immunological probes incardiovascular disease*immunologicalprobes in cardiovascular disease...

Br Heart J 1982; 47: 1-10

Immunological probes in cardiovascular disease*EDGAR HABERFrom the Cardiac Unit, Massachusetts General Hospital and HarvardMedical School, Boston, Massachusetts, USA

SUMMARY The immune system has long been recognised as playing a central role in the organism'sdefence against infectious diseases and possibly the development of neoplasia. The active stimula-tion of the immune system by immunisation and the passive administration of antitoxins have a

venerable history in medicine. Yet the concept that antibodies may be used to modify physiologicalor pharmacological effects or may act as diagnostic agents in the living organisms has only recentlycome to be recognised. Advances, both in an understanding of the structural chemistry of theantibody molecule and in the ability to culture antibody-producing cells, now permit the selectionand production of homogeneous antibodies and their smaller fragments in quantity by means otherthan conventional immunisation. These innovations will allow the development of a new phar-macology based on the remarkable resolving power of the antibody combining site. Antibodies or

their fragments are shown to inhibit the pressor action of renin, to neutralise the pharmacologicalactions of digitalis, to block the beta-adrenergic receptor, and to detect and image myocardialinfarcts.

Antibodies represent a set of molecules that exhibitextraordinary specificity and selectivity. This hasbeen appreciated since the work of Landsteiner,'who, in the early years of this century, was able toprovide examples such as the ability of an antiserumto resolve the d and 1 stereoisomers of a simple alipha-tic molecule, tartaric acid. More complex organiccompounds that vary only by a single substituentgroup (such as the steroids, or digoxin and digitoxin,which differ in one hydroxyl group) may be readilyseparated.2 The general structure of the antibodymolecule is remarkably conserved, so that it isdifficult for the chemist to tell the difference betweentwo antibodies of varying specificity. There is a smallregion of the molecule, however, that is highly vari-able in its structure (the complementarity region) andmakes up the surface that binds antigen. It is theamino acid sequence in this complementarity regionthat determines the nature of the antigen recognised.Recent work has uncovered the mechanisms respon-sible for variation in the amino acid sequence of thecomplementarity region.The antibody molecule comprises four polypeptide

chains, two identical light and heavy chains. The vari-able region of the light chain is the product of two*Modified from the text of the International Lecture of the International Societyand Federation of Cardiology delivered at the Royal College of Physicians, Lon-don on 4 December 1980.

Accepted for publication 7 September 1981

genes: V, which occurs in several hundred copies, andJ, of which four copies have been identified.3-6 V andJ of the light chain may occur in any combination.The heavy chain's variable region is the product ofthree genes: V, in several hundred copies, D, in 10 ormore copies, and J, in four copies.78 As in the lightchain, any permutation and/or combination of thesegenes may occur. These mechanisms alone accountfor 107 different antibodies. In addition, somatic-point mutation has been shown to occur. Thus thenumber of possible antibodies must actually exceed1010.

Clinicians have not neglected the immune system,but its deliberate manipulation has been restricted to alimited number of areas. Of course, immunisation haslong been a cornerstone of preventative medicine.The passive infusion of pneumococcal-specific anti-sera had a brief vogue in pre-antibiotic days, whiletetanus and Rh antibodies are still of great valuetoday. By providing a method for measuring sub-stances in biological fluids that would not be possibleby any other means, the immunoassay has produced aliteral explosion in the use of antibodies as in vitrodiagnostic agents. Yet, aside from these few uses, thisremarkable group of compounds has been largely neg-lected in its potential utility as a source of phar-maceutical agents with both diagnostic and therapeu-tic applications.There may be several explanations for this neglect.

1

on Novem

ber 17, 2020 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.47.1.1 on 1 January 1982. D

ownloaded from

http://heart.bmj.com/

Haber

Before very recent times, antibodies were only avail-able as complex mixtures of proteins, isolated in verylimited quantities from animal sera. Standardisationof an antibody is not possible since each immunisedanimal provides an antiserum of different properties.There is even variation among specimens of sera ofsingle animals at different times after immunisation.An additional problem is the potential immunogenic-ity ofheterologous antibodies in the recipient. Each ofthese problems has been overcome by application of anew technology that has only been available in thepast several years. Antibodies may now be producedby cell culture methods in infinite quantities with pre-viously unimagined homogeneity and reproducibil-ity.9 Antibody fragments may be created that residewithin the body for shorter periods of time and areless likely to provoke an immune response.'0 We areon the threshold of producing human antibodies bythe same cell-culture approach."I 12

I shall review the potential application of antibodiesto human pharmacology in four different areas: inhi-bition of the action of a hormone that is secreted inexcess, blockade of a physiological receptor, detectionof a newly exposed antigen in the living organism, andneutralisation of the toxic effect of a drug.

