antigen recognition by t-cells and its suppression · ann. rheum. dis. (1971), 30, 565 antigen...

Ann. rheum. Dis. (1971), 30, 565

Antigen recognition by T-cellsand its suppressionSignificance and origin of rosette-forming cells

JEAN-FRANtOIS BACHHopital Necker, Paris

Antigen recognition is one of the most debatedaspects of the immune response and its mechanismis central to most theories of antibody formation.The clonal selection theory of Bumet (1959), whichseems to be widely accepted at the moment, holdsthat lymphocyte clones are already in existence atbirth, each bearing receptors specific for one antigen.Recent studies have shown that three cell popula-tions are involved in immune responses, at least forsome antigens:(1) Macrophages;(2) Thymus-dependent (T) lymphocytes;(3) Bone-marrow-dependent (B) lymphocytes.The existence and mechanism of co-operation ofthese three cell types are still matters of discussion.One of the most debated points, and one which willinterest us here, is the specificity at T and B-celllevels.

Several techniques have been proposed for study-ing antigen binding by lymphoid cells, the mostfrequently used being bacterial adherence, auto-radiography of cells having bound isotope-labelledantigens and, especially, immunocytoadherence (therosette test), which was described by Nota, Liaco-poulos-Briot, Stiffel, and Biozzi (1964) and Zaalberg(1964) and developed by Biozzi, Stiffel, Mouton,Bouthillier, and Decreusefond (1968). This antigenrecognition in vitro has been shown to be a model ofrecognition in vivo. In other words, cells which bindantigens in vitro are the same as those which bindantigens in vivo, this being the first step in antigenrecognition (Bach, Muller, and Dardenne, 1970).Antigen binding can be prevented in several ways,including removal of antigen, neutralization of thereceptors responsible for the binding and membranemodification making the receptors unavailable forantigen.

The Clones

The clonal selection theory of Burnet (1959) issupported by the fact that antibody-forming cells (asstudied by the local haemolysis in gel technique) arerestricted in specificity, class, and allotype (Dubiskiand Fradette, 1966). One particular plaque-formingcell generally produces antibodies of one class, oneallotype, one specificity, and one affinity, with a fewexceptions (Makela and Cross, 1970; Liacopoulos,1971). While this is suggestive of clonal populationsof lymphocytes, direct proof of the existence ofclones before immunization is needed to support theclonal selection theory. Such direct proof has beenafforded by a number of workers who have been ableto remove one clone of a given specificity from apopulation of normal lymphoid cells, leaving themunresponsive to the particular antigen. Three mainmethods have been applied for this purpose: cellretention on antigen-coated columns, specific killingby isotope-labelled antigens, and removal ofrosette-forming cells by centrifugation on gradient.

The first experiments were those of Wigzell (1970),showing that it was possible to deplete a cell popula-tion of its antigen-sensitive lymphocytes by passingthe cells down an antigen-coated column. Cells withhigh affinity are retained more easily than cells withlow affinity. The retention is inhibited by thepresence of free antigen in the column. These experi-ments were done primarily with immune cells buthave also been performed with normal cells. Thissystem works for both thymus-dependent antigens(e.g. serum albumin), against which the immuneresponse needs the presence of the thymus, and forthymus-independent antigens (e.g. polyvinyl pyroly-done). The experiments of Truffa-Bachi and Wofsy(1970), using a different type of antigen-coatedcolumn, have confirmed those of Wigzell (1970) and

Lecture given at the VII European Rheumatology Congress, Brighton, June, 1971.

copyright. on July 17, 2021 by guest. P

rotected byhttp://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.30.6.565 on 1 Novem

ber 1971. Dow

nloaded from

http://ard.bmj.com/

566 Annals of the Rheumatic Diseases

have provided a method of eluting highly specificcell populations.The experiments of Ada and Byrt (1969) and of

Humphrey and Keller (1971) showed that it was

possible to kill specifically antigen-sensitive cellswhen the particular antigen was heavily labelled withan isotope and mixed with the cells. In brief, Adaand Byrt mixed normal spleen cells with 125L-labelledflagellin A and non-labelled flagellin B for 30 min.at 400 C. These cells were used for the reconstitutionof irradiated mice, which then proved to be un-

responsive to flagellin A, while retaining normalresponsiveness to flagellin B. Similarly, Dutton andMishell (1967) have shown with the hot-pulse tech-nique that there is a population of cells specificallyinvolved in antigen recognition towards sheeperythrocytes which is completely different from thatreacting towards red blood cells of other species.Lastly, affinity labelling has been used (Plotz, 1969;Segal, Globerson, Feldman, Haimovich, and Givol,1969) to make receptors specifically unreactive.We have performed experiments in which we have

removed from normal cell populations those cellsforming rosettes with sheep erythrocytes. The re-

maining cells were then unresponsive to this antigen.The starting hypothesis was that cells recognizingheterologous red cells formed rosettes. In order toobtain depletion of spontaneous rosette-formingcells (sRFC) from normal spleen cells, we havemixed the latter with red cells so as to form therosettes and then put the cell suspension on a

Ficoll/Triosil mixture. Such a mixture is generallyused to separate lymphocytes from red cells (Bach,Muller, and Dardenne, 1970). After centrifugation,in our particular experiment, rosette-forming lym-phocytes which were coated with red cells behavedlike erythrocytes and went through the Ficoll, thusbecoming separated from the other lymphocytes.The immunological competence of the depletedpopulation was tested by restoring animals madeimmunologically anergic by a cyclophosphamideinjection (300 mg. intraperitoneally). The results(Bach, Muller, and Dardenne, 1970) showed thatdepletion of cells forming rosettes with sheeperythrocytes induced unresponsiveness against sheepred blood cells. This was true both at the humorallevel (haemagglutinins) and the cellular level (plaque-forming cells, rosette-forming cells). As controls, weshowed that the passage on Ficoll in the presence ofred cells but without rosette formation, and rosetteformation and incubation at 37°C. in Ficoll butwithout centrifugation, did not induce a sinilarunresponsiveness, thus excluding tolerance induc-tion in vitro.

