Journal of Human Hypertension (1999) 13, 781–785  1999 Stockton Press. All rights reserved 0950-9240/99 $15.00 http://www.stockton-press.co.uk/jhh ORIGINAL ARTICLE The effect of moxonidine on plasma lipid profile and on LDL subclass distribution

MS Elisaf, C Petris, E Bairaktari, S-A Karabina, C Tzallas, A Tselepis and KC Siamopoulos Department of Internal Medicine, University of Ioannina Medical School, Greece

Moxonidine is a new antihypertensive agent whose cant decrease in both systolic and diastolic blood ,mechanism of action appears to involve specific stimu- pressure (from 147 ؎ 10 to 131 ؎ 11 mm Hg, P Ͻ 0.001 ,lation of imidazoline receptors resulting in an inhibition and from 98 ؎ 4.5 to 86 ؎ 5 mm Hg, P Ͻ 0.001 of the activity of the central and peripheral sympathetic respectively). No significant change in plasma lipid pro- nervous system. The drug seems to behave neutrally file was observed after moxonidine administration. with respect to plasma lipid parameters. However, there Additionally, no change in the susceptibility of LDL sub- are no data on the effects of moxonidine on the low- classes to copper-induced oxidative modification was density lipoprotein (LDL) subclass pattern or on the LDL noticed. Finally, drug therapy was not followed by any oxidation susceptibility, both of which are known to play change in either LDL phenotype or in mass and compo- a prominent role in the pathogenesis of atherosclerosis. sition of the three LDL subfractions. We conclude, that Thus, we undertook the present study to examine the unlike other antihypertensive drugs, such as beta-block- influence of moxonidine on the LDL subspecies profile ers which may predispose to expression of a relatively and their susceptibility to copper-induced oxidative atherogenic lipoprotein subclass pattern, moxonidine modification in 20 hypertensive patients (11 men, 9 does not affect either plasma lipid parameters or lipo- women) aged 38–61 years. Moxonidine administered at protein composition. a dose of 0.4 mg daily for 8 weeks produced a signifi-

Keywords: moxonidine; plasma lipid profile; LDL oxidation susceptibility; LDL subclass pattern; sympathetic nervous system