Blockade of physiological action of renin by specificantibody

Though renin has been known to play a role in cir-culatory control since the classic work of Goldblatt etal. in the 1930's,'3 its precise importance in a numberof specific circumstances has been in doubt. Much hasbeen learned from the application of inhibitorsdirected at several of the steps in the sequence leadingto the production and action of the final product ofrenin, angiotensin II. 14 Most of these compounds lackspecificity. 14 The competitive inhibitors, such assaralysin, of angiotensin II's action on receptors arepartial agonists. The angiotensin-converting enzymeis identical to the enzyme that inactivates bradyki-nin,5 and thus its inhibitors also affect the kinin sys-tem. In addition, compounds such as captopril havebeen shown to stimulate prostaglandin synthesis.'6Since both the kinin and the prostaglandin systemshave an effect on vasoregulation, it is difficult todefine the specific part that renin plays in interpretingexperiments in which these inhibitors are used. Aspecific antibody for renin should be a highly selectiveantagonist. Renin antibodies have been used asphysiological reagents for many years,'7 18 yet theirspecificity has been in doubt since we now know thatthe preparations then used as immunogens containedless than 1% of the enzyme.Dzau et al. 19 in our laboratory purified canine renin

some 600 000-fold in an eight-step process that

yielded a product homogeneous by several criteria.Antibodies specific for purified canine renin producedin a goat inhibited the pressor action of the enzymebut did not modify the capacity of either angiotensin Ior II to raise blood pressure.20 This antibody prepara-tion did not have any effect on the haemodynamics ofthe unanaesthetised, sodium-replete dog, while asignificant hypotensive effect was noted in thesodium-depleted dog when the immunoglobulin Gfraction from an antiserum was injected intraven-ously. Parallel to the fall in blood pressure, a decreasein both plasma renin activity and angiotensin II con-centrations was observed, indicating that the antibodywas exerting its effect by inhibiting the enzymaticaction of renin on its substrate.

Intact antibody has a number of troublesome prop-erties when used as a drug. When the source is aheterologous species, it is an immunogen. After thefirst use, an immune response develops that mayresult in anaphylaxis, serum sickness, or at best accel-erated elimination. Pharmacological effects may bevery persistent, because antibody is eliminated eitherby metabolism or by the immune system if sensitisa-tion has occurred. The half-life for endogenousimmunoglobulins may be measured in days or weeksdepending on the species studied, the immunoglobu-lin's isotype, and whether or not the recipient hasbeen immunised. When there is concern about renalfunction, the presence of immune complexes is likelyto cloud interpretation. In haemodynamic studies,vasoactive peptides released by activation of comple-ment may have independent effects, particularly onhaemodynamics.Some of these problems may be minimised by pro-

ducing antibody fragments. Immunoglobulinmolecules that bind two moles of antigen per molemay be cleaved to smaller molecules by the enzymepapain.2' This results in two Fab that each bind onemole of antigen and one Fc that contains the comple-ment binding site. Fab and Fc may be separated read-ily. Fab has a number of desirable properties whencompared with the intact molecule, IgG: equilibriumdistribution in extracellular fluid is achieved morerapidly; the volume of distribution is greater; and thefragment is eliminated with a far shorter half-life.'0 Inaddition, when injected intravenously, Fab is lessimmunogenic than IgG.'0 The immune complexesthat may be formed are smaller than those that causenephrotoxicity (comprising a single antigen moleculewith several Fab attached), and complement cannotbe fixed because the relevant binding sites on the Fchave been lost.

Renin-specific Fab has been used.to block theaction of renin in both sodium-depleted unanaesthet-ised dogs as well as in uninephrectomised dogs madehypertensive by constriction of the renal artery. In

2

on Novem



/B


ownloaded from


Immunological probes in cardiovascular disease

both models, intravenous injection of Fab results in ahypotensive response (Fig. 1). When compared withintact antibody, the initiation of hypotension by Fabis more rapid, and the duration very much shorter.Repetitive experiments have been possible in the sameanimal without the problem of anaphylaxis that hadbeen experienced previously when intact antibodyhad been used.An antiserum is a mixture of several hundreds of

antibodies, all capable of binding to the immunising

antigen. These antibodies vary considerably both withrespect to affinity for the antigen and specificity ofrecognition. The mixture of antibodies containedwithin an antiserum is different among even inbred(genetically identical) individual animals, as it is in thesera of a single animal collected at varying times afterimmunisation. Supplies of antisera are necessarilylimited. Kohler and Milstein9 showed that antibodiesmight be produced in a very different manner. Thetechnique of somatic-cell fusion allows for the produc-tion of hybrid cells. If the precursors of the hybrid arenormal lymphocytes and cells from a malignant plas-macytoma, one can incorporate both properties ofantibody production and growth in vitro into theproduct (colloquially named "hybridoma"). Cells thatgrow in culture may be cloned, so that all the progenyare daughters of a single precursor cell. A homogene-ous culture of this type produces a single antibodythat is uniform in structure and antigen binding prop-erties. Since the cultures may be stored indefinitely atlow temperatures, the same antibody may always berecovered. The amount of renin-specific antibody

available from conventional sera was severely limited;we set out, therefore, to make monoclonal renin-specific antibodies by the somatic-cell fusionmethod,22 making production in industrial quantitiespossible. Fig. 2 shows the binding to immobilisedrenin by a monoclonal antibody. We look forward tousing monoclonal antibodies as more specific reagentsthat are available in indefinite quantities in the dissec-tion of the physiology and chemistry of the renin-angiotensin system.

