INTERLEUKIN-1, INTERLEUKIN-1 RECEPTOR ANTAGONIST AND MACROPHAGE POPULATIONS IN RHEUMATOID ARTHRITIS SYNOVIAL MEMBRANE

Transcription

1 British Journal of Rheumatology 1997;36:935±940 INTERLEUKIN-1, INTERLEUKIN-1 RECEPTOR ANTAGONIST AND MACROPHAGE POPULATIONS IN RHEUMATOID ARTHRITIS SYNOVIAL MEMBRANE A. CAULI, G. YANNI and G. S. PANAYI Rheumatology Unit, Division of Medicine, UMDS, Guy's Hospital, London SUMMARY To determine whether interleukin-1 (IL-1) and interleukin-1 receptor antagonist (IL-1ra) are produced by di erent macrophage subsets, we applied immunoperoxidase and double-labelling immuno uorescence techniques to 10 rheumatoid arthritis (RA) and 10 osteoarthritis (OA) synovial membranes. In RA, greater numbers of early 27E10+ macrophages were found in the sublining layer while mature 25F9+ macrophages were more abundant in the lining layers. The majority of IL-1a+ cells were also IL-1ra+ (79212% sublining layer, 9822% lining layer). In OA sublining layer, a higher percentage of cells double stained for 25F9 and IL-1ra was detected compared to those double stained for 25F9 and IL-1a (P < 0.004). In OA, 25F9+ macrophages demonstrated a lower percentage of IL-1a+ in the lining and sublining layers compared to RA (P < 0.02 and P < 0.004, respectively). It may be concluded that once monocytes have migrated into the RA joint, they undergo phenotypic and functional changes from an early pro le (27E10+, CD14+, low percentage of IL-1+ and IL-1ra+ cells) to a mature pro le (CD14+/, 25F9+, RM3/1+, high percentage of IL-1+ and IL-1ra+ cells). KEY WORDS: Rheumatoid arthritis, Osteoarthritis, Interleukin-1, Interleukin-1 receptor antagonist, Synovial membrane, Immunohistochemistry, Macrophages. Submitted 5 August 1996; revised version accepted 21 February Correspondence to: Dr G. Yanni, Rheumatology Unit, Department of Medicine, Fourth Floor, Hunt's House, Guy's Hospital, St Thomas Street, London SE1 9RT. RHEUMATOID arthritis (RA) is a chronic immunemediated in ammatory disease of unknown aetiology predominantly a ecting the synovial membrane. The characteristics of chronic synovial in ammation include vascular proliferation, the accumulation of T cells and macrophages, and synovial lining layer thickening. The events responsible for this process are thought to be due to the stimulation of T cells by unknown antigens presented in the groove of HLA- DR4, leading to macrophage activation and production of monokines such as interleukin-1 (IL-1a and IL-1b) and tumour necrosis factor-a (TNF-a) [1±3]. The release of these and other in ammatory mediators, as well as the activation of synoviocytes, chondrocytes and osteoclasts, bring about the in ammatory and destructive changes which are characteristic of the disease. Monocytes play a crucial role in these pathogenetic events. Bone marrow grafting experiments in mice have shown that the resident type A cells of the synovial lining layer are of bone marrow origin [4]. Monocytes migrate via the blood stream, through the high endothelial venules of the synovial membrane, to the lining layer. After entry into tissues, monocytes mature into macrophages and undergo phenotypic and functional changes which have been studied in vitro and in vivo [5, 6]. Peripheral blood monocytes express the CD14 antigen which participates in the binding of monocytes to cytokine-stimulated endothelial cells and thus facilitates the entry of monocytes into tissues [7]. CD14 is expressed in the bone marrow at a very early stage of monocyte di erentiation [8]. A series of monoclonal antibodies have been developed which detect early acute in ammatory macrophages (27E10) [9] as well as mature macrophages, namely 25F9 [10] and RM3/ 1. The expression of the latter is strongly enhanced after