Treatment with R R Vp-specific antibodies

Eur. J. Immunol. 1991. 21: 2899-2905

Gilles Chiocchiaon Marie-Christophe BoissieroA and Catherine FournierO

2899

Therapy against murine collagen-induced arthritis with T cell receptor Vg’specific antibodies*

INSERM U.283O,Hijpital Cachin,

p& and Dept. of ~ e ~ t o l o & , Immunization with native type 11Collagen (CU) of susceptible Strains of mice (H-29) induces a rheumatoid arthritis-like disease. Collagen-induced arthritis H6pital Avicenne, Bob(CIA) is an experimental model for Tcell-mediated autoimmune disease. To investigate the T cell receptor (TcR) repertoire involved in the pathogenesis of CIA, CII-primed DBA/1 mice were treated with various TcR Vg-specific monoclonal antibodies (mAb) using a protocol resulting in a long-term elimination of the target Tcells. Zn vivo treatment with anti-CD4 mAb led to nearly complete protection against CIA. Mice injected with anti-Vg8.1,2or anti-Vg5.1,2 mAb had a reduced incidence of arthritis (respectively28.6% and 50% vs 84.6% for the control group). Administration of anti-Vp2 mAb delayed the onset of the disease whereas injection of anti-Vg6 or anti-Vgll mAb did not alter CIA. Moreover, the combined treatment with anti-Vg2 and anti-Vg5 mAb efficiently reduced the development of CIA. The humoral response to CII was downregulated only in the groups of mice that were improved by the treatment. In v i m proliferative response to CII of lymph node cells from primed DBA/1 was partially blocked by addition of several anti-Vg mAb. Thus, our findings suggest that the overall T cell response to CII may be polyclonal while the Tcell clones involved in the pathogenesis of CIA express a limited number of Vg chains.

[8-lo]. Moreover, in vivo treatment with anti-CD4 mAb prevenk CIA [ l l , 121. In our laboratory,we have generated Collagen-inducedarthritis (CIA) is a chronic inflammatory murine CII-reactiveT cell hybridomas expressing the CD8 arthropathy which mimics rheumatoid arthritis in humans. phenotype and involved in antigen-specific cytotoxicity in The experimental disease is inducible in rats [l], mice [2], association with H-2Kq molecules [13]. When inoculated 3 or primates [3] upon immunization with native type I1 weeks before priming with CII, two of these clones that collagen (CII) from heterologousor homologousorigin [4]. exhibit different epitope specificity protected the mice Susceptibility to the disease in rodents is strongly in- against CIA. Furthermore, one of them also had the ability fluenced by MHC genes. In mouse, H-2 linkage which to down-regulatethe ongoing inflammation (manuscript in varies with the species of collagen used for immunization is preparation).Thus, both CD4+ and CD8+T cell clones may limited to H-2s and H-2‘ haplotypes in inbred strains [5]. participate in the induction and perpetuation of the More precisely, the expression of critical residues on the I-A autoimmune response to CII. Accordingly, the expression p molecule of these haplotypes were shown to predispose of some critical TcR rearrangements would be a prerequisite for the development of CIA. Indeed, the resistance of for susceptibility to CIA [6]. SWR (H-2q) mice to CIA is supposed to rely upon their Although the exact mechanism by which CIA develops is genomic deletion of 50% of the TcR Vg genes [14]. In still unclear, several lines of evidence suggest that both addition, preliminaryresults have implicated theVg6 allele, humoral and cellular responses to CII are intimately a gene which carries polymorphic variations in SWR mice, involved. Indeed, neither cells nor antibodiesalone are able as the susceptibility determinant to CIA [15]. to completely mimic arthritis induced by immunization [7]. Contribution of T cells to the development of arthritis have The aim of the present study was to analyze the TcR been substantiated by the successful transfer of arthritic repertoire which may confer susceptibility to CIA in lesions with CD4+ Tcell lines or clones specific for CII DBA/1 (H-29) mice. Our strategy was based upon the elimination of T cells expressing specific TcR Vg chains by treatment with mAb before and after immunization. [I 95991 Among the mAb available,we chose those specific for Vp5, 8, and 11 molecules because the corresponding genes are * This work was supported by the “Institut National de la Sante et deleted in SWR mice.We also used mAb directed toV 6, a de la Recherche Medicale“ (INSERM) and a grant from the gene that was postulated to be involved in CIA [15f. In addition,we treated mice with two other anti-Vg mAb. The “Association de Recherches sur la Polyarthrite” (ARP). first one was specific for cule without known Recipient of a studentship from ARP. intra-thymic selection by nants and expressed cond reacted with Correspondemce: Catherine Fournier, INSERM U.283, HBpital by all inbred mouse str deleted in mouse strains Cochin, 27, rue du Fg St-Jacques, F-75674 Paris Cedex 14, Vp3+ cells that are intrat expressing MlsCand IE s. Our results suggest a France crucial role for Tcells ssmg the Vg8.1,2 genes in Abbreviations: CIA: Collagen-induced arthritis CII: ’Qpe I1 arthritis development and an influence of at least two other sets of T cells that useVg2 and Vg5.1,2 families. In contrast, collagen

