key: cord-0010010-2zrn6jqg
authors: Goverman, Joan
title: Tolerance and autoimmunity in TCR transgenic mice specific for myelin basic protein
date: 2006-04-28
journal: Immunol Rev
DOI: 10.1111/j.1600-065x.1999.tb01313.x
sha: ad75762334f808b3563f632920a31540738e0fca
doc_id: 10010
cord_uid: 2zrn6jqg

Summary: T‐cell receptor (TCR) transgenic mice provide the ability to follow the maturation and fate of T cells specific for self‐antigens in vivo. This technology represents a major breakthrough in the study of autoimmune diseases in which specific antigens have been implicated. Proteins expressed within the central nervous system are believed to be important autoantigens in multiple sclerosis, TCR transgenic models specific for myelin basic protein (MBP) allowed us to assess the role of tolerance in providing protection from T cells with this specificity Our studies demonstrate that T cells specific for the immunodominant epitope of MBP do not undergo tolerance in vivo and that TCR transgenic mice are susceptible to spontaneous autoimmune disease. The susceptibility to spontaneous disease is dependent on exposure to microbial antigens, MBP TCR transgenic models expressing TCRs specific for the same epitope of MBP but utilizing different V(α) genes exhibit differing susceptibilities to, spontaneous disease. These data support the idea that genetic and environmental differences play a role in susceptibility to autoimmunity MBP TCR transgenic models are playing an important role in defining mechanisms by which infectious agents trigger autoimmune disease as well as defining mechanisms by which tolerance is induced to distinct epitopes within self‐antigens.

Summary: T-cell receptor (TCR) tran,sgemc mice provide the ability to follow the maturation and fate ofT cells specific for self-antigens in vivo. This technology represents a major breakthrough in the study of autoimmune diseases in which specific antigens have been implicated. Proteins expressed within the central nervous system are beheved to be important autoantigens in multiple sclerosis, TCR transgenic models specific for myelin basic protein (MBP) allowed us to assess the role of tolerance in providing protection from T cells widi this specificity Our studies demonstrate that T cells specific for the immunodominant epitope of MBP do not undergo tolerance in vivo and that TCR transgenic mice are susceptible to spontaneous autoimmune disease. The susceptibihty to spontaneous disease is dependent on exposure to microbial antigens, MBP TCR transgenic models expressing TCRs specific for the same epiCope of MBP but utilizing different V,, genes exhibit differing susceptibilities to ,spontaneous disea,se. These data support the idea that genetic and environmental differences play a role in susceptibihty to autoimmunity MBP TCR transgenic models are playing an important role in defining mechanisms by which infectious agents trigger autoimmune disease as well as defining mechanisms by which tolerance is induced to distinct epitopes within self-antigens.

Multiple sclerosis (MS) is a neurological disorder whose underlying cause is unknown. It is a primary demyelinaLing disease characterized by infiltrates of mononuclear cells that are predominantly found in the white matter (1), Patients with MS manifest symptoms of neurological deficit in an acute, chronic relapsing/remitting or chronic progressive course of disease. The disease typically occurs hetween the ages of 20-45 and females have a relative risk of 2:1 compared to males (2), An autoimmune pathogenesis is suspected for MS for several reasons (3), First, celltilar infiltrates in iesions within the central nervous system (CNS) consist primarily of lymphocytes and monocytes. Second, genetic mapping studies have shown that immune response genes are linked to disease susceptibility Third, an animal model of demyelinating disease elicited by stimulating autoimmune responses has heen estabhshed that exhihits many similarities to MS. These observations led to a proposed model for MS in which self-reactivc T lymphocytes are activated and cross the blood braiti barrier into the CNS where they attack cells expressing self-antigens.

Both genetic and environmental factors influence susceptibihty to MS, The irnportance of hoth of these factors to disease is reflected in the 25-35% concordance rate for MS in monozygotic twins (4, 5), Variations in prevalence rates of MS thrcDughout the world support a role of environmental factors in influencing MS. There are almost no cases of MS reported in populations living near the Equator and the prevalence significantly increases in populations living in moderate or colder climates (6) .

Epidemiological studies suggest that an infectious agent triesers MS (7) . Some studies indicate that an individual's relative risk of developing MS is correlated with the environment in which that individual lived prior to the age of twelve and is not associated with the environment inhabited later in life (6), This suggests that exposure to the pathogen occurs much earlier in life than when the disease is clinically manifested. Numerous viruses have heen imphcated in the etiology of MS and relapses of MS have heen associated in some cases with viral infections (3, [8] [9] [10] [11] . Bacterial infections have been less commonly associated with MS, possibly because it is more difficult to obtain evidence for antecedent hacterial infection in patients. While no specific infectious agent has yet heen consistently litiked to MS, an infectious etiology remains an attractive hypothesis.

Several theories have heen proposed to explain how infection cotdd initiate MS (12) . First, an infection in the CNS will produce inflammatory cytokines and chemokines. These factors will cause non-specific recruitment of lymphocytes to the site of infection. Some of the lymphocytes could be specific for CNS antigens and undergo activation in situ, initiating an autoreactive response in the target organ. Second, opening of the blood brain barrier due to inflammation may expose the immune system to normally sequestered CNS antigens. Under these circumstances, naive T lymphocytes that have not undergone tolerance may be activated hy self-antigens in the CNS. Third, molecular mimicry of CNS antigens by microbial antigens may result in activation of cross-reactive T lymphocytes (i3, 14) . In support of the molecular mimicry hypothesis, some T-cell clones and lines established from MS patients demonstrated cross-reactivity for hoth CNS antigens and peptides derived from various human viruses (15, 16) . In addition, a peptide from the bacterium Pseudomonas aeruginosa was also able to activate a T-cei! clone specific for MBP (15) .

