- Email: [email protected]
Association between the rs4880 superoxide dismutase 2 (C>T) gene variant and coronary heart disease in diabetes mellitus
diabetes research and clinical practice 90 (2010) 196–201
Contents lists available at ScienceDirect
Diabetes Research and Clinical Practice jou rna l hom ep ag e: w ww.e lse v ier .com/ loca te /d iab res
Association between the rs4880 superoxide dismutase 2 (C>T) gene variant and coronary heart disease in diabetes mellitus D.A. Jones a,*, S.L. Prior a, T.S. Tang a, S.C. Bain a, S.J. Hurel b, S.E. Humphries c, J.W. Stephens a a
Diabetes Research Group, Institute of Life Sciences, Swansea University, Swansea SA2 8PP, UK Department of Diabetes & Endocrinology, UCL Hospitals, London W1T 3AA, UK c Centre for Cardiovascular Genetics, British Heart Foundation Laboratories, Royal Free & University College London Medical School, London WC1E 6JF, UK b
article info
abstract
Article history:
Mitochondrial superoxide dismutase 2 (SOD2) is an endogenous anti-oxidant enzyme. The
Received 21 April 2010
rs4880 gene variant results in a C>T substitution, influencing SOD enzymatic activity. This
Received in revised form
variant has been associated with micro- and macro-vascular complications in diabetes
19 July 2010
mellitus. Our aim was to examine the association between this variant and coronary heart
Accepted 22 July 2010
disease (CHD) risk in a cross-sectional sample of subjects with diabetes.
Published on line 21 August 2010
776 Caucasian subjects with diabetes were genotyped. CHD risk, oxidised-LDL and plasma total anti-oxidant status (TAOS) were analysed in relation to genotype. In females,
Keywords:
the TT genotype was associated with CHD (CC/CT/TT: No CHD vs. CHD: 22.4/56.0/21.6% vs.
Superoxide dismutase
12.0/50.0/38.0%, p = 0.03; for CC/CT vs. TT, p = 0.01). The odds ratio for CHD associated with
Gene variant
the TT genotype compared to CC/CT was 2.22 [95%CI: 1.17–4.24], p = 0.01. The TT genotype
Oxidative stress
was also associated with significantly lower plasma TAOS. In males, no association was
Diabetes
observed between genotype and CHD risk, but CHD was significantly associated with age,
CHD
lower HDL, higher triglycerides, higher BMI and cigarette smoking. The TT genotype of this variant is associated with increased CHD risk and lower plasma
rs4880 Mitochondria
anti-oxidant defences in females with diabetes. This modest genotype-effect is not apparent
Total anti-oxidant status
in males where traditional risk factors may play a greater role. # 2010 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Oxidative stress is associated with numerous adverse effects on the vascular system [1] and also increased cardiovascular events [2]. Mitochondrial superoxide dismutase 2 (SOD2) is an endogenous anti-oxidant enzyme that plays an important role in limiting oxidative burden [3,4]. The enzyme localises from
the cytoplasm to the mitochondria [4] where it acts as a free radical scavenger enzyme [5] catalysing the dismutation of superoxide to hydrogen peroxide, which is subsequently detoxified further by cytoplasmic catalase to form water. SOD2 is encoded on chromosome 6q25 [6], where any variation within the highly conserved sequence of this gene alters the structure of the mitochondrial target sequence interfering
* Corresponding author. Tel.: +44 0 1792 295073; fax: +44 0 1792 602147. E-mail address: D.A.Jonesa*[email protected]">D.A.Jonesa*[email protected] (). 0168-8227/$ – see front matter # 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2010.07.009
diabetes research and clinical practice 90 (2010) 196–201
with trafficking pathways. Genetic variation within this region has been associated with type 2 diabetes mellitus (T2DM) and cardiomyopathy [6]. The rs4880 gene variant results in a C>T substitution (GCT/GTT), within exon 2 of SOD2 resulting in an amino acid change from alanine to valine (Ala16Val) [6], ultimately causing an alteration in the secondary structure and function of the enzyme [5]. SOD2 enzyme activity is highest in CC (Ala16Ala) subjects [7], whilst subjects with TT genotype (Val16Val) have lower enzyme activity, which is thought to be the result of less efficient targeting of the enzyme to the mitochondria from the cytoplasm [7–9]. Previously, the CC genotype has been associated with better diabetes glycaemic control [5], and the C allele has been associated with lower risks of cardiovascular disease [7], diabetic peripheral neuropathy [10,11], nephropathy [12] and macular oedema [13]. Our aim was to examine the association between the rs4880 gene variant and coronary heart disease (CHD) risk in a cross-sectional sample of subjects with diabetes and plasma measurements of oxidative stress. We chose from the outset not to stratify the sample by type of diabetes.
197
cardiac thallium scan, exercise tolerance test, myocardial infarction or symptomatic/treated angina. Any individual who was asymptomatic or had negative investigations was categorised as ‘no CHD’. Smoking status was defined as current, never or ex-smokers. The latter comprised of those who had stopped smoking for greater than twelve months. Current smokers included subjects who had stopped smoking within a 12-month period. Therefore this is a case–control study. None of the subjects were knowingly taking any form of vitamin supplementation. Plasma samples were collected within a 12-month period and stored immediately at 80 8C. Samples were collected during routine diabetes clinic visit. None of the samples were fasting at the time of collection.
2.2.
Genotyping for the rs4880 gene variant
2.
Methods
Genomic DNA was extracted from 5 mL EDTA blood samples. Genotyping was conducted using polymerase chain reaction (PCR) amplification and restriction fragment length polymorphism (RFLP) analysis. Primer sequences were: forward GCTGTGCTTTCTCGTCTTCAG and reverse TGGTACTTCTCCT CGGTGACG; followed by digest with BsaWI. For all variants, genotype was confirmed by two independent technicians and any discrepancies were resolved by repeat genotyping.
2.1.
Subjects
2.3.
Ethical approval was granted by the UCL and UCL Hospitals ethics committee and all subjects gave written informed consent before recruitment. Patients were recruited from the University College London Diabetes and Cardiovascular Study (UDACS), described elsewhere [14,15]. Briefly, this comprises of 1011 consecutive subjects recruited from the diabetes clinic at University College London Hospitals (UCLH) NHS Trust between the years 2001 and 2002. All patients had diabetes according to WHO criteria [16]. Of these, genotype data was available on 1004 (99.3%). We chose to focus on Caucasian subjects only because of differences in genotype distribution between different ethnic backgrounds (Table 1). This has previously been described (http://www.ncbi.nlm.nih.gov/SNP/ snp_ref.cgi?rs=4880). There were a total of 780 subjects of Caucasian origin. Of these, genotype data was available for 776 (99.5% of 780 subjects). Of the 776 subjects with a genotype, CHD status was available on 768 subjects. Of these, plasma TAOS was measured successfully in 739 and Ox-LDL in 494 subjects. This was due to limited availability of plasma. The presence of CHD was recorded if any patient had positive coronary angiography/angioplasty, coronary artery bypass,
Table 1 – Genotype distribution by ethnic origin. Ethnic group
CC
Afro-Caribbean Caucasian Indian Oriental
16.9% (13) 26.0% (202) 28.8% (32) 0% (0)
CT 40.3% 49.2% 52.3% 22.2%
(31) (382) (58) (4)
Measurement of plasma total anti-oxidant status
Plasma total anti-oxidant status (TAOS), which is inversely related to oxidative stress, was measured by Sampson’s modification of Laight’s photometric microassay [17], using 2.5 mL citrated plasma samples in 96-well ELISA plates. The TAOS of plasma was determined by its capacity to inhibit the peroxidase-mediated formation of the 2,2-azino-bis-3-ethylbensthiazoline-6-sulfonic acid (ABTS+) radical. There are two arms to the assay, a control arm and test arm. In the control arm phosphate buffered saline is used instead of plasma. The difference in absorbance (control absorbance test absorbance) divided by the control absorbance (expressed as a percentage) was used to represent the percentage inhibition of the reaction. The inter- and intra-assay coefficients of variation were 10.1% and 4.3%, respectively. Previously, we have shown that baseline plasma TAOS is associated with prospective risk [2] and has a good correlation with plasma F2isoprostanes (r = 0.65; p = 0.003) [2,18].
