Article ID: CJ-14-0233
Background: Elevated homocysteine (Hcy) levels might play a role in the development of essential hypertension (EH). Telomere dynamics provide valuable insight into the pathogenesis of age-related diseases. The contribution of Hcy to leukocyte telomere length (LTL) shortening in EH and the underlying mechanism was examined.
Methods and Results: LTL (ratio of the copy number of telomere [T] repeats to that of a single [S] gene, T/S ratio) was inversely associated with age in patients with EH (n=258) and healthy controls (n=137), but significantly decreased with the Hcy level only in patients with hypertension after adjustment for age and sex. Age, hypertension and levels of Hcy and low-density lipoprotein combined contributed to LTL shortening; an increased serum folate level could reverse the Hcy effect seen on multivariate regression analysis. In addition, qPCR and methylation-specific PCR assay revealed that LTL shortening and mRNA expression and the methylation ratio of human telomerase reverse transcriptase (hTERT) were lower in patients with EH than in controls, and gradually decreased with increasing Hcy level, but not with blood pressure, in EH patients (Ptrend <0.0001, 0.004 and 0.012, respectively). Furthermore, Hyperhomocysteinemia, but not hypertension, promoted telomerase reverse transcriptase DNA hypomethylation and reduced mRNA levels, which contributed to shortened LTL in the hypertension rat model.
Conclusions: Elevated Hcy but not hypertension was related to hTERT DNA hypomethylation and reduced mRNA level, thus contributing to the shortening of LTL hypertension.
Essential hypertension (EH) is a complex disorder involving approximately one-third of the adult population worldwide.1,2 Epidemiological studies have documented that aging as well as genetic and environmental factors such as physical inactivity, obesity, high sodium and low potassium diet, and alcohol consumption are associated with the development of hypertension.3,4 In addition, dysregulated methionine metabolism leading to high plasma homocysteine (Hcy) levels (≥10 mmol/L), known as hyperhomocysteinemia- type EH (HHcy-EH), might play a role.5 However, the mechanisms underlying HHcy-EH and its adverse complications remain unknown.
Editorial p ????
Telomere dynamics (telomere length and its age-dependent shortening) provide valuable insights into the pathogenesis of aging-related diseases. Particularly, clinical epidemiological studies and studies of telomerase- deficient mice have shown a direct link between telomere shortening and hypertension.6 In addition, experimental evidence reveals that both elevated Hcy level and hypertension are related to DNA methylation of leukocytes and are associated with the burden of oxidative stress, inflammation, altered elasticity of the vascular wall, and renal function; all key contributors to age-dependent shortening of leukocyte telomere length (LTL).7
Our previous in vivo and in vitro studies demonstrated that Hcy could promote the recruitment of methylation-sensitive transcription factors to demethylated promoters of atherosclerosis-related genes by reducing DNA methyltransferase 1 activity.8,9 The human telomerase reverse transcriptase (hTERT) gene might regulate telomerase activity and maintain telomere length by encoding the catalytic subunit of the telomerase holoenzyme.5 The hTERT level is tightly regulated by DNA methylation. For instance, 5-aza-2’-deoxycytodine or trichostatin A allows a CCCTC-binding factor (CTCF; a repressor transcription factor) to bind to the hTERT control region, thus reducing the hTERT level and shortening telomere length.10,11
From the above observations, we hypothesized and investigated whether HHcy-related DNA demethylation and the reduced mRNA level of hTERT might be a target of epigenetic regulation and further accelerated LTL shortening in human patients and a rat model with EH.
This was a random hospital-based, cross-sectional case-control study. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Ethics Committee of Peking Union Medical College Hospital/Chinese Academy of Medical Sciences. Informed consent was obtained from all subjects.
