Diversity in supercoupling of β2-adrenergic receptors in orthostatic hypotension

Diversity in supercoupling ofPradrenergic receptors inorthostatic hypotension

Orthostatic hypotension is a clinical condition that frequently involves abnormal adrenergic control ofcardiovascular function. Adrenergic function was studied in six patients with symptomatic orthostatichypotension and in 11 age-matched healthy subjects. The patients demonstrated higher supine meanarterial pressures (MAP; 103 8 versus 86 ± 4 mm Hg) and orthostatic hypotension (AMAP 70 ± 5versus +15 2 mm Hg, p < 0.001) compared with normal subjects. The AMAP in phase II of theValsalva maneuver was significantly greater ( 31 ± 4 versus 7 ± 4 mm Hg, p < 0.002) and phase IVheart rate response was blunted ( 5 3 versus 30 ± 8 beats/min, p < 0.02) in these patients. Moreisoproterenol was required to increase heart rate by 25 beats per minute in patients with hypotension(810 670 versus 3.1 ± 1.3 p,g, p < 0.05), indicating marked chronotropic hyposensitivity. Leukocyte132-adrenergic receptor densities were similar in patients and controls. 132-Adrenergic receptor coupling,however, was elevated in patients with hypotension when compared with control subjects (ratio of thelow-affinity and high-affmity dissociation constants [1(L/K] 140 7.4 versus 66 ± 4.3, p < 0.001).There were negative correlations between the IQ /1(,, value and the dose of isoproterenol required todecrease MAP by 20 torr (p < 0.02) and between the KL/K. value and the product of the hormonereceptor and MAP (p < 0.01). However, the patients could be subdivided into a group who could mounta nearly normal hormone receptor times MAP response on standing (group 1A), and a group who couldnot (group 1B). The group lA patients had elevated plasma norepinephrine responses associated withmilder Vadrenergic receptor supercoupling, whereas group 1B patients had essentially no orthostaticplasma norepinephrine response and had much higher KL/K}, values. Thus, though a state of biochemicalsupersensitivity existed in both patient subgroups, diminished catecholamine exposure was associated, asexpected, with r32-adrenergic hypersensitivity in group 1B, whereas there was no diminution ofcatecholamine exposure in the 132-adrenergic hypersensitity observed in group lA patients. (CLINPHARMACOL THER 1990;47:371-81.)

Adolph Mares, Jr., MD, Albert O. Davies, MD, and Addison A. Taylor, MD, PhDHouston, Texas

Orthostatic hypotension hydroxy may result when thereflexes of the autonomic nervous system fail to main-tain normal blood pressure. In many patients it may be

From the Pulmonary Section, Department of Intemal Medicine, theSection on Hypertension and Clinical Pharmacology, Departmentof Internal Medicine, the Center for Experimental Therapeutics,the Department of Physiology and Molecular Biophysics, and theDepartment of Pharniacology, Baylor College of Medicine.

Supported by a research award (DK35113) from the National Insti-tutes of Arthritis, Metabolism, and Digestive Diseases, NationalInstitutes of Health, Bethesda, Md., an American Heart Associ-ation predoctoral fellowship (AM), and National Institutes ofHealth General Research Center Grant RR-00350.

Received for publication May 10, 1989; accepted Oct. 20, 1989.Reprint requests: Albert O. Davies, MD, Room 520B, Baylor College

of Medicine, One Baylor Plaza, Houston, Texas 77030.13 /1 /17520

difficult to identify the specific abnormalities of theautonomic nervous system that are responsible for theorthostatic hypotension.' Clinically, dysfunction of theautonomic nervous system may be a manifestation ofan underlying illness such as diabetes, or it may occuras an isolated phenomenon. Specific evaluation of ad-renergically mediated hemodynamic reflexes may beclinically useful in patients who show evidence of dys-function of the autonomic nervous system but who donot have an obvious underlying disorder.

Several studies have suggested that increased ad-renergic receptor responsiveness may be among thecardiovascular manifestations of altered autonomicfunction. For example, in patients with orthostatic hy-potension and atrophy of multiple systems, there isusually diminished release of norepinephrine from sym-

371

372 Mares, Davies, and Taylor

pathetic nerves.2-5 An increased sensitivity of adren-ergic receptormediated responses may occur, and thismay be compensatory. Increased pressor responses toinfusion of -adrenergic agonists may occur as a man-ifestation of ce-adrenergic hypersensitivity,6-8 accom-panied by increased ot-adrenergic receptor density.6,7,9,10P-Adrenergic hypersensitivity to infused P-adrenergicagonists6,11,12 is likewise accompanied by increased 32-

adrenergic receptor density'''''' in these patients.However, it is not clear whether the P-adrenergic hy-persensitivity is caused by these changes in receptordensity because changes in receptor function may beindependent of changes in receptor number.