2.4.
Measurement of plasma Ox-LDL
Plasma oxidised-LDL (Ox-LDL) was measured using a commercially available Enzyme-Linked Immunosorbent Assay (ELISA) kit supplied by Mercodia (Uppsala, Sweden). In this assay a monoclonal antibody is directed against antigenic determinants in the Ox-LDL molecule (mAB-4E6).
TT 42.9% 24.7% 18.9% 77.8%
(33) (192) (21) (14)
Ethnic background was available on 982 of the 1011 recruited subjects. x2 = 40.5, p < 0.001.
2.5.
Statistical analysis
Statistical analysis was performed using SPSS (version 10.1, SPSS Inc., Chicago). Data are reported for those individuals with available genotype. Results are presented as mean standard deviation. Deviations from Hardy–Weinberg equilibrium were considered using Chi-squared tests. Allele
198
diabetes research and clinical practice 90 (2010) 196–201
frequencies are shown with the 95% confidence interval. For data that was normally distributed after log transformation, the geometric mean and approximate standard deviation is shown. This included systolic blood pressure, diastolic blood pressure, body mass index (BMI), HBA1c, random glucose and LDL-cholesterol data. For duration of diabetes, HDL-cholesterol, TAOS and Ox-LDL, the data could not be transformed to a normal distribution and so the data is shown as median and interquartile range. Analysis of variance (ANOVA) was used to assess the association between genotype and baseline characteristics for data that was normally distributed after log transformation. For duration of diabetes, HDL-cholesterol, TAOS and Ox-LDL, data was analysed by the Kruskal–Wallis test and Mann–Whitney tests. Chi-squared tests were used to compare differences in categorical variables by genotype. In all cases a p-value of less than 0.05 was considered statistically significant. Two sided statistical testing was performed. The relationships between baseline parameters and TAOS were tested by Pearson rank correlation coefficient. As discussed in the results section, this revealed that TAOS was correlated positively with HDL-cholesterol and negatively with triglyceride and glucose. Ox-LDL was correlated with LDL and HDL. Analysis of covariance (ANCOVA) was performed to test the association between genotype and TAOS or Ox-LDL after adjustment for these potential confounders using multiple regression analysis to obtain a residual. Within this analysis stepwise regression was initially performed to look for the final factors influencing TAOS (glucose, triglyceride) or Ox-LDL (LDL, HDL). Correction for multiple comparisons was not applied to the results because the study design was predominantly ‘hypothesis testing’. Whilst making such an adjustment reduces the type I error, it leads to increases in the type II
error, and fewer errors of interpretation occur when no adjustment is made [19].
3.
Results
Of the 780 Caucasian subjects, genotype data was available for 776 (99.5% of 780 Caucasian subjects). The genotype distribution for the rs4880 variant (CC/CT/TT) was in Hardy–Weinberg equilibrium (CC/CT/TT: 202/382/192, x2 = 0.18, p = 0.37) with a C allele frequency of 0.49 [0.47–0.52]. As shown in Table 2, the mean age of the CC subjects was approximately 3 years less than the CT/TT subjects. Furthermore, there were a significantly higher proportion of males with the CC genotype. Therefore, further analysis was performed after stratifying by gender. There were also a higher percentage of CC subjects with type 1 diabetes mellitus (T1DM) (CC vs. CT/TT: 27.7% vs. 20.5%, p = 0.04). This would therefore explain the higher insulin use observed with the CC genotype (Table 2).
3.1.
Association with coronary heart disease
In the group as a whole there was no association between genotype and CHD risk (CC/CT/TT, No CHD vs. CHD: 82.7/80.5/ 74.9% vs. 17.3/19.5/25.1%, p = 0.13; for CC/CT vs. TT, p = 0.06). Because of the genotype difference by gender, the sample was stratified by gender. Within the females there was a significant association between the TT genotype and CHD risk (No CHD vs. CHD: 22.4/56.0/21.6% vs. 12.0/50.0/38.0%, p = 0.03; for CC/CT vs. TT, p = 0.01). In females the odds ratio (OR) for CHD associated with the TT genotype compared to CC/CT was 2.22 [95%CI: 1.17–4.24], p = 0.01. As shown in Table 3, no association
Table 2 – Baseline differences in subjects by genotype. Trait
Age (years) Duration (years)a Systolic blood pressure (mmHg)b Diastolic blood pressure (mmHg)b Body mass index (kg/m2)b HbA1c (%)b Random plasma glucose (mmol/L)b LDL-cholesterol (mmol/L)b Triglycerides (mmol/L)b HDL-cholesterol (mmol/L)a TAOS (%)a Ox-LDL (U/L)a Males % (n) Type 1 diabetes % (n) Current smokers % (n) ACE inhibitor % (n) Aspirin % (n) Insulin % (n) Statin % (n) a
Genotype
p
CC (202)
CT (382)
TT (192)
60.1 (14.6) 12 [6–20] 137 (19) 79 (11) 28.8 (6.0) 7.9 (1.6) 9.7 (5.0) 2.7 (0.4) 1.6 (1.0) 1.3 [1.0–1.6] 44.9 [36.9–52.1] 46.7 [38.2–60.1] 68.8% (139) 28% (56) 19.9% (40) 47.5% (96) 41.0% (83) 51.5% (104) 27.6% (56)
63.2 (12.8) 11 [5–19] 139 (20) 80 (11) 28.4 (5.2) 7.8 (1.6) 9.6 (4.6) 2.6 (0.4) 1.7 (1.0) 1.4 [1.1–1.7] 45.0 [35.6–51.2] 47.2 [35.5–56.9] 56.1% (214) 20% (76) 16.3% (62) 42.6% (163) 44.9% (172) 40.9% (156) 28.0% (107)
63.6 (13.9) 9 [5–20] 138 (20) 78 (12) 28.3 (5.6) 7.6 (1.6) 9.9 (5.0) 2.6 (0.4) 1.8 (1.1) 1.3 [1.1–1.7] 42.1 [35.0–50.8] 48.0 [36.1–57.9] 61.5% (118) 22% (42) 14.7% (28) 47.1% (90) 50.0% (96) 41.4% (80) 23.0% (44)
0.02 0.32 0.43 0.40 0.62 0.39 0.83 0.92 0.36 0.20 0.40 0.87 0.01 0.10 0.37 0.43 0.20 0.04 0.42
Median and interquartile range shown for duration of diabetes, HDL, TAOS and Ox-LDL. Analysis performed by ANOVA after appropriate transformation of non-normally distributed data and by Kruskal–Wallis for duration of diabetes and Ox-LDL. x2-test was used to compare groups. Significant associations are highlighted in bold. b Log transformed data. Mean and standard deviation shown or geometric mean and approximate standard deviation for log transformed data.
