Immune regulation of myasthenia - jnnp.bmj.com · Studies on specific monoclonal antibodies, anti-idiotypes, and on the effect of measles virus on EAMG are being described and their

Journal of Neurology, Neurosurgery, and Psychiatry, 1980, 43, 634-643

Immune regulation of experimental myastheniaSARA FUCHS, DANIEL BARTFELD, ZELIG ESHHAR,CARMELA FEINGOLD, DARIA MOCHLY-ROSEN,DANIELA NOVICK, MICHAL SCHWARTZ ANDREBECA TARRAB-HAZDAI

From the Department of Chemical Immunology, The Weizmann Institute of Science, Rehovot, Israel

S UMM A R Y Experimental autoimmune myasthenia gravis (EAMG) is an appropriate model forstudying the molecular origin, immunological mechanism and regulation of myasthenia gravis.Several approaches are being utilised for the regulation of the immune response to AChR andfor immunosuppression of EAMG: Corticosteroids and azathioprine can suppress EAMG con-

comitantly with suppression of immune responses to AChR. High dose cyclophosphamide treat-ment in mice facilitates the onset of EAMG and results in a selective suppression of the humoralresponse to AChR whereas the cellular response is enhanced. Specific immunosuppression ofEAMG is achieved by using a nonmyasthenic, denatured AChR preparation which cross reactswith the intact receptor. Various degradations and modifications of AChR are being performedin order to identify the smallest molecular entity responsible for the myasthenic activity ofAChR. Studies on specific monoclonal antibodies, anti-idiotypes, and on the effect of measlesvirus on EAMG are being described and their possible significance in regulating myasthenia are

being discussed.

Two decades have passed since it was first sug-gested that "myasthenia gravis is an 'auto-immune'response of muscle in which an antibody to endplate protein may be formed".' This hypothesis,which was originally based on attempts to explainclinical and experimental phenomena observed inmyasthenia gravis (MG), has undoubtedly provedto be correct. Today, 20 years after the veryfirst seeds of the autoimmune hypothesis weresown,' 2 it is well established that MG is an auto-immune disease in which acetylcholine receptor(AChR) is a major autoantigen.A large part of the recent advances in our

understanding of the nature and origin of MGis due to the ability to induce experimental my-asthenia by immunisation of animals with purifiedAChR. This experimental disease which was des-ignated experimental autoimmune myastheniagravis (EAMG) and which was first observedalmost accidentally in 1973, has become an ex-tensively studied model system. We have accu-mulated considerable knowledge concerning boththe molecular structure of AChR and the natureof the autoimmune response which it produces.Address for reprint requests: Dr S Fuchs, Department of ChemicalImmunology, The Weizmann Institute ofScience, Rehovot, Israel.

The availability of an experimental model diseasefor MG induced by a well defined antigen (AChR),provides a valuable tool for attempting regulationand suppression of the disease. Two general ap-proaches for suppression of EAMG have beenused in our laboratory: non specific and specificimmunosuppressive treatment. While in the firstapproach we applied non-specific drugs whichaffect the whole immune system, in the latter weare aiming at specific suppression of that im-mune response to AChR which is responsiblefor its myasthenic activity. In this paper wedescribe experiments from our laboratory attempt-ing regulation of the immune response to AChRand suppression of EAMG.

REGULATION OF EXPERIMENTAL MYAS1HENIA BYNON-SPECIFIC IMMUNOSUPPRESSIVE DRUGSCorticosteroids and azathioprine: Corticosteroidsand the antimetabolite azathioprine were empiri-cally found to be effective in the treatment ofMG.3-6 We have shown previously that myasthenicpatients have both cellular7 and humoral8 im-munological sensitivity to AChR and that thecellular immune response to AChR is immuno-suppressed following steroid therapy.9 We have

634

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from

Immune regulation of experimental myastheniia

used the same drugs in order to suppress or ratherto prevent the onset of EAMG in rabbits and tofollow the mechanism of suppression. 10-19 Thissystem allows us to vary the course of the drugtreatment and to analyse concomitantly immun-ological parameters affected by the treatment.

Rabbits injected with purified Torpedo cali-fornica AChR were treated with hydrocortisoneeither by early continuous administration of highdoses of the drug or by administration of graduallyincreasing doses of hydrocortisone.1" When hydro-cortisone was administered in high doses from thebeginning, some suppressive effect was observedalthough EAMG appeared earlier and sometimesin a more severe form than in the control animals.In the group of rabbits treated with graduallyincreasing doses of hydrocortisone, EAMG was

completely suppressed. The schedule of steroidsadministration seems to be of great importancealso in MG patients specially for minimising sideeffects; in spite of the overall beneficial effect ofsteroids in MG, increasing wcakness usually occurs

early in treatment, and can be avoided by a

gradually increasing dosage schedule.The effect of hydrocortisone in supprcssing

EAMG was paralleled by a diminished cellularsensitisation to AChR in vitro." This is in agree-

ment with our findings that in myasthenia gravisdecreased cellular sensitisation to AChR corre-

lated with clinical improvement during prednisonetherapy.9We have studied in more detail the immuno-

suppression of EAMG by prolonged treatmentwith azathioprine (Az) and have studied the cor-

relation between clinical effects and severalimmunological parameters resulting from thisimmunosuppressive treatment.'10'4 Rabbits vere

injected on day 0 with AChR (80 ,ug, in ccmpleteFreund's adjuvant, intradermally) and with Az(4 mg/kg, intramuscularly). Similar administra-tions of Az were given to the rabbits daily for15 days and then every two to three days for an

additional five months. In a control group theAz injections were replaced by saline injections.The Az treatment effectively suppressed the onsetof EAMG in rabbits, for at least 12 months, even

after discontinuing the drug treatment for 7months (table 1).12 14 A second injection of AChRto such immunosuppressed rabbits, 12 monthslater, led to the onset of EAMG after a furthersix to ten days, as is the case after a secondaryinjection with AChR.The suppressive effect of Az was accompanied

by decreased cellular and humoral immunologicalreactivity against both the immunising Torpedo