Blockade of beta-adrenergic receptor

The hormone receptors of the cell's plasma cell mem-brane are present in a very small number of copies(10 000 to 50 000 per cell), which makes their isola-tion exceedingly difficult. Antibodies to parially puri-fied receptor preparations have been elicited,23 buttheir utility is limited because they probably containantibodies to other membrane constituents. The hor-mones or drugs that bind to receptors, however, are

usually readily available in quantity. Frequently theyare either peptides of modest size or synthetic organiccompounds. For some receptors, several antagonistshave been synthesised.

Could an antibody for the receptor be obtained byusing the ligand (agonist or antagonist) as a template?The vast diversity of antibody combining sites, as

already suggested, provides the potential for creatinga complementary fit to almost any shape. If a figura-tive plaster mould could be cast upon the surface ofthe ligand that bound to the receptor, then a second

Renal artery constriction

Fab

.

-Al ~ III.

A).

s1

-.-O

H

110

100

90

80Control

0 20 40 60 80 100Minutes

Control L

0 20 40M irutes

60

Fig. 1 Renin-s*ic Fab was administered intravenous4y to either uninephrectomised dogs that were hyprtensive asa result ofconstrictim of the renal atery (left) or sodium-depleted dogs (ngk). In both expenments the rate offaU inblood pressure is more rapid and the duration shorter than observed with intact antibody.

140 F-Go

a 120

t-0~

100-

i: 50r

Fab

4

3

on Novem



/B


ownloaded from


Haber

E 2000-a\z

1000_

ao000 0 001 0 01 0-1 l 10Renin concentration M x106)

Fig. 2 Competitive radioimuunometric assay ofantibody2B3-1A6. Renin is bound to the plastic microtitre plate and thenincubated with the monoclonal antibody soluitio. The ordinaterepresents binding of 12SI goat anti-mouse Fab to IgMmonoclonal antibody 2B3-lA6. Varying concentrations ofsoluble renin (abscissa) are added to the antibody solution beforeits application to immobilised renin. (Reprinted with pemnissionofthe American Heart Association, Inc. Hypertension, in press,1981.)

mould made from the first one should have a perfectfit to the receptor. The well-known immunologicalprinciple of raising antibodies specific for anotherantibody's combining site (anti-idiotypic antibodies)may be used as the vehicle for moulding the desiredshape. Certain refinements are needed to achieve thedesired end. Only some of the atoms of either a hor-mone or an antagonist bond to the receptor. The restof the structure is needed to stabilise the active site orpossibly to protect it from degradation. To achievethe desired result, the first antibody must bind theligand generally in the same way as the receptor does,and thus the same atoms and interatomic interactionsmust be used. There are two ways of achieving thisend short of depending on fortuitous probability.When conventional immunisation is employed,

many antibodies to the inmunogen are formed, eachbinding to it in a somewhat different manner. Someare relevant to the part of the ligand that binds to thereceptor, while others are not. The antibodies of thispolyclonal response may be fractionated by the use ofligands of different structure. By virtue of beingreceptor ligands, they are all capable of binding to thereceptor and must have some common structures. Ifappropriately selected, structures irrelevant to recep-tor binding are not shared. Those components of thepolyclonal antibody mixture that bind to a number ofdissimilar ligands are likely to be similar to the recep-

tor. The most practical way of achieving fractionationwith polyclonal antibodies is sequential affinitychromatography, a technique that employs a varietyof ligands to capture the antibodies that bind to them.This general approach has now been applied success-fully to the insulin receptor24 and in our own labora-tory to the beta-adrenergic receptor.25-27

100

e. 80-2

60X9Q

IX 40

I 20

10 9 8 7 6 5 4Log ligand concentration (M)

Fig. 3 Compelitive inhibition of(-)-[3Hdihydroalprenololbinding to unfracionated anti-alprenolol antsernn. Rabbitimmune serun (100 pi) was diluted 1:10 in buffer and incubatedfor one hour at 250 C with (-)-[3HJdihydroalprenolol andincreasing concentraions of( t)-propranolol (A), ( t)-alprenolol(o), (±)-hydroxybenzylpindolol (*), (±)-norepinephrine (o),(t+)-isoprenaline (A), or catechol (v). Each detennination wasperformed in triplicate and correctedfor the nonspec;6c(-)-[3HJdihydroalprenolol binding to rabbit preimmune serwn.The results are expressed as a percentage of the totalantibody-binding capacity for the tritiated antigen. (Reprintedwith permission of the American Heart Association, Inc. CircRes 1980; 46: 808-13.)

There is a wide variety of structurally differentbeta-antagonists available. All, however, share acommon structure, a propanolamine side chain. Rab-bits were immunised with the beta-antagonistalprenolol conjugated to a protein. The resulting anti-serum was passed over an affinity column to whichanother beta-antagonist had been linked and aU thatwas not bound was discarded. The column was thenwashed with I-propranolol and the eluted antibodyretained. Fig. 3 shows the specificity profile of thisantibody fraction. Several beta-adrenergic antagonistsas well as agonists are bound with considerable affin-ity, in some cases similar to that of the receptor. Thisfraction also resolved 1 and d stereoisomers of iso-prenaline (not shown). Thus the antibody fractioncould be viewed as a qualitative, but not a strictlyprecise quantitative, model for the beta-adrenergicreceptor.

