Acute hyperhomocysteinaemia affects pulse pressure but not microvascular vasodilator function

Abstract

Aims

Raised homocysteine (hcy) levels are associated with premature coronary artery disease, but the underlying vascular mechanism and the extent to which hcy affects small vessel vasodilator responses (especially non-nitric oxide mediated pathways) are unclear.

Methods

This double-blind, placebo-controlled crossover study in 14 healthy male subjects evaluated the effects of single-dose oral methionine 15 g (to induce acute hyperhomocysteinaemia) on cutaneous microvascular vasodilator responses to incremental-dose iontophoretic administration of acetylcholine (Ach) and sodium nitroprusside (SNP) using laser Doppler fluximetry (LDF), and the effects on von Willibrand factor (vWF) levels and systemic haemodynamics.

Results

Methionine administration produced a three fold rise in plasma hcy levels at 8 h, which was accompanied by a significant increase in pulse pressure (53 vs 49 mmHg, P < 0.05) but no change in heart rate. Acute hyperhomocysteinaemia had no significant effect on incremental microvascular vasodilator dose–response curves to Ach and SNP, or circulating levels of vWF.

Conclusions

The present study shows that acute hyperhomocysteinaemia increases pulse pressure (a marker of large vessel stiffness) but has no effect on endothelial-dependent (non-NO-mediated) microvascular vasodilation.

Introduction

In large population studies, mild-to-moderate increases in plasma homocysteine (hcy) concentration are associated with a higher incidence of coronary heart disease [1, 2], peripheral arterial disease [3, 4], stroke [5], and venous thrombosis [6]. An overall analysis estimates the relative risk to be 1.4 for the difference between hcy levels > 15 µmol l⁻¹ compared with hcy levels < 10 µmol l⁻¹ after adjustment for other cardiovascular risk factors [7]. In high risk groups, e.g. patients with acute coronary syndromes, plasma hcy concentration is an independent predictor of long-term survival [8, 9].

Hcy is a sulphur-containing aminoacid and, via demethylation, an intermediary product of methionine metabolism. Circulating hcy levels are determined by genetic and nutritional factors, and oral administration of a methionine load provides a convenient diagnostic test to identify individuals with enzyme defects or poor vitamin status (e.g. vitamin B₆, B₁₂ or folate deficiency) in whom there is an exaggerated rise in plasma homocysteine levels [10]. Oral methionine induces early atherosclerotic lesions in animals [11], and single-dose administration has been used to detect endothelial dysfunction associated with acute hyperhomocysteinaemia in healthy volunteers [12–14].

Whether hcy itself is a cardiovascular risk factor, or simply a marker for some other disease process, remains open to further investigation [15], but clinical and experimental studies have shown that hcy has undesirable effects on endothelial function [12–14] (especially nitric oxide mediated pathways [16]), lipoprotein oxidation [17], thrombogenesis [18] and free radical formation [19]. Previous clinical studies have focused entirely on the effects of hcy on large artery function, but recent studies have associated hyperhomocysteinaemia with small vessel complications, e.g. in dementia [20], and microangiopathy is described as part of the hereditary disorder of homocystinuria [21]. Accordingly, the aims of the present study were, firstly, to evaluate the effects of methionine-induced hyperhomocysteinaemia on cutaneous microvascular responses to iontophoretic administration of incremental doses of the endothelial-dependent vasodilator, acetylcholine (Ach), and the endothelial-independent vasodilator, sodium nitroprusside (SNP), and secondly to evaluate the effects of oral methionine loading on von Willibrand factor (vWF) levels as a biochemical marker of endothelial cell activation.

Methods

Fourteen healthy male volunteers participated in this randomised, double-blind, placebo-controlled crossover study to assess the microvascular, haemodynamic and biochemical effects of a single-dose of oral methionine 15 g. Each subject gave written informed consent prior to any study-related procedures, and the detailed protocol was approved by the Southern Derbyshire Research Ethics committee.

Clinical protocol

Following an initial screening visit, which included a clinical examination and routine blood tests, including fasting lipids and plasma hcy concentration, each subject attended two 9 h study days in the clinical research centre, 1 week apart. On each occasion, following an overnight fast, subjects reported to the unit at 08.00 h. After a 20 min period of quiet supine rest, a venous blood sample was collected prior to baseline measurements of blood pressure (BP) and microvascular reactivity. Methionine 15 g, or matching placebo powder, dissolved in 100 ml of fresh orange juice, was then administered orally. Blinded sachets of methionine and placebo were supplied by SHS International (Liverpool, UK), and the placebo powder was matched for colour, volume and taste.

