Myelin autoreactivity in multiple sclerosis: Recognition ... - Europe PMC

1 downloads 0 Views 1MB Size Report
MARTIN PETTEt, KYOKO FUJITAt, DAVID WILKINSONt, DANIEL M. ALTMANNt,JOHN ... tMax Planck Society, Clinical Research Unit for Multiple Sclerosis, D-8900 Wurzburg, ..... We thank Dr. J. N. Whitaker for providing hMBP fragments, Dr.
Proc. Natl. Acad. Sci. USA Vol. 87, pp. 7968-7972, October 1990 Immunology

Myelin autoreactivity in multiple sclerosis: Recognition of myelin basic protein in the context of HLA-DR2 products by T lymphocytes of multiple-sclerosis patients and healthy donors (autoimmunity/epitope mapping)

MARTIN PETTEt, KYOKO FUJITAt, DAVID WILKINSONt, DANIEL M. ALTMANNt, JOHN TROWSDALE*, GERHARD GIEGERICH§, ARI HINKKANEN§, J6RG T. EPPLEN§, LUDWIG KAPPOSt, AND HARTMUT WEKERLEt§¶ tMax Planck Society, Clinical Research Unit for Multiple Sclerosis, D-8900 Wurzburg, Federal Republic of Germany; tThe Imperial Cancer Research Fund Laboratories, London WC2A 3PX, United Kingdom; and §Max Planck Institute for Psychiatry, D-8033 Martinsried, Federal Republic of Germany Communicated by Hans J. Muller-Eberhard, July 5, 1990 (received for review May 3, 1990)

Products of HLA-DR loci have been shown to act as the dominant restriction elements for myelin basic protein (MBP)-specific T lymphocytes isolated from the peripheral blood (11, 12). Most DR haplotypes contain genes encoding two functional products differing with respect to their polymorphic 3 chains (13). These different gene products are concomitantly expressed on the surface of conventional antigen-presenting cells (APCs). To circumvent this problem and to address whether both gene products of a given DR haplotype may be involved in the presentation of the autoantigen MBP, we used mouse L-cell lines transfected with the DR2a or DR2b genes of the DR2Dw2 haplotype, respectively. The results indicate that human MBP (hMBP) can be presented and recognized in the context of either DR2a or DR2b heterodimers. Most lines, however, were restricted by DR2a rather than by DR2b molecules. With a set of 10 proteolytically obtained hMBP fragments as well as four synthetic peptides, at least five different T-cell epitopes were shown to be presented by the DR2a heterodimer of the DR2Dw2 haplotype. We found a preferential recognition of a peptide comprising amino acids (aa) 139-153.

A panel of 20 human myelin basic protein ABSTRACT (hMBP)-specific T-lymphocyte lines was generated from the peripheral blood of eight multiple sclerosis (MS) patients and two healthy donors, most of them expressing the HLA-DR2 haplotype, which is associated with an increased susceptibility to MS. Using HLA-DR gene-transfected mouse L-cell lines as antigen-presenting cells, we established that of the 20 hMBPspecific T-lymphocyte lines, 7 were restricted by the DR2a gene products of the DR2Dw2 haplotype. Four T-cell lines recognized hMBP in the context of the DR2b products of the DR2Dw2 haplotype. DR2b-restricted T-cell responses were demonstrable only in T-cell lines derived from MS patients. The hMBP epitopes presented by the DR2a heterodimer were mapped to peptides covering amino acid residues 144, 76-91, 131-145, or 139-153 and to a region spanning the thrombincleaved bond at Argl3 Alat3I. DR2b-restricted T-cell lines recognized epitopes within amino acids 80-99 and 148-162. Peptide 139-153 was also presented in the context of HLA-DR1 molecules. Our data show that (i) in MS patients both the DR2a and DR2b products of the DR2IDw2 haplotype function as restriction elements for the myelin autoantigen hMBP, (ii) the DR2a molecule presents at least five different epitopes to hMBP-specific T lymphocytes, and (iu) anti-hMBP T-cell lines derived from individual donors can differ in their antigen fmne specificity as well as in their HLA restriction.