Introduction pheral vascular resistance. Moreover, the reduced sympathetic drive results in lower concentrations of It is well known that some antihypertensive agents, catecholamines and renin.10,11 Randomised com- especially diuretics in high dosage and beta-block- parative studies have shown that the efficacy of ers, have unfavourable effects on plasma lipids, moxonidine as monotherapy is similar to that of including elevation of triglycerides and total choles- other antihypertensive drugs.12 In clinical studies terol and reduction of high-density lipoprotein chol- moxonidine has been found to behave neutrally 1–4 esterol (HDL-C). Additionally, it has been shown with respect to plasma lipid parameters.12–14 How- that beta-blocker use may predispose to expression ever, there are no data on the effects of moxonidine of a relatively atherogenic lipoprotein subclass pro- on the LDL subclass pattern or on the LDL oxidation file, as it is associated with a predominance of susceptibility, both of which are known to play a smaller, denser, low-density lipoprotein (LDL) par- 5,6 prominent role in the pathogenesis of atheroscler- ticles and less HDL mass. Since an overview of all osis. Thus, we undertook the present study to exam- hypertension trials has shown that antihypertensive ine the influence of moxonidine on the LDL sub- treatment does lead to a less than expected species profile and their susceptibility to copper- reduction in coronary disease (CHD) event induced oxidative modification in patients with 7,8 rates, it is tempting to suggest that the adverse idiopathic hypertension. effects of the study drugs on lipid metabolism may have offset the potential benefit of Materials and methods (BP) reduction. The selective imidazoline-I1-recep- tor , a recently introduced class of antihy- Twenty patients (11 male, 9 female) aged 38–61 pertensive agents, offers an innovative therapeutic years with mild/moderate arterial hypertension approach to the treatment of hypertension.9 Moxoni- were studied. No patient had clinical or biochemical dine, the first drug of this class, stimulates imidazo- evidence of diabetes mellitus, thyroid, hepatic, or line I receptors in the medulla, thereby reducing renal disease. Additionally, none of them was taking central sympathetic drive and attenuating peri- any lipid lowering drugs or any other medication known to affect lipid metabolism, including hor- monal therapy. Individuals known to ingest more Correspondence: Dr Moses S Elisaf, Associate Professor of Medi- cine, Department of Internal Medicine, University of Ioannina, than two alcoholic drinks or to take vitamin sup- Medical School, GR 451 10 Ioannina, Greece plements were excluded from the study. To avoid Received 10 January 1999; accepted 25 February 1999 any possible effects of diet on the susceptibility of Effect of moxonidine on LDL subclass distribution MS Elisaf et al 782 the LDL subfractions to oxidation, all subjects were dation is divided into three consecutive phases, lag advised to avoid altering their dietary habits phase, propagation phase and decomposition phase. throughout the study. The lag time, the maximal rate of conjugated dienes After a wash-out period of at least 4 weeks for the formation, and the total amount of dienes formed patients on antihypertensive drugs, moxonidine was were calculated as previously described.16 The rela- given at a dose of 0.4 mg/day for 8 weeks. Before tive electrophoretic mobility (REM) of the oxidised and after treatment BP was measured by a calibrated subfractions was determined by agarose gel electro- sphygmomanometer as triplicate measurements phoresis on agarose gels [Lipo + Lp(a), Sebia]. after the patients had been sitting for 5 min. In addition to BP measurements, pulse rate was Analytical methods recorded and venous blood was obtained after a 14 h overnight fast for the determination of plasma Total plasma cholesterol and triglyceride levels lipid parameters. were measured by enzymatic colorimetric assays Subfractionation of LDL with density gradient using an RA 1000 analyser (Technicon Instruments, ultracentrifugation was performed and the suscepti- NY, USA). HDL-cholesterol levels were also deter- bility of LDL subclasses to copper-induced oxidative mined by the same method after precipitation of modification was tested. apoB-containing lipoproteins with magnesium-dex- tran sulphate. LDL-cholesterol levels were calcu- lated using the Friedewald formula.17 Plasma lipo- Subfractionation of low-density lipoprotein protein (a) [Lp(a)] levels were measured in duplicate Venous blood samples were collected from over- using a monoclonal anti-Lp(a) antibody technique night fasting donors in tubes containing 0.001% by the enzyme immunoassay Macra Lp(a) (Terumo Na2EDTA. Immediately after collection of plasma, Medical Corporation Diagnostic Division, Elkton, 1.3 mM EDTA and 50 ␮g/ml of gentamicin were MD, USA). The lower limit of detectability was added and plasma was stored at 4°C in order to pre- 0.8 mg/dL. Plasma apoAl and apoB were measured vent metal cation catalysed lipoprotein oxidation by immunonephelometry with the aid of a Beckman and microbial growth. Within 24 h from its collec- array analyser (Beckman Instruments, CA, USA). tion, plasma was submitted to density gradient ultra- The cholesterol, phospholipid and triglyceride con- centrifugation in a Beckman ultracentrifuge at tent of each lipoprotein subfraction was determined 40 000 rpm, 14°C for 24 h, using a Beckman SW41 enzymatically using the BioMerieux kits. The pro- Ti rotor as previously described.15 Construction of a tein content of the gradient fractions and lipoprotein discontinuous density gradient at ambient tempera- subfractions was determined by the BCA method. ture was initiated by pumping 3 ml of plasma into Lipoprotein mass in each subfraction was calculated the bottom of the tube. The density of the plasma as the sum of the concentrations of the individual was raised to 1.10 g/ml by dissolution of 0.42 g KBr. components (cholesterol, triglyceride, phospholipid The plasma was then successively overlayered by and protein) and allowed the determination of the four solutions of decreasing density (2 ml of d = percent chemical composition.16 1.065 g/ml, 3 ml of d = 1.035 g/ml, 3 ml of d = 1.019 g/ml and 1.0 ml of d = 1.006 g/ml). After ultra- Statistical analysis centrifugation, 30 fractions of 400 ␮l were collected by successive aspiration with a precision pipette Data were expressed as mean ± s.d., except for Lp(a), from the meniscus downwards. All fractions were which was expressed in terms of median and range. analysed for their protein content. Subsequently, Statistical analysis was performed using paired or equal volumes of specific gradient fractions corre- unpaired Student’s t-test or Wilcoxon signed-rank sponding to d = 1.030–1.034 g/ml, d = 1.034– test. A P value of less than 0.05 was considered to 1.041 g/ml and d = 1.041–1.048 g/ml were pooled to be significant. constitute the LDL1, LDL2 and LDL3 subfractions, respectively. Results Moxonidine produced a significant decrease in both Oxidation of low-density lipoprotein subfractions systolic and diastolic BP (from 147 ± 10 to 131 ± 11 Before oxidation, LDL subfractions were dialysed to mm Hg, P Ͻ 0.001, and from 98 ± 4.5 to 86 ± 5 remove EDTA against two changes of a 200-fold vol- mm Hg, P Ͻ 0.001, respectively) while no significant ume of 10 mM phosphate buffered saline solution change in heart rate was noticed. As shown in Table (PBS), pH 7.4 for 24 h at 4°C in the dark. From the 1, drug administration was not followed by any dialysed LDL subfractions, a volume of 1 ml con- significant change in plasma lipid parameters. taining 100 ␮g protein/ml PBS was oxidised in the presence of CuSO ,5␮M final concentration. The 4 Mass profile and chemical composition of the LDL kinetics of the oxidation were determined by moni- subfractions toring the increase in the 234 nm absorbance on a Perkin-Elmer L15 spectrophotometer, every 10 min The total LDL mass (calculated as the sum of the for 3 h. The increase in absorbance is due to the for- mass of the 3 LDL subfractions) was 225.6 ± mation of conjugated fatty acid hydroperoxides dur- 75.3 mg/dL and was not significantly altered after ing the peroxidation of the polyunsaturated fatty moxonidine treatment, 204.7 ± 56.3 mg/dL. The acid content of LDL. The sigmoidal curve of the oxi- major LDL subfraction in our patients was LDL2 hav- Effect of moxonidine on LDL subclass distribution MS Elisaf et al 783 Table 1 Effect of moxonidine on plasma lipid parameters