1 Introduction

0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991

0014-2980/9111212-2899$3.50+ .2510

2900

G. Chiocchia, M.-C. Boissier and C. Fournier

Eur. J. Immunol. 1991. 21: 2899-2905

treatment with anti-Vg6 or anti-Vgll mAb failed to modify the course of CIA.

2.3 I n vivo treatment with mAb and experimental protocol

The mAb used in this study are listed in Table 1. The hybridoma cells (kind gifts from the scientists mentioned in Table 1) were grown as ascites in nude mice. mAb were purified from ascites by two consecutiveammonium sulfate precipitations and chromatography on DEAE-Sepharose (Pharmacia, Uppsala, Sweden). For FCM analyses, mAb were biotinylated with NHS-d biotin (Sigma Chemicals, St. Louis, MO) according to the manufacturer’s recommendations.

TcR Vg-specific mAb (500 pg/mouse) were injected (i.p.) on two occasions: 1day before immunization with CII/CFA and on day 22 post-priming. Anti-CD4 mAb-treated mice received an additional injection of 500 pg of mAb on day 9 after immunization. Combined treatment with two mAb was performed with 250 pg of each mAb. Two series of experiments were conducted. The first one included 5 groups of 5 mice each: 4 groups were injected, respectively with mAb specific for TcR Vg2,3,6 or 11and control group was not treated. The second series was composed of 63 DBA/1 male mice randomly allocated to each group and mice from different groups were housed in each cage. Nine groups of 7 mice were performed. The two control groups were (a) untreated and (b) treated with rat IgG. Six experimental groups were given singly treatments with mAb to CD4 and to TcR Vg2, 5, 6, 8 or 11. Finally, one group received a combination of anti-Vg2 and anti-Vg5.

2.2 Induction and assessment of arthritides

2.4

Immunization and arthritis evaluation were performed as previously described in detail [26]. Briefly, DBA/1 male mice were injected (id.) at different sites on the back and at the root of the tail with 1OOpg of native bovine CII (Bioetica, Reims, France) emulsified in CFA. On day 20 after priming, mice received a boost injection (id.) with 1OOpg of CII in PBS. From the 3rd week after initial immunization on, mice were examined 3 times a week for signs of arthritis by an observer who had no knowledge of the treatment. In each mouse, the date of disease onset was recorded and clinicalseverity of each joint or group of joints (toes, tarsus, ankle, wrist, knee) was graded: 0 (normal), 1 (erythema), 2 (swelling), 3 (deformity), 4 (necrosis). The scores were summed to yield the arthritic score.

Spleen cells were isolated from individual mice and enriched in T cell by elimination of sIg+ cells using complement-mediated cytotoxicity with sheep anti-mouse IgG Ab (Biosys, CompBgne, France). Cell suspensions were then incubated with appropriate dilutions of biotinylated mAb. After washes, FITC-conjugated streptavidin (Immunotech, Marseille, France) was added and the cells were fixed and analyzed using a cytofluorograph FACScan (Becton Dickinson, Mountain View, CA) as previously described [13]. Percentages of residual B lymphocytes in the cell suspensions were always < 10% as evaluated with FITCconjugated goat anti-mouse IgG (Cappel, West Chester, PA).

2 Materials and methods 2.1 Mice and mAb DBA/1 J male mice, 7- to 10-week-old, were purchased from the Jackson Laboratory (Bar Harbor, ME).