Experimental allergic encephalomyelitis (EAE) is an animal model of autoimmune disease with many similarities to MS (3), EAE is an inflammatory, demyehnating disease that causes acute, chronic or chronic-relapsing paralysis (17) . It is characterized hy perivascular inflammatory lesions in the white matter of the CNS (18), demonstrating a similar pathology to MS. EAE can be induced in many species either by immunization with CNS antigens in complete Freund's adjuvant or by adoptive transfer of activated CNS-specific T cells into naive recipients (19, 20) , Adoptive transfer experiments demonstrated that MHC class Il-restricted CD4+ T cells secreting inflammatory cytokines (Thl helper cells) are the primary agents mediating EAE (21), Like MS, susceptibihty to EAE is correlated with expression of particular MHC class II genes. Among mice, the strains that have been most commonly employed in studies of EAE are SJL mice that express H-2^ MHC alleles and Bl O.PL or PL/J mice that express H-2" MHC alleles. These strains demonstrate some interesting differences. Bl O.PL mice respond primarily to myehn basic protein (MBP). The T-cell response to MBP in this strain is very restricted. There appears to be a single immunodominant epitope consisting of the amino-terminal peptide MBPl-11 (22) .Thereisalsolitdeheterogeneity in the primary structures of the T-cell receptors (TCRs) that recognize this epitope. The majority of MBP 1-11 -speciflc T cells employ the Vp8.2 gene segment paired with either the V<,2.3 or 4.2 gene segment (23, 24) . Subdominant responses to MBP31-50 and MBP121-140 have also been reported, representing a minor fraction of the T-cell proliferative response to MBP in these mice (25, 26) .

In SJL mice, proteolipid protein (PLP) is the predominant autoantigen although T-cell responses to MBP are also generated. The immunodominant epitope of PLP is PLP139-i50 and the immunodominant MBP epitope is MBP89-100 (27, 28) . SJL mice more commonly exhibit a relapsing/remitting disease than BIO,PL mice (29), which may reflect the spreading of T-cell responses from PLP to MBP epitopes (30) .

EAE has been a very useful model to investigate the effector cells contributing to CNS autoimmune disease. Studies of EAE demonstrated that autoreactive. CNS-specific T cells are a normal part of the T-cell repertoire in healthy animals. These cells only become pathogenic following exposure to an exogenous stimulus. Differentiation of CNS-specific T cells into Thl cells that .secrete inflammatory cytokhit!> is an important requirement for disease induction. Activation nf these T cells results in expression of adhesion markers that facihtate their interaction with endothelial cells comprising the blood brain barrier and promote entry into the CNS (3 1-34), Entry of Thl T cells into the CNS and secretion of pro-inflammatory cytokines such as interleukin (IL)-] 2, interferon (IEN)-Yand tumor necrosis factor (TNE)-a elicit the production of chemokines that cause a further influx of monocytes and non-speciflc T cells (35) (36) (37) (38) (39) (40) . The toxic effects of TNE-a and nitric oxide produced by T cells and macrophages play an important role in damaging the myelin sheath (41) (42) (43) (44) (45) , Recovery mechanisms from this inflammatory response are not well understood. They appear to involve in part the production of anti-inflammatory cytokines such as IL-10. IL-4 and transforming growth factor (TGE)-p (46) (47) (48) (49) . Production of these cytokines may be the result of generating Th2 T cells which have been shown in some cases (50) (51) (52) hut not in others (53, 54) to suppress EAE. Apoptosis of CNS antigen-specific T cells that infiltrated the tissue has also heen imphcated in recovery from EAE (55) (56) (57) (58) . EAE is a model system to test therapies for MS In addition to defining many of the events that are critical to the effector stage of disease, EAE has been a valuahle model for developing therapeutic strategies. Approaches to inhihit EAE have targeted different steps in the pathogenesis of the disease. Several strategies have used either antihodies directed against pro-inflammatory cytokines, soluhle cytokine receptors or inhibitors of specific cytokine activity to prevent EAE (59) (60) (61) (62) (63) (64) (65) , Direct administration of anti-inflammatory cytokines has also been used to regulate EAE (66) (67) (68) (69) (70) . Encephahtogenic T cells have been targeted in a number of ways. Antibodies against V genes expressed on encephalitogenic T cells have been used to eliminate T cells mediating disease (71) , Vaccination of animals with inactivated encephahtogenic T cells (72) or specific peptides derived from V genes expressed on these cells (73, 74) has also been effective in inhibiting EAE in some systems. Other strategies have utilized knowledge of the antigen specificity of T cells to inhibit disease. Several studies have shown that exposing encephalitogenic T cells to altered peptides, variants of the wild-type peptide sequence normally recognized hy the T cell, can change the pattern of cytokine secretion from inflammatory to anti-inflammatory cytokines. This approach has been used effectively to inhibit EAE (75) (76) (77) (78) . Exposing animals to CNS antigens administered by different routes, such as intravenous, oral or intranasal, also prevents induction of EAE (79) (80) (81) . Several of these studies have led to the design of therapies and clinical trials that may be useful in the treatment of MS.

Studies tlSing EAE have contributed enormously to our understanding of CNS autoimmune disease. A great deal has been learned about the cell types and effector mechanisms involved in demyelination. This knowledge has laid the foundation for developing therapies to prevent and/or regulate the autoimmune responses. The methods used to induce EAE are artificial, however, and therefore preclude investigation of potential triggers of CNS autoimmune disease that may be relevant to MS. The development of TCR transgenic models of EAE described below offers new potential to investigate these factors.

