0022-202X/85/850 ls-0 ll 5s$02.00/ 0 TliE J OURNAL OF INVESTIGATIVE DERMATOLOGY, 85:1 15s- 120s, 1985 Vol. 85, No. 1 Supplement Copyright © 1985 by The Williams & Wilkins Co. Printed in U. .A. Pathogenesis of Graft-Versus-Host Reactions (GVHR) and GVH-like Diseases

ERNST GLEICHMANN, M.D. AND HELGA GLEICHMANN, M.D. Division of Immunology, Medico /Institute of Environmental Hygiene and Diabetes Research Institute, University of Dii.sseldor{, Dusseldorf, F.R.G.

The graft-versus-host reaction (GVHR) in both mice causa ignota or after exposure to certain viruses or and humans can lead to the development of a broad drugs, are caused b y T lymphocytes reacting against spectrum of clinical and pathological symptoms. These self-MHC structures on lymphohemopoietic cells that symptoms are strikingly similar to those of a number of were rendered "foreign". By analogy to GVHD, it is diseases of proven or presumed immunolog ical origin, conceivable that the development of either stimulatory such as systemic lupus erythematosus (SLE), other col­ or suppressive GVH-like symptoms in individuals ex­ lagen vascular diseases, lymphoproliferative disease, posed to a given virus or sensitizing drug depends not and aplastic anemia. on the etiologic agent per se, but on whether the predom­ 8 The purpose of our investigation was to describe the inant r esponse is made by the individual's TH or T 1K immunological and pathological events that take place cells. This, in turn, might depend on whether the agent in the course of graft-versus-host disease (GVHD) and becomes immunogenic in combination with class-II or to gain insight into the cellular mechanisms underlying class-! alloantigens. these events. To this end, a model was employed in which nonirradiated F, mice were used as recipients of paren­ tal lymphoid cells. By pathological manifestations, 2 basic forms of GVHD can be distinguished in such non­ Lymphocytes are highly speciali zed in recognizing and re­ irradiated Ft recipients: One is acute GVHD which is sponding to foreign elements invading from the outside world often lethal. It is characterized b y a variety of suppres­ while no apparent, or at least no harmful , immune reactions sive (hypoplastic) pathological symptoms, including a develop towards self-constituents. Previously, it had been pro­ severe hypoplasia of the lymphohemopoietic system ac­ posed that t his state of self-tolerance is achieved by deletion of companied by aplastic anemia and hypogammaglobuli­ self-reactive clones during ontogeny a nd autoimmunity results nemia. The other basic form is characterized by stimu­ from the effect of "forbidden clones" [1] . From more recent latory symptoms, such as persistent lymphoid hyperpla­ experimental scrutiny, however, we know that potentially au­ sia, formation of autoantibodies, 'and development of toreactive B lymphocytes do exist in the intact, healthy immune pathological symptoms reminiscent of SLE and other system (2-8] but are normally controlled by mechanisms pre­ collagen vascular diseases. The suppressive pathological cluding a utoimmune reactivity. The designation autoimmune graft-versus-host (GVH) symptoms are caused by T sup­ disease is unambiguous where the functional relevance of auto­ 81 antibodies to unaltered autoantigens has been established, e.g., pressor/killer (T K) cells of the donor which react to­ antibodies reacting with t he acetylcholine receptor in myas­ wards allogeneic class-1-structures of the F 1 recipient's major histocompatibility complex (MHC). The stimula­ t henia gravis, with erythrocytes in autoimmune hemolytic ane­ tory pathological GVH symptoms, by contrast, are mia, with the insulin receptor in some patients with insulin­ caused by donor T helper (T'') cells which react toward resistant diabetes, with thyroglobulin in Hashimoto's thyroid­ the recipient's allogeneic class-II-MHC structures. itis, with basement membrane structures in Goodpasture's syn­ The possible implications of these observations for the drome, with the intrinsic factor in pernicious a nemia, or with pathogenesis of a number of GVH-like diseases in hu­ various a utoantige ns as in systemic lupus e rythematosus mans are discussed. The hypothesis is advanced that (SLE). Other di sease processes are also immunologically me­ some of these GVH-like conditions, which arise either e diated, alt hough they are not strictly autoimmune in t he sense that the immune reaction is directed to unaltered self-antigen. For instance, immune complexes that contain foreign material, Supported in part by a grant f~om Deutsc he Forsc hungsgemeinsc haft such as microbial antige ns, can cause deleterious lesions such (Hu 290/1), Bonn, Fede ral Republic of Ge rm any. as glomerulonephritis. Thus, in this type of immunological Rep rint requests to: Ern st Gleichmann , M.D., Division of Immunol­ disea e t he target antige ns need not consist