CELLULAR

IMMUNOLOGY

140,444-452 (1992)

Suppression of Collagen Arthritis with Antibodies to an Arthritogenic, Oligoclonal T Cell Line’ DEREK J. PEACOCK,*GRACEKu,*,? MONA LISA BANQUERIGO,* AND ERNESTBRAHN*,* *Department of Medicine, Division of Rheumatology, UCLA School of Medicine, Los Angeles, California 90024-1670; and tDepartment of Microbiology and Molecular Genetics, University of California, Los Angeles, California 90024-1489 Received October 22, 1991; accepted December 2, 1991 Rats immunized with type II collagen (CII) develop an immunologically mediated polyarthritis. T cells have been implicated in the pathogenesisof this model since they can adoptively transfer the disease.A CII-specific T cell line (VA), consisting of three distinct clones by Southern blot analysis, has been shown to be arthritogenic. Antibodies specific for this line were generated by immunizing rabbits. In an attempt to prevent collagen-induced,arthritis (CIA), Louvain rats were injected with 1 ml of anti-VA ip on Days - 1, + 1, +3 and 0.5 ml on Day +5 (early treatment). To evaluate its effect on existing disease,rats received anti-VA on the day of arthritis onset and subsequently on 4 successivealternate days using the same dosage protocol (late treatment). Control rats received no therapeutic injections or were administered normal rabbit serum. All rats were immunized with CII on Day 0 to induce CIA. Rats administered antibodies using the early anti-VA treatment protocol had a significantly diminished incidence of arthritis compared to controls. Established arthritis was significantly diminished compared to controls in rats given the late anti-VA treatment. In both protocols, radiographic evidence of joint destruction was significantly reduced compared to controls. T cell phenotyping using flow cytometry analysis demonstrated that the anti-VA antibody therapy selectively eliminated a small subset of T cells since there was little difference in total T cell counts in the experimental versus control groups. Delayed type hypersensitivity and IgG antibody titers to CII were minimally decreasedin the experimental versus control group. These results suggestthat antibodies raised to an oligoclonal arthritogenic T cell line can suppresscollagen arthritis. This may have implications with respect to 1) the size of the T cell receptor repertoire involved in the pathogenesis of collagen arthritis and 2) immunospecific protocols for CIA and other autoimmune diseases. 0 1992 ~&tic press,IX.

INTRODUCTION Collagen-induced arthritis (CIA) is an animal model of rheumatoid arthritis that can be induced in susceptible strains of rodents and nonhuman primates after immunization with native type II collagen (CII) (1, 2). T cells have been implicated in 1This work was supported in part by National Institutes of Health Grants AR 38844 and AR 36834 and by grants from the Southern California Chapter of the Arthritis Foundation (Meyer Arthritis Investigator Award), the Bertram A. Maltz, M.D., Laboratory for Molecular Rheumatology, and the Senate Committee on Research at UCLA. * To whom correspondenceshould be addressedat UCLA School of Medicine, Division of Rheumatology, 1000 Veteran Ave., Rm. 32-47, Los Angeles, CA 90024-1670. 444 0008-8749192$3.00 Copyright 0 1992 by Academic Press,Inc. All rights of reproduction in any form reserved