The first TCR transgenic model of EAE utihzed genes encoding a TCR specific for MBP 1-11 presented hy I-A" isolated from a BIO.PL mouse (82) . Two transgenic lines were generated, one expressing the transgenic TCR a-chain (V^2.3 paired with J,,l i) and one expressing the transgenic TCR p-chain (V(,8,2 paired with Jp2.6) that together encode the MBPl-11-speciflc TCR. When these two transgenic lines were bred together, MBP 1-11 TCR transgenic mice were generated that allowed us to follow the fate of MBP 1-11 -specific T cells in vivo. We were able to study maturation of these cells in the thymus as well as their ability to survive in the periphery and mediate disease.

It was already clear that some MBPl -11 -specific T cells are normally present in the peripheral repertoire in BIO.PL mice because these cells are activated by immunization with MBP and are the predominant effector cells in EAE in this strain of mouse. It was not understood, however, why these autoreactive T cells were present in the periphery One possibihty was that these T cells are not subject to the mechanisms of tolerance induction that normally eliminate self-reactive T cells. MBP was beheved to be sequestered behind the blood brain barrier and therefore not presented to T cells under conditions that would induce tolerance. Alternatively, exposure to MBP may induce T-cell tolerance, but the MBPl-11-specific T cells may represent a specific subset ofT cells that escape tolerance.

Analyses of the thymocytes in MBPl-l 1 TCR transgenic mice demonstrated that MBP 1-11 -speciflc T cells do not undergo donal deletion in the diymus. Strong skewing in the thymocyte populations toward mature CD4+ thymocytes expressing the transgenic TCR indicated that these cells undergo very efficient positive rather than negative selection in the thymus (82) . Indeed, the selection of a TCR with this particular YjYp combination is so efficient on the B1 O.PL hackground that mice expressing only the transgenic p-chain generate a high precursor frequency ofT cells paired with endogenous a-chains en-Governian -TCR transgenic modcl5 of EAE coded by the same V^2,3-Jnl 1 gene segments that are expressed in [he parental MBPl-11-specific TCR (83) . The frequency ofT cells expressing the MBP-specific TCR in the TCR P-chain mice is sufflciently high to allow bttlk lymph node cells from unimmunized TCR P-chain mice to prohferate when cultured with MBP 1 -11 peptide (83, 84) . In contrast, T cells from TCR a-chain mice do not prohferate when stimulated with MBPl-11 peptide (T, Brabb, J. Goverman, unpublished observations) ,

The strong positive selection of transgenic tbymocytes on the Bl O,PL background indicates that these self-reactive T cells are not suhject to central tolerance induction. One reason that MBPl-11-specific T ceils may escape negative selection is that the MBPl-11 epitope displays a very low affinity for its I-A" MHC ligand and therefore may be a poor mediator of negative selection (85, 86) , Observations from a different MBPl-11specific TCR transgenic model are consistent with the notion that these cells would not undergo negative selection in the thymus even if they were exposed to their cognate antigen hecause of the instahility of the peptide/MHC complex (87), Thymocytes in these MBPl-il TCR transgenic mice did not undergo clonal deletion when the wild-type MBP 1-11 was administered intraperitoneally to the mice hut did exhihit deletion when a variant of MBPl-11 with a inuch higher affmity for I-A^ (MBPAcl-1 i[4Y]) was administered. Similar experiments carried out in different MBP 1-11 TCR transgenic mice resulted in some deletion of transgenic thymocytes following administration of the native MBPl-11 peptide; however, higher doses of peptide were used in these experiments (88),

In addition to detnonstrating that endogenous expression of MBP does not result in negative selection of MBP 1-11 -specific thymocytes. our analyses of thymocyte maturation in the MBPl-11 TCR transgenic mice revealed an unexpected observation related to the strong positive selection of the thymocytes with this antigen specificity The MBP 1-11 TCR transgenic mice exhibit a disrupted thymic architecture in which the stromal elements do not organize into a central medulla. Instead, small medullary foci are dispersed throughout the thymus surrounded by regions of cortical epithelium (89), Bone marrow-derived cells, dendritic cells and macrophages are still preferentially associated with the medullary foci. Analyses of other TCR transgenic mice revealed that disruption of the thymic architecture is increased in models that exhibit strong positive and negative selection compared to models that exhihit weak positive selection. Thns, the organizadon of medullary foci into a central medulla is impeded w-hen the normal balance of signals between thymocytes and stromal cells is skewed by high-avidity interactions (89) , Despite the lack of organization of the medullary foci into a central medulla, thymocyte maturation and export to the periphery occurs normally

The MBPl-11-specific T cells exported from the thymus do not undergo tolerance induction in the periphery (82) . T-cell numbers in the periphery are comparahle to wild-type mice and consist primarily of CD4-T cells expressing the transgenic TCR, These T cells are not anergic as they respond vigorously to stimulation with antigen in vitro and can mediate EAE in vivo. Thus, the peripheral T-cell repertoire in the transgenic animals is dominated by fully functional MBPl-1 1-speciflc T cells.