of unaltered self­ ogy, Medical Institute of Environmental Hygiene, Un iversity of Dues­ antigens. Moreover, patients with aplastic anemia may have T seldorf, Auf'm Hennekamp 50, D-4000 Duesseldorf 1, Federal Republic lymphocytes that suppress t he growth of erythroid and myeloid of Germany. colonies in cultures [9,10] . Here, the eliciting antigen is not Abbreviations: ANA: antinuclea r antibodies' known at all. DPH: diphenylhyda ntoin To analyze the mechanisms of autoimmune diseases 2 ge n ­ dsDNA: doub le-stranded DNA era lly different types of experimental animal models have been GV H: graft-versus- host used: t hose in which the diseases occur spontaneou.sly and those GVHD: graft-versus-host di sease in which t he diseases are induced in recipients that usually GV HR: graft-ve rsus- host reaction(s) would not suffer from autoimmune attacks. The spontaneous ICGN: immune co mplex glome rulonep hritis lupu ~- lik e disease of New Zealand Black mice is a n example of MCH: major hi stocompatibi li ty complex the l!rst type. Vanous techmques have been applied for the MuLV: murine leukem ia virus e.xperi~ e ntal induction of aut?immuni PLN: popliteal lymph node ty: these include injec­ SLE: systemic lupus erythematosus tiOn of syngenetc or a ll ogenetc tissues mixed with Freund's T DTH ce ll : T cell ca usin g del ayed hyperse nsitivity reaction complete adjuvant into normal animals, manipulation of the Tg: thyroglobu lin lymphocyte population by thymectomy at a stage of immatu­ TH cell: T-helper cell ri ty, or modification of normal cells by chemicals and viruses. T51K ce ll: T-all osuppressor/ ki ll er ce ll The present paper fo cuses first on the well -defined cellula r

115s 116s GLEICHMANN AND GLEICHMANN Vol. 85, No. 1 Supplement pathogenesis of a s pectrum of immunologic diseases which can be induced b y T lymphocytes in ge neticall y normal mice undergoing systemic graft-versus-host reactions (GVHR). In t his model T lymphocytes are responding to allogeneic struc­ tures of the major hi stocompat ibility complex (MHC) (i.e., H - 2) . The second part of the paper deals with t he possible patho­ ge nic mechanism(s) of GVHR-like immunologic disorders which arise after exposure to a fore ign compound and might be due to graft-versus-host-(GVH) like reactions ofT lymphocytes towards self-structures rendered n on -self by the etiologic agent.

GRAFT-VERSUS-HOST REACTIONS (GVHR) B DONOR Experimental Design GVHR was induced by injection of parental strain lympho­ .l>.l!R ..... cytes in to adult, otherwise untreated, F1 hybrid recipients. For rH _..------TS/K genetic reasons, t he grafted cells are tolerated and not rejected Lyt n- Lyt ,.2. by the semiallogeneic F, host. The injected T lymphocytes, t however, recognize alloantigens whi ch the F1 host has inherited Lyt ,-z· ...... ______,. from the other parental strain. These a ll oreactive donor T ly mphocytes mediate the development of various immunologi­ soppress\of\ I cal lesions .' For t he induction of immunologic diseases by I GVHR 2 conditions are required: 1) all oreactive T lymphocytes It: rlo·- have to be present in the inoculum and 2) a histoincompatibility ,r:: .. has to be present in the F 1 recipient in order to activate the I'- donor T lymphocytes. In non irradiated F, mi ce, the histoincom­ I //., I patibility has to be provided by t he recipient's major histocom­ I patibility complex MHC, i. e., H-2 in t he mouse.

Spec;l;rum of GVHR-Induced Immunological Disorders F1 RECIPIENT Fig 1A shows the spectrum of pat hological symptoms which can be induced by GVHR. The majority of t hese alterations han.: been observed as sequelae ofGVHR in both human (11,12] ~ I o I ana laboratory a nimals (12- 14] . Two basic forms of GVHR­ in duced pathological lesions have been distinguished. One form ·--- is characterized by stimulatory pathological symptoms, as shown H-2=mijor histocompitibility complex !MHC) in t he middle and on the left hand side o f Fig lA. These symptoms include a persistent lymphoid hyperplasia of mainly FIG l. Spectrum and co ncept of ce llular pathogenesis of immuno· B lymphocytes resul ting in hypergammaglobulinemia and the logical disorders resulting from GVHR in F, mice. Not listed is toxic formation of autoantibodies characteristic of SLE, such as epidermal nec rolysis which is a se rious GVH complication in species antibodies against nuclear antige ns (ANA), double-stranded other than mouse, such as hamster and man (see [1 2] for references). A, Spectrum of pathological lesions DNA (dsONA), erythrocytes, and thymocytes [3,15- 20]. In that. may develop in GVH F, mice. B, Co ncept of the basic cellular and MHC requirements for induction addition, stimulatory GVH symptoms include a severe immune­ of stimulatory and middle (left) and suppressive (right) GVHR lesions, complex glomerulonephritis (ICG N) in which antibodies respectively. The F1 cell