between the PRKAG3 R200Q mutation and RN alleles among unrelated pigs from different breeds.

View this table:

Functional characterization of theRN mutation is complicated by the fact that it occurs in a regulatory subunit and by the expression of several AMPK isoforms in skeletal muscle whose functional differences have not yet been established (22). On the basis of the established roles of yeast SNF1 in glycogen utilization and of mammalian AMPK in regulation of energy metabolism, activated AMPK is expected to inhibit glycogen synthesis and stimulate glycogen degradation. Furthermore, AMPK activation in muscle leads to translocation of glucose transporter 4 (GLUT4) from an intracellular location to the plasma membrane, increased glucose uptake, and increased glycogen content in skeletal muscle (23–26). We found that AMPK kinase activity in muscle extracts was about three times higher in normal rn+ pigs than in RN pigs, both in the presence and absence of AMP (Table 3). Thus, R200Q may be a dominant negative mutation inhibiting AMP activation and glycogen degradation, but only if it interferes with multiple isoforms because the major AMPK activity in muscle appears to be associated with the PRKAG1 and 2 isoforms (22). Alternatively, it may be a gain-of-function mutation constitutively activating the holoenzyme, leading to an increased glucose transport and/or glycogen synthesis. If so, the reduced AMPK activity in RN animals is likely to reflect feedback inhibition due to the high-energy status of the muscle. More detailed functional studies are needed to distinguish between these possibilities.

Table 3

AMPK activity in muscle extracts from rn+ and RN pigs. Phosphorylation was measured as counts per minute of incorporated 32P/0.1 μl muscle extract (38). Least square means ± standard errors are reported; the significance values were obtained by an F-test with a linear regression model; n, number of animals.

View this table:

The distinct phenotype of the RN mutation indicates that PRKAG3 plays a key role in the regulation of energy metabolism in skeletal muscle. Further characterization of PRKAG3 may shed new light on muscle physiology including the adaptation to physical exercise, which is associated with increased glycogen storage (27). Finally, it will be of interest to determine whether PRKAG3 or other AMPK genes are involved in the pathogenesis of noninsulin-dependent diabetes mellitus, a disorder associated with impaired glycogen synthesis.

  • * These authors contributed equally to the work.

  • To whom correspondence should be addressed. E-mail: Leif.Andersson{at}


Stay Connected to Science

Navigate This Article