In recent studies, we demonstrated that expression of MBP in vivo does induce T-cell tolerance and that the lack of tolerance to MBPl-11 is epitope specific (90) , We compared the immune response to MBP in both H-2" wild-type mice and H-2" MBPdeficient shiverer mice (MBP-'-) tbat do not syntbesize intact MBP Wild-type mice primed with intact MBP responded only to the dominant MBPl-11 epitope and intact MBP in vitro. We did not detect significant responses to MBP3i-50 or MBPl 21-1 40 that had been reported by other investigators as subdominant MBP epitopes (2 5, 91), In contrast to wild-type mice, immunized MBP-'-mice generated very strong T-cell responses to two distinct epitopes witliin the MBP121-150 region of the protein. The response to MBPl-11 was very weak in comparison to the MBP121-150 epitopes. These data demonstrate that the true immunodominant epitopes of MBP are within MBP121-150 and that endogenous expression of MBP induces tolerance in T cells responding to this region (90) . These observations also exclude the hypothesis that MBPl-11-specific T cells do not undergo tolerance hecause MBP is a sequestered antigen that is invisible to the immune system. Tolerance to MBP has also been demonstrated by comparing T-cell responses to MBP in C3H MBP-' and wild-type mice (92) , Further characterization of the MBPi21-150-speciflc T cells found in MBP-'-mice revealed that this response is unexpectedly complex (93) . Two distinct, non-overlapping epitopes represented by MBPl25-135 and MBP 136-146 are present in this region that form very stable complexes with I-A" MHC molecules in vitro. The half-times of dissociation of these peptides are 270 and 180 h respectively Analyses of the residues that function as TCR contacts in these epitopes demonstrated tbat the antigenic surfaces of these tw-o peptide/MHC complexes lack any structural similarity Thus, it was very surprising that most of the T cells that respond to this region of MBP are cross-reactive for both epitopes. Even more intriguing was the observation that recognition by the cross-reactive T cells was lost when functional TCR contacts were interchanged between the two epitopes. Thus, the T cells appear to adopt mutually exclusive conformations to achieve specific re-Immumiiugirai RCTTCVVS 169/1999 Goverman • TCR trdnssenic models of EAE cognition of iwo distinct epitopes that present very different antigenic surfaces (93) . Interestingly, the few T cells speciflc for this region of MBP that escape tolerance in wild-type mice recognize peptides that contain the first hut not the second epitope (90), Very few T cells that recognize the epitopes within MBP121-150 escape tolerance in wild-type mice. To understand how tolerance to the imtnunogenic epitopes of this selfantigen occurs, we turned again to the powerful system of antigen-specific TCR transgenic mice. Using two sets of genes that encode TCRs cross-reactive for both MBP 125-135 and MBP136-146, we generated several lines of TCR transgenic mice. Our prehminary data indicate that the transgenic T cells may undergo some tolerance during maturation in the thymus but a number of transgenic T cells are exported to the periphery (E, Husehy, J. Goverman, unpuhlished ohservations). In prehminary experiments, the transgenic T cells on a heterozygous H-2'"''^"'"'' background retain the ability to respond to one epitope within MBP i 21 -1 5 0 but not to the MBP 131-150 peptide that contains the second epitope. This response is similar to the reactivity of the few remaining MBP121-140-specific T cells that escape tolerance in wild-type mice (90), These models will be very valuable in defining the mechanisms of tolerance induced by endogenous expression of MBP in vivo.

Dissecting CNS autoimmune disease in TCR transgenic mice Our studies of the MBP 1-11 TCR transgenic mice showed that these mice are highly susceptible to induction of EAE using the standard protocol of immunization with MBP peptide and injection of pertussis toxin (82) . This result w-as expected given the high precursor frequency of functional MBPl-11-specific T cells in the periphery. Immunization of the TCR p-chain transgenic mice with MBP1-! I peptide also induced EAE, consistent with the high precursor frequency of MBPl-i l-specific T cells in these mice (84) . Surprisingly, administration of pertussis toxin alone to the MBPl -11 TCR transgenic mice without immunization with MBP also resulted in a high incidence of EAE (82), The mechanism by which pertussis toxin induces EAE in the transgenic mice is not known. One reported function of pertussis toxin is to promote access of lymphocytes to the CNS by increasing permeability of the blood brain harrier (94, 95) . The ability of pertussis toxin alone to trigger EAE in the transgenic mice suggested that entry into the CNS might be the critical checkpoint in the iniliation of autoimmune disease. We tested this idea by determining whether injection of transgenic MBPl-11-speciflc T cells directly into the cerebral spinal fluid (CSF) of non-transgenic recipient mice was sufflcient to induce EAE, Injection of transgenic splenocytes that had been activated in vitro by exposure to MBPl-11 induced a high incidence of EAE (96) , Surprisingly, injection of non-stimnlated transgenic splenocytes as w-ell as non-stimulated purified T cells also induced EAE although at a lower frequency These results suggested that the presence of high numbers of CD4-M BP-specific T cells behind the blood brain barrier where MBP is expressed might he sufficient to trigger disease (96) , These experiments also demonstrated that EAE induced hy injection of pertussis toxin is manifested differently than EAE induced by intrathecal injection of MBP 1 -11-specific transgenic T cells into the CSF. EAE induced hy injection of pertussis toxin in MBPl-11 TCR transgenic mice was significantly more severe and demonstrated fewer relapses than EAE induced by intralhecal injection of activated or resting transgenic T cells (96) , A difference in the course of disease induced by these two methods w-ould not be expected if the only effect of pertussis toxin was to facilitate access ofT cells lo the CNS, Thus, exposure to pertussis toxin appears to influence EAE in multiple w-ays.