ANTIBODY

SUPPRESSION OF COLLAGEN ARTHRITIS

445

the pathogenesis of CIA by studies showing that arthritis can be induced in naive animals by the transfer of lymph node and spleen cells from rats immunized with CII (3). Additionally, CIA does not occur in congenitally athymic rats (4) or animals treated with either antithymocyte or CD4+ antibodies (5, 6). T cell lines have been developed from CII-immunized Louvain (LOU) rats (7). These lines proliferate in responseto CII in vitro, although the level of reactivity does not predict arthritogenicity in vivo (8). Southern blot analysis of T-T hybridomas developed from an arthritogenic line, designated VA, has shown that it contains three CII-reactive clones, suggesting limited T cell receptor (TCR) VP-gene rearrangements (9). The purpose of the current study is to examine the effect of antibodies to line VA (anti-VA) on prevention and treatment of established arthritis in vivo. These studies may help clarify the size of the TCR repertoire involved in CIA and could have clinical relevance for immunospecific therapies of other autoimmune diseases. MATERIALS AND METHODS Animals Syngeneic LOU rats were used in all experimental protocols. Animals were housed at the UCLA Vivarium facility and received standard laboratory chow and water. New Zealand white rabbits were used to generate antibodies to line VA. Antibodies to Line VA The formation of collagen-specific arthritogenic T cell lines (W3/25+, 0X8- phenotype) has previously been described (7). Antibodies specific to an arthritogenic line VA were generated by immunizing rabbits with 4 X lo8 line VA cells in complete Freund’s adjuvant (CFA) (Difco Laboratories, Detroit, MI) and giving two booster injections of 1 X lo8 line cells in CFA at 3-week intervals. Anti-VA serum was then sequentially adsorbed with rat liver and kidney to remove antibodies to public rat antigens. Two distinct activated CD4+ ovalbumin-specific T cell lines (W3/25+, 0X8-) (10) were then used to adsorb nonspecific T cell antibodies. Arthritis Induction Rats were immunized id on the back during ether anesthesia with 0.5 mg native chick CII (Genzyme, Boston, MA) solubilized in 0.1 A4 acetic acid and emulsified in incomplete Freund’s adjuvant (IFA) (Difco Laboratories, Detroit, MI) (1). Experimental Design All rats were immunized with CII on Day 0 to induce arthritis which usually develops 9- 14 days later. Two experimental protocols were evaluated. In an attempt to prevent CIA, rats received 1 ml of anti-VA ip on Days - 1, + 1, +3 and 0.5 ml on Day 5 (early treatment). To evaluate its effect on established arthritis, rats received anti-VA on the day of arthritis onset and subsequently on four successivealternate days, using the same dosage as described above (late treatment). Control rats in the early treatment group received no therapeutic injections or were administered normal rabbit serum (Whittaker Bioproducts, Walkersville, MD). Rabbit serum was not given to the late treatment control group since prior use in the early treatment control group had no

446

PEACOCK ET AL.

effect on arthritis. Clinical assessmentof arthritis was begun on Day 9. Antibody levels to CII were measured on Days 14 and 25 and rats were sacrificed on Day 28 after completion of DTH assaysto CII. Hind limbs were harvested for subsequent radiographs.

Arthritis Assessment Both the incidence and the severity of the arthritis were evaluated. Incidence was defined as the number of rats that had clinical evidence of synovitis during the study period, and severity was quantified by daily scoring each paw in integers from 0 to 4 (0 = normal, 4 = maximum) based on increasing levels of swelling and periarticular erythema. The sum of the scoresfor all four paws was used to calculate the arthritic index (1, 11). Scoresof 6-8 represent severearthritis since CIA primarily affects hind limbs. Radiographic scoring was assessedby the extent of soft tissue swelling, joint spacenarrowing, periosteal new bone formation, and bone destruction ( 1, 12). Scores were assignedas integers from 0 to 3 per limb (0 = normal, 3 = maximum score) and were determined by a blinded investigator. The radiographic joint index represents the mean of the hind paw scores from all rats (12), with a maximum possible score of 6 per rat.

Cellular Immunity to CII DTH to native CII was quantified in vivoby a radiometric ear assay( 13). Rats were pulsed with tritiated thymidine on a weight basis and, 24 hr later, CII was injected id in one ear and control buffer in the other. After another 24 hr, the ears were harvested using a punch biopsy and their relative radioactivity was determined with a scintillation counter. The radiometric ear index was defined by the ratio of counts per minute (cpm) for the CII-injected ear/cpm for the buffer-injected ear. A radiometric ear index 2 1.4, which represents 2 standard deviations above the mean for naive controls, is considered a significant DTH response to CII.

Humoral Immunity to CII IgG antibodies to CII were measured in serum samples obtained on Days 14 and 25 by an enzyme-linked immunosorbent assay (ELISA) (14). Based on a previously standardized curve, titers were expressedas the optical density (O.D.) at 490 ti of a I:2560 serum dilution.

Complement-DependentCytotoxicity Studies VA line cells were incubated for 45 min at 37°C in 1 ml of Dulbecco’s modified Eagle’s medium containing 2% rabbit serum as a complement source (Mediatech, Washington, DC). Anti-VA in serial dilutions was added to VA line aliquots. Cell viability was determined by Trypan blue exclusion (5).