MATERIALS AND METHODS Patients and Healthy Donors. Seven patients (codes BB, CK, EB, HH, MA, WI, and WR) suffering from clinically definite MS [according to the criteria of Schumacher et al. (14)], one patient with probable MS (code PS), and two healthy individuals (codes ER and HW) participated in the present study (Table 1). Establishment of hMBP-Speciflic T-Lymphocyte Lines. Peripheral blood leukocytes were separated by Ficoll/Hypaque density gradient centrifugation. The cells were resuspended (2 x 106 per ml) in complete culture medium [RPMI 1640 (GIBCO) with 2 mM L-glutamine, 100 units of penicillin per ml, 100 ,g of streptomycin per ml, and 5% heat-inactivated autologous or pooled AB serum] and seeded (100 ,ul per well) into 96-well round-bottom microtiter plates (Nunc). hMBP was added to a final concentration of -30 ,ug/ml. The cultures were incubated for 3-4 days in a humidified, 5% CO2 atmosphere before the addition of 100 ,ul of culture medium containing 3 units of highly purified human interleukin 2 (IL-2) (Biotest, Frankfurt) to each well. The cultures were fed with fresh IL-2-containing medium every 3-4 days, for a total culture period of 18-20 days. After this period, primary cultured cells were washed in the plate to remove IL-2 and

As in other human diseases with a putative autoimmune pathogenesis, susceptibility to multiple sclerosis (MS) is linked to certain HLA haplotypes. In some Caucasian populations, HLA-DR2 is associated with an enhanced disease probability (1). Although the existence of other HLA-Dassociated "disease genes" cannot be excluded (2, 3), it is most probable that the HLA class II genes, which control the myelin reactivity of T lymphocytes, are the key factors involved in this association (4). HLA class II products are believed to affect (auto)immune responses by selectively binding certain fragments of the relevant protein (auto-) antigen on the one hand and by being recognized together with the antigen by the variable region of the T-cell receptor on the other (5). Our approach to study these interactions in detail is based on the isolation of antigen-specific T-lymphocyte lines and the characterization of the molecular elements involved in the antigen-recognition process that may participate in the pathogenesis of MS. In animal models of myelin autoimmunity, detailed analyses of these interactions have been used as a basis for specific immune intervention (6-10).

Abbreviations: aa, amino acid(s); APC, antigen-presenting cell; IL-2, interleukin 2; MBP, myelin basic protein; hMBP, human MBP; MHC, major histocompatibility complex; MS, multiple sclerosis. $To whom reprint requests should be addressed.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 7968

Immunology: Pette et al.

Proc. Natl. Acad. Sci. USA 87 (1990)

Table 1. Age and HLA-DR haplotypes of MS patients and healthy donors from whom MBP-specific T-cell lines were established Initials Sex Age HLA-A HLA-B HLA-C HLA-DQ HLA-DR HLA-DR2' MS patients 2 EB F 30 3/24 7 7 W1 HH M 37 ND 2/w11 3/10 35/45 wl/w3 MA F 48 26/29 w7 2 7/44 wl/w2 M 25 7 2 PS ND wl 2/3 2/7 WI F 27 3 7 w3 wl/w2 2/3 WR M 37 1/3 7/8 w7 wl/w2 HLA-DR2' healthy donors wi ER F 35 2/w8 2/24 7/44 w5/w7 HW M 42 7 2/6 2/3 7/44 W1 HLA-DRI+ MS patients BB M 29 7 ND 1/7 1/28 7/8 w4 1/5 CK F 49 3/32 35/51 wl/w3 ND, not determined.

