Article Text
Abstract
Objective: Studies in white people have shown that telomere length, a marker of biological ageing, is shorter in individuals with coronary artery disease (CAD). South Asian Indians have a high prevalence of CAD, especially premature CAD. We examined the association of telomere length with CAD in Indian subjects.
Design: Case-control study.
Setting: Mumbai, India.
Subjects: 238 consecutive patients (aged 29–82 years), admitted to Cumballa Hill Hospital for coronary investigations or treatment and 238 control subjects (aged 30–87 years) from the same area without any clinical evidence of CAD.
Methods: Mean leucocyte telomere length was measured using a polymerase chain reaction (PCR)-based assay and expressed as a ratio (T/S ratio) of the telomere signal to that of a control single copy gene.
Results: T/S ratio was significantly lower in CAD cases compared with controls (cases 1.21 (95% CI 1.16 to 1.26); controls 1.33 (1.28 to 1.38); p = 0.0003), equivalent to approximately 166 base pairs. The difference remained significant after adjustment for other clinical characteristics (p = 0.002). There were trends towards longer telomeres in vegetarian subjects compared with subjects on a mixed diet (vegetarians 1.31 (1.25 to 1.38); mixed 1.25 (1.18 to 1.33); p = 0.088) and shorter telomeres in subjects with a positive family history (FH) for CAD (+ve FH 1.25 (1.18 to 1.32); -ve FH 1.31 (1.24 to 1.38); p = 0.094).
Conclusion: Subjects of Indian ethnicity with CAD have shorter telomeres than subjects without such a history. The finding provides further evidence that telomere biology is altered in subjects with CAD.
Statistics from Altmetric.com
While epidemiological studies have identified several cardiovascular risk factors that potentially explain a large portion of coronary artery disease (CAD) risk in a population,1 at an individual level there is wide variation in both the occurrence of CAD and its age of manifestation, even in subjects with similar risk factor profiles. The reasons for this wide inter-individual variation in susceptibility are poorly understood. CAD is an age-associated disease2; cellular senescence is a major feature of atherosclerotic plaques,3 4 and in vitro induction of senescence in coronary endothelial cells causes expression of molecules implicated in atherogenesis.5 This has resulted in the hypothesis that, at least to some extent, inter-individual variation in risk of CAD may reflect variation in rate of biological ageing.6
Telomeres comprise the very distal ends of eukaryotic chromosomes, and consist of a large number of tandem repeats (TTAGGG in humans). Although the full functions of telomeres are not fully understood, they are known to be involved in maintaining chromosomal stability and in cell cycle regulation.7 Because DNA polymerase cannot fully complete the replication of the 3′ end of linear DNA, telomeres progressively shorten with repeated cell division.8 Olovnikov was the first to suggest that this is a potential mechanism for a biological clock determining cellular behaviour,9 and this concept has since been supported experimentally.10 11 In many cell types, senescence and subsequent cell death often occur when the mean telomere length reaches a critical value.12 Therefore mean telomere length provides a marker for biological age, at least at the cellular level, with shorter telomeres indicating increased biological age.
In previous studies, we have shown that mean telomere length in leucocytes is shorter in patients with severe triple vessel CAD compared with subjects with angiographically normal coronary arteries,13 and in patients who have had a premature myocardial infarction (MI) compared with age-matched and sex-matched subjects without such a history.14 We have also demonstrated that shorter mean telomere length at baseline is a predictor of future coronary events in middle-aged men in a large prospective analysis.15
Studies to date examining the association of telomere length with CAD or atherosclerosis have been done in subjects of white European origin or Japanese subjects.13–20 CAD, especially premature CAD, is particularly prevalent in South Asians subjects.21 22 Consistency of finding of association of shorter telomere with CAD in different ethnic groups would add weight to the hypothesis that it may have a primary role in the disease. In the present study, we examined the association of mean leucocyte telomere length in Indian patients with CAD and control subjects recruited in India.
