Association Between Heme Oxygenase-1 Promoter Polymorphisms and the Development of Albuminuria in Type 2 Diabetes

Study Subjects

A total of 536 Korean type 2 diabetic patients were recruited from the Diabetes Center in Severance Hospital. According to the criteria of the American Diabetes Association,¹⁷ diabetes was diagnosed with any 1 of the following criteria: fasting blood glucose level of 126 mg/dL or higher; random blood glucose level of 200 mg/dL or higher and symptoms of diabetes such as increased thirst, urination, or weight loss; blood glucose level of 200 mg/dL or higher after 2 hr of 75 g oral glucose tolerance test; and glycated hemoglobin (HbA1c) level of 6.5% or higher. Patients were excluded if they were using glucocorticoid medication or had type 1 diabetes mellitus, abnormal thyroid stimulating hormone levels, a severe infectious disease or a history of cancer, kidney transplantation, ketoacidosis, impending or undergoing dialysis, or liver disease. This research was carried out according to The Code of Ethics of the World Medical Association (Declaration of Helsinki). All patients received adequate information about the study and provided written informed consent. The protocol of this study was approved by the ethics committee of the Yonsei University College of Medicine.

Clinical and Biochemical Measurements

A detailed medical history, including medications, and duration of diabetes, was recorded. All anthropometric and laboratory parameters including fasting plasma glucose (FPG), insulin, and lipid profiles were measured after 8 hr of fasting. HbA1c was measured 4 times over a 12- to 18-month period using high-performance liquid chromatography (Variant II; Greencross, Seoul, Korea) and mean HbA1c was used to represent overall glycemic control. The homeostasis model assessment (HOMA) of insulin resistance (IR) and HOMA-β were calculated by using the following formula: HOMA-IR = (insulin [μIU/mL] × FPG [mg/dL])/405, HOMA-β = 360 × insulin (μIU/mL)/(FPG [mg/dL] − 63). Hypertension was defined by a blood pressure ≥ 140/90 mm Hg or the use of antihypertensive medications.

As glomerular filtration rate is usually normal or increased during the early stage of diabetic nephropathy, albuminuria is considered as a marker for development of nephropathy in type 2 diabetes.¹⁸ Therefore, we used the presence of albuminuria as an indicator for developing diabetic nephropathy in this study. A random or 24-hr urine specimen was collected from all patients and the concentrations of albumin and creatinine were analyzed. The urinary albumin and creatinine concentration were determined by a nephelometric assay with sensitivity to 2.3 mg/L (Beckman Image, Fullerton, CA) and a colorimetric assay (Roche Diagnostic, Indianapolis, IN), respectively. The albumin/creatinine ratio (ACR) in randomly collected urine samples was measured more than twice during the follow-up. Significant albuminuria was defined as follows: urinary ACR ≥ 30 mg/g or albumin excretion ≥ 30 mg/day in a 24 hr urine analysis.

DNA Extraction and Analysis of Polymorphisms in the HO-1 Gene Promoter

Genomic DNA was isolated from peripheral blood lymphocytes. The genotyping of the HO-1 promoter polymorphism T(-413)A (rs2071746) was performed using the TaqMan fluorogenic 5′ nuclease assay (ABI, Foster City, CA) as previously described.¹⁹ Forty-eight duplicate samples and negative controls were included to ensure the accuracy of genotyping, and 100% of the duplicates replicated the original genotype. The rate of successful genotyping was 96.5% for rs2071746 and 98.2% for rs3761439.

The length of the (GT)_n microsatellite polymorphism in the HO-1 gene promoter was determined as previously described with slight modification.²⁰ The (GT)_n repeat segment was amplified in a volume of 20 μL containing 50 ng of genomic DNA, 0.5 pmol of each primer, 1× polymerized chain reaction (PCR) buffer, 200 μM dNTP, 1.5 mM MgCl₂, 1× Band Doctor, and 0.5 U Taq polymerase (Solgent, Daejeon, Korea). The reaction consisted of denaturation at 95°C for 2 min, followed by 32 cycles of 95°C for 20 sec, 53°C for 40 sec, and 72°C for 1 min, with a final extension at 72°C for 5 min using PTC-200 thermal cycler (Bio-Rad, Hercules, CA). The PCR products were analyzed by capillary on a MegaBACE500 (GE Healthcare, Piscataway, NJ) together with an allelic ladder after denaturation at 96°C for 5 min.

Expression Study of HO-1

To explore the regulatory effects of the T(-413)A polymorphism in the promoter region, we constructed HO-1 promoter/luciferase fusion genes. The promoter region between −1876 and +75 was amplified by PCR with a sense primer, 5′-TCCACCTCCACCTTCCCTTAAAGTCGGC-3′ and an antisense primer, 5′-ACGCTCGAGAGGAGGCAGGCGTTGACT-3′, and sub-cloned into pGL3-Basic or pGL3-Enhancer DNA (Promega, Madison, WI), which does not contain any promoter sequence. Site-specific mutation was made using Pfu-X DNA polymerase (SPX16-R250, Solgent). Transfection was performed in human embryonic kidney cells (HEK293, ATCC CRL-1573) and murine mesangial cells (SV40 MES-13, ATCC CRL-1927) using lipofectamine 2000 (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. Assays were performed 48 hr after transfection using the Dual-Luciferase Reporter Assay System (Promega). Luciferase activity was read with a microplate luminometer (Centro XS3 LB960, Berthold Technologies, Bad Wildbad, Germany). Transfections were carried out in triplicate in independent experiments.

