IL-1 molecules are encoded by two distinct genes, IL-1α and IL-1β. Both isoforms possess essentially identical activities and potencies, whereas IL-1α, in contrast to IL-1β, is known to act as a membrane-associated IL-1 (MA-IL-1) and plays an important role in a variety of inflammatory situations. The transgenic (Tg) mouse line (Tg1706), which was generated in our laboratory, overexpresses human IL-1α (hIL-1α) and exhibits a severe arthritic phenotype characterized by autonomous synovial proliferation with subsequent cartilage destruction. Because the transgene encoded Lys64 to Ala271 of the hIL-1α amino acid sequence, Tg mice may overproduce MA-IL-1 as well as soluble IL-1α. The present study investigated whether MA-IL-1 contributes to synovial proliferation and cartilage destruction in the development of arthritis. Flow cytometric analysis revealed that both macrophage-like and fibroblast-like synoviocytes constitutively produce MA-IL-1. D10 cell proliferation assay revealed MA-IL-1 bioactivity of paraformaldehyde-fixed synoviocytes and the further induction of endogenous mouse MA-IL-1 via autocrine mechanisms. MA-IL-1 expressed on synoviocytes triggered synoviocyte self-proliferation through cell-to-cell (i.e., juxtacrine) interactions and also promoted proteoglycan release from the cartilage matrix in chondrocyte monolayer culture. Interestingly, the severity of arthritis was significantly correlated with MA-IL-1 activity rather than with soluble IL-1α activity or concentration of serum hIL-1α. Moreover, when the Tg1706 line was compared with the Tg101 line, which selectively overexpresses the 17-kDa mature hIL-1α, the severity of arthritis was significantly higher in the Tg1706 line than in the Tg101 line. These results suggest that MA-IL-1 contributes to synoviocyte self-proliferation and subsequent cartilage destruction in inflammatory joint disease such as rheumatoid arthritis.

Rheumatoid arthritis (RA)3 is characterized by a permanently proliferative synovium, leading to the formation of hyperplastic synovial tissue (pannus) that invades both cartilage and bone. Human IL-1α (hIL-1α) transgenic (Tg) mice overexpressing hIL-1α exhibit macrophage- and neutrophil-dominant arthritis characterized by marked synovial proliferation and progressive cartilage destruction, resembling RA with a progressive phenotype. Histopathological analysis of synovial joints from hIL-1α Tg mice has demonstrated that proliferative synovium directly invades the cartilage, ultimately destroying both cartilage and underlying bone (1). As IL-1 is known to play a pivotal role in the pathogenesis of RA, analysis of IL-1-mediated synovial proliferation and subsequent invasion of the cartilage may elucidate the mechanisms of joint destruction and suggest new therapies for RA.

IL-1 molecules are encoded by two distinct genes, IL-1α and IL-1β. Both genes initially produce precursor polypeptides with a predicted Mr of 31 kDa. IL-1α precursor is fully biologically active and acts as a membrane-associated IL-1 (MA-IL-1), whereas IL-1β precursor displays no biological activity until it has been processed to form the 17-kDa mature form (2, 3). Unlike other secreted proteins, IL-1α precursor lacks a hydrophobic leader sequence (4) and is never found in organelles involved in the classical secretory pathway. The processing and release of IL-1α demonstrate atypical regulation through a number of post-translational modifications (5, 6, 7), and the exact processes vary between different cell types (8, 9, 10). Our detailed analysis of hIL-1α Tg mice revealed that among various cell types, synoviocytes are the predominant cells producing both precursor and processed forms of hIL-1α despite the use of ubiquitous CAG promoter. This preferential distribution of hIL-1α in synoviocytes seems at least partially due to the extended retention of MA-IL-1 in these cells (1).

In certain situations, IL-1α reportedly acts preferentially as MA-IL-1 (11), which was first described as IL-1 bioactivity within paraformaldehyde (PFA)-fixed macrophage or purified macrophage membranes (12). The presence of IL-1α has subsequently been demonstrated on the surface of various cell types (13, 14, 15, 16, 17, 18, 19). A wide spectrum of biological properties has also been reported, including induction of autonomous proliferation in vascular smooth muscle cells (20), T cell activation during Ag presentation (21), up-regulation of monocyte/macrophage-mediated tumor cytotoxicity (22), and stimulation of osteoclast formation (23), where cell-to-cell (i.e., juxtacrine) interactions play a key role in these actions.