depicted at the bottom represents a stimulator/ again st ANA and envelope antige ns of murine leukemia virus target cell situated at various anatomical sites, such as lympho-hemo­ (MuLV) are involved (21]. This ICGN may be accompanied by poietic tissue, endothelium, synovia, skin, salivary gland, bile ducts, deposition of immunoglobulin along the basement membrane gut, and lungs. of the s kin [4]. The SLE-Iik e GVH symptoms may occur together with symptoms c haracteri stic of other collagen vas­ GVHD. One approach was to separate the donor T cells into cular diseases including arteritis (1 5,22,231, chronic polyarthri­ Lyt subsets before transferring them into F, recipients. In a t is [23 ], Sjogren- and scleroderma-like lesions, and li ve r second approach unseparated T cell s were injected into F 1 changes resembling sclerosing cholangitis (22,23]. Usually, t he recipients differing only at class I and/or II H-2 antigens stimulatory GVH symptoms present themselves as chronic [17,18]. ln addition, in vitro studies were performed to dissect GV H disease (G VHD). the cellular and immunogenetic requirements of allohelp and The other basic form of GVHR-induced lesions is character­ allosuppression [26]. Based on the results of these experiments ized by suppressive (hypoplastic) pathological symp toms (see a general concept of the cellular pathogenesis ofG VHR-induced right hand side of Fig lA) which clinica ll y manifest themselves diseases has been elaborated, as illustrated in Fig lB. as an acute GVHR. The s uppressive GV H symptoms consist Thus, when the donor's Lyt 1+2- T-helper (TH) cells [19] are of a seve re hypoplasia of the lympho-hemopoietic compartment activated by class-ll MHC structures of the F, recipient per­ leading to apl astic anemia and hypogammaglobulinemia [1 8- sistent lymphoid hyperplasia and SLE-like GVHD develop 20,24,25]. The microbiological status of the recipients bas an [1 7,18]. Whether t hese class-Il-reactive donor T cells a lso influence on whether or not thi s form of GVHR will terminate trigger the GVH-associated malignant lymphomas [14,27] and as lethal GV HD. symptoms of co ll agen vascular disease, such as chronic polyar­ thritis, Sjogren- and scleroderma-like disease, has not yet been Subsets of Alloreactive T Lymphocytes Induce Either established. Stimulato ry or S uppressive GVHR Syndromes The cellular in teractions leading to suppressive pathological Several experimental approaches have been taken to analyze GVH symptoms are more complex. Optimal induction of the if functionall y different s ubsets of all oreactive donor T ce ll s suppressive GVH symptoms required non-Lyt-separated donor are activated in the F1 recipients undergo ing different forms o[ T ce ll s [19] and incompatibility in the F, recipients at both July 1985 GV H DISEASE AND GVH-LIKE DISEASES 117s

A. normal T-B cell coope=r=a.::ti.:::;O:..:"------class I and class II structures [1 8]. In such parent - F 1 combinations t here is a sequential alloactivation first of donor TH cells and then T-allosuppressor/ killer (Ts1K) cells that cause t he suppressive pathological symptoms (Fig 1B) [18,20,24]. Thus, after a brief initial activation of Lyt 1 +2- donor TH cells, which r esults in a transient lymphoproliferation a nd B-cell stimulation in t he first week of GVHR, class !-specific Lyt 2+ donor T 51K cells are activated [20,25,28]. The detection of t hese

F 1 -reactive T s;K effector cells in GVHR, mice was strictly confined, both in time and in t he parent - F, combinations B. abnormal T-B cell cooperation (GVHR) being compared, to mice exhibiting the suppressive pathologica l symptoms [18,20,25]. This establi shed t hat F,-reactive donor T s;K cells are the effector cells causing t he hypoplastic GVH syndrome (Fig 1A ). absence of @ antigen Mechanism of Autoantibody Formation During GVHR absent or There is general agreement t hat potentially autoreactive inadequate signal G): B ·------® ······ no cross-linking of lymphocytes exist in healt hy individua ls, including geneticall y lg receptors normal F 1 mice (5]. Although IJOrmally silent t hese F, B cells are able to secrete IgG autoantibodies upon adequate activation. During GVHR such an activation i s delivered by t he donor's alloreactive TH cells. Host (F1)-derived T cells are not required 20. Tg for t he activation of t hose autoreactive B cells a nd, by contrast, counteract autoimmunization by t he GVHR [3,21] . The mech­ ·. i t1 .... ___ ·®···.... anism underlying t he GVHR-induced autoimmunization most likely is abnorma l T-B-cell cooperating, as illustrated in Fig 2B. During normal T-B-cell cooperation for antibody production 3 both TH and B cells react to antigenic sites on the same molecule (Fig 2A ). A stable carrier-hapten bridge is formed between the 2 specific cells. The haptenic part of t he immunogen is recog­ 0 adequate signal G) : ni zed by t he B cell and delivers signal 1, whereas the carrier ·------®····· cross-linking of part is recognized by t he TH cell. The TH cell secretes lg receptors helper factor(s) , here designated signal 2 [29 ]. Macrophages partici­ pate in t his activation process as antigen-presenting cells. 