Polymorphism in the 5′-Flanking Region of Human Glutamate-Cysteine Ligase Modifier Subunit Gene Is Associated With Myocardial Infarction
Background— Human glutamate-cysteine ligase (GCL) is a rate-limiting enzyme for the synthesis of glutathione that plays a crucial role in antioxidant defense mechanisms in most mammalian cells, including vascular cells. Oxidants transcriptionally upregulate GCL genes for glutathione synthesis, providing a protective mechanism against oxidative stress-induced cellular dysfunction. This study examined the hypothesis that variation in the GCL genes may be associated with coronary artery disease in which oxidative stress plays a pathogenetic role.
Methods and Results— We searched for the common variants in the 5′-flanking region of the GCL modifier subunit (GCLM) gene in patients with myocardial infarction (MI). We found a polymorphism (−588C/T) in which the T allele showed lower promoter activity (40% to 50% of C allele) in response to oxidants in the luciferase reporter gene assay. Allele frequencies were determined by polymerase chain reaction-based analysis of restriction fragment length polymorphism in 429 patients with MI and 428 control subjects (as defined by angiography) in Kumamoto Prefecture, Japan. The frequency of the T polymorphism was significantly higher in the MI group than in the control group (CT and TT genotypes: 31.5% in MI group versus 19.2% in control group; P<0.001). In multiple logistic regression analysis, the T polymorphism was a risk factor for MI independent of traditional coronary artery disease risk factors (odds ratio, 1.98; 95% confidence interval, 1.38 to 2.83; P<0.001).
Conclusions— These findings suggest that the −588T polymorphism of the GCLM gene may suppress GCLM gene induction in response to oxidants and that it is a genetic risk factor for MI.
Received March 20, 2002; revision received April 11, 2002; accepted April 12, 2002.
Oxidative stress plays a crucial role in the pathogenesis of atherothrombotic complications in various occlusive vascular diseases including coronary artery disease (CAD).1–4⇓⇓⇓ Cells have several antioxidant defense mechanisms, such as the induction of several cellular antioxidants, including glutathione (GSH).5,6⇓ Thus, a weakened defense system against oxidative stress might partly contribute to the development of CAD. GSH (tripeptide thiols) is a major, naturally occurring antioxidant, and it has a predominant role in the regulation of the intracellular redox state and protects cells from oxidative injury in systems that scavenge radicals, eliminate lipid peroxidation products through GSH peroxidases, and repair oxidant damage.7 GSH is synthesized within most mammalian cells by the action of human glutamate-cysteine ligase (GCL) and GSH synthetase.7,8⇓ GCL catalyzes the rate-limiting step in de novo GSH synthesis.7,8⇓ GCL is a heterodimer composed of catalytic and modifier (GCLM) subunits.7,8⇓ GSH production is paralleled with the induction of GCL gene expression, which is regulated primarily at the level of transcription.5,8–11⇓⇓⇓⇓ Both genes have putative oxidative stress-responsive elements in their promoter/enhancer regions.9–11⇓⇓ Thus, GCL genes are transcriptionally upregulated for GSH synthesis by oxidants through the cis-acting DNA elements, providing a protective mechanism against oxidative stress.5,8–11⇓⇓⇓⇓
We hypothesized that the ability of individuals to regulate antioxidant defense mechanisms, such as the increase in antioxidant levels in response to oxidants, may genetically vary because all of individuals with coronary risk factors, which impose oxidative stress as a common mechanism for their atherogenic effects, may not necessarily develop CAD. In the present study, we examined the hypothesis that functional variations in the 5′-flanking region of GCL genes may be present and that they may be implicated in myocardial infarction (MI). Previous reports showed that several mutations in the catalytic GCL gene caused hemolytic anemia and neurological disorders, but there was no description of cardiovascular changes in the patients with the gene mutations.12,13⇓ Thus, the present study focused on the GCLM gene.