Table 1 Suppression of AChR-induiced EAMG byazathioprine

Treatment* EAMG Survivingafter

Onset Clinical 12month(days) signs (%)

(%!

Saline 21-30 100 (8/8)t 0(control)Azathioprine 100 10 (1/10) 80 (8/10)(5 months)

All rabbits were injected with 80Ag Torpedo californica AChR on day 0.tThe numbers in parentheses represent the number ofanimals.

AChR and self AChR, with a significantly morepronounced effect on the response to self re-ceptor.12 The cellular sensitivity towards theTorpedo AChR as measured by the in vitrolymphocyte transformation technique, was sup-pressed by about 45% as a result of Az treatment,whereas the reactivity towards self receptor wasessentially abolished. Az treatment was shown tobe effective also on macrophage bound antibodiesand to a lesser extent on lymphocyte bound anti-bodies. The effect of Az on the humoral responsetowards AChR was measured lby analysing totalcirculating antibodies as well as cytophilic anti-bodies. The antibody titre against self AChR,representing an autoimmune response, wasessentially abolished as was the case also for thecellular sensitivity. However, the titre against theforeign fish receptor was reduced only to a smallextent.12Az treatment resulted in qualitative changes in

the antibody class, specificity, and affinity. Theassociation between the immunosuppressive effectof Az on AChR-induced EAMG and the affinityof the antibodies was studied, utilising purifiedanti-AChR antibodies.13 Binding experiments wereperformed with constant amounts of purified anti-bodies reacted with increasing amounts of iodin-ated AChR. The data of these experiments wereplotted using Sips and Scatchard equations.15 Fromthe mathematical analysis obtained from thesebinding experiments it could be concluded thatAz treatment leads to quantitative rather thanqualitative differences. Thus it appears that themain immunosuppressive effect of Az is on thelevel of anti-AChR antibodies possessing highaffinity values.13The experiments with azathioprine allowed also

the evaluation of the immunological memoryfollowing such an immunosuppressive treatment.14As mentioned earlier a second injection of AChRto Az immunosuppressed rabbits, 12 months after

635

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from

Sara Fuchs et al

the first challenge led to the development of acuteclinical signs of EAMG six to ten days later. Thisbehaviour suggests that although EAMG was pre-vented by Az treatment, immunological memorywas maintained. However, it is of interest to pointout that no clinical myasthenic signs were observedin rabbits which were originally injected withAChR, immunosuppressed with Az and boosted12 months later with denatured AChR preparation(RCM-AChR), which cross-reacts with AChRbut does not induce EAMG.'6 Thus, the anamn-estic response to the myasthenic determinants wasnot awakened by this cross reactive, AChRderivative.

In addition to the clinical manifestation of thesecondary antigenic challenge in the immuno-suppressed rabbits the humoral immune responseexhibited also an anamnestic pattern. The anti-body titre in AChR-injected and Az treated rabbitsdecreased gradually during the first 12 months toa level of about 1% (3 X 10-9 M, expressed as molesof AChR precipitated by litre of serum,14) of thetitre obtained 30 days after the primary injection(2X10-7M). Following the booster AChR injec-tion, there was a marked and fast increase of theantibody titre, to about 10-6 M. It should be notedthat rabbits that were given a booster injection ofRCM-AChR exhibit also an increase in anti-AChR antibody titre (to 10-7 M), although thisincrease was lower than that obtained in AChRboosted rabbits. The fast kinetic and extent ofanti-AChR antibody in the boosted animals issimilar to or even higher than a secondary re-sponse obtained in rabbits following a boosterinjection of AChR without any immunosuppress-ive treatment. The affinity of the antibodieselicited in the suppressed rabbits by the secondimmunisation, was at least as high or higher thanthat of rabbits immunised twice with AChR. Thespecificity of antibodies elicited in the Az immuno-suppressed rabbits following the second antigenicstimulation is also identical to that of antibodiesof non treated rabbits.The utilisation of non-specific immunosuppress-

ive drugs in an experimental animal system enablesus to elucidate various aspects in their mechanismof action. However, there are no previous reportsavailable on how immunological memory isaffected in autoimmune diseases following im-munosuppressive treatment. Our data14 demon-strate the existence of immunological memoryfollowing a prolonged Az treatment of AChR-injected rabbits. It seems that such rabbits stillhave the memory cells that could differentiate intoeffector cells (killer cells) and antibody producing

cells, leading to a concomitant appearance of thedisease in a very short period of time. These par-ticular memory cells are specific ones, and cannotbe activated by the cross-reactive antigen RCM-AChR.

In spite of its immunosuppressive activity, itappears that even prolonged Az treatment does notaffect the antigen sensitive cells that carrymemory. This fact should be very carefully con-sidered when immunosuppressive drugs are usedin the treatment of autoimmune diseases since avery small challenge may recall an enhanced ap-pearance of the disease even in patients who haveundergone immunosuppressive treatment.