The lack of T-cell tolerance that was observed in the MBP 1-11 TCR transgenic mice in vivo suggested that these mice have the potential to develop spontaneous autoimmune disease. Spontaneous EAE was ohserved in some transgenic mice as early as one hackcross to BIO,PL mice when the mice were heterozygote for H-2" MHC class II molecules (82) , We continued to observe cases of spontaneous EAE as the mice were further backcrossed onto the BIO,PL background, A few cases of spontaneous EAE have also been observed in TCR p-chain mice, consistent with the idea that these mice are enriched for TCRs that are speciflc for MBPl-11 compared to non-transgenic mice (T, Brahb. J, Goverman, unpublished observations). The incidence of spontaneous EAE in the p-chain transgenic mice is very low but has not been precisely determined hecause we have not monitored them consistently for symptoms of EAE,

The window of age in which MBPl-11 TCR transgenic mice are suscepdble to spontaneous EAE is similar to MS in that the disease is manifested primarily during adolescence and early adulthood (96) , Fij]. 1 shows the age of mice at the onset of symptoms in a colony of 228 conventionally housed transgenic mice in which 107 developed EAE sp(Dntaneously. No cases of EAE were seen at less than S weeks of age and few cases were ohserved in mice older than 12 weeks of age. Unlike MS, however, the MBP I -11 TCR transgenic males have a higher relative risk (1,8/1) than females (96) , While females have a higher relative risk for MS than males, gender differences in both the human disease and the transgenic mouse model sup- (82) were housed in microisoiators in either an SPF or conventional animal facility with the same water, food and bedding, MBPl-11 TCR transgenic mice expressing the V^4,2/V|i8,2 TCR (97) were housed in the same conventional facility as the Y^l.i/V^H.l TCR transgenic mice. All mice were monitored for symptoms of EAE from weaning until I 2 weeks of age. Forthe VB2.3/Vp8,2 TCR transgenic mice, six cases among 40 mice were observed in the SPF facility and ten cases among 24 mice in the conventional facility. For the Va4,2/Vp8,2 TCR transgenic mice, five cases among 45 mice were observed in the conventional facility. port the idea that hormonal changes play a role in susceptibility to CNS autoimmune disease.

One of the most interesting observations made with this TCR transgenic model is that spontaneous EAE was seen only in mice that were housed under conventional conditions and was not ohserved when the TCR transgenic mice were housed under specific-pathogen-free conditions (SPF) (82) . The ohservation that the incidence of spontaneous EAE depended on environmental conditions under which the mice were housed represents a potentially important parallel with MS, Therefore, we sought to determine if spontaneous EAE was triggered by differences in microbial exposure or some other environmental factor. In our initial studies, several environmental differences in addition to microbial flora, such as food, water and caging, existed between the two colonies. To investigate specifically the role of microbes in triggering EAE in the transgenic mice, we established two cohorts of genetically equivalent MBPl-11 TCR transgenic mice that had been backcrossed onto BIO,PL for ten generations and had either conventional or limited microbial flora. Mice in both cohorts were housed in microisolators with identical food, water and bedding and were monitored over a 12-week period for symptoms of EAE. The incidence of spontaneous EAE shown in Fig. 2 was significantly higher in the conventional mice (43%) compared to the SPF mice (15%, p=0,017), indicating that exposure to microbes facilitates the induction of EAE (96) . This observation is consistent with the hypothesis that MS is triggered by exposure to pathogens.

During our studies of the incidence of spontaneous EAE in the MBPl-11 TCR transgenic mice that we had established, we introduced MBP 1-11 TCR transgenic mice into our conventional colony that had been developed independently and utilized a different TCR specific for MBP 1 -1 1 (97). These mice express a transgenic TCR specific for MBP 1-11 utilizing Vo4.2 and V,,8.2 obtained from PL/J mice (98) . We backcrossed these MBPl-11 TCR transgenic mice 11 generations onto the BIO.PL background and compared spontaneous disease in a cohort of these conventionally housed mice to our conventionally housed MBP 1-11 TCR transgenic mice that had been backcrossed for ten generations onto BIO.PL. Surprisingly, the data in Fig. 2 show that the incidence of spontaneous EAE was significandy lower (11,1% versus 41.78%, p = 0.003) in the MBPl-l 1 TCR transgenic mice expressing the Vj,4,2/V(i8,2 transgenic TCR than the MBPl-11 TCR transgenic mice expressing the Vo2,3/Va8,2 TCR. Because these mice had both been backcrossed extensively onto BIO.PL and were housed in the same 152 Immunological Reviews 169/1999 Immunoregulation in MBPl-11 TCR transgenic mice We compared thecharacteristicsof EAE that occurs in MBPl-11 TCR transgenic mice either spontaneously or following administration of pertussis toxin, as well as EAE that is induced by intrathecal injection of transgenic T cells into non-transgenic recipients (96) . In both spontaneous EAE and EAE induced hy pertussis toxin, the maximum severity of symptoms is fairly high, with over half of the cases reaching grade three severity (Fig. 3A) , The disease is usually chronic, with improvement observed in some cases, but complete recovery seen less often. Relapses are infrequently observed in either spontaneous or pertussis toxin-induced EAE (Fig, 3B) . Manifestation of EAE induced by intrathecal injection, however, is strikingly different. The incidence of disease is high, but the severity of EAE induced by this method never exceeded grade one. In addition, relapses were observed in almost half of the cases. Thus, EAE that is tnediated hy MBP-specific transgenic T cells is manifested very differently depending on how these T cells are triggered and/or how they enter the CNS, These observations suggest that the different methods of disease induction using transgenic T cells stimulate different mechanisms of immune regulation. Interestingly, the CNS lesions in mice with intrathecally induced EAE were much more severe than lesions seen in mice with either spontaneous or pertussis toxin-induced EAE despite the very mild clinical symptoms (96) . On a scale of 0 to 5, the average severity of cellular infiltrates was 4,3 in intrathecally-injected mice and i in mice with spontaneous EAE, A lack of correlation between severity of clinical symptoms and pathology in EAE has been noted in other studies (99, 100), One of the most intriguing ohservations regarding spontaneous EAE in the TCR transgenic mice is that, despite a very large population of functional MBP-specific T cells in the periphery, disease incidence does not reach 100% even in mice that are housed together in a conventional facihty (96) , This suggests that peripheral tolerance mechanisms function in these mice to prevent the occurrence of autoimmune disease and that these mechanisms fail with some stochastic frequency in a percentage of the mice.