Similar results have been obtained in immunizedmice. The procedure was similar to that describedabove for normal cells, the only difference being that

immune cells were taken instead of normal cells. It isthus suggested that memory cells for heterologouserythrocytes bind antigen and form rosettes.

These experiments and the others quoted abovestrongly suggest that there are specific clones in theunimmunized mouse. The size of these clones isdifficult to estimate. Several figures have beenproposed for sheep red blood cells (SRBC). Makino-dan and Albright (1963), using cell dilution tech-niques, proposed 150 antigen-sensitive cells perspleen; Playfair, Papermaster, and Cole (1965) andKennedy, Till, Siminovitch, and McCulloch (1966),by the colony technique, gave figures of 5,000;Biozzi and others (1968), by extrapolating the kine-tics of rosette-forming cells (RFC), proposed similarfigures of 2,000 to 6,000. Looking at spontaneousRFC, we have found significantly higher figures.Similarly, Byrt and Ada (1969) and Humphrey andKeller (1971) also obtained high figures, around1 antigen-binding cell per 1,000 cells. These twokinds of evaluation may be reconciled if one acceptsthat antigen-binding cells represent a mixture ofmacrophages, B and T-cells (as will be detailed later).Moreover, it is known that antigen-binding cellshave a wide range of affinity and that probably onlyhigh affinity cells are involved in one particularresponse. In fact, when rosettes are formed with alow erythrocyte/leucocyte ratio, fewer RFC arefound (- 0- 1 per cent. for a ratio of 0 1) (Bach,Reyes, Dardenne, Fournier, and Muller, 1971).This explains why, when looking at rosettes orantigen-binding cells, one should consider onlythe high affinity B or T-cells as the clone involvedin one response. On the whole, it seems reason-able to evaluate the size of the clones at about1 per 104 lymphocytes for red cells (which have acomplex mosaic of antigens) and perhaps five totwenty times less for simpler antigens.

CeLlular co-operationIt has been shown for heterologous red cells and forsome other antigens, such as serum proteins, thatantibody production needs the co-operation of twopopulations of lymphocytes: one having beenprocessed by the thymus (T-lymphocytes), the othercoming directly from the bone marrow (B-lympho-cytes). Macrophages also play a role at the first stepof the immune response by transmitting 'superantigens' to lymphocytes. The mechanism of thiscellular co-operation is still uncertain. From all thepublished work the following scheme may bepresented: the T-cells recognize the antigen andpresent it to the B-cells. It might be, at least incertain systems, that the T-cells will recognizeantigenic determinants different from those recog-nized by the B-cells. In a hapten-carrier system, theT-cells would recognize the carrier and the B-cells



j.com/

Ann R

heum D


ber 1971. Dow

nloaded from

http://ard.bmj.com/

Antigen recognition by T-cells and its suppression 567

the hapten. T-cells would mainly have a function ofantigen concentration on B-cells (Miller andMitchell, 1969; Mitchison, 1971). The mechanismof antigen concentration by T-cells is still a matterfor speculation. One possibility is that the T-cellshave a lower threshold of sensitivity to antigen witha wider range of specificity. The B-cells wouldproduce the anti-hapten antibody. The effects ofneonatal thymectomy, which suppresses the T-cellpopulation, can be overcome by the injection of highdoses of antigen, short-circuiting the simple helperfunction of T-cells for B-cells. In the first days afterantigen stimulation T-cells divide, as has been shownby Davies (1969) using a chromosome marker. Someof the T-cells can recirculate, which differentiatesthem from B-cells. The main methods used todeplete T-cells are neonatal thymectomy, anti-lymphocyte serum (ALS) treatment, and thoracicduct drainage. B-cells are responsible for antibodyproduction, as shown by chromosome or antigenicmarkers (Miller and Mitchell, 1969; Nossal, Cun-ningham, Mitchell, and Miller, 1968). The mechan-ism of antigen recognition is not well understood.One of the main problems is the location of speci-ficity. Is specificity the same at the B and T level or isit peculiar to one of these two populations?

SPECIFICITY AT THE B-CELL LEVELThere are several arguments in favour of this. DNAsynthesis can be specifically induced in bone marrowcells and can be specifically eliminated by passage ofthe cells on antigen-coated columns (Singhal andRichter, 1968; Wigzell, 1970). Bone marrow RFCare specific for the antigen (Bach, Reyes, and others,1971) and the depletion of RFC from the bonemarrow makes it incapable of co-operating with thethymus (Brody, 1970). Moreover, in a carrier-haptensystem, Wigzell (1970) has shown that the depletioninvolves anti-hapten cells rather than anti-carrierT-ells. Also, responses against antigen not needingthe presence of thymus can be suppressed by passageof the cells on antigen-coated columns (Wigzell,1970). Lastly, Playfair (1969), Chiller, Habicht, andWeigle (1970), and Argyrls (1968) have shown that,when a mouse is tolerant towards certain antigens,the tolerant cells can be found in the bone marrow.On the other hand, it is not possible to induce toler-ance in a thymus-marrow co-operation system byincubating bone-marrow with antigens heavilylabelled with radioisotope (Basten, Miller, Warner,and Pye, 1971).

