Usage of Vb3.3 T-Cell Receptor by Myelin BasicProtein-Specific Encephalitogenic T-Cell Linesin the Lewis RatAndrew Chan,1 Ralf Gold,1* Gerhard Giegerich,1 Thomas Herrmann,2 Stefan Jung,1Klaus V. Toyka,1 and Hans-Peter Hartung1

1Department of Neurology, Clinical Research Group for Multiple Sclerosis and Neuroimmunology,Julius-Maximilians-University, Wu¨rzburg, Germany2Department of Virology and Immunobiology, Julius-Maximilians-University, Wu¨rzburg, Germany

T-cell receptor (TCR) b-chain usage in experimentalautoimmune encephalomyelitis (EAE) seems to bemuch more heterogeneous than previously assumedeven for a single autoantigen in an inbred animalstrain. Owing to the lack of suitable antibodies, thishas been demonstrated only at the RNA level so far. Tocharacterize further the TCR elements used in theLewis rat for the recognition of the main encephalito-genic region of myelin basic protein (MBP), BALB/cmice were immunized with T-cell hybridomas express-ing non-Vb8.2 TCR specific for guinea pig MBPpeptide aa 68–88. Two B-cell hybridomas (clonesC-A11 and F-D6) producing TCR Vb3.3-specific mono-clonal antibodies were selected. Specificity was demon-strated by RT-PCR and cloning of PCR productsobtained from sorted T-cell lines and naive T cells.MBP-specific Vb3.3 T cells used the L-S motif in theVDJ region, were associated with Va2 or Va8 chains,and specifically recognized MBP peptide aa 68–88.Vb3.3 TCR-positive T cells were detected in all of apanel of six MBP-specific T-cell lines, although to alesser degree than Vb8.2 TCR-positive T cells. Afterintravenous injection of sorted Vb3.3 T cells, animalsdeveloped EAE, and Vb3.3-positive cells were foundby immunocytochemical analyses in the spinal cord.Furthermore, treatment of EAE induced by immuni-zation with MBP was more effective when a combina-tion of anti-V b3.3 and anti-Vb8.2 mAbs was used.These results confirm the functional role of TCRVb3.3 and thus underscore the heterogeneity of TCRusage in MBP-associated autoimmunity. J. Neurosci.Res. 58:214–225, 1999.r 1999 Wiley-Liss, Inc.

Key words: autoimmunity; multiple sclerosis; T-cellreceptor variable region

INTRODUCTIONExperimental autoimmune encephalomyelitis (EAE)

is an inflammatory disease of the central nervous system

(CNS) mediated by autoantigen-specific CD4-positive Tcells. EAE is widely investigated as an animal model forsome aspects of multiple sclerosis (MS). In susceptibleanimals, EAE can either be actively induced by immuni-zation with autologous or heterologous CNS proteins orby the adoptive transfer of activated, autoantigen-specificT cells (Martin et al., 1992). Myelin basic protein (MBP)is a major CNS autoantigen, and most of the encephalito-genic MBP-reactive T lymphocytes in the Lewis ratrecognize a dominant encephalitogenic MBP epitopespanning amino acids (aa) 68–88 of guinea pig MBP,complexed to RT1.Bl class II MHC molecules (Martin etal., 1992).

MBP-specific, encephalitogenic T cells from Lewisrats and different mouse strains were initially reported touse a highly restricted repertoire of T-cell receptor (TCR)variable genes, preferentially the TCR Vb8.2 element,often in combination with the Va2 variable region (Burnset al., 1989; Chluba et al., 1989). In addition, a markedconservation of the first two amino acid residues (asparticacid-serine, DS motif) of the junctional complementarydetermining region 3 (CDR3) associated with the Vb8.2receptors has been described with few or no nongerm-linemutations (Gold et al., 1991; Hafler et al., 1996). Conse-quently, the paradigm of a restricted, pathogenic T-cellresponse in EAE and the human disease MS has led to anumber of immunotherapeutic strategies targeted at the

Contract grant sponsor: University of Wu¨rzburg and the Bundesminis-terium fur Bildung und Forschung (Interdisziplina¨res Zentrum fu¨rKlinische Forschung, Wu¨rzburg); Contract grant number: project C2.

Gerhard Giegerich’s current address is Department of Neurology,University of Regensburg, D-93053 Regensburg, Germany.

Hans-Peter Hartung’s current address is Department of Neurology,Karl-Franzens-University, A-8036 Graz, Austria.

*Correspondence to: Ralf Gold, MD, Neurologische Universita¨tsklinik,Josef-Schneider-Straße 11, D-97080 Wu¨rzburg, Germany.E-mail: r.gold@mail.uni-wuerzburg.de

Received 29 April 1999; Revised 1 June 1999; Accepted 4 June 1999

Journal of Neuroscience Research 58:214–225 (1999)

r 1999 Wiley-Liss, Inc.

TCR, e.g., using antibodies or T-cell/TCR vaccination(Hohlfeld, 1997). The successful modulation of MBP-induced EAE by Vb8.2 TCR-directed therapy (Howell etal., 1989; Offner et al., 1991; Imrich et al., 1995) furthersupported an important pathogenic role of T cells withsuch biased TCR element usage, at least in certain stagesof the disease.

However, recent data suggested a more diversespectrum of TCR usage in Lewis rat EAE: TCRb-chainusage of different encephalitogenic T-cell lines specificfor the dominant MBP epitope (aa 68–88) was notrestricted to the Vb8.2 element, and also non-Vb8-bearing cells were highly encephalitogenic (Sun et al.,1993). Furthermore, TCLs isolated directly from brainsof Lewis rats during the early invasive phase of EAEexhibited a broad spectrum of TCR Vb elements (Sun etal., 1994), which could also be demonstrated during latestages of EAE (Karin et al., 1993; Offner et al., 1993).This was shown at the RNA level using reverse transcrip-tase-polymerase chain reaction (RT-PCR). Finally, en-cephalitogenicity of TCR Vb8.2-negative, MBP-aa-68–88-specific T cells even in Lewis rats that had beendepleted of host-derived Vb8.2 T cells confirmed thatVb8.2-positive cells are not essential for the inductionand progression of the disease (Gold et al., 1995).