In conclusion, the immunosuppressive treatmentof EAMG provides a valuable tool for elucidatingmany aspects of the mechanism of action of nonspecific immunosuppressive drugs in therapy ofautoimmune diseases in general, and, in particular,for establishing the optimal conditions for such atherapy in MG.Cyclophosphamide: Cyclophosphamide (CYP) isan immunosuppressive drug17 18 known to have aselective effect on distinct cell populations involvedin the immune response. Its effect is highly de-pendent on the dose and timing of administrationwith respect to the antigen injection. Given athigh doses (100-300 mg/kg) CYP inactivates bothB cells and suppressor T cells,17 and enhancessome forms of delayed type hypersensitivity.19 20CYP was shown to be a potent suppressor of anti-body response in the mouse21 due to its effect onB-cells. There are reports in the literature thatCYP is also involved in the induction of endo-genous viruses22; Tindall et al have observed in-creased titres of cytomegalovirus antibodies in MGpatients given CYP or Az, and appearance ofvirus in the urine.2' It thus seemed of interest tostudy the effect of CYP on regulating the immuneresponse to AChR and on the induction andcourse of EAMG in mice.

Different mouse strains possessing different H-2haplotypes or on different genetic background suchas C57BL/6J, BALB/c, SWR and ASW, wereinjected subcutaneously (between the ears) withCYP in saline (150-200 mg/kg). Forty hours laterthose mice were immunised intradermally withAChR (10 jug/mouse) in complete Freund's adju-vant (CFA). C57BL/6J mice were found to be themost sensitive to the CYP pretreatment and inmany cases showed clinical signs of EAMG 6-8weeks following the single immunisation withAChR. On the other hand such CYP pretreatmenthad no effect in mouse strains which were reportedas not susceptible to the disease.'4

636

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from

Immune regulation of experimental myasthenia

The humoral and cellular response of the CYP-treated and AChR-injected C57BL/6J mice, aswell as of the control mice injected with AChR,without CYP pretreatment, was studied on variousdays following AChR administration. Humoralimmune response was determined by micropassivehaemagglutination test24 and cell mediated im-munity was determined by the ear test.25 For thistest the tested mice were injected subcutaneouslywith AChR (8 ,ug) in the right ear and with dilut-ing buffer in the left ear. Ten hours later FUdR(0.1 ,mole) followed by 125I-UdR(2 ,uCi, given 30minutes after FUdR) were administered intra-peritoneally.25 The ears were cut 26 hours afterthe AChR sensitisation and the radioactivity wascounted. The results were expressed as the ratioof radioactivity of the right (experimental) to theleft (control) ear (stimulation index-SI). As canbe seen in fig 1 CYP treatment resulted in sup-pression of humoral response and enhancement ofcellular response.

3c

2En

6 8 1113 21

Days after AChR injection31 46 64

16r1

014N

12

10 la0

8

6

4 -1

Fig 1 Kinetics of delayed type hypersensitivity andhumoral response in C57BL/6J mice. Open bars andempty circles represent SI and log2 haemagglutinationtitres, respectively, in C57BL/6J mice treated withCYP 48 hrs before AChR injection. Dark bars anddark circles represent SI and log>, haemagglutinationtitres, respectively, in C57BL/6J mice injected withAChR and not treated with CYP.

It is thus suggested that the induction of un-responsiveness at the humoral level and the in-crease in the cellular reactivity associated withCYP treatment may have facilitated the onset ofEAMG iu C57BL/6J mice following one injectionof AChR. However, it should be pointed out thatwhereas the effect on humoral and cellular activityby CYP was observed in all experiments, its effecton the onset of EAMG was not always repro-

ducible. It is still not clear what is the mode ofaction of CYP that makes the animal more prone

to EAMG but nevertheless it may be helpful in

637

the elucidation of the immunological mechanisminvolved in this disease.

MODIFICATION OF THE AChR MOLECULE ANDSPECIFIC IMMUNOSUPPRESSION OF EAMGAlthough immunosuppressive drugs seem to beuseful in the treatment of both the human andexperimental diseases, there are disadvantages inusing such agents which may suppress the wholeimmune system. The ideal treatment of choice, ifpossible, should be a specific one, namely, a drugwhich will affect selectively the immunological re-activity which leads to the neuromuscular dis-order, leaving the overall immune response intact.Having a well defined antigen (AChR) which in-duces EAMG and which is the autoantigen inMG, one may try by molecular modifications andderivations, and by immunological analysis of thereceptor molecule to isolate and characterisesmall fragments of the molecule which are re-sponsible for the specific cholinergic or the path-ologic myasthenic activity of AChR, or both.Moreover, by such an analysis one may design andachieve derivatives of AChR with a potential toregulate specifically the immune response toAChR and to immunosuppress EAMG.Studies with denatured AChR: We have shownpreviously that a denatured AChR preparationwhich by itself does not induce EAMG has atherapeutic potential on this disease.10 16 26 De-naturation of AChR was achieved by completereduction and carboxymethylation of the receptorin 6M guanidine hydrochloride and the obtainedderivative was designated reduced-carboxymethy-lated AChR or RCM-AChR. RCM-AChR is de-void of the pharmacologic activity of the intactAChR and it does not bind 1251-a-bungarotoxin. Italso lost the myasthenic activity of AChR asrabbits immunised repeatedly with RCM-AChRdid not develop any clinical signs of EAMG. How-ever, such rabbits developed high titres of anti-bodies which cross-react with AChR. Differencesbetween the antigenic specificities of AChR anddenatured AChR were observed by both immuno-diffusion and radioimmunoassay (fig 2). AChR andRCM-AChR bind in an identical manner to anti-RCM-AChR whereas there is only a partial crossreactivity between anti-AChR and RCM-AChR,suggesting the presence in this antiserum of anti-bodies against some antigenic determinants whichdo not exist on the denatured receptor. Thus, themajor difference between AChR and RCM-AChRleading to their different pathogenicity, resides intheir different antigenic specificity. Some antigenicdeterminants in the AChR molecule were abolished