45-89, and 90-190 were obtained by cleavage of purified hMBP with brain cathepsin D (19) and purified by gel filtration and ion-exchange chromatography. hMBP fragments 1-97, 14-53, 59-75, 76-91, 98-170, 98-130, and 131170 were obtained as described (20). hMBP peptides 80-99, 131-145, 139-153, and 148-162 were synthesized using an Applied Biosystems automatic peptide synthesizer (model 431A) according to the fluorenylmethoxycarbonyl method and were purified by HPLC. RESULTS Determination of the Major Histocompatibility Complex (MHC) Restriction. Seventeen hMBP-specific T-lymphocyte lines were isolated from eight HLA-DR2+ donors [six patients and two healthy donors (HD)]. All of these lines were restricted by HLA-DR gene products as judged from inhibition using monoclonal anti-DR antibodies (data not shown). Seven T-cell lines (three MS and four HD) recognized hMBP in the context of the DR2a product, and four (all MS) were restricted to the DR2b product of the DR2Dw2 haplotype. In addition, six hMBP-specific T-cell lines (four MS and two HD) isolated from three HLA-DR2+ MS patients and one HLA-DR2+ healthy donor recognized hMBP when presented by autologous APCs but not when presented by the DR2a or DR2b transfectants (Fig. 1). Thus they may be restricted by the other DR allele (e.g., T-cell line WR BP2) or, in the case of the DR2-homozygous donors, by other DR2 subtypes (e.g., MA BP2). Moreover, three T-cell lines from two HLA-DR1+ MS patients reacted to hMBP presented by L cells expressing the HLA-DR1 heterodimer (data not shown). Purified protein derivative from Mycobacterium tuberculosis and nontransfected Ltk- cells were used as negative controls and did not induce a proliferative response in any of the T-cell lines (data not shown). Epitope Mapping with Proteolytic Fragments of hMBP. In an attempt to localize the epitopes recognized by the hMBPspecific T-cell lines in the context of the DR2a and DR2b products, we screened the T-cell lines for reactivity against a set of proteolytically split hMBP fragments (Fig. 2). At least four different hMBP epitopes were found to be presented by the DR2a product. These were mapped to aa 1-44 (ER BP4, Fig. 2a) and 76-91 (PS BP7, Fig. 3; ER BP4, Fig. 2a), to a region overlapping the thrombin-cleaved bond at Arg130Ala13' (HH BP1, Fig. 2b), and to aa 131-170 (ER BP3, Fig. 2c; HW BP3, Fig. 2d; PS BP1, data not shown). Only T-cell line ER BP4 recognized two independent fragments, indicating oligoclonality in this line (Fig. 2a). Fragments 76-91 and 131-170 were recognized by T-cell lines from MS patients as well as from healthy donors. Three T-cell lines from two MS patients (WI BP2, WI BP4, and WR BP4, data not shown), restricted by the DR2b heterodimer, recognized an epitope

resuspended in 100 Al of culture medium. The contents (100 ,ul) of each well were split and subcultured by transfer to adjacent wells on a new culture plate. Autologous irradiated (40 Gy) peripheral blood leukocytes were suspended (4 x 106 per ml) in complete medium. Fifty microliters of this suspension was added to both subcultures and 10 ,ul of hMBP (3 ,ug) was then added to only one. By this means, cell populations grown in single wells in the presence of hMBP and IL-2 could be screened for antigen-dependent proliferation during the first restimulation. Corresponding cultures were examined every day and those showing an antigen-specific cell proliferation were identified. After 3-4 days, IL-2containing medium was added. Antigen-specific populations were then expanded by several restimulations in the presence of antigen and autologous APCs. L-Cell Transfectants. Construction of the L-cell transfectants expressing the DR2a or DR2b molecules (DRBS*010J or DRBI*1501; ref. 15), respectively, has been described in detail (16). To obtain HLA-DR1 transfectants the corresponding DRB1 cDNA (DRBI*0101; refs. 15 and 17), together with a full-length DRA cDNA clone and the hygromycin B-resistance gene, was transfected into mouse Ltk- cells. Proliferation Assays. Transfected L cells grown to confluence were trypsinized and treated with mitomycin C (16). They were transferred (5 x 104 cells) along with 2 x 104 T lymphocytes to 96-well flat-bottom microtiter plates (Nunc) in 100 ,ul of complete medium with 5% fetal bovine serum (Biochrom, Berlin). hMBP was added at a final concentration of 3 ,ug/ml. Preparation and Purification of Antigens. hMBP was prepared by a standard procedure (18). hMBP fragments 1-44, MS patients

Healthy donors

HH BP1 PS BP1

PS BP7

ER BP3 ER BP4 HW BPb HW BP3

. ---_

_

_

EB BP6 a WI BP2 WI BP4 WR BP4

HW BP2 HW BP10

MA BP2

MA BP4 PS BP6 WR BP2 4 CPM

8 X

1-

12

In

u

7969

4

cpm

8 X

10-4

12

FIG. 1. MHC restriction of hMBP-specific T-cell lines derived from HLA-DR2' MS patients (Left) or from HLA-DR2' healthy donors (Right). Antigens were presented by mouse L cells transfected with the DRA/DRB5 (LDR2a; upper, black bar for each T-cell line) or DRA/DRB1 (LDR2b; lower, light bar) genes of the DR2Dw2 haplotype. Nontransfected mouse L cells (Ltk-) were used as control APCs. Results represent [3H]thymidine in-

corporation (cpm).

