Venous response to nitroglycerin isenhanced in young, healthy carriers of the825T allele of the G protein �3 subunitgene (GNB3)

Objective: The ultimate mode of action by which nitroglycerin elicits vasodilation remains elusive. Animalstudies point to an involvement of pertussis toxin–sensitive G proteins. The 825T allele of the GNB3 C825Tpolymorphism in the gene encoding the G protein �3 subunit is associated with enhanced signal transductionvia pertussis toxin–sensitive pathways in vitro. We hypothesized that G proteins have a role in nitroglycerin-mediated vasodilation and that carriers of the 825T allele would exhibit a stronger response to nitroglycerin.Methods: We used the linear variable transducer technique to compare dorsal hand vein compliance in 28young, healthy men with and without the T allele (n � 15 CC, n � 8 CT, and n � 5 TT). After individualdose-response curves to phenylephrine had been established, veins were preconstricted to 70% of the maximalphenylephrine-induced constriction. Nitroglycerin was then infused in ascending doses (0.02-2000 ng/min),and the vasodilatory response was measured.Results: The vasodilatory response to nitroglycerin was significantly greater in carriers of the 825T allele. Themaximal response to nitroglycerin was 102% � 6% venodilation in the CT/TT group and 78% � 5% in the CCcontrol group (P � .0045) (mean difference, �24% � 8%; 95% confidence interval, 8%-40%). Comparison ofthe nitroglycerin dose-response curves by ANOVA confirmed an enhanced nitroglycerin-induced venodila-tion in 825T-allele carriers (P < .0001).Conclusion: Our results suggest that the GNB3 C825T polymorphism determines venous response to nitro-glycerin and that G proteins may be involved in the signal transduction pathway. (Clin Pharmacol Ther2003;74:499-504.)

Anna Mitchell, MD, Sandra Buhrmann, BSc, Anne Seifert, Jens Nurnberger, MD,Rene R. Wenzel, MD, Winfried Siffert, MD, Thomas Philipp, MD, and Rafael F.Schafers, MD Essen, Germany

Nitrates have been the backbone of antianginousdrug therapy for more than 100 years.1 They presum-ably have a direct relaxant effect on vascular smoothmuscle: Their dilation of peripheral veins and, at higherdoses, peripheral arteries reduces cardiac preload and

afterload, thereby lowering myocardial oxygen con-sumption. Whereas organic nitrate esters have an effecton all types of vessels, veins are most sensitive.2 Theexact mechanism of action is still not known, andpatients’ response to nitrates is notoriously variable.3

Circumstantial evidence from the literature suggeststhat the action of nitroglycerin could be mediated, atleast in part, by heterotrimeric G proteins. Studies onisolated bovine mesenteric arteries have shown that thedilating effect of nitroglycerin, but not that of isosor-bide dinitrate or isosorbide-5-mononitrate, is attenuatedby pretreatment of the vessels with pertussis toxin.4,5

As pertussis toxin prevents the interaction and activa-tion of heptahelical receptors with G proteins, espe-cially of the Gi/Go type, these findings would suggestthat the action of nitroglycerin is not mediated simplyby nitric oxide release but may constitute a more com-

From the Departments of Nephrology & Hypertension and Pharma-cology, Essen University.

This study was supported in part by the Deutsche Forschungsgemein-schaft (DFG, WE 1772/3-1 and 3-2) and the Oertel Foundation.

Received for publication June 9, 2003; accepted July 18, 2003.Reprint requests: Anna Mitchell, MD, Department of Nephrology &

Hypertension, Essen University Hospital, Hufelandstrasse 55, Es-sen, Germany.