METHODS
Subjects and recruitment
A total of 504 subjects (252 cases and 252 controls) were recruited between 2000 and 2002, in Mumbai, India, following approval of the local ethics committee. The cases comprised consecutive subjects (195 male and 57 female aged 29–82 years), admitted to Cumballa Hill Hospital and Heart Institute, Mumbai, for coronary investigations or treatment. Seventy-four per cent of the cases had been admitted for diagnostic angiography, 20% for percutaneous transluminal coronary angioplasty (PTCA) and 6% for coronary artery bypass graft surgery (CABG). Of the 252 patients, 84 (33%) (73 male and 11 female) had had an MI previously. The diagnosis of MI was confirmed by direct examination the patients’ medical records and based on clinical symptoms, ECG changes and recorded raised levels of cardiac enzymes in the circulation.
The control group of 252 individuals (195 male and 57 female aged 30–87 years) was drawn from the general population. None of the controls had ever had an MI or stroke, none had a personal history of CAD or was ever required to undergo coronary angiography. None was on any anti-anginal medication, and all were negative for angina-type symptoms as assessed by Rose’s angina questionnaire.23 As with the patients, known cases of diabetes mellitus or hypertension among controls were not excluded. In order to facilitate their enrolment, circulars from the hospital management explaining the purpose of the study and inclusion/exclusion criteria were sent to people living and working in the neighbourhood of Cumballa Hill Hospital.
All patients and controls gave their written informed consent to participate in the study. All the patients and controls were of Indian origin, unrelated to one another and of similar socioeconomic status. Demographic information, anthropometric and diet data, personal medical history, details of current medication, lifestyle and coronary risk factors and history of cardiovascular disease, diabetes mellitus or hypertension among first-degree relatives were recorded on a questionnaire, and entered in a computer database. Hypertension and diabetes were based on reported history. Smoking was classified as never smoked or current/ex-smoker. Family history of cardiovascular disease was defined as a history of fatal or non-fatal heart attack/stroke/sudden death in at least one first-degree relative before the age of 60 years.
Coronary atherosclerosis
All case subjects had undergone angiography, either during the index admission or previously. All coronary angiograms were reviewed by the responsible cardiologist. Presence of stenosis of >50% coronary arterial diameter reduction assessed visually was considered as significant disease, and classified as single, two or triple vessel disease according to the American College of Cardiology/American Heart Association guidelines for coronary angiography.24
Blood sampling
Overnight fasting samples (12–14-hour fast) of venous blood were collected with minimum stasis in EDTA and citrate anticoagulants (Vacuette, Greiner Bio-One GmbH, Austria). The blood plasmas were separated by centrifugation, aliquoted and stored at −80°C until assay, and the cell pellets were stored at −20°C until DNA extraction in batches by standard salting out method. The methods for determining total, high-density lipoprotein (HDL)-cholesterol, and low-density lipoprotein (LDL)-cholesterol and triglyceride have been described previously.25
Real-time PCR analysis of telomere length
Mean telomere length was assessed using a real-time polymerase chain reaction (PCR)-based technique26 which compares telomere repeat sequence copy number to single-copy gene (36b4) copy number in a given sample, with minor modifications.15 Duplicate DNA samples were amplified in parallel 20 μl PCR reactions; both reactions included 20 mM TRIS-HCl, pH 8.4, 50 mM KCl 200 μM each dNTP, 1% DMSO, 2.5 mM DTT, 0.4× Sybr Green I (Invitrogen, Paisley, UK) and 20 ng of sample DNA. The telomere-specific reaction included 1.5 mM MgCl2, 0.8 units of platinum Taq DNA polymerase (Invitrogen, Paisley, UK) and 300 nM of each of the telomere-specific primers (forward: 5’CGGTTTGTTTGGGTTTGGGTTTGGG TTTGGGTTTGGGTT 3’; reverse: 5’GGCTTGCCTTACCCTTACCCTTACCCTTACCCTTACCCT3’). The 36b4 reaction included 3.5 mM MgCl2, 0.5 units of platinum Taq, and 300 nM of the forward primer (5’CAGCAAGTGGGAAGGTGTAATCC3’) primer and 500 nM of the reverse primer (CCCATTCTATCATCAACGGGTACAA). All PCRs were performed on the Corbett Research Rotor-Gene 3000 Real-time Thermal Cycler (Corbett Research, Cambridge, UK). The thermal cycling profile for both amplicons began with a 95°C incubation for 10 minutes to activate the PlatinumTaq DNA polymerase. For telomere PCR, there followed 25 cycles of 95°C for 15 seconds, 52°C for 1 minute. For 36B4 PCR, there followed 35 cycles of 95°C for 15 seconds, 58°C for 1 minute. The specificity of all amplifications was determined by melting curve analysis. Thirty-four study samples, a calibrator sample (see below) and one no-template control sample (all in duplicate) were processed per run.