A total of 258 patients (130 males, mean age 55.16±9.89 years; 128 females, 57.33±10.17 years) with EH were recruited when they were first admitted to Peking Union Medical College Hospital from June 2012 to May 2013. Arterial hypertension was considered present when there was a systolic blood pressure (SBP) ≥140 mmHg or a diastolic blood pressure (DBP) ≥90 mmHg, which was in accordance with the 2013 European Society of Hypertension and European Society of Cardiology Guidelines.12 From the detailed medical history, 121 cases were caused by primary hypertension and 137 cases were caused by recurrent hypertension with a history of hypertension of approximately 4–12 months. However, SBP, DBP and the Hcy level did not differ between patients with primary and recurrent hypertension. Only 63 of 137 patients with recurrent hypertension received antihypertensive drugs, including propranolol, atenolol, hydrochlorothiazide, nifedipine or nimodipine. Regardless, hypertension was consistent (SBP: 150.1±16.5 and DBP: 97.4±11.0), which might be related to the short treatment time and inconsistent use of medication.
Patients were divided into 4 groups according to their Hcy level: H0 (63 patients, Hcy <10.0 μmol/L), HI (101 patients, Hcy 10.0–15.0 μmol/L), HII (62 patients, Hcy 15.0–30.0 μmol/L) and HIII (32 patients, Hcy ≥30.0 μmol/L). Hypertension complications, at least 1 atherosclerotic lesion or myocardial infarction were defined according to clinical signs and symptoms by physicians and confirmed by neurologic examination, computed tomography, or magnetic resonance imaging.
We also recruited 137 controls (70 males, mean age 57.46±12.03 years; 76 females, 57.74±10.40 years), who were visiting the hospital for a health examination and had a SBP <140 mmHg and a DBP <90 mmHg. Controls had no history of diabetes mellitus, secondary hypertension, myocardial infarction, stroke, renal failure, drug abuse or other serious diseases, and no family history of hypertension in first-degree relatives.
Blood Sample CollectionWe collected 2 blood samples from each subject before any treatment in our hospital, for leukocyte and serum measurements, into 2 vacutainers with and without ethylenediaminetetraacetic acid, respectively. After centrifugation (2,000 g, 5 min, 4°C), serum was separated and placed into a new tube for biochemical measurements. Leukocyte samples were separated into new tubes for DNA and RNA analysis. The leukocyte and serum samples were stored at –80˚C until use.
Biochemical MeasurementsBiochemical variables, such as serum total Hcy, blood glucose, liver and kidney function parameters, and levels of high-sensitivity C-reactive protein (hs-CRP), folate and vitamin B12 were measured by use of a commercial reagent (Beckman). Levels of triglycerides (TG), total cholesterol (TC), high-density lipoprotein (HDL) and low-density lipoprotein (LDL) were measured (Sekisni Medical Co) by use of an automatic analyzer (Beckman AU5800). Coefficients of variation were <10%.
Measurement of Telomere LengthLeukocyte telomerase length was measured by a quantitative PCR (qPCR) method comparing the ratio of the telomere repeat copy number (T) to the single-copy gene copy number (S) (expressed as the telomere length ratio, T/S ratio) in a given DNA sample.13 DNA samples were extracted by using the TIANamp Genomic DNA kit (Beijing), measured by the use of Nanodrop-1000 (NanoDrop Technologies) and further amplified with the use of LightCycler 480 (Roche, Switzerland). A reference DNA sample (transformed human epithelial kidney 293T cells) was paired with each measurement to control for interassay variability. Correlation of a dilution series from 1.56 to 100.00ng (2-fold dilution; 7 points) in telomeres and β-globin qPCR was linear (R2 =0.98). Each DNA sample was measured in triplicate. The mean SD for these measurements was 6.8%. The T/S values were converted to kilobases (Kb) by using the following formula: telomere base pair (Kb) = 1.585 × T/S ratio + 3.582.
Reverse-Transcription qPCR (RT-qPCR)Total leukocyte RNA was extracted by using the Trizol reagent method (Invitrogen, CA, USA) and RNA concentrations were measured by using Nanodrop-1000 and this was converted to cDNA with 0.5 μg RNA samples for hTERT mRNA quantification. RT-qPCR amplification involved the use of LightCycler 480 (Roche, Switzerland), with β-actin as a normalization control and water as a negative control. Both hTERT and β-actin primer sequences were described previously.14 All cDNA samples were assayed in triplicate.
