|
A correction has been published: N Engl J Med 1993;328(15):1136.
|
| |||||||||||||||||||||||||||||||||||||||||
Background The storage of glucose as glycogen in skeletal muscle is frequently impaired in patients with non-insulin-dependent diabetes mellitus (NIDDM) and their nondiabetic relatives. Despite an intensive search for candidate genes associated with NIDDM, no data have been available on the gene coding for the key enzyme of this pathway, glycogen synthase.
Methods and Results Using a human complementary DNA probe, the restriction enzyme XbaI, and Southern blot analysis, we identified two polymorphic alleles, A1 and A2, in the glycogen synthase gene. The gene was localized to chromosome 19. The A1A2 or A2A2 genotype was found in 30 percent of 107 patients with NIDDM but in only 8 percent of 164 nondiabetic subjects without a family history of NIDDM (P<0.001). The diabetic patients with the A2 allele had a stronger family history of NIDDM (P = 0.019), a higher prevalence of hypertension (P = 0.008), and a more severe defect in insulin-stimulated glucose storage (P = 0.001) than the diabetic patients with the A1 allele. The concentration of the glycogen synthase protein in biopsy specimens of skeletal muscle from the patients with the A2 allele was normal, however, suggesting that expression of the gene was unaltered. The XbaI polymorphism was due to a change of a single base in an intron.
Conclusions The XbaI polymorphism of the glycogen synthase gene identifies a subgroup of patients with NIDDM characterized by a strong family history of NIDDM, a high prevalence of hypertension, and marked insulin resistance.
The purpose of this study was to determine whether polymorphism of the human glycogen synthase gene is associated with NIDDM in general and with insulin resistance in particular. To accomplish this, we screened patients with NIDDM and subjects with no family history of NIDDM for polymorphisms in the glycogen synthase gene with a human glycogen synthase complementary DNA (cDNA) probe.
Methods
Subjects
We studied 107 unrelated patients with NIDDM and 164 unrelated nondiabetic subjects with no family history of NIDDM. The patients with NIDDM were consecutive patients seen in an outpatient clinic. The nondiabetic subjects were from the same ethnic group and were recruited from the spouses of the diabetic patients (67 subjects), members of the staff of Helsinki University (66 subjects), and nondiabetic patients in the outpatient clinic (31 subjects). In addition, 18 members of three families were examined to establish whether the possible polymorphism was inherited in a mendelian fashion. All the patients with NIDDM were treated with diet, oral hypoglycemic drugs, or both. A subgroup of 32 patients with diabetes underwent studies with the euglycemic-hyperinsulinemic clamp, and 15 diabetic patients underwent muscle biopsies for the measurement of glycogen synthase protein. None of these patients had clinical evidence of cardiac, hepatic, or renal disease or endocrine disease other than diabetes. Informed consent was obtained from all the subjects, and the study protocol was approved by the local ethics committee.
DNA Studies
Genomic DNA was extracted from frozen peripheral-blood leukocytes by standard methods11. Aliquots of DNA (5 µg) were digested with 13 different restriction enzymes (ApaI, BamHI, BglII, EcoRI, EcoRV, HindIII, MspI, NcoI, PstI, PvuII, TaqI, XmnI, and XbaI) to search for polymorphisms. The resulting fragments were separated by agarose-gel electrophoresis, transferred to nitrocellulose filters, hybridized with a human glycogen synthase cDNA probe,12 and labeled with phosphorus-32 (specific activity, 0.6 micro Ci per nanogram of DNA). The filters were prehybridized with a solution containing 4 x saline sodium citrate buffer (SSC) (1 x SSC is 0.15 M sodium chloride and 0.015 M sodium citrate), 10 x Denhardt's solution, 0.1 M phosphate buffer (pH 6.8), 1 mg of herring-sperm DNA per 20 ml of solution, and 5 percent dextran sulfate at 65 °C for 24 hours. The filters were then hybridized with the 32P-labeled glycogen synthase probe for 24 hours with the same solution and were washed twice for 20 minutes each with 3 x SSC and 0.1 percent sodium dodecyl sulfate at 65 °C. The bands were visualized by autoradiography with Fuji Rx film within two days and quantitated by densitometry.
Localization of the Chromosome for Glycogen Synthase
To determine the chromosomal location of the glycogen synthase gene, Bios chromosome panels I and II (containing human and hamster somatic-cell hybrid DNA digested with the restriction enzyme EcoRI) (Bios, New Haven, Conn.) were hybridized with the 32P-labeled glycogen synthase cDNA probe. The signals obtained were compared with the fragment pattern in control lanes containing either human or hamster DNA. The fragments specific to humans were present only in cell lines that retained human chromosome 19.
Localization of the XbaI Polymorphism
To determine the location of the restriction-fragment-length polymorphism, genomic glycogen synthase clones were isolated from the Lambda Dash genomic library (Stratagene, La Jolla, Calif.) with the glycogen synthase cDNA probe12. The putative area for a new XbaI cleavage site was sequenced directly from genomic DNA with the double-stranded DNA (dsDNA) Cycle Sequencing System (GIBCO BRL, Gaithersburg, Md.).