The study included 429 consecutive patients with previous MI and 428 consecutive control subjects who underwent coronary angiography and left ventriculography in Kumamoto University Hospital; all subjects were analyzed with regard to allele frequency. Clinical characteristics of the study patients and controls are shown in Table 1. All of the patients and controls were ethnic Japanese and partly overlapped with those studied in our previous reports.14 Criteria for MI included chest symptoms, characteristic ECG changes, and elevation of creatine kinase enzyme to more than twice the upper limit of normal. The findings in coronary angiography and left ventriculography supported the diagnosis of MI. Evaluation of risk factors was performed in the overnight fasting condition in the chronic phase of MI in all MI patients. The control subjects underwent cardiac catheterization for atypical chest pain during a study period overlapping that of the patients with MI. The control subjects had angiographically documented normal coronary arteries (<10% stenosis), normal ventriculography, and no clinical evidence of coronary artery spasm and syndrome X. All patients and subjects gave informed consent for this study. The study protocol followed the national guidelines for genetic analysis in Japan and was approved by the ethics committee at Kumamoto University Hospital.
Identification of Polymorphisms in the Promoter Region
Genomic DNA was extracted from peripheral blood lymphocytes with the use of the phenol chloroform protocol.14 The promoter region of the GCLM gene was amplified by polymerase chain reaction (PCR) from genomic DNA obtained from 12 patients with MI and 12 control subjects using 10 sets of primer pairs covering the region from −1900 to 85 on the basis of the published sequence of the GCLM promoter sequence.10,11⇓ The PCR products were sequenced with the Dye Terminator Cycle Sequencing Ready Reaction Kit (ABI PRISM) on an ABI Genetic Analyzer 373S (PE Biosystem). Sequences were analyzed and compared among the patients and the control subjects to detect polymorphisms.
We identified two polymorphisms, −588C/T and −23G/T, that were completely linked (Figure 1). Genotypes of each polymorphism were determined by PCR-based method of restriction fragment length polymorphism by investigators (S.S. and S.K.) who had no knowledge of the angiographic and clinical data of MI cases and control subjects. The −588C/T polymorphism creates a novel site for the MspI restriction enzyme (C/CGG) in the presence of the C allele. A set of primers was designed to amplify a 329-base pair (bp) fragment of the GCLM promoter by PCR (forward 5′-CTCAAGGGCAAAGACTCA-3′; reverse 5′-CCGCCTGGT-GAGGTAGACAC-3′), encompassing the −588C/T polymorphic site and an additional site for MspI as a control. Subjects with the CC genotype were identified by the presence of 200-, 84-, and 45-bp products on gel electrophoresis; those with the CT genotype were identified by 200-, 129-, 84-, and 45-bp bands; and those with TT genotypes were identified by 200- and 129-bp bands, as shown in Figure 2.
Similarly, the −23G/T polymorphism creates a novel site for the HinP1I restriction enzyme (G/CGC) in the presence of the G allele. A set of primers was designed to amplify a 127-bp fragment of the GCLM promoter by PCR (forward 5′-GCTCCCTCTCGGGTCTCTCT-3′; reverse 5′-GCTGTCGGTGCCCATGGCAC-3′), encompassing the −23G/T polymorphic site. Subjects with the GG genotype were identified by the presence of 88- and 39-bp bands; those with the GT genotype were identified by 127-, 88-, and 39-bp bands; and those with TT genotypes were identified by 127-bp bands, as shown in Figure 2.
HeLa (human cervical carcinoma cell line) cells and THP-1 (human monocytic cell line) cells were obtained from the Health Science Research Resources Bank. They were cultured in DMEM and RPMI 1640, respectively, with fetal calf serum. Primary cultures of human umbilical vein endothelial cells (HUVECs) were obtained as previously described.15 HUVECs at the second passage were used in the present study.