E-mail: anna.mitchell@uni-essen.deCopyright © 2003 by the American Society for Clinical Pharmacol-

ogy & Therapeutics.0009-9236/2003/$30.00� 0doi:10.1016/S0009-9236(03)00230-3

499

plicated mechanism potentially involving the activationof either receptors or G proteins or both. G proteinsconsist of a family of �, �, and � subunits,6 the ��subunits forming a functional monomer. On activationof heptahelical membrane receptors, the � subunitsrelease guanosine diphosphate in exchange forguanosine triphosphate and both � and �� subunitsdissociate and can modulate a huge variety of targetssuch as ion channels, phospholipase C isoforms, andadenylyl cyclase isoforms, to name but a few.7 Hence itis expected that mutations or genetic polymorphisms,which ultimately affect the function of G proteins, musthave a significant effect on intracellular signal trans-duction. In 1998 a C825T polymorphism was describedin the gene GNB3, which encodes the �3 subunit ofheterotrimeric G proteins.8 In vitro the 825T allele isassociated with increased intracellular signal transduc-tion and increased cell responsiveness8-10 in pathwayslinked to pertussis toxin–sensitive G proteins.

Genetic host factors are increasingly being recog-nized as major determinants of drug responses. TheGNB3 C825T polymorphism is associated with hyper-tension and other risk factors for cardiovascular mor-bidity.8,11,12 Recent investigations have established thatGNB3 C825T allele status determines coronary vaso-constriction in response to an �2–adrenergic receptoragonist13 and vasoconstriction in skin microvessels oninjection of endothelin 1, angiotensin II, and norepi-nephrine.14 On the other hand, we were able to showthat the blood pressure decrease after administration ofclonidine15 and moxonidine (R. F. Schafers, MD, un-published data, Essen, Germany, 2003) is significantlyenhanced in 825T-allele carriers, and 825T-allele car-riers were reported to show an increased responsivenesstoward antidepressants.16,17 Taken collectively, theseresults point to a generally increased response to vaso-active drugs in carriers of the 825T allele.

We performed this study to further characterize themode of action of nitroglycerin in humans, especiallythe potential contribution of G protein activation and apossible involvement of the GNB3 C825T polymor-phism. We hypothesized that vasodilation to nitroglyc-erin would be stronger in 825T-allele carriers. An as-sociation of vasodilation with genotypes would suggestthat the action of this compound may directly or indi-rectly involve G protein activation.

We determined the vasodilatory effect of nitroglyc-erin in the dorsal hand vein in young, healthy malevolunteers genotyped for the GNB3 C825T polymor-phism. We report here for the first time a differentialeffectiveness of nitroglycerin in 825T-allele carrierscompared with homozygous 825C-allele carriers.

METHODSStudy population. Twenty-eight young, white male

volunteers (aged 20-36 years) participated in this study.All were healthy judged on the basis of medical history,physical examination, electrocardiogram, and routinelaboratory screening. All subjects were nonsmokersand drug-free. Body mass index had to be �25 kg/m2

for subjects to be included in the study.Each subject gave written informed consent before

taking part in the study, which had been approved bythe Essen University Medical School Ethics Commit-tee, Essen, Germany. The study was conducted in ac-cordance with the principles laid down in the Declara-tion of Helsinki.

Genotyping. Genotyping with respect to the GNB3C825T polymorphism was performed as described pre-viously.8 Subjects and investigators were blinded withregard to genotype at the GNB3 locus.

Drugs. The following drugs diluted in physiologicsodium chloride solution were used: phenylephrine(Neo-Synephrine; Abbott Laboratories, North Chicago,Ill) and nitroglycerin (Trinitrosan; Merck Kg, Darm-stadt, Germany).