incubation of macrophages with dexamethasone [11]. In RA synovial membrane, macrophages are found abundantly [12]. Greater numbers of CD14+ and CD68+ macrophages correlate with a worse radiological outcome [13]. This is related in part to the production of high levels of pro-in ammatory monokines such as IL-1 and TNF which result in joint destruction [2]. There is increasing evidence that the function of macrophages is dependent on the stage of di erentiation reached by these cells. In in vitro studies, monocytes are potent producers of IL-1. In the presence of immobilized IgG and granulocytemacrophage colony-stimulating factor (GM-CSF), monocytes mature and produce high levels of IL-1 receptor antagonist (IL-1ra) [14±16]. The aim of this work was to compare IL-1 and IL-1ra expression by di erent macrophage populations in the RA and osteoarthritis (OA) synovial membranes, and to determine whether these macrophage subsets have a pro- or anti-in ammatory phenotype, and whether the cytokines examined are produced exclusively by a particular macrophage subset. MATERIALS AND METHODS Patients Synovial tissue was obtained from 10 patients with RA who ful lled the revised American College of 935 # 1997 British Society for Rheumatology

2 936 BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 9 TABLE I Macrophage staining in rheumatoid arthritis (RA) and osteoarthritis (OA) synovial membrane. Results are expressed on a vepoint scale (outlined in Materials and methods), as mean2s.d. Lining layer Sublining layer RA OA RA OA 27E CD F RM3/ liquid nitrogen. Samples were stored at 708C until sectioned for immunohistological staining. Fivemicrometre-thick sections were cut with a cryostat (Leitz) at 228C. Sequential sections were mounted on poly-l-lysine-coated slides, xed in acetone for 10 min, wrapped in tin foil and stored at 708C until further use. The monoclonal antibodies used to determine the cellular markers were 27E10, RM3/1, 25F9 (biotinylated monoclonal mouse IgG, a kind gift of Dr G. Burmeister, BMA Biomedicals, Switzerland) and anti-cd14 (mouse monoclonal IgG2a, a kind gift of Dr P. Beverly and Dr N. Hogg, ICRF, UK). The monoclonal antibodies used to determine cytokine expression were anti-il-1a (rabbit polyclonal 4/ 86BM, NIBSC, UK), anti-il-1b (mouse monoclonal antibody, a kind gift of Dr P. Ghiara, Italy) and anti-il-1ra (rabbit polyclonal, a kind gift of Dr W. Arend, USA). Mouse monoclonal antibody anti-il- FIG. 1.ÐDouble-immuno uorescence staining for IL-1a and IL- 1ra in rheumatoid arthritis synovial membrane. Magni cation 650. (A) IL-1a stained with TRITC (red uorescence), (B) IL-1ra stained with FITC (green uorescence), (C) double exposure which shows a high percentage of cells positive for IL-1a and IL-1ra. Rheumatology (ARC) criteria for RA [17], four men and six women, mean age 51 yr (range 20±65), and from 10 patients with OA, four men and six women, mean age 61 yr (range 50±72). One RA patient was on i.m. gold, one on penicillamine and the remaining eight were on non-steroidal anti-in ammatory drugs only. Tissue preparation and staining Synovial samples were embedded in optimal temperature cutting compound (OCT; Miles, CA, USA) and snap frozen in isopentane cooled in FIG. 4.ÐDouble-immuno uorescence staining of RM3/1 and IL- 1a (A, B and C) and RM3/1 and IL-1ra (D, E and F) in osteoarthritis synovial membrane. Magni cation 650. (A) RM3/1 stained with TRITC (red uorescence); (B) IL-1a stained with FITC (green uorescence); (C) double exposure of RM3/1 and IL- 1a; (D) RM3/1 stained with TRITC (red uorescence); (E) IL-1ra stained with FITC (green uorescence); (F) double exposure of RM3/1 and IL-1ra. Greater RM3/1/IL-1ra (C) staining is seen in the sublining layer compared to RM3/1/IL-1a (F).

