Levels of Soluble CD27 in Sera and Synovial Fluid and Its Expression On

Clinical and Experimental Rheumatology 2002; 20: 863-866. PEDIATRIC RHEUMATOLOGY Levels of soluble CD27 in ABSTRACT man inflammatory arthritides, includ- sera and synovial fluid O b j e c t ive. CD27 is a member of ing juvenile idiopathic arthritides (JIA) tumour necrosis factor receptor family. (1-3). and its expression on Its ex p ression is pre d o m i n a n t ly con - CD27 is a member of the tumour ne- memory T cells in fined to mat u re ly m p h o cytes and is crosis factor receptor (TNFR) family. patients with juvenile strongly enhanced after cell activation. Its ex p ression is pre d o m i n a n t ly con- idiopathic arthritides Shedding of the CD27 from the surface fined to mat u re ly m p h o cytes and is of activated cells is re l ated to their strongly enhanced after cell activation effector phase.The aim of the present (4). This molecule has been shown to 1 2 M. Gattorno , I. Prigione , study was to evaluate the levels of solu - stimulate T cell proliferation and B cell S. Vignola3, F. Falcini4, ble CD27 in sera and synovial fluids, differentiation through the interaction S. Chiesa2, F. Morandi2, together with its expression on circulat - with its specific ligand, namely CD70. P. Picco1, A. Buoncompagni1, ing and synovial fluid (SF) memory T Pe rsistent antigenic stimu l ation in- cells, in children with JIA. duces the proteolytic cleavage of the 1 2 A. Martini , V. Pistoia Methods. Sera from 40 patients with soluble 32 kDa form of CD27 from the a c t ive JIA we re studied for solubl e cell surface (4). According to previous 1 Second Department of Pediatrics CD27. Paired SF samples were avail - studies such process correlates with the (Rheumatology Unit), 2Laboratory of Oncology and 3Third Department of Pedi- able in 20 patients. Sera from 12 age- degree of T cell activation, both at the atrics (Gastroenterology Unit); “G. Gasli- matched patients affected with various systemic and the local level. In fact, ni” Institute for Children, Genoa; 4Clinics acute infectious diseases and 12 age- increased levels of soluble CD27 have of Pediatrics, Florence, Italy. matched healthy subjects were used as been found both in sera of renal trans- Marco Gattorno, MD; Ignazia Prigione, controls. In 8 JIA patients freshly iso - plant recipients (5) and in cerebrospi- PhD; Silvia Vignola, MD; Fernanda Falci- lated circulating and SF lymphocytes nal fluid of patients affected with mul- ni, Associate Professor of Pediatrics; Sab- were stained for CD27 in CD4+CD45 tiple sclerosis (6). rina Chiesa, PhD; Fabio Morandi; PhD; Paolo Picco, MD; Antonella Buoncompag- RO+ T cell subpopulation and ana - According to the model proposed by ni, MD; Alberto Martini, Professor of Pedi- lyzed by cytometry. Van Lier et al., naïve T cells constitu- atrics; Vito Pistoia, MD. R e s u l t s . S o l u ble CD27 serum leve l s tively express CD27, which is up-regu- Please address correspondence and reprint we re signifi c a n t ly higher in pat i e n t s l ated fo l l owing antigen pre s e n t at i o n requests to: Marco Gattorno, MD, 2nd with polyarticular JIA and acute sys - and T cell activation. Interactions be- Department of Pediatrics, “G. Gaslini” temic infectious diseases than in tween the T cell co-stimulatory mole- Scientific Institute for Children and Uni- patients with active oligoarticular or cule CD28 and its ligands (CD80/CD versity of Genoa, Largo G. Gaslini 5, h e a l t hy controls. Both polya rt i c u l a r 86) on antigen presenting cells (APC) 16147, Genoa, Italy. and oligoarticular JIA patients showed is followed by the release of pro-in- E-mail: marcogattorno@ospedale-gaslini.ge.it increased levels of soluble CD27 in SF flammatory cytokine from APC, which when compared with paired seru m eventually results to the expression of Received on February 28, 2002; accepted in revised form June 4, 2002. samples (p = 0.01). In all the patients CD70 on activated T cells. The CD27- tested a significant enri chment of CD70 interaction will further support © Copyright CLINICAL AND EXPERIMENTAL RHEUMATOLOGY 2002. CD27Ð T cells was seen in the SF clonal expansion of antigen specific T (median 39.5%, range 18-56%) when cells, the majority of which will subse- Key words: Juvenile idiopathic compared to paired CD4+CD45RO+ quently differentiate into effector cells, arthritides, memory T cells, activation, pe ri p h e r al lym p h o c ytes (median 19.5%, shedding the CD27 from their surface CD27. range 5-43%; p = 0.01). in fluid phase. C o n cl u s i o n s . A clear enri chment of Conversely, a subset of antigen-specific CD4+ memory SF T cells with a CD27- T cells that does not differentiate into phenotype is observed when compared effector cells circulate in the blood as to correspondent circulating T lympho - CD27+ “resting” memory T cells, rea- cytes. This issue is conceivably related dy to undergo to re-activation upon sec- to re - a c t ivation and re c ruitment of o n d a ry ex p o s u re to antigen (“early ” memory T cells to the site of inflamma - m e m o ry) and subsequent diffe re n t i a- tion, and to the subsequent expansion tion into “ e ffe c t o r ” C D 2 7- m e m o ry of a subpopulation of “effector” memo - cells (7). ry T cells. The concept the CD4+CD45RO+CD27- cell subset associates with the effector Introduction phase of memory T cells differentiation T lymphocytes are thought to play a is supported by the following observa- pivotal role in the initiation and perpet- tions: i) it is a potent inducer of B cell uation of chronic synovitis in the hu- differentiation (and Ig production) (8);

863 PEDIATRIC RHEUMATOLOGY CD27 in juvenile idiopathic arthritides / M. Gattorno et al. ii) it is enriched for cells producing IL- Pa i red synovial fluid (SF) samples weeks before the examination, together 4 or IFN-g ( 9 , 10); iii) it ex p re s s e s were collected after therapeutic arthro- with clinical or laboratory (i.e. elevated organ-specific homing receptors, such centesis in 20 patients (7 with poly- ESR or CRP) signs of inflammation at as CLA and the a Eb7 integrin (11, 12). a rticular and 13 with oligo a rt i c u l a r the moment of the study, were consid- Soluble CD27 has been found to be sig- course). ered as criteria of exclusion. n i fi c a n t ly increased in the synov i a l A number of clinical (no. of active S o l u ble CD27 in sera and SF wa s fluid of patients affected with rheuma- joints, no. of joints with limited range determined using a commercial ELISA toid art h ritis (RA) in comparison to of motion and physician global assess- kit (CLB,The Netherlands), according patients with osteoarthritis (13). More- ment of overall disease activity) and to the manufacturer’s instructions. over, CD27 has been shown to be ex- laboratory parameters (erythrocyte sed- Expression of CD27 on peripheral and pressed in different proportion and dis- imentation rate, ESR; C-reactive pro- SF memory T cells. Peripheral blood tribution, particularly in the subset of tein, CRP; white blood cell, WBC and (PB) and synovial fluid (SF) mononu- me m o r y T cells (CD4+CD45RO+) both platelet, PLT, counts; hemoglobin ser- clear cells were isolated from heparin- in the synovial fluid and in the inflam- um concentration, Hb) of disease activ- ized blood and synovial samples