Recombinant Plasmid Constructions
Recombinant expression vectors were created by cloning restriction fragments isolated from the 5′-flanking sequences of the GCLM gene into pGL3-Basic (Promega) to determine promoter activity. Two DNA fragments covering the region from −1338 to 85 were amplified by PCR with genomic DNA (from −588C/−23G wild-type homozygote or −588T/−23T variant homozygote) and sequenced. The fragments were isolated by KpnI/XhoI restriction digestion and cloned into the KpnI/XhoI sites of pGL3-Basic, thus creating the recombinant plasmids pGL3-GCLMwildtype (with −588C/−23G) and pGL3-GCLMvariant (with −588T/−23T). These plasmids were subjected to digestion with additional restriction enzymes to generate other constructs containing only 1 of the 2 polymorphisms. The KpnI/SacII DNA fragment (−1338 to −345; containing −588C) from pGL3-GCLMwildtype was inserted into the KpnI/SacII-digested pGL3-GCLMvariant (containing −23T) to create pGL3-GCLM-588C/-23T (containing the major C allele at −588 and the minor T allele at −23). The SacII/PstI DNA fragment (−345 to 62; containing −23G) from pGL3-GCLMwildtype was ligated into the SacII/PstI-digested pGL3-GCLMvariant (containing −588T) to create pGL3-GCLM-588T/-23G (containing the minor T allele at −588 and the major G allele at −23).
Luciferase Reporter Gene Assay
DNA was introduced into the cells by the liposome-mediated transfection method. Briefly, subconfluent cultures of HUVECs, THP-1, and HeLa cells were transfected with cationic lipid reagent. The transfection used 1 μg of pGL3 Luciferase Reporter Vector with various GCLM promoters and 0.1 μg of pRL-TK control vector (Promega) with the herpes simplex virus thymidine kinase promoter to provide constitutive expression of Renilla luciferase expression. LIPOFECTIN Reagent and PLUS Reagent (GIBCO BRL) were used for the experiments with HeLa cells and HUVECs, and Effectene Reagent (QIAGEN) was used with THP-1 cells. Twenty-four hours after the transfection, the transfected cells were treated for 18 hours with or without one of the following oxidants: 10 to 100 μmol/L of tert-butylhydroquinone (tBHQ), 10 μmol/L of menadione, or 100 μmol/L of H2O2. The treated cells were harvested with passive lysis buffer (Promega). Luciferase activity was measured with the dual-luciferase assay system (Promega) and a luminometer. Luciferase levels were expressed in arbitrary units after normalization to Renilla luciferase levels.
Preparation and Culture of Human Monocytes
To examine the effects of −588C/T polymorphism on GCLM mRNA expression, human peripheral blood monocytes were isolated16 from 38 subjects (CC genotype, n=19; CT genotype, n=19) who were selected from the subjects examined for allele frequency. Coronary risk factors and frequencies of previous MI were matched between the selected subjects with the CC genotype and CT genotype. The mononuclear cells were plated in 6-well dishes for 2 hours. The nonadherent cells were then removed by washing. After 5 days of culture, the monocytes-macrophages were washed and then treated for 18 hours with or without 50 μmol/L of tBHQ. The treated cells were assayed for GCLM mRNA expression.
GCLM mRNA Expression in Human Monocytes-Macrophages
Total RNA was extracted from the treated human monocytes-macrophages with a Qiagen RNeasy kit and DNase I (QIAGEN). The expression of GCLM mRNA was quantified by a real-time one-step reverse transcriptase PCR assay based on the 5′ nuclease activity of Taq polymerase and with the use of an ABI PRISM 7700 Sequence Detector (PE Applied Biosystems). The PCR primers and the TaqMan fluorogenic probe were designed with the use of Primer Express (PE Applied Biosystems) from the GenBank databases (NM002061) to amplify a 113-bp product in the GCLM cDNA sequence as follows: sense, 5′-TTGACATGGCCTGTTCAGTCCT-3′; antisense, 5′-CCAGTAAGGCTGTAAATGCTCCA-3′; and TaqMan probe, 5′-FAM-AGTTGCACAGCTGGATTCTGTGATC- ATTGC-TAMRA-3′. The GAPDH housekeeping gene was used for normalization of GCLM gene expression. The levels of mRNA expression of GCLM were expressed as a ratio of its own expression to GAPDH expression.