3 CAULI ET AL.: IL-1 AND IL-1ra EXPRESSION IN RA 937 antibody of the same type; no staining was noted in these sections. Tonsil was used as a positive control. Con rmation of the speci city of the cytokine monoclonal antibody has been extensively investigated by the providers of the antibodies [19±21]. FIG. 2.ÐPercentage of interleukin-1 alpha (IL-1a)-positive and interleukin-1 receptor antagonist (IL-1ra)-positive macrophages in rheumatoid arthritis (RA) lining layer (upper panel) and in osteoarthritis (OA) lining layer (lower panel). 1a (Dr P. Ghiara) was used for the double-labelling immuno uorescence experiments with anti-il-1ra. As previously described by our group, a standard singlestaining indirect immunoperoxidase technique was used to examine the distribution of cytokines and macrophages [18]. For the double-labelling experiments, an indirect double-immuno uorescence technique was used. The tissue sections were incubated for 10 min with a 1:20 dilution of normal goat serum in a humidi ed chamber at room temperature. The sections were then incubated with the primary monoclonal antibody at appropriate dilutions in phosphate-bu ered saline (PBS) for 90 min. The excess monoclonal antibody was removed by washing in PBS. The second layer, tetramethylrhodamine B isothiocyanate (TRITC)-conjugated streptavidin or TRITC-conjugated goat anti-mouse (Sera Lab, UK) diluted in PBS 1:50, was added and sections were incubated for 30 min. After rinsing in PBS, the rabbit polyclonal antiserum to IL-1ra or to IL-1a was added for an incubation time of 90 min followed, after washing in PBS, by uorescein isothiocyanate (FITC)-conjugated swine anti-rabbit immunoglobulins (Dako) for 30 min. Finally, after rinsing in PBS, the sections were mounted in Vector mounting medium and analysed with a Zeiss uorescence microscope. Negative control staining was performed both in the immuno uorescence and immunoperoxidase experiments on all the specimens studied. An identical procedure was followed except that the primary antibody was substituted with an irrelevant Microscopic evaluation Microscopic evaluations were performed by one observer (AC) who was blinded to the name of the patient. Furthermore, to rule out intra-observer variation, samples from ve patients were read on three di erent occasions after randomization. No statistically signi cant di erences were observed between the three readings, with <10% variability. Samples were also assessed by another observer (GY); the results were in agreement with those of AC, with <10% variability. Distribution and single staining of the macrophage markers examined were carried out on the immunoperoxidase-stained sections. The numbers of positive macrophages were counted out of a total of 500 cells (all nucleated cells in the eld) in each of the lining and sublining layers, and expressed on a ve-point scale where 0 indicates no staining, 1 is 1±10% positively staining cells, 2 is 11±25%, 3 is 26±50%, 4 is 51±75% and 5 is 76±100%. Double staining for macrophages and cytokines was assessed on doubleimmuno uorescence-stained sections in the lining and sublining layer areas. An average of 400 cells was evaluated for each subset in all patients. Cells double stained for a macrophage marker and IL-1a or IL- 1ra were expressed as a percentage of the cells stained for the macrophage markers. The same assessment FIG. 3.ÐPercentage of interleukin-1 alpha (IL-1a)-positive and interleukin-1 receptor antagonist (IL-1ra)-positive macrophages in rheumatoid arthritis (RA) sublining layer (upper panel) and in osteoarthritis (OA) sublining layer (lower panel).

4 938 BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 9 was applied to the cells double stained for IL-1a and IL-1ra. Statistics Results are expresssed as a mean 2 S.D. Histological scores were compared using the Mann± Whitney test. RESULTS Distribution of macrophages and cytokines throughout RA and OA synovial membranes (Table I) In RA, greater 27E10 staining was noted in the sublining layer than in the lining layer (P < 0.004). CD14+ cells were expressed in equal amounts in the sublining and lining layers. 