The 3-adrenergic receptoradenylyl cyclase systemhas been shown to consist of at least three components:the receptor (R), the stimulatory nucleotide regulatoryprotein (Gs), and the catalytic moiety of adenylyl cy-clase. Hormone (H) first binds to the receptor and formsa low-affinity complex, HR, which subsequently reactswith the nucleotide regulatory protein to form the high-affinity ternary complex, HRGs (formerly termedHRI\is). I6"7 The ternary complex then activates adenylylcyclase. Hence, the stimulatory G-protein participatesin the coupling of receptor occupation with adenylylcyclase activation. '8.'9

Regulation of 3-adrenergic receptor action is de-monstrable as an alteration in the coupling of receptoroccupation to enzyme activation. The degree of cou-pling is estimated by the ratio of dissociation constantsK., and KM for the low- and high-affinity forms of thereceptor, respectively.16'20 The magnitude of the KL/KHratio directly corresponds to coupling with enzyme ac-tion, with a linear relationship between KJ KM and ac-tivation of adenylyl cyclase. Further, the ratio KL/KMalso corresponds to more distant coupling to physiologicaction, with a log-linear relationship between KL/ KH

and pharmacologic sensitivity to isoproterenol in humansubjects."'" However, in distinct contrast, changes inthe values of K, / K. do not necessarily or even usuallycorrespond to changes in biochemical or physiologicfunction.' Enhanced p-adrenergic sensitivity and cou-pling (increased KL/KH) may occur without a changein receptor density in a number of states, includingpatients with 3-adrenergic supersensitivity and in-creased adenylyl cyclase activation' or patients whohave been exposed to glucocorticoids.17.22''23 Desensi-tization, or diminished (3-adrenergic receptor action,may consist of either diminished receptor couplingalone (decreased K, / KH) or both diminished receptorcoupling and diminished receptor density."'' Thus thesequence of receptor occupation, formation of the high-affinity complex, and enzyme activation is a carefully

CLIN PHARMACOL THERMARCH 1990

controlled process whose regulation may be of majorsignificance in human disease.

We recently reported that (32-adrenergic receptor su-percoupling correlated with hypersensitivity to injectedisoproterenol in a group of patients with mitral valveprolapse (MVP) and with symptoms of 3-adrenergichypersensitivity, a syndrome termed MVP dysauto-nomia.2125 Those patients generally had elevated plasmanorepinephrine concentrations. In contrast, the presentstudy focused on the evaluation of patients with ortho-static hypotension, at least some of whom would beexpected to have diminished plasma norepinephrineconcentrations. In preliminary testing we noted thatmany of the orthostatic hypotension patients had ex-aggerated vasodilatory responses to isoproterenol thatwere similar to those observed in patients with MVPeven though the two groups are clinically quite distinct.Our hypothesis was that, in a group of patients withorthostatic hypotension that was not attributable toMVP dysautonomia, the orthostatic blood pressureresponse involved an element of vasodilatoryadrenergic hypersensitivity, corresponding to biochem-ical Vadrenergic receptor supercoupling.

MATERIAL AND METHODSPatients. The study protocol was approved by the

Insitituional Review Boards for Human Research ofBaylor College of Medicine and the Methodist Hospital,Houston, Texas. Written consent was obtained fromeach subject before testing. Patients were deemed eli-gible for this study on the basis of two major criteria:(1) the presence of the primary presenting complaint oforthostatic lightheadedness and (2) the absence of ex-clusionary criteria. This study was aimed at patientswho were more than 50 years old and whose primarymedical problems were unexplained orthostatic light-headedness , presyncope, or syncope. Thus many suchpatients came to us with labels such as "idiopathic or-thostatic hypotension." However, we found that, amongindividual patients, the physiologic and catecholaminefindings could change rather dramatically, and a varietyof terms (such as Shy-Drager and others) could be ap-plied to an individual patient at various points in time.We therefore determined to avoid labeling the patients,and we simply viewed them with regard to their symp-toms of orthostatic lightheadedness at the time of study.The patients selected generally had experienced symp-toms that included presyncopal or syncopal episodes,suggesting orthostatic hypotension, for months or foryears. Exclusion criteria were as follows: (1) evidencefor the MVP dysautonomia such as compatible history,compatible echocardiogram, or pharmacologic and

VOLUME 47NUMBER 3

Table I. Clinical and biochemical data for individual patients

*Clinical diagnosis: preganglionic or postganglionic.

physiologic responses consistent with MVP dysauto-nomia upon invasive quantitative autonomic functiontesting (see below)''' and (2) the presence of majorintercurrent illnesses known to cause orthostatic symp-toms. Orthostatic hypotension caused by volume de-pletion, myocardial conduction abnormalities, acutelyunstable hemodynamic processes, carotid sinus syn-drome, and autonomic dysfunction secondary to otheridentifiable diseases such as amyloidosis, diabetes mel-litus, or malignancy were thus specifically excluded.

Two women and four men between the ages of 57and 89 years (mean age, 71 ± 5 years) met these cri-teria and were recruited as they came to the Hyperten-sion and Clinical Pharmacology Clinic at Baylor Col-lege of Medicine (Houston, Texas) for evaluation. Asummary of some of the clinical and biochemical dataobtained from these individuals is contained in TableI. Medications known to affect the adrenergic nervoussystem were carefully withdrawn, usually 2 weeks be-fore testing. Eleven healthy subjects (two women andnine men) between the ages of 42 and 89 years (meanage, 59 ± 4 years; p = NS versus patients) who werehistorically free of unexplained orthostatic lighthead-edness, presyncope, or syncope were recruited as age-matched normal control subjects. Once recruited for thestudy, both patients and control subjects underwentscreening examination of their systolic blood pressureresponse to standing. Patients manifested at least a 10ton orthostatic systolic blood pressure decrease duringthe screening visit, and clinic and hospital records in-dicated abnormal orthostatic systolic blood pressure de-creases on previous visits. Control subjects had littledecline or an actual increase in blood pressure afterstanding.