Methylation-Specific PCRAs we previously described, blood genome DNA was modified by use of the BisulFlash DNA Modification kit (Epigentek Group Inc, NY, USA) for methylation-specific PCR (MSP) amplification.15 We used human methylated and unmethylated placental genomic DNA (gDNA; Sigma-Aldrich) in vitro with CpG Methyltransferase (M. Sssl, New England Biolabs, St. Louis, UK) and converted this with sodium bisulfite as a positive and negative control, respectively. MSP was repeated 3 times for each sample. PCR products were separated on agarose gels, visualized by staining with ethidium bromide and quantified by densitometry by using ImageJ. The ratio of DNA methylation was calculated as: methylation/(methylation and unmethylation)×100%.14
Animal ModelThe animal experimental protocol conformed to the Guide for the Care and Use of Laboratory Animals by the US National Institutes of Health (NIH Publication No. 85–23, revised 1996) and was approved by the Peking Union Medical College Hospital Animal Care and Use Committee. Male 4-week-old Sprague-Dawley rats (180–220 g) were maintained in a 12-h light/dark cycle at a controlled room temperature and had free access to standard chow and tap water. We divided rats into 4 experimental groups for treatment (n=8 each): (1) control, standard chow and water; (2) HHcy: standard water and 2% (wt/wt) L-methionine (Sigma, USA) added to standard chow; (3) EH: standard chow and 1.5% NaCl in drinking water; and (4) HHcy-EH: 2% (wt/wt) L-methionine (Sigma, NJ, USA) added to standard chow and 1.5% NaCl in drinking water. After 8 weeks, systolic blood pressure was measured by tail-cuff plethysmography, then all animals were anesthetized (with an intraperitoneal injection of 100 mg/kg ketamine and 10 mg/kg xylazine), blood was collected, and serum was harvested for measurement of total Hcy level by the enzymatic recycling assay. Total RNA and genomic DNA were extracted from circulating leukocytes for qRT-PCR, qPCR or MSP assay.16,17
Statistical AnalysisHomocysteinemia values were positively skewed and were log-transformed for normality and denoted as Lg-Hcy. Data for clinical and biological characteristics of subjects are described as mean±SD. Comparison of the telomere length between patients and controls involved independent-samples t-test (unpaired) or Mann-Whitney 2-sample test for continuous variables and a chi-squared test or one-way ANOVA for categorical variables (according to the normality of variables). Linear regression analysis was used to examine the association of clinical and biological characteristics and LTL, adjusting for age and sex or mutivariables (Lg-Hcy, glucose, lipids, liver and kidney function parameters, hs-CRP, folate and vitamin B12), as well as the contribution of levels of LDL, hs-CRP, folate and vitamin B12 with serum Hcy level to LTL shortening. The correlation between Lg-Hcy, LTL, and hTERT DNA methylation and mRNA level was analyzed by Spearman correlation coefficient with adjustment for age and sex. All analyses involved the use of SPSS 16.0 (SPSS Inc, Chicago, IL, USA). A 2-sided P<0.05 indicated statistical significance.