Li 0 i

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from

Sara Fuchs et al

/

6 '

5 '

* s oz~~

I ~~~~~~~~~~~~~~~..

Fig 2 Specificity of anti-A ChR andantibodies. (a) Immunodiffusion of Xserum (well 1) and anti-AChR serurRCM-AChR (wells 2, 3 and 5) and,7). (From Bartfeld and Fuchs.26) (b)binding of `I5-AChR to anti-A ChRanti-RCM-AChR serum (right) by ARCM-AChR ( --- ). (From Bartfelc

by the denaturation procedure.ditional determinants which werthe intact molecule became imidenaturation.The difference in specificity be

and anti-RCM-AChR is reflecteffects on the binding of bungarwhereas anti-AChR antibodies tbinding of "25I-a-bungarotoxineffectively, anti-RCM-AChR antbinding only to a very limited eiThe altered antigenic specific

to RCM-AChR along with theirblocking toxin binding to AChR

that the denaturation of AChR destroyed one ormore antigenic determinants, important for theinduction of EAMG, and which may be locatedclose to the toxin-binding site.The immunological, pharmacological and path-

ological properties of RCM-AChR made it ane'''°-:>Ww '' appropriate potential candidate for trying specific2 immunosuppression of EAMG. Indeed, RCM-

AChR which by itself is devoid of any myasthenicactivity was shown to have both preventive andtherapeutic effects on the experimental neuro-

3 muscular block induced by the intact receptor.'0 16/ The onset of EAMG was either delayed or com-

pletely prevented in rabbits preimmunised withRCM-AChR. In addition to the protective poten-tial of RCM-AChR, its therapeutic effect on my-asthenic rabbits was observed as well. A singleadministration of RCM-AChR in completeFreund's adjuvant to myasthenic rabbits duringthe initial stages of clinical development of EAMGled to a suppression of EAMG in at least 15 outof 30 myasthenic rabbits.'0 16 27There seemed to be a correlation between the

clinical conditions of the RCM-AChR treatedrabbits and the antigenic specificity of their im-mune response. There were differences both inantibody titres and antigenic specificity betweensera of rabbits in which EAMG was prevented orreversed by RCM-AChR and those of sick non-treated rabbits.'1 27 28 The preventive as well as thetherapeutic effect of RCM-AChR was accom-panied by a change in the specificity of the

I anti-RCM-A C/iR humoral immune response, from a specificity ofanti-RCM-AChR AChR-injected rabbits towards that of non-mn (well 4) with myasthenic RCM-AChR injected rabbits16 26-284ChR (wells 6 and (fig 3). All the rabbits in which EAMG wasInhibition of the eliminated following an injection of RCM-AChRserum (left) and exhibited a similar change of antibody specificity.4ChR ( ) and The cross-reactivity between AChR and RCM-I and Fuchs.'") AChR and the non-pathogenicity of the latter

appear to be crucial in governing the immuno-However, no ad- suppressive and therapeutic effects of RCM-AChRe not expressed in on EAMG. It is possible that the humoral re-munopotent after sponses which accompany the prevention, appear-

ance or suppression of EAMG represent relativetween anti-AChR levels of two different antibody populations againstted also by their AChR, one of which is specific to a myasthenicotoxin to AChR: determinant in the AChR molecule and is in-block the in vitro volved in the disease, whereas the other one isto AChR very directed to other antigenic determinants which

-ibodies block this are not involved in the disease. AChR can elicitKtent.26 antibodies to both types of determinants and it isity of antibodies the immune response against the myasthenic deter-r altered effect in minants which is responsible for induction ofled us to propose EAMG.

638

:.-

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from

Immune regulation of experimental myasthenia

Bleedings (a) (b) (c)

Days 0 EAMG 80 120 160Injections I I

AChR RCM-AChR

100l(a)AChR

RCM-AChR

(b)EAMG

V

.',o

1 100 10000 1Inhibitor added (ng)

100 10000 1

Fig 3 Therapy of EAMG by RCM-AChR. Correlation between the clinicalsymptoms and the antigenic specificity of the antibodies in a representativemyasthenic rabbit in which the disease had been reversed by treatment withRCM-AChR. The inhibition of the binding of "'I-AChR to the rabbit'sserum by AChR ( ) and RCM-AChR ( --- ) was measured for eachbNeeding. The course of immunisations, bleedings and development of EAMGare described schematically at the top of the figure. (From Bartfeld andFuchs.'6)