4

9

\\k

on Novem



/B


ownloaded from


Immunological probes in cardiovascular disease

Table Inhibition of ['H laprenolol binding to antibodiesand receptors by anti-idiotypic antibody*

Maximal binding In presence ofwith preimmune anti-idiotype Inhibition(DPM) (DPM) (%)

Original idiotype 10853 t 250 488 t 250 96%(alprenolol)

Turkey erythrocyte 19643 t 540 6195 t 525 68%Canine lung 6385 t 283 4315 t 190 34%

*Specific binding at 4-6 nM [3H]alprenolol.

(Reprinted with permission ofWiiams and Wilkins Co., Baltimore.Pharm Rev, in press, 1981.)

Antibodies specific for the antigen binding regions(complementarity regions) of the first antibody setwere then raised by immunisation of rabbits that weregenetically identical with respect to immunoglobulinconstant region structures. Because of tolerance toself-determinants, the imjmunised animals made anti-bodies only to the variable regions, the unique struc-tures on the immunogen. The antigenic determinantof an immunoglobulin that is characteristic of a singleantibody species is called an idiotype, and antibodiesdirected towards that determinant are named anti-idiotypes. The Table shows the inhibition of bindingof a labelled beta-adrenergic antagonist, [3H]-alprenolol, to several kinds of binding sites by theanti-idiotypic antibody. Binding of the ligand,alprenolol, to the idiotype (the first generation anti-body that had been raised in response to alprenolol) islargely inhibited by anti-idiotype (second-generationantibody). Of greater interest is the inhibition of thespecific binding of alprenolol to plasma membranepreparation in two different animal species. Quantita-tive analysis of binding shows it to be of a competitivenature, analogous to that shown by conventionalbeta-antagonists.The anti-idiotype is also an inhibitor of adenylate

cyclase activation by beta-adrenergic agonists.Increasing concentrations of isoprenaline progres-sively inhibit adenylate cyclase production in turkeyerythrocyte membranes at 5 x 10-7M isoprenalineconcentration. At a higher concentration of iso-prenaline (10-4M), less inhibition is observed, sug-gesting a competitive nature for this interaction aswell. Thus the anti-idiotype behaves as a true beta-adrenergic antagonist, competing with both agonistsand antagonists for the receptor site. The obviouspotential uses of such receptor-specific antibodies are:the recognition of structural differences among sub-sets of beta-adrenergic receptors28; a more rigorousexamination of their respective physiological roleusing reagents of greater resolution; and the isolationof receptors with antibody affinity chromatography.

Imaging of myocardial infarcts using cardiacmyosin-specific antibodies

The hallmark of cell death is loss of the membraneintegrity: when there is no longer a physical separa-tion between inside and outside, the cell ceases toexist. This principle is commonly employed in clinicalpractice in diagnosing tissue infarction. Intracellularenzymes are lost to extracellular fluid, and theirincreased concentration in the blood may be used as ameasure of tissue necrosis. An alternative approach isthe inward diffusion of a marker that is normallyexcluded from cells. We have selected myocardialinfarction as a model and radioactively labelledmyosin-specific antibody as a marker.Myosin is the principal protein of the cardiac cell.

Its covalent structure is unique to the heart,29 allow-ing the development of antibodies that differentiatebetween cardiac and either skeletal- or smooth-musclemyosins. In the intact organism, cardiac myosin isprotected from extracellular fluid by the cell's plasmamembrane. When cell death occurs and the mem-brane breaks down, myosin is exposed to extracellularfluid. It is then available to react with labelled anti-bodies or antibody fragments.A visual demonstration of this phenomenon is seen

in Fig. 4. The figure shows a scanning electron micro-graph of part of a neonatal mouse cardiac myosite thathad been rendered ischaemic by prolonged exposureto nitrogen and then incubated with 1 A diameter,polystyrene spheres covalently bound to myosin-specific antibody. Myofibrils are seen protruding froma hole in the cell membrane. Antibody-coated micro-spheres clearly bind to the myofibrils. Not only ismyosin very insoluble in physiological fluids so thatmembrane breakdown does not result in antigen loss,but it persists for considerable periods after ischaemicnecrosis.30 This should allow for identification ofinfarcted myocardium in days to, perhaps, severalweeks after the initial event. In our initial explorationof this concept, a canine myocardial infarction modelwas used.31 The left anterior descending coronaryartery was ligated, and four hours later I25I-labelled,myosin-specific antibody or antibody fragments wereinjected intravenously. At varying times after theinjections, the animals were killed, the hearts wereperfused with triphenyltetrazolium chloride, and themyocardia examined.32 Fig. 5 (left) shows a section ofa heart treated in this way. The light area represents alargely subendocardial infarction. The central panelof the figure is a tracing of the slice superimposed onan autoradiograph. The exposed area corresponds tothe infarct as disclosed by the triphenyltetrazoliumstain. The right panel shows specific radioactivity inthe area of the infarct relative to myocardium on theposterior wall. It is apparent that the major concentra-

