Association of a Missense Glu298Asp Mutation of the Endothelial Nitric Oxide Synthase Gene with End Stage Renal Disease

Glomerular microcirculation is involved in the deterioration of renal function. Among several factors that regulate renal hemodynamics, NO has been reported to be critical (1)(2). In the vascular endothelium, NO is produced by endothelial NO synthase (eNOS). NO production can be influenced by polymorphisms of the eNOS gene. Polymorphisms in exons may alter the three-dimensional structure of the enzyme, and those in introns may change the transcriptional activity. These can lead to a decrease in NO production and, subsequently, an increase in arterial pressure or intraglomerular hypertension. Several studies have shown that the polymorphisms of the eNOS gene are associated with hemodynamics. Homozygosity of the eNOS4 a allele has been shown to be a risk factor for coronary artery disease among smokers (3). We (4) and Yokoyama et al. (5) demonstrated that a polymorphism of the variable numbers of tandem repeats in intron 4 of the eNOS gene is associated with the progression of nondiabetic end stage renal disease (ESRD), but we failed to demonstrate the association between diabetic renal failure and the polymorphism. The genetic background for the progression of diabetic nephropathy to ESRD is still unclear. Furthermore, in some patients diabetic nephropathy does not show any association with conventional risk factors such as hyperglycemia (6). We speculate that there must be a genetic factor that causes the deterioration of diabetic nephropathy to ESRD. Recently, a Glu298Asp missense polymorphism within exon 7 of the eNOS gene was reported to be associated with essential hypertension, vasospastic angina, and myocardial infarction in a Japanese population (7)(8)(9). In this study, we explore the association between the Glu298Asp mutation and ESRD. To evaluate whether the Glu298Asp mutation was associated with ESRD, we examined 159 patients with ESRD undergoing maintenance hemodialysis (96 men and 63 women; mean age, 57.1 years; age range, 17–85 years) and 270 genetically unrelated, apparently healthy control subjects (195 men and 75 women; mean age, 49.3 years; urine or blood examinations of the subjects were normal). All patients and controls were Japanese. Informed consent was obtained from each person enrolled in this study. The underlying causes of end stage renal failure were chronic glomerulonephritis (CGN; n = 68), diabetes mellitus (DM; n = 48), hypertension (n = 17), polycystic kidney disease (n = 13), lupus nephritis and vasculitis (n = 7), reflux and obstructive nephropathy (n = 5), and interstitial nephritis (n = 1; Table 1A ). Genomic DNA was extracted from peripheral mononuclear cells with a DNA extraction reagent set (Wako Pure Chemical) and was stored at 4 °C in Tris-EDTA buffer until analysis. To determine the genotypes of eNOS codon 298, we originally designed a set of primers (forward, 5-gaaacggtcgcttcgacgtgct; reverse, 5-ccacccagtcaatccctttg) that produced 140-bp amplified fragments. Approximately 100 ng of genomic DNA was amplified in a total volume of 25 μL containing 2.5 μL of Thermophilic DNA Polymerase Buffer (supplied with Taq polymerase; Promega), 3.0 mmol/L MgCl2, 200 μmol/L deoxynucleotide triphosphates, 0.5 μmol/L each primer, and 1 U of Taq polymerase. The PCR amplifications were 35 cycles performed in a GeneAmp PCR System 9700 thermocycler (Applied Biosystems) under the following conditions: denaturation at 94 °C for 30 s, annealing at 56 °C for 30 s, extension at 72 °C for 1 min. Predenaturation was at 94 °C for 5 min, and final extension was for 7 min. PCR products were then digested by MboI, which produces two fragments 50 and 90 bp in length from alleles with glutamic acid (G) at codon 298 (G, wild type), and by BanII, which produces two fragments 55 and 85 bp in length from alleles with aspartic acid (T, mutant) at the same codon. The digested products were analyzed by 3% agarose gel electrophoresis. The frequencies of homozygous or heterozygous Glu298Asp mutations (T/T + T/G) were 7.4% (20 of 270) in the control group, 22.0% (35 of 159) in all patients, 22.5% (25 of 111) in the nondiabetic group, 23.5% (16 of 68) in the CGN group, and 20.8% (10 of 48) in the diabetic group. The distribution of Glu298Asp genotypes in controls was in the Hardy-Weinberg equilibrium. To correct for the contribution of age and sex, we also performed a multiple logistic regression analysis with the genotype at codon 298 of the eNOS gene as the dependent variable, and entered age, sex, and underlying disease (diabetic or nondiabetic renal disease) into the model. Statistical analyses were performed with SPSS, Ver. 6.1 for Macintosh. Multiple logistic regression analysis revealed that the frequencies of the Glu298Asp mutation in all patients, in the nondiabetic group, in the CGN group, and in the diabetic group were significantly