199
diabetes research and clinical practice 90 (2010) 196–201
Table 3 – Baseline differences by CHD and gender. Variable
Age (years) Duration of diabetes (years)a Systolic blood pressure (mmHg)b Diastolic blood pressure (mmHg)b Body mass index (kg/m2)b LDL-cholesterol (mmol/L)b HDL-cholesterol (mmol/L)a Triglyceride (mmol/L)b TAOS (%)a Ox-LDL (U/L)a Current smoker (%) (n) ACE inhibitor (%) (n) Aspirin (%) (n) Insulin (%) (n) Statin (%) (n) Genotype (%) (n) CC/CT/TT
Females
Males
No CHD (251)
CHD (50)
p
No CHD (361)
CHD (106)
63.4 (13.9) 11 [5–11] 141 (21) 81 (11) 28.5 (6.2) 2.8 (2.6) 1.5 [1.2–1.9] 1.7 (0.9) 43.9 [34.6–50.4] 48.3 [38.5–57.8] 13.7% (34) 40.6% (102) 34.7% (87) 44.6% (22) 23.1% (58) 22.4/56.0/21.6% (56/141/54)
70.5 (8.4) 10 [6–19] 142 (18) 77 (10) 29.6 (4.6) 2.5 (2.3) 1.4 [1.2–1.8] 1.9 (0.9) 46.6 [35.4–53.8] 44.5 [39.4–57.3] 14.0% (7) 41.2% (21) 70.6% (35) 33.3% (17) 58.0% (29) 12.0/50.0/38.0% (30/25/19)
0.001 0.99 0.58 0.49 0.20 0.005 0.22 0.14 0.35 0.27 0.95 0.94 <0.001 0.14 <0.001 0.03
59.4 (13.8) 11 [4.3–20.8] 137 (19) 81 (11) 28.1 (5.0) 2.7 (0.4) 1.3 [1.1–1.6] 1.6 (1.0) 48.4 [37.3–52.7] 45.4 [35.2–57.8] 20.9% (76) 43.6% (157) 39.8% (354) 46.5% (168) 14.7% (53) 29.8/45.6/24.6% (108/165/88)
67.7 (10.9) 12 [6–18] 137 (23) 77 (12) 29.4 (5.7) 2.2 (0.8) 1.2 [0.9–1.4] 1.9 (1.1) 45.1 [33.0–49.1] 48.4 [32.9–58.1] 14.0% (15) 63.2% (67) 76.9% (82) 36.8% (39) 62.3% (66) 27.4/45.3/27.4% (29/48/2)
p <0.001 0.71 0.82 <0.001 0.03 <0.001 <0.001 0.02 0.03 0.94 0.03 <0.001 <0.001 0.08 <0.001 0.81
a Median and interquartile range shown for duration of diabetes, HDL, TAOS and Ox-LDL. Analysis performed by ANOVA after appropriate transformation of non-normally distributed data and by Kruskal–Wallis for HDL, TAOS and Ox-LDL. x2-test was used to compare groups. Significant associations are highlighted in bold. b Log transformed data. Mean and standard deviation shown or geometric mean and approximate standard deviation for log transformed data.
was observed between CHD and genotype for the males (No CHD vs. CHD: 29.8/45.6/24.6% vs. 27.4/45.3/27.4%, p = 0.81).
3.2. Association of baseline risk factors with CHD status after stratifying for gender As shown in Table 3, the risk factors in females associated with CHD were age and genotype. In males, risk factors included age, lower HDL-cholesterol, higher triglycerides, higher BMI and cigarette smoking. However, in males there was no association with CHD risk and genotype. A possible explanation for this gender difference might be that the combination of these traditional risk factors makes a greater contribution to overall CHD risk, masking the modest genotype-effect. Of interest, significantly more males were taking ACE inhibitors, which lower blood pressure, have cardio-protective effects and also possible anti-oxidant and anti-inflammatory properties. In both males and females, a significant proportion of those with CHD were taking aspirin and statin therapies. This may account for the paradoxical lower LDL levels seen in those subjects.
3.3.
Plasma total anti-oxidant status by genotype
Since SOD2 is known to modulate oxidative stress, we examined the association between plasma TAOS and genotype. As described previously [2,14], plasma TAOS correlated positively with plasma HDL-cholesterol, and negatively with triglyceride and glucose (correlation coefficient r = 0.13, 0.14, and 0.12; all p < 0.05). As observed in Table 2, there was no association between genotype and plasma TAOS for the whole group. After stratifying by gender no differences were observed (for males CC/CT/TT: 43.5 [36.8–50.6]% vs. 45.6 [36.1–55.0]% vs. 44.0 [35.8–50.5]%, p = 0.38; for females 46.7 [36.7–56.0]% vs. 44.8 [35.4–49.7]% vs. 40.8 [32.7–51.1]%, p = 0.13).
No difference was observed after adjustment for the correlates of plasma TAOS. As shown in Table 3, with respect to CHD status, plasma TAOS was lower in the males with CHD. No difference was observed in the females (no difference was observed after adjustment for the correlated of TAOS). Fig. 1 shows TAOS stratified by gender, median TAOS and genotype. As may be observed there was a significant difference in genotype distribution for the females, with the TT genotype being associated with a lower median plasma TAOS. The OR for a plasma TAOS below the median for the TT vs. CC/CT genotype was 2.07 [1.20–3.60], p = 0.009. No association was observed in the males. For plasma Ox-LDL, as previously described within this sample [2], plasma Ox-LDL was independent of pharmacotherapy but was correlated with LDL and HDL (correlation coefficient r = 0.32, 0.26, respectively, p < 0.05). As observed in Table 2, there was no association between genotype and plasma Ox-LDL for the whole group. After stratifying by gender no differences were observed (CC vs. CT vs. TT: males 47.6 [38.2–60.7] vs. 45.1 [33.7–56.5] vs. 47.0 [36.1–57.3], p = 0.22; females 45.9 [36.9–53.3] vs. 48.5 [41.6–57.8] vs. 47.1 [36.6–58.5], p = 0.49). No difference was observed after adjustment for the correlates of plasma Ox-LDL. As shown in Table 3, with respect to CHD status, in both genders, Ox-LDL did not differ by CHD status. No difference was observed in the females (no change was observed after adjustment for the correlated of TAOS).
4.
Discussion
In line with previous studies, we describe a deleterious effect associated with the TT genotype of the rs4880 SOD2 variant. The TT genotype was associated with a greater than twofold increase in CHD risk compared to the C allele in females with diabetes. We observed that the TT genotype was also
200
diabetes research and clinical practice 90 (2010) 196–201
[(Fig._1)TD$IG]
Fig. 1 – rs4880 Genotype distribution by median plasma TAOS. (A) Females: the median plasma TAOS in the group of females was 44%. In this sample, 30.2% of those subjects with the TT genotype were in the lower TAOS group compared to 17.7% in the higher TAOS group. This was statistically significant (x2 = 7.00, p = 0.03, for CC/CT vs. TT, p = 0.01). Plasma TAOS is inversely proportional to oxidative stress (the lower the plasma TAOS, the greater the oxidative burden). (B) Males: the median plasma TAOS in the group of males was 47.5%. No significant difference was seen in the genotype distribution when the group was divided by the median plasma TAOS (x2 = 0.82, p = 0.66).