. _:

Immunology: Pette et al.

7970 'I

Proc. Natl. Acad. Sci. USA 87 (1990)

.-R 131 {Z4IID I.I)

C

.\tR

\ I ),( 'S:, I. 1.) R2 1 70

89

:4

1 _0

74 t£-~9~

~

44

_

08889899

0990

109

*

53

II D1\

'2.R~~~~M" ,

59

o

rZ4

91

>

cpm'11

U)

I1'

i._

;.', 1.1

(1

H1II B11 1NIS pl'ticliti *\P)( : I.1)R4.170

l

1 70 94

9999o7

Kr..FF,;,0,.'. .m:'.:.'...>-XezrE^..-\.-.-.wnf-: a .--.

° -8,,,eS>!-,?

090 99991

- C B-' H-}';'v

-

.

gS

s>';'.- ARUS'"'"$;FNn

,_i.,-E

7C -

:3

11

1>(. s: I .15I_:

=

t'

1 3

98

39**

Yl

I IV :z) I 11I

07

58

ZOO`.g,'

M

,T

!

I

CI1V11 ". 10

FIG. 2. Antigen fine specificities of hMBP-specific T-cell lines from MS patients and healthy donors (HD). The hMBP fragments tested are shown at left. Results represent [3H]thymidine incorporation (mean ± SD). Antigens were used at optimal concentrations [hMBP (aa 1-170), 3 ,ug/ml; 1-97 and 98-170, 1 ,ug/ml; 1-44, 45-89, 98-130, and 131-170,0.3 ,g/ml; 14-53, 59-75, and 76-91, 1 ,ug/ml] except for the experiments depicted in A and D, where 0.3 ,tg was used.

T-cell line HH BP1 (see also Fig. 2c) responded to none of the synthetic peptides tested (Fig. 3). With the exception of the T-cell lines specific for peptide 80-99 (WI BP2, WI BP4, and WR BP4), all lines showed a maximal response in the presence of the recognized synthetic peptides that corresponded to the maximal response against the whole hMBP molecule (Fig. 3 and data not shown). This indicated that the synthetic peptides function efficiently as epitopes for each T-cell line and that these lines are

within hMBP aa 1-97, but not within aa 1-44, 14-53, 45-89, or 76-91, in the context of the DR2b product. One MS patient-derived T-cell line (EB BP6, data not shown), also restricted by the DR2b heterodimer, recognized the hMBP fragment 131-170. Epitope Fine Mapping with Synthetic Peptides. Overlapping peptides were synthesized that contained sequences of the thrombic fragment 131-170. Each of these peptides was recognized by at least one hMBP-specific T-lymphocyte line.

8HH BP1 (LDR2a) *-. MBP o..---.o Ml A---A M2 *--- M3.6

4 '--4

HW BPb (LDR2a) .-. MBP o-o Ml -40 A---& M2 **-- M3.6/

6

\

.

CL

2(1

20 I

*

0.01

0.1

10

i

MBP 76-91 .o---- o M7.2

/

4

x u.

PS BP7 (LDR2a)

1

2

o 0.01

06

---

0.1

1

0.1

0.01

10

10

1

l5

60- PS- BP1 (LDR2a) MBP O ..o Ml A'-',&M2 ° 40- .--. M3.6

A-

_

4

If

x C

------

30 EB BP6 (LDR2b) -. MBP T-I O----o Ml , /L/ -A M2 20- ----. M3.6 T."I

"T

/

20-

10-

/

A

/----0.1

Antigen,

1

tg/ml

10

0.001

0.01

-

*....

5-

,,-

11,\~~~~~~~~I

nj

bOi0

WI BP2 (LDR2b) MBP M7.2 10 A 76-91

-

s.

--

0.1

Antigen, ,g/ml

.