25F9+ cells were found in greater numbers in the lining layer compared to the sublining layer (P < 0.02). In OA, there were no di erences in the subset staining in the sublining and lining layers. All samples clearly reacted with anti-il-1a, IL-1b and IL-1ra antibodies, giving a di use pattern of staining. As expected, the mononuclear cells and the blood vessels within the sublining and lining layers were positively stained for cytokines. Double staining of macrophages and cytokines IL-1a and IL-1b were co-expressed on all the macrophages examined in both RA and OA. Exclusive staining of cells with either IL-1a or IL-1b was not seen. Therefore, for all the double-labelling experiments, IL-1a and not IL-1b was used. Comparison of IL-1a and IL-1ra staining by the same macrophage subsets. In RA lining layer, 9322% of IL-1a+ cells were also IL-1ra+ (Fig. 1). In RA and OA lining layer, on each of the macrophage markers examined, there was no statistically signi cant di erence in the expression of IL-1a and IL-1ra (Fig. 2). In RA sublining layer, 79212% of IL-1a+ cells were also IL-1ra+ (Fig. 1). There was no di erence in the percentage of IL-1a+ and IL-1ra+ macrophages for each subset. In OA sublining layer, a higher percentage of cells double stained for 25F9 or RM3/1 and IL-1ra was detected compared to the one double stained for 25F9 or RM3/1 and IL-1a (Fig. 3, lower panel, and Fig. 4) (25F9 P < and RM3/1 P < 0.02). Comparison of IL-1a and IL-1ra staining by di erent macrophage subsets. In RA lining and sublining layers (Figs 2 and 3, upper panels), fewer 27E10+ cells were IL-1a+ compared to CD14+ (lining layer P < 0.004, sublining layer P < 0.02), 25F9+ (lining layer and sublining layer P < 0.004) and RM3/1+ macrophages (lining layer P < and sublining layer P < 0.04). Fewer 27E10+ cells were IL-1ra+ compared to CD14+ (lining layer P < 0.004, sublining layer P < 0.05), 25F9+ (lining layer and sublining layer P < 0.004) and RM3/1+ (lining layer P < 0.004, sublining layer P < 0.007). There was no di erence in the expression of IL-1a and IL-1ra on RM3/1+ and 25F9+ cells. In OA lining and sublining layers, as in RA (Figs 2 and 3, lower panels), 27E10+ cells showed a lower percentage of IL-1ra positivity compared to 25F9+ cells (P < 0.05). Greater numbers of CD14 + IL- 1a+ (P < 0.03), 25F9 + IL-1a+ (P < 0.04) and RM3/1 + IL-1a+ cells were counted in the lining layer compared to the sublining layer. Comparison of RA and OA staining. In OA, 25F9+ macrophages demonstrated a lower percentage of IL- 1a+ cells both in the lining layer and sublining layer compared to RA (P < 0.02 and P < respectively). However, the percentages of IL-1ra+ 25F9+ and IL-1ra+ RM3/1+ cells in the lining layer and sublining layer were similar in OA and RA. DISCUSSION Our observations of greater 27E10 macrophage staining in the sublining layer compared to the lining layer, and the prevalence of 25F9 in the lining layer compared to the sublining layer in RA but not in OA, are interesting in the light of the information emerging about the synovial lining cells. Our results support the suggestion that bone marrow-derived monocytes migrate from the blood stream through the endothelial venules of the sublining layer, and there undergo maturation and phenotypic changes, from an early macrophage phenotype (27E10+, CD14+) to a late macrophage phenotype (CD14+/, 25F9+, RM3/1+). Double-immuno uorescence analysis revealed co-expression of IL-1a and IL-1b on all the macrophages examined in the lining and sublining layers, suggesting that the stimuli which trigger IL-1a and IL-1b synthesis act on all macrophage subsets. The polyclonal IL-1ra antibody used detects both intracellular and secretory IL-1ra. Isolated synovial macrophages produce primarily secretory IL-1ra, while broblast-like synoviocytes produce primarily intracellular IL-1ra [22]. Our analysis of IL-1a and IL-1ra expression by the di erent macrophage subsets showed that in RA the percentages of IL-1a+ and IL-1ra+ cells, in the lining and sublining layers, staining the same macrophage subset were not statistically di erent. In OA sublining layer, there was a higher percentage of cells stained for mature macrophages (25F9, RM3/1) and IL-1ra when compared to the double staining with IL-1a. These results con rm and extend previous in vitro observations [23] that the same cells (early or late macrophages) which produce IL-1 (a or b) also produce its antagonist, IL-1ra. In in vitro cultures, monocytes produce greater amounts of IL-1 compared to IL-1ra. Exposure of these cells to adherent IgG or GM-CSF promotes the expression of IL-1ra, indicating a quantitative switch in the production of the two forms of the cytokine during maturation [16]. Using immunohistochemical techniques, we were not able to demonstrate such clear-cut switching in cytokine production in RA or OA synovium. The di erence

