Multiple sclerosis (MS) is believed to be an autoimmune disease mediated by T cells specific for CNS Ags. MS lesions contain both CD4+ and CD8+ T lymphocytes. The contribution of CD4+ T cells to CNS autoimmune disease has been extensively studied in an animal model of MS, experimental autoimmune encephalomyelitis. However, little is known about the role of autoreactive CD8+ cytotoxic T cells in MS or experimental autoimmune encephalomyelitis. We demonstrate here that myelin basic protein (MBP) is processed in vivo by the MHC class I pathway leading to a MBP79–87/Kk complex. The recognition of this complex by MBP-specific cytotoxic T cells leads to a high degree of tolerance in vivo. This study is the first to show that the pool of self-reactive lymphocytes specific for MBP contain MHC class I-restricted T cells whose response is regulated in vivo by the induction of tolerance.

The ability to activate autoreactive T cells in the periphery of healthy animals in models of autoimmune disease demonstrates that central and peripheral tolerance is incomplete. This is illustrated in experimental autoimmune encephalomyelitis (EAE),3 an animal model of multiple sclerosis (MS) (1) that is induced by immunization with myelin basic protein (MBP) or by adoptive transfer of activated, MBP-specific T cells into naive recipients (2, 3, 4, 5). Although MS lesions contain both CD4+ and CD8+ T lymphocytes (6), studies of EAE focus on the role of CD4+ T cells. Thus, little is known about the role of autoreactive CD8+ cytotoxic T cells in the development and manifestation of EAE.

Previous studies suggested that CD8+ T cells might participate as effector or regulatory cells in EAE (7, 8, 9, 10). The existence of MHC class I-restricted T cells specific for naturally processed MBP in vivo has not been demonstrated, although human CD8+ T cells specific for a peptide of MBP have been isolated in vitro (11). These issues motivated us to study the MHC class I-restricted immune response to MBP and examine the potential role of endogenous MBP in shaping the CTL repertoire specific for this Ag.

C3HeB/FeJ and C3HeB/FeJ-MBPshi/+ were purchased from The Jackson Laboratory (Bar Harbor, ME). MBPshi/+ and MBPshi/shi (MBP−/−) mice (12, 13, 14) were identified by PCR (15) and whole body tremor.

An E1 inserted, Ad5 recombinant adenovirus expressing MBP (Ad/MBP) was generated by inserting the plasmid pXCJL.1 containing an MBP cDNA (16) into an E1-deficient Ad5 adenovirus pJM17 (17, 18). In addition, a L929 cell line expressing MBP (L/MBP) was generated using the expression vector pcDNA3.1 (Invitrogen, Carlsbad, CA) containing the MBP cDNA (16). Expression of MBP mRNA was detected in cells infected with Ad/MBP and L/MBP cells by RT-PCR (19). A recombinant vaccinia virus expressing MBP (Vac/MBP) (20), was obtained from Therion Biologics (Cambridge, MA).

Peptides were synthesized using TBOC chemistry on a model 430A peptide synthesizer (Applied Biosystems, Foster City, CA). The peptides were purified by reverse-phase HPLC, and all peptides were analyzed for purity by mass spectrometry.

Mice were infected with 107 pfu Ad/MBP virus i.p. or 106 pfu Vac/MBP i.v. After 3 wk infection, immunized mice were harvested and 3 × 107 splenocytes were stimulated in vitro in 10-ml cultures with 1 × 106 irradiated target cells in RPMI 1640 media supplemented with 10% FCS. All procedures have been approved by the animal care committee at the University of Washington.

Target cells were infected with virus at a multiplicity of infection of 10 and incubated for 72 h for adenovirus and 12 h for vaccinia virus in growth media before labeling with chromium. Target cells were then incubated with 100 μCi of (51Cr)O4 (Amersham, Arlington Heights, IL) for 60 min, washed, and incubated with effector cells in a standard 4-h 51Cr release assay. The percent lysis was calculated as (51Cr release in the presence of CTLs – spontaneous 51Cr release) × 100/(total 51Cr release in 2% Nonidet P-40 – spontaneous 51Cr release).

MBP-specific T cell lines and clones were tested in 51Cr release assays with L cells pulsed for 18 h with 30 μM overlapping 20- to 23-mer synthetic peptides in RPMI to allow processing of suboptimal peptides before chromium labeling. The 9-mer peptides were preloaded to L cells, RMA-S cells, and RMA-S-Kk (kindly provided by Dr. Peter Cresswell, Yale University, New Haven, CT) at known dilutions for 30 min before the addition of effector cells.