Because heterogeneous TCR element usage hasbeen demonstrated only at the RNA level so far, we setout to characterize non-Vb8.2 TCR T cells in the Lewisrat at the protein level and to study their functionalrelevance. We have raised monoclonal antibodies (mAbs)by immunization with non-Vb8.2 T-cell hybridomasgenerated from encephalitogenic T-cell lines and identi-fied TCR Vb3.3 as the recognition element. Encephalito-genicity of sorted Vb3.3 TCR-positive cells and therapeu-tic modulation of active MBP-induced EAE by theanti-Vb3.3 mAb argues for the pathogenic importance ofthis recognition element and confirms heterogeneousTCR usage of MBP-specific T cells even in an inbredrodent strain.

MATERIALS AND METHODSReagents and Antibodies

All tissue culture supplements were purchased fromGibco-BRL (Eggenstein, Germany), with exception ofbovine serum albumin (BSA) fraction V (Roth, Karls-ruhe, Germany) and concanavalin A (Sigma, Deisen-hofen, Germany). MAb R78 (IgG1) specific for Vb8.2TCR and mAb R73 (IgG1) recognizing thea/b TCRwere kindly provided by Dr. Hu¨nig (University of Wurz-burg; Hunig et al., 1989; Imrich et al., 1995). IgG1 andIgG2a antibody isotype controls were obtained fromSigma and phycoerythrin (PE)-labeled donkey anti-mouse Ig from Dianova (Hamburg, Germany). MBP was

prepared as described previously (Eylar et al., 1979). Allreagents used for molecular biology were obtained fromRoche Diagnostics Boehringer Mannheim (Mannheim,Germany), Gibco-BRL, Merck (Darmstadt, Germany),Pharmacia-BRL (Freiburg, Germany), Promega (Heidel-berg, Germany), Serva (Heidelberg, Germany), or Sigma.

T-Cell Lines and T-Cell HybridomasMBP-specific CD41 rat T-cell lines were estab-

lished from lymph nodes of female Lewis rats immunizedin the hind footpad with 100 µg of MBP emulsified incomplete Freund’s adjuvant (CFA; Difco, Detroit, MI) asdescribed before (Gold et al., 1995). Antigen-specific Tcells were selected by repeated cycles of propagation inmedium with 7.5% supernatant of concanavalin A-treatedmouse spleen cells and 5% fetal calf serum (FCS),followed by antigen-specific restimulation using irradi-ated thymic antigen-presenting cells. Cells expressing therespective TCR Vb genes were enriched or depletedusing the magnetic cell sorting MACS system accordingto the manufacturer’s instructions (Miltenyi Biotech,Bergisch-Gladbach, Germany; Miltenyi et al., 1990). Thegeneration of the Vb8.2/Vb10-depleted, MACS-sortedMBP-specific T-cell line MBP 13 has been reportedpreviously (Gold et al., 1995). Sorted T cells were thenpropagated in interleukin 2 (IL-2)-containing medium. Ifnecessary, cells were subjected to repeated sorting. T-cellhybridomas were generated by fusion of activated T-cellblasts from T-cell lines with a variant of the mousethymoma BW5147 (BW1100.129.237; Torres-Nagel etal., 1993) and screened for TCR expression by immuno-fluorocytometric analysis.

For proliferation studies, resting T cells of unsortedor sorted T-cell lines were added at 23 104 /well with 106

irradiated thymic antigen-presenting cells and antigen in100 µl medium to 96-well round bottom microtiter plates(Nunc). Triplicate cultures were maintained for 48 hr(37°C, humidified atmosphere with 5% CO2/95% air) andharvested after a 16-hr pulse with 0.2 µCi/well [3H]thymi-dine (Amersham-Buchler, Braunschweig, Germany). Cellswere collected on glass fiber filters with a Betaplate96-well harvester (Pharmacia-LKB), and radioactivityassociated with the dried filter was quantified using a96-well Betaplate liquid scintillation counter (Pharmacia-LKB).

Generation of Monoclonal Antibodies (mAbs)MAbs directed to the rat TCR variable region were

generated as described before (Torres-Nagel et al., 1993;Stienekemeier et al., 1999). Briefly, T-cell hybridomasderived from Vb8.2/10-depleted encephalitogenic MBP-specific T-cell lines (Gold et al., 1995) were used forimmunization of BALB/c mice (Charles River, Sulzfeld,

T-Cell Receptor Heterogeneity in EAE 215

Germany). After five weekly intraperitoneal injectionsof 1 3 107 cells, mice were rested for 3 weeks andchallenged i.v. 3 days before polyethylene glycol-mediated fusion of spleen cells with the Ag8 nonproducermyeloma cell line. Supernatants of hybridomas obtainedafter selection in HAT-containing medium were screened

for binding to encephalitogenic T-cell lines by cytofluoro-metric analysis. For determination of mAb isotypes, themouse monoclonal antibody isotyping kit was used(Roche Diagnostics Boehringer Mannheim). For anti-body therapy studies, B-cell hybridomas were cultured inRPMI 1640 (2.5%g-globulin-free FCS, 2 mM glutamine,

Fig. 1. a: Representative two-color flow cytometric analysis oflymph node cells of a naive Lewis rat. Double staining usingmAb R78 (Vb8.2-specific) or the new mAbs C-A11/F-D6(FL2) and T-cell subset-specific antibodies (FL1) as indicated.Percentages of labeled cells as defined by quandrant markersafter subtraction of the respective isotype-specific controls.

b: Immunoprecipitation of T-cell receptor by mAb C-A11. Theprecipitated probe from T-cell hybridoma MBP-H1 was submit-ted to a reducing 12.5% SDS-polyacrylamide gel. Western blotanalysis was performed using the biotinylated rata/b TCR-specific mAb R73 for detection. Relative molecular weight(kDa) as indicated by the respective markers.