SPECIFICITY AT THE T-CELL LEVELThis has also been demonstrated by a number ofexperiments. Double transfer experiments (Clamanand Chaperon, 1969; Miller and Mitchell, 1969) andthe study of T-cell mitoses (Davies, 1969) have shown

that stimulation of T-cells is specific. In the carrier-hapten system already mentioned, the carrier effect,which is mainly thymus-dependent, is specific for theantigen. When tolerance is induced in vivo, thethymus is made specifically tolerant and unable toco-operate with bone marrow for this antigen(Isakovic, Smith, and Waksman, 1965; Miller andMitchell, 1969; Taylor, 1969; Chiller and others,1970). Tolerance in T-cells is easier to induce thantolerance in B-cells. Greaves (1970) has shown that,when irradiated mice are reconstituted with syn-geneic bone marrow cells and semi-syngeneicthymus cells, a large portion of RFC is of thymicorigin (indicated by specific inhibition using anti-H2serum). These data indicate that specificity exists inT-immune cells, although this specificity might beless evident than in B-cells. One of the main argu-ments in favour of this low specificity is the highnumber of antigen-sensitive cells found in a pureT-cell system, like graft versus host (GVH) (Nisbetand Simonsen, 1967) and mixed lymphocyte culture(MLC) (Wilson, Klyth, and Nowell, 1968). This highpercentage, 1 to 2 per cent. of lymphocytes, is amatter of speculation. It is fundamental to the newtheory proposed by Jerne (1971) for antibodyformation.

Origin of rosette-forming cells (RFC)Rosette-formation represents an interesting modelfor the study of the origin of antigen-sensitive cells;as rosette-forming lymphocytes present in bonemarrow, spleen, or thymus are specific for anantigen (Bach, Reyes, and others, 1971), determina-tion of the origin of RFC gives information aboutthe specificity of the lymphocyte population.

Bone marrow sRFCThese bear immunoglobulin receptors for the anti-gen. They probably include:(1) Cells co-operating with thymus, since theirremoval specifically induces inability of bone mar-row cells to co-operate with thymus (Brody, 1970);(2) Precursors of T-cells, since incubation of bonemarrow cells with thymosin produces cells able toform rosettes and having the attributes of T-cells,namely theta antigen and sensitivity to azathioprineand ALS (Bach, Dardenne, Goldstein, Guha, andWhite, 1971).

Thymus sRFCThese are present only in small numbers in normalthymus (20 to 100 per 106 cells). However, after aninjection of hydrocortisone, whereas the total num-ber of thymus cells is drastically reduced, thepercentage of RFC increases equivalent to that ofthe spleen (Bach and Dardenne, 1971a). This is a



j.com/

Ann R

heum D


ber 1971. Dow

nloaded from

http://ard.bmj.com/


general feature of thymus cells, since all the thymuscells able to induce graft versus host reaction(Blomgren and Andersson, 1969; Cohen, Fischbach,and Claman, 1970) and to co-operate with bonemarrow (Andersson and Blomgren, 1970) are alsopresent in this hydrocortisone-resistant pool.

Spleen sRFCThere are also T-RFC in the spleen, since the unre-sponsiveness induced by removal of sRFC can bepartially reconstituted by thymus cells (Bach andDardenne, 1971a). Such spleen sRFC are present at anormal level in neonatally thymectomized animals(Bach and Dardenne, 1971a) but are then insensitiveto azathioprine and ALS (Bach and Dardenne,1971b), suggesting that T-cells possess receptorsbefore being processed by the thymus. Moreover,75 per cent. of spleen sRFC bear the theta antigen(Bach, Muller, and Dardenne, 1970; Bach andDardenne, 1971a) which is characteiistic of T-cells,as described below.Adult thymectomy eliminates theta-positive aza-

thioprine sensitive RFC within 5 days, suggestingthat spleen T-RFC are short lived (Bach, Dardenne,and Davies, 1971).

Theta-negative spleen sRFCThese are probably antigen-sensitive B-cells, sincethe unresponsiveness induced by removal of anti-sheep sRFC is partially reconstituted by bonemarrow cells. This reconstitution with bone marrowcells becomes total when studying anti-chickenRFC (Bach and Dardenne, 1971a).

Lymph node sRFCThese are 90 per cent. theta-positive. They areprobably long lived and they recirculate since theyare not modified by adult thymectomy. Theydisappear 6 hours after an injection of ALS. (Ourunpublished data.)

T-cell markersTheta (0) is an iso-antigen, genetically determined inmice according to two alleles, theta AKR and thetaC3H, Schlesinger and Yron (1970) and Raff (1969)have shown that theta is a marker of T-cells. It ispresent in thymus and brain and also in thymus-derived cells. Neonatal thymectomy and anti-lymphocyte serum (ALS) treatment lead to thedisappearance of theta-positive cells in lymph nodes(Raff and Wortis, 1970; Schlesinger and Yron, 1969).We have shown that azathioprine and ALS inhibit

sRFC exactly as anti-theta serum does in all systemstested: neonatal thymectomy (Bach and Dardenne,1971b), adult thymectomy (Bach, Dardenne, andDavies, 1971), and distribution in the lymphoidorgans (Bach and Dardenne, 1971b). The superiorityof azathioprine over theta as a marker of T-cells is

the probable lack of species specificity of azathio-prine and its possible application in man. Moreover,the use of a chemical substance avoids the difficultyof standardization needed for biological products,which is particularly true for anti-theta serumcontaining several categories of antibodies.

Suppression of antigen recognitionAntigen recognition occurs by means of antibody-like specific receptors located in the lymphocytemembrane; hence a number of experimental condi-tions can suppress antigen recognition: removal ofantigen (injection of passive antibodies), neutraliza-tion of receptors (anti-Ig serum), and suppression ofreceptor availability without direct action on thereceptors themselves (azathioprine, ALS). Otherpossibilities could be put forward, but only thoselisted above will be discussed here.