Parameters Before treatment After treatment P (mg/dl)

T CHOL 181 ± 27 175 ± 30 NS HDL CHOL 49 ± 10 49 ± 11 NS LDL CHOL 98 ± 28 92 ± 25 NS TRG 131 ± 42 134 ± 44 NS ± ± Apo A1 157 24 159 21 NS Apo B 92 ± 19 90 ± 24 NS Lp(a) 7 (0.8–27) 7.5 (0.8–26) NS

TCHOL: total cholesterol, HDL CHOL: HDL cholesterol, LDL CHOL: LDL cholesterol, TRG: triglycerides, Apo: apolipoprotein, Lp(a): lipoprotein (a). ing a mass of 112.2 ± 50.6 mg/dL, significantly Ͻ higher compared to LDL1 and LDL3 (P 0.001 for both comparisons). The mass of LDL1 was also sig- Ͻ nificantly higher compared to LDL3 (P 0.001) (Table 2). Moxonidine administration did not alter the mass of each LDL subfraction (Table 2) and Figure 1 Distribution of LDL mass between density gradient consequently, the mass distribution profile was not subfractions before and after moxonidine treatment. The % of significantly influenced (Figure 1). Additionally, as total mass of LDL in individual gradient subfractions is plotted shown in Table 2, moxonidine did not alter the on the ordinate against each LDL subfraction on the abscissa. Data are expressed as the means ± s.d. The mass of each chemical weight % content of cholesterol, triglycerides, phos- component in individual LDL gradient subfractions was deter- pholipids and protein of each LDL subfraction. mined by chemical analysis; the total mass of each subfraction was then calculated as the sum of the mass of each component. Oxidation profile of LDL subfractions The LDL subfractions were oxidised in the presence measured was significantly altered after treatment of 5 ␮MCu++ under continuous monitoring of the (Table 3) absorbance at 234 nm. As shown in Table 3, the LDL1 subfraction either before or after treatment was Discussion more resistant to oxidation, since it had a greater lag time, lower rate of oxidation, lower total amount of Because of the frequent association of dyslipidaemia dienes and REM values compared with the other with hypertension, plasma lipid parameters should two subfractions (P Ͻ 0.05 for comparisons of all be measured in hypertensive patients. In subjects parameters with those observed in LDL2 or LDL3 who display any lipid abnormalities, subfraction, either before or after treatment with diuretics and beta-blockers may not be an appropri- moxonidine). As it was expected, the subfraction ate first-step treatment, since they appear to alter the more susceptible to oxidation was the dense LDL3 lipid profile unfavourably, at least in short-term either before or after moxonidine treatment. It is studies.1–4 Additionally, beta-blocker use is associa- important to note that drug treatment did not influ- ted with a predominance of smaller, denser LDL par- ence the susceptibility to oxidation of each LDL ticles and less HDL mass,5,6,18 lipoprotein changes subfraction, since none of the oxidation parameters that might be expected to increase coronary artery