from 8 ed synovial tissue from RA pat i e n t s ity we re re c o rd e d, t ogether with the JIA patients (2 poly and 6 oligo) by (13, 14). ongoing treatment, at the time of the Ficoll (Biochrom KG-Germany) densi- Aim of the present study was to evalu- study (Table I). ty gradient centrifugation. PB mononu- ate the levels of soluble CD27 in sera Disease activity was defined by the pre- clear cells were also isolated from two and synovial fluids, together with its sence of active arthritis (swelling or, if age - m at ched healthy controls. Cells expression on circulating and synovial swelling was ab s e n t , l i m i t ation of were washed and a three-color staining fluid memory T cells, f rom ch i l d re n motion with tenderness) at least in one was performed with CD45RO-tricolor affected with JIA. joint at the moment of clinical exami- mAb (Caltag - B u rl i n ga m e - C a ) , C D 4 - nation and a physician global estimate FITC or -PE mAb (BD Biosciences- Patients and methods of disease activity (measured on a 0-10 Mountain View-Ca) and CD27-PE or - Soluble CD27 serum and SF cm visual analog scale) higher than 1 FITC mAb (PharMingen, San Diego, concentrations (16,17). CA). Cells were incubated with saturat- Sera from 40 consecutive patients af- Se r a from 12 previ o u s l y healthy age- ing amounts of mAbs for 30’ at 4¡C, fected with active JIA according to mat c hed patients affected with vari o u s then cells we re washed in PBS Durban criteria were studied between acute feb rile infectious conditions (4 (Biowhittaker-Walkersville-USA) with (15). Fifteen patients displayed a pol- with bacterial pneumonia, 3 with strept o - 1% FCS (Biochrom) and analyzed by yarticular course of the disease (5 with coccal pharyn gi t i s , 2 with phyel o n ep h r i- flow cytometry (FACScan-BD Biosci- systemic onset, 4 with RF-negative pol- ti s , 2 with acute mononucl e o s i s , 1 with ences) gating on the CD4+CD45RO+ T ya rticular onset, 1 with RF-positive Salmonella gas t ro e n t e r itis) wer e used as cell subset. PE-, FITC- and tri c o l o r- polyarticular onset and 5 with extended po s i t i ve, “i n fl a m m at o ry ” co n t r ols. c o n j u gated mouse Ig control mAbs oligoarticular course). Twenty-five had Twe l ve age - m at ched healthy subjects ( C a l t ag) we re used as negat ive con- an oligoarticular course. A serum sam- attending at our clinic for routinary pre- t rols. CellQuest softwa re (BD Bio- ple was collected with permission at o p e rat ive (orch i o p ex y, phimosis cor- sciences) was used for analyses. each control visit or at the moment of rection …) examination were used as h o s p i t a l i z ation for a disease fl a re - u p negative controls after informed con- Statistical analysis and stored at -80¡ immediat e ly after sent of the parents. History of inflam- S e rum levels of soluble CD27 we re centrifugation. matory or infectious disorders in the 4 compared among four subgroups of pa-

Table I. Clinical and laboratory features (mean, range) of JIA patients at the moment of the study.

Disease No. of joints Physician’s Erythrocyte duration active/limited global sedimentation Age (yrs.) range of motion index rate Treatment

NSAID, MTX (8 pts.) Polyarticular course 7.2 3.7 7.2 /12.2 8.5 76 NSAID, CS, MTX (3 pts.) (15 pts) (3.3 Ð 16.1) (0.5 Ð 11.2) (1-16)/(1-29) (4.6 Ð 10) (23-128) NSAID, CS (2 pts.) NSAID alone (2 pts.)

NSAID alone (12 pts.) Oligoarticular course 8.5 2.4 1. 3/1.5 6.8 28.5 NSAID, CS (5 pts.) (25 pts) (3 Ð 17.9) (0.3 Ð 14) (1-3)/ (1-3) (3 Ð 9.1) (7 Ð 76) NSAID, MTX (3 pts.) Nihil (5 pts.)

NSAID: non steroidal anti-inflammatory drugs; MTX: methotrexate; CS: corticosteroids.