Cell Phenotyping Mouse monoclonal antibodies W3/25 (anti-rat CD4) (15) and MRC OX-8 (antirat CD8) ( 16) were purchased asascites(Accurate Corp., Westbury, NY). A monoclonal mouse anti-rat crfi TCR antibody was obtained from culture supematants generated

ANTIBODY

SUPPRESSION OF COLLAGEN ARTHRITIS

447

from the R73 hybridoma (provided by Dr. D. P. Gold, La Jolla Institute for Experimental Medicine, La Jolla, CA) (17). All antibodies had the IgG, isotype. Affinitypurified, fluorescein isothiocyanate (FITC)-labeled goat anti-mouse IgG antibody (Caltag Laboratories, Inc., San Francisco, CA) or FITC F(aby2 goat anti-rabbit Fc antibody (Cappel Inc, Westchester, PA) was used as the second antibody. The cell surface phenotype of splenic cells obtained from antibody-treated and control LOU rats was assessedby standard fluorescent antibody staining and flow cytometry techniques. Ficoll-Hypaque-purified LOU rat spleen cells ( 1 X 106)were incubated for 30 min at 4°C with the various dilutions of antibody in complete medium: RPM1 1640 (Fisher/Cellgro, Tustin, CA) supplemented with 5% fetal bovine serum (FBS) (J. R. Scientific, Woodland, CA), 5 X lo-* mM 2-mercaptoethanol, 2 mM L-glutamine, 100 U/ml penicillin, 100 pg/ml streptomycin, 1 r&4 sodium pyruvate, 0.1 mM nonessential amino acids, and 0.1% sodium azide (NaN3). Glutamine, penicillin, streptomycin, sodium pyruvate, and nonessential amino acids were obtained from Irvine Scientific (Irvine, CA). Based on prior standardization experiments, the R73 antibody was used at a 1:2.5 dilution and the W3/25 and MRC OX-8 antibodies were used at a 1:200 dilution. The cells were then incubated with a I:25 dilution of the FITC goat anti-mouse IgG antibody, washed three times with RPM1 complete media, and resuspended in phosphate-buffered saline (pH 7.4) and 1% paraformaldehyde. Line VA-derived T-T hybridomas (9), bearing the VP4 gene segment in their expressed TCR, were incubated with a 1:4 dilution of anti-VA antibodies and then with a 1:25 dilution of the FITC F(aby, goat anti-rabbit Fc antibody prior to flow cytometry analysis. Fluorescein-labeled cells were analyzed on a FACScan flow cytometer (Becton-Dickinson Immunocytometry Systems,San Jose,CA) equipped with a 15-mW 488-nm air-cooled argon-ion laser. Five thousand events were acquired for each sample and populations of lymphocytes were distinguished from monocytes and granulocytes by using the FACScan Autocomp software (Becton-Dickinson). This FACScan software optimized forward light scatter (FSC) and side (right angle) light scatter (SSC)and automatically adjusts the fluorescence photomultiplier tube settings and compensation. FITC emission was detected using a 530-nM short pass filter. Statistics Continuous variables were analyzed by their group means (Student’s t test) and dichotomous variables by their proportionate group frequencies (X2 test). The Yates correction factor was used where appropriate and results were considered significant at the P < 0.05 level. RESULTS Incidence and Severity of Arthritis Experimental rats were administered anti-VA starting on Day - 1 (early) or on the day of arthritis onset (late) (Table 1). There was a significant decreasein the incidence of arthritis (P < 0.00 1) in rats receiving anti-VA (early) compared to controls (Table 1). Only one rat developed mild arthritis in the early anti-VA antibody group. All rats in the late anti-VA protocol developed arthritis by Day 10, comparable to untreated controls. By Day 14,4 days (two doses)after the initiation of anti-VA (late) treatment, the rats had a significantly diminished arthritic index (P < 0.0 1) compared to controls.

448

PEACOCK ET AL. TABLE 1 Incidence of Arthritis and Radiographic Scores Treatment group

Number of rats

Control Anti-VA” starting Day - 1 (early) Anti-VAb starting at arthritis onset (late)

30 8 9

Incidence of arthritis

Radiographic index 4.6 0’ 1.3’

93% 12?7ac NA

a Antibody to line VA, given ip on alternate days for a total of four doses. b All rats had synovitis by Day 10, NA = not applicable. ’ P < 0.00 1 compared to controls.