216 Chan et al.

100 units/ml penicillin, 100 µg/ml streptomycin, 0.12%glucose). MAbs were purified from supernatants onProtein A-Sepharose following standard procedures.

Immunocytochemical StudiesFlow cytofluorometric analysis (FACScan, Becton

Dickinson, San Jose, CA) of TCR expression was per-formed as described elsewhere (Jung et al., 1992) usingthe CellQuest or LYSIS II software (Becton Dickinson).For FACS analysis of naive T-lymphocytes, single-cellsuspensions of popliteal and mesenterial lymph nodesfrom naive Lewis rats were prepared. Briefly, cells werestained using a primary mouse mAb and PE-labeledsecondary donkey anti-mouse Ig, followed by anotherprimary mouse FITC-labeled mAbs. For three-color flowcytometric analyses, biotin-labeled mAbs were added anddetected by Streptavidin 670 (Life Technologies, Egger-stein, Germany). In some cases PE-conjugated primarymAbs were taken. To avoid capture of primary mAb byPE-labeled secondary donkey anti-mouse Ig, normalmouse Ig was used in excess to block potential freebinding capacity on the secondary antibody. Light scatter

gates were set to include all viable nucleated cells, andmore than 40,000 events were analyzed.

For histological analysis, various segments of thelumbar spinal cord were removed and processed asdescribed elsewhere (Zettl et al., 1995). Cellular infil-trates were characterized by incubation with mAb R73(Hunig et al., 1989), R78 (Imrich et al., 1995), or thenewly generated antibody C-A11 at 4°C overnight andvisualized using the PAP technique and DAB as substrateas described elsewhere (Zettl et al., 1995). T-cell hybrid-oma MBP-H1 was used for immunoprecipitation of theT-cell receptor by the new mAb C-A11; 83 107 cellswere treated with lysis buffer (1% Nonidet P40/1 mMPMSF/PBS), and immunoprecipitation was performedaccording to the manufacturer’s instructions (Dynal,Hamburg, Germany) and Stienekemeier et al. (1999). Theimmunoprecipitate was then electrophoresed on a 12.5%SDS-polyacrylamide gel. Precipitated TCR were ana-lyzed by semidry Western blotting using the biotinylatedrat a/b TCR-specific mAb R73 (1:100; Hu¨nig et al.,1989) and detected using streptavidin-conjugated alka-line phosphatase (Dako, Hamburg, Germany) with NBT/BCIP as chromogenic substrate (Roche Diagnostics Boeh-ringer Mannheim).

TCR Sequence Analysis by RT-PCR and DNASequencing

Extraction of total RNA was performed using theRoti-Quick kit (Roth, Karlsruhe, Germany) according tothe manufacturer’s instructions. Approximately 20 µg oftotal RNA was reverse transcribed into first-strand cDNAby using NotI-d(T) 18 primers and a cDNA synthesis kit(Pharmacia-LKB). The cDNA was used at a dilution of1:50 for enzymatic amplification with specific TCRV-element primers and a common constant region Ca orCb primer. The primers used detect the known rat Va andVb families in lymph node cells of Lewis rats (Gold et al.,1995; Stienekemeier et al., 1999). The PCR mixtureconsisted of 2.0 µl of diluted cDNA, 5.0 µl of 103 PCRbuffer (Applied Biosystems, Weiterstadt, Germany), 1.25U Taq polymerase (Applied Biosystems), 10 nmol dNTP(Pharmacia), and 25 pmol of the respective TCR C- andV-element primers at a final volume of 50 µl H2O.Amplifications were performed with a model 9600 ther-mocycler (Perkin-Elmer, Langen, Germany), with dena-turation at 95°C (40 sec), annealing at 55°C (120 sec),and extension at 72°C (60 sec) over five cycles. This wasfollowed by denaturation at 95°C (60 sec), annealing at55°C (20 sec), and extension at 72°C (60 s) over 20 or 30cycles with a prolonged 10-min extension during the lastcycle. Reaction products purified by QIAquick spin PCRpurification kits (Qiagen, Hilden, Germany) were cloned

b

Figure 1. (Continued.)

T-Cell Receptor Heterogeneity in EAE 217

into PCRscript vectors (Stratagene) and sequenced (T7sequencing kit; Pharmacia-LKB).

Animals, Induction of EAE, and Therapy StudiesAll experiments were conducted according to Bavar-

ian state regulations for animal experimentation andapproved by the local authorities. Adoptive transfer(AT)-EAE was induced in 4–6- week-old Lewis rats(Charles River) by intravenous injection of 53 106

freshly activated MBP-specific T-cell blasts or by intrave-nous injection of 107 sorted, Vb3.3-positive T cells. Ratswere weighed daily, and clinical disease severity wasassessed by using a scale from 0 to 5 (Gold et al., 1995).Clinical scores used were: 0, normal; 1, limp tail; 2, mildparaparesis of the hind limbs, ataxia; 3, moderate parapa-resis, voluntary movements still possible; 4, para- ortetraplegia; 5, moribund. Active EAE was induced byinoculation of hind footpads with 75 µg MBP in anemulsion of equal volumes of saline and CFA (1 mg/ml;Difco). For therapeutic studies in active EAE, animalswere injected intravenously (tail vein) beginning on day 8after disease induction before clinical symptoms orweight loss had emerged. Treatment groups consisting ofsix weight- and age-matched male animals each receivedeither the combination therapy with anti-Vb8.2/Vb3.3antibodies [mAb R78 (IgG1)/CA11 (IgG2a), 250 µg eachin 500 µl PBS] or anti-Vb8.2 monotherapy [mAb R78/isotype control IgG2a, 250 µg each]. Control animalswere injected with the respective isotype controls (IgG1/IgG2a, 250 µg each). For therapeutic studies on Vb8.2-depleted animals, neonatal Lewis rats were injected i.p.biweekly with mAb R78 (50 µg in 100 µl PBS) or therespective isotype control beginning 2 days postpartum.After 21 days postpartum the dose was doubled, andinjections were stopped on day 49 post partum concomi-tantly with the induction of active MBP-EAE (75 µgMBP in saline/CFA). On day 8, weight-matched animals(n 5 23) received either 250 µg mAb C-A11 in 250 µlPBS or the respective isotype control i.v. Statisticalanalysis of disease courses and body weights was per-formed using the two-way ANOVA test (GraphPadSoftware Inc., San Diego, CA).