We first compared the baseline characteristics of 258 patients with EH and 137 age- and sex-matched healthy normotensive controls (Table 1). As compared with normotensive controls, EH patients had higher levels of serum Hcy, creatinine, blood urea nitrogen (BUN), TC, LDL and hs-CRP but lower levels of vitamin B12 and folate. Moreover, LTL was lower for EH patients than controls [5.07±4.59 vs. 5.48±4.82 kb; odds ratio –0.262 (95% confidence interval –4.06 to –0.117), P<0.001], as was previously observed.18,19
| Characteristics | Controls (n=137) | Patients with EH (n=258) | P value |
|---|---|---|---|
| Age (years) | 57.57±12.23 | 56.41±11.11 | 0.357 |
| Sex (M/F) | 70/67 | 130/128 | 0.894 |
| SBP (mmHg) | 110.97±17.93 | 149.40±14.17 | <0.001 |
| DBP (mmHg) | 80.89±7.27 | 95.23±11.52 | <0.001 |
| LTL (T/S value) | 1.20±0.78 | 0.94±0.64 | 0.001 |
| LTL (kb) | 5.48±4.82 | 5.07±4.59 | 0.001 |
| Serum Lg-Hcy level (μmol/L) | 1.05±0.18 | 1.16±0.23 | <0.001 |
| Folate (ng/ml) | 11.1±3.7 | 6.3±5.6 | 0.021 |
| Vitamin B12 (pg/ml) | 512.2±205.7 | 377.5±170.4 | 0.017 |
| Creatinine (μmol/L) | 71.83±14.80 | 84.80±35.54 | <0.001 |
| BUN (mmol/L) | 4.52±1.30 | 4.91±3.85 | 0.012 |
| TC (mmol/L) | 4.72±1.37 | 5.18±1.16 | 0.002 |
| TG (mmol/L) | 1.84±2.78 | 1.80±1.19 | 0.886 |
| HDL (mmol/L) | 1.30±0.37 | 1.28±0.59 | 0.606 |
| LDL (mmol/L) | 2.99±0.85 | 3.44±1.02 | 0.031 |
| hs-CRP (ng/ml) | 1.64±1.92 | 4.21±8.63 | 0.003 |
| ALT (U/L) | 27.96±27.04 | 29.43±46.22 | 0.735 |
| GGT (U/L) | 38.15±47.15 | 37.03±48.55 | 0.840 |
| Glucose (mg/dl) | 5.64±1.85 | 5.92±1.17 | 0.112 |
Data are presented as mean±SD or number.
ALT, alanine aminotransferase; BUN, blood urea nitrogen; DBP, diastolic blood pressure; EH, essential hypertension; GGT, γ-glutamyl transpeptidase; Hcy, homocysteine; HDL, high-density lipoprotein; hs-CRP, high-sensitivity C-reactive protein; LDL, low-density lipoprotein; Lg-Hcy, log-transformed Hcy; LTL, leukocyte telomerase length; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride; T/S ratio, ratio of telomere repeat copy number (T) to single-copy gene copy number (S).
LTL (T/S value) did not differ between males and females (0.96±0.68 vs. 1.09±0.73, P=0.067) but was inversely correlated with age for EH patients and controls (r=–0.188, P=0.011; r=–0.159, P=0.039; Figure 1A) and decreased with age at a mean yearly rate of 0.011±0.005 and 0.009±0.004, respectively. However, the addition of squared or cubed age terms gave non-significant results, which indicates a linear relationship. Relative LTL decreased with SBP and DBP in all subjects after adjustment for age and sex (Figures 1B,C). Furthermore, linear regression analysis revealed that the Lg-Hcy level was negatively correlated with age- and sex-adjusted LTL in EH patients (r=–0.179, P=0.015) but not in normotensive controls (r=–0.115, P=0.185) (Figure 1D). However, there were no associations of SBP and DBP in determining LTL for EH patients (r=–0.128, P=0.086 for SBP and r=–0.101, P=0.094 for DBP) or normotensive controls (r=–0.004, P=0.911 for SBP and r=–0.010, P=0.782 for DBP). Moreover, age- and sex-adjusted LTL was negatively associated with Lg-Hcy and LDL levels, but positively associated with the folate level in EH patients (Table 2). Only glucose was negatively associated with age- and sex-adjusted LTL in controls.
Linear regression analysis of the association between leukocyte telomere length (LTL) and age (A) and homocysteine (Hcy) level adjusted for age and sex (D) in controls and patients with essential hypertension, as well as systolic blood pressure (B) and diastolic blood pressure (C) in all subjects. Telomere length is plotted as the T/S value (ratio of the copy number of telomerase (T) repeats to that of a single (S) gene) and Hcy as the log-transformed Hcy (Lg-Hcy).