C3H MBP−/− and MBP+/+ mice were infected i.p. with Ad/MBP. Splenocytes from infected mice were harvested and stimulated in vitro for 5 days with L/MBP. The stimulated splenocytes were then tested for their ability to lyse syngeneic target cells expressing MBP. From 14 of 16 MBP−/−mice, MBP-specific killing was observed. In contrast, no MBP-specific killing by T cells from wild-type mice was observed (0 of 15 mice; data not shown).

Because adenovirus has a restricted tropism, a second protocol was used to assess MBP-specific cytotoxic T cell responses. Splenocytes from mice infected with Vac/MBP were stimulated in vitro with irradiated L cells infected with Vac/MBP. From all MBP+/+ and MBP−/− mice, a potent vaccinia-specific T cell response was generated. In contrast, an MBP-specific cytotoxic T cell response was generated only in MBP−/− mice (13 of 13) but not in MBP+/+ mice (0 of 15) (Fig. 1, A and B).

FIGURE 1.

MBP-specific CTL responses are generated in MBP−/− but not MBP+/+ mice. C3H MBP−/− (A) and C3H MBP+/+ (B) mice were i.v. infected with 106 pfu Vac/MBP. Spleens were removed after 3 wk and stimulated in vitro with irradiated L cells infected with Vac/MBP. B gal, β galactosidase.

FIGURE 1.

MBP-specific CTL responses are generated in MBP−/− but not MBP+/+ mice. C3H MBP−/− (A) and C3H MBP+/+ (B) mice were i.v. infected with 106 pfu Vac/MBP. Spleens were removed after 3 wk and stimulated in vitro with irradiated L cells infected with Vac/MBP. B gal, β galactosidase.

Close modal

Twenty-one MBP-specific T cell clones were established from three MBP−/− mice after one in vitro stimulation by limiting dilution cloning. All clones were of the αβ TCR+, CD8+ lineage (data not shown). The fine specificity of three MBP-specific T cell clones was determined using target cells pulsed with a panel of overlapping peptides comprising the entire MBP protein. All three clones specifically lysed target cells pulsed with MBP68–91 but no other peptides (Fig. 2 A). The remaining 18 T cell clones were tested and specifically lysed target cells presenting this peptide (data not shown).

FIGURE 2.

MBP-specific T cell lines and clones were used to define MHC class I epitopes within MBP. (A) T cell clones were used at an E:T ratio of 10:1, and (B) primary T cell lines derived from MBP−/− and MBP+/+ mice were used at an E:T ratio of 50:1 in a 51Cr release assay. This experiment was performed twice with three MBP clones and two MBP−/− and MBP+/+ mice in each group. B gal, β galactosidase.

FIGURE 2.

MBP-specific T cell lines and clones were used to define MHC class I epitopes within MBP. (A) T cell clones were used at an E:T ratio of 10:1, and (B) primary T cell lines derived from MBP−/− and MBP+/+ mice were used at an E:T ratio of 50:1 in a 51Cr release assay. This experiment was performed twice with three MBP clones and two MBP−/− and MBP+/+ mice in each group. B gal, β galactosidase.

Close modal

To assure that the exclusive specificity of the T cell clones for MBP68–91 was representative of all CTLs specific for MBP, T cell lines derived from MBP+/+ and MBP−/− mice after one in vitro stimulation were tested for their ability to lyse target cells coated with the panel of overlapping MBP peptides. Only T cells derived from MBP−/− mice were able to specifically lyse MBP-expressing targets. The MBP-specific response was directed solely at targets coated with MBP68–91, indicating that the dominant MBP epitope is contained in this region. (Fig. 2 B).

To identify the core 9-mer epitope that is targeted by MBP-specific CTLs, we first evaluated the sequence within MBP68–91 for an H-2k class I binding motif (21). The overlapping 9-mer peptides MBP78–86, MBP79–87, and MBP80–88 were tested. All MBP-specific CTL clones recognized target cells coated with the MBP79–87 peptide epitope in a dose-dependant manner (Fig. 3, and data not shown). These results indicate that MBP79–87 represents the naturally processed MBP-specific CTL epitope in C3H mice.

