Plasma l-[3H]Norepinephrine, d-['4C]Norepinephrine,and d,l-[3H]Isoproterenol

Kinetics in Essential Hypertension

DAVID S. GOLDSTEIN, DAVID HORWITZ, HARRYR. KEISER, RONALDJ. POLINSKY,and IRWIN J. KOPIN, National Heart, Lung, and Blood Institute and NationalInstitute of Mental Health, National Institutes of Health, BethesdaMaryland 20205

A B S T R A C T We infused tracer-labeled 1-[3H]-norepinephrine, d-['4C]norepinephrine, and d,l-[3H]-isoproterenol simultaneously into patients with essentialhypertension and into normotensive control subjects,in order to determine whether abnormalities in thedisappearance kinetics of these substances characterizedthe hypertensive patients. The mean preinfusion venousplasma norepinephrine concentration was somewhathigher in the hypertensive group (260 vs. 194 pg/ml,P = 0.06), but the groups did not differ in the disap-pearance kinetics of 1- or d-norepinephrine or of iso-proterenol. Preinfusion plasma norepinephrine wassignificantly positively correlated with calculated spill-over rates in both the hypertensive and normotensivegroups, but not with norepinephrine clearances. Thedli ratio in plasma norepinephrine was the same as inthe infusate during and after the infusion, even afterpretreatment with the neuronal norepinephrine uptakeblocker, desipramine. Because isoproterenol is not takenup by nerve endings, the ratio of [3H]isoproterenol tol-[3H]norepinephrine increased after the infusion ended.This increase was almost completely abolished by pre-treatment with desipramine. These results indicate that(a) increased plasma norepinephrine levels seen in somepatients with essential hypertension result from in-creased sympathetic neural activity and not from de-creased clearance of norepinephrine, (b) changes inthe isoproterenol/norepinephrine ratio after simulta-neous infusion of both provide an index of neuronalnorepinephrine uptake in man, and (c) neuronal nor-epinephrine uptake is not stereospecific.

Received for publication 25 June 1982 and in revisedform 20 July 1983.

INTRODUCTION

The results of recent extensive literature reviews, aswell as several large-scale comparative studies of pa-tients with essential hypertension and of normotensivecontrol subjects, have suggested that a subgroup of hy-pertensives have elevated resting, supine, venous plasmanorepinephrine (NE)' levels, consistent with augmentedsympathetic nervous system activity in some patientswith essential hypertension (1-6).

The evidence, however, that plasma NE measuressympathetic neural activity in man is mainly indirect.Patients who have undergone sympathectomy for Ray-naud's phenomenon show decreased venous plasma NEin the venous drainage from the sympathectomizedlimb (7). A variety of stimuli known to increase sym-pathetic neural activity in animals, e.g., exercise andenvironmental stress, also increase plasma NE in man(8, 9). Diabetic patients with autonomic neuro-pathy and patients with idiopathic orthostatic hypo-tension show decreased levels of NE either at rest orin response to standing (10, 11). Infusions of amphet-amine or tyramine, which release NE from sympatheticnerve endings in vitro, produce increases in venousplasma NE in man associated with the pressor re-sponse-presumably a measure of the increment in NEat the neuroeffector junction (12). Finally, when sym-pathetic nerve activity has been recorded directly in

' Abbreviations used in this paper: COMT, catechol-O-methyltransferase; I, isoproterenol; LCED, liquid chroma-tography with electrochemical detection; MAO, monoamineoxidase; NE, norepinephrine; NMDA, N-methyldopamine;U, and U2, Uptake, and Uptake2.

The Journal of Clinical Investigation Volume 72 November 1983- 1748-17581748

man, peripheral plasma NE has correlated fairlywell (13).

These findings indicate that venous plasma NEdoesbear some relationship, albeit indirect, to synaptic cleftNE. Even so, it is well known that venous plasma NEis the product of several complex, interacting pro-cesses, some of which are incompletely understood,intervening between sympathetically mediated NErelease from presynaptic sites and norepinephrinemeasured in venous blood. Fig. 1 shows some of theseprocesses, and serves as a reference for our discussionof the rationale for the study reported here.

Most endogenously released NEdisappears from thesynaptic cleft by reuptake (Uptake,; UI) into the pre-synaptic axon (14). U1 is an active process that is notstereoselective (15-16), but inside the nerve ending,NE is catabolized by monoamine oxidase (MAO) orincorporated into storage vesicles, both processes beingstereospecific.

NE also may be removed from extraaxonal sites bya nonneuronal, nonstereospecific uptake process (Up-take2; U2), e.g., into smooth muscle cells (14-15). Onceinside these cells, NE may be catabolized by catechol-O-methyltransferase (COMT), which is not stereospe-cific.

Finally, a proportion of NE in the synaptic cleftdiffuses through the blood vessel wall or out the ad-ventitia and appears in the general circulation. Cir-culating NE is predominantly conjugated (17), takenup by platelets (18), and excreted by the kidney (19),as well as metabolized in the liver and other organs.The fact that plasma NE increases from artery to vein(20) suggests that there is a net release of NE into thegeneral circulation along the vascular tree.

Whereas endogenously released NEis removed fromthe synaptic cleft mainly by U1, U2 and O-methylationappear to be more important removal mechanisms forexogenously administered NE (14).

Exogenously administered NE also can be removedby Ul, since pretreatment with the specific U1 blockerdesipramine prolongs the disappearance of NE fromthe bloodstream in man (21). It must be emphasizedthat details of this schema derive mainly from in vitroor in vivo studies in animals, not in people.

In an attempt to measure separately the contribu-tions of these factors in determining plasma NE inpatients with essential hypertension, Esler and asso-ciates (22, 23) measured the plasma disappearance oftracer amounts of infused radioactivity labeled NE.They reported that labeled NE disappeared with an

(reeptor

LUMEN

FIGURE 1 Diagrammatic cross-section of a blood vessel, showing several factors influencingthe relationship between NE release from sympathetic nerve endings and circulating NE.