The group of patients was thus clearly distinct fromthe control group, as well as distinct from other patientswith explained cardiovascular dysautonomias. No sat-isfactory specific term exists for our heterogeneous

Diversity of supercoupling of p-receptors 373

group of patients but, for simplicity in this report, wewill refer to them as the orthostatic hypotension group.

Evaluation of autonomic control of cardiovascularfunction. Subjects were studied in the Baylor Auto-nomic Testing Laboratory of the Methodist Hospital(Houston, Texas). A venous catheter was placed in aforearm vein and a radial artery was catheterized fordirect blood pressure measurements. Patency of bothcatheters was maintained by periodic flushing with 10units per milliliter of sodium heparin in normal salinesolution. The electrocardiographic tracing was moni-tored continuously. Phasic arterial pressure was mon-itored by a Statham P23ID pressure transducer attachedto the chest at the level of the right atrium and connectedto a Gould 2800 series ink-writing recorder (Gould,Cleveland, Ohio). Electronically derived mean arterialpressure, electrocardiographic tracing, and heart ratederived through a biotach preamplifier were each re-corded separately.

Physiologic cardiovascular responses. A venousblood sample was obtained from each subject (whilerecumbent) for measurement of catecholamines and for(3-adrenergic receptor analysis. The subject then stoodquietly for 5 minutes or for as long as could be tolerated.A second blood sample was obtained for catecholaminedetermination. Blood pressure and heart rate were mea-sured while subjects were in both supine and standingpositions.

Valsalva maneuver. During continuous recording ofelectrocardiogram, heart rate, and intra-arterial bloodpressure, each subject forcibly exhaled into a mouth-piece attached to a mercury manometer, maintaining atransthoracic pressure of 40 ton throughout the Valsalvamaneuver. Maximal changes in blood pressure and heartrate during (phase II) and immediately after (phase IV)the maneuver were compared with pre-Valsalva (phase0) values.

Pharmacologic stimulation tests. The oc-adrenergic

Supine Standing Supine StandingPatient Age norepinephrine norepinephrine epinephrine epinephrine

no. yr. PrelPost* (pglml) (pglml) (pglml) (pg1m1) KLIKH

1 89 ? NA NA NA NA 1892 76 ? 612 1223 23 15 1303 75 Pre 917 2372 90 102 1204 68 Pre + Post 86 69 NA NA NA5 59 Post 283 274 35 22 2586 57 Pre then Post 50 64 111 136 443

3 74 Mares, Davies, and Taylor

agonist, phenylephrine, was given by bolus injectionin increasing pulse doses of 10, 25, 50, 75, and 100p,g until a blood pressure increment of 25 ton- wasobserved or until the subject experienced drug-relatedobjectionable symptoms. The I3-adrenergic agonist,isoproterenol, was given in increasing pulse doses of0.1, 0.2, 0.5, 1.0, 1.5, and 2.0 11,g until a heart rateincrement of 20 to 30 beats per minute was observedor until the subject experienced drug-related objection-able symptoms. Hemodynamic values were allowed toreturn to baseline values before injection of the sub-sequent agonist dose.

Catecholamine measurements. Plasma norepineph-rine and epinephrine concentrations were determinedby electrochemical detection after high performance liq-uid chromatographic separation by use of a minor mod-ification of the method of Goldstein et al."

Neutrophil I32-adrenergic receptor assay. The meth-ods for preparing and assaying polymorphonuclear leu-kocyte membranes for Pradrenergic receptors havebeen described previously. 22 Radioiodinated cyanopin-dolol (1'211-CYP, 2200 Ci/ mmol; Amersham Corpo-ration, Arlington Heights, Ill.) was used. Briefly,neutrophils were isolated from heparinized blood, anda plasma membrane preparation that contained f32-adrenergic receptors was obtained by polytron actionand differential centrifugation. The binding assaywas performed in a total volume of 150 p,1 that con-

CLIN PHARMACOL THERMA_RCH 1990

tained 25 RI each of cyanopindolol and either 6 x 10'mol/L ascorbic acid or a series of dilutions of ( )-isoproterenol in 6 x 10' mol/L ascorbic acid. Ap-propriate isoproterenol concentrations are indicated inthe figures. The mixture was incubated for 40 minutesat 37° C, and the reaction was terminated by rapidfiltration and washing. The filters were counted at ap-proximately 80% efficiency.

Receptor density. 32-Adrenergic receptor densitywas determined from saturation binding curves and re-ported as femtomoles per milligram of protein iso-lated. 18'22

132-Adrenergic receptor binding data analysis. Com-petition binding curves were analyzed individually forthe orthostatic hypotension group and the controlgroup. The details of this analysis are elaborated else-where. 16,29,20 Brie,. yn iterative curve modeling tech-niques were used to fit equations expressing the law ofmass action to the empiric competition binding curves,which resulted in estimates of the binding parameters,including binding affinities and receptor amounts.'"This procedure tested for the existence of one or morebinding states of the receptor. The several fits werecompared by F test to search for improvement ofgoodness-of-fit by successively more complicated mod-els. A two-state model was accepted only if it statis-tically significantly improved the fit over the one-statemodel. When two-state models were observed, their

50

o

MAP HR

30 30EE

cc7 10-co

turzo.