The PCR data were analysed using the comparative quantitation approach.27 Briefly, this approach, implemented using the Corbett Research Rotor-Gene 6.0.1 analysis software, calculates the second derivative of the real-time amplification curve. The peak of the second derivative curve represents the maximum exponential growth of the product. The point 80% back from the peak is taken as the take-off point, and the amplification efficiency calculated from the section of curve between these two points. This was done for all the samples in a run and a mean amplification efficiency (MAE) calculated. The concentration value for each unknown sample was calculated relative to the calibrator sample using the mean amplification efficiency value and the respective take-off points using the formula:
Relative concentration = MAE(calibrator takeoff – sample takeoff)
The same calibrator sample was used in all runs to allow comparison of results across runs.
This process was performed for both the telomere and single-copy gene reactions, and telomere length expressed as a ratio of the two (telomere/single-copy gene, referred to as the T/S ratio), using the averaged data from the duplicate runs. All analysis was undertaken blinded to case-control status of the subject. Validation and reproducibility of the PCR method used and correlation with the Southern blot measurement of telomere length have previously been described.15
Statistical analyses
DNA was not of sufficient quality for telomere measurements for 14 cases and 14 controls. Analysis was therefore performed on data from 476 subjects (238 cases and 238 controls). χ2 Tests of association (categorical data) or t tests (continuous data) were used to compare the characteristics of the two groups. All continuous variables were examined using standard methods and were found to have come from a normal distribution. A generalised linear model in SAS version 9.1 was used to compare the mean differences in telomere length between cases and controls. Linear regression was used to investigate the relation between covariates and telomere length. To relate observed differences in T/S ratio to differences in telomere length in base pairs, we used the equation (y = 1.38x + 5.24) of the regression line from our previous correlation of telomere length measured by the PCR method and by Southern blotting.15
RESULTS
Characteristics of cases and controls are shown in table 1. Subjects were well matched for age, sex, body mass index and smoking history. There was no difference in total or LDL cholesterol between cases and controls. However, triglyceride levels, hypertension, diabetes and family history of CAD were significantly higher in cases. The lack of difference in LDL cholesterol level between cases and controls may reflect a higher prevalence of statin treatment at the time of sampling in cases (25.0%) versus controls (0.02%). A dyslipidaemia marked by higher triglycerides and a low HDL and a smaller difference in LDL is also said to be more typical of Indian patients with CAD.28
The distributions of mean leucocyte telomere length assessed by real-time PCR and presented as the T/S ratio in CAD cases and controls are shown in figure 1. Mean leucocyte telomere was significantly shorter in subjects with CAD compared with controls (cases 1.21 (95% CI 1.16 to 1.26); controls 1.33 (1.28 to 1.38); p = 0.0003). The difference remained significant after adjustment for all the characteristics shown in table 1 (cases 1.22 (1.15 to 1.29); controls 1.34 (1.27 to 1.42); p = 0.002). Telomere length was also shorter in the subgroup of cases (n = 76) with MI (p = 0.008).
Thirty-three (14%) case subjects did not have significant CAD at angiography defined as at least one vessel with a stenosis >50%. Exclusion of these subjects did not affect the association of CAD with shorter telomeres (p = 0.0004). The proportions of cases with one, two and three-vessel disease were 22%, 25% and 39%, respectively. There was no relation between T/S ratio and number of diseased vessels (p = 0.327).
Analysis of the relation between telomere length with other characteristics showed no significant association with age, gender or other cardiovascular risk factors (table 2). Notably, there was no significant correlation of telomere length with age in either cases (r = −0.0464, p = 0.476) or controls (r = 0.0646, p = 0.327). Interestingly, there was trend to suggest that individuals reporting a vegetarian-only diet had longer telomeres than those with a mixed diet, taking other characteristics into account (vegetarians 1.31 (1.25 to 1.38); mixed 1.25 (1.18 to 1.33); p = 0.088). There was also a trend towards shorter telomeres in subjects with a positive family history (FH) for CAD, although this did not reach statistical significance either (+ve FH 1.25 (1.18 to 1.32); -ve FH 1.31 (1.24 to 1.38); p = 0.094).