Norepinephrine and Isoproterenol Kinetics in Essential Hypertension 1749

early, rapid half-life (t1/20) of -2 min and later witha slower half-life (t1/2b) of >30 min; that desipramineprolonged t1/2a; and that a subgroup of hypertensivepatients showed prolongation of t1/2a. They concludedthat in some patients with essential hypertension,faulty U, could account for elevated circulating levelsof NE and the patients' high blood pressure. Paradox-ically, however, in these studies U1 blockade with de-sipramine did not appear to increase plasma NE. Theseinvestigators also reported that apparent spillover ratesof NE into the bloodstream correlated well withresting plasma NE but NE clearance rates did not,indicating that differences across individuals in restingplasma NE depend more on differences in sympa-thetic tone than on differences in NE removal mech-anisms (22).

The evidence that the technique used to measureUi actually did so is circumstantial. Wehave evalu-ated, in patients with essential hypertension and innormotensive controls, two somewhat similar tech-niques that we thought might be more specific.

The techniques we used exploited two generally ac-cepted points: The first is that MAOactivity and ve-sicular uptake are stereospecific, whereas COMT, U2,other removal processes, and probably renal excretionare not. The second is that isoproterenol (I) is not takenup by U, (24), but undergoes COMTdegradation andrenal excretion in a manner similar to NE. If one weresimultaneously to administer tritiated I-NE and d-['4C]-NE, any differences in the disappearance kinetics ofthese substances-most simply reflected in the dll ra-tio-would provide an index of the contribution of ste-reospecific processes. Similarly, any differences in thedisappearance kinetics of I and i-NE-most simply re-flected in the I/NE ratio-could be attributed to Ul.

On the other hand, if neither the dll ratio nor theI/NE ratio changed in man during the plasma disap-pearance of simultaneously infused tracer amounts ofI-[3H]NE, d-[14C]NE, and [3H]I, one would concludethat the disappearance of exogenously administeredNEwas not so much due to U1 as to U2, O-methylation,renal excretion, or other removal processes.

To test whether changes in either the dll or I/NEratios provide an index of U, in man, we measuredthese ratios in normotensive individuals before andafter treatment with desipramine.

In summary, we infused labeled stereoisomers of NEand labeled I simultaneously in patients with essentialhypertension and in normotensive controls, and com-pared the plasma disappearance of these substancesin these groups, in order to determine whether highNE levels seen in some hypertensive result from ex-cessive sympathetic neural activity or from prolongedplasma disappearance of NE. We used the ratio of

[3H]I to 1-[3H]NE after the infusion as an index of neu-ronal uptake of NE.

METHODS

Subjects. 22 patients with essential hypertension partic-ipated in this study. Clinical data about them are shown inTable I. All the patients had systolic blood pressures averaging>140 mmHgor diastolic pressures averaging >90 mmHgduring at least five outpatient visits. The patients either hadnever received antihypertensive medication or had not takenany for at least 2 wk. Secondary forms of hypertension wereexcluded by blood and urine tests (complete blood count,serum electrolytes, liver function tests, blood urea nitrogen,creatinine, glucose, uric acid, urinalysis, urine culture), med-ical history and physical examination, and, when appropriate,intravenous pyelography, renal scanning, abdominal com-puterized transaxial tomography, renal venous sampling, andarteriography. Patients were excluded and immediatelytreated if their diastolic pressures exceeded 120 mmHg.

13 subjects served as age-matched normotensive controls.None had shown systolic blood pressures >140 mmHgordiastolic pressures >90 mmHg. Most of them had no familyhistory of hypertension. Medical history, physical exami-nation, and routine laboratory blood and urine testing ex-cluded any concurrent illness at the time of study. None ofthe controls were hospital or laboratory personnel. Most hadbeen recruited through the National Institutes of Health(NIH) Normal Volunteer program.

All but two of the patients and all the controls were studiedas outpatients. All subjects consented in writing to participatein this protocol, which was approved by the National Heart,Lung, and Blood Institute Institutional Review Board. In onehypertensive and one control subject, technical problemsforced a premature end to the study, so that only base-lineNE values from these subjects were included in the dataanalysis. These subjects were in addition to the 32 who un-derwent the entire infusion protocol.

Protocol. Most of the subjects were studied in the morn-ing, after a light breakfast. With the subject supine, intra-venous plastic catheters or butterfly needles were insertedin each forearm and isotonic saline infused slowly. After atleast 15 more minutes, base-line blood pressure was mea-

TABLE ISummary Clinical Data for the Hypertensive

and Normotensive Groups

Hypertensive Normotensive

n 22 13Age (yr) 42±11 43±14Mean arterial pressure

(mmHg) 108±8e 83±9Pulse rate

(beats per minute) 75±81 65±6Weight (kg) 76±17 70±11Creatinine (mg/dl) 1.1±0.2 1.0±0.1

All values expressed±1 SD.Significant hypertensive-normotensive difference, P < 0.001.

I Significant hypertensive-normotensive difference, P < 0.01.

1750 Goldstein, Horwitz, Keiser, Polinsky, and Kopin

sured in the arm that would be used for blood sampling. Justafter this, two 10-ml blood samples were drawn from thesampling catheter, without use of a tourniquet, through astopcock and directly into chilled, evacuated, heparinizedtubes. About 10-15 seconds were required to fill each col-lection tube.

After the base-line sampling, an intravenous infusion wasbegun in the other arm. The infusate contained 50 IACi tri-tiated I-NE, 10 ACi "C-labeled d-NE, and 50 MCi tritiatedd,l-I in 65 ml 5%dextrose. The infusion, which was controlledby an IMED (San Diego, CA) pump set to run at 180 ml/h, lasted 20 min.