5

on Novem



/B


ownloaded from


Haber

Fig. 4 Scanning electron micrograph ofan ischaemic neonatal mouse n.yosite(Marcation x 2400). The spheres adherent to the ceU are covalenty bonded tomyosin-specjic antibody. Non-ischaemic ceUs do not bind significant numbers ofspheres.

r Vf

I_. .. 4-s+s;

Fig. 5 Histochemicaly delinead myocardial infarction in a ventricular slide (left) seen as white or lighter-colowed regions,and nomalMocardium seen as darer regims. The conespondingmacroautoradiograph is shown in the centre with the outlineofthe ventricular slide. The right panel shows relative antibody uptake in the indicated areas. Ratios ofantibody uptake weredeternined as specjic radioatvity in relation to normal posterior veticular miyocrdiw . (Reprinted with permission of theAmerican Heart Assocaion, Inc. Proc N Engl Cardiovasc Soc 1979: Vol. II; 314.)

tion of radioactivity is in the subendocardial region,with lesser concentration in the subepicardial regionthat had been subject only to spotty necrosis as indi-cated in the triphenyltetrazolium stain.

Microautoradiographs showed that individual ne-crotic myosites could be identified and differentiatedfrom adjacent viable cells.33 In order to show thatantibody concentration was specific and not simplythe result of passive diffusion of a macromolecule intoinfarcted cells, specific antibody labelled with 1311 andnonimmune globulin labelled with 125I were injected

simultaneously into the coronary arteries.32 At thecentre of the infarct, the antibody has concentrated34-fold in relation to normal myocardium, while thenonimmune globulin is only sevenfold in excess. Innormal tissue, as expected, concentrations are equal.When compared to a marker of relative flow, the dis-tribution of radioactive microspheres which had beeninjected into the left atrium, it was clear that the con-centration of labelled antibody was inversely relatedto relative blood flow.31 32 There seems to besufficient collateral circulation in this ischaemic model

6

on Novem



/B


ownloaded from


Imnological probes in cardiovascular disease

to provide for delivery of antibody, even to the re-gions of minimal blood flow. It would be very desir-able to apply this method to the detection, localisa-tion, and quantification of myocardial infarcts in theliving subject.

Optimal imaging with a gamma or positron camerarequires the labelling of antibodies with radionucidesof appropriate half-life and affinity. We have accom-plished this by covalently linking diethylene triaminepenta-acetic acid (DTPA) to antibody or antibodyfragment and then binding cationic radionuclides bychelation.34 Successful images have been obtainedusing the following radionuclides linked to antibody:"lIn-DTPA-Ab Fab,34 68Ga-DTPA Fab,35 and99mTc-DTPA-Ab Fab.34 An example of a positronimage of an anterior infarct in a living dog using68Ga-labelled antibody is shown in Fig. 6 (left). Thepositron imaging technique allows for tomographicreconstruction and, thus, the heart is visible oncross-section in the "3N-ammonia image in the centrepanel of Fig. 6 (ammonium ion is a potassiumanalogue and concentrates in normal tissue). On theleft side of the image there is clearly less density,reflecting the region of infarction. Superimposition ofthe 13N and the 68Ga images (right panel) showed thatthe antibody was concentrated directly in the area ofdiminished ammonium ion uptake. The potential forapplying this method to the evaluation of myocardialinfarcts in a clinical setting seems very great indeedsince it offers high specificity and resolution.

Reversal of digitalis glycoside intoxication

If the effect of an endogenously produced hormone

can be reversed, it should also be possible to coun-teract the undesirable effects of an exogenouslyadministered drug or toxin. The digitalis glycosidesare of great value in the treatment of congestive heartfailure and consequently are frequently used in clini-cal medicine. Unfortunately, they are characterisedby a very close toxic-therapeutic ratio. Digitalis in-toxication is one of the most frequent adverse drugreactions reported. There is no specific antidote, andthe cardiac arrhythmias that are a feature of digitalisintoxication are not uncommonly fatal.We reasoned that if an antibody specific to the

digitalis glycosides had a higher affinity for the drugthan the physiological receptor, it should be possibleto transfer the ligand from the receptor to antibody,simply by mass action. For optimal effectiveness dif-fusion distances should be minimal, and the antibodyshould be in high concentration in extracellular fluidin proximity to the receptor. It would also be desir-able to remove the antibody-drug complex rapidlyfrom the body. Both aims may be accomplished bythe use of digitalis-specific Fab. As indicated above,the volume of distribution of Fab is greater than intactantibody since it readily enters extracellular fluid. It isalso of sufficiently small size (50 000 Daltons) to passthrough the glomerular filter and be largely excretedin the urine.36

Digoxin-specific antibody was purified from sheepantiserum using immobilised ouabain Fab isolatedafter papain cleavage. 3 After demonstration of safetyand effectiveness in animal studies, clinical investiga-tions were initiated. At the time of this writing, 17patients with life-threatening digitalis intoxicationhave been studied. The patients ranged in age from

Fig. 6 68Ga and 13NH4+ were obtained sequentially using a positron camera; S mCi of 13NH4+ was administeredintraveo(y to a dog that had had occlusion ofthe distal segment ofthe left anterior descending arteryforfour hours andwas imaged (centre); 30 mitmes later ) mCi of 68Ga was injected into the same artery and imaged (left panel). Right panelis a composite of both images. (Reprinted with permission of the American Heart Association, Inc. Proc N EngiCardiovasc Soc 1979: Vol. II; 31-8.)