associated with a lower median plasma TAOS and hence increased plasma oxidative stress. Of interest within the female group the only other statistically significant risk factor was age. Conversely in the males, no association between CHD and genotype was observed. However, risk in males was associated with age, lower HDL-cholesterol, higher triglycerides, higher BMI and cigarette smoking. One possible explanation for the gender difference might therefore be that these risk factors make a greater contribution to overall CHD risk and might mask the modest genotype-effect. Of importance, the males were also prescribed more secondary preventative drugs which may also contribute to this effect. Another speculative explanation for the lack of genotype risk in males may be related to the genotype difference observed by gender. It may be that the higher proportion of males with the CC genotype results from survival bias as a result of increased mortality in those with the TT genotype. Further examination is difficult in this case–control study. There are limitations to our study. We have chosen from the outset to examine one gene variant in SOD2 in relation to intermediate biochemical markers of oxidative stress. We have focused on plasma TAOS and Ox-LDL as intermediate biochemical phenotypes. There are limitations to these
measurements as discussed elsewhere [1,2]. Further work should be performed to look at other variants within the SOD2 regulatory region to allow haplotype based analysis. This would also allow linkage disequilibrium across the region to be examined. At first sight, the lack of any significant difference in genotype distribution between those with and without CHD in males might appear to conflict with that observed in females. However, there are several possible explanations for this as discussed. In addition, prospective gene-association studies are more powerful than case–control studies. Case– control cross-sectional studies are prone to intrinsic bias, for example due to possible altered rates of disease progression, subsequent progression of secondary phenotypes, or genotype associations with death or treatment changes. Furthermore, diabetes with its increased obesity, oxidative burden, inflammation and hyperglycaemia might all overwhelm the SOD2 genotype ‘strength of signal’ and hence the genotype-effect on CHD risk might be lower than in a non-diabetic sample. Therefore, this SNP should be examined in a non-diabetic sample. Such influences are well-recognised confounders [20,21]. Oxidative stress is an important in the underlying pathophysiology of CHD [1,2]. It is likely to play a major role
diabetes research and clinical practice 90 (2010) 196–201
in patients with diabetes, where no more than 25% of the excess CHD risk may be accounted for by traditional risk factors [22,23]. The C>T substitution of the rs4880 gene variant alters the structure of the mitochondrial target sequence, interfering with trafficking pathways resulting in a deleterious effect on overall enzyme activity [5,6]. We show that this change is associated with a modest CHD risk and also that the gene variant is associated with a deleterious change in a global measure of oxidative stress. Further prospective analysis of the gene variant is required in diabetic and no diabetic cohorts with measured plasma markers of oxidative stress however access to such cohorts can be difficult [24].
Acknowledgements Diabetes UK supported JWS (BDA: RD01/0001357) and the creation of UDACS. SEH is supported by the British Heart Foundation (RG2005 014).
Conflict of interest The authors declare that they have no conflict of interest.
references
[1] Stephens JW, Khanolkar MP, Bain SC. The biological relevance and measurement of plasma markers of oxidative stress in diabetes and cardiovascular disease. Atherosclerosis 2009;202:321–9. [2] Stephens JW, Gable DR, Hurel SJ, Miller GJ, Cooper JA, Humphries SE. Increased plasma markers of oxidative stress are associated with coronary heart disease in males with diabetes mellitus and with 10-year risk in a prospective sample of males. Clin Chem 2006;52:446–52. [3] Kiritoshi S, Nishikawa T, Sonoda K, Kukidome D, Senokuchi T, Matsuo T, et al. Reactive oxygen species from mitochondria induce cyclooxygenase-2 gene expression in human mesangial cells: potential role in diabetic nephropathy. Diabetes 2003;52:2570–7. [4] Kinnula VL, Lehtonen S, Koistinen P, Kakko S, Savolainen M, Kere J, et al. Two functional variants of the superoxide dismutase genes in Finnish families with asthma. Thorax 2004;59:116–9. [5] Nomiyama T, Tanaka Y, Piao L, Nagasaka K, Sakai K, Ogihara T, et al. The polymorphism of manganese superoxide dismutase is associated with diabetic nephropathy in Japanese type 2 diabetic patients. J Hum Genet 2003;48:138–41. [6] Hiroi S, Harada H, Nishi H, Satoh M, Nagai R, Kimura A. Polymorphisms in the SOD2 and HLA-DRB1 genes are associated with nonfamilial idiopathic dilated cardiomyopathy in Japanese. Biochem Biophys Res Commun 1999;261:332–9. [7] Flekac M, Skrha J, Hilgertova J, Lacinova Z, Jarolimkova M. Gene polymorphisms of superoxide dismutases and catalase in diabetes mellitus. BMC Med Genet 2008;9:30.
201
[8] Kang D, Lee KM, Park SK, Berndt SI, Peters U, Reding D, et al. Functional variant of manganese superoxide dismutase (SOD2 V16A) polymorphism is associated with prostate cancer risk in the prostate, lung, colorectal, and ovarian cancer study. Cancer Epidemiol Biomarkers Prev 2007;16:1581–6. [9] Zejnilovic J, Akev N, Yilmaz H, Isbir T. Association between manganese superoxide dismutase polymorphism and risk of lung cancer. Cancer Genet Cytogenet 2009;189:1–4. [10] Chistyakov DA, Savost’anov KV, Zotova EV, Nosikov VV. Polymorphisms in the Mn-SOD and EC-SOD genes and their relationship to diabetic neuropathy in type 1 diabetes mellitus. BMC Med Genet 2001;2:4. [11] El Masry TM, Abou Zahra MAM, El Tawil MM, Khalifa RA. Managanese superoxide dismutase Alanine to Valine polymorphism and risk of neuropathy and nephropathy in Egyptian type 1 diabetic patients. Rev Diabetic Stud 2005;2:70–4. [12] Lee SJ, Choi MG, Kim DS, Kim TW. Manganese superoxide dismutase gene polymorphism (V16A) is associated with stages of albuminuria in Korean type 2 diabetic patients. Metabolism 2006;55:1–7. [13] Lee SJ, Choi MG. Association of manganese superoxide dismutase gene polymorphism (V16A) with diabetic macular edema in Korean type 2 diabetic patients. Metabolism 2006;55:1681–8. [14] Dhamrait SS, Stephens JW, Cooper JA, Acharya J, Mani AR, Moore K, et al. Cardiovascular risk in healthy men and markers of oxidative stress in diabetic men are associated with common variation in the gene for uncoupling protein 2. Eur Heart J 2004;25:468–75. [15] Stephens JW, Hurel SJ, Acharya J, Humphries SE. An interaction between the interleukin-6-174G>C gene variant and urinary protein excretion influences plasma oxidative stress in subjects with type 2 diabetes. Cardiovasc Diabetol 2004;3:2. [16] Alberti KG, Zimmet PZ. New diagnostic criteria and classification of diabetes—again? Diabet Med 1998;15:535–6. [17] Sampson MJ, Gopaul N, Davies IR, Hughes DA, Carrier MJ. Plasma F2 isoprostanes: direct evidence of increased free radical damage during acute hyperglycemia in type 2 diabetes. Diabetes Care 2002;25:537–41. [18] Stephens JW, Humphries SE, Miller GJ, Cooper JA, Hurel SJ. Increased oxidative stress is associated with coronary heart disease in males with diabetes mellitus and with 10-year prospective risk in males without diabetes. Diabet Med 2005;22(suppl. 2):P145. [19] Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology 1990;1:43–6. [20] Risch NJ. Searching for genetic determinants in the new millennium. Nature 2000;405:847–56. [21] Sing CF, Stengard JH, Kardia SL. Genes environment, and cardiovascular disease. Arterioscler Thromb Vasc Biol 2003;23:1190–6. [22] Stephens JW, Ambler G, Vallance P, Betteridge DJ, Humphries SE, Hurel SJ. Cardiovascular risk and diabetes. Are the methods of risk prediction satisfactory? Eur J Cardiovasc Prev Rehabil 2004;11:521–8. [23] Pyorala K, Laakso M, Uusitupa M. Diabetes and atherosclerosis: an epidemiologic view. Diabetes Metab Rev 1987;3:463–524. [24] Stephens JW, Bain SC, Humphries SE. Gene–environment interaction and oxidative stress in cardiovascular disease. Atherosclerosis 2008;200:229–38.