At 1, 3, 5, 10, and 15 min and at the end of the infusion,blood samples were drawn through the stopcock into collec-tion tubes as described above. In the few instances wherethe infusion ended just before 20 min, a sample was drawnimmediately and the stopwatch restarted. After the infusion,blood samples were drawn at 1, 2, 3, 5, 10, 20, and, in severalcases, at 40 and 60 min.

Drugs and dosimetry. Weused labeled substances withhigh specific activity, so that the concentrations of physio-logical active NE and I would be too small to affect endog-enous NE release. Tritiated 1-NE (5 Ci/mmol) was obtainedfrom Amersham Corp. (Arlington Heights, IL) and preparedby the NIH Radiopharmacy for use in humans. The -46MCi of 1-NE infused consisted of 1.6 jug, resulting in an 1-NE infusion rate of -78 ng/min. During the study, ring-labeled 1-NE of even higher specific activity was obtained,but the amount of radioactivity infused was not changed.The d-['4C]NE (33 mCi/mmol) also was obtained from Amer-sham Corp. and tested and prepared for human use in thesame way. The 9.2 MCi of d-NE (47 ,ug) resulted in a d-NEinfusion rate of 2.4 ,ug/min. Of the increment in total plasmaNE observed in this study, -97% was due to physiologicallyinactive d-NE. The tritiated I (11.2 Ci/mmol) was infusedat a rate of -43 ng/min.

Solutions of these substances were aliquoted and stored ina -80°C freezer. When assayed for catecholamine contentby the liquid chromatographic-electrochemical techniquedescribed below, the NEwas free of any contamination withI and vice versa, and at least 95% of the radioactivity inthese substances was contained in the appropriate chro-matographic peaks.

Radiation dosimetry was calculated on the basis of a bio-logical half-life of 1 d (19). Wecalculated that the patientswould be exposed to -1 mrad/infusion.

To assess whether inclusion of I or d-NE influenced thedisappearance of tritiated 1-NE, for five hypertensives theinfusate contained only the tritiated 1-NE.

Sample collection and assay technique. The blood sam-ples were placed on ice until centrifuged at 4°C within 2h. The plasma was transferred to plastic sample tubes with-out additives and stored either in a -80°C freezer or in atank containing liquid nitrogen until the time of assay. Allthe samples from a given infusion were assayed at the sametime.

Plasma NE and I were separated by liquid chromatog-raphy and quantified with electrochemical detection (LCED).Since we had previously validated and reported this tech-nique for measuring plasma catecholamines (25), the follow-ing description will deal mainly with those aspects used forthe first time in this study.

Either 1 or 2 ml freshly thawed and recentrifuged plasmawas added to a plastic sample tube containing - 10 mgacid-washed alumina. To this were added the internal standard-1 ng N-methyldopamine (NMDA), 2 ng unlabeled I, or

both-and then 400 Al Tris-EDTA buffer at pH 8.6. Thetube was shaken vigorously for 20 min and the alumina thenwashed twice with water. The catecholamines were elutedwith 100 MI of 0.1 Mperchloric acid or 100 MI of 0.2 Maceticacid. In most cases, 90 Ml of eluate were injected into theLCEDapparatus.

The chromatographic and detection equipment includeda Waters M6000 pump, U6K injector, and MicrobondapakC18 reverse phase column (Waters Associates, Milford, MA).The mobile phase consisted of sodium acetate, acetonitrile,EDTA, and heptane-sulfonic acid as previously described(25). The mobile phase was pumped at 1.0 ml/min and re-cycled. The electrochemical detector was set at 0.5 V at asensitivity in most cases of 1.0 nA/V. The limit of detectionof this technique is -10 pg/ml plasma. A sample chro-matogram of an injection of a mixture of NE, epinephrine,dopamine, NMDA,and I standards is shown in Fig. 2. Samplechromatograms of plasma-derived eluates before and duringthe infusion are shown in Fig. 3.

As shown in the figures, NE and I were easily separatedand detected by LCED. The NE and I peaks for each eluatewere collected from the postdetector outlet tube into glassscintillation vials. For NE, collection began 20 s before theexpected appearance of the peak and ended 70 s after thepeak, for a total of 90 s (1.5-ml mobile phase). I was collected

0

0

2 mn

FIGURE 2 Chromatographicpg each of NE, epinephrineand I standards.

4

0

z

tracing after injection of 500(E), dopamine (DA), NMDA,

Norepinephrine and Isoproterenol Kinetics in Essential Hypertension 1751

2min

Ezz 0

FIGURE 3 Chromatographic tracings after injection ofplasma-derived eluates (a) before and (b) during infusion of1-[3H]NE, d-[14C]NE, and [3H]I. Same abbreviations as inFig. 2.

in two successive 1.5-ml aliquots beginning either 30 s beforethe expected appearance of the I peak or when the I internalstandard peak began to rise (whichever came sooner) until150 s after the I peak. The reason for this difference incollection times was that the I peak, being further from thesolvent front, was wider than the NE peak. Whereas >95%of the radioactivity in NEoccurred within the 90-s collectionperiod, only 71% of the radioactivity in I occurred withinthe first 90 s of collection of the I peak. More than 95% ofthe radioactivity in I did appear, however, by the end of the180-s collection. Accordingly, for I, the radioactivity for thetwo consecutive 1.5-ml aliquots was summed for each bloodsample. In several cases, only one 1.5-ml aliquot was collectedfor I. In these cases, the quantity of radioactivity in I wasback-calculated by dividing the counts per minute by 0.71,since this was the proportion of radioactivity in 3-ml aliquotsthat occurred in the first 1.5-ml aliquot in 12 individuals,and it also was the proportion of I standard occurring in thefirst 1.5-ml aliquot collected after the standard had undergonethe alumina sample preparation step.

Plasma NE was calculated from the peak heights of NEand the internal standards and expressed in picograms permilliliter.