7

on Novem



/B


ownloaded from


Haber

infants to septuagenarians and in each case a dramaticreversal of the signs and symptoms of intoxicationoccurred.38-40The history of a recently reported patient is typical

of the group.39 She was a 34-year-old wife of a physi-cian, who, because of marital difficulties, took 20 mgdigitoxin, a massive overdose. She appeared well onadmission to the hospital except for nausea, but soonlapsed into a series of life-threatening arrhythmiasthat included multiple episodes of ventricular fibrilla-tion, which were treated with countershock, as well asasystole, which was treated by ventricular pacing. Atthe time the antibody Fab became available to thephysicians treating her, the patient was in shock andanuric, and she had dilated pupils. Her serum potas-sium was raised, a grave prognostic sign in digitalisintoxication.4' Within an hour after the intravenousadministration of antibody Fab, her atrioventricularconduction had returned, and she was soon in normalsinus rhythm. No further arrhythmias occurred. Shewas discharged from hospital without sequelae severaldays later. Fig. 7 shows the initial distinct increase ofserum digitoxin concentration in this patient as

10o

75

Digitoxin_ngrr) 5

25

1 2Day

3 4 5 6 14

60000

Fab40 0Q (ngmik)

(o.--o)

20000

Fig. 7 Blood levels ofdigioxin and Fab after intravenousadministaion ofantbody Fab to a 34year-old womansuffering a seres of life-threatening arrhythmias as a resuk ofamassive overdose (20 mg) ofdigowxin. Within an hour ofintravenous adinistration ofantibody Fab, arioventricularconductin had returned. (Reprinted from Contributions ofchemical biology to the biomedical siences withpeswn ofAcademc Press, New York. In press, 1981.)

tissue-bound drug equilibrated with the antibody(antibody-bound drug is pharmacologically inactive),followed by rapid clearance of both drug and Fab. Itshould be noted that the half-life of digitoxin in man isnormally three and a half days when hepatic metab-olism of the drug is the major source of removal. It is

apparent that excretion was much accelerated by theantibody Fab, the half-life apparently reduced toabout 12 hours. Fig. 8 shows that both antibody andFab appear in the urine in largest amounts within thefirst day after Fab administration.

Initiation oftreatment

1000 _ 60000A

Digtoxin L40000(,ugl/h) Fabulh

500 ~~~~~~~~~~(o0---o )

Fig. 8 Urinary excretion rates ofdigitoxin and Fab afterintravenous administration ofantibody Fab to a 34year-oldwoman sufferingfrom digitalis intoxication (20 mg). (Reprintedfrom Contributions of chemical biology to the biomedicalsciences with permission ofAcademic Press, New York. Inpress, 1981).

The digitalis glycosides in most common use aredigoxin and digitoxin. They differ fror one anothervery little in their chemical structure, yet most anti-bodies differentiate between them, with differences inaffinity of 50- to 100-fold.2 It would be desirable touse an antibody that bound both digoxin anddigitoxin with equal affinity. A monoclonal antibody(designated 26-10 in our laboratory) was selected thathad an affinity of 5 x 109 for digoxin and a nearly equalaffinity for digitoxin.42 Since a monoclonal antibody isa homogeneous protein that recognises only a singleaspect of an antigen molecule, it appears that the hy-droxyl group is not a part of the antigenic recognitionsite of this molecule. This antibody, which can beproduced in quantities without limit, has been shownto reverse digoxin intoxication in experimentalanimals.

It should be apparent that this approach could beapplied to many other drugs or toxins. Of particularinterest is the acceleration of excretion of a substancethat is normally metabolised slowly. Fab is capable ofaltering the route of disposal from hepatic metabolismto renal excretion.43

Iniiation oftreatment

I'

I'\;o &

o L \L A

8

50(

on Novem



/B


ownloaded from


Imrmnological probes in cardiovascular disease

Antibodies as drugsThe development of a new drug is often the result ofempiricism, serendipity, or the deliberatemodification of an existing natural product. Theseproducts of the organic chemist's laboratory are usu-ally characterised by a lack of specificity with conse-quent abundance of adverse reactions. If antibodieswere the basis of identification of a specific site withinthe body for binding, blockade, or modification, onecould choose among some 1011 different recognitionsites and thus much increase specificity. Adverse reac-tions to drugs are most often the result of reactionwith a site other than the desired target. Selectiveinteraction should, of necessity, diminish'these unde-sired effects. The considerably larger size of the anti-body combining site in comparison to conventionaldrugs allows for the establishment of a greaternumber of specific interactions among atoms and thusboth greater affinity and specificity.