RESULTSGeneration and Characterization of mAbs C-A11and F-D6 on Naive T Cells and T-Cell Hybridomas

We first generated monoclonal antibodies to Vb3.3TCR by immunization of BALB/c mice with non-Vb8.2TCR-expressing T-cell hybridomas that recognized themain encephalitogenic MBP peptide aa 68–88 (Gold etal., 1995). Two clones (F-D6 and C-A11, both IgG2aisotype) recognizing the TCR of non-Vb8.2 T cells in a

stoichiometric manner were selected by immunofluoro-metric analysis with double staining using the rata/bTCR-specific mAb R73 (Hu¨nig et al., 1989). Figure 1adepicts a representative cytofluorometric analysis oflymph node cells from a naive Lewis rat. Quantitativeanalysis of R731 (a/b TCR) CD4/CD8 T cells wasperformed on three Lewis rats. Analysis with the Vb8.2-specific mAb R78 resulted in proportions of 5.5%60.2% R781/CD41 and 4.0%6 0.6% R781/CD81 cells(mean6 SD). The new mAb C-A11 detected frequenciesof 4.0% 6 0.3% C-A111/CD41 and 2.4% 6 0.2%C-A11

1/CD81 cells (mAb F-D6: 3.6%6 0.4% F-D61/

CD41, 2.3%6 0.1% F-D61/CD81 cells).The T-cell hybridoma MBP-H1 specific for guinea

pig MBP was used for immunoprecipitation studies (Fig.1b). Specific binding of rata/b TCR by the new mAb wasconfirmed by immunoprecipitation of the TCR with mAbC-A11 and detection in Western blot analysis using thea/b TCR-specific mAb R73. Results were confirmed at alater time using the Vb4 idiotype-specific mAb 8G8 onthe T-cell hybridoma P2.48.1 as a positive control(Stienekemeier et al., 1999). Although mAb F-D6 wassuitable only for cytological studies, mAb C-A11 alsostained Vb3.3 TCR-positive T cells in cryosections ofsplenic tissue of naive Lewis rats (Fig. 2C) and spinalcord of animals with AT-EAE that was passively inducedby injection of mAb C-A11 sorted T cells (Fig. 2D).

RT-PCR Determination of Vb3 TCR Usage ofLymph Node Cells Activated by the New mAbsC-A11 and F-D6

The new mAbs C-A11 and F-D6 were immobilizedon plastic dishes and used to activate lymph node cells ofnaive animals. T-cell blasts were then purified by gradientcentrifugation and expanded in IL-2-enriched medium.After RNA extraction, TCR usage was characterized byRT-PCR using a panel of primers that amplify all knownrat Vb and Va chains as described previously (Gold et al.,1995; Stienekemeier et al., 1999). RT-PCR on cellsactivated with mAb C-A11 or F-D6 displayed only aspecific Vb3 signal compared to RT-PCR from the samecell preparation before activation (Fig. 3; only mAbC-A11 is shown). These findings indicate that both mAbsC-A11 and F-D6 recognized the Vb3 TCR.

Specific Recognition of TCR on EncephalitogenicT-Cell Lines by mAbs C-A11 and F-D6

To determine further the specific usage of TCRchains as recognized by the new mAbs, a panel of sixencephalitogenic T-cell lines was investigated by cytofluo-rometric analysis (Fig. 4). In early stages of antigen-specific activation cycles, up to 20% of the T cells werestained by mAb C-A11. In some cell lines, the frequency

218 Chan et al.

of TCR usage as detected with the new mAb almostreached the level of Vb8.2-positive cells (Fig. 4). Usingguinea pig MBP as antigen, a selection towards Vb8.2TCR could be observed in later stages of propagation(data not shown). In other nonencephalitogenic, antigen-specific cell lines (P2 neuritogenic protein, ovalbumin,PPD), the new mAbs consistently recognized,5% of Tcells (Stienekemeier et al., 1999; data not shown).

MAb-Sorted, Vb3-Positive MBP-Specific T-CellLines Are Highly Encephalitogenic and Recognizethe Immunodominant, MHC RT1.B l -RestrictedMBP-68–88 Epitope

To characterize Vb3 TCR usage further, cell lineswith a high proportion of Vb3 T cells were sorted to.98% homogeneity by repetitive MACS sorting andantigen-specific or mitogen activation as described previ-

ously (Gold et al., 1995). Two of our T-cell lines, MBP13and LK3, showed a strong proliferative response toguinea pig MBP (Fig. 5). Encephalitogenicity of theseVb3-positively sorted T-cell lines was proven by induc-tion of AT-EAE in Lewis rats. Injection of 107 sortedMBP13 or LK3 cells resulted in severe disease starting onday 3, with a tetraplegia in all (n5 4) animals on day 4and a weight loss of 20% (data not shown). A severeinfiltration of the lumbar spinal cord bya/b T lympho-cytes was observed on days 3 and 4. As is depicted inFigure 2D, Vb3-positive cells could be demonstrated inspinal cord of these animals using mAb C-A11.