Without added unlabeled I, the plasma concentration oflabeled I during the infusion was too small to produce adetectable peak in the obtained chromatograms, indicatinga concentration <10 pg/ml.

When both I and NMDAwere used as internal standardson the same specimen, the NE concentrations calculated

from them were virtually identical. The I peak height de-creased as the volume injected increased, even with the samenumber of picograms injected, so that standards had to beinjected in identical volumes to obtain equivalent NMDAand I recoveries. The average recovery through the aluminapreparation step was -70% in samples from both the hy-pertensive and normotensive subjects.

Assay of 3H and 14C. The amount of d-['4C]NE, tritiated1-NE, and tritiated I were measured by assay of the tritiumand "4C in the NE and I peaks. The samples were assayedby liquid scintillation spectrometry, counting disintegrationsduring 100 min, which was required because of the smallamount of radioactivity per sample and the simultaneouscounting of '4C and tritium on one channel. The counts perminute were adjusted for background, crossover from the14C to the tritium channels (the crossover from tritium to14C being negligible), the volume of plasma assayed, and therecovery through the alumina extraction step. Backgroundwas determined from the results obtained from the bloodsample drawn just before the infusion.

Spillover and clearance calculations. Apparent spilloverand clearance rates of NE into and out of the circulationwere calculated by applying previously described techniques(21, 22). Since 97% or more of the infused NE was the d-form, the steady-state endogenous NE concentration wascalculated by subtracting the d-[14C]NE concentration fromthe total in 15 subjects. This procedure demonstrated that,as expected, the steady-state endogenous NE concentrationwas unchanged from the preinfusion base-line level. Ac-cordingly, apparent spillover rates were calculated in eachsubject from the 1-[3H]NE infusion rate, the amount of ra-dioactivity in [3H]NE at the steady state, and the preinfusionNE concentration. Because the infusion rate of the labeledsubstances was the same for each subject, calculated spilloverand clearance rates varied inversely with the steady statelevel of labeled NE. Values of t1/2a for the disappearance ofNE were calculated from the slope of the line relating thelogarithm of the plasma tritium content in NE to time overthe first 5 min after the infusion ended. This technique wasused because after -20 min after the infusion levels ofradioactivity were too low to yield reliable points forgraphical analysis. Our procedure probably slightly over-estimated tl/2a.

Data analysis. In each subject, for each sample, the ratioof d-['4C]NE to tritiated 1-NE was determined from the netlevels of radioactivity in each; the ratio of I to 1-NE wasdetermined similarly. The I/NE and d/l ratios were ex-pressed as percentages of the corresponding ratios in theinfusate. The statistical significance of differences in trendsbetween the hypertensive and normotensive groups was as-sessed by means of analyses of variance for repeated mea-sures, Pearson correlation coefficients, and independent-means t tests (26).

Desipramine. In five normotensive and three hypertensivesubjects, the infusion was repeated after at least a 2-wk in-terval. Three h before this second infusion, the subjects re-ceived a single oral dose of 125 mg desipramine.

RESULTS

Total NE. The hypertensive subjects had a highermean preinfusion NE level than the normotensive sub-jects (260 vs. 194 pg/ml, t = 1.99, P = 0.06). Theincrement in total NE during the infusion, virtuallyentirely due to d-['4C]NE, was similar in the hyper-

1752 Goldstein, Horwitz, Keiser, Polinsky, and Kopin

TABLE IILevels of Radioactivity in 1-[3H]NE, d-['4C]NE, and [3H]I and Total Plasma NE during and after

Simultaneous Infusion of the Three Substances in Patients with EssentialHypertension (EH) and in Normotensive Controls (NT)

l-[3'HNE d-["CINE ('H]I Total plasma NE

Time EH NT EH NT EH NT Eli NT

nin net cpm/ml net cpm/ml net comp/ml pg/ml

1 37±44 19±33 10±13 6±9 30±33 17±24 328±145 228±863 170±91 172±119 52±28 47±32 153±77 143±120 533±150 454±1845 202±83 186±83 59±26 56±30 189±77 171±80 594±209 480±173

10 248±106 261±116 73±29 71±31 241±81 253±113 664±223 540±17415 253±93 258±103 73±27 70±32 258±72 273±87 637±209 522±17020 249±93 239±92 73±27 66±26 260±73 269±83 646±226 508±189

21 189±58 217±87 56±17 58±24 221±52 249±93 526±156 458±13922 126±32 150±53 36±10 40±16 160±42 196±63 473±168 393±13523 89±26 103±34 24±8 28±9 124±34 151±42 378±172 366±13025 58±21 62±22 16±6 17±4 99±31 104±24 337±158 276±8130 32±13 31±13 8±3 9±3 58±24 61±17 301±138 230±7540 19±12 21±14 6±2 5±3 31±17 36±17 285±141 227±72

All values expressed±1 SD.

tensive and normotensive groups, so that at all timepoints during and after the infusion, the mean totalNE level was higher in the hypertensive subjects (Ta-ble II). In the five hypertensive subjects who receivedonly tritiated I-NE, total plasma NE at 20 min afterthe start of the infusion was the same as before theinfusion (194 vs. 216 pg/ml). Pretreatment with de-sipramine did not influence base-line plasma NE (182pg/ml without desipramine, 193 pg/ml with desipra-mine among the eight subjects tested).

1-[3H]NE. As indicated in Fig. 4, tritiated i-NE ac-cumulated rapidly in plasma during the infusioin anddisappeared rapidly after the infusion ended. Duringthe first 5 min after the infusion, 1-[3H]NE disappearedwith a t, 2 of '2.5 min, after which the disappearancerate declined and a second component with a muchslower disappearance rate predominated. By 60 minafter the infusion, the radioactivity in l-[3H]NE couldnot be distinguished from background, despite the 100-min period of scintillation counting.