By Day 15 the difference in arthritis severity between the groups was greater (P < 0.001) and was sustained throughout the remainder of the study (Fig. 1). More than half (55%) of experimental rats in the late treatment group had no clinically detectable arthritis by Day 19 (9 days after institution of anti-VA antibody therapy). Radiographic Scores At the conclusion of the studies on Day 28, the mean radiographic indexes of rats administered anti-VA early or late were 0 and 1.3, respectively. The mean radiographic index for control rats was 4.6 (P < 0.001 compared to both experimental groups) (Table 1). Rats without clinical arthritis by Day 28 had no radiographic evidence of joint destruction. This included five rats in the anti-VA (late) treatment group that initially had clinical synovitis prior to therapy. DTH and Antibody Responses Sensitization to CII, as measured by DTH and IgG antibodies, was evaluated in all rats. Both anti-VA treatment groups had a similar DTH responsecompared to controls (Table 2). Antibodies to CII, as measured by ELISA, were evident in all rats by Day 14. Both anti-VA groups had a significantly diminished antibody responseon Day 25 compared to controls. TABLE 2 Immune Responsesto CII DTH response”

Antibody tite@

Treatment group

Day 28

Day 14

Day 25

Control Anti-VA starting day - 1 (early) Anti-VA starting at arthritis onset (late)

3.61 3.03 3.54

0.202 0.152 0.203

0.203 0.165 0.183

’ Mean radiometric ear index. ’ Mean optical density at I:2560 dilution,

ANTIBODY

SUPPRESSION OF COLLAGEN ARTHRITIS

449

0 ~~l~I’i’i~l~,‘,--rT0

2 DAY

4

6 POST

8 10 12 14 16 18 20 ARTHRITIS

ONSET

FIG. 1. Severity of collagen arthritis in the anti-VA (late) treatment group (mean score f SEM). Rats received anti-VA starting on the day of arthritis onset. Q Control rats; q , Late anti-VA group. *P < 0.01 compared to controls beginning 4 days after initiation of anti-VA therapy.

Complement-DependentCytotoxicity In complement-dependent cytotoxicity assays,the addition of anti-VA antibodies to line VA in dilutions greater than 1:10,000 resulted in cell death compared to controls (line VA and rabbit serum). These findings were not observed if anti-VA antibodies were incubated with nonpathogenic ovalbumin-specific CD4+, CD8- T cell lines (data not shown).

T Cell Phenotyping The surface phenotype of splenic cells from antibody-treated and control LOU rats was examined by indirect fluorescent antibody staining. The percentage of splenic lymphocytes reactive with antibodies specific for the cell surface antigens CD4, CD8, and &TCR was determined separately for each of three antibody-treated and three control LOU rats. Experimental rats had only a slight decreasein TCRf lymphocytes compared to controls (mean of 65 and 67%, respectively) (Fig. 2). In experimental rats the percentage of CD4+ and CD8+ cells was increased 3% and decreased 6%, respectively, compared to control rats. Therefore, anti-VA antibodies did not have a major public deletional effect on total T cells or T cell subsetsin treated animals. Line VA-derived T-T hybridomas, bearing VP4 TCR, showed intense staining with antiVA (data not shown). DISCUSSION The administration of anti-VA antibodies to an oligoclonal arthritogenic T cell line not only suppressed,but also reversed both clinical and radiographic evidence of CIA. Only one rat developed diseasewhen anti-VA was given prior to arthritis onset and, by Day 28, it completely eliminated clinical arthritis in 55% of rats with established

450

PEACOCK ET AL.