Glycogen Synthase Protein in Skeletal Muscle
Twenty to 40 mg of muscle was obtained by biopsy of the vastus lateralis muscle in each of eight diabetic patients with the A1A1 genotype and each of seven patients with the A1A2 genotype. The biopsy specimens were immediately placed in liquid nitrogen and were kept frozen at -70 °C until the concentration of glycogen synthase protein was measured by immunoblotting13 with a polyclonal antibody. The antibody was raised in a rabbit with use of an oligopeptide composed of 12 amino acids and specific for the carboxy terminal of the enzyme (Ser-Pro-Thr-Ser-Ser-Leu-Gly-Glu-Arg-Asn-Cys). After the blot was washed, it revealed a single 86-kd band, compatible with the molecular size of glycogen synthase. The blots were quantitated by densitometry and by a count of the radioactivity of the excised bands with a gamma counter (Wallac, Turku, Finland). DNA was quantitated with a fluorometric assay. The results were expressed as a fraction of the mean value in a control sample -- i.e., as the relative number of optical-density units per nanogram of DNA.
Clinical and Metabolic Investigations
Blood pressure was measured with the patients in the sitting position after 30 minutes of rest. Hypertension was defined as the presence of blood pressure above 160/95 mm Hg or the use of a known treatment for hypertension. Blood samples for determinations of fasting plasma glucose and serum insulin, C peptide, total and high-density lipoprotein (HDL) cholesterol, triglyceride, and hemoglobin A1c were collected after an overnight 12-hour fast. In the diabetic patients, 
-cell function was also assessed by measuring serum C-peptide concentrations six minutes after the intravenous administration of 1 mg of glucagon. The body-mass index was calculated as the weight in kilograms divided by the square of the height in meters.
The action of insulin was measured by means of a two-hour study with the euglycemic-hyperinsulinemic clamp, combined with indirect calorimetry and the infusion of [3-3H]glucose in 32 patients with NIDDM7,14. Total glucose disposal was calculated by adding the mean value of residual hepatic glucose production during the final 60 minutes of the insulin-clamp study to the mean rate of glucose infusion during the same period. The rate of glycogen synthesis was calculated as the difference between the rate of glucose disposal and the rate of glucose oxidation, as measured by calorimetry.
Assays
Plasma glucose was measured with a glucose oxidase method adapted for the Beckman Glucose Analyzer II (Beckman Instruments, Fullerton, Calif.). Serum insulin and C peptide were measured by double-antibody radioimmunoassays. The specific activity of [3-3H]glucose was measured in duplicate in supernatants of 1 M perchloric acid extracts of plasma samples, after the evaporation of radiolabeled water. Serum cholesterol, HDL cholesterol, and triglycerides were measured by specific enzymatic assays. Hemoglobin A1c was measured by high-performance liquid chromatography (reference level, 4 to 6 percent).
Statistical Analysis
The Mann-Whitney rank-sum test was used to test for differences in the distribution of continuous variables, and chi-square analysis with Yates' correction was used to test the significance of differences in frequency.
Results
XbaI Polymorphism of the Glycogen Synthase Gene
Two constant 23-kb and 5.3-kb bands were found in all the study subjects when DNA was digested with XbaI and hybridized with the glycogen synthase cDNA probe (Figure 1). In addition, two polymorphic patterns were found. Many subjects had a single 6.8-kb band, referred to as allele A1; some subjects, however, did not have this band but instead had a 5.1-kb and a 1.7-kb band, together referred to as allele A2. Subjects who had only the A1 or the A2 allele were considered homozygous for the corresponding genotypes (A1A1 and A2A2 in Figure 1), whereas subjects with all three bands were considered heterozygous (A1A2). It was evident from the family studies that the alleles were inherited in a mendelian fashion. The polymorphic A2 allele was due to a change of a single base in an intron, in which thymidine was substituted for cytidine (CCTAGA to TCTAGA), creating a new XbaI site 302 base pairs upstream from position 1970 of the cDNA.
|
|
|
Studies using the euglycemic-hyperinsulinemic clamp were performed in 32 patients with NIDDM, 16 of whom had the A1A1 genotype and 16 of whom had either the A1A2 or the A2A2 genotype. These groups were matched with respect to sex, age (56 and 58 years, respectively), and body-mass index (27.4 ±1.0 and 27.9 ±1.0). The mean total glucose disposal was lower in the patients with the A1A2 or the A2A2 genotype than in those with the A1A1 genotype (4.0 ±0.3 vs. 6.0 ±0.6 mg [22.3 ±1.8 vs. 33.4 ±3.6 micromol] per kilogram of lean body mass per minute; P = 0.026) (Figure 2). Among the patients with NIDDM, the glucose oxidation rate did not differ between those with the A2 allele and those without it (2.7 ±0.2 vs. 2.5 ±0.2 mg [15.5 ±1.0 vs. 14.1 ±0.8 micromol] per kilogram of lean body mass per minute). The rate of nonoxidative glucose metabolism (glycogen synthesis) was reduced by half in the diabetic patients with the A1A2 or the A2A2 genotype as compared with those with the A1A1 genotype (1.3 ±0.3 vs. 3.4 ±0.5 mg [7.0 ±1.4 vs. 19.0 ±2.8 micromol] per kilogram of lean body mass per minute; P<0.001).