Table II and Fig. 4 demonstrate that the hyperten-sive and normotensive groups showed virtually iden-tical plasma 1-[3H]NE levels at all points during andafter the infusion. Furthermore, the mean t 1 2a for thedisappearance of 1-['3H]NE during the first 5 min afterthe infusion also was similar in the hypertensive andnormotensive groups (Table III). The plasma level of1-[3H]NE at 20 min after the start of the infusion wassimilar in the patients receiving only l-[3H]NE in theinfusate as in those receiving 1-[3H]NE, d-['4C]NE, andtritiated I simultaneously.

d,l-[3H]. The plasma disappearance rate of [3H]Iwas slower than that of 1-[3H]NE. This difference wasmost apparent during the first several minutes afterthe infusion ended. The tl/2a for tritiated I was -3.5min. As with 1-[3H]NE, the overall mean disappearancefrom plasma of [3H]I was similar in the hypertensiveand normotensive groups.

300r

100

zw

z

0

i0

50

10

(*0

0

0

0

E-

Do

S

Do

0

F

-U

a

135 10 15 20 25 30TIME min

40

Fi(GURE 4 Plasma mean levels of radioactivity in 1-[3H]NEand d-['4C]NE during and after the infusion in the hyper-tensive (EH) and normotensive (NT) groups. *, EH 1-[3H]NE;0, NT, 1-[3H]NE; *, EH d-['4C]NE, E, NT, d-['4C]NE.

Norepinephrine and Isoproterenol Kinetics in Essential Hypertension

r

.0

5L

1753

TABLE IIIIndividual Patient Norepinephrine Kinetic Data

Patient Base-line NE Spillover rate Clearance I/NE % slope t,/2 NE/Us

pg/ml

Hypertensive134678

101319202324252627282930313233

Mean±1 SD

Normotensive259

111214151617182122

Mean±1 SD

304291338267301471207204274226407109177140379310

75247

90215356

257±105

199257132138177386131221274179131203

202±76

mcg/min liters/min

0.710.880.951.011.022.381.030.741.111.353.260.431.260.780.951.050.180.670.241.431.90

1.11±0.70

0.500.690.860.601.181.490.430.710.990.680.921.03

0.84±0.30

2.353.032.803.793.385.055.003.644.045.998.004.037.135.562.503.402.412.782.706.655.34

4.27+1.67

2.502.696.524.336.673.863.303.223.603.797.035.05

4.38+1.58

min

0.150.090.050.230.030.060.210.080.160.140.040.11

0.130.280.220.11

0.13±0.07

0.070.070.070.060.090.310.220.150.370.040.060.10

0.13±0.11

2.742.152.151.513.343.012.512.742.152.516.022.322.742.742.012.011.771.581.672.012.18

2.47±0.94

2.152.013.762.323.012.152.742.511.513.013.762.15

2.59±0.69

2,0273,2336,76001,161

10,033-7,850°

9862,5501,7131,614

10,175°991

1,900321977

3,236

3470±3305

2,8433,6711,8862,3001,9671,245

5951,473

7414,4752,1832,030

2,117±1,128

I/NE % slope = slope of line relating the ratio of counts per minute per milliliter tritiated isoproterenol tocounts per minute per milliliter tritiated L-NE expressed as a percentage of the ratio in the infusate, across thetime interval from 0 to 5 min after the infusion ended; t1/2" = first half-life of plasma disappearance of NEdetermined by net counts per minute per milliliter tritiated I-NE across the time interval from 0 to 5 min afterthe infusion ended. Patient No. indicates sequence in which subjects actually were studied. Five hypertensivesubjects received only tritiated [-NE, accounting for the unreported data.e Individual value > 2 SD from the normotensive mean.

I/NE ratio. In marked contrast to the lack ofchange in the dll ratio during or after the infusion,the I/NE ratio increased slowly during and rapidly

after the infusion in both the hypertensive and nor-motensive groups (Fig. 5). Pretreatment with desipra-mine virtually abolished this change in ratios (Fig. 5).

1754 Goldstein, Horwitz, Keiser, Polinsky, and Kopin

200

0

z

0

U

LL

150

100

90

80

70

NT I NE

/ = EH l:NE

/NT D L

- --------- EH D:L

NT lNE pDESIP

EH I:NE p DESIP

nfusion

i i , ,_ __- 1__

3 5 1 0 1 5 20 25 30

TIME min

FIGURE 5 Ratio of I to 1-[3H]NE (1/NE) and of d-NE to 1-NE (d/l) expressed as percentages of the ratios in the infusatein the hypertensive and normotensive groups. I/NE meanvalues for five normotensive and 3 hypertensive subjects pre-treated with desipramine (pDESIP) are also shown.

The degree of increase in the I/NE ratio after theinfusion ended did not differ significantly between thehypertensive and normotensive groups, as assessed bythe interaction effect in an analysis of variance fordiagnostic category and time.

Table III shows individual values for the slope ofthe regression line relating to time the I/NE ratio (ex-pressed as a proportion of that in the infusate) duringthe first 5 min after the infusion ended. U, activityestimated usiIng this technique was significantly neg-

atively correlated with the t1 2aof 1-[TH]NE (r = -0.57in the hypertensive and -0.63 in the normotensivegroups, P < 0.01). Table III also shows individual valuesfor the ratio between preinfusion plasma NE and U1activity (NE/UJ1) by the I/NE ratio technique. As in-dicated, four hypertensive subjects (25% of the studiedgroup) had a NE/UI ratio > 2 SD from the normotensivemean. No clinical characteristics (in particular, serum

creatinine, body weight, and severity of hypertension)differentiated these patients from the other hyperten-sives.

Apparent spillover and clearance rates. The con-

gruity of plasma 1-[3H]NE and of plasma d-['4C]NEdisappearance curves in the hypertensive and nor-

motensive groups meant that their overall NE clear-ances were identical (Table III). Preinfusion plasmaNE was unrelated to clearance in both groups (Fig. 6).