Statistical Analyses

The genotype frequencies were tested for Hardy–Weinberg equilibrium using the χ² test. Continuous variables were expressed as the mean ± standard deviation; variables not normally distributed were expressed as median and interquartile ranges. The Student t test or χ² test were used to compare the continuous or categorical parameters of groups distinguished by different genotypes or the presence of albuminuria. Mann–Whitney U test was used to compare the variables not having a normal distribution. Multivariable logistic regression tests were performed to control confounding variables, and odds ratio (OR) and 95% confidence intervals (CIs) were calculated. Genetic analysis using haplotypes was performed to evaluate the combined effects of both the T(-413)A SNP and the (GT)_n repeat polymorphism in the HO-1 promoter on the prevalence of albuminuria. Based on previous studies,^13,21 the distribution of lengths of (GT)_n repeats was categorized into 2 subclasses: the short (S) allele consisted of ≤25 repeats, and the long (L) allele consisted of >25 repeats. Association between haplotypes and the prevalence of albuminuria in type 2 diabetes patients was analyzed using Haploview version 4.2 (Massachusetts Institute of Technology, Cambridge, MA). A P value <0.05 was considered statistically significant. Statistical analyses except analysis of haplotype were performed using SPSS for Windows software (version 18.0; SPSS, Chicago, IL).

Clinical Characteristics of the Subjects and Association Between Genotype and Albuminuria

As shown in Table 1, the mean age, body mass index, and duration of diabetes were significantly higher in patients with albuminuria than in those without albuminuria. Patients with albuminuria also had higher levels of HbA1c, HOMA-IR, and triglycerides and an increased quantity of urinary albumin excretion. Compared with diabetic patients without albuminuria, patients with albuminuria had a higher prevalence of hypertension, carotid plaques, retinopathy, and neuropathy.

Genotyping of the T(-413)A SNP in the HO-1 gene promoter was replicated for 297 patients without albuminuria and 239 patients with albuminuria. The distribution of the T(-413)A SNP was in agreement with Hardy–Weinberg equilibrium and the T was the major allele (53.4%). As shown in Table 2, the T(-413)A SNP was associated with albuminuria in additive and dominant models. Analysis of the (GT)_n microsatellite polymorphism was completed in only 427 patients due to lack of samples. A total of 25 different alleles for the HO-1 (GT)_n microsatellite were identified in patients with type 2 diabetes. The distribution of alleles for the (GT)_n microsatellite in HO-1 promoter ranged from (GT)₁₅ to (GT)₄₀. A bimodal pattern of (GT)_n alleles was observed with a peak at 23 (18.6%) and 30 (31.6%) repeats, which is consistent with other reports.^13,22 Based on their distribution and previous studies,^13,21 we divided the allelic repeats into 2 subclasses: short (S) allele with 25 or less (GT)_n repeats and long (L) allele with more than 25 (GT)_n repeats. The distribution of the (GT)_n polymorphism was in agreement with Hardy–Weinberg equilibrium and the L was the major allele (59.8%). The (GT)_n polymorphism was not associated with albuminuria development in additive, dominant, or recessive models (Table 2). The haplotype analysis of HO-1 promoter polymorphisms did not show any difference (data not shown).

Clinical Characteristics of the Subjects According to the T(-413)A SNP

Based on the analysis of genetic models (Table 2), we compared the general characteristics in the patients according to the genotypes of the T(-413)A in the HO-1 promoter. Albuminuria was significantly more prevalent in patients with the TT genotype than in those with either the AA or AT genotype (52.5% vs 41.2%, P < 0.05). Fasting and postprandial glucose, HbA1c levels, and urinary albumin excretion were higher in patients with the TT genotype than in those with either the AA or AT genotype (each P < 0.05). No other clinical or biochemical parameters significantly differed between AA + AT and TT groups (Table 3).

Analysis of the Contribution of HO-1 Promoter Polymorphisms to Albuminuria Development

To evaluate the association of albuminuria with HO-1 genotypes or other conventional risk factors, multivariate logistic regression analysis was performed (Table 4). After adjusting for established variables, patients with the TT genotype were significantly more susceptible to albuminuria development compared with those carrying the A allele with an OR of 1.616 (95% CI, 1.007–2.593; P = 0.047). Other known risk factors, such as obesity, hypertension, and status of glycemic control, were also significantly associated with albuminuria, as expected. However, no significant relationship between carriers of the LL genotype and albuminuria was observed (data not shown). We then stratified the study population according to baseline clinical and biochemical characteristics and performed subgroup analyses to further investigate the association of the T(-413)A SNP with the risk of albuminuria in each subgroup (Fig. 1). After categorizing the patients based on the duration of diabetes, the OR of having albuminuria for patients with diabetes ≥20 years and with the TT genotype rose from 1.403 to 3.325 and reached statistical significance (95% CI, 1.057–10.400; P = 0.040). Furthermore, patients with poor glycemic control and male sex had higher ORs than those without these risk factors, and the ORs for patients with the TT genotype and these factors reached statistical significance. Although there was no significant association between (GT)_n repeats and albuminuria, patients with hypertension carrying the LL genotype had a higher OR of 1.825 (95% CI, 1.095–3.042; P = 0.021) than those carrying an SS or SL genotype.

Functional Significance of HO-1 Promoter Polymorphisms

To evaluate functional significance of HO-1 promoter polymorphisms, we compared levels of HO-1 expression according to the HO-1 promoter polymorphisms using an in vitro luciferase reporter assay. As shown in Figure 2, the in vitro promoter activity of the A allele of the T(-413)A SNP was significantly higher than that of the corresponding T allele (P < 0.01). Regardless of the T(-413)A SNP allele, the promoter activity of the (GT)₂₃ allele was significantly higher than that of the (GT)₃₀ allele (P < 0.05).