By using an immunoadsorbent of RCM-AChRbound to Sepharose it is possible to fractionatethese two antibody populations from antisera ofAChR-injected rabbits.29 The unadsorbed antibodyfraction (effluent) is of antibodies which bind onlyto AChR and not to RCM-AChR. This antibodyfraction is designated anti-native AChR (n-AChR)antibodies. It blocks the in vitro binding of toxinto AChR and probably contains antibodies againstmyasthenic determinants. The second antibodypopulation (eluate) which adsorbs to RCM-AChRdoes not block toxin binding to AChR and itrepresents antibodies against non-structural deter-minants of AChR present also on the denaturedreceptor. Although the latter antibody fraction(eluate, or anti-denatured AChR (d-AChR) anti-bodies) may be present in myasthenic AChR-injected rabbits, it is probably not involved in themyasthenic activity and may even have a regulat-ing effect on it.The therapeutic potential of denatured AChR

derivatives may have practical implications ifsimilar approaches can be worked out also forAChR from other species or whether denaturedTorpedo AChR can exert a therapeutic effect on

myasthenia induced by AChR from other sources.

Trypsin ated AChR: Enzymic degradations of

AChR provide a useful approach for attemptingthe isolation and analysis of fragments of thereceptor molecule which retain the specific cholin-ergic binding site or the pathologic myasthenicactivity or both. Isolation and characterisation ofthe minimal structural entity which governs themyasthenic activity of AChR may provide apowerful tool towards designing specific drugs formyasthenia.To this end we have treated AChR with trypsin

and have studied the biological activity of the re-sulting derivative.30 Tryptic digestion of AChRfrom Torpedo californica did not change thepharmacological specificity and the pathologicalmyasthenic activity of the receptor molecule. Theproduct obtained after tryptic digestion was re-

purified by affinity chromatography on a Najanaja siamensis toxin-Sepharose resin and was

designated T-AChR. T-AChR has a sedimentationcoefficient of 8-0 S and in SDS acrylamide gelelectrophoresis shows one major band with amolecular weight of about 27 000. Immunologicalanalysis reveals that T-AChR bind to anti-AChRantibodies directed against conformational deter-minants of AChR (anti-n-AChR) and binds very

weakly to antibodies directed against non-struc-tured determinants (anti-d-AChR)29 30 (fig 4). This

_80

c 600

D 40,C

20

0

-,

100 10000

639

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from

Sara Fuchs et al

is in contrast to the denatured receptor (RCM-AChR) which binds only to anti-d-AChR anti-bodies30 (fig 4). These results suggest that theexposed unfolded regions of AChR were digestedby the trypsin, retaining only the folded regionswhich are resistant to proteolytic activity. Thecholinergic binding sites as well as the myasthenicsites of AChR seem to reside within this struc-tured region. Immunisation of rabbits with T-AChR (one or two injections of 80 ,ug protein incomplete Freund's adjuvant) results in EAMGsymptoms identical to those obtained followingimmunisation with AChR.

100

80'

$ 60-0

40C

20'

0

(a)

T-AChR

RCM-AChR

1 100 10000Inhibitor added (ng)

-A

--1-

1 100

Fig 4 Antigenic specificity of T-AChR. Inhibition ofthe binding of "lI-AChR to anti-n-AChR antibodies(a) and to anti-d-AChR antibodies (b) by AChR( ), T-AChR ( --- ) and RCM-AChR ( ... ).

T-AChR represents an active receptor moleculewith a lower structural complexity than that ofthe intact detergent purified receptor. It seems thatthe polypeptide chains in T-AChR result from the40000 polypeptide of the intact receptor, which isknown to contain the cholinergic site. If this is thecase it may be concluded that the myasthenicactivity of AChR resides also within the 40000subunit of the receptor; this simplifies furtherstudies for the elucidation of the molecular originof the myasthenic activity of AChR and for de-signing specific drugs for myasthenia.Other modifications: Additional manipulationsof AChR are achieved by chemical modificationsof specific functional groups in the molecule.Polyalanylation of free amino groups in AChR byreacting them with N-carboxy-DL-alanine anhy-dride resulted in a derivative designated poly-DL-alanyl AChR (PA-AChR).31 32 This chemicalmodification did not affect the physiologic activity

of the receptor as measured by toxin bindingcapacity but it abolished its myasthenic activity.Immunological studies demonstrated that poly-alanylation resulted in a marked change in theimmunogenicity but not in the antigenic specificityof AChR. Antibodies from rabbits immunisedwith PA-AChR were directed mainly to the poly-alanine determinants and bound AChR only to avery limited extent. On the other hand the anti-genic reactivity of PA-AChR towards anti-AChRantibodies was as efficient as that of AChR. Suchimmunological specificity may render PA-AChRan effective agent for immunosuppression ofEAMG by antigenic competition or, alternatively,by neutralising the immune response to AChR.

Nitrated AChR obtained by reacting AChR withtetranitromethane was demonstrated to be devoidof both myasthenic and pharmacologic activity(Mochly-Rosen and Fuchs, in preparation). Thischemical modification is known to affect specific-ally tyrosine residues in proteins33 34 and it is thussuggested that at least some tyrosine residues inAChR are crucial for its biological activity.

Additional chemical modifications of AChR andits further enzymic or chemical degradation willcontribute to the detailed structure and functionanalysis of this molecule. Such an approach maylead to the elucidation of a small therapeutic frag-ment and its synthesis can be then attempted. Ifsuch a fragment would be common to variousnicotinic receptors its therapeutic potential mightbe broader.