The idea that immunoregulatory cells exist in vivo that function to prevent EAE was proposed in studies of a separate line of MBP 1-11 TCR transgenic mice bred onto the Rag-''-background (97) . A low incidence of spontaneous EAE was observed in these transgenic mice under SPF conditions but the incidence increased to 100% when the transgenic TCR was crossed onto the Rag-'-background. Recently, two studies of this MBPl-11 TCR transgenic model have implicated CD4-*-T cells expressing endogenous TCRs as the cells responsible for providing protection against the development of spontaneous EAE in MBP TCR transgenic mice on the Rag+' ' + background (101, 102) , Crosses of the transgenic TCR to mice deficient in Govorman • TCR transgenic models of EAE B ceils, CDS' T cells, NK-T cells or yb T cells did not increase the incidence of spontaneous EAE. How-ever, crossing the TCR otito a Rag-'-, Q,"'" or C^-''" hackground significantly increased spontaneous EAE. Adoptive transfer of CD4-T cells from nontransgenic mice into MBPl-11 TCR transgenic mice on the Rag-''-hackground decreased the incidence of spontaneous EAE. These adoptive transfer experiments also indicated that TCR Cransgenic Rag-''-mice had to receive the CD4-^ T cells from non-transgenic mice at an early age (40 days old or younger) to he fully protected frcmi spontaneous EAE (101, 1 02). The mechanism hy which these regulatory T cells protect MBPl-1 1 TCR transgenic mice from EAE is not known.

A different form of immunoregulation of MBP-specific T cells has recendy been demonstrated using the TCR P-chain Cransgenic mice. In these studies, it was found that the TCR pchain mice produce a population ofT ceils that are specific for determinants expressed on the transgenic p-chain itself (103). can traffic to the CNS and whether MBP-specific T cells that have not been previonsly activated can be stimulated by encountering endogenous MBP in this tissue. Iti analyzing T lymphocytes isolated from the CNS of healthy MBP TCR transgenic and non-transgenic mice, we first made the sttrprising ohservacion that comparable numbers of T cells are present in this tissue in transgenic and non-transgenic mice. This finding was unexpected because only activated T cells are believed to traffic to the CNS, and there are significantly fewer activated T cells in the periphery of TCR transgenic mice that do not have EAE compared to non-transgenic mice. TCR transgenic mice have fewer activated T cells in the periphery because the TCR repertoire is very restricted and therefore the chance that environmental antigens can activate T cells is reduced. Also unexpected was the observation that, while the majority of T cells in the CNS in non-transgenic mice exhihit an activated phetiotype, most of the T cells isolated from the CNS of young MBP TCR transgenic mice exhibited a naive phenotype (T, Brabb, J. Goverman, manuscript in preparation). We have now analyzed several other TCR transgenic models that were not specific for CNS antigens. We found that the number ofT cells in the CNS of these TCR transgenic mice was also comparahle to the number in non-transgenic mice and that their phenotype was predominantly naive. These data suggest that there is a steady state numher of T cells chat traffic through the CNS in a healthy animal. While activated T cells appear to have a preferential advantage in trafficking to this tissue, naive T cells can traffic to the CNS in the ahsence of "competition" from activated T cells in the periphery. One of the most interesting ohservations from these studies is that increased numbers of memory MBP-specific T cells are found in the CNS of older MBP TCR transgenic mice that do not exhihit any signs of clinical disease. Thus, an increase in the numher of activated and memory T cells in the CNS of MBPl-11 TCR transgenic mice does not appear to be sufficient for the induction of HAE.

Studies of MBP TCR transgenic mice have heen invaluable in answering questions about the induction of tolerance to CNS antigens, as well as the events that lead to the hreakdown in tolerance and the induction of autoimmune disease. Studies with these models are continuing as the answers to many questions typically lead to new areas of inquiry. New transgenic models specific for epitopes of MBP that undergo tolerance in vivo as well as f <Dr MHC class I-restricted epitopes of MBP will also be very informative. These models, as well as TCR transgenic mice specific for different CNS antigens, will be valuable tools in the future to understand the pathogenesis of CNS autoimmune disease and to test new therapies for the treatment of MS,

The neuropatliology of multiple sclerosis

The increased suscepcibiLity of women to multiple sclerosis

Immunology of multiple sclerosis and experimental allergic encepbalomyelitis

A population-based study of multiple sclerosis in twins: update

Genetic influences in multiple sclerosis

Multiple sclerosis: clinical and pathogenic basis

Kurtzke JF Fpidemiologic evidence for multiple sclerosis as an infection

Viral infections trigger multiple sclerosis relapses: a prospective seroepidemiological study

Cbnical viral infections and multiple sclerosis

Treatment of multiple sclerosis witli y interferon: exacerbations associated with activation of tbe immiane system

Tbe proposed viral etiology of multiple sclerosis and related demylinatbig diseases

Immunological aspects of demyelinating diseases

Amino acid bomology between tbe encephalitogenic site of myebn basic protein and virus: mechanism for autoimmunity

Sequence homology betw^een certain viral proteins and proteins related to enc epbalomyelitis and neuritis

Molecular mimicry in T ceO-mediated autoimmunity: viral peptides activate buman T cell clones specific for myelin basic protein

Myelin basic protein and buman coronavirus 229E cross-reactive T cells in multiple sclerosis

Experimental autoimmune (allergic) enceplialomyebtis: induction, patbogenesis, and suppression

Tbe lesion in multiple sclerosis and cbronic relapsing experimental allergic encephalomyelitis: a structural camparison

Adoptive transfer of experbuental allergic encephalomyelitis in SJL/J mice after in vivo activation of lympb node ceUs by myelin basic protein: requirement for Lyt-1-2"' T lympbocyies