higher than those in the control subjects (P = 0.0001, 0.0019, 0.0021, and 0.0117, respectively). Odds ratios were 2.8–3.2 (Table 1B ). The patients enrolled in this study had been on hemodialysis for 9.2 (mean) years in the nondiabetic group, and 4.2 (mean) years in the diabetic group. To evaluate the survival bias in analyzing the relationship between the Glu298Asp mutation and progression to chronic renal failure, we divided the patients into two groups according to the duration of hemodialysis; among the diabetics, patients who had experienced >5 years of hemodialysis were placed in the long-term survival group, whereas patients with hemodialysis <5 years were placed in the short-term group. A similar division into two groups was made among the nondiabetic patients: the long-term survival group (hemodialysis ≥10 years) and the short-term group (hemodialysis <10 years). The two groups were compared with regard to the frequencies of the Glu298Asp mutation by the χ2 test. Statistical analysis revealed that there were no significant differences between of the two groups. In the present study, the Glu298Asp mutation of the eNOS gene was increased in patients with ESRD (22.0%), whose underlying causes of chronic renal failure included not only nondiabetic renal diseases (22.5%) but also DM (20.8%), compared with controls (7.4%; Fig. 1 ). The results demonstrate that patients with renal diseases, including DM, who have the Glu298Asp mutation are at higher risk for ESRD. This is the first report, as far as we can tell, that refers to the significance of the Glu298Asp mutation in patients with renal diseases. We could not evaluate the influence of hypertension because most of the patients had received antihypertensive drugs by the time dialysis was initiated. Presence of a family history of hypertension was not significantly different according to the genotype. In previous reports, the relative risk for the Glu298Asp variant was 1.5 in myocardial infarction (9), 2.3 in essential hypertension (7), and 2.73 in coronary spasm (8). The odds ratio in our study [3.2 (CI, 1.8–5.8)] indicates a similarly strong association between ESRD and the Glu298Asp variant, perhaps independent of hypertension or other disorders. We also evaluated the effects of the Glu298Asp mutation on mortality. Although the frequency of the Glu298Asp mutant in the diabetic long-term survival group was as low as 14.3%, we could not find a statistically significant difference in the frequencies of the Glu298Asp mutation between short-term survival groups and long-term survival groups in both diabetic and nondiabetic patients. Our results may indicate that the Glu298Asp mutation is not a lethal factor but a risk factor for ESRD, although its association with relatively lethal diseases such as myocardial infarction has been reported. Further examination will be necessary. We did not measure the enzyme activity of the variant. Yoshimura et al. (8) suggested that according to a computer analysis, the Glu298Asp mutation might affect enzyme activity because glutamic acid at codon 298 is in the middle of an α helix and the mutation produced a tight turn. Enzyme activities of the polymorphic variants should be measured to confirm the hypothesis that the polymorphism is responsible for the progression of renal diseases through decreased endothelial NO production. However, to date there have been no reports that confirm the diminished eNOS activity in vivo or in vitro. Tsukada et al. (10) reported that plasma NOx (nitrate + nitrite) concentrations of subjects with the eNOS4 a allele of the variable number of tandem repeats region in intron 4 of the eNOS gene were significantly lower than those without the a allele. Although a decrease in serum NOx according to eNOS gene polymorphisms is suggestive, it is still controversial because serum NOx can vary according to renal function or dietary intake of nitrite or nitrate (11). Measurement of the enzyme activity in vivo or in vitro will solve the controversy. In conclusion, Glu298Asp mutations of the eNOS gene are significantly increased not only in nondiabetic ESRD but also in ESRD associated with diabetic nephropathy. The Glu298Asp mutation may be a risk factor, but additional studies, such as in vitro measurement of eNOS activity using cultured endothelium, are required. Characteristics of patients and controls in this study. β, beta coefficient. CI, confidence interval. One degree of freedom. Characteristics of patients and controls in this study. β, beta coefficient. CI, confidence interval. One degree of freedom. Frequencies of the Glu298Asp mutation. ▦, TG + TT; bars, SD; ∗, P <0.01. This work was supported in part by the Scientific Research Funds of the Ministry of Education, Science and Culture of Japan and a Grant for Research Project, University of Tsukuba. We are grateful to Dr. B. Cohen for valuable assistance in preparing the manuscript.