(1) REMOVAL OF ANTIGEN (Uhr and Moller, 1968;Wigzell, 1969) (Suppression of antigen recognitionby passively administered antibody).Primary antibody formation can be prevented bymixing the antigen with excess antibody beforeinjection, or injecting antibody before, simultan-eously, or even after the antigen. The phenomenonseems to be a very general one. It might represent afeed-back function of antibody on its own formation.The mechanism by which antibody synthesis issuppressed by passively administered antibody in-volves primarily a blocking of antigen site andelimination of the antigen in the form of an antigen-antibody complex. Some authors have suggested acentral action at the lymphoid cell level. In favourof this hypothesis is the observation that incubationof normal spleen cells with anti-red blood cellantibodies makes these cells unresponsive to redcells. However, these results have been contradictedin other experiments. The action of passivelyadministered antibody is specific for the antigen andresides in the Fab fragment of the immunoglobulin(Lg). Inhibition of antigen recognition can beobtained even when injecting the antibody severaldays after the antigen, indicating that the majority ofantibody-producing cells during the early stage ofthe immune response are short-lived in the absenceof antigen. Another point which merits comment isthe much greater efficiency of high affinity antibodiesin suppressing the immune response, which fits withthe antigen removal hypothesis generally accepted toexplain this suppression (Uhr and Moller, 1968).7S antibodies (IgG) which are usually more avidthan 19S antibodies (IgM) are the more efficient insuppressing immune responses. The main clinicalapplication of antibody-induced immunosuppres-sion is the injection of anti-D antibodies to pregnantwomen, but enhancement in kidney grafts is alsobeing attempted.



j.com/

Ann R

heum D


ber 1971. Dow

nloaded from

http://ard.bmj.com/


(2) ANTI-IMMUNOGLOBULIN SERUMThe receptors responsible for antigen recognition,present on the surface of antigen-sensitive cells, havebeen extensively studied in several situations. Whenanti-allotype antibodies are injected into pregnantmice, the newborn are not able to produce antibodiesof the allotype that they have in their genome andagainst which the anti-allotype serum is directed(Dubiski and Fradette, 1966). Primary stimulationin vitro against SRBC (Greaves, 1970) or againstallogeneic lymphocytes in the mixed lymphocytereaction (Greaves, Torrigiani, and Roitt, 1969) canbe inhibited by anti-light chain antibodies; moreover,in Wigzell's system of antigen-coated columns, theretention of specific cells is abolished by preincuba-tion of the cells with anti-Ig antibodies (Wigzell,1970). Lastly, the graft versus host reaction can betransferred to irradiated mice but the transfer maybe prevented if the cells are incubated with lightchain antibodies before transfer (Mason and Warner,1970). The nature of the Ig responsible for thisantigen recognition system is not known. It seemsthat, for delayed hypersensitivity or transplantationimmunity, only anti-light chain antibodies areefficient in inhibiting the response, whereas inhumoral responses, anti-heavy chain sera are alsoefficient. In fact, the two antigen-binding reactionsdescribed above (isotope-labelled antigen bindingand rosette formation) are at present the best toolsfor the study of receptors of antigen recognition.Greaves (1970) has shown that, in normal animals,RFC are inhibited only by anti-light chain sera, noinhibition being obtained with antisera directedagainst heavy chains. This observation is in contrastwith what is observed with immune RFC, where alarge proportion of the cells is inhibited by anti-itchain sera (at the early stage of the response) andanti-gamma chain sera (at the late stages of theresponse). Warner, Byrt, and Ada (1970) haveobtained some slightly different results with inhibi-tion of flagellin-binding cells, where anti-IA seraproved to be efficient in normal mice. On the otherhand, Basten, and others (1971) have obtainedinhibition of bovine serum albumin-bindingcells in the thymus with anti-light chain seraonly.On the whole it is possible to inhibit immune

reactions by incubating normal or immune sensitivecells with anti-Ig serum. For B-ells it appears thatinhibition can be obtained by both anti-light chainand anti-heavy chain sera (p chain especially). Forthymus derived (T) cells inhibition is obtainedonly with anti-light chain sera. No positive resultshave been reported for anti-heavy chains. Onepossibility is that receptors on T-ells have heavychains of a new class, IgX; another is that the anti-genic portion of the Ig receptor is buried in the

lymphocyte membrane. Some authors doubt thepresence of antibody-like receptors on T-cells. Ourfinding that hydrocortisone-resistant thymus sRFC,which are probably pure specific T-cells, are inhibitedby anti-Ig antisera (Bach and Dardenne, 1971a) is infavour of the existence of T-cell receptors.

(3) ANTILYMPHOCYTE SERUM (ALS)ALS inhibits cell-mediated immune responses andthose humoral antibody responses depending on thepresence ofthe thymus (Bach, 1970a). The inhibitionof such humoral antibody production is obtainedonly when ALS is injected before the antigen orsimultaneously. ALS appears to act mainly onT-cells (at least at low dosage), since:(i) Immuno-incompetence induced by ALS treat-ment can be partially reconstituted by injection ofthymus cells (Martin and Miller, 1968) or thymosin,a cell-free thymus extract (Quint, Hardy, andMonaco, 1969);(ii) Transfer experiments show that antigen-sensitivecells are more readily inactivated than antibody-producing cells (Moller and Zukoski, 1968);(iii) ALS immunosuppressive action is enhancedby adult thymectomy (Monaco, Wood, and Russell,1966);(iv) ALS treatment induces a selective depletionof thymus-dependent areas in lymph nodes andspleen (Taub and Lance, 1968);(v) Lastly, ALS is mostly active on cell-mediatedimmune responses, much more than in humoralantibody responses, which are mainly dependent onB-cells (Lance, 1970).The mechanism of inactivation of T-cells by ALS

is a matter of debate. It has been suggested (Lance,1970) that ALS could influence T-cells much morereadily than B-cells because T-cells are moreaccessible in the peripheral blood and are renewedmore slowly than B-cells. This is supported by thefact that ALS is found in high concentrations in theperipheral blood, whereas it is at a lower concentra-tion in the central lymphoid organs. Elimination ofT-ells might occur by cytotoxicity or, alternatively,by opsonization as has been demonstrated in vitroand in vivo.