Demyelinating diseases of tbe ceniral and peripheral nervous systems

Tbe T lymphocyte in experimental allergic encephalomyebtis

T-cell epitope of the autoantigen myelin basic protein that induces encepbalomyelitis

Restricted use ofT cell receptor V genes in murine autoimmune encephalomyelitis raises possibilities for antibody therapy

Limited heterogeneity of T cell receptors from lympliocytes mediating autoiinmune encepbalomyelitis allows specific immune intervention

Multiple discrete encephalitogenic epitopes of tbe aucoantigen myelin basic protein include a determinant for I-E class Il-restricted T cells

Spreading of T-celi autoimmunity to cryptic determinants of an auto-antlgen

Identification of an encephalitogenic determinant of myelin proteolipid protein for SJL mice

Cbaracterization of a major encepbalitogenic T cell epitope in SJL/J mice with synthetic oligopeptides of myelin basic protein

Brown AM, McFarlin DF. Relapsing experimental allergic encephalomyebtis in tbe SJL/J mouse

Functional evidence for epitope spreading in tbe relapsing pathology of experimental autoimnnnie encephalomyelitis

Homing to central nervous system vasculature by antigen specific lymphocytes. II. Lymphocyte/endotbelial cell adhesion during the initial stages of autoimmune demyelination

Adhesion-related molecules in the central nervous system. Upregulation correlates witb inflammatory cell influx during relapsing experimental autoimmune encephalomyelitis

Evidence for involvement of ICAM-i and VCAM-1 in lympbocyte interaction wiili ciidodiehum in experimental autoimmune encepbalomyelitis in tbe central nervous sysieni m the SJL/J mouse

Expression of vascular addressins and [CAM-1 by endotbelial cells in tlie spinal cord during chronic relapsing experimental allergic encephaloniyelitis in the Biozz! AB/H mouse

Interaciion of T lymphocytes with cerebral endothelial cells in vitro

Inflammatory leukocytes and cytokines in the peptideinduced disease of experimciital allergic encepbalomyehtis in SJL and Bl O.PL mice

Cytokine production by cells in cerebrospinal fluid during experimental allergic encephalomyelitis in SJL/J micu

An imponant role for the chemokine macrophage inflammatory protein-1 a hi the pathogenesis of the T cellmediated autoimmune disease, experimental autoiinmune encepbalomyeliiis

Cytokine production in the central nervous system of Lewis rats witb experimental autoimmune encephalomyelitis: dynamics of mRNA expression for interleukin-10. interleukin-12, cytolysin, tumor necrosis factor a and tumor necrosis factor p

Cytokines in relapsing experimental auloimmune encephalomyelitis in DA rats: persistent mRNA expression of proin flam mat ory cytokines and absent expression of interleukin-10 and transforming growth factor

TNF-a expression by resident microglia and infiltrating leukocytes in the central nervous system of mice witb experimental allergic encepbalomyelitis

Staphylococcal enterotoxin B and rumornecrosis factor-a-induced relapses of experimental allergic encephalomyebtis: protection by transforming growib factor-P and interleukin-10

Lympbotoxin and tumor necrosis factor-cc production by myelin basic protein-speciiic T cell clones correlates witb encephalitogenicity

Tumor necrosis factor mediates myelin and oligodendrocyte damage in vitro

Differential tumor necrosis factor a. expression by astrocytes from experimental allergic encepbalomyelitis-susceptible and -resistant rat strains

Analysis of cytokine mRNA expression in the central nervous system of mice witb experimental autoimmune encephalomyelitis reveals tbat IL-10 mRNA expression correlates with recovery

Inflammatory cytokines in the brain: does the CNS shape bnmune response? Inimunol Today

Oral tolerance to myehn basic protein and natural recovery from experimental autoimmune encephalomyehtis are associated with downregulation of inflammatory cytokines and differential upregulation of transforming growth factor p, interleukin 4, and prostaglandin E expression in the brain

Fncephalitogenic Th 1 cells arc inhibited b)' Tb2 cells with related peptide specificity: relative roles of interleukin (IL)-4 and IL-10

Regulatory T cell clones induced by oral tolerance and suppression of autoimmune encephalomyelitis

B7-1 andB7-2 costimulatory molecules activate differentially the Thl/Tli2 developmental pathways: application to autoimmune disease therapy

Neuroantigen-specific ThZ cells are inefficieni suppressors of experimental autoimnnine encephalomyelitis induced by effector Thl ceEs

Myelin basic protein-specific T helper 2 (Th2) ceEs cause experimental autoimmune encephalomyehtis in immunodeficient hosts rather than protect them from the disease

Apoptosis in brain-specific autoimmune disease

Apoptotic ebmination of Vp 8,2+ cells from the central nervous system during recovery from experimental autoimmune encephalomyelitis induced by tbe passive transfer of W^ 8.2" encepbalitogenic T cells

Apoptosis of ap T lymphocytes in tbe nervous system in experimental autoimmune encephalomyelitis: its possible implications for recovery and acquired tolerance

Antigen presentation by astrocytes primes rat T lymphocytes for apoptotic cell death. A model for T-cell apoptosis in nvo

An antibody to lympbotoxin and tiunor necrosis factor prevents transfer of experimental allergic encepbalomyelitis

Prevention of experimental autoimmune encephalomyelitis by antibodies against interleukin 1 2

Prevention of chronic relapsing experimental autoimmune encepbalomyehtis by soluble tumor necrosis factor receptor 1

Tbe antidepressant rolipram suppresses cytokine production and prevents auto immune encephalomyelitis

Prevention of autoimmune demyelination in non-human primates by a cAMP-specific phosphodiesterase inhibitor