|
Glycogen synthase protein was measured in muscle-biopsy specimens from eight diabetic patients with the A1A1 genotype and seven patients with the A1A2 genotype. The protein content was similar in the patients with the two genotypes (Figure 3). The mean concentrations of glycogen synthase protein in the patients with the A1A1 and A1A2 genotypes were 2.4 ±0.5 and 3.4 ±0.9 relative units per nanogram of DNA, respectively.
|
The rationale for including the glycogen synthase gene in the list of candidate genes for NIDDM was the finding that activation of glycogen synthase by insulin was impaired in relatives of patients with NIDDM10. The glycogen synthase cDNA used in this study includes the vast majority of the coding sequence of the gene (nucleotides 414 to 3535). The restriction enzyme XbaI has two cleavage sites, corresponding to nucleotides 3260 and 3333. Both sites are outside the coding region of the gene; the stop codon is at position 2374 of the cDNA12. By subcloning and direct sequencing of genomic DNA, the XbaI polymorphism was found to be due to a point mutation (CCTAGA to TCTAGA) in an intron 302 base pairs upstream from position 1970 of the cDNA. A recent screening of eight Danish patients with NIDDM to identify mutations in the glycogen synthase gene that used single-stranded conformational polymorphism with primers covering the entire coding sequence of the gene did not reveal any mutations,15 but it is not known whether any of the Danish patients had the A2 allele. It is therefore unlikely that the mutation we found would result in any structural change of the glycogen synthase enzyme. This does not exclude the possibility that it could influence expression of the gene in skeletal muscle.
To address this question, we measured the concentrations of the glycogen synthase protein in skeletal-muscle-biopsy specimens. The results suggest that expression of the glycogen synthase gene is normal in skeletal muscle of patients with the A1A2 allele. There was a fourfold variation in protein concentrations among these patients and a four- to fivefold variation in rates of glucose uptake. It should be noted that these results only suggest that the A2 allele is a marker associated with NIDDM, not that a mutation in the glycogen synthase gene is the cause of NIDDM in these patients. The possibility that the association with the restriction-fragment-length polymorphism is due to a disequilibrium of linkage with another gene should also be kept in mind.
Hypertension was twice as frequent in the patients with NIDDM who had the A2 allele as in the patients who did not. More important, 36 percent of the normal subjects with the A2 allele had hypertension, as compared with 9 percent of those with the A1 allele. Both hypertension and NIDDM have been considered components of the so-called metabolic syndrome, the common denominator of which is insulin resistance16. In fact, normal subjects with hypertension have insulin resistance that is caused by impaired glycogen synthesis17. In patients with NIDDM who had the A2 allele, total glucose metabolism was reduced, mainly because of a reduction in the rate of glycogen synthesis in skeletal muscle. It should be remembered, however, that all the diabetic patients were severely insulin-resistant as compared with normal subjects7.
In conclusion, the A2 allele of the human glycogen synthase gene on chromosome 19 identifies a subgroup of patients with NIDDM who have a strong family history of NIDDM and in whom hypertension and insulin resistance are prevalent. The A2 allele can thus be considered a genetic marker for NIDDM.
Supported by grants from the Sigrid Juselius Foundation, the Perklen Foundation, Finska Lakaresallskapet, and the Nordisk Insulin Foundation.
We are indebted to Michelle Browner, Ph.D., for kindly providing us with the glycogen synthase probe and to Leena Peltonen, M.D., for help with the chromosomal localization of the glycogen synthase gene.
Source Information
From the Fourth Department of Medicine (L.C.G., C.S.-J., A.E., E.W., J.E., A.F.-K., C.S.) and the Department of Biochemistry (M.K., P.N.-I., E.K.), Helsinki University, and the Finnish Red Cross Blood Transfusion Service (S.K.), both in Helsinki, Finland.
Address reprint requests to Dr. Groop at the Fourth Department of Medicine, Helsinki University Hospital, Unioninkatu 38, SF-00170 Helsinki, Finland.
References
-cell, muscle, liver: a collusion responsible for NIDDM. Diabetes 1988;37:667-687. [Medline]
| |||||||||||||||||||||||||||||||||||||||||
Related Letters:
Polymorphism of the Glycogen Synthase Gene and Non-Insulin-Dependent Diabetes Mellitus
Zouali H., Velho G., Froguel P., Kadowaki T., Kadowaki H., Yazaki Y., Groop L., Schalin-Jantti C., Lehto M.
Extract |
Full Text
N Engl J Med 1993;
328:1568-1569, May 27, 1993.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