Alternative explanations include antigen competi-tion, sterile activation and, especially, 'blind-folding': ALS could coat the lymphocytes and thusprevent them from recognizing the antigen. Data onrosette formation support this last hypothesis.

All ALS studied inhibit rosette formation (Bach,1970c; Bach, Dardenne, Dormont, and Antoine,1969; Bach and Antoine, 1968). Some inhibition isobtained without complement, but this is enhancedby the presence of complement. The percentage of



j.com/

Ann R

heum D


ber 1971. Dow

nloaded from

http://ard.bmj.com/


ALS-sensitive spleen cells is about 75 per cent.;inhibition appears after 60 min. The complement isonly needed for its first two components, in contrastwith cytotoxicity which needs all nine components ofcomplement (Bach, Gigli, Dardenne, and Dormont,1972). RFC from normal mice are much more sensi-tive to ALS than those from immunized mice.T-RFC from thymus and spleen are more sensitivethan spleen B-rosettes and bone marrow rosettes(Bach and Dardenne, 1971b).

In summary, ALS could act by eliminating someof the T-antigen sensitive lymphocytes by opsoniza-tion (Cl, C4, C2, C3), by cytotoxicity, and perhapsby inactivating the others by blind-folding, i.e. coat-ing as in rosette inhibition with Cl, C4. Such ascheme would be compatible with the fact that ALSretains all its action in CS-depleted mice (Barth andCarroll, 1970; Cinader, Jeejeebhoy, Koh, andRabbat, 1971).

(4) AZATHIOPRINE

This drug inhibits rosette formation after 90 min.incubation at 37°C. (Bach and Dardenne, 1971c;Bach, Dardenne, and Foumier, 1969). This inhibi-tion is total for high concentrations, but only 75 percent. RFC are inhibited at lower concentrations.Azathioprine-sensitive cells have proved to beT-cells, since RFC lose their sensitivity to azathio-prine after neonatal thymectomy (Bach and Dar-denne, 1971b). Rosette inhibition by azathioprine isreversible within 20 to 30 min. which is not com-patible with inhibition of receptor synthesis, butrather with modifications of the lymphocyte mem-brane, making the receptor no longer available forthe antigen (Bach and Dardenne, 1971c). As withALS, RFC from unimmunized animals are moresensitive to azathioprine than RFC from immunizedanimals. Bone marrow RFC are less sensitive toazathioprine than spleen or thymus RFC but canacquire identical sensitivity after incubation withpurified thymosin, a thymic hormone (Bach,Dardenne, Goldstein, and others, 1971). As RFCfound in normal non-immunized animals areantigen-recognizing cells (Bach, Muller, and Dar-denne, 1970), and as azathioprine-sensitive cells aretheta-positive T-cells (Bach and Dardenne, 1971b),one may suggest that azathioprine is immunosup-pressive in vivo by virtue of its action on antigen-sensitive T-cells. This hypothesis is corroborated bythe fact that azathioprine suppresses thymidineincorporation in mixed lymphocyte culture, which isa pure T-cell reaction, and that this inhibition isobtained only when azathioprine is put in the culture-before the 24th hour (Bach and Bach, 1972). Thisaction of azathioprine on antigen recognition byT-cells is in contrast with the antiproliferative action-suggested by Berenbaum (1967). However, the main

argument used by Berenbaum (the optimum anti-metabolite action when injected 1 to 2 days afterantigen stimulation) is not valid if one considers thatantimetabolite action is reversible, as shown in vivoand in vitro in rosette inhibition. If given before theantigen, or on day 0 the antimetabolite would nolonger be active when the antigen is still at theimmunogeneic level, whereas, given at day 2, theantimetabolite can suppress antigen-lymphocytecontact and prevent the response. We have shown(Bach and Fournier, unpublished) that azathioprineselectively alters the homing of T-cells: when lymphnodes are labelled with 51Cr and injected intra-venously after incubation with azathioprine intonormal mice, the cell portion which normallymigrates to lymph nodes goes to spleen and liver, asdo cells taken from neonatally thymectomized mice(Lance and Taub, 1969). The selective action ofazathioprine on T-cells would explain why this drugis mostly active on cell-mediated reactions and doesnot prevent antibody production. Such absence ofaction on humoral responses might be favourable inorgan transplantation, where blocking antibodiesmay have a very favourable action.