FIG. 2. T cell phenotyping by FACS analysis. Column A: Control group. Column B: Anti-VA (early) group. Pooled mononuclear cells from three spleens in both groups were stained with R73 (mouse anti-rat TCR Mab) (Row 1), W3/25 (mouse anti-rat CD4 Mab) (Row 2) or MRC OX-8 (mouse anti-rat CD8) (Row 3). X axis, mean and peak channel log,, green fluorescence; Y axis, percentage of positive cells detected with each Mab.

arthritis. The majority of animals in the latter group had no radiographic evidence of joint destruction- This significant reversal of established CIA has not previously been found with any other single-agent treatment modality. Although it is difficult via adsorption techniques to generate clone- or TCR-specific antibodies, several lines of evidence suggestthat the anti-VA antibodies are relatively specific. Previous studies have shown that hybridomas derived from line VA contain three clones by Southern Blot analysis (9) and it was this oligoclonal line that was used to produce anti-VA antibodies. To increase the specificity of the anti-VA serum, it was sequentially adsorbed against 1) rat liver and kidney to remove public rat

ANTIBODY

SUPPRESSION OF COLLAGEN ARTHRITIS

451

antibodies and 2) two activated CD4+ ovalbumin-specific T cell lines to remove nonspecific T cell antibodies. FACS analysis of spleen ceils from anti-VA-treated rats showed only a small decreasein the R73+ (@ TCR) cells compared to controls. In contrast, a 98% decreasein W3/ 13+ cells (pan T cell) was observed when heterologous anti-thymocyte serum (ATS) was given using a similar protocol (5). Moreover, antiVA antibodies, at titers greater than 10e4,were lethal for line VA cells in complementdependent cytotoxicity studies but were not cytotoxic for CD4+ ovalbumin-specific T cell lines. Northern blot analysis of line VA hybridomas has shown that five of the seven hybridomas that have been characterized use VP4 in their expressedTCR ( 18). It is interesting to note that anti-VA antibodies were able to bind to a VP4 hybridoma. In the prior ATS study, both DTH and the total number of T cells were significantly reduced in experimental animals (5). This is in contrast to the current study in which the elimination of a few clones that recognize a small epitope on the CII molecule (9, 18) does not preclude other sensitized T cells from manifesting a strong DTH response to different (nonarthritogenic) epitopes. In an additional comparison of the specificity of anti-VA and ATS, a significant diminution in arthritis was observed in anti-VA (late)-treated rats, which was not observed if ATS was begun even as early as 5 days prior to arthritis onset (5). The above findings suggestthat arthritis reduction in experimental, anti-VA-treated rats occurred by the elimination of a T cell subset that was critical in the pathogenesis of CIA and imply that anti-VA is relatively selective for T cells with a limited TCR repertoire. At present, TCR Va! and VP usage in CIA is not clearly defined. VP6 has been reported to be preferentially used in mm-meCIA ( 19), but this has not been confirmed by other groups (20,21). Experimental allergic encephalomyelitis (EAE) and CIA are both CD4+ T cell-dependent autoimmune diseases.In trying to elucidate the TCR repertoire of these animal models, various immunospecific protocols have been evaluated. Mab to CD4 and MHC II have been successful in treating EAE and CIA (6, 22-25). Pathogenic T cell lines in rat and murine EAE are specific for different epitopes on myelin basic protein and have different MHC restrictions but both use Vfl8.2 and Va4.3 gene segments in their expressed TCR (26-28). Mab specific to these TCR subsetshave been shown to prevent encephalomyelitis in rats and mice (26, 27, 29). Since anti-VA antibodies successfullydiminished CIA, and since the VP4 genesegment is expressedby the majority of line VA hybridomas, future protocols will evaluate the effectivenessof Mab to VP4 in preventing arthritis. The development of Mab targeted to V/34 could offer a more immunospecific therapy for CIA. The concept that a subset of T cells with a limited TCR repertoire may be critical in the pathogenesis of CIA may have clinical implications for rheumatoid arthritis. There are no published studies of the V/3 repertoire in naive LOU rats, although unpublished investigations (Daniel P. Gold, Ph.D., personal communication) indicate that VP4 usage may represent approximately 8% of c@’ T cells. Mab to CD4+ have been used successfully to treat rheumatoid arthritis (30) and studies are in progressto define T cell clonality in rheumatoid pannus (31). A better understanding of the TCR repertoire involved in autoimmune diseasesshould result in more effective and safertherapeutic interventions. REFERENCES 1. Trentham, D. E., Townes, A. S., and Kang, A. H., J. Exp. Med. 146, 857, 1977. 2. Cathcart, E. S., Hayes, K. C., Gonnerman, W. A., Lazzari, A. A., and Franzblau, C., Lab. Invest. 54, 26, 1986.

452

PEACOCK ET AL.