A

+oni-e10 >o

A_t -10 -10 xlicç

A

pÇ 0.01 o

-30 -30 3

-50

-70 fOH CON OH CON

Fig. 1. Mean arterial pressure and heart rate responses to change of posture in patients withorthostatic hypotension (OH) and in control subjects (CON). Change in mean arterial pressure andheart rate from the resting supine to the standing positions in individual subjects are depicted asMAP and HR, respectively. Open triangles represent group IA patients and closed trianglesrepresent group 1B patients (see Fig. 2). Mean values ± SEM (and a p value) are shown for group1B patients (closed triangles) and control subjects (closed squares).

dissociation constants were designated K. for the high-affinity state and KL for the low-affinity state. The ratioof the dissociation constants (K, /K.) could then becalculated. It has previously been shown that the mag-nitude of this parameter is a correlate of high-affinitystate (HRG,) formation and reflects coupling of receptoroccupation with activation of adenylyl cyclase.

Statistical analysis. Statistical analyses of biochem-ical parameter estimates were conducted as previouslydescribed." Physiologic data were compared by use ofthe unpaired Student t test or ANOVA. Dose-responsecurve data were subjected to linear regression analysis.Statistical analyses were performed by use of BMDPstatistical programs (BMDP Statistical Software, Inc.,Los Angeles, Calif.) on a DEC PDP-11 computer (Dig-ital Equipment Corp., Maynard, Mass.).

RESULTS

The patient population. Table I contains a summaryof the clinical and biochemical data for the patients inthis study. The third column represents the clinicalimpression at the time of the autonomic function stud-ies. In two patients, the clinical picture did not conformto published criteria for any particular named dysau-tonomia. In two patients a simple assignment of pre-ganglionic or postganglionic defect could be made. Inone patient who had been followed for a number ofyears, there were elements of both "central" and "pe-ripheral" defects simultaneously; therefore the term,"pre and post" was assigned. Finally, one patient whohad also been observed for a number of years evolvedfrom normal to "central" to "peripheral" defects andthus was assigned the term, "pre then post."

Physiologic cardiovascular responses. The individ-ual hemodynamic responses to a change of posture inpatients with orthostatic hypotension and in control sub-jects were significantly different. The resting supinemean arterial pressure (MAP) of the patients was higherthan that of control subjects (103 ± 8 versus 86 ± 4mm Hg), and four of six patients were hypertensive,whereas the mean heart rates for the two groups werenot different (67 ± 4 versus 70 -± 2 beats / mm). Therewas a significant orthostatic mean arterial blood pres-sure reduction in patients compared with control sub-jects (standing MAP, 57 ± 14 versus 103 -± 6 mm Hg;p < 0.01) representing a mean decrease of 37 ± 15mm Hg in the patients compared with a mean incrementof 17 3 mm Hg in control subjects (p < 0.01; Fig.1). These blood pressure recordings were generallymade after the patient had stood for 5 minutes or whenthe patient could no longer stand if orthostasis wasprominent. When the MAP remained near baseline, wetypically observed the systolic blood pressure decrease

o

supine standing

OHn.6

supine standing

CONn.7

Fig. 2. Product of heart rate and mean arterial pressure beforeand after standing in patients with orthostatic hypotension(OH) and in control subjects (CON). Measurements weretaken at rest and after standing quietly for 5 minutes or untilthe patients became symptomatic. Open triangles representgroup lA patients, with a postural rise in HR MAP product.Closed triangles represent group 1B patients, with a posturaldecline in HR MAP product. Mean values -± SEM are re-ported for control subjects.

to be associated with a diastolic blood pressure increase.This pattern was not observed in control subjects,whose diastolic blood pressures typically rose in concertwith a systolic blood pressure increase. The three pa-tients with profound orthostatic decrease failed to mounta substantial increase in diastolic blood pressure.

To evaluate both cardiac and vascular elements ofthe physiologic response to change of posture, the prod-uct of the heart rate (HR) and MAP was determined.The products of these physiologic responses in boththe supine and standing positions for the patients andcontrol subjects are illustrated in Fig. 2. This mathe-matic manipulation reflected the degree of heart rate-dependence in the successful maintenance of bloodpressure after standing, and it identified two distinctsubgroups within the patient population. Group IA

VOLUME 47NUMBER 3 Diversity of supercoupling of 13-receptors 375

376 Mares, Davies, and Taylor

SUPINE STANDINGCON

na6

Fig. 3. Plasma norepinephrine concentration before and afterstanding in patients with orthostatic hypotension (OH) and incontrol subjects (CON). Blood samples were obtained at restand after standing quietly for 5 minutes or until the patientsbecame symptomatic. Open triangles, group IA patients;closed triangles, group 1B patients. Mean values ± SEMare reported for control subjects. Values for one patient ingroup 1 A were unavailable.

manifested some compensatory increase in heart ratewhen standing blood pressure decreased, resulting innearly normal products of standing HR MAP. Group1B had a conspicuously absent heart rate response tostanding, with much lower standing HR MAP prod-ucts. Notably, the supine HR MAP product washigher in group 1B than in group 1A. In spite of thesephysiologic distinctions between these patient sub-groups, patients could not be distinguished by otherclinical signs or symptoms.

As a whole group, patients with orthostatic hypo-tension had about the same supine plasma norepineph-rine concentrations as the control subjects (Fig. 3).However, patients in group 1 A tended to have highervalues than their counterparts in group 1B. Afterstanding, patients in group 1 A had remarkably greaterincrements in plasma norepinephrine concentrations

CLIN PHARMACOL THERMARCH 1990

than did the control subjects (1000 rt 430 versus480 ± 51 pg/ ml), and both were greater than group1B (-13 -± 3 pg/ml).