Clinical applicationsAntigen-binding techniques can be used in man.Rosette formation appears to be much simpler(Bach, 1970b) than autoradiography after antigenlabelling. The main applications of the rosettetechnique have been in rheumatoid arthritis (Bachand Delbarre, 1968; Bach, Delrieu, and Delbarre,1970), thyroiditis (Perrudet-Badoux and Frei, 1969),drug allergy (Cruchaud and Frei, 1967), glomerulo-nephritis (Mahieu, Dardenne, and Bach, 1971),pigeon breeder's disease (Bach, Bach, Foumier,Texier, Dardenne, Laborde, and Drouet, 1971), andanti-D immunization (Elson and Bradley, 1970).The significance of rosettes found in these conditionsis not yet established. The fact that the lymphocytescome from the peripheral blood suggests that RFCmight be antigen-sensitive cells rather than antibody-producing cells which do not circulate much.Another clinical application has been evaluation

of immunosuppressive potency of ALS (Bach,Dormont, Dardenne, and Balmer, 1969; Bach,1970b; Bach and Dormont, 1971) and of immuno-suppressive drugs (Bach, Dardenne, and Fournier,1969). The rosette inhibition technique proved to besensitive enough to detect serum metabolites ofantimetabolites in normal subjects (Bach, Dardenne,and Fournier, 1969) and in pathological conditions,such as renal failure (Bach and Dardenne, 1971d)and liver diseases (Bach and Dardenne, 1971e).Moreover, we have shown that abnormal sensitivityto azathioprine may be due to slow metabolicdegradation (Bach, 1971).



j.com/

Ann R

heum D


ber 1971. Dow

nloaded from

http://ard.bmj.com/


References

ADA, G. L.,AND BYRT, P. (1969) Nature (Lond.), 222, 1921 (Specific inactivation of antigen reactive cells with1261 labelled antigen).

ANDERSSON, B., AND BLOMGREN, H. (1970) Cell. Immunol., 1, 362 (Evidence for a small pool of immunocompetentcells in the mouse thymus. Its role in the humoral antibody response against sheep erythrocytes, bovineserum albumin, ovalbumin, and the nip determinant).

ARGYRIS, B. F. (1968) J. Immunol., 100, 1255 (Role of thymus in adoptive tolerance).BACH, C., BACH, J. F., FOURNIER, C., TEXIER, J. L., DARDENNE, M., LABORDE, M. A., AND DROUET, E. (1971)

Presse med., in press (La maladie des eleveurs d'oiseaux chez l'enfant. Mise en evidence d'une sensibilisationdes lymphocytes sans hypersensibilite retardee).

BACH, J. F. (1970a) Eur. J. clin. biol. Res., 15, 28, 258 (Les serums antilymphocytes. I and II).(1970b) Lancet, 1, 565 (Antigen-binding lymphocytes).(1970c) Fed. Proc., 29, 120 (In vitro assay for antilymphocyte serum).

- (1971) Transpl Proc., 3, 27 (Immunosuppression by chemical agents).- AND ANToiNE, B. (1968) Nature (Lond.), 217, 658 (In vitro detection of immunosuppressive activity of anti-

lymphocyte sera).-AND DARDENNE, M. (1971a) Cell. lmmunc., in press (Antigen recognition by T lymphocytes. I. Thymus andbone marrow dependence of spontaneous rosette-forming cells in normal and neonatally thymectomized mice).-- (1971b) Ibid., in press (Antigen recognition by T lymphocytes. II. Similar effects of Azathioprine,

antilymphocyte serum, and anti-theta serum on rosette-forming lymphocytes in normal and neonatallythymectomized mice).-- (1971c) Eur. J. clin. biol. Res., in press (Activities of immunosuppressive agents in vitro. I. Rosetteinhibition by Azathioprine).

(1971d) Transplantation, in press (The metabolism of azathioprine in renal failure).(1971e) Proc. roy. Soc. Med., in press (Serum immunosuppressive activity of azathioprine in normal

subjects and patients with liver diseases).AND DAVIES, A. J. S. (1971) Nature (Lond.), 231, 110 (Early effect of adult thymectomy).Do,RMoNT, J., AND ANToINE, B. (1969) Transpl. Proc., 1, 403 (A new in vitro test evaluating

antilymphocyte serum potency).-, AND FOURNIER, C. (1969) Nature (Lond.), 222, 998 (In vitro evaluation of immunosuppressive drugs).-, GOLDSTEIN, A., GuHA, A., AND WHIT, A. (1971) Proc. nat. Acad. Sci. (Wash.), in press (Appearance

of T-cell markers in bone marrow rosette-forming cells after incubation with purified thymosin, a thymichormone).

-- AND DELBARRE, F. (1968) C. R. Acad. Sci. Paris, Ser, D, 267, 134 (Nouvelle methode de detection du facteurrhumatoide au niveau cellulaire).DELRIEU, F., AND DELBARRE, F. (1970) Amer. J. Med., 49, 213 (The rheumatoid rosette. A diagnostic test

unifying seropositive and seronegative rheumatoid arthritis).-- AND DoRMoNT, J. (1971) Transplantation, 11, 96 (Further developments of the rosette inhibition test for the

testing of antihuman lymphocyte serum).-, , DARDENNE, M., AND BALNER, H. (1969) Ibid., 8, 265 (In vitro rosette inhibition by anti-human ALS.

Correlation with skin graft prolongation in subhuman primates).- GIGGu, I., DARDENNE, M., AND DORMONT, J. (1972) Immunology, in press (Mechanism of complement

action in rosette inhibition and lymphocyte-toxicity by antilymphocyte serum)., MULLER, J. Y., AND DARDENNE, M. (1970) Nature (Lond.), 227, 1251 (In vivo specific antigen recognition byrosette-forming cells).REYES, F., DARDENNE, M., FoURNE, C., AND MULLER, J. Y. (1971) 'Rosette formation a model for

antigen recognition', in 'Cell Interactions and Receptor Antibodies in Immune Responses', ed. 0. Makela,A. Cross, and T. U. Kosunen, p. 111. Academic Press, New York.

BACH, M. A., AND BACH, J. F. (1972) Clin. exp. Immunol., in press (Different timing of azathioprine andamethopterin in suppression and stimulation of mixed lymphocyte reaction).