4 During abnormal T-B-cell cooperation (Fig 2B) F1 B cells receive signa l 1 by specific antigen, or hapten, and signal 2 (i .e., help), by a ll oreactive TH cells of t he parental donor strain 0/. [5,29,30]. The TH cells are Lyt 1 +2 - cells recognizing allogeneic ·------®··· ··· class-II MHC ant icrens [26]. Antibodies resulting from abnor­ mal T-B-cell coope~ation in vivo belong mainly to t he IgG class FIG 2. Hypothes is s howing essential differences between normal [30,31] a nd have high affinity [32 ]. and abnorma l T-8-cell cooperation in order to ex plain the selected The fi rst investigations of abnormal T-B-cell cooperation autoantibody fo rm ation seen in SLE- Iik e GVHD. All B ce lls depicted during GVHR were performed with non-self compounds as here are unprimed; hence, their ave rage receptor affinity is low and antigen such as sheep red blood cells, or haptens such as stable binding to the epitopes on the a ntige n is c ru cial. A, In norm al trinitrophenyl and dinit rophenyl, coupled to special carrier T-B-cell cooperation the 8 and the TH ce ll s react to antigenic sites on molecules. T hese studies demonstrated t hat a positive all oge­ the same mol ec ul e, thus a stable carri er-hapten bridge betwee n the two neic e ffect, i.e., a maximall y e nhanced lgG a ntibody response, specific cel ls is form ed. The hapte nic part of the immunoge n is recog­ induced by alloreactive TH cell s [30,31], only occurred if the nized by the 8 ce ll resultin g in signa l 1 to the 8 ce ll [29 ]; the ca rrier foreign antigen was present at t he onset of GVHR, thus pro­ part of the immunoge n is recogni zed by the TH ce ll s whi ch secrete viding signal 1 to the F B cells (16,31 ] (for rev1ew see (33]). helper factor( s), here designated signal 2 [29 ]. Mac rophages also pa r­ 1 ticipate in normal T-B cooperation, but are not depicted here. B, In Hence, in addition to signal 2 (a.llohelp), signal I (provided by ab normal T-8-cell cooperation the all oreactive TH cell reacts to a epitopes on the antigen) is required for abnormal T-B-cell co­ molecule, i.e. , all oantigen, whi ch is physically separate from the mole­ operation to become optimally effective. cule the B ce ll reacts to. This all oreaction prov ides help, or signal 2, to During GVHR, a ll oreactive donor TH cell s provide s igna l 2, all F, B ce ll s irrespective of their antibody specificity. Other than in or help, to t he F 1 B cell s. This type of help is indiscriminatory normal T-B-ce ll cooperation, t.h ere is no TH ce ll that foc uses the because a ll F 1 B cells are semiallogeneic and can thus receive epitopes to the 8 cel l. Hence the 8 ce ll has to interac t with the hapten all ohelp (Fig 2B). Nevertheless, t he resulting B-cell stimulation all by itself. For an effective signal 1, mul ti point binding betwee n the is a highl y selective one. The reason for t his is t hat t he mere epitopes on the hapten and the 8 ce ll is req uired. (1) If there is no presence of a self-antigen is not enough to provide a n adequate hapten, no signal I delive red. The B ce ll ca nnot be triggered into optimal proliferation and lgG production. (2) Antige n i s a globular signal 1 to the corresponding autoreactive B cell. Apparently, protein, such as autologous thyroglobu lin (Tg) whose epitopes ca n onl y on ly t hose kinds of self-antigen depicted in Fig 28.3, 2B.4 seem hind in a mo nova lent fa shion to the lg receptors of the 8 ce ll. Mon o­ to supply an effective signa l 1. By t his reasoning we t ry to va lent binding is in effective, because it is of low av id ity and doe ~ not explai n the observation t hat autoantibody formation in SLE­ cross-l ink the lg receptors. (3 and 4) Antige n p o~sesses repeatin g like GVHD fa ils to be a random polyclonal event [29,33]. identical epitopes located on either a ri gid backbone, such as DNA, or a ce ll su rface; in addition, epitopes may be electri ca lly charged. T hi s allows for mu ltipo in t hi gh-av idi ty-binding to and cross- linking of the lg receptor and thu s provides an adequate signal 1. B ce ll s are driven accounts for the phenomenon of selective lgG (auto)antibody form ation into clonal proliferation and max im al l gG sec retion. The difference found in SLE-like GVHD were autoantibodies to DNA and ce ll -surface betwee n 1 and 2 on the o ne hand and 3 and 4 o n the other hand epitopes, but not to globul ar protein antige ns, such as Tg, are formed. 118s GLEICH MANN AND GLEICHMANN Vol. 85, No. 1 Supplement