Monoclonal antibodiesThe involvement of anti-AChR antibodies and ofspecific cell mediated immunity in the pathogenesisof MG and EAMG has been demonstrated onclinical and experimental grounds. Passive transferof immunoglobulins from myasthenic patients35and from animals with EAMG36 resulted in my-asthenia gravis-like symptoms. AChR is a com-plexed multideterminant immunogen which raisesa heterogeneous immune response and it is stillnot known which and whether certain specificity

E (ies) is (are) responsible for the myasthenic activityof the molecule. In order to analyse the structureof AChR and for identifying the molecular entitywhich governs its myasthenic activity, large quan-tities of antibodies with defined specificity are re-quired. Such antibodies provide a useful tool forstudying the role of antibodies in myasthenia andtheir mechanism of action. To this end we haveused the method of lymphocyte hybridisationdeveloped by Kohler and Milstein37 and haveprepared hybridomas by hybridisation of mouse

r-,.......

64-A0

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from

Immune regulation of experimental myasthenia

anti-AChR antibody producing cells with a non-secreting myeloma line. Such hybridomas secretemonoclonal anti-AChR antibodies with restrictedspecificities towards defined determinants ofAChR.38Our findings suggest that some of the hy-

bridomas secrete antibodies which recogniseselectively conformational antigenic determinantsof AChR, which are present in the trypsinatedreceptor30 and are absent in the denatured re-

ceptor.26 The myasthenic activity of the receptormolecule resides within this group of determinants.Thus it may be expected that antibodies directedagainst these specificities may have pathologicaleffect associated with myasthenic activity. On theother hand, other hybridomas secrete antibodieswhich recognise selectively denatured AChR,which expresses the non-structural antigenicdeterminants of the receptor molecule. Such anti-bodies are probably not associated with any my-asthenic activity and may even have a perventiveor therapeutic activity.We have recently developed one hybridoma

which secretes antibodies directed against thecholinergic binding site of AChR. The binding ofthese monoclonal antibodies to AChR is inhibitedby bungarotoxin and by several cholinergic agents.Although this antibody specificity is not detectedin immune anti-AChR sera, it may be presentthere in minute quantities and may even play a

crucial role in the pathogenesis of myasthenia.The ability to select preferentially for one speci-ficity, even a minor one, is one of the advantagesoffered by the lymphocyte hybridisation method.

Thus, monospecific antibodies appear to be a

useful tool for structural and functional analysisof AChR; they may help to define the sites in-volved in the various biological activities relatedto AChR. Moreover, they may enable the elucida-tion of the role of antibodies of defined specificity,in regulating the immune response to AChR andthe disease resulting from it.

A nti-idiotypesAnti-idiotypes may have a regulatory role in auto-immune diseases or other disorders of an immun-ological nature. The application of anti-idiotypicantibodies furnishes a possible approach forspecific regulation of the immune response toAChR and of myasthenia. Anti-idiotypic serum

specific to anti-AChR idiotypes was prepared inC57BL/6J mice by repeated injections with puri-fied C57BL/6J anti-AChR antibodies or withsyngeneic spleen cells educated with AChR.39 Theanti-idiotypic serum reacted specifically with anti-

AChR antibodies of several mouse strains and ofother species. The anti-idiotypic serum was alsoable to inhibit the binding of 125I-AChR to mouseanti-AChR antibodies, suggesting that idiotypicdeterminant(s) against which the serum is directedare associated with the antigen combining site.39The broad cross-reactivity among anti-AChRidiotypic determinants of different mouse strainsas well as those of different species suggest thatsimilar idiotypic determinants exist in the anti-AChR antibodies tested.The approach of anti-idiotypes is hampered by

the use of a heterogenous population of anti-AChR antibodies for the elicitation of anti-idiotypic serum. Antibodies obtained from AChR-immunised animals contain a whole variety ofdifferent idiotypes, part of which may not berelevant for myasthenia. Now that monoclonalanti-AChR antibodies are available it may beadvantageous to raise anti-idiotypic serum againstmonoclonal antibodies of characterised definedspecificity.

Measles virus (possible viral involvement?)The etiology of MG is still not clear; a viral in-fection has been proposed as one of the possibletriggering factors.40 Certain diseases of the centralnervous system such as subacute sclerosing pan-encephalitis4' and possibly multiple sclerosis42 areassociated with measles virus infection. In view ofthis and also since we faced some difficulties inobtaining reproducible EAMG with high incidencein mice, we have studied recently the effect ofmeasles virus vaccine on the induction of EAMGin mice (Feingold, Tarrab-Hazdai and Fuchs, inpreparation). Our preliminary results indicatethat injection of measles virus vaccine in combin-ation with AChR facilitate the onset of EAMG inC57BL/6J mice. The results of three experimentsare summarised in table 2. In all three experimentsan early onset of EAMG was observed folowingone injection of AChR together with measlesvirus vaccine. The specificity of the disease wasconfirmed by the Tensilon test. With AChR aloneat least two injections were required for the induc-tion of EAMG.24 The effect of the measles virusappeared to be restricted to mice susceptible toEAMG, since immunisation of SJL/J mice (anon-susceptible strain24) with a combination ofAChR and measles virus vaccine had no effect onthe induction of the disease. Also, the stimulatingeffect of measles virus vaccine on C57BL/6J micewas not observed by another type of virus such asbacteriophage T4.The mechanism by which measles virus regulates