In contrast, apparent spillover rates correlated sig-

FIGURE 6 Preimifusion plasma NE as a function of NE clear-ance (in liters per minute). *, individual hypertensives, 0,

individual normotensives.

nificantly with preinfusion plasma NE in the hyper-tensive (r = 0.72, P < 0.01) and normotensive(r = 0.69, P < 0.01) groups (Fig. 7).

Although the hypertensive mean spillover rate ex-

ceeded the normotensive mean spillover rate, the dif-ference was not statistically significant (Table 111).Hypertensive subjects with preinfusion NE levelsgreater than the median hypertensive value (250 pg/ml) had a higher mean spillover rate than hypertensivesubjects with preinfusion NE less than the medianvalue (1384 vs. 815 ng/min, t = 2.00, P = 0.06) andalso a higher mean spillover rate than the normotensivesubjects (839 ng/min, t = 2.20, P < 0.05).

Desipra mine. As mentioned above, desipraminepretreatment did not alter preinfusion plasma NE levelsbut did profoundly influence plasma NE kinetics. Afterdesipramine pretreatment, the infusion produced an

extremely rapid accumulation of 1-[3H]NE and d-['4C]NE in plasma, with plateau levels averaging aboutdouble the levels seen without desipramnine pretreat-ment (Fig. 8). Desipramine prolonged the disappear-ance rates of both d- and 1-NE. The rapid initial de-crease in plasma levels was slowed to an extent con-

sistent with the elevated steady-state levels, i.e., abouttwofold, since the later, slower disappearance rate was

Norepinephrine and Isoproterenol Kinetics in Essential Hypertension

500r.

400*

300wLuz

0.z

Sn4-ia.

00

0

00* :0O

0

00

0

0 0

0

0

0

20*

o00o

0o

2 4CLEARANCE

1755

wz

m

n

ui

0.

0z

01

0.

400

3000

200 -

0

100 _

: /a=120 ,=0.12

o / r=O. 7 1

I

50

E 40

X 300.0

0

20 _

0 00

3000

10

5

4

3

SPILLOVER RATE ng/min

FIGURE 7 Preinfusion plasma NE as a function of NE spill-over rate.

unchanged by the drug. Desipramine had no effect onthe d/l ratio but, as mentioned above, virtually com-pletely abolished the increase in the I/NE ratio duringand after the infusion. None of these effects of desi-pramine differentiated the hypertensive and normo-tensive groups.

DISCUSSION

Several studies involving large numbers of individualshave suggested that a minority of patients with essentialhypertension have elevated levels of plasma NE. Therelationship between sympathetic neural activity andcirculating NE levels, however, is complex. This studywas designed to examine factors that influence plasmaNE levels, in an attempt to explain why those levelsare high in some patients with hypertension.

Concentrations of circulating, free NE depend notonly on sympathetically mediated neural secretion, butalso on metabolic degradation, uptake and bindingprocesses, the site of sampling, and renal excretion. Inthis study, we infused tracer-labeled NE to determinewhether hypertensives show decreased plasma clear-ance of NE. To assess the contribution of stereospecificprocesses, we compared the disappearance kinetics ofd- and I-NE, most simply reflected in the dll ratio. Toexamine differences that might depend on neural NEuptake, we compared the disappearance of infused I

21 3 5 10 15 20 25 30 40 60

TIME min

FIGURE 8 Effect of desipramine (1 Des) on plasma 1-[3H]NEand d-['4C]NE levels during and after the infusion in fivenormotensive subjects.

and NE, most simply reflected in the I/NE ratio, be-cause I is not taken up by sympathetic nerve endings,but NE is inactivated to an important degree by neu-ronal uptake.

Preinfusion resting NE levels tended to be higheron average in the patients with essential hypertensionthan in the age-matched normotensive controls, yet noaspect of the kinetics of infused l-[3H]NE, d-['4C]NE,or [3H]I significantly differentiated the groups. PlasmaNE levels were significantly positively correlated withapparent spillover rates but not at all with NE clear-ances. Relatively hypernoradrenergic hypertensivesubjects tended to have higher spillover rates than nor-monoradrenergic hypertensive subjects. These resultsindicate that the increased resting NE levels seen insome patients with essential hypertension do not resultfrom prolonged plasma disappearance of NE so muchas from increased NE release.

Because no change in the d/l ratio occurred in hy-pertensive or normotensive subjects or after the in-fusion, the disappearance of exogenously administeredNE from plasma is probably not influenced to anyimportant extent by stereospecific processes such asMAOoxidation or incorporation of NE into storage

1756 Goldstein, Horwitz, Keiser, Polinsky, and Kopin

vesicles, which have been reported to be stereospecific(15, 16). The lack of effect of desipramine, a specificblocker of neuronal NE uptake (Ul), on dll ratios sug-gests that in man, U1 is not stereoselective.

The clearcut, progressive, and large increases in theI/NE ratio after termination of the infusion in all nor-motensive and most hypertensive subjects indicate thatneural NE uptake does play a role in the plasma dis-appearance of exogenously administered NE. The va-lidity of changes in the I/NE ratio as an index of neu-ronal NE uptake in man was tested by repeating theinfusion in five healthy and three hypertensive subjectsafter treatment with desipramine. U1 blockade virtu-ally abolished the increase in the I/NE ratio in thefirst 5 min after the infusion, confirming that a sub-stantial proportion of injected NE is taken up by nerveendings in man.