GVHR-LIKE IMMUNOLOGICAL DISORDERS TABLE 1. Selected reports on HLA associations with induced INDUCED BY MODIFIED SELF autoimmune phenomena Symptoms Induced by H LA associatio n Reference Concept of Pathogenesis Drug-induced SLE Hydralaz ine DR4 [41] The basic principles established for T cell alloreactivity, as Nephropathy D-penicillamine DR3 and B8 [51] operating in GVHR, might also operate in the ce llular patho­ Nephropathy Aurothiomalate DR3 and B8 [51) ge nesis of GVHR-like diseases developing after exposure to an Selective lgA defi - Diphenylhydantoin A2 [54 ) etiologic agent. GVHR-like cellular interactions may be stim­ ciency ulated in autologous or syngeneic systems provided cell -surface SLE-like autoanti- Spani sh toxic oil DR3 and DR4 [53 ) determinants of lymphohemopoietic ce lls and/or antigen-pre­ bodies senting cells were modified by a given etiologic agent. These Severe scleroderma- Vinyl -chl oride DR5 [52) "altered-self' structures might trigger reactions by autologous like lesions T cells comparable to the reactions of parental strain T cells toward the alloge neic structures on F , recipient cells. Thus, remains to be tested. It is unlikely that the majority of these autologous or synge neic TH cells may be preferent iall y activated ant ibodies are directed against the small haptenic co mpound if a foreign ant igen X is recognized in association with class II 81 DPH. This assumption is consistent with the finding that MHC determinants [30,34,35]. In contrast, T K cells may be DPH-specific antibodies were only rarely seen in patients sen­ preferentially activated if the same foreign antigen X is recog­ sitized to the drug [44,46 ]. A DPH-spec ific lymphoproliferation nized in association with class I MHC determinants [36]. Thus, in vitro was demonstrated in several patients sensitized to the depending on the T cell subpopulation activated, stimulatory drug [38,45). or suppressive pathological symptoms may develop, as shown The data on DPH is consistent with the co ncept that the in Fig lA. Certain drugs and viruses such as Epstein-Barr, parent molecule, or its metabolites, can render membrane stru c­ cytomegalovirus, and rubella virus, are suspected to create such tures of potentiall y stimulatory cells, such as dendritic cells non-self structures because they have been found to be associ­ and B lymphocytes, "non-self'. T lymphocytes would then ated with various GVHR-like diseases in man. recognize these altered membrane determinants in co njunction Drugs Known to Induce GVHR-Lihe Diseases wi th autologous MHC structures and respond in a GVHR-like manner, resulting in the development of the respective immu­ An array of drugs have been reported to cause a variety of nological lesions. pathological symptoms out of the broad spectrum of GVHR­ The proposed immunopathogeneses induced by an etiologic like condit ions shown in F ig l A. For example, the anticonvul­ agent may also be relevant for the mec hanisms of mercury sant drug diphenylhydantoin (DPH, phenytoin, dilantin) has chloride (HgC]z)-induced lesions in rats [47,48]. Autoimmune been reported to cause hypergammaglobulinemia, drug- induced glomerulonephritis, splenomegaly, antinuclear, and ant i- DNA SLE, art hritis, dermatitis, toxic epidermal necrolysis, B-cell ant ibodies were observed in HgCh-susceptible strains, but not lymphomas, aplastic anemia, and hypogammaglobulinemi a in strains carrying other MHC haplotypes. HgC12 -induced B­ [1 4,3 7,38]. Furthermore, lymphadenopathy, a feature of drug­ cell activation was observed, both in vivo and in vitro, only in induced SLE, or thrombocytopenia and granulocytopeni a de­ the susceptible strain. Interestingly, the B-cell activation in veloped during therapy of rheumatoid arthritis with D-penicil­ vitro wasT-cell-dependent [47 ]. lamine (39] or captopril , an antihypertensive drug- inhibiting No PLN enla rgement was found in several mouse strains angiotensin-I -coverting enzyme [40 ]. Hydralazine, another which had been t reated with hydralazine or methyldopa, drugs ant ihypertensive drug, also causes drug-induced SLE [41] . that are we ll known to induce SLE and autoimmune hemolytic anemia, respectively [49,50]. The pathogenic mechanisms by Studies on the Pathogenic Mechanism of Drug-Induced which these drugs induce the immunological injuries remain to GVHR-Lihe Symptoms be evaluated. · The popli teal lymph node (PLN) assay, an establi shed Interestingly, in human, a high association has been reported method fo r measuring local GVH and host-versus-graft (HVG) between certain genetic factors and several etiologic agents reactivity, was used to explore t he immunological effects of causing GVHR-like symptoms (Table I). Thus, here is a pre­ sensitizing drugs in vivo. It was found that subcutaneous injec­ ponderance of both HLA-DR4 positivity and slow acetylation tion of DPH, D-penicillamine, or captopril into the hind fo ot in patients with hydralazine-induced SLE [ 41). Moreover, sig­ pad of mice resulted in a significant