While some of the advantages of antibodies areobvious, the disadvantages, discussed in part above,must also be considered. When binding to a specificsite is all that is desired, the activation of the biologi-cal effects that are an intrinsic part of an immunereaction is undesirable. The smaller Fab that was dis-cussed above as a' possible solution to some of theseproblems still carries with it an entire domain (onehalf of its mass) that is of no relevance to antigenbinding. Fab from a heterologous species, while ofdiminished immunogenicity, is likely to causehypersensitivity when used on a long-term basis. Thesolutions to these problems are almost at hand. Amuch smaller fragment of antibody has been pro-duced that retains all the binding energy andspecificity of the intact molecule. Fv comprises asingle domain and has a molecular size of 25 000 Dal-tons.44 While its pharmacokinetics have not as yetbeen mined, I anticipate that it will be very rapidlydeared and distributed. The advent of the hybridomamethod will make possible not only the selection andlarge-scale production of antibodies of uniform prop-erties, but also the reduction or elimination of thepotential problem of hypersensitivity, since it will bepossible, as indicated above, to produce human anti-bodies. Will the formation of anti-idiotypic antibodiesdefeat the long-term utility of antibody therapy? Ibelieve that the well-recognised difficulty in produc-ing anti-idiotypes, even deliberately, indicates thatthis problem will be rare. The next decade will see anew pharmacology based on the antibody combiningsite.

References1 Landsteiner K. The specificity of serological reactions.Revised ed. Cambridge, Massachusetts: Harvard Uni-versity Press, 1945.

2 Smith TW, Butler VP Jr, Haber E. Characterization ofantibodies of high affinity and specificity for the digitalisglycoside digoxin. Biochemistry 1970; 9: 331-7.

3 Brack C, Hirama M, Lenhard-Schuller R, Tonegawa S.A complete immunoglobulin gene is created by somaticrecombination. Cell 1978; 15: 1-14.

4 Sakano H, Kurosawa Y, Weigert M, Tonegawa S.Identification and nucleotide sequence of a diversityDNA segment (D) of immunoglobulin heavy-chaingenes. Nature 1981; 290: 562-70.

5 Seidman JG, Leder A, Nau M, Norman B, Leder P.Antibody diversity: the structure of cloned immunoglo-bin genes suggests a mechanism for generating new se-quences. Scence 1978; 202: 11-7.

6 Seidman JG, Max EE, Leder P. A K-immunoglobulingene is formed by site-specific recombination withoutfurther somatic mutation. Nature 1979; 280: 370-5.

7 Early'P, Huang H, Davis M, Calamine K, Hood L. Animmunoglobulin heavy chain variable region gene is gen-erated from three segments of DNA: VH, D and JH.Cell 1980; 19: 981-92.

8 Weigert M, Gatmaitan L, Loh E, Schilling J, Hood L.Rearrangement of genetic information may produceimmunoglobulin diversity. Nature 1978; 276: 785-90.

9 Kohler G, Milstein C. Continuous cultures of fused cellssecreting antibody of predefined specificity. Nature1975; 256: 495-7.

10 Smith TW, Lloyd BL, Spicer N, Haber E. Immunogen-icity and kinetics of distribution and elimination of sheepdigoxin-specific IgG and Fab fragments in the rabbit andbaboon. Clin Exp Immunol 1979; 36: 384-96.

11 Croce CM, Linnenbach A, Hall W, Steplewski Z, Kop-rowski H. Production of human hybridomas secretingantibodies to measles virus. Nature 1980; 288: 488-9.

12 Olsson L, Kaplan HS. Human-human hybridomas pro-ducing monoclonal antibodies of predefined antigenicspecificity. Proc Natl Acad Sci USA 1980; 77: 5429-31.

13 Goldblatt H, Lynch J, Hanzal RF, Ramon F, Summer-vile WW. Studies on experimental hypertension. I. Theproduction of persistent elevation of systolic blood pres-sure by means of renal ischemia. J Exp Med 1934; 59:347-79.

14 Haber E. Specific inhibitors of renin. Clin Sci 1980; 59,suppl 6: 7s-19s.

15 Yang HYT, Erdos EG, Levin Y. A dipeptidylcarboxypeptidase that converts angiotensin I and inacti-vates bradykinin. Biochim Biophys Acta 1970; 214:374-6.

16 Zusman RM. Captopril stimulates prostaglandin E2synthesis in vitro: possible mechanism of antihyperten-sive action. Clin Res 1981; 29: 362A.

17 Deodhar SD, Haas E, Goldblatt H. Production of anti-renin to homologous renin and its effect on experimentalrenal hypertension. J Exp Med 1964; 119: 425-32.

18 Wakerlin GE. Antibodies to renin as proof of thepathogenesis of sustained renal hypertension. Circula-tion 1958; 17: 653-7.

19 Dzau VJ, Slater EE, Haber E. Complete purification ofdog renal renin. Biochemistry 1979; 18: 5224-8.

20 Dzau VJ, Kopelman RI, Barger AC, Haber E. Renin-specific antibody for study of cardiovascular homeostasis.

9

on Novem



/B


ownloaded from


Haber

Science 1980; 207: 1091-3.21 Nisonoff A. Enzymatic digestion of rabbit gamma globu-

lin and antibody and chromatography of digestion pro-ducts. Methods in Medical Research 1964; 10: 134-41.