864 CD27 in juvenile idiopathic arthritides / M. Gattorno et al. PEDIATRIC RHEUMATOLOGY tients (active polyarticular JIA patients, In polyarticular JIA, sCD27 serum con- enrichment of CD27Ð T cells was seen active oligoarticular JIA patients, acute centrations displayed a significant cor- in the SF (median 39.5%, range 18- febrile infectious disease, healthy con- relation with some clinical (physician 56%) when compared to paired CD4+ t rols) using the non-para m e t rical U global estimate of disease activity r = C D 4 5 RO+ peri p h e ral ly m p h o cy t e s Ma n n - Wh i t n e y test. Correl a tions among 0.85, p = 0.001) and laboratory (ESR: r (median 19.5%, range 5-43%; p = 0.01) all the variables considered were ana- = 0.61, p = 0.03; CRP: 0.63, p = 0.02) (Fig. 2). No difference was found in the lyzed using the non-parametric Spear- p a ra m e t e rs of disease activ i t y. Con- ex p ression of CD27 on CD4+CD45 man rank test. Differences in soluble versely, no statistically significant cor- RO+ circulating T cells among JIA pa- CD27 concentrations between paire d re l ations we re found in the active tients and age - m at ched healthy con- serum and SF and difference of the per- oligoarticular subset. trols (81% in control 1, 86% in control centage of CD27 positive cells in the Both polya rticular and oligo a rt i c u l a r 2). CD 4 + C D 4 5 R O+ subpopulation betwee n JIA patients showed a signifi c a n t ly p e ri p h e ral and SF ly m p h o cytes we re increased levels of sCD27 in SF (medi- Discussion d e t e rmined with the Wi l c oxon ra n k an: 1013 U/mL, range 323-2580 and Activated T cells are thought to play a test. 1318 U/mL, range 117-2410; respec- crucial role in the pathogenesis of JIA, t ive ly) when compared with paire d through their role in the regulation of Results serum samples (p = 0.01). the inflammatory response after anti- Soluble CD27 serum and SF concen- gen presentation. In the present study, trations Expression of CD27 on peripheral CD27 was used both as a marker of Soluble CD27 serum levels were sig- blood and SF memory T cells ac t i vation and functional cha ra c t e ri z a - nificantly higher both in patients with As expected, an enrichment of CD4+ tion of T cells during the active phase of polyarticular JIA (median 415.7 U/mL, CD45RO+ memory T cells was found the disease. In parti c u l a r , ac c o r ding to ra n ge 206-864) or in acute systemic in the SF of patients with JIA (median some Aut h o rs , the irrevers i b le sheddi n g infectious diseases (437 U/mL, range 34%, range 22-50%) when compared of the CD27 from the cellular surface is 220-750) than in patients with active to paired peripheral blood mononuclear co n s i d e r ed as a rel i a ble marker of their o l i go a rticular JIA (247 U/mL, ra n ge cells (median 8%, range 4-14%, p = e ffector phase for re c e n t ly activat e d 100-800; p=0.05 and p= 0.004, respec- 0.01). Expression of CD27 was detect- na ï v e and memory T cells (7, 8, 18). tively) or healthy controls (198 UI/mL, ed both on peripheral and SF memory In the present study, similarly high lev- range 88-383 U/mL; p = 0.04 and p = (CD4+CD45RO+) T cells. The CD27+ els of soluble CD27 were seen both in 0.004, respectively) (Fig. 1, panel A). subset was the most represented pheno- the JIA patients with a polya rt i c u l a r No statistical diffe rences we re noted type, both in peripheral blood (median course and in the subjects with acute between patients with active polyartic- 8 0 . 5 . 