Valsalva maneuver. The reduction in MAP duringphase II of the Valsalva maneuver compared with base-line (phase 0) values was greater in the patient groupthan in the group of normal control subjects (AMAP,31 -± 4 versus 7 -± 4 mm Hg; p <0.003),whereas the increment in heart rates in patients andcontrol subjects were similar Although the change inMAP from phase II to phase IV was similar in bothpatients and control subjects, the bradycardic responsewas significantly diminished in the patient group( 5 -± 3 versus 30 -± 8 beats/min; p < 0.02).

Pharmacologic stimulation. ec-Adrenergic receptorhypersensitivity was not detected in most of these pa-tients with orthostatic hypotension. The average dose

OH CONn=6 n=7

Fig. 4. Dose of isoproterenol required to decrease mean ar-terial pressure by 20 mm Hg (IS08,20) in patients with ortho-static hypotension (OH) and in control subjects (CON). Opentriangles, group 1 A patients; closed triangles, group 1B pa-tients. Mean dose ± SEM is reported for control subjects.

SUPINE STANDINGOHn=5

of phenylephrine required to increase MAP by 25 mmHg (PHE25) was similar in patients and in controlsubjects (5.3 2.1 versus 8.6 ±- 3.5 1.tg/ kg bodyweight). However, two patients in group 1B were foundto have PHEHp25 values of 0.1 and 0.34 Rg/kg bodyweight, respectively. The greater than 10-fold decreasein dose represents aradrenergic hypersensitivity inthose two patients.

The response of heart rate to isoproterenol was di-minished in the patients with orthostatic hypotensioncompared with control subjects. The dose of isopro-terenol required to increase heart rate by 25 beats/minfrom baseline was 810 ± 670 p.g in patients versus3.1 ± 1.3 1,tg in control subjects (p <0.05). In grouplA patients the vasodilatory effect of isoproterenol,assessed as the dose required to lower blood pressureby 20 mm Hg (ISOop2o), was similar to that of controlsubjects. In contrast, the ISOBp2o values for patients ingroup 1B were much lower, indicating an augmentedsensitivity to isoproterenol (Fig. 4).

Neutrophil 132-adrenergic receptor analysis. 112-

Adrenergic receptor parameters were measured in fivepatients with orthostatic hypotension and in four controlsubjects (only two of these control subjects also con-sented to undergo physiologic studies). 132-Adrenergicreceptor densities derived from saturation binding datawere similar between patients with orthostatic hypo-tension and control subjects (25 ± 7 versus 20 ± 5fmol / mg protein). Similary, the mean antagonist affin-ity (KJ estimates were statistically indistinguishablefrom one another (22 3 pmol/L versus 23 ±- 5pmol/L).

Coupling of the 132-adrenergic receptor was assessedby estimation of the ratio KL/KH derived from ligandcompetition data. The composite competition curvesfor the patients with orthostatic hypotension and for thecontrol subjects are illustrated in Fig. 5. The compositecompetition curve for the normal control subjects wasfairly typical of other normal control (young adults)groups"' and for our control elderly population.' Notethat the control curve begins its downslope at a muchhigher ligand concentration, has a less shallow slope,and is located to the right of the patient composite curve;computer analysis of this ligand binding curve is con-sistent with a two-state model. The composite com-petition curve for the data obtained from the patientswith orthostatic hypotension has the characteristics ofa rapid downsloping at low concentrations of competingligand, a shallow slope, and a position at the far leftof the graph. Iterative computer curve modeling anal-ysis of this ligand binding curve also detects a two-state model. However, the magnitude of the KL/KE{ ratio

100

75-

50-

25-

.o 1:0 2!0 31:1 41.0

Log [Isoproterenol], nM

Fig. 5. Polymorphonuclear leukocyte r32-adrenergic receptorcoupling in patients with orthostatic hypotension (circles) andin control subjects (squares). Abscissa respresents the con-centration of isoproterenol competing for [`"I]-CYP bound.Curves represent composites obtained by use of each indi-vidual competition curve from each of the groups and are thecomputer-generated best fits to the data

in the patients is more than twice the magnitude ofthe KL/KH ratio determined for the control subjects(140 ± 7.4 versus 66 ± 4.3;p < 0.001). The controlKL/KH ratio is very similar to previously determinedvalues.'"

The individual competition binding curves for eachpatient and control subject were also separately exam-ined, and mean K, /K. values were derived for patientand control groups. Again, supercoupled Vadrenergicreceptors were noted in the patients with orthostatichypotension, and the mean of individual K, /K. ratioswas 190 -±- 28 versus 58 -± 7; p <0.01 (Fig. 6).

Correlation of pharmacologic and biochemical as-sessments. To assess the relationship between the doseof isoproterenol required to decrease blood pressure by20 mm Hg and (3-adrenergic receptor coupling, regres-sions were performed. This analysis involved the in-dividual values for both ISOH,20 (the dose of isopro-terenol required to decrease blood pressure by 20 mmHg) and K,/KM determined in the patients with ortho-static hypotension. The pharmacologic index of 132-adrenergic receptor sensitivity (ISOH,N) was linearlycorrelated with the biochemical parameter (KL/KH)used to assess P-adrenergic receptor coupling (r =0.94, p <0.02). This linear relationship is similarto that of patients with mitral valvular prolapse andsymptoms of P-adrenergic dysfunction.'