We next tested the antigen specificity of the Vb3-positively sorted, encephalitogenic T-cell lines. Reactiv-ity towards different MBP peptides that are recognized byvarious encephalitogenic T-cell lines (Chluba et al., 1989)or that dominate the immune response to MBP in naive

Fig. 2. Representative serial sections of splenic tissue of a naive Lewis rat stained for thea/bTCR (mAb R73;A), Vb8.2 TCR (mAb R78;B), and the new mAb C-A11 (C). Scale bar in A5100 µm for A–C. Note the weaker staining signal with mAb C-A11 and thus the relativequantitative underestimation of labeled cells compared to FACS analysis.D: Spinal cord of aLewis rat with AT-EAE after passive transfer of mAb C-A11-sorted T cells. Immunohistochem-istry using mAb C-A11. Scale bar5 25 µm.

T-Cell Receptor Heterogeneity in EAE 219

rats (Mor and Cohen, 1993) was investigated in T-cellproliferation assays (Fig. 5). Among the sorted T-celllines, only the encephalitogenic lines LK3 and MBP13recognized the dominant encephalitogenic, MHC RT1.Bl-restricted epitope aa 68–88 of guinea pig MBP. Prolifera-tion of these cell lines could be inhibited by mAb OX6directed against a determinant on the RT1.Bl molecule(data not shown). T-cell line MBP14 did not exhibitsignificant reactivity to MBP after sorting and could bepropagated only by mitogen. MAbs C-A11 and F-D6adsorbed to tissue culture plastic also activated sortedT-cell lines (data not shown).

We then sequenced the TCR variable regions of theVb3-positively sorted encephalitogenic cell lines LK3and MBP13 and the nonencephalitogenic cell line MBP14.Cloning and subsequent sequencing of PCR motifsdemonstrated the exclusive usage of the Vb3.3 TCRelement (Fig. 6). Theseb chains were associated withVa2 or Va8 elements (data not shown). Interestingly, celllines LK3 and MBP13 both used the L-S motif as the firsttwo amino acid residues of the junctional complementar-ity determining region 3 (CDR3).

Combined Preventive Anti-Vb3.3/Vb8.2 Therapy IsSuperior to Anti-V b8.2 Monotherapy in ActiveMBP-Induced EAE

To substantiate the pathogenic role of Vb3.3 TCR-positive T cells in EAE actively induced with MBP, threegroups (n5 6) of age- and weight-matched male animalsreceived either a combination therapy with mAb C-A11(anti-Vb3.3, IgG2a)/R78 (anti-Vb8.2, IgG1), R78 mono-therapy/IgG2a, or the respective isotype controls (IgG1/IgG2a). Preventive treatment was started on day 8, before

first signs of the disease had emerged. As is shown inFigure 7a, disease severity could be ameliorated inanimals treated with the anti-Vb8.2 monotherapy com-pared to the control animals (P , 0.0001, ANOVA test),as has been reported (Imrich et al., 1995). The combina-tion anti-Vb3.3/Vb8.2 proved to be superior to theanti-Vb8.2 monotherapy with regard to mean clinicalscores (Fig. 7a;P , 0.05) and loss of body weight (Fig.7b, P , 0.01). Furthermore, although two animals eachdied in the isotype control and the anti-Vb8.2 mono-therapy group, only one animal of the Vb8.2/Vb3.3combination therapy group succumbed at the peak of thedisease. To test whether anti-Vb3.3 therapy could confercomplete resistance to MBP-EAE in animals devoid ofVb8.2 TCR, we depleted neonatal Lewis rats of Vb8.2-positive T cells by biweekly injections with mAb R78(Imrich et al., 1995; Gold et al., 1995). Administration ofanti-Vb3.3 mAb C-A11 after immunization with MBPdid not lead to a significant reduction of disease incidenceand severity in these Vb8.2 TCR-depleted rats comparedto the IgG2a-treated control group (data not shown).

DISCUSSIONPCR analysis of inflamed tissue and the generation

of T-cell hybridomas may not reflect the actual frequencyof individual T-cell subsets in a heterogeneous T-cellpopulation, so we have raised the new mAbs C-A11 andF-D6 to characterize functionally the TCR repertoire ofMBP-specific, encephalitogenic T cells on the proteinlevel. Exclusive usage of the Vb3.3 element could bedemonstrated on naive and mAb-sorted, MBP-68–88-specific, encephalitogenic T-cell lines. Furthermore, theproportion of Vb3.3-positive T cells increased after

Fig. 3. DNA gel electrophoresis of RT-PCR fragments. RT-PCR was performed after RNAextraction from naive lymph node cells activated by immobilized mAbs C-A11 or F-D6 (onlymAb C-A11 is shown).Top depicts RT-PCR fragments of lymph node cells before activation.Bottom shows the results after activation by immobilized mAb C-A11, where only a specificVb3 signal is observed. Vb-specific primers are indicated on top of each lane. DNA sizemarkers (in bp) are shown on the left.

220 Chan et al.

immunization with MBP. A broad panel of MBP-specificpathogenic T-cell lines displayed varying degrees ofVb3.3 usage. An active pathogenic role of Vb3.3 TCR-positive cells was also demonstrated by the high encepha-litogenicity of mAb C-A11-sorted T cells that wereidentified in EAE spinal cord by immunohistochemistry.The higher efficacy of preventive combined anti-Vb8.2/Vb3.3 immunotherapy in actively induced EAE supports

a causative role of Vb3.3 immune response and arguesagainst a unique role of Vb8.2 in the induction ofMBP-EAE.

These results extend and corroborate our previouswork on encephalitogenicity of a MBP-68–88-specificT-cell line depleted of TCR Vb8.2 and Vb10 T cells evenin Lewis rats devoid of host-derived Vb8.2 T cells (Goldet al., 1995). T-cell hybridomas generated from this cell

Fig. 4. FACS analysis of TCR usage by a panel of encephalitogenic, MBP-specific T-cell lines.T-cell lines stained fora/b TCR vs. isotype control (left) TCR Vb8.2 (middle), or mAb C-A11(right ). Some T-cell lines were followed during different cycles of restimulation (R1–6) invitro. Note that in some T-cell lines [MBP8 (R2), MBP21 (R4)] the amount of cells stained bymAb C-A11 almost reached the level of Vb8.2-positive T cells.