BARTH, R. F., AND CARROLL, G. F. (1970) J. Immunol., 104, 522 (Immunosuppressive effects of antilymphocyteserum in complement-deficient mice: evidence that immune cytolysis is not essential for ALS activity).

BASTEN, A., MILLER, J. F. A. P., WARNER, N. L., AND PYE, J. (1971) Nature (Lond.), 231, 104 (Specific inactivationof thymus derived (T) and non-thymus derived (B) lymphocytes by 125I labelled antigen).

BERENBAUM, M. C. (1967) 'Immunosuppressive agents and the cellular kinetics of the immune response' in'Immunity, Cancer, and Chemotherapy', ed. E. Mihic, p. 217. Academic Press, New York.

BIozzI, G., STIFPEL, C., MOUTON, D., BOUTHILLIER, Y., AND DECREUSEFOND, C. (1968) Immunology, 14, 7 (Akinetic study of antibody producing cells in the spleen of mice i.v. injected with sheep erythrocytes).

BLOMGREN, H., AND ANDERSSON, B. (1969) Exp. cell. Res., 57, 185 (Evidence for a small pool of immunocompetentcells in the mouse).

BRODY, T. (1970) J. Immunol., 105, 126 (Identification of two cell populations required for mouse immunocompetence).BURNETT M. (1959) 'The Clonal Selecdion Theory of Acquired Immunity'. University Press, Cambridge.



j.com/

Ann R

heum D


ber 1971. Dow

nloaded from

http://ard.bmj.com/


BYRT, P., AND ADA, G. L. (1969) Immunology, 17, 503 (An in vitro reaction between labelled flagellin orhaemocyanin and lymphocyte-like cells from normal animals).

CiLLER, J. M., HABICHw, G. S., AND WEIGLE, W. 0. (1970) Proc. nat. Acad. Sci. Wash., 65, 551 (Cellular sites ofimmunologic unresponsiveness).

CINADER, B., JEEEBHOY, H. F., KOH, S. W., AND RABBAT, A. G. (1971) J. exp. Med., 133, 81 (Immunosuppressiveand graft-rejecting antibodies in heterologous anti-lymphocytic serum).

CLAMAN, H. N.. AND CHAPERON, E. A. (1969) Transpl. Rev., 1, 92 (Immunologic complementation between thymusand marrow cells. A model for the two-cells theory of immunocompetence).

COHEN, J. J., FISCHBACH, M., AND CLAMAN, H. N. (1970) J. Immunol., 105, 1146 (Hydrocortisone resistance ofgraft versus host activity in mouse thymus, spleen, and bone marrow).

CRuCHAUD, S., AND FREi, P. C. (1967) Int. Arch. Allerg., 31, 455 (Demonstration of specific antibodies on humancirculating lymphocytes by a new technique).

DAvImE A. J. S. (1969) Transpl. Rev., 1, 43 (The thymus and the cellular basis of immunity).DuBnsKi, S., AND FRADETrE, K. (1966) Proc. Soc. exp. Biol. (N. Y.), 122, 126 (The feed back mechanism in

immunoglobulin synthesis).DUTrON, R. W., AND MISHELL, R. I. (1967) J. exp. Med., 126, 443 (Cell populations and cell proliferation in the

in vitro response of normal mouse spleen to heterologous erythrocytes. Analysis by the hot pulse technique).ELSON, C. J., Am BRADLEY, J. (1970) Lancet, 1, 798 (Human peripheral blood leucocytes forming rosettes with

rhesus (o) isoantigen).GREAVFs, M. F. (1970) Transpi. Rev., 5, 45 (Biological effects of anti-immunoglobulins: evidence for

immunoglobulin receptors on 'T' and 'B' lymphocytes)., ToRluoINm, G.,Am Rorrr, I. M. (1969) Nature (Lond.), 222, 885 (Blocking of the lymphocyte receptor sitefor cell-mediated hyper-sensitivity and transplantation reactions by anti-light chain sera).

HUMPHREY, J. H., AND KELLER, M. V. (1971) 'Some evidence for specific interaction between immunologicallycompetent cells and antigens', in 'Symposium on Developmental Aspects of Antibody Formation, 1969,Prague'. Academic Press, New York.

ISAKOVId, K., SMrTH, S. B., AND WAKSMAN, B. H. (1965) Science, 148, 1333 (Immunologic tolerance inthymectomized, irradiated rats grafted with thymus from tolerant donors).

JERNE, N. K. (1971) Eur. J. Immunol., 1, 1 (The somatic generation of immune recognition).KENNEDY, J. C., TILL, J. E., SmIovrrcH, L., AND MCCULLOCH, E. A. (1966) J. Immunol., 96, 973 (The

proliferative capacity of antigen-sensitive precursors of hemolytic plaque-forming cells).LANCE, E. M. (1970) Clin. exp. Immunol., 6, 789 (The selective action of antilymphocyte serum on recirculating

lymphocytes: a review of the evidence and alternatives).- AND TAUB, R. N. (1969) Nature (Lond.), 221, 841 (Segregation of lymphocyte populations through differentialmigration).