5 CAULI ET AL.: IL-1 AND IL-1ra EXPRESSION IN RA 939 may lie in the amount of agonist or antagonist secreted by each single cell. Although immunohistochemistry is speci c, it does not allow accurate estimation of the amounts of IL- 1a, IL-1b and IL-1ra proteins produced and secreted by each macrophage subset. It is possible that some di erences may be detected when using in situ hybridization or reverse transcriptase±polymerase chain reaction on isolated cells. In RA, a lower percentage of 27E10+ cells was positive for IL-1a and IL-1ra in the lining and sublining layers compared to 25F9+ cells, suggesting that with the phenotypic changes of maturation, more macrophages produce both the pro-in ammatory and anti-in ammatory forms of IL-1. As previously shown [3], we found that in OA sublining layer, compared to RA, a lower proportion of mature macrophages were IL-1a+. On the other hand, the percentage of IL-1ra+ macrophages was similar in the two disease groups. These ndings may be responsible for the lower degree of in ammation seen in OA. The distribution and the localization of IL-1a and IL-1ra in the synovial membrane, and the balance between IL-1a and IL-1ra, are important in determining the degree of in ammation and the disease process within the joint. Studies have indicated that in order to block the IL-1-mediated e ects, a 100- fold excess of IL-1ra is needed [24±27], suggesting that in our RA patients the levels of IL-1ra are too low to neutralize IL-1. Furthermore, IL-1 needs to bind to a small number of receptors to activate the target cell [28, 29]. Studies in animal models suggest that IL-1 is involved in the development of joint erosions [30]. The di erences seen between RA and OA in the expression of IL-1a may be one determinant in swinging the in ammatory process towards and away from erosive disease in RA and OA, respectively. Furthermore, the greater macrophage numbers in RA synovial membrane, resulting in greater production of in- ammatory mediators, would correlate with radiological damage [13]. In conclusion, our ndings suggest that in RA synovitis, monocytes migrate into the joint and undergo phenotypic and functional changes, from an early macrophage pro le (27E10+, CD14+, low percentage of IL-1+, IL-1ra+ cells) to a late macrophage pro le (CD14+/, 25F9+, RM3/1+, high percentage of IL-1+, IL-1ra+ cells). The degree of in ammation seen may result from the balance between IL-1a and IL-1ra. ACKNOWLEDGEMENTS This study has been supported by a core support grant (U9) from the Arthritis and Rheumatism Council of Great Briatin. The authors wish to thank Dr Len Poulter and Dr Aida Condez for their kind help in setting up the immuno uorescence technique. REFERENCES 1. Panayi GS. The immunopathogenesis of rheumatoid arthritis. Rheumatol Rev 1992;1:63± Arend WP, Dayer J-M. Cytokine and cytokine inhibitors or antagonists in rheumatoid arthritis. Arthritis Rheum 1990;33:305± Deleuran BW, Chu CQ, Field M, Brennan FM, Katsikis P, Feldmann M et al. Localisation of interleukin-1, type 1 interleukin-1 receptor and interleukin-1 receptor antagonist in the synovial membrane and cartilage/pannus junction in rheumatoid arthritis. Br J Rheumatol 1992;31:801±9. 4. Edwards JCW, Willoughby DA. Demonstration of bone marrow derived cells in synovial lining by means of giant intracellular granules as genetic markers. Ann Rheum Dis 1982;41:177± Janossy G, Panayi GS, Duke O, Bo ll M, Poulter LW, Goldstein G. Rheumatoid arthritis: a disease of T lymphocyte/macrophage immunoregulation. Lancet 1981;ii:839± Ridley MG, Kingsley GH, Pitzalis C, Panayi GS. Monocyte activation in rheumatoid arthritis: evidence for in situ activation and di erentiation in joints. Br J Rheumatol 1990;29:84±8. 7. Beekhuizen H, Blokland I, Corsel-van Tilburg AJ, Koning F, van Furth R. CD14 contributes to the adherence of human monocytes to cytokine-stimulated endothelial cells. J Immunol 1991;147:3761±7. 8. Goyert SN, Tesio L, Ashman LK et al. M3. 