T-Cell Receptor Heterogeneity in EAE 221

line showed a predominant usage of the TCR Vb3.3element. Moreover, among several other Vb elements,Vb3.3 TCR could be detected in early spinal cord lesionsof these animals by RT-PCR. However, using this ap-proach, preferential attraction of Vb3.3 bystander inflam-matory cells to the lesion could not be excluded (Cross etal., 1990). Encephalitogenicity of our Vb3.3-positivelysorted MBP-68–88-specific T cells clearly argues againstthis possibility and proves a pathogenic role of thisreceptor element in Lewis rat EAE.

Notably, both Vb3.3-positively sorted, MBP-68–88-specific encephalitogenic T-cell lines MBP13 and LK3used the L-S motif in the first two amino acid residues ofthe junctional complementarity determining region 3(CDR3), which is most intimately linked with antigenicpeptide recognition (Garcia et al., 1996). Conversely,T-cell line MBP 14, which did no longer show significantreactivity to MBP after Vb3.3 enrichment, exhibited adifferent VDJ sequence. The L-S sequence bears similari-ties to the conserved D-S motif in the CDR3 associated

Fig. 5. Antigen specificity and mitogen activation of sorted, Vb3-positive T-cell lines. T-celllines were activated by syngeneic thymic antigen-presenting cells in the presence of antigens, asindicated. The y axis gives the mean cpm6 SD of triplicate wells (logarithmic scale). Note thatT-cell line MBP13 was already in a later restimulation cycle and showed only a weak overallresponse. Counts,1,000 were not considered significant in our system. pep. AA: MBP-peptidecomprising amino acids 44–67, 68–88, or 87–99, respectively. ConA, concanavalin A.

Fig. 6. Nucleic acid and predicted amino acid sequences of rat TCR chains in the Vb3-sorted,MBP-68–88-reactive, encephalitogenic T-cell lines MBP13 and LK3 and the nonencephalito-genic T-cell line MBP14. Db elements are underscored.

222 Chan et al.

with Vb8.2 receptors in Lewis rats (Gold et al., 1991).Moreover, data of Mor et al. (1996) indicate that severalother amino acids can be found in the first position of theDS consensus motif. These data could indicate similarconserved CDR3 elements in both Vb8.2 and Vb3.3 TCRreactive to MBP.

We observed a preferential coexpression of Vb3.3with Va2/Va8-recognition elements in our Vb3.3-positively sorted, encephalitogenic T-cell lines. This biastowards Vb3.3/Va2 association had also been notedbefore in T-cell hybridomas generated from the Vb8.2/Vb10-depleted cell line of our previous study (Gold et al.,1995). Insofar as a predominant association also of Vb8.2with Va2 has been reported previously (Chluba et al.,1989; Burns et al., 1989), it is tempting to hypothesizethat the immune response against MBP in the Lewis ratmay be restricted to specialized Va elements rather thanVb TCR chains. A similar bias towards predominant Vausage irrespective of the associated Vb chain has been

noted in P2-specific neuritogenic T cells in Lewis ratexperimental autoimmune neuritis, where TCR Vb4 andVb13 elements were frequently associated with Va11(Stienekemeier et al., 1999; F.X. Weilbach and Giegerich,unpublished data).

In naive animals Vb8.2-positive T cells make upabout 5% of the T-cell population in peripheral blood andlymph nodes (Tsuchida et al., 1993; Smith et al., 1992),whereas determination of TCR usage in peripheral lym-phoid tissues using the new mAb C-A11 demonstratedbetween 3–4% Vb3.3-positive cells. This quantitativedifference may, at least in part, explain the overrepresen-tation and thus the earlier recognition of Vb8.2-positiveT-cell lines generated ex vivo; the protocols used for theraising of T-cell lines may select for the most dominantresponse (Mor et al., 1996; Sun et al., 1993). Indeed,several other lines of evidence suggest rather heteroge-neous TCR Vb usage in EAE also in vivo: Althoughanti-Vb8.2 immunotherapy in Lewis rats reportedly ledto a strong reduction both in incidence and severity ofEAE symptoms, a significant number of the treatedanimals remained susceptible with severe clinical symp-toms, indicating the recruitment of non-Vb8.2 encephali-togenic cells (Imrich et al., 1995). Susceptibility todisease induction in these anti-Vb8.2 treated rats wasmost pronounced in male animals. Although the basis ofthis gender bias is not understood, possible explanationsinclude the influence of the neuroendocrine system onTh1/Th2 balance (Imrich et al., 1995; Rook et al., 1994).Therefore, to select for these ‘‘nonresponders’’ and toavoid an additional heterogeneity in the sensitivity toEAE induction or prevention, our therapy studies were allconducted with male animals. The higher efficacy of thecombined anti-Vb8.2/Vb3.3 therapy argues for a contri-bution also of Vb3.3-positive T cells in MBP-inducedEAE, yet, even with the combination of anti-Vb8.2/Vb3.3,one animal succumbed, and, though significant, ameliorationof clinical disease severity was only partial. Moreover,neonatal depletion of Vb8.2-positive T cells by biweeklyinjections of anti-Vb8.2 mAb R78 followed by anti-Vb3.3immunotherapy (mAb C-A11) after immunization with MBPdid not confer complete resistance to EAE induction.Thus, these in vivo data suggest an even broader reper-toire of TCRb-chain usage in addition to Vb8.2/Vb3.3 inactive MBP-EAE and are in line with RT-PCR datademonstrating the expression of various TCRb elementsin EAE lesions (Sun et al., 1994; Gold et al., 1995).