FIGURE 3.

The MBP-specific CTL epitope is defined by MBP79–87. MBP-specific T cell clones were tested in a 51Cr release assay with target cells pulsed with the overlapping 9-mer peptides MBP78–86, MBP79–87, and MBP80–88 at an E:T ratio of 10:1. Four independent T cell clones recognized only MBP79–87 in a dose-dependant manner; two representative T cell clones are shown (open symbols, filled symbols). This experiment was performed four times.

FIGURE 3.

The MBP-specific CTL epitope is defined by MBP79–87. MBP-specific T cell clones were tested in a 51Cr release assay with target cells pulsed with the overlapping 9-mer peptides MBP78–86, MBP79–87, and MBP80–88 at an E:T ratio of 10:1. Four independent T cell clones recognized only MBP79–87 in a dose-dependant manner; two representative T cell clones are shown (open symbols, filled symbols). This experiment was performed four times.

Close modal

To identify the MHC allele that presents MBP79–87 to CD8+ cytotoxic T cells, Con A blasts from the B10 MHC congenic strains B10.A(4R), B10.MBR, and C3H were used as target cells. The results suggested that the MBP79–87 peptide is presented by the MHC class I allele Kk (data not shown). To confirm this result, we tested the ability of RMA-S cells and RMA-S cells transfected with Kk to present the MBP epitope. RMA-S-Kk cells were able to present the MBP epitope, while untransfected RMA-S cells were not (Fig. 4).

FIGURE 4.

MBP79–87 is presented by H-2Kk. L, RMA-S, and RMA-S cells transfected with H-2Kk pulsed with MBP79–87 were used as targets for MBP-specific T cell clones at an E:T ratio of 10:1. These experiments were performed three times.

FIGURE 4.

MBP79–87 is presented by H-2Kk. L, RMA-S, and RMA-S cells transfected with H-2Kk pulsed with MBP79–87 were used as targets for MBP-specific T cell clones at an E:T ratio of 10:1. These experiments were performed three times.

Close modal

The results described above demonstrate that MBP-specific, MHC class I-restricted cytotoxic T cells are present in the periphery of MBP-deficient mice and that these T cells undergo tolerance in wild-type mice that express endogenous MBP. These observations raised the question of where tolerance to MHC class I epitopes of MBP occurs in vivo. To address this issue, we used two independent protocols to assess whether peripheral mechanisms are responsible for tolerance of MBP-specific CTLs. Results are shown in Table I. In the first experiment, MBP-specific T cells (group 1) and vaccinia-specific T cells (group 2) were transferred into SCID mice to test for retention of Ag-specific responses. 51Cr release assays were performed with the T cell lines just before transfer to confirm their CTL activity and the recipient SCID mice were bled 1 wk after transfer to assure survival of T cells after transfer (data not shown). After 4 wk, spleens from recipient mice were harvested and tested in CTL assays. Vaccinia-specific T cells were easily detected in recipients of vaccinia-specific T cells. In contrast, MBP-specific CTL activity was not detected in any recipients of MBP-specific CTLs. In a separate approach, we asked whether naive peripheral T cells from MBP−/− mice that have not been exposed to MBP would undergo tolerance when transferred into the periphery of MBP+/+ mice. SCID mice were reconstituted with naive lymphocytes isolated from MBP−/− (group 3) and MBP+/+ (group 4) mice. The mice were bled 1 wk after transfer to assure that the lymphocytes had reconstituted all mice equally. Four weeks after reconstitution, mice were infected with either Ad/MBP or Vac/MBP. Potent vaccinia-specific responses were generated in mice that received either MBP−/− or MBP+/+ lymphocytes when Vac/MBP was used both as the immunogen and to restimulate the T cells in vitro. We attempted to generate MBP-specific CTL responses by immunizing the mice with either Ad/MBP or Vac/MBP. However, no responses were detected in SCID mice reconstituted with lymphocytes from either MBP−/− or MBP+/+. Control MBP−/− mice (group 5) but not MBP+/+ mice (group 6) infected and restimulated using these protocols at the same time as the recipient mice generated MBP-specific CTL responses. Therefore, although these data do not exclude a role for central tolerance mechanisms, they indicate that peripheral mechanisms eliminate functional MBP-specific CTL responses from mice expressing endogenous MBP.

Table I.