The hIL-1α Tg mouse line established in our laboratory was designed to integrate a 660-bp HindIII/HincII restriction fragment of hIL-1α cDNA coding Lys64 to Ala271 of the hIL-1α amino acid sequence in an attempt to overproduce both pro and mature forms of IL-1α. As the transgene includes a nuclear localization sequence (aa 79–86) that has been shown to be important for IL-1α association with the plasma membrane (24), MA-IL-1 is expected to express in Tg mice and play an important role in the development of joint destruction. The present study investigated whether biologically active hIL-1α derived from the transgene appears on the surface of synoviocytes, and whether MA-IL-1 contributes to synovial proliferation and cartilage destruction in the development of arthritis in hIL-1α Tg mice. MA-IL-1 was found to be expressed on the surface of synoviocytes from Tg mice and triggered synoviocyte self-proliferation and cartilage destruction in vitro. Interestingly, the activity of MA-IL-1, but not soluble IL-1, in synoviocytes displayed correlations with both macroscopic and histological severity of arthritis in Tg mice. These results suggest that blocking the activities of both membrane-associated and soluble IL-1 may be required to effectively neutralize the pathogenic potential of this cytokine in inflammatory arthropathy such as RA.

The generation of hIL-1α Tg mice has been described previously (1). A 660-bp HindIII/HincII restriction fragment of hIL-1α cDNA (Immunex, Seattle, WA) coding Lys64 to Ala271 of the hIL-1α amino acid sequence was inserted into the EcoRI site of the third exon of the rabbit β-globin gene in the expression plasmid, pBsCAG-2. pBsCAG-2 possesses CAG containing the first intron of the chicken β-actin gene and a portion of the rabbit β-globin gene. The resulting construct was excised and microinjected into pronuclei of fertilized one-cell eggs from B6×B6C3F1 mice. The established Tg mouse line (designated Tg1706) was backcrossed with C3H/HeJ mice for six to eight generations and used in all experiments. The Tg101 line, which was designed to integrate 420 bp of mature hIL-1α cDNA coding Ser113 to Ala271, was used in a histological examination, and the macroscopic and histological scores were compared with those of Tg1706 (Fig. 1).

FIGURE 1.

Schematic representation of human IL-1α polypeptide. The transgenes of Tg1706 and Tg101 coded Lys64 to Ala271 and Ser113 to Ala271 of human IL-1α amino acid sequence, respectively: ▦, N-terminal conserved region (Met1-Gly78); □ (+), multiple basic region known as nuclear localization sequence (Lys79-Arg86); ▨ (−), negatively charged region (Leu86-Arg112); ▪, mature 17-kDa IL-1α (Ser113-Ala271).

FIGURE 1.

Schematic representation of human IL-1α polypeptide. The transgenes of Tg1706 and Tg101 coded Lys64 to Ala271 and Ser113 to Ala271 of human IL-1α amino acid sequence, respectively: ▦, N-terminal conserved region (Met1-Gly78); □ (+), multiple basic region known as nuclear localization sequence (Lys79-Arg86); ▨ (−), negatively charged region (Leu86-Arg112); ▪, mature 17-kDa IL-1α (Ser113-Ala271).

Close modal

Synovial specimens obtained from knee joints of 6- to 8-wk-old Tg mice were treated using 120 U/ml Streptomyces sp. C-51 collagenase (Sanko Junyaku, Tokyo, Japan) at 37°C for 30 min. Dispersed synovial cells were allowed to adhere to dishes in DMEM (Life Technologies, Gaithersburg, MD) containing 10% FBS (Life Technologies), 100 U/ml penicillin, and 100 μg/ml streptomycin (Life Technologies). Fifth-passage cells were used in all experiments.

Clinical symptoms of arthritis in all four limbs were macroscopically evaluated according to a visual scoring system. Arthritic joints were graded on a scale of 0–4: 0 = no change, 0.5 = swelling and erythema of 1 digit, 1 = swelling and erythema of ≥2 digits, 2 = mild swelling and erythema of the limb, 3 = gross swelling and erythema of the limb, and 4 = gross deformity and inability to use the limb. Scoring was performed in a blinded fashion by two observers, and the macroscopic score for each mouse comprised the sum of scores for all four limbs, for a maximum score of 16. In histological evaluations, ankle and knee joints were dissected and fixed in formalin. Sagittal sections (6 μm) were prepared and stained using H&E. Using the method described by van den Berg et al. (25), synovial infiltration and cartilage destruction were scored on four semiserial sections of each specimen spaced 10 sections apart. Neutrophil infiltration was graded on a scale of 0–3, according to the number of neutrophils in synovial tissue. Cartilage destruction was also graded on a scale of 0–3: 0 = no change, 1 = dead chondrocytes (empty lacunae) or focal loss of cartilage, 2 = loss of 25–50% of cartilage, and 3 = complete loss of cartilage. Scoring was performed again in a blinded fashion by two observers, and histological scores for each mouse comprised the sum of scores for two hind limbs, for a maximum score of 24.