Table 2 Mean weight % chemical composition and lipoprotein mass of LDL subfractions before and after moxonidine treatment

LDL subfraction

123 d 1.030–1.034 g/ml d 1.034–1.041 g/ml d 1.041–1.048 g/m

Before After Before After Before After moxonidine moxonidine moxonidine moxonidine moxonidine moxonidine

Component, % Cholesterol 31.8 ± 9.6 36.5 ± 5.2 41.8 ± 12.2 39.8 ± 13.9 40.1 ± 15.2 43.4 ± 9.5 Triglyceride 16.9 ± 7.1 15.9 ± 8.1 7.9 ± 3.3 9.0 ± 5.2 10.6 ± 5.9 10.2 ± 4.8 Phospholipid 25.8 ± 7.7 27.6 ± 7.9 27.8 ± 2.9 28.0 ± 4.5 25.8 ± 7.7 27.6 ± 7.9 Protein 29.0 ± 8.4 24.8 ± 3.8 29.7 ± 11.5 28.8 ± 9.5 36.8 ± 10.9 37.7 ± 14.1 Lipoprotein mass, mg/dl plasma 70.9 ± 22.4 66.6 ± 16.1 112.2 ± 50.6 102.4 ± 58.4 42.6 ± 15.5 35.5 ± 14.0

Values are means ± s.d. of duplicate determinations of each component. Analyses were performed as described in Materials and methods. Effect of moxonidine on LDL subclass distribution MS Elisaf et al 784 Table 3 Oxidation parameters of LDL subfractions before and after moxonidine treatment

LDL subfraction

123 d 1.030–1.034 g/ml d 1.034–1.041 g/ml d 1.041–1.048 g/ml

Before After Before After Before After moxonidine moxonidine moxonidine moxonidine moxonidine moxonidine

Lag time, min 72.2 ± 9.4 75.0 ± 18.4 58.3 ± 13.0 56.4 ± 12.2 51.0 ± 8.0 50.6 ± 12.4 Rate of oxidation, nmol/mg 4.2 ± 0.4 4.1 ± 0.5 4.9 ± 0.6 4.9 ± 0.6 5.5 ± 0.3 5.3 ± 0.4 protein min−1 Total amount of dienes, 296.7 ± 25.0 276.7 ± 47.9 334.3 ± 59.6 326.8 ± 58.8 353.7 ± 60.5 336.5 ± 60.1 nmol/mg protein REM 2.7 ± 1.0 2.6 ± 1.2 3.2 ± 1.1 2.9 ± 1.2 3.4 ± 1.3 3.3 ± 1.5

Oxidation was performed by incubating of 100 ␮g/ml of LDL protein with Cu++,5␮M final concentration at 37°C. The kinetics of oxidation were determined by monitoring the increase in absorbance at 234 nm every 10 min for 3 h. Values represent the means ± s.d.