Plasma GSH Assay
EDTA-plasma was obtained after an overnight fast and stored at −80°C until assay after the addition of 50 μmol/L acivicin (γ-glutamyltranspeptidase inhibitor). Plasma GSH levels were measured by the high performance liquid chromatography-electrochemical method17 in 84 subjects (CC genotype, n=48; CT genotype, n=34; TT genotype, n=2). These subjects were selected from the study subjects examined for allele frequency to match the risk factors and frequency of previous MI between patients with CT and TT genotypes and those with the CC genotype.
To evaluate the −588T polymorphism as an independent risk factor between patients with MI and control subjects, multiple logistic regression analysis was performed with the following factors as categorical covariates: age (≥70 years), sex (men), smoking history (defined as smoking ≥10 cigarettes per day for ≥10 years), hypertension (>140/90 mm Hg or taking antihypertensive medication), diabetes mellitus (according to American Diabetes Association guidelines18), hypercholesterolemia (>220 mg/dL or the use of lipid-lowering medications), body mass index (>26 kg/m2), and −588T polymorphisms (TT and CT genotypes). Mean values of continuous variables with normal distributions and frequencies between groups were compared with an unpaired t test and χ2 analysis or Fisher’s exact test, respectively. Statistical significance was defined as P<0.05. Statistical analysis was performed with StatView 5.0 (SAS Institute, USA).
Identification of GCLM Polymorphism and GCLM Genotypes
Two novel polymorphisms (−588C/T and −23G/T) were identified in the 5′-flanking region of the GCLM gene (Figure 1). Figure 2 shows representative agarose gels loaded with PCR products after digestion with MspI (−588C/T) or HinP1I (−23G/T). The 2 polymorphisms were completely linked. The −588TT, CT, and CC genotypes were present in 16 (3.7%), 119 (27.8%), and 294 (68.5%) of 429 patients with MI, respectively, and they were present in 2 (0.5%), 80 (18.7%), 346 (80.8%) of 428 control subjects, respectively. The genotype distribution was consistent with the population being in Hardy-Weinberg equilibrium. In the analysis of the recessive, additive, and dominant effects of the −588T polymorphism, each frequency was significantly higher in the MI group than in the control group (P<0.001 in all), as shown in Table 2. The −588T polymorphism (TT and CT genotypes), male sex, high age (≥70 years), hypercholesterolemia, diabetes mellitus, and hypertension were variables differing significantly and independently between patients with MI and control subjects in multiple logistic regression analysis, as shown in Table 3. This polymorphism remained a significant risk factor for MI independently of other traditional risk factors in multivariate statistical analysis after matching sex and age between cases and controls (data not shown).
Promoter Activities of the GCLM Gene
HUVECs, THP-1 cells, and HeLa cells transfected with pGL3-GCLMvariant (−588T/−23T) had luciferase levels comparable to those transfected with pGL3-GCLMwildtype (−588C/−23G) at the nontreated control condition, as shown in Figure 3. The luciferase levels were induced in cells with pGL3-GCLMwildtype but not those with pGL3-GCLMvariant when cells were incubated with 10 to 100 μmol/L of tBHQ for 18 hours, as shown in Figure 3. Similar results were obtained with the experiments that used other oxidants (ie, menadione or H2O2; data not shown). The induced luciferase levels in response to tBHQ were significantly lower in cells transfected with pGL3-GCLM-588T/-23G (containing the minor T allele at −588 and the major G allele at −23) than those with pGL3-GCLMwildtype (HeLa cells), as shown in Figure 3. The induced levels were comparable between cells with pGL3-GCLM-588C/-23T (containing the major C allele at −588 and the minor T allele at −23) and those with pGL3-GCLMwildtype. Thus, the −588T polymorphism reduced the promoter activity of the GCLM gene at the condition exposed to tBHQ, whereas the −23T polymorphism did not have a significant effect.