Contents lists available at ScienceDirect
Diabetes Research and Clinical Practice jou rna l hom ep ag e: w ww.e lse v ier .com/ loca te /d iab res
Association between the rs4880 superoxide dismutase 2 (C>T) gene variant and coronary heart disease in diabetes mellitus D.A. Jones a,*, S.L. Prior a, T.S. Tang a, S.C. Bain a, S.J. Hurel b, S.E. Humphries c, J.W. Stephens a a
Diabetes Research Group, Institute of Life Sciences, Swansea University, Swansea SA2 8PP, UK Department of Diabetes & Endocrinology, UCL Hospitals, London W1T 3AA, UK c Centre for Cardiovascular Genetics, British Heart Foundation Laboratories, Royal Free & University College London Medical School, London WC1E 6JF, UK b
article info
abstract
Article history:
Mitochondrial superoxide dismutase 2 (SOD2) is an endogenous anti-oxidant enzyme. The
Received 21 April 2010
rs4880 gene variant results in a C>T substitution, influencing SOD enzymatic activity. This
Received in revised form
variant has been associated with micro- and macro-vascular complications in diabetes
19 July 2010
mellitus. Our aim was to examine the association between this variant and coronary heart
Accepted 22 July 2010
disease (CHD) risk in a cross-sectional sample of subjects with diabetes.
Published on line 21 August 2010
776 Caucasian subjects with diabetes were genotyped. CHD risk, oxidised-LDL and plasma total anti-oxidant status (TAOS) were analysed in relation to genotype. In females,
Keywords:
the TT genotype was associated with CHD (CC/CT/TT: No CHD vs. CHD: 22.4/56.0/21.6% vs.
Superoxide dismutase
12.0/50.0/38.0%, p = 0.03; for CC/CT vs. TT, p = 0.01). The odds ratio for CHD associated with
Gene variant
the TT genotype compared to CC/CT was 2.22 [95%CI: 1.17–4.24], p = 0.01. The TT genotype
Oxidative stress
was also associated with significantly lower plasma TAOS. In males, no association was
Diabetes
observed between genotype and CHD risk, but CHD was significantly associated with age,
CHD
lower HDL, higher triglycerides, higher BMI and cigarette smoking. The TT genotype of this variant is associated with increased CHD risk and lower plasma
rs4880 Mitochondria
anti-oxidant defences in females with diabetes. This modest genotype-effect is not apparent
Total anti-oxidant status
in males where traditional risk factors may play a greater role. # 2010 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Oxidative stress is associated with numerous adverse effects on the vascular system [1] and also increased cardiovascular events [2]. Mitochondrial superoxide dismutase 2 (SOD2) is an endogenous anti-oxidant enzyme that plays an important role in limiting oxidative burden [3,4]. The enzyme localises from
the cytoplasm to the mitochondria [4] where it acts as a free radical scavenger enzyme [5] catalysing the dismutation of superoxide to hydrogen peroxide, which is subsequently detoxified further by cytoplasmic catalase to form water. SOD2 is encoded on chromosome 6q25 [6], where any variation within the highly conserved sequence of this gene alters the structure of the mitochondrial target sequence interfering
* Corresponding author. Tel.: +44 0 1792 295073; fax: +44 0 1792 602147. E-mail address: D.A.Jonesa*[email protected]">D.A.Jonesa*[email protected] (). 0168-8227/$ – see front matter # 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2010.07.009
diabetes research and clinical practice 90 (2010) 196–201
with trafficking pathways. Genetic variation within this region has been associated with type 2 diabetes mellitus (T2DM) and cardiomyopathy [6]. The rs4880 gene variant results in a C>T substitution (GCT/GTT), within exon 2 of SOD2 resulting in an amino acid change from alanine to valine (Ala16Val) [6], ultimately causing an alteration in the secondary structure and function of the enzyme [5]. SOD2 enzyme activity is highest in CC (Ala16Ala) subjects [7], whilst subjects with TT genotype (Val16Val) have lower enzyme activity, which is thought to be the result of less efficient targeting of the enzyme to the mitochondria from the cytoplasm [7–9]. Previously, the CC genotype has been associated with better diabetes glycaemic control [5], and the C allele has been associated with lower risks of cardiovascular disease [7], diabetic peripheral neuropathy [10,11], nephropathy [12] and macular oedema [13]. Our aim was to examine the association between the rs4880 gene variant and coronary heart disease (CHD) risk in a cross-sectional sample of subjects with diabetes and plasma measurements of oxidative stress. We chose from the outset not to stratify the sample by type of diabetes.
197
cardiac thallium scan, exercise tolerance test, myocardial infarction or symptomatic/treated angina. Any individual who was asymptomatic or had negative investigations was categorised as ‘no CHD’. Smoking status was defined as current, never or ex-smokers. The latter comprised of those who had stopped smoking for greater than twelve months. Current smokers included subjects who had stopped smoking within a 12-month period. Therefore this is a case–control study. None of the subjects were knowingly taking any form of vitamin supplementation. Plasma samples were collected within a 12-month period and stored immediately at 80 8C. Samples were collected during routine diabetes clinic visit. None of the samples were fasting at the time of collection.
2.2.
Genotyping for the rs4880 gene variant
2.
Methods
Genomic DNA was extracted from 5 mL EDTA blood samples. Genotyping was conducted using polymerase chain reaction (PCR) amplification and restriction fragment length polymorphism (RFLP) analysis. Primer sequences were: forward GCTGTGCTTTCTCGTCTTCAG and reverse TGGTACTTCTCCT CGGTGACG; followed by digest with BsaWI. For all variants, genotype was confirmed by two independent technicians and any discrepancies were resolved by repeat genotyping.
2.1.
Subjects
2.3.
Ethical approval was granted by the UCL and UCL Hospitals ethics committee and all subjects gave written informed consent before recruitment. Patients were recruited from the University College London Diabetes and Cardiovascular Study (UDACS), described elsewhere [14,15]. Briefly, this comprises of 1011 consecutive subjects recruited from the diabetes clinic at University College London Hospitals (UCLH) NHS Trust between the years 2001 and 2002. All patients had diabetes according to WHO criteria [16]. Of these, genotype data was available on 1004 (99.3%). We chose to focus on Caucasian subjects only because of differences in genotype distribution between different ethnic backgrounds (Table 1). This has previously been described (http://www.ncbi.nlm.nih.gov/SNP/ snp_ref.cgi?rs=4880). There were a total of 780 subjects of Caucasian origin. Of these, genotype data was available for 776 (99.5% of 780 subjects). Of the 776 subjects with a genotype, CHD status was available on 768 subjects. Of these, plasma TAOS was measured successfully in 739 and Ox-LDL in 494 subjects. This was due to limited availability of plasma. The presence of CHD was recorded if any patient had positive coronary angiography/angioplasty, coronary artery bypass,
Table 1 – Genotype distribution by ethnic origin. Ethnic group
CC
Afro-Caribbean Caucasian Indian Oriental
16.9% (13) 26.0% (202) 28.8% (32) 0% (0)
CT 40.3% 49.2% 52.3% 22.2%
(31) (382) (58) (4)
Measurement of plasma total anti-oxidant status
Plasma total anti-oxidant status (TAOS), which is inversely related to oxidative stress, was measured by Sampson’s modification of Laight’s photometric microassay [17], using 2.5 mL citrated plasma samples in 96-well ELISA plates. The TAOS of plasma was determined by its capacity to inhibit the peroxidase-mediated formation of the 2,2-azino-bis-3-ethylbensthiazoline-6-sulfonic acid (ABTS+) radical. There are two arms to the assay, a control arm and test arm. In the control arm phosphate buffered saline is used instead of plasma. The difference in absorbance (control absorbance test absorbance) divided by the control absorbance (expressed as a percentage) was used to represent the percentage inhibition of the reaction. The inter- and intra-assay coefficients of variation were 10.1% and 4.3%, respectively. Previously, we have shown that baseline plasma TAOS is associated with prospective risk [2] and has a good correlation with plasma F2isoprostanes (r = 0.65; p = 0.003) [2,18].