641

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from

Sara Fuchs et al

Table 2 Effect of measles virus vaccine on EAMG

Treatment Day Score oftested clinical

signs

Exp. I AChR+measles on day 0; 30 0AChR on day 140

AChR on day 0; AChRon day 140

40 ++45 +++50 +++145 +++180 +++30 050 0145 ++180 ++

Measles on day 0; AChR 30 0on day 140 50 0

145 0180 0

Exp. 2 AChR+measles on day 0; 15 +measles on day 75 30 + + +

45 ++85 +++110 +++125 +++

AChR on day 0; imeasles 30 0day 75 45 0

85 ++110 +125 I

Measles on day 0; 30 0Measles on day 75 45 0

30110 0125

Exp. 3 AChR+nieasles on day 0; 5 +AChR+measles on day 35 30 + +

40 +++50 +++

AChR on day 0; AClhRon day 35

Measles on day 0;measles on day 35

30 040 060 0

30 040 060 0

C57BL/6J mice (7 mice in each group) were injected in the footpads inCFA with AChR (10 fig), measles virus vaccine (25 gA from a I mlsolution oflive attenuated virus preparation (Merck, Sharp and Dohm))or a combination of both. The mice were observed daily for signs ofmuscular weakness and other visual symptoms of the disease whichwere graded according to the following score: 0=no weakness; ± =weak grasping and fatigability in part oft}e group; + =weak graspingand generalised weakness in the whole group; ++=ruffled fur;hunched posture, uncoordinated movements and sometimes paralysisoflimbs; + + + =loss ofability to grasp, tremulous, severe generalisedweakness.

EAMG is still not worked out. Its effect may beexerted by an adjuvant effect or by cross reactivitywith AChR. Preliminary experiments support thelatter possibility, as it was demonstrated that a

second immunisation of mice with measles virus,following a first challenge with AChR or AChRand measles virus, has a boosting effect on theanti-AChR antibody titre. Furthermore, in some

experiments mice immunised with measles virusalone had a low but significant anti-AChR anti-

body titre. Additional data should be accumulatedbefore a conclusion can be drawn concerning themechanism of the measles virus effect, andwhether it can somehow explain its involvementin the aetiology of the disease.

This work was supported by grants from theMuscular Dystrophy Association of America, theLos Angeles Chapter of the Myasthenia GravisFoundation and the United States-Israel BinationalScience Foundation (BSF).

References

1 Simpson JA. Myasthenia gravis: a new hypothesis.Scot Med J 1960; 5:419-36.

2 Nastuk WL, Plecia OJ, Osserman KE. Changesin serum complement activity in patients withmyasthenia gravis, Proc Soc Exp Biol Med 1960;105:177-84.

3 Jenkins RB. Treatment of myasthenia gravis withprednisone. Lancet 1972; 1:767-7.

4 Warmolts JR, Engel WK. Benefit from alter-native day prednisone in myasthenia gravis. NewEngl J Med 1972; 286:12-20.

5 Seybold ME, Drachman DB. Gradually increasingdoses of prednisone in myasthenia gravis: reduc-ing the hazards of treatment. New Engl J Med1974; 290:81-4.

6 Matell G, Bergstrom K, Franksson C, et al.Effects of some immunosuppressive procedureson myasthenia gravis. Ann NY Acad Sci 1976;274:659-76.

7 Abramsky 0, Aharonov A, Webb C, Fuchs S.Cellular immune response to acetylcholinereceptor-rich fraction in patients with myastheniagravis. Clin Exptl immunol 1975;19:11-6.

8 Aharonov A, Abramsky 0, Tarrab-Hazdai R,Fuchs S. Humoral antibodies to acetylcholinereceptor in patients with myasthenia gravis.Lancet 11:340-2.

9 Abramsky 0, Aharonov A, Teitelbaum D, FuchsS. Myasthenia gravis and acetylcholine receptor:Effect of steroids in clinical course and cellularimmune response to acetylcholine receptor. ArchNeurol 1975; 32:684-7.

10 Fuchs S. Immunosuppression of experimentalmyasthenia. In: Dau PC, ed. Plasmapheresis andthe immunobialogy of myasthenia gravis. Hough-ton Mifflin 1979; 20-31.

11 Abramsky 0, Tarrab-Hazdai R, Aharonov A,Fuchs S. Immunosuppression of experimentalautoimmune myasthenia gravis by hydrocortisoneand azathioprine. J Immunol 1976; 117:225-8.

12 Tarrab-Hazdai R, Abramsky 0, Fuchs S. Im-munosuppression of experimental autoimmunemyasthenia gravis by azathioprine. II. Evaluationof immunological mechanism. J Immunol 1977;119:702-6.

642

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from

Immune regulation of experimental myasthenia

13 Schwartz M, Lancet D, Tarrab-Hazdai R, FuchsS. Effect of azathioprine on the affinity of anti-bodies against acetylcholine receptor: analysiswith purified antibodies. Mol Immunol 1979; 16:483-7.

14 Tarrab-Hazdai R, Schwartz M, Fuchs S. Immuno-logical memory following immunosuppression ofexperimental autoimmune myasthenia gravis withazathioprine. Immunology Lett 1979; 1:101-4.

15 Werblin TP, Siskind GW. Distribution of anti-body affinities: Technique of measurement. Im-munochemistry 1972; 9:987-1011.

16 Bartfeld D, Fuchs S. Specific immunosuppressionof experimental autoimmune myasthenia gravisby denatured acetylcholine receptor. Proc NatlAcad Sci 1978; 75:4006-10.