DISCUSSION
This is the first study to investigate the relation between telomere length and CAD in Indian subjects. Consistent with findings in other ethnic groups,13–16 18–20 the study shows that Indian subjects with CAD have significantly shorter telomeres than matched controls without such a history. Based on our previous correlation of telomere length measured by the PCR method and by Southern blotting,15 the observed difference in T/S ratio between cases and controls of 0.12 is equivalent to a difference in telomere restriction fragment length of around 166 base pairs. This should be viewed as illustrative, as the correlation between telomere length measured by the two methods is based on observations made in white subjects and could be different in Indian subjects. None the less, the findings provide further evidence for the association between shorter telomeres and CAD and demonstrate that this association is seen in different ethnic groups.
A strength of the present study is that the subjects were unselected consecutive cases attending a single institution with well-matched controls from the same geographic area. A surprising finding was that neither the cases nor the controls showed the age-associated attrition in telomere length seen in most studies of this type.13–15 29 30 The reasons for this are unclear. Both cases and controls spanned a substantial age range. Each cohort was of modest size and it is possible that this mitigated against detecting an age-related effect on telomere length, although we have readily detected such an effect in previous studies of a similar size or even smaller.13 14 Similarly, it is unlikely that the lack of observed age-related attrition reflects imprecision of the assay. We and others have previously seen age-related attrition using the same assay.15 30 Furthermore, each sample was assayed in duplicate and the coefficient of variance between assays was only 1.5%. Finally, the intriguing possibility that the lack of significant age-related attrition in this study reflects ethnic differences in this parameter remains to be investigated.
Although we observed a highly significant difference between subjects with CAD and control subjects, we did not see any relation between telomere length and the number of diseased coronary arteries, defined as arteries with a >50% maximal stenosis. Maximal stenosis which is widely used to categorise the number of diseased coronary arteries is an imprecise measure of the severity or extent of coronary atheroma as most of the atheroma is accommodated within the vessel wall with little impact on luminal size until the process is very advanced by external remodelling of the vessel wall. Therefore, the lack of correlation between telomere length and number of coronary arteries with >50% stenosis does not exclude an association of shorter telomeres with greater atheroma burden. Demonstration of such an association will require more precise methods such as intravascular ultrasound for quantifying the amount of coronary atherosclerosis.
Consistent with previous studies,14 15 we did not see any association of telomere length with other conventional cardiovascular risk factors. This suggests that the association of shorter telomeres with CAD is independent of these risk factors. However, there was a trend to suggest that telomere length may be marginally longer in those subjects who were vegetarians compared to those who had a mixed diet. This exploratory finding opens the possibility that dietary content and/or habits may affect telomere attrition. This intriguing possibility requires investigation in larger cohorts specifically characterised to test this hypothesis. There was also a trend towards shorter telomeres in subjects with a positive family history of premature CAD. This finding is similar to our previous observations in white subjects with CAD and controls,14 and also consistent with the recent finding that telomere length was shorter in offspring of subjects with CAD.31 The finding points to a possible primary involvement of telomere length in the pathogenesis of CAD.6 31
CAD is more prevalent in Indian subjects and occurs at an earlier age than in white subjects.21 22 This raises the question as to whether there are ethnic differences in telomere length. Our study was not designed to address this question, which requires head-to-head comparison of matched subjects, ideally randomly selected from the general population rather than disease cases or controls, from the two ethnic groups.
In summary, we show that subjects of Indian ethnicity with confirmed CAD have shorter telomeres than matched subjects without such a history. Our case-control study design does not permit a cause-effect nature of the association to be established. None the less, the finding provides further evidence that telomere biology is altered in subjects with coronary artery disease.
Acknowledgments
We are grateful to the medical and research staff of Cumballa Hill Hospital and Heart Institute, Mumbai, India, for their help with the study. The work was supported by the Baun Foundation Trust, Mumbai India, and the British Heart Foundation. NJS holds a chair supported by the British Heart Foundation.
REFERENCES
Footnotes
-
MM and SS contributed equally to this work.
-
Competing interests: None.
-
MM’s present address: Department of Food Engineering and Technology, Institute of Chemical Technology (Autonomous), University of Mumbai. N Parekh Marg, Matunga, Mumbai, India.