T Lymphocytes in the Synovial Fluid of Patients with Active Rheumatoid

Home , CD134, RANK

Clinical and Experimental Rheumatology 2001; 19: 317-320. BRIEF PAPER T lymphocytes in the ABSTRACT in the pathogenesis of the disease (4). Objective. To assess the percentage of The human CD 134 (OX40) cell sur- synovial fluid of patients T ly m p h o cy t e s , b e a ring CD134, a faceantigen, a 50-kDa membrane-asso- with active rheumatoid member of the TNF receptor superfam - ciated glycoprotein, is a member of the arthritis display CD134- i ly, p ri m a ri ly found on autore a c t ive tumor necrosis factor (TNF) receptor CD4+ T cells in the peripheral blood superfamily, which is found primarily OX40 surface antigen (PB) and synovial fluid (SF) of rheu - on activated CD4+ cells and not on matoid arthritis (RA) patients. normal resting peripheral blood lym- R. Giacomelli, M e t h o d s . The surface ex p ression of phocytes (5). Its interaction with the A. Passacantando, CD134 on SF and PB mononu cl e a r specific ligand (CD134L Ð OX40L), a 1 cells was performed by flow cytometry type II membrane protein (6) generally R. Perricone , in 25 RA patients and correlated to the expressed on activated B lymphocytes I. Parzanese, M. Rascente, disease activity. (7), antigen presenting cells, and acti- G. Minisola2, G. Tonietti Results. CD134 expression on CD3+, vated endothelial cells (8) is, on one CD4+, CD8+ and CD25+ cells was hand, an efficient costimulatory signal Clinica Medica, University of L’Aquila; higher in SF than in PB of RA patients for CD4+ T cell-dependent humora l 1Immunologia Clinica, University of Tor ( P < 0.001). No diffe rences we re ob - immune responses (9), resulting in the Vergata; 2Santo Spirito-Villa Betania served in the percentage of CD134+/ induction of IL-4 production by naive T Hospital, Rome, Italy. CD4+ T lymphocytes in the PB of RA cells and their development into effec- Please address correspondence and p atients and controls. Patients with tor cells producing higher levels of this reprint requests to: R. Giacomelli, MD, active RA had significantly higher per - cytokine (10), thus enhancing B-cell Clinica Medica, Department of Internal centage of CD3+/CD134+, CD4+/ CD proliferation and immunoglobulin se- Medicine, University of L’Aquila, 134+,CD8+/CD134+and CD25+/ CD cretion (9) and, on the other hand, an via Vetoio, 67100 L’Aquila, Italy. 134+ than those with inactive disease. event mediating the adhesion of lym- E-mail: [email protected] C o n clusion. These findings sugge s t phocytes to endothelial cells and leuko- Received on July 11, 2000; accepted in that CD134+ T cells are involved in the cyte recruitment (8). revised form on November 7, 2000. i m mu n o p at h o l ogical process of RA The CD134 molecule shows a distinct © Copyright CLINICAL AND synovitis, maybe mirroring some other p at t e rn of ex p ression in autoimmu n e EXPERIMENTAL RHEUMATOLOGY 2001. autoimmune disease in which autore - d i s e a s e s , in that it is ex p ressed on active T cell infiltrating the target tis - CD4+ autoreactive T cells, in the in- Key words: Rheumatoid arthritis, sues largely coexpress CD134. flammatory infiltrates of experimental CD134, TNF-receptors superfamily. autoimmune encephalomyelitis (EAE) Introduction ( 1 1 ) , i n fl a m m at o ry bowel disease Rheumatoid arthritis (RA) is a disease (IBD) (12) and acute graft-versus host of unknown aetiology, in which several disease (aGVHD) (13), but not in their T cell-dependent immune mechanisms ci rc u l a ting CD4+ cells. Furt h e rm o re, are involved in the pathogenesis, inclu- s e l e c t ive depletion of OX 4 0 + / C D 4 + ding T cell migration and their recruit- c e l l s , in EAE, induces