2.4.
Measurement of plasma Ox-LDL
Plasma oxidised-LDL (Ox-LDL) was measured using a commercially available Enzyme-Linked Immunosorbent Assay (ELISA) kit supplied by Mercodia (Uppsala, Sweden). In this assay a monoclonal antibody is directed against antigenic determinants in the Ox-LDL molecule (mAB-4E6).
TT 42.9% 24.7% 18.9% 77.8%
(33) (192) (21) (14)
Ethnic background was available on 982 of the 1011 recruited subjects. x2 = 40.5, p < 0.001.
2.5.
Statistical analysis
Statistical analysis was performed using SPSS (version 10.1, SPSS Inc., Chicago). Data are reported for those individuals with available genotype. Results are presented as mean standard deviation. Deviations from Hardy–Weinberg equilibrium were considered using Chi-squared tests. Allele
198
diabetes research and clinical practice 90 (2010) 196–201
frequencies are shown with the 95% confidence interval. For data that was normally distributed after log transformation, the geometric mean and approximate standard deviation is shown. This included systolic blood pressure, diastolic blood pressure, body mass index (BMI), HBA1c, random glucose and LDL-cholesterol data. For duration of diabetes, HDL-cholesterol, TAOS and Ox-LDL, the data could not be transformed to a normal distribution and so the data is shown as median and interquartile range. Analysis of variance (ANOVA) was used to assess the association between genotype and baseline characteristics for data that was normally distributed after log transformation. For duration of diabetes, HDL-cholesterol, TAOS and Ox-LDL, data was analysed by the Kruskal–Wallis test and Mann–Whitney tests. Chi-squared tests were used to compare differences in categorical variables by genotype. In all cases a p-value of less than 0.05 was considered statistically significant. Two sided statistical testing was performed. The relationships between baseline parameters and TAOS were tested by Pearson rank correlation coefficient. As discussed in the results section, this revealed that TAOS was correlated positively with HDL-cholesterol and negatively with triglyceride and glucose. Ox-LDL was correlated with LDL and HDL. Analysis of covariance (ANCOVA) was performed to test the association between genotype and TAOS or Ox-LDL after adjustment for these potential confounders using multiple regression analysis to obtain a residual. Within this analysis stepwise regression was initially performed to look for the final factors influencing TAOS (glucose, triglyceride) or Ox-LDL (LDL, HDL). Correction for multiple comparisons was not applied to the results because the study design was predominantly ‘hypothesis testing’. Whilst making such an adjustment reduces the type I error, it leads to increases in the type II
error, and fewer errors of interpretation occur when no adjustment is made [19].
3.
Results
Of the 780 Caucasian subjects, genotype data was available for 776 (99.5% of 780 Caucasian subjects). The genotype distribution for the rs4880 variant (CC/CT/TT) was in Hardy–Weinberg equilibrium (CC/CT/TT: 202/382/192, x2 = 0.18, p = 0.37) with a C allele frequency of 0.49 [0.47–0.52]. As shown in Table 2, the mean age of the CC subjects was approximately 3 years less than the CT/TT subjects. Furthermore, there were a significantly higher proportion of males with the CC genotype. Therefore, further analysis was performed after stratifying by gender. There were also a higher percentage of CC subjects with type 1 diabetes mellitus (T1DM) (CC vs. CT/TT: 27.7% vs. 20.5%, p = 0.04). This would therefore explain the higher insulin use observed with the CC genotype (Table 2).
3.1.
Association with coronary heart disease
In the group as a whole there was no association between genotype and CHD risk (CC/CT/TT, No CHD vs. CHD: 82.7/80.5/ 74.9% vs. 17.3/19.5/25.1%, p = 0.13; for CC/CT vs. TT, p = 0.06). Because of the genotype difference by gender, the sample was stratified by gender. Within the females there was a significant association between the TT genotype and CHD risk (No CHD vs. CHD: 22.4/56.0/21.6% vs. 12.0/50.0/38.0%, p = 0.03; for CC/CT vs. TT, p = 0.01). In females the odds ratio (OR) for CHD associated with the TT genotype compared to CC/CT was 2.22 [95%CI: 1.17–4.24], p = 0.01. As shown in Table 3, no association
Table 2 – Baseline differences in subjects by genotype. Trait
Age (years) Duration (years)a Systolic blood pressure (mmHg)b Diastolic blood pressure (mmHg)b Body mass index (kg/m2)b HbA1c (%)b Random plasma glucose (mmol/L)b LDL-cholesterol (mmol/L)b Triglycerides (mmol/L)b HDL-cholesterol (mmol/L)a TAOS (%)a Ox-LDL (U/L)a Males % (n) Type 1 diabetes % (n) Current smokers % (n) ACE inhibitor % (n) Aspirin % (n) Insulin % (n) Statin % (n) a
Genotype
p
CC (202)
CT (382)
TT (192)
60.1 (14.6) 12 [6–20] 137 (19) 79 (11) 28.8 (6.0) 7.9 (1.6) 9.7 (5.0) 2.7 (0.4) 1.6 (1.0) 1.3 [1.0–1.6] 44.9 [36.9–52.1] 46.7 [38.2–60.1] 68.8% (139) 28% (56) 19.9% (40) 47.5% (96) 41.0% (83) 51.5% (104) 27.6% (56)
63.2 (12.8) 11 [5–19] 139 (20) 80 (11) 28.4 (5.2) 7.8 (1.6) 9.6 (4.6) 2.6 (0.4) 1.7 (1.0) 1.4 [1.1–1.7] 45.0 [35.6–51.2] 47.2 [35.5–56.9] 56.1% (214) 20% (76) 16.3% (62) 42.6% (163) 44.9% (172) 40.9% (156) 28.0% (107)
63.6 (13.9) 9 [5–20] 138 (20) 78 (12) 28.3 (5.6) 7.6 (1.6) 9.9 (5.0) 2.6 (0.4) 1.8 (1.1) 1.3 [1.1–1.7] 42.1 [35.0–50.8] 48.0 [36.1–57.9] 61.5% (118) 22% (42) 14.7% (28) 47.1% (90) 50.0% (96) 41.4% (80) 23.0% (44)
0.02 0.32 0.43 0.40 0.62 0.39 0.83 0.92 0.36 0.20 0.40 0.87 0.01 0.10 0.37 0.43 0.20 0.04 0.42
Median and interquartile range shown for duration of diabetes, HDL, TAOS and Ox-LDL. Analysis performed by ANOVA after appropriate transformation of non-normally distributed data and by Kruskal–Wallis for duration of diabetes and Ox-LDL. x2-test was used to compare groups. Significant associations are highlighted in bold. b Log transformed data. Mean and standard deviation shown or geometric mean and approximate standard deviation for log transformed data.