A few other groups of investigators have comparedplasma NE disappearance in hypertensive and nor-motensive groups (22, 23, 27, 28) after an infusion tosteady state, with discrepant results. Esler and co-workers found that the first half-life of NE disappear-ance is prolonged in a minority of patients with essentialhypertension (22), which they attributed to decreasedU1 (29). Our results complement theirs, because U1activity using the I/NE ratio was significantly positivelycorrelated with U1 activity using the t1/2" for the plasmadisappearance of 1-[3H]NE, and a few hypertensivesubjects showed decreased U, activity by both tech-niques. Also consistent with the findings of Esler et al.(22), hypertensive subjects with relatively high restingplasma NE levels also tended to have decreased U1activity. Four hypertensives (25% of the tested group)had a NE/Ul ratio > 2 SD from the normotensivemean. No clinical characteristics distinguished thesepatients from other hypertensive subjects.

The proportion of the rapid disappearance of 1-[3H]NE due to U1 can be estimated from the data pre-sented in Table III as follows: First-order disappearancekinetics for I (I) and NE (N) means that I = I_e-k" andN = Noe-k2 , where k is the proportion of radioactivitylost in time t. At time t, the ratio of tritium in I to thatin NE is I/N = Io/Noe(k2-k,)t, and so the I/NE ratio asa proportion of that in the infusate is (I/N)/(IO/NO)= exp[(K2 - KI)t]. Since for low values of a, ea = 1+ a, the values for the slopes for I/NE reported inTable III were roughly equal to k2 - k. From TableIII, k2 = 0.693/2.48 = 0.28, and k2 - k1 = 0.13. Theproportion of the early, rapid disappearance of NEdueto U1, therefore, is 0.13/0.28, a bit under 1/2. Thesecalculations demonstrate that the two techniques formeasuring U1 activity were not independent, account-ing for the correlation between their values. Further-more, the slope of the I/NE ratio more accurately mea-sures U1 than does the t1I2a of 1-[3H]NE, since the latter

is determined by other factors (U2, metabolism, etc.)besides U,.

Our interpretation of the findings of the present studydiffers somewhat from the model used by Esler et al.(22). According to their conception, faulty neuronalreuptake of norepinephrine causes higher NEspilloverrates into plasma in hypernoradrenergic hypertensives.This hypothesis can account neither for the lack ofeffect of extensive U1 blockade with desipramine onplasma NEnor for the fact that directly measured sym-pathetic activity correlates well with venous plasmaNE (13). Our working hypothesis is that in some hy-pertensive subjects, there is both decreased U1 activityand increased NE release from sympathetic nerve end-ings, and that these may be due to the same basicneuronal defect, the net result being increased stim-ulation of postsynaptic adrenoceptors. Weagree withEsler et al. (22) that it is possible that decreased U1activity accentuates circulating NE responses to stress.

Neither Esler et al. (22) nor we noted significantlyincreased apparent spillover rates in the groups of hy-pertensive subjects as a whole. Wedid find evidencefor increased spillover rates in the hypertensive subjectswith preinfusion NE levels greater than the hyperten-sive median value, although with substantial scatter inthe data. The calculation of apparent spillover ratesincludes errors due to variability in the specific activitiesof the infused NE, assay of tritium where the amountof radioactivity is small, inexact timing of sample col-lection, inconstancy of the infusion rate, and accuracyof the NE assay. It is therefore possible that differencesin preinfusion NE concentrations between the hyper-tensive and normotensive groups may have been ob-scured by errors inherent in the calculation of spilloverrates.

Apparent spillover rates and clearances, being phar-macokinetic rather than physiologic entities, need notbe related to specific processes influencing circulatingNE. For instance, calculated NE clearances may ormay not correlate with Ul, and apparent spilloverrates, which decrease after desipramine (because thespecific activity of labeled NE increases), may or maynot be associated with a decreased concentration ofNE in the synaptic cleft.

Because antecubital venous samples provided thebasis for the spillover and clearance calculations, ef-fects of pulmonary and arm extraction of NEprobablyled to overestimation of the spillover and clear-ance rates due to lower steady-state levels of thelabeled NE.

The results of the present study lead to the followingconclusions: (a) Most patients with essential hyperten-sion have normal venous plasma NE levels. (b) No ab-normality in the disappearance kinetics of either 1- ord-NE occurs in essential hypertension, even in the mi-

Norepinephrine and Iso proterenol Kinetics in Essential Hypertension 1757

nority of patients with relatively high resting plasmaNE levels. (c) Apparent NE spillover rates correlatewell with resting plasma NE, but NE clearances donot. (d) U1 activity can be measured in man by meansof the I/NE ratio. (e) U1 is not stereospecific in manbut does contribute importantly to the disappearancefrom plasma of exogenously administered NE. And (f)a minority of patients with essential hypertension showcoexistent decreased U1 activity and enhanced NE re-lease, resulting in increased circulating levels of NE.

REFERENCES

1. Goldstein, D. S. 1981. Plasma norepinephrine in essentialhypertension: a study of the studies. Hypertension.3:48-52.

2. Kopin, I. J., D. S. Goldstein, and G. Z. Feuerstein. 1981.The sympathetic nervous system and hypertension. InFrontiers in Hypertension Research. J. J. Laragh, F. R.Buhler, and D. W. Seldin, editors. Springer-Verlag, NewYork. 283-289.

3. Goldstein, D. S. 1981. Plasma norepinephrine duringstress in essential hypertension. Hypertension. 3:551-556.

4. Bertel, O., F. R. Buhler, W. Kiowski, and B. E. Lutold.1980. Decreased beta-adrenoreceptor responsiveness asrelated to age, blood pressure, and plasma catechol-amines in patients with essential hypertension. Hyper-tension. 2:130-138.

5. Ogawa, K., T. Matsuno, S. Tsuchiya, B. Masaaki, and K.Minaguchi. 1981. Plasma level of norepinephrine, cyclicAMPand cyclic GMPin essential hypertension. Jpn.Circ. J. 45:654-660.

6. Sever, P. S., M. Birch, B. Osikowska, and R. D. G. Tun-bridge. 1977. Plasma-noradrenaline in essential hyper-tension. Lancet. I:1078-1081.