In summary, our data support the notion that thepathogenic T-cell repertoire in terms of epitope specific-ity, TCR Vb composition, and VDJ sequence in Lewis ratEAE is less restricted than was previously assumed(Wilson et al., 1993). The seemingly biased immuneresponse, summarized in the ‘‘V-region disease hypoth-esis,’’ provided the basis for T cell-specific therapies such

Fig. 7. Anti-Vb8.2 vs. combined anti-Vb8.2/Vb3.3 treatmentin active MBP-EAE. Disease course (a) and body weight (b) ofLewis rats treated with anti-Vb8.2 monotherapy, combinedanti-Vb8.2/Vb3.3 therapy, or the respective isotype controlsIgG1 and IgG2a. Groups consisted of six male animals each.Animals were treated intravenously on day 8, before anyclinical symptoms had emerged. One (1) or two (11) animalssuccumbed. In comparison to the isotype controls, anti-Vb8.2monotherapy was effective with regard to mean clinical scores(a;P , 0.0001, ANOVA test) and body weight (b;P , 0.0001).Combined anti-Vb3.3/Vb8.2 therapy proved to be superior toanti-Vb8.2 monotherapy [mean clinical scores (a),P , 0.05;body weight (b),P , 0.01].

T-Cell Receptor Heterogeneity in EAE 223

as the use of antagonist peptides, TCR peptides, T-cellclones, and anti-Vb-mAb (Ben-Nun and Cohen, 1981;Howell et al., 1989; Vandenbaark et al., 1989; Kumar etal., 1995; Imrich et al., 1995). Our data demonstrate that,among others, Vb3.3 is a recognition element used byencephalitogenic, MBP-68–88-specific T-cell lines in thecontext of RT1Bl MHC restriction. Thus, the heteroge-neous composition of potential idiotypic target structureson pathogenic T cells specific for a single autoantigeneven in an inbred animal strain discourages overly highexpectations for immunotherapies with anti-TCR specific-ity (Vandenbark et al., 1996; Gold et al., 1997; Hohlfeld,1997; Zipp et al., 1998). In the future, antigen-specificapproaches may turn out to be a more promising strategyin the development of specific immunotherapies for EAEand MS (Nicholson et al., 1995; Weishaupt et al., 1997;Liblau et al., 1997).

ACKNOWLEDGMENTSWe gratefully acknowledge the excellent technical

assistance of Alexandra Bunz and Helga Bru¨nner. We areindebted to Dr. T. Hu¨nig for helpful discussions and forproviding mAb R73 and mAb R78.

REFERENCES

Ben-NunA, Cohen IR. 1981. Vaccination against autoimmune encepha-lomyelitis (EAE): attenuated autoimmune T lymphoctes conferresistance to induction of active EAE but not to EAE mediatedby the intact T lymphocyte line. Eur J Immunol 11:949–952.

Burns FR, Li X, Shen N, Offner H, Chou YK, Vandenbark AA,Heber-Katz E. 1989. Both rat and mouse T cell receptorsspecific for the encephalitogenic determinant of myelin basicprotein use similar Va and Vb chain genes even though themajor histocompatibility complex and encephalitogenic determi-nants being recognized are different . J Exp Med 169:27–39.

Chluba J, Steeg C, Becker A, Wekerle H, Eppplen JT. 1989. T cellreceptorb chain usage in myelin basic protein-specific rat Tlymphocytes. Eur J Immunol 19:279–284.

Cross AH, Cannella B, Brosnan CF, Raine CS. 1990. Homing to centralnervous system vasculature by antigen-specific lymphocytes. I.Localization of 14C-labeled cells during acute, chronic, andrelapsing experimental allergic encephalomyelitis. Lab Invest63:162–170.

Eylar EH, Kniskern PJ, Jackson JJ. 1979. Myelin basic proteins.Methods Enzymol 31:323–341.

Garcia KC, Degano M, Stanfield RL, Brunmark A, Jackson MR,Peterson PA, Teyton L, Wilson IA. 1996. Anab Tcell receptorstructure at 2.5 A and its orientation in the TCR-MHC complex.Science 274:209–219.

Gold DP, Offner H, Sun D, Wiley S, Vandenbark AA, Wilson DB. 1991.Analysis of T cell receptorb chains in Lewis rats withexperimental allergic encephalomyelitis: conserved complemen-tarity determining region 3. J Exp Med 174:1467–1476.

Gold DP, Smith RA, Golding AB, Morgan EE, Dafashy T, Nelson J.1997. Results of a phase I clinical trial of a T-cell receptorvaccine in patients with multiple sclerosis. II. Comparativeanalysis of TCR utilization in CSF T-cell populations before and

after vaccination with a TCRVb6 CDR2 peptide. J Neuroimmu-nol 76:29–38.

Gold R, Giegerich G, Hartung H-P, Toyka KV. 1995. T-cell receptor(TCR) usage in Lewis rat experimental autoimmune encephalo-myelitis: TCR b-chain-variable region Vb8.2-positive T cellsare not essential for induction and course of disease. Proc NatlAcad Sci USA 92:5850–5854.

Hafler DA, Saadeh MG, Kuchroo VK, Milford E, Steinman L. 1996.TCR usage in human and experimental demyelinating disease.Immunol Today 17:152–159.

Hohlfeld R. 1997. Biotechnological agents for the immunotherapy ofmultiple sclerosis. Brain 120:865–916.

Howell MD, Winters ST, Olee T, Powell HC, Carlo DJ, Brostoff SW.1989. Vaccination against experimental allergic encephalomyeli-tis with T cell receptor peptides. Science 246:668–670.

Hunig T, Wallny H-J, Hartley JK, Lawetzky A, Tiefenthaler G. 1989. Amonoclonal antibody to a constant determinant of the rat T cellantigen receptor that induces T cell activation. J Exp Med169:73–86.