Dorsal hand vein technique. The dorsal hand veintechnique was used as described previously.18,19 Inbrief, subjects reported to the laboratory at 8 AM after anovernight fast. They lay down on a bed in atemperature-controlled room (24 � 2°C). One arm wasplaced on a padded support sloping upward at an angleof 30° to ensure complete emptying of the dorsal handveins. A 23-gauge butterfly needle was inserted into asuitable dorsal hand vein, and saline solution was in-fused for at least 45 minutes at a constant rate of 0.3mL/min. A tripod holding a linear variable differentialtransducer (LVDT) (Venograph VEG 91; Boucke, Re-utlingen, Germany) was mounted onto the back of thehand over the vein that was undergoing evaluation 10mm downstream from the tip of the needle. The veindiameter was measured by recording the displacementof the freely movable core of the LVDT. Baselinerecordings of the position of the LVDT core during theinfusion of saline solution were made both before andafter inflation of a sphygmomanometer cuff on theipsilateral arm to 45 mm Hg. Dilation of the vein atbaseline after 45 to 60 minutes of saline infusion wasdefined as 100%. Given the low venous tone underthese conditions, venodilatory effects can be assessedonly in preconstricted vessels. We used phenylephrine,a selective �1-adrenergic receptor agonist, in ascendingdoses ranging from 3.2 to 10,000 ng/min to obtainpreconstriction of the hand veins. We established max-imal constriction and from there calculated the dose

CLINICAL PHARMACOLOGY & THERAPEUTICS500 Mitchell et al NOVEMBER 2003

needed to procure approximately 70% of maximal con-striction (ED70). Hand veins were allowed to return tomaximal dilation under infusion of saline solution. Sub-sequently, phenylephrine was infused at the rate of theED70 to achieve a stable vasoconstriction, which wasdefined as 0% dilation. Nitroglycerin was then infusedat increasing doses (0.02-2000 ng/min), and the localvenous vasodilatory response was measured. Venousrelaxation was defined as percentage of the baselinepreconstriction by the following formula: (A � Y)/(X � Y) � 100, in which A is the vein diameter withinfusion of nitroglycerin, X is baseline diameter (undersaline solution infusion), and Y is preconstricted veindiameter.

Statistical analysis. Data from in vitro experi-ments8,20 and indirect evidence from population stud-ies8,11,12,21 demonstrate that the presence of the T al-lele, whether homozygous or heterozygous, changescardiovascular and neurohumoral regulation relative tothe CC genotype. Therefore our primary statisticalcomparisons have determined the significance of dif-ferences between T-allele carriers and subjects with theCC genotype (ie, CT and TT subjects were pooled foranalyses). Results are expressed as mean � SEM andwere analyzed by unpaired, 2-tailed t test and 2-wayANOVA. Significance levels were set at P � .05.

RESULTSOf the 28 subjects, 13 were carriers of the 825T

allele (5 TT and 8 CT) and 15 individuals had thehomozygous CC genotype. Anthropometric variablesdid not differ by GNB3 genotype (Table I), nor didbaseline vein diameter (P .48, data not shown).

Infusion of phenylephrine induced dose-dependentvasoconstriction in all subjects with no significant dif-ference between genotypes (Fig 1). Phenylephrinedoses calculated from the complete dose-responsecurves to obtain approximately 70% venoconstrictionbefore infusion of nitroglycerin were 2.762 � 0.133ng/min for CC homozygous individuals and 2.697 �0.144 ng/min for T-allele carriers (data given as log

dose, P .74). The preconstriction achieved was 67%� 3% and 64% � 5% from baseline for the CC and theCT/TT group, respectively (P .59).

After preconstriction of dorsal hand veins to approx-imately 70% of their maximum phenylephrine-inducedconstriction, infusion of nitroglycerin produced a sig-nificantly greater vasodilation in carriers of the 825 Tallele: Maximum vasodilation in 825T-allele carrierswas 102% � 6% from baseline, whereas CC homozy-gous subjects showed a maximum vasodilation of 78%� 5% from baseline (mean difference, �24% � 8%;95% confidence interval, 8%-40%; P .0045 versusCC) (Fig 2, A). The dose-response curve of T-allelecarriers showed a marked shift to the left as comparedwith CC subjects (P � .0001) (Fig 2, B). The differencebetween genotypes in the effective dose needed toproduce 50% of maximal constriction (ED50) for nitro-glycerin was borderline significant (P .07), with a logED50 of 4.043 � 0.118 pg/min in subjects with the CCgenotype and 3.681 � 0.1478 pg/min in carriers of theT allele.