GCLM mRNA Expression in Human Cultured Monocytes-Macrophages
GCLM mRNA expression levels in monocytes-macrophages from patients with the CT genotype were comparable to those from the risk factor-matched patients with the CC genotype at the nontreated control condition. The mRNA levels were induced in both genotypes in response to 50 μmol/L of tBHQ for 18 hours, but the induced levels were significantly lower in cells with the CT genotype than those with the CC genotype after incubation with tBHQ, as shown in Figure 4.
Plasma GSH Levels
Plasma GSH levels were significantly lower in CT and TT genotypes than in the CC genotype, as shown in Figure 5.
The present study found 2 polymorphisms (−588C/T and −23G/T), which are completely linked, in the 5′-flanking region of GCLM gene. The present reporter gene assay showed that the variants in the GCLM gene suppressed promoter activity in response to oxidants and that the −588T polymorphism was responsible for the functional change between the 2 polymorphisms. The association study showed that the frequency of the −588T polymorphism was significantly higher in the MI group than in the control group and that this polymorphism was an independent risk factor for MI. Furthermore, the present study showed that GCLM mRNA expression in the cultured human monocytes-macrophages was increased in response to oxidants, but the induction was less in the cells from CT subjects compared with CC subjects. Thus, these data indicate that the −588T polymorphism may suppress the increase in GCLM gene expression to oxidative stress and that it is a genetic risk factor for MI. In addition, plasma GSH levels were significantly lower in subjects with the T allele than in those without the T allele. The T allele may possibly weaken the intracellular production of GSH in response to oxidative stress, leading to the increase in susceptibility to the oxidant-induced vascular injury that is thought to occur as part of the pathogenesis of CAD.
GSH is synthesized from its constituent amino acids in 2 sequentials by GCL and GSH synthetase.7,8⇓ GCL is the rate-limiting enzyme in de novo GSH synthesis, whereas GSH synthetase apparently has no regulatory role. The GCLM gene has several putative oxidative stress-responsive elements in the promoter/enhancer region.10,11⇓ When cells are challenged with sublethal oxidative stress or GSH depletion, GCLM mRNA expression was induced through activation of oxidative stress-responsive elements in the promoter region.10,11,19⇓⇓ These result in de novo GSH synthesis and provide a protective/adaptive mechanism against oxidative stress. Although there is no putative enhancer element that contains the −588 position on the computer-based data research, Moinova et al10 showed that the nucleotides −718 to −344 contribute to the increase in the promoter activity in response to β-naphthoflavone, a potent inducer of several phase II enzymes. Thus, it is possible that the −588T allele might modify the binding of nuclear proteins to unidentified cis-elements around the −588 position, leading to suppression of oxidant-induced expression of the GCLM gene.
A recent study showed that low serum levels of GSH are a risk factor for CAD.20 Previous in vitro studies demonstrated that GSH depletion inhibits nitric oxide production in endothelial cells.21 We and others previously showed that GSH supplementation improves endothelial vasomotor dysfunction, an early event observed in atherosclerotic arteries, in the coronary arteries of patients with coronary risk factors or coronary spastic angina that imposes oxidative stress.22–24⇓⇓ Furthermore, GSH suppresses the induction of proatherothrombotic molecules in the cultured human monocytes exposed to oxidative stress.15 These atheroprotective effects of GSH support the hypothesis that low GSH levels may potentially contribute to the development of atherosclerotic CAD.
The present study is limited by the relatively small number of studied patients. Furthermore, the precise mechanisms by which −588C/T polymorphism in the GCLM gene is linked to the genesis of MI remain largely undefined. We cannot exclude the possibility that this polymorphism is a marker for another functional variant within this or adjacent genes. In addition, only survivors of acute MI were enrolled as MI patients in the present study. A longitudinal study with a large number of study patients with homogeneous risk is thus required to assess the precise role of this gene variant in the pathogenesis of CAD.