Measurements of BP and microvascular function, together with blood sampling for plasma methionine and hcy concentrations, were repeated at 3 h and 8 h after methionine administration to coincide with the anticipated peaks in plasma methionine and hcy concentrations, respectively. Although subjects were ambulant in the clinical research centre, each set of measurements was recorded in a quiet, temperature controlled room (25 °C) after 30 min supine rest.

Laser Doppler fluximetry and iontophoresis

At baseline, 3 and 8 h, cutaneous microvascular blood flow was measured on the forearm using laser Doppler fluximetry (LDF, DRT4 system, Moor Instruments, UK) combined with local iontophoretic administration of incremental doses of Ach and SNP, as descibed previously [22]. In brief, after cleaning the skin surface, two laser-emiting probes (780 nm) with platinum electrodes that record temperature and flux were applied to the forearm, avoiding any underlying veins, with each probe surrounded by an iontophoresis chamber containing 1% Ach (anode) and 1% SNP (cathode) dissolved in purified water.

Incremental iontophoretic currents were applied, each for 10 s, with 7 min of LDF recording after each current stimulus to evaluate the microvascular vasodilator responses to ACh and SNP. The 10 s iontophoretic currents were administered as follows: 8, 16, 32, 64 and 128 µA. Thus, iontophoretic charges (current × time) were, respectively, 80, 160, 320, 640 and 1280 µC. The LDF recordings for time-dependent responses to each iontophoretic dose of Ach and SNP were analysed using dedicated software within the DRT4 unit to derive peak vasodilator responses to each agonist for individual subjects at each timepoint.

Blood pressure and heart rate

Supine BP and heart rate were recorded in duplicate at baseline, 3 and 8 h (prior to the microvascular measurements) using a semiautomatic sphygmomanometer (Dinamap, Criticon, UK).

Blood sampling and laboratory measurements

At baseline, 3 and 8 h, venous blood samples were collected for measurement of plasma concentrations of methionine and total hcy. Samples were centrifuged immediately and stored at −80 °C until assay. Hcy levels were measured using high-performance liquid chromatography, as described by our group previously [23], and methionine was measured by an amino acid analyser [23]. At baseline and 8 h, additional blood was collected for measurement of vWF levels according to the method of Short et al. [24]. All samples were analysed in duplicate and samples from each subject were analysed within the same batch.

Statistical analysis

Peak microvascular vasodilator responses to Ach and SNP (in arbitrary units) for individual subjects at each of the six timepoints were derived using the DRT4 software (Moor Instruments, Exeter, UK) that provides an automated dose–response analysis of the LDF profiles. The microvascular, haemodynamic and biochemical parameters for individual subjects at each timepoint were compared between the two study days using one-way analysis of variance (anova, STATISTICA software package).

Results

Fourteen healthy subjects participated in the study: mean age 27 ± 4.9 years, BP 120/75 ± 6/8 mmHg, BMI 24.1 ± 2.0 kg m⁻². The mean plasma hcy concentration at baseline was 9.0 µmol l⁻¹ (range 6–12 µmol l⁻¹). As expected, oral administration of methionine produced a marked increase in plasma methionine and hcy concentrations, with methionine levels higher at 3 h and hcy levels highest at 8 h (Figure 1). The acute increase in plasma hcy concentration was accompanied by a significant increase in pulse pressure at 8 h (P < 0.05), with no change in HR (Figure 1). Mean systolic and diastolic BP values at baseline were 117/72 ± 8/9 and 119/68 ± 8/7 mmHg on methionine and placebo days, respectively, and correspondingly 120/67 ± 8/6 vs 118/69 ± 10/7 mmHg at 8 h.

Figure 1.

Serum concentrations of methionine and hcy, and heart rate and pulse pressure (measured semiautomatically) at baseline, 3 h and 8 h after oral methionine administration (mean ±s.e. mean). *P < 0.0001.

There was no evidence of systemic endothelial cell activation in response to methionine loading. vWF levels at 8 h were no different between the two study days: e.g. 0.99±0.3 vs 1.02±0.3 U ml⁻¹.

Microvascular LDF responses

Cutaneous microvascular responses for an individual subject are illustrated in Figure 2a. Oral methionine had no significant effect on peak vasodilator responses to Ach and SNP, either at 3 h (when methionine is maximal) or 8 h (coinciding with the peak in hcy levels) (Figure 2b). At 8 h, the 95% confidence intervals for the differences in maximal Ach and SNP responses are −18.1% to + 23.8% and −22.5% to + 16.4%, respectively.

Figure 2.

a) Cutaneous microvascular vasodilator responses to Ach (upper panel) and SNP (lower panel) in a representative subject using iontophoretic currents, eAch applied for 10 s, delivering 80, 160, 320, 640 and 1280 µC. Software built into the LDF system provides automated analysis of the cumulative dose response for eAch subject. b) Vasodilator dose–response curves for iontophoretic administration of Ach (upper panel) and SNP (lower panel) at baseline, 3 h and 8 h after methionine (open bars) or placebo (black bars) (n = 14).