LIACoPOULOs, P. (1971) Personal communication.MAHIEU, P., DARDENNE, M., AND BACH, J. F. (1971) Amer. J. Med., in press (Detection of humoral and cell-

mediated immunity to kidney basement membranes in human renal diseases).MAKELA, O., AND CRoss, A. M. (1970) Progr. Allerg., 14, 145 (The diversity and specialization of immunocytes).MAKINODAN, T., AND ALBRIGHT, J. F. A. (1963) 'Cytokinetics of antibody response', in 'Immunopathology: III

International Symposium, 1963', ed. P. Grabar and P. A. Miescher, p. 99. Schwabe, Basel.MARTIN, N. J., AND MILLER, J. F. A. P. (1968)J. exp. Med., 128, 865 (Cell to cell interaction in the immune

response. IV. Site of action of antilymphocyte globulin).MASON, S., AND WARNER, N. L. (1970)J. Immunol., 104, 762 (The immunoglobulin nature of the antigen

recognition site on cells mediating transplantation immunity and delayed hypersensitivity).MILLER, J. F. A. P., AND MrrCHELL, G. F. (1969) Transpl. Rev., 1, 3 (Thymus and antigen-reactive cells).MITCHISON, N. A. (1971) Eur. J. Immunol., 1, 10, 18 (The carrier effect in the secondary response to hapten

conjugates. I and II).MOLLER, G., AND ZUKOSKI, C. (1968)J. Immunol., 101, 325 (Differential effect of heterologous ALS on

antibody-producing cells and antigen-sensitive cells).MONACO, A. P., WOOD, M. L., AND RUSSELL, P. S. (1966) Ann. N.Y. Acad. Sci., 129, 190 (Studies on heterologous

ALS in mice.m. Immunologic tolerance and chimerism produced across the H-2 locus with adultthymectomy and ALS).

NISBET, N. W., AND SIMONSEN, M. (1967)J. exp. Med., 125, 967 (Primary immune response in grafted cells.Dissociation between the proliferation of activity and the proliferation of cells).

NossAL, G. J. V., CuNNINGHAM, A., M1TCHELL, G. F., AND MILLER, J. F. A. P. (1968) Ibid., 128, 839 (Cell to cellinteraction in the immune response.m. Chromosomal marker analysis of single antibody-forming cells inreconstituted, irradiated, or thymectomized mice).

NOTA, N. R., LIACOPOULOS-BRIOT, M., STIMFL, C., AND BIozzI, G. (1964) C. R. Acad. Sci. Paris, 259, 1277(L'immunocyto-adh6rence: une m6thode simple et quantitative pour l'etude in vitro des cellules productricesd'anticorps).

PERRUDET-BADOUX, A., AND FREi, P. C. (1969) Clin. exp. Immunol., 5, 117 (On the mechanism of 'rosette' formationin human and experimental thyroiditis).



j.com/

Ann R

heum D


ber 1971. Dow

nloaded from

http://ard.bmj.com/


PLAYFAIR, J. H. L. (1969) Nature (Lond.), 222, 882 (Specific tolerance to sheep erythrocyte in mouse bone marrowcells).

- , PAPERMASTER, B. W., AND CoLE, L. J. (1965) Science, 149, 998 (Focal antibody production by transferredspleen cells in irradiated mice).

PLOTZ, P. H. (1969) Nature (Lond.), 223, 1373 (Specific inhibition of an antibody response by affinity labelling).QUINT, J., HARDY, M. A., AND MONACO, A. P. (1969) Surg. Forum, 20, 252 (Modification of the immunosuppresive

effects of rabbit antimouse lymphocyte serum by a thymic humoral factor: thymosin).RAFF, M. C. (1969) Nature (Lond.), 224, 378 (Theta isoantigen as a marker of thymus-derived lymphocytes in mice).

AND WORTIS, H. H. (1970) Immunology, 18, 931 (Thymus dependence of 0-bearing cells in the peripherallymphoid tissues of mice).

SCHLESINGER, M., AND YRON, I. (1970) J. Immunol., 104, 798 (Serologic demonstration of a thymus-dependentpopulation of lymph-node cells).- (1969) Science, 164, 1412 (Antigenic changes in lymph-node cells after administration of antiserum tothymus cells).

SEGAL, S., GLOBERSON, A., FELDMAN, M., HAIMOVICH, J., AND GIVOL, D. (1969) Nature (Lond.), 223, 1374(Specific blocking in vitro of antibody synthesis by affinity labelling reagents).

SINGHAL, S. K., AND RICHTER, M. (1968) Int. Arch. Allerg., 33, 493 (The response of normal rabbit bone marrowcells to antigen in vitro).

TAUB, R. N., AND LANCE, E. M. (1968) J. exp. Med., 128, 1251 (Histopathological effects in mice of heterologousantilymphocyte serum).

TAYLOR, R. B. (1969) Transpi. Rev., 1, 114 (Cellular cooperation in the antibody response of mice to two serumalbumins. Specific function of thymus cells).

TRUFFA-1BACHI, P., AND WOFSY, L. (1970) Proc. nat. Acad. Sci. Wash., 66, 685 (Specific separation of cells inaffinity columns).

UHR, J. W., AND MOLLER, G. (1968) Advanc. Immunol., 8, 81 (Regulatory effect of antibody on the immuneresponse).

WARNER, N. L., BYRT, P., AND ADA, G. L. (1970) Nature (Lond.), 226, 942 (Blocking of the lymphocyte antigenreception site with anti-immunoglobulins sera in vitro).

WIGZELL, H. (1969) Antibiot. Chemother. (Basel), 15, 82 (Specific fractionation of immuno-competent cells).- (1970) Transpl. Rev., 5, 76 (Use of humoral or cell-bound antibody to achieve depression of immunological

reactivity).WILSON, D. B., KLYTH, J. L., AND NOWELL, P. C. (1968) J. exp. Med., 128, 1157 (Quantitative studies of the mixed

lymphocyte interaction in rats. M. Kinetics of the response).ZAALBERG, 0. B. (1964) Nature (Lond.), 202, 1231 (A simple method for detecting single antibody-forming cells).



j.com/

Ann R

heum D


ber 1971. Dow

nloaded from

http://ard.bmj.com/

antigen recognition by t-cells and its suppression · ann. rheum. dis. (1971), 30, 565 antigen...

Documents