| Clinical characteristics | LTL | |||||
|---|---|---|---|---|---|---|
| Controls | Patients with EH | |||||
| B | 95% CI | P value | B | 95% CI | P value | |
| Lg-Hcy (μmol/L) | –0.518 | –1.287 to 0.250 | 0.185 | –0.423 | –0.752 to –0.094 | 0.012 |
| SBP (mmHg) | –0.006 | –0.014 to 0.003 | 0.167 | –0.013 | –0.006 to –0.021 | 0.001 |
| DBP (mmHg) | 0.001 | –0.021 to 0.023 | 0.926 | 0.001 | –0.007 to 0.009 | 0.289 |
| ALT (U/L) | –0.002 | –0.008 to 0.003 | 0.36 | 0.015 | –0.03 to 0.01 | 0.368 |
| GGT (U/L) | –0.002 | –0.005 to 0.001 | 0.243 | 0.005 | –0.002 to 0.01 | 0.704 |
| Creatinine (μmol/L) | –0.009 | –0.02 to 0.003 | 0.13 | –0.005 | –0.03 to 0.02 | 0.559 |
| BUN (mg/dl) | –0.059 | –0.16 to 0.043 | 0.255 | 0.000 | –0.011 to 0.1 | 0.931 |
| Glucose (mg/dl) | –0.099 | –0.168 to –0.03 | 0.005 | –0.059 | –0.126 to 0.008 | 0.082 |
| TC (mmol/L) | 0.039 | –0.075 to 0.153 | 0.501 | 0.01 | –0.076 to 0.97 | 0.818 |
| TG (mmol/L) | –0.02 | –0.068 to 0.027 | 0.398 | –0.018 | –0.091 to 0.055 | 0.636 |
| HDL (mmol/L) | 0.12 | –0.276 to 0.515 | 0.55 | 0.134 | –0.015 to 0.284 | 0.078 |
| LDL (mmol/L) | 0.058 | –0.096 to 0.212 | 0.46 | –0.214 | –0.413 to –0.014 | 0.038 |
| hs-CRP (ng/ml) | –0.039 | –0.135 to 0.057 | 0.424 | –0.116 | –0.221 to 0.011 | 0.131 |
| Folate (ng/ml) | 0.1445 | –0.021 to 0.268 | 0.075 | 0.2235 | 0.037 to 0.410 | 0.024 |
| Vitamin B12 (pg/ml) | 0.104 | –0.097 to 0.305 | 0.112 | 0.201 | –0.075 to 0.477 | 0.098 |
95% CI, 95% confidence interval. Other abbreviations as in Table 1.
We further examined whether Lg-Hcy, LDL and folate explained the shortening of LTL in EH patients (Table 3). Multifactorial linear regression analysis revealed that age, hypertension, Lg-Hcy and LDL levels accounted for 12.5% (P<0.001), 5.9% (P=0.028), 6.2% (P=0.021) and 3.9% (P=0.015), respectively, of the LTL variability when entered into the model step by step. More importantly, age, hypertension and LDL level but not Lg-Hcy level (6.2%, P=0.110) persistently accounted for variability in LTL after folate was entered into the model (accounting for 8.5%, P=0.022). The entry of creatinine, BUN, TC and hs-CRP into the model did not further explain the variability in LTL.
| Independent variable | Estimate* | P value | Change in R2† | Model R2 |
|---|---|---|---|---|
| Intercept | 1.684±0.193 | <0.001 | ||
| Age | –0.011±0.003 | 0.001 | 12.467 | 12.467 |
| Intercept | 1.897±0.198 | <0.001 | ||
| Age | –0.012±0.003 | <0.001 | 12.467 | |
| Hypertension | –0.285±0.078 | 0.028 | 5.862 | 17.634 |
| Intercept | 2.534±0.275 | <0.001 | ||
| Age | –0.012±0.003 | <0.001 | 12.467 | |
| Hypertension | –0.225±0.079 | 0.028 | 5.862 | |
| Lg-Hcy | –0.605±0.184 | 0.013 | 6.234 | 23.868 |
| Intercept | 2.897±0.303 | <0.001 | ||
| Age | –0.012±0.003 | <0.001 | 12.467 | |
| Hypertension | –2.226±0.079 | 0.028 | 5.862 | |
| Lg-Hcy | –0.601±0.184 | 0.021 | 6.234 | |
| LDL | –0.207±0.033 | 0.015 | 3.878 | 27.746 |
| Intercept | 3.197±0.303 | <0.001 | ||
| Age | –0.012±0.003 | <0.001 | 12.467 | |
| Hypertension | –2.226±0.079 | 0.025 | 5.862 | |
| Lg-Hcy | –0.601±0.184 | 0.110 | 6.234 | |
| LDL | –0.207±0.033 | 0.045 | 3.878 | |
| Folate | 0.265±0.024 | 0.022 | 8.502 | 36.248 |
*Estimates are unstandardized β coefficients±SE.