stab i l i z at i o n ment and activation in the synovia with and/or disease prevention, via elimina- consequent cytokines release (1). The tion of autoantigen re a c t ive T cells majority of T lymphocytes infiltrating (14), suggesting a potential role for tar- RA joints express CD4 molecules and geting OX40+ cells as a strategy in the specific autoreactive T cells are present control of autoimmune diseases. in this CD4+ population (2). In this study, given the specificity of In the last years, on the basis of their CD134 for autoreactive CD4+ T cells, cytokines production,T cells have been we evaluated whether this molecule is divided into two distinct subsets, mod- overexpressed in RA. Thus, we asses- ulating different immune responses (3): sed the percentage of CD134+ T lym- T helper 1 (Th1) cells, mainly produc- phocytes in peripheral blood (PB) and ing interleukin-2 (IL-2) and interferon- synovial fluid (SF) of RA patients and (IFN- ) and Th2 cells, mainly pro- we analysed the possible correlations ducing IL-4, IL-5 and IL-10. Recent of these findings with clinical and labo- studies have shown that CD4+ T lym- ratory parameters of disease activity. phocytes, isolated from RA joints, pro- duced high levels of IFN- and low lev- Patients and methods els of IL-4, suggesting a dominance of Patients Th1 immune responses with conse- Twenty-five patients, 20 women and 5 quent proinflammatory role for T cells males, median age 48 years (range 17-

317 BRIEF PAPER CD134 in active RA / R. Giacomelli et al.

75 years), who met the revised criteria (PBMC) we re sep a rated from paire d ried out with samples of PB and SF of the American Rheumatism Associa- samples of SF and PB by standard from 25 patients with RA. As shown in tion for RA (15) were enrolled in the Ficoll-Hypaque density gradient centri- Fi g u re 1, the perc e n t age of CD3+/ study. Fifteen displayed early RA (dur- fugation. Their viability, as assessed by CD134+ cells is significantly increased ation of disease less than two years); in trypan blue dye exclusion, was routine- in SF (12.5%, range 5-34%) compared 10 patients the disease had been present ly > 95%. The SF had been pre-treated to PB (2%, range 0-5%) (p < 0.001). for 4 to 25 years. Eighteen patients had for 30 min at 37¡C by the addition of CD134 expression on CD4+ cells was IgM rheumatoid factor, median titer 54 10 units/ml of ovine hyaluronidase. higher in SF (12%, ra n ge 5.7-28%) IU/ml (range 26-400) and 97.8 IU/ml than in PB (0.8%, range 0.1-4.4%) (P < (range 68-600) in the serum and SF, Flow cytometric analysis of PBMC 0.001). Similarly, the perc e n t age of respectively. The surface coexpression of OX-40 on CD134+/CD8+ cells from SF (7%, Disease activity was defined by the SF and PB mononu clear cells wa s range 0.5-12%) was significantly high- presence of at least 3 of the following determined by two-color immunofluo- er than the percentage observed in PB criteria: > 6 joints tender or painful on rescence staining, after incubation with (0.25%, range 0.1-4%) (p<0.001). Fur- motion, > 3 swollen joints, increase of a FITC or PE-conjugated anti-OX 4 0 thermore, in RA patients,an increase of We s t e rgren ery t h ro cyte sedimentat i o n MoAb (PharMinge n , La Jo l l a , C A ; CD134+ cells in SF displaying the acti- rate (ESR) > 28, morning stiffness > 45 Becton Dick i n s o n , Mountain Vi ew, vation marker CD25 (p55 IL2 receptor) minutes in duration, both patient’s and CA; respectively) in combination with was observed compared to PB cells (SF physician’s analog scale for the global A n t i - C D 3 , - C D 4 , - C D 8 , - C D 2 5 CD134+/CD25+ cells: 9%, range 1.3- assessment of disease activity > 50. MoAbs FITC or PE-conjugated (all 