dose-dependent enl arge­ nificant associations have been found between go ld- or D­ ment of t he draining PLN. T lymphocytes were required for penicillamine-induced glomerular nephropathy and HLA- DR3 triggering t his drug-induced enl arge ment. Co nge nitally a thymic and -B8 [51], between vinyl-chloride-induced severe sclero­ nu/ nu recipients, i. e., T ce ll-deprived mice, fai led to react to derma-like lesions and DRS [52], and between DR3 and DR4 the inoculation of t hese drugs, whereas their+ / nu counterparts and SLE-like autoantibodies after ingestion of the Spanish readily did so [42] and unpublished results). Al though T cell s toxic oil [53]. DPH-induced lgA-deficiency is associated with were required for the DPH-induced PLN enla rge ment the HLA-2 (54] . majority of t he proliferating cells in the PLN were not T cells [42,43]. These observations exclude that t he PLN reactions were ca used by a direct mitogenic effect of DPH on B lympho­ CONCLUSIONS cytes. Furthermore, DPH fail ed to exert nonspecific mitogenic The genetic and cellular requirements for induction by or comitogenic effects on lymphocytes in vivo and in vi tro [43- GVHR of broad spectrum immunological disorders on the one 45 ] and unpublished resul ts). Subcutaneous injection of phe­ hand and the GVHR-like features induced by DPH [49,50] and nobarbital, a drug t hat is pharmacologicall y and structurally HgC lz [54,55] on the other hand are considered relevant for the related to DPH, but does not cause GVHR-like side effects, pathogenesis that may underly some of the human GVHR-Iike failed to induce PLN enlargement [43). diseases associated with certain HLA alleles. A given etiologic Additional studies revealed that subcutaneous inoculation of agent is assumed to induce an antigenic site X on cell-surface DPH induced a more t han 100-fold increase of IgG -secreting determinants ("altered-self'). In a susceptible individual au­ cells in the draining PLN as compared to the uninjected co n­ tologous T lymphocytes might recognize and respond to these tralateral PLN [43]. Syngeneic spl ee n cell s which had been antigenic sites akin to the way parental strain T lymphocytse pretreated with DPH in vit ro also induced significant PLN recognize and respond to alloge neic MHC structures ofF, celL enlargement. The specifi city of t he DPH-induced antibodies in the GVHR model [33]. Thus, TH cells are preferentially July 1985 GVH DISEASE AND GVH-LIKE DISEASES 119s

activated if the antigenic s ite X is m a inly associated or seen 16. van Rappard-van der Veen FM, Kiesel U Poels L Schuler W Melief CJM, Landegent J, G l eic hm a n~ E: Further evidenc~ · togethe r wit h class-II HLA structures. In contrast, TS/ K cell s agamst random polyc lonal antibody formation in mice with are prefere nt ia lly activated if t he antigenic site X is m a inly lupus-like graft-versus host disease. J Immunol 132:1814- 1820 associated or seen togethe r with class-I HLA structures. Cor­ 1984 , respondingly, immunological disorders cons isting of stimula­ 17. van Rappard-van der Veen FM, Rolink AG, Gleichmann E: Dis­ eases caused by reactions ofT lymphocytes towards incompatible tory or suppressive GVHR-like symptoms would e nsue as illus­ structures of t he major histocompatibility complex . VI. Auto­ t rated in Fig 1. Certainly, a numbe r of other immunolouical antibodies characteristic of systemic lupus erythematosus in­ a nd nonimmunological factors suc h as ly mphokines, sex h o r­ duced by abnormal T-8-cell cooperation across I-E. J Exp Med mones, a nd m etaboli c pathways of a given etiologic agen t, a lso 155:1555- 1560, 1982 con tribute to t h e complex processes eventua lly result ino- in 18. Rolink AG, Pals ST, Gleichmann E: Allosuppressor- and all ohel ­ per-T ce ll s in acute and chronic graft-versus- host disease. II. F1 immunopathological lesio ns. Therefore, in most cases theb in­ recipients carrying mutations at H-2K and/or I-A. J Exp Med duction of disease by a n etiologic agen t depends on t he presence 157:755- 771, 1983 of suscep t ibility a ll eles at a number of differen t gen etic loci. 19. Rol ink AG, Gleichmann E: Allosuppressor- and a ll ohelper-T cell s Assessment of t he pathogenic pathways by which ide n tifi ed in acute and chronic graft-versus-host di seases. III. Different Lyt subsets of donor T cell s induce di ffe rent pathological syn­ etiologic agents induce immunological diseases s hould increase dromes. J Exp Med 158:546- 558, 1983 our knowledge of t he etiopathology of the same diseases devel­ 20. Pals ST, Radaszkiewiczt T, Gleichmann E: Allosuppressor- and oping s p onta n eous ly . These investigatio ns m ay help to ela bo­ allohelper-T ce ll s in acute and chronic graft-versus-host disease. rate m ethods to even t u a lly prevent a nd/or cure those diseases. IV. Activation of donor allosuppressor ce ll s is con fined to acute GVHD. J Immunol 