disease risk19 and offset the beneficial effects of anti- during copper-induced oxidation,28,31 suggesting hypertensive therapy on cardiovascular morbidity that calcium antagonists may potentially exert anti- and mortality.7,8 It has recently been suggested that atherosclerotic properties via a reduction of the oxi- moxonidine seems to be a logical choice for hyper- dative modification of LDL. tensive patients with coexistent intolerance We conclude that unlike other antihypertensive or dyslipidaemia,20 as in clinical studies moxonid- drugs moxonidine does not affect plasma lipid para- ine has been proved to have neutral or even ben- meters, LDL subfraction distribution, or LDL sub- eficial effects on lipid and carbohydrate metab- classes composition and oxidability. olism.12–14 In this setting, we have shown for the first time that the drug does not affect plasma lipid para- meters, LDL subfraction distribution, or LDL sub- References classes composition and oxidability. 1 Lardinois CK, Neuman SL. The effects of antihyperten- The neutral effect of moxonidine on plasma lipid sive agents on serum lipids and lipoproteins. Arch parameters is in agreement with the findings pre- Intern Med 1988; 148: 1280–1288. viously reported with other centrally acting drugs 2 Ames RP. The effects of antihypertensive drugs or and especially with , which has been serum lipids and lipoproteins. 1. Diuretics. Drugs found to minimally influence lipid levels.21,22 1986; 32: 260–278. The absence of a significant change in plasma trig- 3 Weinberger MH. Mechanisms of diuretic effects on lycerides, unlike the significant increase in plasma carbohydrate tolerance, insulin sensitivity and levels. triglycerides observed after beta-blockage use could Eur Heart J 1992; 13 (Suppl G): 5–9. partly explain the lack of a significant change in LDL 4 Ames RP. A comparison of blood lipid and blood pressure responses during the treatment of systemic pattern, taking into account that plasma triglycer- hypertension with indapamide and with . Am ides are emerging as the most important determinant J Cardiol 1996; 77: 12B–16B. 18,23 of the LDL subfraction profile. 5 Superko HR, Haskell WL, Krauss RM. Association of It has recently been reported that LDL isolated lipoprotein subclass distribution with use of selective from patients with exhibits and non-selective beta-blocker medications in patients increased propensity for oxidation.24 The increased with coronary heart disease. Atherosclerosis 1993; 101: LDL oxidability in hypertensive patients may be 1–8. related to angiotensin II, which has been shown to 6 Superko HR, Wood PD, Krauss RM. Effect of alpha- have a stimulatory effect on copper-induced oxi- and selective beta-blockade for hypertension control dation of LDL, as well as on LDL degradation by on plasma lipoproteins, apoproteins, lipoprotein sub- 24 classes and postprandial lipemia. Am J Med 1989; 86 macrophages. Furthermore, the increased suscepti- (Suppl 1B): 26–29. bility of hypertensive patients’ LDLs to lipid peroxi- 7 Collins R et al. Blood pressure, stroke, and coronary dation decreased significantly after treatment with heart disease. Part 2. Short-term reductions in blood angiotensin-converting enzyme inhibitors, which pressure: overview of randomized drug trials in their may act through the suppression of local production epidemiological context. Lancet 1990; 335: 827–838. of angiotensin II in the arterial wall.24 However, 8 MacMahon S et al. Blood pressure, stroke, and coron- even though moxonidine could lower serum con- ary heart disease. Part 1. Prolonged differences in centrations of catecholamines and renin,25–27 no sig- blood pressure: prospective, observational studies cor- nificant change in the susceptibility of LDL sub- rected for the regression dilution bias. Lancet 1990; classes to copper-induced oxidative modification 335: 765–774. 9 Head GA. Importance of imidazoline receptors in the was found. Similarly, atenolol, which can also cardiovascular actions of centrally acting antihyper- reduce plasma renin activity had no effect on the tensive agents. Ann NY Acad Sci 1995; 763: 531–540. 