3. Trentham, D. E., Dynesius, R. A., and David, J. R., J. Clin. Invest. 62, 359, 1978. 4. Klareskog, L., Holmdahl, R., Larsson, E., and Wigzell, H., Clin. Exp. Zmmunol. 51, 117, 1983. 5. Brahn, E., and Trentham, D. E., Cell. Zmmunol. 86,421, 1984. 6. Ranges, G. E., Sriram, S., and Cooper, S. M., J. Exp. Med. 162, 1105, 1985. 7. Brahn, E., and Trentham, D. E., Cell. Zmmunol. 109, 139, 1987. 8. Brahn, E., and Trentham, D. E., Cell. Zmmunol. 118, 491, 1989. 9. Ku, G., Brahn, E., and Kronenberg, M., Cell. Zmmunol. 130,472, 1990. 10. Helfgott, S. M., Dynesius-Trentham, R., Brahn, E., andTrentham, D. E., J. Exp. Med. 162, 1531, 1985. 11. Schoen, R. T., Green, M. I., and Trentham, D. E., J. Zmmunol. 128,717, 1982. 12. Brahn, E., and Trentbam, D. E., Clin. Zmmunol. Zmmunopathol. 31, 124, 1984. 13. McCune, W. J., Trentham, D. E., and David, J. R., Arthritis Rheum. 23, 932, 1980. 14. Trentham, D. E., and Dynesius-Trentham, R. A., J. Zmmunol. 130,2689, 1983. 15. Jeffiies, W. A., Green, J. R., and Williams, A. F., J. Exp. Med. 162, 117, 1985. 16. White, R. A. H., Mason, D. W., Williams, A. F., Galfre, G., and Milstein, C., J. Exp. Med. 148, 664, 1978. 17. Hilnig, T., Wallny, H.-J., Hartley, J. K., Lawetzky, A., and Tiefenthaler, G., J. Exp. Med. 169, 73, 1989. 18. Kronenberg, M., Brahn, E., Gold, D., and Ku, G., FASEB 5, A606, 1991. 19. Haqqi, T. M., and David, C. S., J. Autoimmunity 3, 113, 1990. 20. Smith, L. R., Plaza, A., Singer, P. A., and Theofilopoulos, A. N., J. Zmmunol. 144, 3234, 1990. 21. Goldschmidt, T. J., Jansson, L., and Holmdahl, R., Immunology 69, 508, 1990. 22. Walder, M. K., Sriram, M. K., Hardy, R. L., Herzenberg, A., Herzenberg, L. A., Lanier, L., Lim, M., and Steinman, L., Science 227, 4, 1985. 23. Sriram, S., and Roberts, C. A., J. Zmmunol. 136,4464, 1986. 24. Wofsy, D., and Seaman, W. E., J. Exp. Med. 161, 378, 1985. 25. Billingham, M. E., Hicks, C., and Camey, S., Agents Actions 29, 77, 1990. 26. Acha-Orbea, H., Mitchell, D. J., Timmermann, L., Wraith, D. C., Trausch, G. S., Walder, M. K., Zamvil, S. S., McDevitt, H. O., and Steinman, L., Cell 54, 263, 1988. 27. Urban, J. L., Kumar, V., Kono, D. H., Gomez, C., Horvath, S. J., Clayton, J., Ando, D. G., Sercarz, E. E., and Hood, L., Cell 54, 577, 1988. 28. Bums, F. R., Li, X., Shen, N., Offner, H., Chou, Y. K., Vanderbark, A. A., and Heber-Katz, E., J. Exp. Med. 169, 27, 1989. 29. Owhashi, M., and Heber-Katz, E., J. Exp. Med. 168, 2153, 1988. 30. Reiter, C., Kakavand, B., Rieber, E. P., Schattenkirchner, M., Riethmtiller, G., and Kruger, K., Arthritis Rheum. 34, 525, 1991. 31. Cooper, S. M., Dier, D. L., Roessner,K. D., Budd, R. C., and Nicklas, J. A., Arthritis Rheum. 34, 537, 1991.

Suppression of collagen arthritis with antibodies to an arthritogenic, oligoclonal T cell line.

Rats immunized with type II collagen (CII) develop an immunologically mediated polyarthritis. T cells have been implicated in the pathogenesis of this...
616KB Sizes 0 Downloads 0 Views