In addition, a general relationship was noted in the

VOLUME 47NUMBER 3 Diversity of supercoupling of 0-receptors 377

50 60

100 _

0_

_10._

a10-200

300

400A

500

CLIN PHARMACOL THERMARCH 1990

50 100 150 200 250

KL/KH

Fig. 7. Relationship of biochemical sensitivity (KL/K.)to physiologic responsiveness (HR MAP/ ANE). The ab-scissa represents the individual KL/KH values for controlsubjects (closed squares), group lA (open triangles), andgroup 1B (closed triangles) patients. The ordinate representsHR MAP/ ANE, in units of beats ton milliliter perminute picograms. Linear regression parameters are

r = 0.98;p <0.01.

ANE increased even more, and the physiologic respon-siveness (HR MAP/LINE) decreased relative to con-trol subjects (Fig. 7). In group 1B the HR MAP prod-uct decreased, but the ANE was miniscule or evennegative, so the ratio HR MAP/ LINE decreased evenmore. There was a negative yet statistically significantcorrelation (p < 0.01) between physiologic respon-siveness (HR MAP/ ANE) and biochemical sensitiv-ity (IC., / KH), as seen in Fig. 7. Noting that this phys-iologic responsiveness relates to maintaining or increas-ing vasoconstriction and that the biochemical sensitivityrelates to vasodilatory Pradrenergic receptors, the in-verse relationship is comprehensible. We are not as-serting that, with six patients, this relationship is de-finitively linear; we are simply reporting the correlationbetween the parameters and noting that a linear ap-proximation is statistically satisfactory.

DISCUSSIONThis study demonstrates in a population of somewhat

older patients with otherwise unexplained orthostatichypotension that physiologic abnormalities of bloodpressure and heart rate control are associated with bio-chemical abnormalities of 32-adrenergic receptor cou-pling. The patient group was not only diverse betweenindividuals, several had been diverse with respect tothemselves as they had been monitored for a numberof years before the study. The consistent clinical featurewas a primary and otherwise unexplained orthostaticlightheadedness that was severe enough for the patient

378 Mares,

350-

300-'

250-

200-

150-

100-

50 -

Davies, and Taylor

AA

* p< 0.05

OH CONn.5 n=4

Fig. 6. Polymorphonuclear leukocyte Pradrenergic receptorcoupling in patients with orthostatic hypotension (OH) andin control subjects (CON). Coupling is assessed by the ratioof the dissociation constants of the high-affinity and low-affinity states of the receptor (KL/KH). Open triangles, grouplA patients; closed triangles, group 1B patients. Error barsrepresent mean values ± SEM.

two orthostatic hypotension patient subgroups, depictedby the HR MAP product and the magnitude of K, / KH.

Group 1A, which at least partially compensated forchange in posture, was associated with K, /KH ratiosranging from 120 to 190 (Fig. 6). Group 1B, whichcould not mount an effective postural response, wasassociated with KL/KH ratios whose magnitudes weremuch greater, ranging from 240 to 260. Importantly,group 1B patients had both a smaller change of plasmanorepinephrine in response to change of posture and agreater sensitivity to the vasodilatory effect of isopro-terenol .

The standing HR MAP parameter reflects the ef-fectiveness of cardiovascular reflexes to defend againsta gravitational challenge. This effectiveness correlatedto the biochemical sensitivity as reflected in the param-eter, KL/ KH (r = 0.98). Physiologic responsivenesscould be conceived as the effectiveness of response(HR MAP) relative to the postural rise in circulatingnorepinephrine associated with, and perhaps requiredto produce, that response. This physiologic respon-siveness could be expressed as the value: HRMAP/ ANE, in which NE is norepinephrine. In grouplA patients, HR MAP increased after standing, but

to seek medical attention. These patients demonstratedphysiologic responses that were diverse with respect toeach other, but the patients as a group were quite easilyclinically distinguishable from both normal subjects andpatients with other types of dysautonomias. When com-pared with patients with the dysautonomia of MVP,these patients were older, were more free of other symp-tomatic complaints, did not have valvular prolapse, andhad distinct physiologic patterns during examination ofthe autonomic nervous system (including diminishedor absent orthostatic cardioacceleration and exaggeratedhypotensive but diminished bradycardic responses toValsalva maneuver).

Within the patient group, two subgroups could beidentified on the basis of orthostatic cardioaccelerationand blood pressure responses. Once subgrouped assuch, it became clear that those who were least able tomaintain blood pressure were least prone to have acardioacceleration response. These patients also had theleast catecholamine response, had the most vasodilatoryhypersensitivity, and had the greatest biochemicalsupercoupling. In viewing the relationship betweenphysiologic responsiveness and biochemical sensitivity( Fig. 7), the possibility is raised that in both groups1 A and 1B there was an underlying common processof progressive 132-adrenergic receptor hypersensitivity.In this view, the increased catecholamine response tostanding represents an almost effective attempt to main-tain perfusion; as this response begins to fail further,the vasodilatory hypersensitivity becomes most prom-inently expressed physiologically. One would then pos-tulate that the process of release of catecholamines wasconnected in some way to the process of developinghypersensitivity. Denervation hypersensitivity is rea-sonably a candidate mechanism, particularly if periph-eral adrenergic neurons are losing their release capacityearlier than are the adrenals, and the predominantsource of the group IA norepinephrine is adrenal. Inter-estingly, it was our clinical impression that the patientsin group 1 A had been ill for a shorter duration or hadslower clinical progression than those in group 1B .