17 Bach JF. Alkylating agents. Frontiers of Biology.1975; 41:173-225.

18 Otternes IG, Chang YH. Comparative study ofcyclophosphamide, 6-mercaptopurine, azathio-prine and methotrexate. Relative effects on thehumoral and cellular response in the mouse. ClinExp Immunol 1976; 26:346-54.

19 Easmon CSF, Glynn AA. Effect of cyclophos-phamide on delayed hypersensitivity to Staphy-lococcus aureus in mice. Immunology 1977; 33:767-76.

20 Parker D, Turk JL. The effect of cyclophospha-mide pretreatment on B-cell stimulation. Dissoci-ation of action on homocytotropic antibody andother B-cell functions. Immunology 1978; 34:115-21.

21 Stockman GD, Heim LR, South MA, TrentinJJ. Differential effects of cyclophosphamide onthe B and T cell compartments of adult mice.J Immunol 1973; 110:277-82.

22 L'Age Stehr J, Diamanstein T. Induction of auto-reactive T lymphocytes and their suppressor cellsby cyclophosphamide. Nature 1978; 271:663-5.

23 Tindall RSA, Cloud R, Luby T, Rosenberg RN.Serum antibodies to cytomegalovirus in my-asthenia gravis. Neurology 1978; 28:273-7.

24 Fuchs S, Nevo D, Tarrab-Hazdai R, Yaar I.Strain differences in the autoimmune response ofmice to acetylcholine receptors. Nature 1976;263:329-30.

25 Vadas MA, Miller JFAP, Gamble J, Whitelaw A.A radioisotopic method to measure delayed typehypersensitivity in the mouse. I. Studies on sen-sitized and normal mice. Int Arch Allergy ApplImmunol 1975; 49:670-92.

26 Bartfeld D, Fuchs S. Immunological character-ization of an irreversibly denatured acetylcholinereceptor. FEBS Lett 1977; 77:214-8.

27 Fuchs S, Tarrab-Hazdai R, Nevo D, Bartfeld D.Experimental auto-immune myasthenia gravis: atool for studying the pathogenesis and therapy ofmyasthenia gravis. In: Lunt G, Marchbanks RM,edh The biochemistry of myasthenia gravis and

643

muscular dystrophy. Academic Press 1978; 189-216.

28 Fuchs S. Immunochemical analysis of acetylcho-line receptor and its relevance to specific treat-ment of experimental autoimmune myastheniagravis. Adv Cytopharmacol 1979; 3:279-85.

29 Bartfeld D, Fuchs S. Fractionation of antibodiesto acetylcholine receptor according to antigenicspecificity. FEBS Lett 1979; 105:303-6.

30 Bartfeld D, Fuchs S. Active acetylcholine re-ceptor fragment obtained by tryptic digestion ofacetylcholine receptor from Torpedo californica.Biochem Biophys Res Commnun 1979; 89:512-9.

31 Schmidt-Sole J, Tarrab-Hazdai R, Fuchs S. Im-munological and physiological activity of poly-alanyl acetylcholine receptor. Israel J Med Sci1977; 13:1043.

32 Fuchs S. Immunology of the nicotine acetylcholinereceptor. Curr Top Microbiol Immunol 1979; 85:1-29.

33 Sokolovsky M, Riordan JF, Vallee BL. Tetranitro-methane. A reagent for the nitration of tyrosylresidues in proteins. Biochemistry 1966; 5:3582-9.

34 Fuchs S, Gurari D, Silman I. Chemical modifi-cation of electric eel acetylcholinesterase by tetra-nitromethane. Arch Biochem Biophys 1974; 165:90-7.

35 Toyka KV, Drachman DB, Griffin DE, PestronkA, Winkelstein JA, Fischbeck KH, Kao I. My-asthenia gravis. Study of humoral immune mech-anisms by passive transfer to mice. New Eng JMed 1977; 296:125-31.

36 Lindstrom JM, Engel AG, Seybold ME, LennonVA, Lambert EH. Pathological mechanisms inexperimental autoimmune myasthenia gravis. II.Passive transfer of experimental autoimmune my-asthenia gravis in rates with anti-acetylcholine re-ceptor antibodies. J Exp Med 1976; 144:739-53.

37 Kohler G, Milstein C. Continuous cultures offused cells secreting antibody of predefined speci-ficity. Nature 1975; 256:495-7.

38 Moshly-Rosen D, Fuchs S, Eshhar Z. Monoclonalantibodies against defined determinants of acetyl-choline receptor. FEBS Lett 1979; 106:389-92.

39 Schwartz M, Novick D, Givol D, Fuchs S. In-duction of anti-idiotypic antibodies by immuniza-tion with syngeneic spleen cells educated withacetylcholine recep:or. Nature 1978; 273:543-5.

40 Dalla SK, Schwartz RS. Infectious(?) myasthenia.New Engl J Med 1974; 291:1304-5.

41 Horta-Barbosa L, Fuccillo DA, Sever JL, ZemanW. Subacute sclerosing panencephalitis. Isolationof measles virus from a brain biopsy. Nature1969; 221:974.

42 Brown P, Cathala F, Gajdusek DC, Gibbs CJ.Measles antibodies in the cerebrospinal fluid andpatients with multiple sclerosis. Proc Soc Exp BiolMed 1971; 137:956-61.

Protected by copyright.

on 15 August 2019 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.7.634 on 1 July 1980. Dow

nloaded from