22 Dzau V, Mudgett-Hunter M, Kapler G, Haber E.Monoclonal antibodies binding renal renin. Hyperten-sion, in press 1981.

23 Wrenn S, Haber E. An antibody specific for propranololbinding site of cardiac muscle. J Biol Chem 1979; 254:6577-82.

24 Kahn CR. Membrane receptors for hormones andneurotransmitters. J Cell Biol 1976; 70: 261-86.

25 Rockson S, Homcy C, Haber E. Anti-alprenolol anti-bodies in the rabbit: a new probe for the study of beta-adrenergic receptor interaction. Clin Res 1980; 46:808-13.

26 Homcy C, Rockson S, Haber E. An anti-idiotypic anti-body which recognizes the beta-adrenergic receptor.Clin Res 1981; 29: 207A.

27 Schreiber AB, Couraud PO, Andre C, Vray B, StrosbergAD. Anti-alprenolol anti idiotypic antibodies bind toB3-adrenergic receptors and modulate catecholamine-sensitive adenylate cyclase. Proc Natl Acad Sci USA1980; 77: 7385-9.

28 Burges RA, Blackburn KJ. Adenyl cyclase and the dif-ferentiation of p-adrenoreceptors. Nature New Biol1972; 235: 249-50.

29 Katz AM. Contractile proteins of the heart. Physiol Rev1970; 50: 63-158.

30 Fetisova TV, Zabolotnyuk IP. Changes in heart muscleand liver proteins in experimental myocardial infarction.Fed Proc 1965; 24: T960-2.

31 Khaw BA, Belier GA, Haber E, Smith TW. Localizationof cardiac myosin-specific antibody in myocardial infarc-tion. J Clin Invest 1976; 58: 439-46.

32 Khaw BA, Gold KH, Leinbach RC, et al. Early imagingof experimental myocardial infarction by intracoronaryadministration of 1311-labelled anticardiac myosin (Fab')2fragments. Circulation 1978; 58: 1137-42.

33 Khaw BA, Fallon JT, Belier GA, Haber E. Specificity oflocalization of myosin-specific antibody fragments inexperimental myocardial infarction: histologic,histochemical, autoradiographic and scintigraphicstudies. Circulation 1979; 60: 1527-31.

34 Khaw BA, Falion JT, Strauss HW, Haber E. Myocardialinfarct imaging of antibodies to canine cardiac myosinwith Indium-ll 1-diethylenetriamine pentaacetic acid.Science 1980; 209: 295-7.

35 Khaw BA, Fallon JT, Strauss W, et al. Indium-Illdiethylenetriamine pentaacetic acid-antimyosin Fabfragments localization and imaging in experimentalmyocardial infarction (abstract). J Nucl Med 1979; 20:605.

36 Epstein M, Haber E, Re R. A kinetic assessment ofaldosterone responsiveness in secondary hyperaldoste-ronism and in anephric man. In: James VHT, Serio M,Giusti G, Martini L, eds. The endocrine function of thehuman adrenal cortex; proceedings of the Serono sym-posia. Vol. 18. London: Academic Press, 1978: 493-508.

37 Curd J, Smith TW, Jaton J-C, Haber E. The isolation ofdigoxin specific antibody and its use in reversing theeffects of digoxin. Proc Natl Acad Sci USA 1971; 68:2401-6.

38 Smith TW, Haber E, Yeatman L, Butler VP Jr. Reversalof advanced digoxin intoxication with Fab fragments ofdigoxin-specific antibodies. N Engl J Med 1976; 294:797-800.

39 Aeberhard P, Butler VP, Smith TW, et al. Le traitementd'une intoxication digitalique massive (20 mg de digitox-ine) par les anticorps anti-digoxine fractionnes (Fab).Arch Mal Coeur 1980; 73: 1471-8.

40 Hess T, Stucki P, Barandun S, Scholtysik G, Riesen W.Treatment of a case of lanatoside C intoxication withdigoxin-specific F(ab')2 antibody fragments. Am HeartJ1979; 98: 767-71.

41 Bismuth C, Gaultier M, Conso F, Efthymiou ML.Hyperkalemia in acute digitalis poisoning: prognosticsignificance and therapeutic implications. Clin Toxicol1973; 6: 153-62.

42 Margolies MN, Mudgett-Hunter M, Novotny J, HaberE. Monoclonal antibodies to the cardiac glycosidedigoxin. In: Hammerling G, Hammerling U, KearneyJF, eds. Monoclonal antibodies and T Cell hybridomas.New York: Elsevier/North-Holland, in press.

43 Ochs HR, Smith TW. Reversal of advanced digitoxintoxicity and modification of pharmacokinetics by specificantibodies and Fab fragments. J Clin Invest 1977; 60:1303-13.

44 Hochman J, Inbar D, Givol D. An active antibody frag-ment (Fv) composed of the variable portions of heavyand light chains. Biochemistry 1976; 12: 1130-5.

Requests for reprints to Dr Edgar Haber, CardiacUnit, Massachusetts General Hospital, Boston, Mas-sachusetts 02114, USA.

10

on Novem



/B


ownloaded from


immunological probes incardiovascular disease*immunologicalprobes in cardiovascular disease...

Documents