15% vs PB CD134+/CD25+ cells: Thirteen patients with early RA had not MoAbs were from Becton Dickinson, 0.5%, range 0.1-6.3%) (p < 0.001). Of p rev i o u s ly taken disease-modifying Montain View, CA, USA). Phenotypic interest, no differences were observed antirheumatic drugs (DMARDs), 1 was analysis was performed by flow cytom- in the percentage of CD134+/CD4+ T re c e iving sulphasalazine and another etry (FACScan; Becton Dickinson). lymphocytes in the PB of RA patients was taking 10 mg/die of prednisone. and controls (Fig. 2). All patients with chronic disease were Statistical analysis Patients with active RA had significant- receiving DMARDs (5 sulphasalazine, Data was expressed as median (range). ly higher perc e n t age of CD3+/CD 2 hy d rox y ch l o roquine and 3 metho- Because of non-normal distribution of 134+, CD4+/CD134+, CD8+/CD134+ trexate). samples, the Mann-Whitney U test for and CD25+/CD134+ than those with Twenty sex- and age-matched healthy n o n - p a i red samples and Wi l c ox o n ’s inactive disease (Fig. 3). No differences volunteers, with the same demographic s i g n e d - rank test for paired data and were observed in RA patients accord- findings of the studied populat i o n , Spearman’scorrelation coefficient were ing to the kind of therapy. Furthermore, served as controls. used for statistical analysis of the data. no diffe rences we re observed in the p e rc e n t age of CD4+/CD134+ and Isolation of synovial fluid and PBMC Results CD8+/CD134+ ly m p h o cytes betwe e n Pe ri p h e ral blood mononu clear cells I m munophenotypic studies we re car- new onset and chronic synovitis.

Table I. Clinical characteristics of RA patients. Discussion Early Chronic In this study we demonstrated increased percentages of CD134+ cells in the Age 39 ± 7 41 ± 6 synovial fluid of RA patients, due to an Disease duration (yrs) 1.6 ± 4 13 ± 11 Female 12/15 8/10 expansion of both CD4+ and CD8+ cells ex p ressing this surface marke r Groups values Active patients Non-active patients and confirm that in normal conditions (18/25) (7/25) peripheral blood T cells do not display Tender joint count Early 18.1 ± 1.3 18.7 ± 1.6¡ 1.5 ± 1.4¡ this surface antigen (5). These findings Chronic 20.0 ± 2.6 19.5 ± 2.4¡ 2.8 ± 1.1¡ are partially in agreement with a previ- Swollen joint count Early 13.6 ± 1.5 13.5 ± 1.4¡ 0.6 ± 1.3¡ ous report in which an increased per- Chronic 12.1 ± 3.1 11.8 ± 2.1¡ 0.0 ± 1.0¡ centage of SF CD134+ T cells, mainly Physician global assessment Early 60.7 ± 4.2 59.3 ± 5.5¡ 28.2 ± 2.7¡ CD4+ cells, without significant expres- Chronic 48.2 ± 4.3 45.5 ± 6.2¡ 21 ± 0.9¡ sion on CD8+ cells, were observed in a small group of RA patients (16). On the Patient global assessment Early 58.1 ± 4.1 58 ± 4.2 27.5 ± 2.2 Chronic 46.2 ± 5.2 45.1 ± 3.2 24 ± 2.9 contrary, in our study a small but significant amount of CD8+ T cells, dis- ESR (mm/hr) Early 46 ± 16 45 ± 9 27 ± 3 Chronic 48 ± 15 47 ± 4 24 ± 5 p l aying the CD134 antige n , can be observed.These data confirm that this Morning stiffness Group A 120 ± 36 125 ± 11 27 ± 10 molecule, first identified on activated Group B 80 ± 17 69 ± 4 26 ± 15 CD4+ T cells (5), can be detected, to a

318 CD134 in active RA / R. Giacomelli et al. BRIEF PAPER