132: 1669- 1678, 1984 21. Rolink AG, Gleichmann H, Gleichmann E: Diseases caused by We are obli ged to our co ll eaf,rues Drs. A. G. Rolink, E. H. van Elven, reactions of T lymphocytes to incompatible structures of t he S. T . Pals, and T. Radaszkiewicz whose in tell ectual contributions and major hi stocompatibility complex. VII. Immune-complex glo­ dedicated work made t hi s publication possible. merulonephritis. J Immunol 130:209- 215, 1983 22. van Rappard-van der Veen FM, Radaszkiewicz T, Terraneo L, Gleichmann E: Attempts at standardization of lupus-like graft­ REFERENCES versus- host disease: inadvertent repopulation by DBA/2 spleen ce lls of H-2-different non-irradiated F, mice. J Immunol 1. Burnet FM: The Clonal Selection Theory of Aquired Immunity. 130:2693-2701, 1983 Nashville, Tennessee, Vanderbilt Uni ve rsity Press, 1959 23. Pals ST, Radaszkiewicz T, Roozendaal L, Gleichmann E: Chronic 2. Fialkow PJ, Gdchnst C, Al li son AC: Autoimmunity in chroni c progressive polyarthritis and other symptoms of coll agen vascu­ graft-versus-hostdisease. Clin Exp lmmunol 13:479- 486, 1973 lar di sease induced by graft-versus-host reaction. J lmmunol, in 3. Gleichmann E, Gle1chmann H: Diseases caused by reactions ofT­ press lymphocytes to incompatible structures of the major hi stocom­ 24. Rolink AG, Radaszkiewicz T, Pals ST, van der Meer W, Gleich­ patibility co mplex: I. Autoimmune hemolytic anemia. Eur J mann E: Allosuppressor- and allohelper-T cell s in acute and Immunol 6:899- 906, 1976 chronic graft-versus-host disease. I. Alloreactive suppressor cell s 4. van Elven EH, Agterberg J , Sadal S, Gleichmann E: Diseases rather t han ki ll er T ce ll s appea r to be t he decisive effector cell s caused by reactions ofT lymphocytes to in co mpatible structure in lethal graft-versus-host disease. J Exp Med 155:1501- 1522, of the major hi stocompatibility complex : II. Autoantibodies de­ 1982 posited along t he basement membrane of skin and t heir rela tion­ 25. Pals ST, Gleichmann H, Gleichmann E: Allosuppressor- and allo­ ship to immunecomplex glomerulonephritis. J Immunol helper-T cells in acute and chroni c graft-versus-host disease. 126: 1684- 169 1, 1981 V. F, mice with secondary chronic GVHD co ntain F,-reactive 5. Allison AC, Denma~ AM , Barnes RD: Hypothesis. Cooperating all ohelper but no all osuppressor T cell s. J Exp Med 159:508- and controlli ng functiOns of t hymus-den ved lymphocytes in 523, 1984 relation to autoimmunity. Lancet 2:135- 140 1971 26. Rolink AG, van der Meer WGK, Melief CJM, Gleichmann E: 6. Rose NR, Bacon LD , ? undick RS, Kong YM, 'Esquivel P, Bigazzi Jn tra-H-2 requirements for t he induction of maximal positive PE: Genetic regulatiOn 111 autoimmune t hyroiditis, Autoimmun­ and negative a lloge neic effects in vitro. Eur J Immunol 13:191- ity Genetic, Immunological and Clinical Aspects. Edited by N 197, 1983 Tala!. New York, Academic Press, 1977, pp 63- 87 · 27. Pals ST, Zij lstra M, Radaszkiewicz T, Cuypers T , Schoenmakers 7. Fournie GJ, Lambert PH, Miescher PA: Release of DNA in ci rcu ­ HJ, Me lief CJM, Berns A, Gleich mann E: Immunological induc­ lating blood and induction of anti-DNA antibodies after injection tion of malignant lymphoma: graft-versus-host reaction-in­ ofbacteriallipopolysaccl1arides. J Exp Med 140:1189- 1206, 1974 duced-B-cell lymphomas contain reintegrations of several types 8. Fernandez C, Hammerstrom L, Moller G, Primi D Smith CJE: of murine leukemia virus sequences. Submitted for publication, Immunological tolerance affects onl y a subpopu'lation of t he 1984 antige n-specifi c B lym phocytes: evidence against clonal delection 28. Hurtenbac h U, Shearer GM: Analysis of murine T lymphocyte as t he mechani sm of tolerance induction. Immunol Rev 43:3- 41 markers during t he early phases of GVH-associated suppression 1979 , of cytotoxic T-lymphocyte response. J lmmunol 130:1561- 1566, 9. Bacigalupo A, Podesta M, van Lint MT, Vimercati R Ce rri R 1983 Rossi E, Ri sso M, Ca rella A, Santim G, Damasio E,'Giordan~ 29. Bretscher P: Hypothesis: a model for generalized autoimmunity. D, Marmont AA: Severe aplastic anaemia: correlation of in vitro Ce ll Immunol 6: 1- ll, 1973 tests wi th clinical response to immunosuppression in 20 patients! 