A

,~

_

1

10

0.1

1

10

Antigen, gg/ml

FIG. 3. Distinct epitope recognition patterns of T-cell lines specific for the hMBP thrombic peptide 131-170, determined by proliferation in the presence of hMBP; synthetic peptide Ml (aa 131-145), M2 (aa 139-153), M3.6 (aa 148-162), or M7.2 (aa 80-99); or proteolytic fragment 76-91. APCs (LDR2a or LDR2b) are indicated in parentheses. Results represent [3H]thymidine incorporation (mean + SD). assays

Immunology: Pette et al. Table 2. Antigen fine specificity and HLA-DR restriction of MBP-specific T-cell lines isolated from the peripheral blood of MS patients and healthy donors (HD) T-cell line Status APCs Target sequence HH BP1 MS LDR2a 98-170 PS BP7 MS LDR2a 76-91 PS BP1 MS LDR2a 139-153 ER BP4 HD LDR2a 1-44 ER BP3 HD LDR2a 139-153 HW BPb HD LDR2a 131-145 HW BP3 HD LDR2a 139-153 WI BP2 MS LDR2b 80-99 WI BP4 MS LDR2b 80-99 WR BP4 MS LDR2b 80-99 EB BP6 MS LDR2b 148-162 BB BPc9 MS LDR1 139-153 CK BP4 MS LDR1 139-153 CK BP14 MS LDR1 139-153

homogeneous. The epitope for the T-cell lines that are specific for aa 80-99 may contain posttranslational modifications, since the proteolytic fragment 1-97 stimulates as well as the whole hMBP molecule, but the synthetic peptide 80-99 does not. A second explanation for this phenomenon may be that the absence of the amino acids N-terminal to aa 80 leads to a conformational variation of the epitope recognized by WI BP2, WI BP4, and WR BP4. This conjecture is supported by the fact that the proteolytic fragments 45-89, 76-91, and 90-170 did not have any stimulatory effect on the T-lymphocyte lines WI BP2, WI BP4, and WR BP4 (data not shown). An additional T-cell epitope within hMBP aa 76-99 stimulated PS BP7 (Fig. 3). None of the lines specific for hMBP aa 80-99 was stimulated by the overlapping peptide 76-91, demonstrating their strict specificity (WI BP2, Fig. 3; WI BP4 and WR BP4, data not shown). The data are summarized in Table 2.

DISCUSSION The exact cause of MS remains to be defined. For several reasons, however, we consider hMBP a prime candidate as a myelin target autoantigen involved in the pathogenesis of this disease. (i) MBP is the dominant encephalitogenic autoantigen in all species examined, ranging from chicken to primates (21). (ii) In a virus-triggered model of subacute rat encephalomyelitis, the actual pathogenic mechanism was suggested to be a myelin-directed autoimmune T-cell response focused on guinea pig MBP (22). (iii) In children with measles encephalomyelitis, an enhanced cellular reactivity against hMBP has been observed, indicating an autoantigenic potential of hMBP in humans (23). (iv) In patients with encephalomyelitis following treatment with myelin-containing rabies vaccine, MBP was reported to be the encephalitogen (24). Clearly, however, other myelin determinants must and will be considered as potential autoantigens in MS. A previous report (25) showed that hMBP-specific T-cell lines could be established from 80%o of the peripheral blood samples obtained from both MS patients and healthy donors. The presence of hMBP-reactive T-cell clones within the immune repertoire of healthy donors is not too surprising (26, 27), since lethally encephalitogenic MBP-specific T-cell lines can be isolated from the immune system of completely normal Lewis rats, which never spontaneously develop experimental autoimmune encephalomyelitis (EAE)-like disease (28). Hence, the presence of hMBP-reactive T-cell precursors in healthy human beings does not argue per se against a pathogenic potential of these cells in MS.

Proc. Natl. Acad. Sci. USA 87 (1990)