1. Report on CD14 cluster workshop. In: Knapp W et al. eds. Leucocyte typing IV. White cell di erentiation antigens. Oxford: Oxford University Press, 1989:789± Zwadlo G, Schlegel R, Sorg C. A monoclonal antibody to a subset of human monocytes found only in the peripheral blood and in ammatory tissues. J Immunol 1986;137:512± Zwadlo G, Brocker EB, von Bassewitz DB, Feige U, Sorg C. A monoclonal antibody to a di erentiation antigen present on mature human macrophages and absent from monocytes. J Immunol 1985;134:1487± Zwadlo G, Voegeli R, Ostho KS, Sorg C. A monoclonal antibody to a novel di erentiation antigen on human macrophages associated with the down-regulatory phase of the in ammatory process. Exp Cell Biol 1987;55:295± Highton J, Palmer DG, Panayi GS, eds. The mononuclear phagocyte and rheumatoid arthritis. In: Immunology of the connective diseases. Dordrecht: Kluwer Academic Press, 1993:43± Yanni G, Whelan A, Feighery C, Bresnihan B. Synovial tissue macrophages and joint erosions in rheumatoid arthritis. Ann Rheum Dis 1994;53:39± Arend WP, Joslin FG, Massoni RJ. E ects of immune complexes on production by human monocytes of interleukin 1 or interleukin 1 inhibitor. J Immunol 1985;134:3868± Roux-Lombard P, Modoux C, Dayer JM. Production of Interleukin 1 (IL-1) and a speci c IL-1 inhibitor during human monocyte-macrophage di erentiation: in uence of GM-CSF. Eur Cytokine Netw 1989;1:45± Janson RW, Hance KR, Arend WP. Production of IL- 1 receptor antagonist by human in vitro-derived macrophages. E ects of lipopolysaccharide and granulocytemacrophage colony-stimulating factor. J Immunol 1991;147:4218± Arnett FC, Edworthy SM, Bloch DA et al. The

6 940 BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 9 American Rheumatism Association 1987 revised criteria for the classi cation of rheumatoid arthritis. Arthritis Rheum 1988;31:315± Farahat MN, Yanni G, Poston R, Panayi GS. Cytokine expression in synovial membranes of patients with rheumatoid arthritis and osteoarthritis. Ann Rheum Dis 1993;52:870± Thorpe R, Wadh WA, Glearing AJH, Mahon B, Poole S. Sensitive speci c immunoradiometric assay for human interleukin-1 alpha. Lymphokine Res 1988;7: 119± Boraschi D, Volpini G, Villa L, Nencioni L, Scapigliati G, Nucci D et al. A monoclonal antibody of the Interleukin-1 beta peptide blocks adjuvanticity but not pyrogenicity of IL-1 beta in vivo. J Immunol 1989;143:131± Malyak M, Joslin FG, Verderber EL, Eisemberg SP, Arend WP. IL-1ra ELISA: reduction and alkylation of synovial uid eliminates interferences by IgM rheumatoid factors. J Immunol Methods 1991;140:281± Firestein GS, Boyle DL, Yu C, Paine MM, Whisenand TD, Zvai er NJ et al. Synovial interleukin-1 receptor antagonist and interleukin-1 balance in rheumatoid arthritis. Arthritis Rheum 1994;37:644± Andersson J, BjoÈ rk L, Dinarello CA, Towbin H, Andersson U. Lipopolysaccharide induces human interleukin-1 receptor antagonist and interleukin-1 production in the same cell. Eur J Immunol 1992;22:2617± Arend WP, Welgus HG, Thompson RC, Eisenberg SP. Biological properties of recombinant human monocytederived interleukin 1 receptor antagonist. J Clin Invest 1990;85:1694± Arend WP, Coll BP. Interaction of recombinant monocyte-derived interleukin 1 receptor antagonist with rheumatoid synovial cells. Eur Cytokine Netw 1991; 3:407± Smith RJ, Chin JE, Sam LM, Justen JM. Biologic e ects of an interleukin 1 receptor antagonist protein on interleukin-1 stimulated cartilage erosion and chondrocyte responsiveness. Arthritis Rheum 1991;34:78± Seckinger P, Klein-Nulend J, Alander C, Thompson RC, Dayer JM, Raisz LG. Natural and recombinant human IL-1 receptor antagonists block the e ects of IL-1 on bone resorption and prostaglandin production. J Immunol 1990;145:4181± Dower SK, Call SM, Gillis S, Urdal DL. Similarity between the interleukin 1 receptors on a murine T-lymphoma cell line and on a murine broblast cell line. Proc Natl Acad Sci USA 1986;83:1060± Orencole F, Dinarello CA. Characterization of a subclone (D10S) of the D10.G4.1 helper T-cell line which proliferates to attomolar concentrations of interleukin- 1 in the absence of mitogens. Cytokine 1989;1:14± Pettipher ER, Higgs GA, Henderson B. Interleukin 1 induces leucocyte in ltration and cartilage proteoglycan degradation in the synovial joint. Proc Natl Acad Sci USA 1986;83:8749±53.