28–30 oxidation resistance of LDL, while nifedipine 10 Prichard BNC, Graham BR. The use of moxonidine in and other calcium channel blockers displayed a sig- the treatment of hypertension. J Hypertens 1997; 15 nificant dose-dependent capacity to protect LDL (Suppl 1): S47–S55. Effect of moxonidine on LDL subclass distribution MS Elisaf et al 785 11 Morris STW, Reid JL. Moxonidine: a review. J Hum lipid subfractions, and apolipoproteins in mild hyper- Hypertens 1997; 11: 629–635. tension. Am Heart J 1990; 120: 172–179. 12 Chrisp P, Faulds D. Moxonidine. A review of its phar- 23 Griffin BA et al. Role of plasma triglyceride in the macology, and therapeutic use in essential hyperten- regulation of plasma low density lipoprotein (LDL) sion. Drugs 1992; 44: 993–1012. subfractions: relative contribution of small, dense LDL 13 Kaan EC et al. Effects of and moxonidine on to coronary heart disease risk. Atherosclerosis 1994; glucose metabolism: an integrated approach towards 106: 241–254. pathophysiological mechanisms in cardiovascular 24 Keidar S et al. Low density lipoprotein isolated from metabolic disorders. Cardiovasc Risk Factors 1995; 5 patients with essential hypertension exhibits (Suppl 1): 19–27. increased propensity for oxidation and enhanced 14 Michel MG, Ernsberger P. Keeping an eye on the I site: uptake by macrophages: a possible role for angiotensin imidazoline-preferring receptors. Trends Pharmacol II. Atherosclerosis 1994; 107: 71–84. Sci 1992; 13: 369–370. 25 Kirch H, Hutt H-J, Pla¨nitz V. Pharmacodynamic action 15 De Graaf J et al. Enhanced susceptibility to in vitro oxi- and of moxonidine after single oral dation of the dense low-density lipoprotein subfrac- administration in hypertensive patients. J Clin Pharm- tion in healthy subjects. Arterioscler Thromb 1991; 11: acol 1990; 30: 1088–1095. 298–306. 26 Mitrovic V, Patyna W, Hu¨ ting J, Schlepper M. Hemo- 16 Tselepis AD et al. PAF-degrading acetylhydrolase is dynamic and neurohumoral effects of moxonidine in patients with essential hypertension. Cardiovasc preferentially associated with dense LDL and VHDL1 in human plasma. Catalytic characteristics and Drugs Ther 1991; 5: 967–972. relation to the monocyte-derived enzyme. Arterioscler 27 Mitrovic V et al. Haemodynamic and neurohormonal effects of a single oral dose of the imidazoline I recep- Thromb Vasc Biol 1995; 15: 1764–1773. tor moxonidine in patients with idiopathic 17 Friedewald, WT, Levy RI, Fredrickson DS. Estimation dilated cardiomyopathy and . Eur J Clin of the concentration of low-density lipoprotein choles- Res 1996; 8: 149–161. terol in plasma, without use of the preparative ultra- 28 Lesnik P et al. Impact of a combination of a calcium centrifuge. Clin Chem 1972; 18: 499–502. antagonist and a ␤-blocker on cell- and copper- 18 Campos H et al. Low density lipoprotein particle size mediated oxidation of LDL and on the accumulation and . Arterioscler Thromb and efflux of cholesterol in human macrophages and 1992; 12: 187–195. murine J774 cells. Arterioscler Thromb Vasc Biol 1997; 19 Austin MA, King MC, Vranizan KM, Krauss RM. 17: 979–988. Atherogenic lipoprotein phenotype. A proposed gen- 29 Jenkins RR, Del-Signore CM, Sauer P, Skelly C. The etic marker for coronary heart disease risk. Circulation effect of beta blocking drugs on lipid peroxidation in 1990; 82: 495–506. rat heart in vitro. Lipids 1992; 27: 539–542. 20 Krentz AJ, Evans AJ. Selective 30 Yue TL et al. Carvedilol, a new antihypertensive agent, agonists for metabolic syndrome. Lancet 1998; 351: prevents lipid peroxidation and oxidative injury to 152–153. endothelial cells. Hypertension 1993; 22: 922–928. 21 Lardinois CK, Johns JP. Cardiac drug effects on lipids. 31 Lupo E, Locher R, Weisser B, Vetter W. In vitro antioxi- Curr Opin Lipidol 1992; 5: 274–278. dant activity of calcium antagonists against LDL oxi- 22 Houston MC et al. The effects of clonidine hydrochlor- dation compared with a-tocopherol. Biochem Biophys ide versus atenolol monotherapy on serum lipids, Res Com 1994; 203: 1803–1808.