MBP-specific T cells are tolerized in the periphery of wild-type mice

GroupMouseTransferred CellsImmun.Restim.% Specific Lysis
L cellsaMBP79–87Vac/MBPVac/control
SCID MBP T cell lineb N /A Vac /MBP 8, 2 7, 4 7, 1 6, 1 
SCID MBP T cell lineb N /A Vac /MBP 1, 3 2, 6 3, 4 1, 2 
SCID Vac T cell lineb N /A Vac /MBP 4, 6 5, 7 41, 36 40, 42 
SCID Vac T cell lineb N /A Vac /MBP 0, 2 2, 6 36, 25 32, 18 
SCID Naive MBP−/−c Ad /MBP MBP79–87 14, 8 14, 10 18, 9 20, 5 
SCID Naive MBP−/−c Vac /MBP Vac /MBP 1, 3 3, 3 48, 34 51, 40 
SCID Naive MBP+/+c Ad /MBP MBP79–87 2, 0 2, 3 1, 0 2, 0 
SCID Naive MBP+/+c Vac /MBP Vac /MBP 2, 3 3, 5 43, 28 52, 28 
MBP−/− N/A Ad /MBP MBP79–87 7, 4 40, 28 40, 22 11, 4 
MBP+/+ N/A Ad /MBP MBP79–87 16, 2 16, 4 16, 2 16, 3 
MBP−/− N/A Vac /MBP Vac /MBP 1, 3 16, 8 40, 42 52, 46 
MBP+/+ N/A Vac /MBP Vac /MBP 2, 2 4, 3 42, 28 56, 43 
GroupMouseTransferred CellsImmun.Restim.% Specific Lysis
L cellsaMBP79–87Vac/MBPVac/control
SCID MBP T cell lineb N /A Vac /MBP 8, 2 7, 4 7, 1 6, 1 
SCID MBP T cell lineb N /A Vac /MBP 1, 3 2, 6 3, 4 1, 2 
SCID Vac T cell lineb N /A Vac /MBP 4, 6 5, 7 41, 36 40, 42 
SCID Vac T cell lineb N /A Vac /MBP 0, 2 2, 6 36, 25 32, 18 
SCID Naive MBP−/−c Ad /MBP MBP79–87 14, 8 14, 10 18, 9 20, 5 
SCID Naive MBP−/−c Vac /MBP Vac /MBP 1, 3 3, 3 48, 34 51, 40 
SCID Naive MBP+/+c Ad /MBP MBP79–87 2, 0 2, 3 1, 0 2, 0 
SCID Naive MBP+/+c Vac /MBP Vac /MBP 2, 3 3, 5 43, 28 52, 28 
MBP−/− N/A Ad /MBP MBP79–87 7, 4 40, 28 40, 22 11, 4 
MBP+/+ N/A Ad /MBP MBP79–87 16, 2 16, 4 16, 2 16, 3 
MBP−/− N/A Vac /MBP Vac /MBP 1, 3 16, 8 40, 42 52, 46 
MBP+/+ N/A Vac /MBP Vac /MBP 2, 2 4, 3 42, 28 56, 43 
a

T cells were used at an E:T ratio of 50:1 and 12:1. This experiment was repeated twice with similar results.

b

A total of 3 × 107 MBP or vaccinia-specific primary T cells were transferred i.v. after one in vitro stimulation.

c

A total of 2 × 108 spleen and lymph node cells from nonmanipulated MBP−/− or MBP+/+ mice were transferred i.v.

Tolerance in MBP-specific CTLs could be maintained in wild-type mice by encountering the MBP79–87 epitope in the periphery. The MBP gene locus is complex and contains at least two additional promoters 5′ of the classical promoter, which transcribe a unique family of gene products termed golli-MBP (19, 22, 23). These genes are encoded by a combination of exons from classical MBP as well as exons 5′ of classical MBP. Transcripts and protein expression of the golli-MBP isoforms, which contain the MBP79–87 epitope, have been found outside of the nervous system in both lymphoid tissue and in major organs (19, 22, 23, 24).

Recently, it has been demonstrated that some MHC class II-restricted T cells specific for MBP are also efficiently tolerized in wild-type mice (15, 25). The tolerogenic CD4+ T cell epitopes of MBP in H-2u mice have not been found in any of the golli-MBP protein. Therefore, immune tolerance in MHC class II-restricted T cells can be mediated by endogenous expression of classical MBP. Because of differences in tissue distribution and MHC class, it is possible that MHC class I- and class II-restricted T cells specific for MBP undergo tolerance via different mechanisms.