30. Katz DH: Lymphocyte Differentiation, Recognition, and Regula­ Br J Haematol 47:423- 432, 1981 tion. New York , Academi c Press, 1977, pp 410- 482 10. Torok-Storb BJ, Sieff C, Storb R, Adamson J , Thomas ED: In 31. Ordal JC, Grumet FC: Ge netic control of the immune response. vitro tests for di stinguishing possible immune- mediated aplastic The effect of graft-versus-host reaction on the antibody response anemia from transfuswn-mduced sensitization. Blood 55:211 - to poly-L(Tyr, Glu)-poly-D, L-A la-poly-L-Lys in nonresponder 215, 1980 11. Hansen JA, Woodruff JM, Good RA : The graft-versus-host reac­ mice. J Exp Med 136: 1195- 1206, 1972 tion in man. Genetics clinical features and immunopathology 32. Elfenbein GJ, Green I, Paul WE: The allogeneic effect: increased Comprehensive Immunology 7. Edited by B Safai, RA Good: affinity of serum antibody produced during a seco ndary response. New York, Plenum Press, 1981, pp 229- 257 Eur J Jmmunol 3:640- 644, 1973 12 . Bril H, Benner R: Ce ll -mediated immunity to hi stocompatibility 33. Gleichmann E, Pals ST, Rolink AG, Radaszkiewicz T , Gleichmann antigens. CRC Crit Rev Clin Lab Sci, in press H: Graft-versus- host reactiOns: clues to t he etiopathology of a spectrum of immunological di seases. Immunology Today, in 13. Beschorn~r WE,Tutschka PJ, Santos GW: Chronic graft-versus­ host di sease 111 the rat radiatiOn chimera. Transplantation press 33:393- 399, 1982 34. Kindred B, Shref!ler DC: H-2 dependence of cooperation between 14. Gleichmann E, Gleichmann H, Wilke W: Autoimmunization and T and B cell s in vivo. J Immunol 109:940- 943, 1972 lymphomagenes is in !?a rent---> F, combinations differing at t he 35 . Benacerraf B, Germai n RN: The immune response genes of the maJor hi stocompatibility complex: model for spontaneous di sease maJor histocompatibilty complex. Immunol Res 38:70- 119 1978 caused by altered self-antigens? Transplant Rev 31:156- 224 36. Zinkernagal RM, Doherty_ PC: H-2 compatibility r eq uir e m e~t for 1976 , T- ce ll - m e d 1 at~d lysiS of target cells infected with lymphocytic 15. Gleichmann E, van Elven EH, van der Veen JPW: A systemic chon omemng1t1s vi rus. Different cytotoxic T-cell specificities lupus erythematosus (SLE)-like di sease in mice induced by ab­ are associated With structures coded for in H -2K or H-2D. J Exp normal T-B-ce ll cooperatiOn. Preferential formation of auto­ Med 141:1 427- 1436, 1975 antibodies characteristic of SLE. Eur J Immunol 12:1 52- 158 37. Shelby HJ, Rothman SJ , Buckley RH: Phenytoin hypersensitivity. 1982 , J Allergy Clm Immunol 66:166-172, 1980 120s GLEICHMANN AND GLEICHMANN Vol. 85, N o. 1 Supplement

38. R osent hal CJ , Noguera CA, Coppola A, Kapelner SN: Pseudolym­ J Pediatr 92:915- 920, 1978 phoma wit h m ycosis fun go ides manifestations, hyperresponsive­ 47. Druet P , Sapin C, Druet E, Hirsc h F : Genetic cont rol of mercury­ ness to diphenylhydantoin, and lymphocyte di sregul ation. Can­ induced immune response in t he rat, Nephrotoxic Mechanisms cer 49:2305-2314, 1982 of Drugs and Environmental Tox ins. Edited by GA Porter. New 39. Jaffe lA: Penicill ami ne in rheumatoid art hrit is: clinical pharma­ Yo rk/ London, Plenum Medical Book Company, 1983, pp 425- cology and biochemi cal properties. Scand J Rheumatol [Suppl] 435 28:58- 64, 1979 48. Weening J , Hoedemaker PJ, Bakker WW: lmmunoregulation and 40. Editori al: Captopril: benefi ts and risks in severe hypertension. anti-nuclear antibodi es in mercury- induced glomerulopathy i n Lancet 2:129-130, 1980 the rat. 45:64- 71, 1981 41. Batchelor JR, Welsh KI, Mansill a-Tinoco R, Dollery CT , Hughes 49. Perry HM: Possible mechanisms of t he hydralazine-related lupus­ GRV, Bernstein R, Ryan P , Maish PF, Aber GM, Bing RF, like syndrome. Art hrit is Rheum 24:1093- 1104, 1981 Russel G I: Hydralazine- induced systemic lupus e rythematosus: 50. Worlledge SM: Immune drug-induced h emolytic a nemias. Semin influence of HLA-DR and sex o n suscept ibility. Lancet 1:1107- Hematol 10:327-344 , 1973 1109, 1980 51. Wooley PH, Gri ffen J , Panayi GS, Batchelor JR, Welsh Kl, Gibson 42. Gleichmann H : Studies on the mechani sm of drug sensit ization: T TH: HLA- DR ant ige ns and toxic reaction to sodium aurothio­ cell -dependent popli teal lymph node reaction to diphenylhydan­ malate a nd D-penicillamine in patients wi th rheumatoid a rthri­ toin . Clin lmmunol Immunopathol 18:203-211, 1981 tis. N Eng! J M ed 303:300-302, 1980 43. Gleichmann H, Pals ST, Radaszki ewicz T : T cell-dependent B-cell 52. Black CM, Walker A E, Catoggio LJ: Genetic s usceptibility to proli fe ration and activation induced b y a dministration of t he scleroderma- like syndrome induced b y v inyl chloride. Lancet drug diphenylhydantoin in mice. Hematol Onco logy 1:165- 176, 1:53-55, 1. 983 1983 53. Vi cario JL, Serrano-Rios M, San Andres F, Arnaiz- Villena A : 44. Holland P , Mauer A: Drug-induced in vit ro stimulation of periph­ HLA-DR3, DR4 increase in chronic stage of Spanish oil di sease. eral lymphocytes. Lancet 1:1368- 1369, 1965 Lancet 1:276, 1982 45. Schopf E, Schultz K H, Kessler R, Taugner M , Braun W: All ergo­ 54. Shakir RA, Behan PO, Dick H, Lambie DG. Metaboli sm of im­ logische Untersuchunge n beim Lyell -Syndrom. Z Haut kr 50:865- munoglobulin A, lymphocyte fun ction, a nd histocompatibility 873, 1975 antige ns in patients on ant iconvulsants. J Neurol Neurosurg 46. H yman LR, Ballow M , Knieser MR: Diphenylhydantoin intersti­ Psychiatry 41:307-311 , 1978 t ial nephrit is: roles of cellular and humoral immunologic injury.