process, which is characterized by the p e rsistent ex p ression of CD25 (20). These findings suggest that CD134+ T cells are involved in the immunopatho- logical process of RA synovitis, maybe mirroring some other autoimmune diseases such as EAE (11), IBD (12), aGVHD (13), in which activated, autoreactive T cell infiltrating the target tis- sues largely coexpress CD134. The ava i l able literat u re suggests the Fig. 1. The percentage importance of T cells in the induction of CD134+ cells in the and perpetuation of typical RA articu- p e ri p h e ral blood (PB) lar damage: their role in animal mod- and synovial fluid (SF) els, the efficacy of therapies directed of RA patients (*p < 0.001). against these cells and the strong association between the disease and HLA- lesser extent, on CD8+ T cells (17), DR (20). Furthermore, the Th1/Th2 ba- although the specific role of this subset lance could play a pivotal role in modu- needs to be clarified. lating the development and persistence A large amount of SF CD134+ cells of chronic inflammation of the joints c o ex p ress CD25, thus showing their (21). T cell driven cytokine production a c t ivation. Although the activat i o n is a complex mechanism involving sev- mechanisms of SF T cells occurring be- eral cell functions following activation fore entry, during transendothelial mi- (22). CD134, when engaged by agonist gration and following entry into the in- (CD 134L or a specific antibody) dur- flamed tissue are not yet completely ing antigen presentation, induces acti- clarified (18), the presence of CD25, vation of NF- B, via tumor necrosis which seems to be independent of dis- factor receptor-associated factor (23), ease duration and the type of therapy, is resulting in a costimulatory signal as % against a simple non-specific activa- potent as CD28 costimu l ation (24). t i o n , p e r h aps induced by endothelial Th u s , this process leads to enhance- Fig. 2. No differences were observed in the per- contact (19) and suggests that these ment of effector and memory T cell centage of circulating CD134+/CD4+ between cells could be activated via the CD3/ function by up-regulating cytokine pro- patients and healthy controls. TCR complex through an A g - d rive n duction and increasing the life-span of effector T cells (25). Although T cell polarization is dependent on the cytokine milieu present during T cell differ- e n t i at i o n , recent rep o rts have show n that the engagement of the CD134 molecule polarizes these cells to Th2 phe- notype (10, 21). Th u s , although the cytokine pat t e rn seen in CD134+ associated conditions seems to be more complex than with other members of the TNF re c ep t o r superfamily (i.e. CD30+/ Th2 pattern), our data may suggest an in vivo involvement of Th2 cells in the regula- tion of the immu n o l ogical pro c e s s occurring at the level of the RA joint, mirroring those observed for other Th2 markers (22, 23). The idea that the nat- u ral course of RA art h ritis could be modulated by cytokine balance is ap- pealing and although there is evidence Fig. 3. Percentages of CD134+ cells in the synovial fluid of RA patients subdivided according to dis- that Th1 activation prevails in this disease activity (p < 0.001). ease, it is thought that Th2 cells may

319 BRIEF PAPER CD134 in active RA / R. Giacomelli et al. p roduce cytokines in an attempt to 4. CUSH JJ , LIPSKY PE: Phenotypic analysis of 18. FOX DA:The role of T cells in the immuno- d ow n - m o d u l ate infl a m m ation (24). synovial tissue and peripheral blood lympho- pathogenesisofrheumatoidarthritis:New per- cytes isolated from patients with rheumatoid spectives. Arthritis Rheum 1997;40:598-609. Thus,the higher percentage of CD134+ arthritis. Arthritis Rheum 1988; 31: 1230-8. 