The studies reported here describe a model system in which MHC class I-restricted T cells specific for MBP can be generated and analyzed for their contribution to autoimmune disease. The identification of naturally occurring MHC class I-restricted epitopes allows monitoring of MBP-specific CTL responses during the course of disease. Therefore, the ability of MBP-specific CTLs to be activated by (or contribute to) determinant spreading can be investigated (26, 27, 28). This model system provides a novel approach to define the role of CNS Ag-specific CD8+ CTLs in the pathogenesis of autoimmune disease.

We thank Priya Gopaul for technical assistance and Dr. Mark Kay for technical assistance with adenovirus construction.

1

This work was supported by a grant from the Royalty Research Fund of the University of Washington. J.G. is supported in part by a Harry Weaver Junior Faculty Award (2080-A-2) from the National Multiple Sclerosis Society. E.S.H. is supported by a Predoctoral Fellowship from the National Institutes of Health (CA09537-13).

3

Abbreviations used in this paper: EAE, experimental autoimmune encephalomyelitis; MS, multiple sclerosis; MBP, myelin basic protein; Ad/MBP, adenovirus expressing MBP; L/MBP, L929 cell line expressing MBP; Vac/MBP, vaccinia virus expressing MBP.

1
Martin, R., H. F. McFarland.
1997
. Immunology of multiple sclerosis and experimental allergic enchephalomeyelitis. C. S. Raine, and H. F. McFarland, and W. W. Tourtellotte, eds.
Multiple Sclerosis: Clinical and Pathogenic Basis
221
Chapman & Hall, London.
2
Paterson, P. Y., R. H. Swanborg.
1988
. Demyelinating diseases of the central and peripheral nervous systems. M. Samter, and D. W. Talmage, and M. M. Frank, and K. F. Austen, and H. N. Claman, eds.
Immunological Diseases
4th ed.
1877
Little Brown and Co., Boston, MA.
3
Zamvil, S., P. Nelson, J. Trotter, D. Mitchell, R. Knobler, R. Fritz, L. Steinman.
1985
. T-cell clones specific for myelin basic protein induce chronic relapsing paralysis and demyelination.
Nature
317
:
355
4
Swanborg, R. H..
1988
. Experimental allergic encephalomyelitis.
Methods Enzymol.
162
:
413
5
Zamvil, S. S., L. Steinman.
1990
. The T lymphocyte in experimental allergic encephalomyelitis.
Annu. Rev. Immunol.
8
:
579
6
Martin, R., H. F. McFarland, D. E. McFarlin.
1992
. Immunological aspects of demyelinating diseases.
Annu. Rev. Immunol.
10
:
153
7
Koh, D. R., W. P. Fung Leung, A. Ho, D. Gray, H. Acha Orbea, T. W. Mak.
1992
. Less mortality but more relapses in experimental allergic encephalomyelitis in CD8−/− mice.
Science
256
:
1210
8
Jiang, H., S. I. Zhang, B. Pernis.
1992
. Role of CD8+ T cells in murine experimental allergic encephalomyelitis.
Science
256
:
1213
9
Evans, C. F., M. S. Horwitz, M. V. Hobbs, M. B. Oldstone.
1996
. Viral infection of transgenic mice expressing a viral protein in oligodendrocytes leads to chronic central nervous system autoimmune disease.
J. Exp. Med.
184
:
2371
10
Sun, D., Y. Qin, J. Chluba, J. T. Epplen, H. Wekerle.
1988
. Suppression of experimentally induced autoimmune encephalomyelitis by cytolytic T-T cell interactions.
Nature
332
:
843
11
Tsuchida, T., K. C. Parker, R. V. Turner, H. F. McFarland, J. E. Coligan, W. E. Biddison.
1994
. Autoreactive CD8+ T-cell responses to human myelin protein-derived peptides.
Proc. Natl. Acad. Sci. USA
91
:
10859
12
Dupouey, P., C. Jacque, J. M. Bourre, F. Cesselin, A. Privat, N. Baumann.
1979
. Immunochemical studies of myelin basic protein in shiverer mouse devoid of major dense line of myelin.