Anti-tumor necrosis factor therapy abrogates autoitnmune demyelination

AunNeurol 1991:30

Therapeutic potential of phosphodiesterase t}'pe 4 inhibition in chronic autoimmune demyelinating disease

Cytokine-induced immune deviation as a therapy for inflammatory autoimmune disease

Interleukin-10 prevents experimental aEergic encephalomyehtis in rats

Prevention and treatment of cbronic relapsing experimental allergic encephalomyehtis by transforming growth factor-P 1

Macrophage-inactivating IL-1 3 suppresses experimental autoimintuie encephalomyebtis in rats

Retinoid treatment of experimental allergic encepbalomyelitis, IL-4 production correlates with improved disease course

Prevention and treatment of murine experimental allergic encephalomyelitis with T ceE receptor V p-specific antibodies

Vaccination against autoimmune encephalomyelitis with T-l)'mphocy te line ceEs reactive against inyelin basic protein

Brostoff SW Vaccination against experimental allergic encephalomyelitis with T eell receptor peptides

Immunization with a synthetic T-cell receptor V-region peptide protects against experimental autoimmune encephalomyelitis

Treatment of experimental encephalomyelitis with a peptide analogue of myehn basic protein

An altered peptide bgand mediaies immime deviation and prevents autoimmune encephalomyelitis

Reversal of experimental autoimmune encephalomyelitis by a soluble peptide variant ofa myelin basic protein epitope: T cell receptor antagonism and reduction of interferon 7 and tumor necrosis factor a production

A single TCR antagonist peptide inhibits experimental allergic encephalomyelitis mediated by a diverse T cell repertoire

Oral tolerance in experimental autoimmune encephalomyelitis. Ill, Evidence for clonal anergy JImmunol 199 I; 147:2 I 5.5-2 i 63, 80. Metzler B, Wraith DC. Inhibition of experimental autoimmune encephalomyelitis by inhalation but not oral adtnini strati on of the encephalitogenic peptide: influence of MHC binding afTin ity

Peripheral deletion of antigen-reactive T cells bi oral tolerance

Transgenic mice that express a myelin basic protein-specific T cell receptor develop spontaneous a ti to immunity

A molecular map of T cell development

Vaccination with BV8S2 protein amplifies TCR-specific regulation and protection against experimental autoimmune encephalomyelitis in TCR BV8S2 transgenic mice

An autoantigenic T cell epitope forms unstable complexes with class II MHC: a novel route for escape from tolerance induction

Myelin basic protein peptide complexes with the class II MHC molecules l-A" and I-A' form and dissociate rapidly at neutral pH

Pearson CI, van Fwijk V^ McDevitt HO. Induction of apoptosis and T belper 2 (Tb2) responses correlates witb peptide affinity for the major bistocompatibihty complex in se!freactive T cell receptor transgenic mice

Thymic stromal organization is regulated by tbe speciflcity ofT ceE receptor/major histocompatibility complex interactions

Differential tolerance is induced in T cells recognizing distinct epitopes of myelin basic protein

T cell determinant structure of myelin basic protein in Bl O.PL, SJL/J, and their FI s

Endogenous myelin basic protein inactivates the high avidity T cell repertoire

Highly cross-reactive T cdl responses to myelin basic protein epitopes reveal a non-predictable form of TCR degeneracy

Acute experimental autoimmune eiicephalomyeiilis in mice, I, Adjuvant aciion of BordctdlH pertussis is due to vasoactive amine sensitization and increased vascular permeabihty of the central nervous system

Flicitation of experimental allergic encepbalomyelitis (EAE) in mice with the aid of pertussigen

Triggers of autoimmune disease in a murine T-cell receptor transgenic model lor multiple sclerosis

High incidence of spontaneous autoimmime encepbalomyelitis in immunodeficient anti-myelin basic proteni T ceE receptor transgenic mice

Surface expression of a 4 integrin by CD4 T cells is required fc:ir tbeir entry into brain paroncliyma

Abrogation c: >f resistance to Theiler's virus-induced dem)'elination in H-2b mice deficient in P Z-micro globulin

Genetic analysis of susceptibility to experimental autoimmune encephalomyelitis in a cross between SJL/J and BlO.Smice

CeUs prevent spontaneous experimental autoimmune encephalomyelitis in anti-myelin basic protein T cell receptor transgenic mice

Regulatory CD4-T cells expressing endogenous T cell receptor cliains protect myelin basic protein-specific transgenic mice from spontaneous autoimmune encephalomyelitis

Myelin basic protein-specific and TCR Yf, 8,2-specLfic T-cell lines from TCR Vp 8,2 transgenic mice utilize the same Valpha and V beta genes: specificity associated with tbe V^CDR3-J(, region

Neonatal exposure of TCR BV8S2 transgenic mice to recombinant TCR BV8S2 results in reduced T cell proliieraiion and elevated antibody response to BV8S2, and increased severity of FAE

T ceE receptor peptide tlierapy triggers autoregulaiion of experimental enceplialomyelitis

Kumar y Sercarz E. Dysregulation of potentially patbogenic self reactivity is crucial for the manifestation of clinical autoimmunity

T cell deletion in high antigen dose therapy of autoimmune encephalomyeliti s

Propriocidal apoptosis of mattire T lymphocytes occurs at S phase of the cell cycle

Tolerance induction and autoimmune encepbalom)'elilis amelioration after administration oi myelm basic protein-derived peptide

Oral tolerance in myeEn basic protein T-cell receptor transgenic mice; suppression of autoimmiuie encepbalomyelitis and dose-dependent induction of regulatory cells

Treatment of autoimmune disease by oral tolerance to autoantigens