19. SANCHO D, YANEZ-MO M, TEJEDOR R, T cells, o b s e rved in the SF of our 5. PATERSON DJ, JEFFERIES WA,GREEN MR et SANCHEZ-MADRID F: Activation of peripher- patients with active RA, could reflect a al.: Antigens of activated rat T lymphocytes al blood T cells by interaction and migration major recruitment of these activated T including a molecule of 50,000 Mr detected t h rough endothelium: Role of ly m p h o cy t e only on CD4 positive T blasts. Mol Immunol function antige n - 1 / i n t e rcellular adhesion cells at the joint levels with a possible 1987; 24: 1281-90. molecule-1 and interleukin-15. B l o o d 1 9 9 ; d ow n - reg u l at o ry rather than pro - 6. M I U R A S, O H TA N I N, N U M ATA M et al.: 93: 886-96. inflammatory effect. In fact, recent data Molecular cloning and characterization of a 20. I A N N O N E F, CORRIGALL V M , K I N G S L E Y rep o rted by one of us (RG) showe d novel glycoprotein, gp34, that is specifically GH, PANAYI GS: Evidence for the continuous induced by the human T-cell leukemia virus recruitment and activation of T cells into the that, during RA, increased amounts of type I transactivator p40tax. Mol Cell Biol joints of patients with rheumatoid arthritis. Th2 activation marke rs , m ay have a 1991; 11: 1313-25. Eur J Immunol 1994; 24: 2706-13. prognostic value in predicting a good 7. STUBER E, NEURATH M, CALDERHEAD D, 21. FELDMANN M, B R E N NAN FM, MAINI RN: FELL HP, S T ROBER W: C ross-linking of Role of cytokines in rheumatoid art h ri t i s . response to second-line therapy, repre- OX40 ligand, a member of the TNF/NGF cy- Annu Rev Immunol 1996; 14: 397-440. senting the levels of anti-inflammatory tokine family, induces proliferation and dif- 22. PAUL WE,SEDER RA:Lymphocyte responses a c t ivity of Th2 cells (25). On these ferentiation in murine splenic B cells. Immu - and cytokines review. Cell 1994; 76: 241-51 grounds CD 134 expression could be a nity 1995; 2: 507-21. 23. KAWA M ATA S, AHORI T, IMURA A , TA K A- 8. IMURA A, HORI T, IMADA K et al.:The hu- ORI-KONDO A, UCHIYAMA T: Activation of helpful marker in evaluating different man OX40/gp34 system dire c t ly mediat e s OX40 signal transduction pathways leads to therapeutic strategies including intraar- adhesion of activated T cells to vascular en- tumor necrosis factor receptor-associated fac- ticular therapy, wh o s e e ffe c t s o n s y n- dothelial cells. J Exp Med 1996;183:2185-95. tor (TRAF) 2- and TRAF5-mediated NF-kap- ovial tissue have been already studied 9. STUBER E, STROBER W: The T cell-B cell paBactivation. J Biol Chem 1998; 273:5808- interaction via OX40-OX40L is necessary for 14. using serial percutaneous synov i a l the T cell-dependent humoral immune re- 24. WEINBERG AD, VELLA AT, CROFT M:OX- biopsy samples (26). sponse. J Exp Med 1996; 183: 979-89. 4 0 : L i fe beyond the effector T cell stage. Alternatively, as observed in EAE, in 10. OHSHIMA Y, YANG LP, UCHIYAMA T et al.: Semin Immunol 1998; 10: 471-80. which CD134+ T cells exhibited T cell OX40 costimulation enhances interleukin-4 25. KA L E E BA JA R , OFFNER H, VA N D E N BA R K (IL-4) expression at priming and promotes AA,WEINBERGAD:The OX-40 receptor pro- receptor CDR3 binding motifs that are the differentiation of naive human CD4(+) T vides a potent co-stimulatory signal capable ch a ra c t e ristic of those described fo r cells into high IL-4-producing effe c t o rs . of inducing encep h a l i t ogenicity in mye l i n - myelin basic protein re a c t ivity (27), Blood 1998; 92: 3338-45. specific CD4+ T cells. Int Immunol 1998; 10: thus identifying autore a c t ive T cells 11. WEINBERG AD, BOURDETTE DN, SULLIVAN 453-61. JT et al.: Selective depletion of myelin-reac- 26. FIRESTEIN GS, PAINE MM, LITTMAN BH: which are capable of transferring EAE t ive T cells with the anti-OX-40 antibody Gene ex p ression (collage n a s e, tissue inhi- into naive Lewis rats (20), in RA, CD a m e l i o rates autoimmune encep h a l o mye l i t i s . bitor of metalloproteinases,complement, and 134+ cells could be directly involved in Nat Med 1996; 2: 183-9. HLA-DR) in rheumatoid arthritis and osteo- 12. SOUZA HS, ELIA CCS, SPENCER J, MACDON- arthritis synovium. Quantitative analysis and the synovial damage. It must be point- ALDTT: Expression of lymphocyte-endothe- effect of intra-articular corticosteroids. Arth - ed out that EAE tre atment with CD lial receptor-ligand pairs, alpha4beta7/MAd- ritis Rheum 1991; 34: 1094-105. 