In conclusion, the present study demonstrated that 2 polymorphisms (−588C/T and −23G/T) occur in the 5′ flanking region of the GCLM gene, that the −588T polymorphism in the GCLM gene suppresses the induction of the GCLM gene expression to oxidative stress, and that this polymorphism is significantly associated with MI. Thus, the −588T polymorphism in the GCLM gene is a genetic risk factor for MI.
This study was supported in part by grants-in-aid C13670728 and C12670680 from the Ministry of Education, Science, and Culture of Japan and a Smoking Research Foundation Grant for Biomedical Research, Tokyo, Japan.
- ↵Griendling KK, Alexander RW. Oxidative stress and cardiovascular disease. Circulation. 1997; 96: 3264–3565.
- ↵Cai H, Harrison DG. Endothelial dysfunction in cardiovascular disease. Circ Res. 2000; 87: 840–844.
- ↵Rahman I, MacNee W. Lung glutathione and oxidative stress: implications in cigarette smoke-induced airway disease. Am J Pathol. 1999; 277: L1067–L1088.
- ↵Mulcahy RT, Wartman MA, Bailey HH, et al. Constitutive and β-naphthoflavone-induced expression of the human γ-glutamylcysteine synthetase heavy subunit gene is regulated by a distal antioxidant response element/TRE sequence. J Biol Chem. 1997; 272: 7445–7454.
- ↵Moinova HR, Mulcahy RT. An electrophile responsive element (EpRE) regulates β-naphthoflavone induction of the human γ-glutamylcysteine synthetase regulatory subunit gene. J Biol Chem. 1998; 273: 14683–14689.
- ↵Galloway DC, Blake DG, Shepherd AG, et al. Regulation of human γ-glutamylcysteine synthetase: co-ordinate induction of the catalytic and regulatory subunits in HepG2 cells. Biochem. J,. 1997; 328: 99–104.
- ↵Beutler E, Gelbart T, Kondo T, et al. The molecular basis of a case of γ-glutamylcysteine synthetase deficiency. Blood. 1999; 94: 2890–2894.
- ↵Ristoff E, Augustson C, Geissler J, et al. A missense mutation in the heavy subunit of γ-glutamylcysteine synthetase gene causes hemolytic anemia. Blood. 2000; 95: 2193–2197.
- ↵Doi H, Kugiyama K, Oka H, et al. Remnant lipoproteins induce proatherothrombogenic molecules in endothelial cells through a redox-sensitive mechanism. Circulation. 2000; 102: 670–676.
- ↵Oka H, Kugiyama K, Doi H, et al. Lysophosphatidylcholine induces urokinase-type plasminogen activator and its receptor in human macrophages partly through redox-sensitive pathway. Arteriorscler Thromb Vasc Biol. 2000; 20: 244–250.
- ↵The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care. 1997; 20: 1183–1197.
- ↵Morrison JA, Jacobsen DW, Sprecher DL, et al. Serum glutathione in adolescent males predicts parental coronary heart disease. Circulation. 1999; 100: 2244–2247.
- ↵Murphy ME, Piper HM, Watanabe H, et al. Nitric oxide production by cultured aortic endothelial cells in response to thiol depletion and replenishment. J Biol Chem. 1991; 266: 19378–19383.
- ↵Kugiyama K, Ohgushi M, Motoyama T, et al. Intracoronary infusion of reduced glutathione improves endothelial vasomotor response to acetylcholine in human coronary circulation. Circulation. 1998; 97: 2299–2301.
- ↵Kugiyama K, Miyao Y, Sakamoto T, et al. Glutathione attenuates coronary constriction to acetylcholine in patients with coronary spastic angina. Am J Physiol. 2001; 280: H264–H271.