Imrich H, Kugler C, Torres-Nagel N, Do¨rries R, Hunig T. 1995.Prevention and treatment of Lewis rat experimental allergicencephalomyelitis with a monoclonal antibody to the T cellreceptor Vb8.2 segment. Eur J Immunol 25:1960–1964.

Jung S, Kra¨mer S, Schluesener HJ, Hu¨nig T, Toyka K, Hartung HP.1992. Prevention and therapy of experimental autoimmuneneuritis by an antibody against T cell receptors-a/b. J Immunol12:3768–3775.

Karin N, Szafer F, Mitchell D, Gold DP, Steinman L. 1993. Selectiveand nonselective stages in homing of T lymphocytes to thecentral nervous system during experimental allergic encephalo-myelitis. J Immunol 150:4116–4124.

Kumar V, Tabibiazar R, Geysen HM, Sercarz E. 1995. Immunodomi-nant framework region 3 peptide from TCR Vb8.2 chaincontrols murine experimental autoimmune encephalomyelitis. JImmunol 154:1941–1950.

Liblau R, Tisch R, Bercovici N, McDevitt HO. 1997. Systemic antigenin the treatment of T-cell-mediated autoimmune diseases. Immu-nol Today 12:599–604.

Martin R, McFarland HF, McFarlin DE. 1992. Immunological aspectsof demyelinating diseases. Annu Rev Immunol 10:153–187.

Miltenyi S, Muller W, Eilchel W, Radbruch A. 1990. High gradientmagnetic cell separation with MACS. Cytometry 11:231–238.

Mor F, Cohen IR. 1993. Shifts in the epitopes of myelin basic proteinrecognized by Lewis rat T cells before, during, and after theinduction of experimental autoimmune encephalomyelitis. JClin Invest 92:2199–2206.

Mor F, Kantorowitz M, Cohen IR. 1996. The dominant and the crypticT cell repertoire to myelin basic protein in the Lewis rat. JNeurosci Res 45:670–679.

Nicholson LB, Greer JM, Sobel RA, Lees MB, Kuchroo VK. 1995. Analtered peptide ligand mediates immune deviation and preventsautoimmune encephalomyelitis. Immunity 3:397–405.

Offner H, Hashim GA,Vandenbark AA. 1991. T cell receptor peptidetherapy triggers autoregulation of experimental encephalomyeli-tis. Science 251:430–432.

Offner H, Buenafe AC, Vainiene M, Celnik B, Weinberg AD, Gold DP,Hashim G, Vandenbark AA. 1993. Where, when, and how todetect biased expression of disease-relevant Vb genes in ratswith experimental autoimmune encephalomyelitis. J Immunol151:506–517.

Rook GAW, Hernandez-Pando R, Lightman SL. 1994. Hormones,peripherally activated prohormones and regulation of the Th1/Th2 balance. Immunol Today 15:301–303.

224 Chan et al.

Smith LR, Kono DH, Kammuller ME, Balderas RS, TheofilopoulosAN. 1992. Vb repertoire in rats and implications for endog-enous superantigens. Eur J Immunol 22:641–645.

Stienekemeier M, Herrmann T, Kruse N, Weishaupt A, Weilbach FX,Giegerich G, Theofilopoulos A, Jung S, Gold R. 1999. Heteroge-neity of T-cell receptor usage in experimental autoimmuneneuritis in the Lewis rat. Brain 122:523–535.

Sun D, Le J, Coleclough C. 1993. Diverse T cell receptorb chain usageby rat encephalitogenic T cells reactive to residues 68–88 ofmyelin basic protein. Eur J Immunol 23:494–4948.

Sun D, Hu X-Z, Le J, Swanborg RH. 1994. Characterization ofbrain-isolated rat encephalitogenic T cell lines. Eur J Immunol24:1359–1364.

Torres-Nagel NE, Gold DP, Hu¨nig T. 1993. Identification of ratTcrb-V8.2, 8.5, and 10 gene products by monoclonal antibodies.Immunogenetics 37:305–308.

Tsuchida M, Matsumoto Y, Hirahara H, Hanawa H, Tomiyama K, AboT. 1993. Preferential distribution of Vb8.2-positive T cells inthe central nervous system of rats with myelin basic protein-induced autoimmune encephalomyelitis. Eur J Immunol 23:2399–2406.

Vandenbaark AA, Hashim G, Offner H. 1989. Immunization with a

syntheticT-cell receptorV-region peptide protects against experimen-tal autoimmune encephalomyelitis. Nature 341:541–544.

Vandenbark AA, Chou YK, Whitham R, Mass M, Buenafe A, LiefeldD. 1996. Treatment of multiple sclerosis with T-cell receptorpeptides: results of a double-blind pilot trial. Nature Med2:1109–1115.

Weishaupt A, Gold R, Gaupp S, Giegerich G, Hartung H-P, Toyka KV.1997. Antigen therapy eliminates T cell inflammation byapoptosis: effective treatment of experimental autoimmuneneuritis with recombinant myelin protein P2. Proc Natl Acad SciUSA 94:1338–1343.

Wilson DB, Steinman L, Gold DP. 1993. The V-region diseasehypothesis: new evidence suggests it is probably wrong. Immu-nol Today 14:376–382.

Zettl UK, Gold R, Toyka KV, Hartung H-P. 1995. Intravenousglucocorticosteroid treatment augments apoptosis of inflamma-tory T cells in experimental autoimmune neuritis (EAN) of thelewis rat. J Neuropathol Exp Neurol 54:540–547.

Zipp F, Kerschensteiner M, Dornmair K, Malotka J, Schmidt S, BenderA, Giegerich G, de Waal Malefyt R, Wekerle H, Hohlfeld R.1998. Diversity of the anti-T-cell receptor immune response andits implications for T-cell vaccination therapy of multiplesclerosis. Brain 121:1395–1407.

T-Cell Receptor Heterogeneity in EAE 225