No adverse effects were observed after the infusionof either phenylephrine or nitroglycerin. Blood pressureand heart rate remained stable throughout, indicatingthat there were no systemic drug effects during thestudy period (P not significant, data not shown).

Table I. Study population: Anthropometricparameters

Age(y)

Height(cm)

Body mass index(kg/m2)

GenotypeCT/TT 25 � 1.2 185 � 2.0 22.5 � 0.4CC 27 � 0.7 183 � 1.6 22.9 � 0.5

P value .173 .358 .515

Data are shown as mean � SEM.

Fig 1. Dose-response curves for mean venoconstriction frombaseline (�SEM) during infusion of phenylephrine (P notsignificant versus CC).

CLINICAL PHARMACOLOGY & THERAPEUTICSVOLUME 74, NUMBER 5 Mitchell et al 501

DISCUSSIONThis in vivo study demonstrates for the first time that

the vasodilatory response to nitroglycerin is increasedin human subjects carrying the GNB3 825T allele.These findings imply that some of the effects of thiscompound may be G protein–mediated. A similar al-though not directly comparable effect has been reportedrecently22: The authors studied nitroglycerin-mediatedvasodilation of the brachial artery in diabetic patientsbefore and after an optimized glucose control protocol.Although nitroglycerin-induced vasodilation did notdiffer by genotype before intervention, 825T-allele car-riers showed a significantly enhanced response to thiscompound after intensive therapy.

Our findings are consistent with those obtained inbovine mesenteric arteries, in which the effect of nitro-glycerin was significantly blunted in specimens pre-treated with pertussis toxin.4 Nevertheless, the molec-ular mechanisms remain to be defined. It seemspossible that the molecule directly activates G protein–coupled receptors. Alternatively, the compound couldmediate the release of yet-unidentified substances fromendothelial or vascular smooth muscle cells, which inturn activate G protein–coupled receptors.

The question arises as to whether nitrates generallyexert their effect via G protein–coupled receptors or

whether G protein involvement is a specific feature ofnitroglycerin. In the studies on bovine mesenteric ar-teries, relaxation by isosorbide dinitrate and isosorbide-5-mononitrate was unimpaired by pertussis toxin, sug-gesting a different mode of action as compared withthat of nitroglycerin.4 Apparently, this may also applyfor humans. The action of isosorbide dinitrate in coro-nary arteries was independent of GNB3 genotype sta-tus.23 These obviously conflicting findings are furthercomplicated by the fact that the ultimate mode of actionof nitrate esters is very similar: Briefly, they elicitrelaxation of vascular smooth muscle cells via the re-lease of nitric oxide, which activates soluble guanylylcyclase, which consequently raises intracellular cyclic3',5'-guanosine monophosphate concentrations, whichin turn lead to a reduction in intracellular freeCa��.2,24-26 Since the studies mentioned investigatedarterial vasculature, the much greater sensitivity ofveins to nitrate treatment may account for some of thedifferences observed. A recent study found a significantassociation between GNB3 genotypes and therapy re-sponse in men with erectile dysfunction treated withsildenafil,27 which prolongs nitric oxide–mediated cy-clic 3',5'-guanosine monophosphate formation throughinhibition of phosphodiesterase type V. In this obser-vational study, significantly higher responder rates

Fig 2. A, Mean maximal venodilation from baseline during infusion of nitroglycerin (P .0024,versus CC). B, Dose-response curves for mean venodilation from baseline (�SEM) during infusionof nitroglycerin (P � .0001, versus CC).

CLINICAL PHARMACOLOGY & THERAPEUTICS502 Mitchell et al NOVEMBER 2003

were found in men with the TT genotype comparedwith those with the CT or CC genotype.