7. Nielsen, S. L., N. J. Christensen, N. Olsen, and N. A.Lassen. 1980. Raynaud's phenomenon: peripheral cat-echolamine concentration and effect of sympathectomy.Acta Chir. Scand. 502:57-62.

8. Lake, C. R. 1979. Relationship of sympathetic nervoussystem tone and blood pressure. Nephron. 23:84-90.

9. Goldstein, D. S., R. Dionne, J. Sweet, R. Gracely, H. B.Brewer, Jr., R. Gregg, and H. R. Keiser. 1982. Circulatory,plasma catecholamine cortisol, lipid, and psychologicalresponses to a real-life stress (wisdom tooth extractions):effects of diazepam sedation and of inclusion of epi-nephrine with the local anesthetic. Psychosom. Med. Inpress.

10. Christensen, N. J. 1972. Plasma catecholamines in long-term diabetics with and without neuropathy and in hy-pophysectomized subjects. J. Clin. Invest. 51:779-787.

11. Ziegler, M. G., C. R. Lake, and I. J. Kopin. 1977. Thesympathetic-nervous-system defect in primary orthosta-tic hypotension. N. Engi. J. Med. 296:293-297.

12. Pickar D., C. R. Lake, R. M. Cohen, D. C. Jimerson, andD. L. Murphy. 1980. Alterations in noradrenergic func-tion during clorgyline treatment. Commun. Psycho-pharmacol. 4:379-386.

13. Wallin, B. G., G. Sundlof, B.-M. Eriksson, P. Dominiak,H. Grobecker, and L. E. Lindblad. 1981. Plasma nor-adrenaline correlates to sympathetic muscle nerve ac-

tivity in normotensive man. Acta Physiol. Scand. 111:69-73.

14. Bevan, J. A., R. D. Bevan, and S. P. Duckles. 1980. Ad-renergic regulation of vascular smooth muscle. Handb.Physiol. Sect. 2:515-566.

15. Henseling, M., and U. Trendelenburg. 1978. Stereose-lectivity of the accumulation and metabolism of nor-adrenaline in rabbit aortic strips. Naunyn-Schmiede-berg's Arch. Pharmacol. 302:195-206.

16. Eckert, E., M. Henseling, A. Gescher, and U. Tren-delenburg. 1976. Stereoselectivity of the distribution oflabelled noradrenaline in rabbit aortic strips after in-hibition of the noradrenaline-metabolizing enzymes.Naunyn-Schmiedeberg's Arch. Pharmacol. 292:219-229.

17. Kuchel, O., N. T. Buu, T. Unger, and J. Genest. 1978.Free and conjugated catecholamines in human hyper-tension. Clin. Sci. Mol. Med. 55:77s-80s.

18. Mattiasson, I., and B. Hood. 1982. Efflux of noradren-aline from platelets in normotensive members of hy-pertensive families. Clin. Sci. (Lond.). 62:151-155.

19. Gitlow, S., S. Dziedzic, L. Dziedzic, and I. Roubein.1981. Metabolism of D,L-3H-norepinephrine in essentialhypertension. Clin. Exp. Hypertens. 3:897-918.

20. Halter, J. B., A. E. Pflug, and A. G. Tolas. 1980. Arterial-venous differences of plasma catecholamines in man.Metab. Clin. Exp. 29:9-12.

21. Esler, M., G. Jackman, A. Bobik, D. Kelleher, G. Jen-nings, P. Leonard, H. Skews, and P. Korner. 1979. De-termination of norepinephrine apparent release rate andclearance in humans. Life Sci. 25:1461-1470.

22. Esler, M., G. Jackman, A. Bobik, P. Leonard, D. Kel-leher, H. Skews, G. Jennings, and P. Korner. 1981. Nor-epinephrine kinetics in essential hypertension. Defectiveneuronal uptake of norepinephrine in some patients.Hypertension. 3:149-156.

23. Esler, M., G. Jackman, P. Leonard, A. Bobik, H. Skews,G. Jennings, D. Kelleher, and P. Korner. 1980. Deter-mination of noradrenaline uptake, spillover to plasmaand plasma concentration in patients with essential hy-pertension. Clin. Sci. (Lond.). 59:311s-313s.

24. Callingham, B. A., and A. S. V. Burgen. 1966. The up-take of isoprenaline and noradrenaline by the perfusedrat heart. Mol. Pharmacol. 2:37-42.

25. Goldstein, D. S., G. Z. Feuerstein, J. L. Izzo, Jr., I. J.Kopin, and H. R. Keiser. 1981. Validity and reliabilityof liquid chromatography with electrochemical detec-tion for measuring plasma levels of norepinephrine andepinephrine in man. Life Sci. 28:467-475.

26. Edwards, A. L. 1967. Statistical Methods. Holt, Rinehart,& Winston, New York.

27. FitzGerald, G. A., V. Hossmann, and C. T. Dollery. 1981.Norepinephrine release in essential hypertension. Clin.Pharmacol. Ther. 30:164-171.

28. Grimm, M., P. Weidmann, G. Keusch, A. Meier, and Z.Gluck. 1980. Norepinephrine clearance and pressor ef-fect in normal and hypertensive man. Klin. Woch-enschr. 58:1175-1181.

29. Esler, M., G. Jackman, P. Leonard, H. Skews, A. Bobik,and P. Korner. 1981. Effect of norepinephrine uptakeblockers on norepinephrine kinetics. Clin. Pharmacol.Ther. 29:12-20.

30. Lake, C. R., M. G. Ziegler, M. D. Coleman, and I. J.Kopin. 1977. Age-adjusted plasma norepinephrine levelsare similar in normotensive and hypertensive subjects.N. Engi. J. Med. 296:208-209.

1758 Goldstein, Horwitz, Keiser, Polinsky, and Kopin