FCM

Table 1. mAb used in this study

mAb

Specificity

Speciedisotypc

Obtained from

GK.1.5

CDJ

RatlIgGlb

B20.6.5

TcR Vp2

RatngGza

KJ 25

TcR Vp3

HamsterAgG

MR9.J

TcR VpS. 1.2

MouseAgG

5.1.22. I

TcR Vp6

RadIgGza

TR 310

TcR Vg7

Rat/lgGZb

KJ 16

TcR Vp8.1.2

RaflgG

MR 10.2

TcR Vp9

MousengG

RR 3-15

TcR Vpl1

RatngG

MR 12-4

TcR V,,13

MouselIgGr

Pierres M. (Marseille, France) Malissen B. (Marseille, France) Marrack I? (Denver, CO) Kanagawa 0. (St-Louis. MO) Hengartner H. (Zurich, Switzerland) Okada C.Y. (Stanford. CA) Marrack I? (Denver. CO) Kanagawa 0. (St-Louis, MO) Kanagawa 0. (St-Louis, MO) Kanagawa 0. (St-Louis. MO)

Eur. J. Immunol. 1991. 21: 2899-2905

Treatment with TcR Vg-specific antibodies

2.5 Proliferation assays

The proliferative response of LN cells after immunization with native CII was measured as described [27] with minor modifications. Briefly, cells were isolated from draining popliteal LN, 6 to 9 days after immunization in the hind paws, with bovine CII. LN cells (4 X lo5 celldwell) were suspended in RPh4I 1640 containing 1% mouse serum and incubated with varying concentrations of CII in the presence or absence of purified mAb (5 and 10 pg/well). In some experiments, LN cells were first enriched in T cells as described above and then used as responder cells in the proliferative assay. All samples were run in quadruplicate. After a 4-day incubation at 37”C, 1 pCi = 37 kBq of [3H]dThdwas added for an additional 18 h of culture and cells were harvested [27]. Antigen-driven proliferation was calculated by subtracting cpm incorporated into the cells incubated with mAb only from cpm of the cells stimulated with CII in the presence of the same mAb. Results were expressed as percentages of inhibition in comparison with CII-stimulated cultures without mAb.

2901

‘1

20

[Intl.VI)Il

41

30

50

60

70

80

40

50

60

70

80

J(

‘1

a 2 3

40

I

-.

2

U

I1

i

0 20

30

40

50

80

70

80

-

02 0

30

Antl.VOZ+sntl-VOS 4 1

2.6 Anti-CII antibody levels Mice were bled on day 15 after immunization and the individual sera were tested for antibodies directed against bovine or mouse native CII using a standard ELISA as already describedin detail [4].The results were expressed as absorbance in experimental wells minus the background absorbance obtained in wells without CII.

Days sftar ImmunlwIon

Days

8ltsr Imrnunlratlon

3 Results 3.1 Effect of in vivo treatment with TcR Vg’specific mAb on CIA development Attempts to modify the course of CIA by elimination of selected T cell subsets were conducted in DBA/l mice after injection of large amounts (twice 500 pg) of mAb specific for different TcR Vg chains. In a first experiment using a limited number of mice (n = 5 per group), we te possible therapeutic efficacy of mAb directed to Vp2,Vg3, Vg6 or Vpll. Inoculation of mAb 1day before and 22 days after priming with CII resulted in the depletion of the relevant T cell subsets except for anti-Vp3mAb which failed to eliminate in vivo its target cells. Thus, no conclusive observation could be derived from the anti-Vg3-treated group. Regarding the other groups, mice that received either anti-Vg6 or anti-Vgll developed a chronic polyarthritis very similar to that observed in untreated controls. In contrast, in vivo treatment with anti-Vp2 n A b reduced the severity and slightly delayed the onset of clinicalsigns of arthritis (data not shown). Thus, the next experiment was monitored to confirm these findings and extend them to other T cell subsets. As five of the six mAb used were raised in rat, a control group injected with purified rat IgG was added. No difference in terms of incidence and severity of arthritis were noted between untreated mice and those given rat IgG. Consequently, the results of both groups were pooled and used as control reference for comparison with the mAb-treated groups. Confirming our first experiment, the clinical course of CIA was not affected by treatment with either anti-Vg6 or anti-Vgll mAb whereas