In summary, there is persuasive evidence that thevenous response to nitroglycerin is determined by theC825T carrier status. Whereas there are numerous stud-ies comparing patients’ responses to various short- andlong-acting nitrates and ethnic influences have beenreported for arterial vasodilation to sodium nitroprus-side,28 no data exist as yet regarding the contribution ofgenetic or ethnic dispositions on the efficacy of nitro-glycerin treatment. Because we did not compare theeffects of nitroglycerin with those of other nitrovaso-dilators, we cannot at this time conclude that the dif-ferences observed between genotypes are specific tonitroglycerin. Nevertheless, in view of the widespreaduse of nitroglycerin to curb angina pectoris and heartfailure, our results may have important clinical impli-cations. Thus further studies are needed to extend theseresults to patients with coronary artery disease. On thelevel of molecular research, future investigationsshould attempt to clarify the possible role of G proteinsin nitroglycerin-mediated vasodilation.

The authors declare that there are no financial or personal rela-tionships connected with the research described herein that wouldlead to a conflict of interest.

References1. Brunton TL. Use of nitrite of amyl in angina pectoris.

Lancet 1867;ii:97-8.2. Torfgard KE, Ahlner J. Mechanisms of action of nitrates.

Cardiovasc Drugs Ther 1994;8:701-17.3. Eichler HG, Hiremath A, Katzir D, Blaschke TF, Hoff-

man BB. Absence of age-related changes in venous re-sponsiveness to nitroglycerin in vivo in humans. ClinPharmacol Ther 1987;42:521-4.

4. Torfgard K, Ahlner J, Axelsson KL. Relaxation of bovinemesenteric arteries by glyceryl trinitrate and other nitro-compounds: evidence for partly different mechanisms ofaction. Pharmacol Toxicol 1990;67:216-21.

5. Ahlner J, Axelsson KL, Ekstram-Ljusegren M, FriedmanRL, Grundstrom N, Karlsson JO, et al. Relaxation ofbovine mesenteric artery induced by glyceryl trinitrate isattenuated by pertussis toxin. Pharmacol Toxicol 1988;62:155-8.

6. Downes GB, Gautam N. The G protein subunit genefamilies. Genomics 1999;62:544-52.

7. Farfel Z, Bourne HR, Iiri T. The expanding spectrum ofG protein diseases. N Engl J Med 1999;340:1012-20.

8. Siffert W, Rosskopf D, Siffert G, Busch S, Moritz A,Erbel R, et al. Association of a human G-protein beta3subunit variant with hypertension. Nat Genet 1998;18:45-8.

9. Virchow S, Ansorge N, Rosskopf D, Rubben H, SiffertW. The G protein beta3 subunit splice variant Gbeta3-s

causes enhanced chemotaxis of human neutrophils inresponse to interleukin-8. Naunyn Schmiedebergs ArchPharmacol 1999;360:27-32.

10. Lindemann M, Virchow S, Ramann F, Barsegian V,Kreuzfelder E, Siffert W, et al. The G protein beta3subunit 825T allele is a genetic marker for enhanced Tcell response. FEBS Lett 2001;495:82-6.

11. Beige J, Hohenbleicher H, Distler A, Sharma AM.G-protein beta3 subunit C825T variant and ambulatoryblood pressure in essential hypertension. Hypertension1999;33:1049-51.

12. Benjafield AV, Jeyasingam CL, Nyholt DR, Griffiths LR,Morris BJ. G-protein beta3 subunit gene (GNB3) variantin causation of essential hypertension. Hypertension1998;32:1094-7.

13. Baumgart D, Naber C, Haude M, Oldenburg O, Erbel R,Heusch G, et al. G protein beta3 subunit 825T allele andenhanced coronary vasoconstriction on alpha(2)-adreno-ceptor activation. Circ Res 1999;85:965-9.