Discussion

Endothelial dysfunction affecting larger conduit arteries and resistance vessels has been described among patients with chronic elevations of homocysteine [25], and in healthy subjects following a single dose of methionine [12–14], but the effects of acute hyperhomocysteinaemia on small vessel vasodilator function in vivo have not been previously reported. Ach-induced (endothelial-dependent) vasodilation in resistance arteries is mediated almost entirely by local release of nitric oxide (NO), whereas in subcutaneous microvessels non-NO pathways predominate and Ach-induced vasodilation is thought to be mediated by vasodilator prostanoids and/or release of EDHF (endothelium-derived hyperpolarizing factor) [26, 27]. This study would suggest that acute increases in homocysteine have no effect on small vessel vasodilator responses to incremental iontophoretic administration of Ach and SNP, which would be consistent with the notion that homocysteine selectively impairs the release of NO (e.g. in flow-mediated dilatation [13]) but spares other pathways of endothelial function [16]. The present study also confirms that hyperhomocysteinaemia has no effect on smooth muscle (endothelial-independent) vasodilator responses to exogenous nitrate, e.g. SNP. Overall, the observations are consistent with clinical findings that in patients with homocystinuria macrovascular complications are more common than microvascular dysfunction [21].

Refinements to the equipment and protocols for LDF, and the development of new systems such as the DRT4 that interlock cutaneous probes with iontophoresis chambers, have improved the sensitivity and reproducibility of clinical studies investigating microvascular responses in vivo [27, 28]. The technique is painless, and in the present study use of a crossover design with automated software analysis of the vasodilator responses, and low iontophoretic currents, resulted in improved sensitivity (with narrower confidence intervals) for detecting possible effects of homocysteine on small vessel function. In particular, this protocol used much lower and shorter iontophoretic currents in the light of previous concerns that higher electrical charge can induce nonspecific vasodilation. We have previously used a similar protocol to demonstrate significant changes in Ach-induced vasodilation in women with pre-eclampsia [28].

Hyperhomocysteinaemia is associated with increased blood pressure, in particular isolated systolic hypertension [29], and it is interesting that in this double-blind study using a semiautomatic BP-measuring device oral methionine produced a modest but statistically significant increase in pulse pressure at the 8 h timepoint in normotensive healthy volunteers. The interpretation of this result is not entirely clear, especially since acute hyperhomocysteinaemia is reported to have no effect on arterial stiffness in eight subjects [30], but increased pulse pressure is of particular prognostic importance and neutral effects of methionine on mean arterial pressure, as reported in previous studies [14, 30], may conceal divergent changes in systolic and diastolic BP. For example, in a similar number of healthy volunteers (n = 20), Nappo et al. [12] reported a nonsignificant increase in systolic BP (105–108 mmHg) accompanied by a decrease in diastolic BP (71–69 mmHg) 4 h after oral methionine. Thus, the haemodynamic effects of hyperhomocysteinaemia, even after acute methionine loading, merit further investigation, and the increase in pulse pressure observed in the present study would be consistent with an effect of homocysteine on large vessel endothelial function.

Pulse pressure provides a surrogate measure of arterial stiffness, and is an independent predictor of cardiovascular mortality, even in normotensive subjects [31]. Since functional as well as structural changes affect pulse pressure and arterial stiffness, the most likely explanation for the result of the present study is that hcy, via reduced endothelial NO release, affects large vessel compliance. This, of course, requires cautious interpretation since there may be other explanations for the acute increase in pulse pressure.

A second objective of the present study was to evaluate the effects of acute hyperhomocysteinaemia on circulating levels of vWF antigen, a well established biochemical marker of endothelial cell activation that has been associated with thrombotic and microvascular complications as well as endothelial dysfunction [32, 33]. It is well established that vWF levels respond to a range of acute phase stimuli associated with endothelial dysfunction [34], but very little is known about the effect of elevated homocysteine levels on circulating markers of vascular cell activation [35]. There were no significant changes in vWF levels in the present study, which suggests that hyperhomocysteinaemia, at least acutely, does not cause widespread endothelial cell activation, although in a similar study methionine loading enhanced release of tissue plasminogen activator [36].

In summary, acute hyperhomocysteinaemia adversely affects pulse pressure, a marker of large artery compliance, but has no effect on microvascular vasodilation. Since microvascular responses to Ach are not mediated by endothelial release of NO, this study provides indirect evidence that hcy impairs endothelial function in large vessels via selective effects on NO pathways. Nevertheless, the clinical association between hcy and microangiopathy remains [21], e.g. in dementia [20] but the underlying mechanism is unclear.