199
diabetes research and clinical practice 90 (2010) 196–201
Table 3 – Baseline differences by CHD and gender. Variable
Age (years) Duration of diabetes (years)a Systolic blood pressure (mmHg)b Diastolic blood pressure (mmHg)b Body mass index (kg/m2)b LDL-cholesterol (mmol/L)b HDL-cholesterol (mmol/L)a Triglyceride (mmol/L)b TAOS (%)a Ox-LDL (U/L)a Current smoker (%) (n) ACE inhibitor (%) (n) Aspirin (%) (n) Insulin (%) (n) Statin (%) (n) Genotype (%) (n) CC/CT/TT
Females
Males
No CHD (251)
CHD (50)
p
No CHD (361)
CHD (106)
63.4 (13.9) 11 [5–11] 141 (21) 81 (11) 28.5 (6.2) 2.8 (2.6) 1.5 [1.2–1.9] 1.7 (0.9) 43.9 [34.6–50.4] 48.3 [38.5–57.8] 13.7% (34) 40.6% (102) 34.7% (87) 44.6% (22) 23.1% (58) 22.4/56.0/21.6% (56/141/54)
70.5 (8.4) 10 [6–19] 142 (18) 77 (10) 29.6 (4.6) 2.5 (2.3) 1.4 [1.2–1.8] 1.9 (0.9) 46.6 [35.4–53.8] 44.5 [39.4–57.3] 14.0% (7) 41.2% (21) 70.6% (35) 33.3% (17) 58.0% (29) 12.0/50.0/38.0% (30/25/19)
0.001 0.99 0.58 0.49 0.20 0.005 0.22 0.14 0.35 0.27 0.95 0.94 <0.001 0.14 <0.001 0.03
59.4 (13.8) 11 [4.3–20.8] 137 (19) 81 (11) 28.1 (5.0) 2.7 (0.4) 1.3 [1.1–1.6] 1.6 (1.0) 48.4 [37.3–52.7] 45.4 [35.2–57.8] 20.9% (76) 43.6% (157) 39.8% (354) 46.5% (168) 14.7% (53) 29.8/45.6/24.6% (108/165/88)
67.7 (10.9) 12 [6–18] 137 (23) 77 (12) 29.4 (5.7) 2.2 (0.8) 1.2 [0.9–1.4] 1.9 (1.1) 45.1 [33.0–49.1] 48.4 [32.9–58.1] 14.0% (15) 63.2% (67) 76.9% (82) 36.8% (39) 62.3% (66) 27.4/45.3/27.4% (29/48/2)
p <0.001 0.71 0.82 <0.001 0.03 <0.001 <0.001 0.02 0.03 0.94 0.03 <0.001 <0.001 0.08 <0.001 0.81
a Median and interquartile range shown for duration of diabetes, HDL, TAOS and Ox-LDL. Analysis performed by ANOVA after appropriate transformation of non-normally distributed data and by Kruskal–Wallis for HDL, TAOS and Ox-LDL. x2-test was used to compare groups. Significant associations are highlighted in bold. b Log transformed data. Mean and standard deviation shown or geometric mean and approximate standard deviation for log transformed data.
was observed between CHD and genotype for the males (No CHD vs. CHD: 29.8/45.6/24.6% vs. 27.4/45.3/27.4%, p = 0.81).
3.2. Association of baseline risk factors with CHD status after stratifying for gender As shown in Table 3, the risk factors in females associated with CHD were age and genotype. In males, risk factors included age, lower HDL-cholesterol, higher triglycerides, higher BMI and cigarette smoking. However, in males there was no association with CHD risk and genotype. A possible explanation for this gender difference might be that the combination of these traditional risk factors makes a greater contribution to overall CHD risk, masking the modest genotype-effect. Of interest, significantly more males were taking ACE inhibitors, which lower blood pressure, have cardio-protective effects and also possible anti-oxidant and anti-inflammatory properties. In both males and females, a significant proportion of those with CHD were taking aspirin and statin therapies. This may account for the paradoxical lower LDL levels seen in those subjects.
3.3.
Plasma total anti-oxidant status by genotype
Since SOD2 is known to modulate oxidative stress, we examined the association between plasma TAOS and genotype. As described previously [2,14], plasma TAOS correlated positively with plasma HDL-cholesterol, and negatively with triglyceride and glucose (correlation coefficient r = 0.13, 0.14, and 0.12; all p < 0.05). As observed in Table 2, there was no association between genotype and plasma TAOS for the whole group. After stratifying by gender no differences were observed (for males CC/CT/TT: 43.5 [36.8–50.6]% vs. 45.6 [36.1–55.0]% vs. 44.0 [35.8–50.5]%, p = 0.38; for females 46.7 [36.7–56.0]% vs. 44.8 [35.4–49.7]% vs. 40.8 [32.7–51.1]%, p = 0.13).
No difference was observed after adjustment for the correlates of plasma TAOS. As shown in Table 3, with respect to CHD status, plasma TAOS was lower in the males with CHD. No difference was observed in the females (no difference was observed after adjustment for the correlated of TAOS). Fig. 1 shows TAOS stratified by gender, median TAOS and genotype. As may be observed there was a significant difference in genotype distribution for the females, with the TT genotype being associated with a lower median plasma TAOS. The OR for a plasma TAOS below the median for the TT vs. CC/CT genotype was 2.07 [1.20–3.60], p = 0.009. No association was observed in the males. For plasma Ox-LDL, as previously described within this sample [2], plasma Ox-LDL was independent of pharmacotherapy but was correlated with LDL and HDL (correlation coefficient r = 0.32, 0.26, respectively, p < 0.05). As observed in Table 2, there was no association between genotype and plasma Ox-LDL for the whole group. After stratifying by gender no differences were observed (CC vs. CT vs. TT: males 47.6 [38.2–60.7] vs. 45.1 [33.7–56.5] vs. 47.0 [36.1–57.3], p = 0.22; females 45.9 [36.9–53.3] vs. 48.5 [41.6–57.8] vs. 47.1 [36.6–58.5], p = 0.49). No difference was observed after adjustment for the correlates of plasma Ox-LDL. As shown in Table 3, with respect to CHD status, in both genders, Ox-LDL did not differ by CHD status. No difference was observed in the females (no change was observed after adjustment for the correlated of TAOS).
4.
Discussion
In line with previous studies, we describe a deleterious effect associated with the TT genotype of the rs4880 SOD2 variant. The TT genotype was associated with a greater than twofold increase in CHD risk compared to the C allele in females with diabetes. We observed that the TT genotype was also
200
diabetes research and clinical practice 90 (2010) 196–201
[(Fig._1)TD$IG]
Fig. 1 – rs4880 Genotype distribution by median plasma TAOS. (A) Females: the median plasma TAOS in the group of females was 44%. In this sample, 30.2% of those subjects with the TT genotype were in the lower TAOS group compared to 17.7% in the higher TAOS group. This was statistically significant (x2 = 7.00, p = 0.03, for CC/CT vs. TT, p = 0.01). Plasma TAOS is inversely proportional to oxidative stress (the lower the plasma TAOS, the greater the oxidative burden). (B) Males: the median plasma TAOS in the group of males was 47.5%. No significant difference was seen in the genotype distribution when the group was divided by the median plasma TAOS (x2 = 0.82, p = 0.66).