†R2 values are increments in R2, as variables are sequentially added to the model.
Abbreviations as in Table 1.
hTERT, which maintains telomere length, is tightly regulated by DNA methylation.8,9 We therefore examined the contribution of hTERT DNA methylation and mRNA level to the short LTL in EH patients. As expected, as compared with controls, EH patients showed a lower hTERT methylation ratio (65±20% vs. 57±21%, P<0.001) and mRNA level (1.00±0.76 vs. 0.82±0.58, P=0.025), which was consistent with short LTL (1.20±0.78 vs. 0.93±0.64, P<0.001; Figures 2A–C).
Human telomerase reverse transcriptase (hTERT) mRNA level (A), methylation of hTERT promoter (B) and leukocyte telomere length (LTL) (C) in patients with hypertension and controls, and among grades of H-type hypertension in patients (by Hcy level) (D – F). Median (horizontal line) and range (box) of telomere length are shown. Whiskers are from the 10% and 90% range.
Because the accumulation of Hcy induces DNA hypomethylation via increased intracellular levels of S-adenocylhomocyesteine, a transmethylation inhibitor,20 we performed linear regression analyses and found that LTL as well as hTERT mRNA level and the methylation ratio were inversely associated with the Hcy level in EH patients (Ptrend <0.0001, 0.004 and 0.012, respectively; Figures 2D–F). Upon adjustment for age and sex, both LTL and the hTERT methylation ratio were inversely related with Lg-Hcy in EH patients (Table 4). We found a positive correlation of hTERT mRNA level with LTL and the hTERT methylation ratio. In addition, only folate, but not vitamin B12 levels were negatively related with Hcy level but positively related with LTL. Neither folate, nor vitamin B12 was not related with DNA methylation and expression of hTERT. However, we found no correlation between SBP and the hTERT methylation ratio (r=0.044, P=0.273) or mRNA expression (r=–0.082, P=0.108).
| R values | Folate | Vitamin B12 | Lg-Hcy | LTL | hTERT methylation ratio | hTERT mRNA level |
|---|---|---|---|---|---|---|
| Folate | 1.000 | 0.068 | –0.158* | 0.109* | 0.079 | 0.082 |
| Vitamin B12 | 1.000 | –0.088 | 0.044 | 0.061 | 0.074 | |
| Lg-Hcy | 1.000 | –0.146* | –0.163* | –0.107* | ||
| LTL | – | 1.000 | 0.106 | 0.156* | ||
| hTERT methylation ratio | – | – | 1.00 | 0.127* | ||
| hTERT mRNA level | – | – | 1.00 |
*P<0.05 as determined by Spearman correlation.
hTERT, human telomerase reverse transcriptase. Other abbreviations as in Table 1.
Among the 258 EH patients, 101 had clinical atherosclerotic lesions or myocardial infarction in the cerebrovascular region (23 cases), carotid artery (21 cases) or coronary artery (57 cases). Patients with EH and clinical atherosclerotic lesions showed a higher level of Lg-Hcy but lower LTL, hTERT mRNA level and methylation ratio than patients without clinical lesions (Figure 3).
Log-transformed homocysteine (Lg-Hcy) level, leukocyte telomere length (LTL) and human telomerase reverse transcriptase (hTERT) mRNA level and methylation ratio in hypertension patients with and without clinical atherosclerotic lesions. Data are presented as mean±SD.