14. Wenzel RR, Siffert W, Bruck H, Philipp T, Schafers RF.Enhanced vasoconstriction to endothelin-1, angiotensinII and noradrenaline in carriers of the GNB3 825T allelein the skin microcirculation. Pharmacogenetics 2002;12:489-95.

15. Nurnberger J, Dammer S, Mitchell A, Siffert W, WenzelRR, Gossl M, et al. Effect of the C825T polymorphism ofthe G protein �3 subunit on the systolic blood pressurelowering effect of clonidine in young, healthy males. ClinPharmacol Ther 2003;74:53-60.

16. Zeltner R, Delles C, Schneider M, Siffert W, SchmiederRE. G-protein beta(3) subunit gene (GNB3) 825T alleleis associated with enhanced renal perfusion in early hy-pertension. Hypertension 2001;37:882-6.

17. Zill P, Baghai TC, Zwanzger P, Schule C, Minov C,Riedel M, et al. Evidence for an association between aG-protein beta3-gene variant with depression and re-sponse to antidepressant treatment. Neuroreport 2000;11:1893-7.

18. Schafers RF, Nurnberger J, Rutz A, Siffert W, WenzelRR, Mitchell A, et al. Haemodynamic characterization ofyoung normotensive men carrying the 825T-allele of theG-protein beta3 subunit. Pharmacogenetics 2001;11:461-70.

19. Aellig WH. A new technique for recording compliance ofhuman hand veins. Br J Clin Pharmacol 1981;11:237-43.

20. Virchow S, Ansorge N, Rubben H, Siffert G, Siffert W.Enhanced fMLP-stimulated chemotaxis in human neutro-phils from individuals carrying the G protein beta3 sub-unit 825 T-allele. FEBS Lett 1998;436:155-8.

21. Schunkert H, Hense HW, Doring A, Riegger GA, SiffertW. Association between a polymorphism in the G proteinbeta3 subunit gene with lower renin and elevated dia-stolic blood pressure. Hypertension 1998;32:510-3.

22. Fernandez-Real JM, Penarroja G, Richart C, Castro A,Vendrell J, Broch M, et al. G protein beta3 gene variant,

CLINICAL PHARMACOLOGY & THERAPEUTICSVOLUME 74, NUMBER 5 Mitchell et al 503

vascular function, and insulin sensitivity in type 2 diabe-tes. Hypertension 2003;41:124-9.

23. Meirhaeghe A, Bauters C, Helbecque N, Hamon M,McFadden E, Lablanche JM, et al. The human G-proteinbeta3 subunit C825T polymorphism is associated with cor-onary artery vasoconstriction. Eur Heart J 2001;22:845-8.

24. Katsuki S, Arnold W, Mittal C, Murad F. Stimulation ofguanylate cyclase by sodium nitroprusside, nitroglycerinand nitric oxide in various tissue preparations and com-parison to the effects of sodium azide and hydroxyl-amine. J Cyclic Nucleotide Res 1977;3:23-35.

25. Katsuki S, Arnold WP, Murad F. Effects of sodiumnitroprusside, nitroglycerin, and sodium azide on levels

of cyclic nucleotides and mechanical activity of varioustissues. J Cyclic Nucleotide Res 1977;3:239-47.

26. Schultz K, Schultz G. Sodium nitroprusside and othersmooth muscle-relaxants increase cyclic GMP levels inrat ductus deferens. Nature 1977;265:750-1.

27. Sperling H, Eisenhardt A, Virchow S, Hauck E, Lenk S,Porst H, et al. Sildenafil response is influenced by the Gprotein beta 3 subunit GNB3 C825T polymorphism: apilot study. J Urol 2003;169:1048-51.

28. Stein CM, Lang CC, Nelson R, Brown M, Wood AJ.Vasodilation in black Americans: attenuated nitric oxide-mediated responses. Clin Pharmacol Ther 1997;62:436-43.

CLINICAL PHARMACOLOGY & THERAPEUTICS504 Mitchell et al NOVEMBER 2003