Our patients were also diverse with regard to boththeir a-adrenergic sensitivity and their blood pressures.Previous investigations have reported that supine hy-pertension occurs in some patients with autonomic in-sufficiency.' Findings in the present group support thisobservation. Hypersensitivity to infused a-adrenergicagonists in some patients with autonomic insufficiencyhas also been observed by other investigators6-8 but wasnot a consistent feature in our patients.

The vasodilatory and cardioacceleration patterns arenot identical within the patient groups. In contrast to

the enhanced vasodilation in the present group of pa-tients with orthostatic hypotension, the chronotropicresponse to bolus infusions of isoproterenol was mark-edly diminished. Cardioacceleration in response to bothphase II of the Valsalva and standing was also dimin-ished, particulary when viewed relative to the decreasein blood pressure. This pattern of diminished reflexcardioacceleration is not consistent with the findings inplasma catecholamines, vasodilation, or I3-adrenergicreceptor binding, suggesting there may be an additionalmechanism operating in addition to f32-adrenergic re-ceptor regulation.

r3-Adrenergic receptor radioligand binding studieshave provided a biochemical method of assessing thesensitivity of the ri-adrenergic receptor to stimulationby hormone and determining whether such sensitivitychanges are caused by altered receptor numbers or al-tered coupling. The 132-adrenergic receptor sensitivitycan be modulated by many extrinsic factors.17,21 24 wehave recently demonstrated supercoupled (32-adrenergicreceptors in patients with mitral valvular prolapse andp-adrenergic hypersensitivity.' By use of peripheralpolymorphonuclear leukocyte 3-adrenergic receptors asa model, f32-adrenergic receptor supercoupling wasdemonstrated biochemically (elevated IC, / K.) in thecurrent study. The credibility of this model as an in-dicator of cardiovascular ß-adrenergic receptor functionhas recently been established in a group of patients withmitral valvular prolapse with symptoms of 13-adrenergichypersensitivity with the existence of 13-adrenergic re-ceptor hypersensitivity both pharmacologically and bio-chemically.' Clinically, the hypersensitivity was man-ifested chiefly as orthostatic tachycardia, with or with-out orthostatic hypotension. In this particular group ofMVP patients, the dose of isoproterenol required toincrease heart rate by 25 beats per minute was shownto be inversely correlated with the K,/ ratio beforeand after exposure to an infusion of isoproterenol suf-ficient to produce desensitization. The pattern of (32-adrenergic receptor supercoupling in the current groupof patients with orthostatic hypotension was similar tothat seen in our previous group of MVP patients withsymptoms of r3-adrenergic supersensitivity; namely, allhad abnormally elevated KL/1(}, ratios. The statisticallysignificant correlation between the ISO., and K, / KHshould be emphasized because it lends support to thehypothesis that 3-adrenergic receptor coupling, as de-termined by radioligand binding techniques, mirrors thephysiologic and pharmacologic changes observed indiseases that involve the 13-adrenergic system.

Previous investigations have shown a relationship be-tween modulation in 3-adrenergic receptor densities

VOLUME 47NUMBER 3 Diversity of supercoupling of p-receptors 379

380 Mares, Davies, and Taylor

and catecholamine concentrations .6,11,14,15,30 Indeed, thisphenomenon has been thought to contribute signifi-cantly to the denervation hypersensitivity observed inpatients with autonomic dysfunction. A comparison ofthe magnitudes of receptor density changes and cou-pling changes obtained in this investigation suggeststhat the changes in coupling are more reflective ofthe physiologic and pharmacologic observations in thein vivo setting. The present data support our previousconjecture that changes in r3-adrenergic activity is bettermanifest in changes in fl-adrenergic receptor couplingthan in receptor density.2,17,21,23,29

In summary, this investigation reports biochemicalalterations in the 32-adrenergic receptor coupling in pa-tients with autonomic nervous system dysfunction man-infested as orthostatic hypotension. Specifically, 132-adrenergic receptor supercoupling is identified, whichcorresponds closely to the pharmacologic 13-adrenergicsupersensitivity observed in these patients. These bio-chemical and physiologic abnormalities of the 13-

adrenergic receptor system may be integral to the patho-physiology of the orthostatic hypotension and symp-toms experienced by some patients.

We express our sincere gratitude to Cathy Rosene, RN,for her nursing assistance, to Laurie Schneider, Wayne Wang,and Georgeanne Massey for their technical assistance in thereceptor laboratory, and to Rebecca Fullen and Duana Jonesfor editorial assistance.

ReferencesReid JL. Central and peripheral autonomic control mech-anisms. In: Bannister R, ed. Autonomic failure: a text-book of clinical disorders of the autonomic nervous sys-tem. Oxford: Oxford University Press, 1983:17-34.Kontos HA, Richardson DW, Norvell JE. Norepineph-rine depletion in idiopathic orthostatic hypotension. AnnIntern Med 1975;82:336-41.Ziegler MG, Lake CR, Kopin U. The sympathetic-nervous-system defect in primary orthostatic hypoten-sion. New Engl J Med 1977;296:293-7.Kopin U. Plasma levels of norepinephrine. Ann InternMed 1978;88:671-80.Nanda RN, Boyle FC, Gillespie JS, Johnson RH, KeoghHJ. Idiopathic orthostatic hypotension from failure ofnorepinephrine releases in a patient with vasomotorinjnervation. J Neurol Neurosurg Psychiatry 1977;40:11-9.Chobanian AV, Tifft CP, Sackal H, Pitruzella A. Alphaand beta adrenergic receptor activity in circulating bloodcells in patients with idiopathic orthostatic hypotensionand pheochromocytoma. Clin Exp Hypertens [A] 1982;4:793-806.