5 % , ra n ge 56-95%) and in SF systemic infections. Convers e l y, no sig- ular JIA and children with infectious (median 60.5%, range 45-82%) memo- n i ficant diffe rence was observed be- d i s e a s e s , and between oligo a rt i c u l a r ry T cells (Fig. 1, panel B). However, in t ween patients with a pre d o m i n a n t ly JIA patients and healthy controls. all the patients tested a signifi c a n t local inflammation (oligoarticular JIA) and healthy controls. Thus, it is conceivable that serum concentrations of s o l u ble CD27 could be aspecifi c a l ly related to the degree of systemic inflammation (7). Even if the JIA sub- types with a polyarticular course could not be properly considered as systemic diseases, in these conditions the acute phase reactants are often related to the degree of disease activity (19). In this line, the correlation of soluble CD27 serum levels with some disease activity parameters in the polyarticular patients may be aspecifically related to the acti- Fig. 1. Panel A. Serum levels of soluble CD27 in active polyarticular and oligoarticular JIA patients, vation of circulating naïve and memory patients with an acute systemic infection and healthy controls. Soluble CD27 serum levels were signif- T cells (7). icantly higher both in patients with polyarticular JIA or in acute systemic infectious diseases than in Thus, in order to better elucidate the patients with active oligoarticular JIA (p = 0.05 and p = 0.004, respectively) or healthy controls (p = 0.04 and p = 0.004, respectively). Panel B. Expression of CD27 on CD4+CD45RO+ synovial fluid role of CD27 as a specific marker of T (SF) lymphocytes in comparison with paired peripheral blood (PB) lymphocytes. cell differentiation at the site of tissue For both figures:bold horizontal lines represent median values. Boxes contain the 50% of values falling inflammation, we have investigated the between the 25th and 75th percentiles, whiskers lines that extend from the boxes represent the highest expression of CD27 on synovial fluid and lowest values for each subgroups. memory CD4+ T cells, that represent

865 PEDIATRIC RHEUMATOLOGY CD27 in juvenile idiopathic arthritides / M. Gattorno et al. the main T cell subpopulation found in PB SF PB SF the synovial tissue in the idiopat h i c chronic arthritides (1). Even on a relative small number of pa- ti e n t s , a cle a r -cut enric hment of CD4+ CD 4 5 RO+CD27- T cells was observed in SF when compared to correspondent circulating T lymphocytes. In this line, the high concentrations of solubl e CD27 found in SF due to an increased s h e dding from the cells surface may represent a marker of local activation of memory T cells at that level (7). This issue is in line with previous studies on adult RA (13, 14) and is conceivably related to re-activation and recruit- Fig. 2. Three-color flow cytometric analysis of paired peripheral blood (PB) and synovial fluid (SF) ment of memory T cells to the site of lymphocytes from two representative JIA patients with oligoarticular (pt. 1) and polyarticular (pt. 2) i n fl a m m at i o n , and to the subsequent course. Cells were stained simultaneously with CD45RO-tricolor mAb in combination with CD4-FITC expansion of a subpopulation of “effec- and CD27-PE (pt. 1),or CD4-PE and CD27-FITC (pt. 2). Panel A.The dot plots show the expression tor” memory T cells (18). of CD45RO and CD4 by PB and SF lymphocytes; gates were set on double positive CD45RO+,CD4+ cells. Panel B.The expression of CD27 by the gated CD45RO+,CD4+ cells shown in A was analysed. N o t ably, the immu n o h i s t o ch e m i c a l Histograms indicate the percentage of CD27+ (right bars) and CD27- (left bars) cells in the gated pop- ch a ra c t e ri z ation of rheumatoid syn- ulations. The percentage of positive and negative cells for the different Ag was calculated on the basis ovial tissue in adult RA has shown a of appropriate isotype-matched control mAbs. p revalent localization of CD4+CD45 RO+CD27+ T cells in the perivascular with juvenile idiopathic art h ritis. A rt h ri t i s 11. BAARS PA, MAUIRCE MM, REP M, HOOIB- lymphocytic aggregates, with a relative Rheum 2000; 43: 765-74 RINK B, VAN LIER RAW: Heterogenity of the increase of the CD4+CD45RO+CD27- 4. HINTZEN RQ, DE JONG R,HACK CE et al.:A circulating human CD4+ T cell population. J soluble form of the human T cell differentia- Immunol 1995; 154: 17-25. T cells in diffuse ly m p h o cytic infi l- tion antigen CD27 is released after triggering 12. DE RIE MA, CAIRO I, VAN LIER RAW, BOS