Neurosci. Lett.
12
:
113
13
Kirschner, D. A., A. L. Ganser.
1980
. Compact myelin exists in the absence of basic protein in the shiverer mutant mouse.
Nature
283
:
207
14
Roach, A., N. Takahashi, D. Pravtcheva, F. Ruddle, L. Hood.
1985
. Chromosomal mapping of mouse myelin basic protein gene and structure and transcription of the partially deleted gene in shiverer mutant mice.
Cell
42
:
149
15
Harrington, C. J., A. Paez, T. Hunkapiller, V. Mannikko, T. Brabb, M. Ahearn, C. Beeson, J. Goverman.
1998
. Differential tolerance is induced in T cells recognizing distinct epitopes of myelin basic protein.
Immunity
8
:
571
16
Kimura, M., H. Inoko, M. Katsuki, A. Ando, T. Sato, T. Hirose, H. Takashima, S. Inayama, H. Okano, K. Takamatsu, K. Mikoshiba, Y. Tsukada, I. Watanabe.
1985
. Molecular genetic analysis of myelin-deficient mice: shiverer mutant mice show deletion in gene(s) coding for myelin basic protein.
J. Neurochem.
44
:
692
17
McGrory, W. J., D. S. Bautista, F. L. Graham.
1988
. A simple technique for the rescue of early region I mutations into infectious human adenovirus type 5.
Virology
163
:
614
18
Hitt, M., A. J. Bett, L. Prevec, F. L. Graham.
1994
. Construction and propagation of human adenovirus vectors. J. E. Celis, ed.
Cell Biology: A Laboratory Handbook
479
Academic Press, San Diego, CA.
19
Mathisen, P. M., S. Pease, J. Garvey, L. Hood, C. Readhead.
1993
. Identification of an embryonic isoform of myelin basic protein that is expressed widely in the mouse embryo.
Proc. Natl. Acad. Sci. USA
90
:
10125
20
Genain, C. P., L. Gritz, N. Joshi, D. Panicali, R. L. Davis, J. N. Whitaker, N. L. Letvin, S. L. Hauser.
1997
. Inhibition of allergic encephalomyelitis in marmosets by vaccination with recombinant vaccinia virus encoding for myelin basic protein.
J. Neuroimmunol.
79
:
119
21
Elliot, T., M. Smith, P. Driscoll, A. McMichael.
1993
. Peptide selection by class I molecules of the major histocompatibility complex.
Curr. Biol.
3
:
854
22
Campagnoni, A. T., T. M. Pribyl, C. W. Campagnoni, K. Kampf, S. Amur Umarjee, C. F. Landry, V. W. Handley, S. L. Newman, B. Garbay, K. Kitamura.
1993
. Structure and developmental regulation of Golli-mbp, a 105-kilobase gene that encompasses the myelin basic protein gene and is expressed in cells in the oligodendrocyte lineage in the brain.
J. Biol. Chem.
268
:
4930
23
Zelenika, D., B. Grima, B. Pessac.
1993
. A new family of transcripts of the myelin basic protein gene: expression in brain and in immune system.
J. Neurochem.
60
:
1574
24
Voskuhl, R. R..
1998
. Myelin protein expression in lymphoid tissues: implications for peripheral tolerance.
Immunol. Rev.
164
:
81
25
Targoni, O. S., P. V. Lehmann.
1998
. Endogenous myelin basic protein inactivates the high avidity T cell repertoire.
J. Exp. Med.
187
:
2055
26
Lehmann, P. V., T. Forsthuber, A. Miller, E. E. Sercarz.
1992
. Spreading of T-cell autoimmunity to cryptic determinants of an auto-antigen.
Nature
358
:
155
27
Miller, S. D., C.L. Vanderlugt, W. S. Begolka, W. Pao, R. L. Yauch, K. L. Neville, L. Y. Katz, A. Carrizosa, B. S. Kim.
1997
. Persistent infection with Theiler’s virus leads to CNS autoimmunity via epitope spreading.
Nat. Med.
3
:
1133
28
Tuohy, V. K., M. Yu, L. Yin, J. A. Kawczak, J. M. Johnson, P. M. Mathisen, B. Weinstock-Guttman, R. P. Kinkel.
1998
. The epitope spreading cascade during progression of experimental autoimmune encephalomyelitis and multiple sclerosis.
Immunol. Rev.
164
:
93