7971

In the present study we have focused on the recognition of autoantigenic target epitopes in the context of defined MHC products by hMBP-specific T-lymphocyte lines. HLA restriction of the hMBP-reactive T-cell lines was determined using a set of mouse L cells transfected with the genes of the DR2Dw2 haplotype. Two previous observations had suggested that this haplotype might be of particular interest with respect to MS. (i) The DR2Dw2 haplotype is associated with an enhanced susceptibility to MS (1). (ii) The majority of hMBP-specific T-cell lines are restricted by HLA-DR products (11, 12, 25). The epitopes recognized by the T-cell lines were mapped initially by using proteolytic fragments of hMBP. This approach allows one to narrow down the potential T-cell epitopes of an antigen in a stepwise fashion. Moreover, it conserves potential posttranslational modifications of the antigen. This may be of greater importance than is commonly assumed; for example, the majority of all MBP-specific, encephalitogenic T-cell clones derived from the PL/J mouse strain recognize the critical epitopes 1-9 solely in the acetylated state (29, 30). The initial mapping study was then extended by using a set of synthetic peptides. With overlapping peptides we assigned at least three independent T-cell epitopes within the thrombic hMBP fragment 131-170. A fourth T-cell epitope recognized by DR2b-restricted T-lymphocyte lines was defined using the peptide 80-99. In addition, we have demonstrated that the DR2a product of the DR2Dw2 haplotype can present at least five different hMBP epitopes to human T lymphocytes. These epitopes were localized to aa 1-44, 76-91, 131-145, and 139-153 and a region including aa 131. As reported by others (12, 31), T-cell lines isolated from single donors and restricted by one MHC product (e.g., PS BP1, PS BP7, ER BP3, ER BP4) were specific for different hMBP epitopes. This is in stark contrast to the MBP-specific T-cell response in rodents, in which single epitopes are immunodominant (29, 30). The limited number of T-cell lines examined in this study does not allow definite conclusions with respect to epitope selection by either of the HLA determinants examined. However, that six of eight T-lymphocyte lines reactive to the hMBP C-terminal sequences (aa 131-170) were specific for peptide 139-153 suggests dominance of this epitope in the context of HLA-DR2 and HLA-DR1. Both of the two nonallelic products of the DR2Dw2 haplotype are used as restriction elements for antigen-specific T lymphocytes. Thus, we present an example of autoantigen recognition in the context of the DR2b product of the DR2Dw2 haplotype. Interestingly, only MS patients, but not healthy donors, had T cells showing DR2b-restricted presentation/recognition of hMBP. In other studies on presentation of a wide range of microbial antigens to DR2-restricted T cells, 15 of 16 clones studied were restricted by the DR2a product of the DR2Dw2 haplotype (D.M.A. and D. Sansom, unpublished observations). We. assume that hMBP-reactive T cells may be involved in the pathogenesis of MS, but we have no information about the actual pathogenic capacity of individual T-cell lines. In the rat, for example, the encephalitogenic potential of MBPspecific T-cell clones is variable and critically dependent on the target epitope recognized (32). We cannot rule out the possibility that human encephalitogenic T-cell lines are restricted to one particular hMBP epitope. However, our data indicate that the human immune response to MBP is more complex than that of rodents. This situation cautions against overly optimistic expectations for simple therapeutic strategies derived from rodent experimental allergic encephalomyelitis model systems.

7972

Immunology: Pette et al.

Note Added in Proof. Since submission of this manuscript, similar epitopes were reported by Ota et al. (33) and Martin et al. (34). We thank Dr. J. N. Whitaker for providing hMBP fragments, Dr. W. Risau for peptide synthesis, Dr. E. Albert for the HLA typing, Dr. C. Linington for reading the manuscript, and Ms. Heide Roth for processing it. This work was supported by the Deutsche Forschungsgemeinschaft (Ep 7/3-2) and the Sander-Stiftung (AZ 88.032.1). The Clinical Research Unit was supported by funds of the Hermann and Lilly Schilling Foundation. 1. Compston, A. (1986) Multiple Sclerosis, eds. McDonald, W. I. & Silberberg, D. H. (Butterworth, London), pp. 56-73. 2. Sargent, C. A., Dunham, I., Trowsdale, J. & Campbell, R. D. (1989) Proc. Natl. Acad. Sci. USA 86, 1968-1972. 3. Spies, T., Blanck, G., Bresnahan, M., Sands, J. & Strominger, J. L. (1989) Science 243, 214-216. 4. Wraith, D. C., McDevitt, H. O., Steinman, L. & Acha-Orbea, H. (1989) Cell 57, 709-715. 5. Davis, M. M. & Bjorkman, P. J. (1988) Nature (London) 334, 395-402. 6. Howell, M. D., Winters, S. T., Olee, T., Powell, H. C., Carlo, D. J. & Brostoff, S. W. (1989) Science 246, 668-670. 7. Vandenbark, A. A., Hashim, G. & Offner, H. (1989) Nature (London) 341, 541-544. 8. Urban, J. L., Kumar, V., Kono, D. H., Gomez, C., Horvath, S. J., Clayton, J., Ando, D. G., Sercarz, E. E. & Hood, L. (1988) Cell 54, 577-597. 9. Wraith, D. C., Smilek, D. E., Mitchell, D. J., Steinman, L. & McDevitt, H. 0. (1989) Cell 59, 247-255. 10. Urban, J. L., Horvath, S. J. & Hood, L. (1989) Cell 59, 257-271. 11. Weber, W. I. J., Vandermeeren, M. P. P. V., Raus, J. C. M. & Buurman, W. A. (1989) Cell. Immunol. 120, 145-153. 12. Chou, Y. K., Vainiene, M., Whitham, R., Bourdette, D., Chou, C. H.-J., Hashin, G., Offner, H. & Vandenbark, A. A. (1989) J. Neurosci. Res. 23, 207-216. 13. Bell, J. I., Denney, D., Jr., Foster, L., Belt, T., Todd, J. A. & McDevitt, H. 0. (1987) Proc. Natl. Acad. Sci. USA 84, 62346238. 14. Schumacher, G. A., Beebe, G. W., Kibler, R. F., Kurland, L. T., Kurtzke, J. F., McDowell, F., Nagler, B., Sibley, W. A., Tourtellotte, W. W. & Welmon, T. L. (1965) Ann. N. Y. Acad. Sci. 122, 552-568.