Rats and humans have the same CTCF binding site on the TERT gene core promoter region (–200~+100bp; Figure 4A), so we considered that the regulation mechanism of TERT by DNA demethylation of CTCF binding sites would be similar. Therefore, we determined the pathophysiological relevance of the EH rat model (BP mean 156.1±10.4 mmHg) with or without HHcy (total serum Hcy level 66.8±11.7 mmol/L) with 1.5% NaCl drinking water with or without 2% (wt/wt) L-methionine in a chow diet for 8 weeks (Figures 4B,C). qPCR revealed decreased LTL in circulating leukocytes, by 37.9±21.2% (P=0.045), 31.7±26.0% (P=0.037) and 43.8±18.4% (P=0.018), in HHcy, EH and HHcy-EH rats, respectively, as compared with control rats (Figure 4D). Interestingly, rat TERT (rTERT) was also downregulated by 27.6±10.46% (P=0.025) in HHcy rats and 46.1±9.4% (P=0.0006) in HHcy-EH rats as compared with controls (Figure 4E). Similarly, a MSP assay revealed decreased methylation ratios of the rTERT promoter (CTCF binding sites) by 36.86±16.40% (P=0.01) in HHcy rats and 53.63±8.75% (P<0.0001) in HHcy-EH rats (Figures 4F,G), with no significant change in rTERT expression and methylation in EH rats. SBP was negatively associated with relative LTL (r=–0.39, P=0.027) but not with the hTERT methylation ratio (r=–0.17, P=0.365) or mRNA expression (r=0.20, P=0.271). Thus, our clinical and animal data suggest that HHcy plays an important role in LTL shortening via DNA hypomethylation on the TERT promoter in hypertension.
Rat model of the elevated level of homocysteine (Hcy) combined with hypertension with shortened leukocyte telomere length (LTL) as determined by telomerase reverse transcriptase (TERT) methylation. (A) Homology analysis of the CTCF (CCCTC-binding factor) binding site on the promoter of TERT in humans and rats by using the University of California Santa Cruz (UCSC) Genome. Browser website Italics are CpG sites, the transcription start sites (TSS+1; the right-angled arrow). (B) Mean total plasma (B) Hcy and (C) systolic blood pressure in rats. (D) Relative circulating LTL expressed as a T/S ratio value, which was detected by using quantitative polymerase chain reaction (qPCR). (E) TERT mRNA expression in leukocytes. Rat β-actin was an internal control. (F) Methylation-specific PCR analysis of the methylation pattern of the CTCF binding site on rat TERT (rTERT) promoter in circulating leukocytes by using ethidium bromide staining. DM, DNA marker; M, methylated; U, unmethylated. (G) Ratio of DNA methylation to total methylation and unmethylation of rat TERT promoters. *P<0.05 and **P<0.01 vs. controls provided with a standard diet and water for 8 weeks.
Age-dependent shortening of LTL represents the accruing burden of environmental injury and genetic predisposition; 2 key contributors are HHcy and hypertension progression.21 Here, we defined LTL shortening in patients with EH and DNA methylation as the underlying mechanism. The novel findings are that: (1) LTL was lower in EH patients than controls and that this decreased with age, Hcy and blood pressure level in EH patients; (2) age, hypertension and levels of Hcy and LDL combined contributed to short LTL, and folate status might reverse the effect of Hcy; and (3) elevated Hcy level, but not hypertension, induced DNA hypomethylation of the hTERT promoter, which contributed to the restored hTERT mRNA level and short LTL in human and rat models with hypertension.
Similar to smoking or hyperlipidemia, the plasma total Hcy level represents an independent risk factor for vascular disease via arteriolar constriction, renal dysfunction, arterial stiffness, and diminishing vasodilation.22,23 Both Hcy level and hypertension could increase oxidative stress and diminish the nitric oxide (NO) level and inflammation.24 The question then is whether a biomarker can indicate the cumulative lifelong burden of oxidative stress and inflammation. Telomere attrition, especially LTL, might be a biomarker that predates leukocyte collection.25 Indeed, LTL is associated with several diseases such as cancer,26 atherosclerotic cardiovascular disease,15 and hypertension.18,19 LTL is also inversely associated with cardiovascular disease risk factors, including cigarette smoking, obesity, sedentary lifestyle, and an unhealthy lifestyle in general.27 However, we lack studies on loss of telomere length and H-type hypertension.