134:Ig-Fc ch i m e ric protein has been CAM-1 and OX40/OX40 ligand in the colon 27. FLYNN S,TOELLNER KM,RAYKUNDALIA C, able to dampen clinical signs of acute and jejunum of patients with inflammatory GOODALL M, LANE P: CD4 T cell cytokine bowel disease. Gut 1999; 45: 856-63. d i ffe re n t i at i o n : the B cell activation mole- disease and inhibit re l ap s e s , m ay b e 13. TI T T L E T V, WEINBERG A D, S T E I N K E L E R cule,OX40 ligand, instructs CD4 T cells to interfering with CD134/CD134L inter- CN, MAZIARZ RT: Expression of the T-cell express interleukin 4 and upregulates expres- action during A g - s p e c i fic T cell pri- activation antigen, OX-40,identifies alloreac- sion of the chemokine receptor, Blr-1. J Exp ming (14). These data suggest that tar- tive T cells in acute graft-versus-host disease. Med 1998; 188: 297-304. Blood 1997; 89: 4652-8. 28. GERLI R,MUSCAT C,BISTONI O et al.: High geting CD134/CD134L intera c t i o n 14. WEINBERG AD, WEGMANN KW, FUNATAKE levels of the soluble form of CD30 molecule might provide a novel therapeutic strat- C , WHITMAM R H: B l o cking OX - 4 0 / OX - 4 0 in rheumatoid arthritis (RA) are expression of egy for eliminating autoreactive CD4+ ligand interaction in vitro and in vivo leads to CD30+ T cell involvement in the inflamed T cells and may be useful in treating decreased T cell function and amelioration of joints. Clin Exp Immunol 1995; 10: 547-50. ex p e rimental allergic encep h a l o myelitis. J 29. VAN DER GRAAFF W L , P R I N S A P, D I J K- some autoimmune diseases. Immunol 1999; 162: 1818-26. MANS BA, VAN LIER RA: Prognostic value of Further studies are in progress in our 15. ARNETT FC, EDWORTHY S M, BLOCK DA et Th1/Th2 ratio in rheumatoid arthritis. Lancet laboratory in order to clarify the role of al.:The American Rheumatism Association 1998; 351: 1931. CD134+ cells in the SF of RA patients. 1987 revised criteria for the classification of 30. MIOSSEC P, VAN DEN BERG W:Th 1 / Th 2 rheumatoid arthritis. Arthritis Rheum 1988; cytokine balance in arthritis. Arthritis Rheum References 31: 315-24. 1997; 40: 2105-15. 1. PA NAY I G S:T- c e l l - d ependent pat h ways in 16. BRUGNONI D, BETTINARDI A, MALACARNE 31. GERLI R,BISTONI O, LUNARDI C et al.: Solu- rheumatoid arthritis. Curr Opin Rheumatol F, AIRO’ P, CATTANEO R: CD134/OX40 ex- ble CD30 in early rheumatoid arthritis as a 1997; 9: 236-40. pression by synovial fluid CD4+ T Lympho- p redictor of good response to second-line 2. TOYO S A K I T, T S U RU TA Y, YOSHIOKA T e t cytes in chronic synovitys. Br J Rheum 1998; t h e rapy. R h e u m at o l ogy (Oxford ) 1999; 38: al.: Recognition of rheumatoid arthritis syn- 37: 584-585. 1282-4. ovial antigen by CD4+, CD8- T cell clones 17. BAU M P R , G AYLE III R B, RAMSDELL S e t 32. BU E NA F E AC , W E I N B E R G A D, C U L B E RT- established from rheumatoid arthritis joints. al.: Molecular characterization of murine and SON NE, VA N D E N BARK A A , OFFNER H: V Arthritis Rheum 1998; 41: 92-100. human OX40/OX40 ligand systems: identifi- beta CDR3 motifs associated with BP recog- 3. MOSMANN TR, SAD S:The expanding uni- cation of a human OX40 ligand as the HTLV- nition are enriched in OX-40+ spinal cord T verse of T-cell subsets: Th1, Th2 and more. 1-regulated protein gp34. EMBO J 1994; 13: cells of Lewis rats with EAE. J Neurosci Res Immunol Today 1996; 17: 138-46. 3992-4001. 1996; 44: 562-7.

320