Proc. Natl. Acad. Sci. USA 87 (1990) 15. Bodmer, J. G., Marsh, S. G. E. & Albert, E. (1990) Immunol. Today 11, 3-10. 16. Wilkinson, D., DeVries, R. R. P., Madrigal, J. A., Lock, C. B., Morgenstern, J. P., Trowsdale, J. & Altman, D. M. (1988) J. Exp. Med. 167, 1442-1448. 17. Tonnelle, C., DeMars, R. & Long, E. 0. (1985) EMBO J. 4, 2839-2847. 18. Eylar, E. H., Kniskern, P. J. & Jackson, J. J. (1979) Methods Enzymol. 32, 323-341. 19. Whitaker, J. N. & Seyer, J. M. (1979) J. Biol. Chem. 254,

6956-6963. 20. Giegerich, G., Pette, M., Fujita, K., Wekerle, H., Epplen, J. T. & Hinkkanen, A. (1990) J. Chromatogr. 528, 79-90. 21. Alvord, E. C. (1984) Experimental Allergic Encephalomyelitis: A Useful Model for Multiple Sclerosis, eds. Alvord, E. C., Kies, M. W. & Suckling, A. J. (Liss, New York), pp. 503-508. 22. Watanabe, R., Wege, H. & TerMeulen, V. (1983) Nature (London) 305, 150-153. 23. Johnson, R. T., Griffin, D. E., Hirsch, R. L., Wolinsky, J. S., Rodenbeck, S., de Soriano, I. L. & Vaisberg, A. (1984) N. Engl. J. Med. 310, 137-141. 24. Hemachudha, T., Griffin, D. E., Giffels, J. J., Johnson, R. T., Moser, A. B. & Phanuphak, P. (1987) N. Engl. J. Med. 316, 369-374. 25. Pette, M., Fujita, K., Kitze, B., Whitaker, J. N., Albert, E., Kappos, L. & Wekerle, H. (1990) Neurology, in press. 26. Bums, J., Rosenzweig, A., Zweiman, B. & Lisak, R. P. (1983) Cell. Immunol. 81, 435-440. 27. Tournier-Lasserve, E., Hashim, G. A. & Bach, M. A. (1988) J. Neurosci. Res. 19, 149-156. 28. Schluesener, H. J. & Wekerle, H. (1989) J. Immunol. 135, 3128-3133. 29. Kumar, V., Kono, D. H., Urban, J. L. & Hood, L. (1989) Annu. Rev. Immunol. 7, 657-682. 30. Acha-Orbea, H., Steinman, L. & McDevitt, H. 0. (1989) Annu. Rev. Immunol. 7, 371-406. 31. Richert, R. R., Robinson, E. D., Deibler, G. E., Martenson, R. E., Dragovic, L. J. & Kies, M. W. (1989) J. Neuroimmunol. 23, 55-66. 32. Sun, D. & Wekerle, H. (1986) Nature (London) 320, 70-72. 33. Ota, K., Matsui, M., Milford, E. L., Mackin, G. A., Weiner, H. L. & Hafler, D. A. (1990) Nature (London) 346, 183-187. 34. Martin, R., Jaraquemada, D., Flerlage, M., Richert, J., Whitaker, J., Long, E. O., McFarlin, D. E. & McFarland, H. F. (1990) J. Immunol. 145, 540-548.