Our study confirmed that LTL was shorter and significantly decreased with age in EH patients as compared with normotensive controls. Particularly, Hcy concentration was negatively correlated with sex-adjusted LTL in EH patients but not normotensive controls. Furthermore, in addition to Hcy combined with hypertension, the LDL level contributed to a short LTL. These findings provide new evidence to support the fact that HHcy is associated with increased cardiovascular risk with hyperlipidemia.28 Previous meta-analysis of 10 trials of no or partial folic acid fortification with stroke (n=43,426 subjects) indicated that folic acid supplementation could prevent stroke by approximately 11%.29 We also found that an increased folate level might reverse the effect of Hcy on shortening LTL. Folate, as the most important dietary determinant of plasma total Hcy level, plays a pivotal role in Hcy-related diseases. Daily supplementation with 0.5–5 mg folic acid typically reduces plasma Hcy concentration by approximately 25%.30
The common thread that links hypertension and elevated plasma Hcy levels with short LTL might be oxidative stress, insulin resistance and inflammation.24,31 Importantly, we found HHcy-EH, but not hypertension, closely related to hTERT DNA demethylation and reduced mRNA level, which contributed to a short LTL. EH patients with clinical atherosclerotic lesions showed a higher level of Hcy but lower LTL and hTERT methylation ratio and mRNA level than EH patients without lesions. This result further highlights the role of DNA methylation in short LTL in HHcy-EH. Evidence of the association of DNA methylation and risk of HHcy-EH is scarce but expected because of a significant decrease in global DNA methylation in EH patients with progression of hypertension. Thus, our data support the findings that hTERT is a novel target gene in HHcy-EH that induces LTL shortening by an epigenetic mechanism on CTCF binding sites of the hTERT promoter in circulating leukocytes. The mechanism is the same as hTERT acting as a candidate target gene by DNA hypomethylation in various cancers.32
Rats and humans have the same CTCF binding site on the core promoter region (–200~+100 bp) of the TERT gene, with high sequence homology. Our 8-week HHcy-EH rat model suggested a similar pattern of TERT expression and regulation in humans and rats. HHcy, but not hypertension could promote DNA hypomethylation of CTCF binding sites of TERT and reduce mRNA expression, which contribute to shortened LTL in rats with or without hypertension. In addition, we also found HHcy and/or hypertension-decreased bio-availability of NO by downregulating endothelial NO synthase (NOS) and upregulating inducible NOS expression in circulating leukocytes (Figure S1) also related to shortened LTL.21 Therefore, HHcy-EH might cause shortened LTL via DNA hypermethylation of the TERT gene promoter, independent of NO availability.
One limitation of our study is that we did not measure telomere activity. Decreasing telomere activity itself but not telomere length is the primary determinant of phenotypes. Second, calculating the effect of antihypertensive drugs on telomere length or Hcy level is difficult, which might be related to the short time of treatment and inconsistent use of medicine (multiple types, overlap, additions or dose). However, no report has investigated the effect of these antihypertensive drugs on telomere length or Hcy level. Our retrospective study catalogued the cross-sectional relationship between telomere length and indices of cardiovascular risk, which might provide valuable information regarding possible confounders in telomere dynamics among humans.
In summary, our findings indicate that LTL is negatively related to age and significantly decreases with Hcy level in EH patients. Age, Hcy and LDL level combined contribute to short LTL in subjects with EH, and only an increased folate level might reverse the effect of Hcy. Hcy, but not hypertension related to hTERT DNA hypomethylation and reduced mRNA level in parallel with hypertension, contributed to the shortened LTL. Therefore, similar to antihypertension, folate and/or vitamin B6/B12 supplementation is important to prevent telomere dynamics in hypertension. Our findings highlight the role of Hcy-related DNA demethylation, an important part of the epigenetic mechanism, especially for hyperhomocysteinemia-type hypertension.
This work was supported by a PUMC Youth Fund, and the Fundamental Research Funds for the Central Universities (no. 3332013004). We acknowledge Laura Smales (BioMedEditing, Toronto, Canada) for critical reading of the manuscript.
The authors declare no actual or potential conflict of interest.
Supplementary File
Figure S1. Hyperhomocysteinemia (HHcy) and/or hypertension upregulated endothelial nitric oxide synthase (eNOS) and downregulated inducible NOS (iNOS) mRNA expression in circulating leukocytes as compared with control rats (n=8).
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-14-0233