associated with a lower median plasma TAOS and hence increased plasma oxidative stress. Of interest within the female group the only other statistically significant risk factor was age. Conversely in the males, no association between CHD and genotype was observed. However, risk in males was associated with age, lower HDL-cholesterol, higher triglycerides, higher BMI and cigarette smoking. One possible explanation for the gender difference might therefore be that these risk factors make a greater contribution to overall CHD risk and might mask the modest genotype-effect. Of importance, the males were also prescribed more secondary preventative drugs which may also contribute to this effect. Another speculative explanation for the lack of genotype risk in males may be related to the genotype difference observed by gender. It may be that the higher proportion of males with the CC genotype results from survival bias as a result of increased mortality in those with the TT genotype. Further examination is difficult in this case–control study. There are limitations to our study. We have chosen from the outset to examine one gene variant in SOD2 in relation to intermediate biochemical markers of oxidative stress. We have focused on plasma TAOS and Ox-LDL as intermediate biochemical phenotypes. There are limitations to these
measurements as discussed elsewhere [1,2]. Further work should be performed to look at other variants within the SOD2 regulatory region to allow haplotype based analysis. This would also allow linkage disequilibrium across the region to be examined. At first sight, the lack of any significant difference in genotype distribution between those with and without CHD in males might appear to conflict with that observed in females. However, there are several possible explanations for this as discussed. In addition, prospective gene-association studies are more powerful than case–control studies. Case– control cross-sectional studies are prone to intrinsic bias, for example due to possible altered rates of disease progression, subsequent progression of secondary phenotypes, or genotype associations with death or treatment changes. Furthermore, diabetes with its increased obesity, oxidative burden, inflammation and hyperglycaemia might all overwhelm the SOD2 genotype ‘strength of signal’ and hence the genotype-effect on CHD risk might be lower than in a non-diabetic sample. Therefore, this SNP should be examined in a non-diabetic sample. Such influences are well-recognised confounders [20,21]. Oxidative stress is an important in the underlying pathophysiology of CHD [1,2]. It is likely to play a major role
diabetes research and clinical practice 90 (2010) 196–201
in patients with diabetes, where no more than 25% of the excess CHD risk may be accounted for by traditional risk factors [22,23]. The C>T substitution of the rs4880 gene variant alters the structure of the mitochondrial target sequence, interfering with trafficking pathways resulting in a deleterious effect on overall enzyme activity [5,6]. We show that this change is associated with a modest CHD risk and also that the gene variant is associated with a deleterious change in a global measure of oxidative stress. Further prospective analysis of the gene variant is required in diabetic and no diabetic cohorts with measured plasma markers of oxidative stress however access to such cohorts can be difficult [24].
Acknowledgements Diabetes UK supported JWS (BDA: RD01/0001357) and the creation of UDACS. SEH is supported by the British Heart Foundation (RG2005 014).
Conflict of interest The authors declare that they have no conflict of interest.
references
[1] Stephens JW, Khanolkar MP, Bain SC. The biological relevance and measurement of plasma markers of oxidative stress in diabetes and cardiovascular disease. Atherosclerosis 2009;202:321–9. [2] Stephens JW, Gable DR, Hurel SJ, Miller GJ, Cooper JA, Humphries SE. Increased plasma markers of oxidative stress are associated with coronary heart disease in males with diabetes mellitus and with 10-year risk in a prospective sample of males. Clin Chem 2006;52:446–52. [3] Kiritoshi S, Nishikawa T, Sonoda K, Kukidome D, Senokuchi T, Matsuo T, et al. Reactive oxygen species from mitochondria induce cyclooxygenase-2 gene expression in human mesangial cells: potential role in diabetic nephropathy. Diabetes 2003;52:2570–7. [4] Kinnula VL, Lehtonen S, Koistinen P, Kakko S, Savolainen M, Kere J, et al. Two functional variants of the superoxide dismutase genes in Finnish families with asthma. Thorax 2004;59:116–9. [5] Nomiyama T, Tanaka Y, Piao L, Nagasaka K, Sakai K, Ogihara T, et al. The polymorphism of manganese superoxide dismutase is associated with diabetic nephropathy in Japanese type 2 diabetic patients. J Hum Genet 2003;48:138–41. [6] Hiroi S, Harada H, Nishi H, Satoh M, Nagai R, Kimura A. Polymorphisms in the SOD2 and HLA-DRB1 genes are associated with nonfamilial idiopathic dilated cardiomyopathy in Japanese. Biochem Biophys Res Commun 1999;261:332–9. [7] Flekac M, Skrha J, Hilgertova J, Lacinova Z, Jarolimkova M. Gene polymorphisms of superoxide dismutases and catalase in diabetes mellitus. BMC Med Genet 2008;9:30.
201
[8] Kang D, Lee KM, Park SK, Berndt SI, Peters U, Reding D, et al. Functional variant of manganese superoxide dismutase (SOD2 V16A) polymorphism is associated with prostate cancer risk in the prostate, lung, colorectal, and ovarian cancer study. Cancer Epidemiol Biomarkers Prev 2007;16:1581–6. [9] Zejnilovic J, Akev N, Yilmaz H, Isbir T. Association between manganese superoxide dismutase polymorphism and risk of lung cancer. Cancer Genet Cytogenet 2009;189:1–4. [10] Chistyakov DA, Savost’anov KV, Zotova EV, Nosikov VV. Polymorphisms in the Mn-SOD and EC-SOD genes and their relationship to diabetic neuropathy in type 1 diabetes mellitus. BMC Med Genet 2001;2:4. [11] El Masry TM, Abou Zahra MAM, El Tawil MM, Khalifa RA. Managanese superoxide dismutase Alanine to Valine polymorphism and risk of neuropathy and nephropathy in Egyptian type 1 diabetic patients. Rev Diabetic Stud 2005;2:70–4. [12] Lee SJ, Choi MG, Kim DS, Kim TW. Manganese superoxide dismutase gene polymorphism (V16A) is associated with stages of albuminuria in Korean type 2 diabetic patients. Metabolism 2006;55:1–7. [13] Lee SJ, Choi MG. Association of manganese superoxide dismutase gene polymorphism (V16A) with diabetic macular edema in Korean type 2 diabetic patients. Metabolism 2006;55:1681–8. [14] Dhamrait SS, Stephens JW, Cooper JA, Acharya J, Mani AR, Moore K, et al. Cardiovascular risk in healthy men and markers of oxidative stress in diabetic men are associated with common variation in the gene for uncoupling protein 2. Eur Heart J 2004;25:468–75. [15] Stephens JW, Hurel SJ, Acharya J, Humphries SE. An interaction between the interleukin-6-174G>C gene variant and urinary protein excretion influences plasma oxidative stress in subjects with type 2 diabetes. Cardiovasc Diabetol 2004;3:2. [16] Alberti KG, Zimmet PZ. New diagnostic criteria and classification of diabetes—again? Diabet Med 1998;15:535–6. [17] Sampson MJ, Gopaul N, Davies IR, Hughes DA, Carrier MJ. Plasma F2 isoprostanes: direct evidence of increased free radical damage during acute hyperglycemia in type 2 diabetes. Diabetes Care 2002;25:537–41. [18] Stephens JW, Humphries SE, Miller GJ, Cooper JA, Hurel SJ. Increased oxidative stress is associated with coronary heart disease in males with diabetes mellitus and with 10-year prospective risk in males without diabetes. Diabet Med 2005;22(suppl. 2):P145. [19] Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology 1990;1:43–6. [20] Risch NJ. Searching for genetic determinants in the new millennium. Nature 2000;405:847–56. [21] Sing CF, Stengard JH, Kardia SL. Genes environment, and cardiovascular disease. Arterioscler Thromb Vasc Biol 2003;23:1190–6. [22] Stephens JW, Ambler G, Vallance P, Betteridge DJ, Humphries SE, Hurel SJ. Cardiovascular risk and diabetes. Are the methods of risk prediction satisfactory? Eur J Cardiovasc Prev Rehabil 2004;11:521–8. [23] Pyorala K, Laakso M, Uusitupa M. Diabetes and atherosclerosis: an epidemiologic view. Diabetes Metab Rev 1987;3:463–524. [24] Stephens JW, Bain SC, Humphries SE. Gene–environment interaction and oxidative stress in cardiovascular disease. Atherosclerosis 2008;200:229–38.