Days snsr Immunlrstlon

Figure 1. Effect of in vivo treatment withTcR Vp-specific mAb on CIA development. Mice were injected with two doses of 500 pg of anti-Vg mAb or 3 doses of anti-CD4 mAb as indicated in Sect. 2.3. Each group represents the clinical course of CIA in mice treated with the mAb indicated (0)or in control group including untreated and rat IgG-injected mice (B).Results are expressed as mean arthritic score for a given day of 7 mice in mAb-treated groups (except for anti-Vp5-treated mice, n = 6) and 13 mice in control group.

elimination of Vg2 expressingTcells delayed by 1week the onset of arthritis (Fig. 1). In addition, we provided evidence that treatment with anti-Vg8.1,2 almost completely protected the mice against CIAgiven that only two of seven mice exhibited macroscopic signs of arthritis as from days 58 and 72 post-priming, respectively.Very similar findings were found in anti-CD4-treated animals in which suppression of arthritis was maintained until day 64 (Fig. 1 and Table 2). Our data also showed that elimination of Vp5 expressing cells induced a reduction of arthritis scores 2 weeks after the disease onset. Interestingly, the combined treatment with anti-Vg2 and anti-Vg5 mAb resulted in the cumulative beneficial effects of the mAb administered singly, namely an impaired severity of CIA at early and late phases of clinical course (Fig. 1).

2902

Eur. J. Immunol. 1991.21: 2899-2905

G. Chiocchia, M.-C. Boissier and C. Fournier

Table 2. Effect of in vivo treatment with TcR Vppecific mAb on incidence, time of onset and severity of CIA In vivo treatments

None or rat IgG (pooled controls) Anti-04 Anti-Vg2 Anti-V$ Anti-Vg2 + anti-vg Anti-Vg6 Anti-Vp8 Anti-Vgll

n

17

7 12 6

7 12 7 11

Incidence ona)

Average day of onsetb)

Arthritis severity)

Day +40 n (%)

Day +76 n (%)

9 (52.9) 0 (0) *

3 (25) 3(50)

15 (88.2) 4(57.1) 11 (91.6) 3 (50)

44.4 f 3.8 68.0 f 2.3*** 48.8 f 4.2 33.3 f 2.0**

5.0 f 0.9 3.2 f 0.6 3.9 f 0.9 4.1 & 0.5

2 (28.5) lO(83.3) 0 (0) * 6(54.5)

5 (71.4) lO(83.3) 2(28.5) 9 (81.8)

49.8 f 5.7

3.5 f 0.6

35.7 f 1.0* 65.0 f 7.0 39.8 f 3.3

5.6 f 1.1 2.8 f 0.6 4.9 & 1.0

Further analysis of CIA parameters, in the two experiments performed, revealed that anti-Vg8-treated mice as well as those receiving anti-CD4 mAb had a very significant reduction of incidence and a delayed onset of arthritis but in addition, the few mice that were arthritic exhibited a milder disease than control mice (Table 2). Treatment with antiVg5 prevented 50% of the mice from CIA. However, the mice that were not protected developed early and severe clinical signs of arthritis. In contrast, depletion of Vg2+ cells delayed the onset without affecting the incidence or the severity of arthritis (Table 2).

3.2 Eliminstion of TcR-expressing cells after in vivo treatment with mAb The selective depletion of Vg-expressingTcells followingin vivo treatment with mAb was evaluated by FCM on day + 8 post-immunization (9 days after the first injection of mAb) and on days 79 and 90 (respectively, 57 and 68 days after the second injection of mAb) and compared to the percentages observed in untreated DBA/1 mice. As shown on Fig. 2 (left panel), immunization with CJI resulted in a slight increase in the percentages of all the R R Vg+ cells tested, while the overall CD4+ population was not altered. Following the first injection of mAb to either CD4 molecule or the different Vg chains we used, an almost complete elimination of the respective T cell subsets was observed. As shown on Fig. 2 (right panel) 8 days post-immunization, the proportion of depleted cells in the spleen represented

Therapy against murine collagen-induced arthritis with T cell receptor V beta-specific antibodies.

Immunization with native type II collagen (CII) of susceptible strains of mice (H-2q) induces a rheumatoid arthritis-like disease. Collagen-induced ar...
783KB Sizes 0 Downloads 0 Views