CLIN PHARMACOL THERMARCH 1990

Davies B, Sudera D, Sagnella G, et al. Increased numbersof alpha receptors in sympathetic denervation supersen-sitivity in man. J Clin Invest 1982;69:779-84.Robertson D, Hollister AS, Carey EL, Tung C, Gold-berg MR, Robertson RM. Increased vascular beta,-adrenoceptor responsiveness in autonomic dysfunction.J Am Col Cardiol 1984;3:850-6.Davies JB, Sudera D, Sever PS. Endogenous angonistregulation of alpha-adrenoceptors in man. Clin Sci1981;61:207S-10S.Kafka MS, Polinsky RJ, Williams A, et al. Alpha-adrenergic receptors in orthostatic hypotension syn-dromes. Neurology 1984;34:1121-5.Bannister R, Boylston AW, Davies IB, Mathias CJ, SeverPS, Sudera D. Beta-receptor numbers and thermodynam-ics in denervation hypersensitivity. J Physiol (Lond)1981;319:369-77.Man in't Veld AS, Schalekamp MADH. Endogenousand exogenous agonist regulation of responses to beta-adrenoceptor stimulation in patients with chronic auto-nomic failure. Clin Exp Pharmacol Physiol 1982;9:271-5.Davies B, Sudera D, Mathias C, Bannister R, Sever P.Beta receptor on orthostatic hypotension. New Engl JMed 1981;305:1017-9.Hui KKP, Conolly ME. Increased numbers of beta re-ceptors in orthostatic hypotension due to autonomic dys-function. New Engl J Med 1981;304:1473-5.Wood AAJ, Feldman R, Nadeau J. Physiological regu-lation of beta-receptors in man. Clin Exp Hypertens [A]1982;4:807-17.Kent R, DeLean A, Lefkowitz RJ. Quantitative analysisof beta-adrenergic interactions: resolution of high andlow affinity states of the receptor by computer modelingof ligand binding data. Mol Pharmacol 1980;17:14-23.Stadel JM, DeLean A, Lefkowitz RJ. Molecular mech-anisms of coupling in hormone receptor adenylyl cyclasesystems. Adv Enzymol 1982;53:1-43.Stadel J, DeLean A, Lefkowitz RJ. A high affinity ag-onist beta-adrenergic receptor complex is an intermediatefor catecholamine stimulation of adenylyl cyclase in tur-key and frog erythrocyte membranes. J Biol Chem1980;255:1436-41.DeLean A, Stadel JM, Lefkowitz RJ. A ternary complexmodel explains the agonist specific binding properties ofthe denylate cyclase coupled beta adrenergic receptor.J Biol Chem 1980;255:7108-17.Davies AO, Lefkowitz RJ. Agonist promoted high affin-ity state of the beta adrenergic receptor in human neu-trophils: modulation by corticosteroids. J Clin EndocrinolMetab 1981;53:703-8.Davies AO, Mares A, Pool JL, Taylor AA. Mitralvalvular prolapse with symptoms of beta adrenergichypersensitivity -beta2 adrenergic receptor supercou-pling with desensitization upon isoproterenol exposure.Am J Med 1987;82:193-201.

Davies AO, Lefkowitz RJ. Corticosteroid-induced dif-ferential regulation of beta-adrenergic receptors incirculating human polymorphonuclear leukocytes andmononuclear leukocytes. J Clin Endocrino! Metab 1980;51:599-605.Davies AO, Lefkowitz RJ. In vitro desensitization of betaadrenergic receptors in human neutrophilsattenuationby corticosteroids. J Clin Invest 1983;71:565-71.Leflcowitz RJ, Caron MG. Adrenergic receptors molec-ular mechanisms of clinically relevant regulation. ClinRes 1985;33:395-405.Taylor AA, Davies AO, Mares A, et al. Spectrum ofdysautonomia in mitral valvular prolapse. Am J Med1989;86:267-74.Goldstein DS, Feuerstein G, Izzo JL, Kopin IJ, KeiserHR II. Validity and reliability of liquid chromatographywith electrochemical detection for measuring plasma

levels for norepinephrine and epinephrine in man. LifeSci 1981;28:467-75.DeLean A, Hancock AA, Lefkowitz RJ. Validation andstatistical analysis of a computer modeling method forquantitative analysis of radioligand binding data for mix-tures of pharmacological receptor subtypes. Mol Phar-macol 1982;21:5-16.Montamat SM, Davies AO. Physiological responses toisoproterenol and coupling of beta-adrenergic receptorsin young and elderly human subjects. J Gerontol.Bannister R. Clinical studies of autonomic function anddysfunction. J Auton Nerv Syst 1983;7:233-7.Frazer JA, Nadeau HJ, Robertson D, Wood JJ. Downregulation of leukocyte beta adrenoceptor density by cir-culating plasma levels of catecholamines in man. J ClinInvest 1981;67:1777-84.

VOLUME 47NUMBER 3 Diversity of supercoupling of p-receptors 381