trates. Thus, it is conceivable that re- of TCR/CD3 complex. J Immunol 1991; 147: JD: Expression of the T-cell activation anti- c e n t ly re - a c t ivated CD4+CD45RO + 29-35. gens CD27 and CD28 in normal and psoria- CD27+ are recruited from peripheral 5. HINTZEN RQ, VAN LIER RAW, KUIJPER KC sic skin. Clin Exp Dermat 1996; 21: 104-11. et al.: Elevated levels of a soluble form of the 13. KOHEM CL, BREZINSCHEK RI, WISBEY H, blood into the synovium to undergo T cell activation antigen CD27 in cere- TO RTO R E L L A C , LIPSKY PE, O P P E N- further cycles of activation and differ- brospinal fluid of multiple sclerosis patients. HEIMER-MARKS N: Enrichment of differen- entiation to CD4+CD45RO+CD27- T J Neuroimmunol 1991; 211-17. tiated CD45RBdim, CD27- memory T cells in cells, followed by migration in the con- 6. C H A M U L E AU M E D, T E N BERGE IJM, the peripheral blood, synovial fluid and syn- SCHELLEKENS PTA,WILMINK JM, HINTZEN ovial tissue of patients with rheumatoid arth- text of synovial tissue, where they are RQ, VAN LIER RAW: Serum levels of soluble ritis. Arthritis Rheum 1996; 39: 844-54. thought to exert their effector functions CD27 in renal transplant recipients. Trans - 14. TA K P P, H I N T Z E N R Q, T E U N I S S E N J J M e t (14). plantation 1992; 54: 932-6. a l .: E x p ression of the activation antige n A further analysis of the functional 7. LENS SMA, TESSELAAR K, VAN OERS MHJ, CD27 in rheumatoid arthritis. Clin Immunol VAN LIER RAW: Control of lymphocyte func- Immunopathol 1996; 80: 129-38. properties of CD27+ and CD27- mem- tion through CD27-CD70 interaction. Semin 15. PE T T Y R E , S O U T H WO O D T R , BAU M J e t ory T cells in JIA is currently ongoing Immunol 1998; 10: 491-9. al.: Revision of the proposed classification in our laboratory. 8. TORTORELLA C, SCHULZE-KOOPS H, THO- criteria for juvenile idiophatic arthritis: Dur- MAS R et al.: Expression of CD45RB and ban, 1997. J Rheumatol 1998; 25: 1991-4. CD27 identifies subsets of CD4+ memory T 16. GIANNINI EH, RU P E RTO N, R AVELLI A , References cells with different capacities to induce B cell LOVELL DJ, FELSON DT, MARTINI A: Pre- 1. PA NAY I G S, L A N C H BU RY J S, K I N G S L E Y di f fere n t i a tion. J Immun o l 1995; 155: 14 9 - 6 2 . liminary definition of improvement in juve- GH:The importance of the T cell in initiating 9. DE JO N G R , B ROUWER M, H O O I B R I N K B, nile art h ritis. A rt h ritis Rheum 1997; 40: and maintaining the ch ronic synovitis of VAN DER POUW-KRAAN T, MIEDEMA F, VAN 1202-9. rheumatoid arthritis. Arthritis Rheum 1992; LIER RAW:The CD27- subset of peripheral 17. RU P E RTO N, RAV E L L I A , M I G L I AVACCA D 35: 729-35. blood memory CD4+ lymphocytes contains et al.: Responsiveness of clinical measures in 2. GATTORNO M, FACCHETTI P, GHIOTTOF et f u n c t i o n a l ly diffe re n t i ated T ly m p h o cy t e s children with oligoarticular juvenile chronic al.: Synovial fluid T cell clones from oligo- that develop by persistent antigenic stimula- arthritis. J Rheumatol 1999; 26: 1827-30. articular juvenile arthritis patients display a tion in vivo. Eur J Immunol 1992; 22: 993- 18. LOENEN WAM: CD27 and TNFR relatives in prevalent Th1/Th0 patterns of cytokine se- 99. the immune system: their role in health and c retion irre s p e c t ive to immu n o p h e n o t y p e. 10. ELSON LH, NUTMAN T B, METCALFE DD, disease. Sem Immunol 1998; 10: 417-22. Clin Exp Immunol 1997; 109: 4-14. P RUSSIN C: F l ow cy t o m e t ric analysis fo r 19. RUPERTO N, RAVELLI A, FALCINI F et al.: 3. WEDDERBURN LR, ROBINSON N, PATEL A, cytokine production identifies T helper 1, T R e s p o n s iveness of outcome measures in VA R S A N I H , WO O P: S e l e c t ive rec ru i t m e n t helper 2 and T helper 0 cells within the hu- j u venile ch ronic art h ritis. Italian Pe d i at ri c of polarized T cells expressing CCR5 and man CD4+CD27- lymphocytes 1995; 154: Rheumatology Study Group. Rheumatology CXCR3 to the inflamed joints of children 4294-301. (Oxford) 1999; 38:176-80.

866