MA-IL-1 synthesis by synoviocytes was analyzed using flow cytometry. Briefly, adherent synoviocytes (1–5 × 105 cells/test) were harvested and placed in ice-cold 5 mM EDTA and 1% BSA in Ca2+/Mg2+-free PBS at 37°C for 15 min. In accordance with the method described by Bailly et al. (26), either with or without 144 h of fixation in 1% (v/v) PFA at room temperature, synoviocytes were incubated for 15 min with unlabeled anti-CD16/32 (2.4G2; BD PharMingen, San Diego, CA) to block nonspecific binding to FcRII/III. Cells were then stained using PE-labeled anti-hIL-1α mAb (BD Immunocytometry Systems, San Jose, CA). In two-color analysis of freshly isolated synoviocytes, cells were further stained with biotinylated anti-F4/80 Ab (Cedarlane Laboratories, Hornby, Ontario, Canada), then incubated with cytochrome-conjugated streptavidin (BD Immunocytometry Systems). PE-conjugated mouse IgG (BD PharMingen) was used as an isotype-matched control to exclude the possibility of nonspecific binding. Stained cells were then analyzed using FACScan (BD Biosciences, Mountain View, CA). In some experiments cells were treated with 0.01 μg/ml trypsin before PFA fixation, then subjected to flow cytometry.

Cultured synoviocytes were maintained in methionine/cysteine-free medium (Life Technologies) for 2 h, then medium was replaced with freshly prepared appropriate deficient medium containing 40 μCi/ml [35S]methionine/cysteine (Amersham Pharmacia Biotech, Little Chalfont, U.K.) for 6 h, and washed three times using ice-cold PBS. The synoviocyte membrane fraction was prepared as previously described (27). Briefly, cultured synoviocytes harvested with ice-cold 5 mM EDTA in PBS were suspended at a concentration of 5 × 106 cells/ml in ice-cold homogenization buffer (20 mM Tris-HCl (pH 7.4), 10 mM NaCl, 0.1 mM MgCl2, 0.1 mM PMSF, and 0.5 mg/ml DNase I), followed by sonication three times for 15 s each time. Homogenate was centrifuged at 95,000 × g for 1 h over 41% (w/v) sucrose solution. The [35S]methionine/cysteine-labeled membrane fraction was recovered from the interface and treated with lysis buffer (150 mM NaCl, 10 mM Tris-HCl (pH 7.5), 1% deoxycholate, 1% Triton X-100, 0.1% SDS, 10 mM EDTA, and 2 mM PMSF). This isolated membrane fraction was concentrated 5- to 10-fold in a Centricon Centrifugal Concentrator (Millipore, Bedford, MA), then subjected to immunoprecipitation with anti-hIL-1α polyclonal Ab (Endogen, Woburn, MA) using an ImmunoPure Protein A IgG Orientation Kit (Pierce, Rockford, IL). In some experiments, 20 μg of unlabeled recombinant hIL-1α (Genzyme, Cambridge, MA) was added during immunoprecipitation. Labeled proteins in immunoprecipitates and 14C-methylated protein Mr marker (Amersham Pharmacia Biotech) were prepared for electrophoresis on 12.5% SDS-polyacrylamide gels, fixed, and treated with ENLIGHTNING (PerkinElmer, Boston, MA). Gels were dried and exposed to film at −80°C for autoradiography.

MA-IL-1 bioactivity in synoviocytes was quantitated by PFA fixation of cells, as described by Bailly et al. (26). Briefly, synoviocytes were inoculated at 5 × 104 cells/well on 96-well, flat-bottom tissue culture plates (BD Biosciences, Franklin Park, NJ). After culturing for 24 h, cells were fixed with 1% PFA in PBS (pH 7.4) at room temperature for 144 h, washed three times, and incubated in 100 μl of medium for 24 h. IL-1-sensitive mouse T cell clone D10.G4.1 (D10) cells (provided by Dr. Tadakuma, National Defense Medical College) were propagated as described previously (28), then used as an indicator for the presence of IL-1. In the synoviocyte proliferation assay, Tg mouse-derived synoviocytes were used as indicators for IL-1. Indicator cells were distributed to wells at a concentration of 4 × 104 cells/well containing fixed synoviocytes in a total volume of 200 μl of medium supplemented with 1 μg/ml Con A (Sigma-Aldrich, St. Louis, MO). In assays for soluble IL-1, indicator cells were similarly distributed to wells in medium containing 25% (v/v) final concentration of samples, instead of fixed cells. The incorporation of [3H]thymidine into indicator cells was measured during the final 4 h of the 48-h culture. In some experiments neutralizing Abs against human IL-1α (20 μg/ml; Endogen) and/or mouse IL-1α (20 μg/ml; R&D Systems, Minneapolis, MN) were added to cultures during assays. Normal rabbit or goat IgGs (R&D Systems) were used as isotype-matched controls for anti-human or anti-mouse IL-1α neutralizing Ab, respectively. The mitogenic activity of 100 pg/ml recombinant human IL-1α (Endogen) was determined to provide a reference for the magnitude of the effects of MA-IL-1 expressed on fixed synoviocytes.

Synoviocytes were inoculated at 1.5 × 105 cells/well on 24-well, flat-bottom tissue culture plates (BD Biosciences). After 24 h of culture, cells were fixed with 1% PFA in PBS (pH 7.4) at room temperature for 144 h. Live synoviocytes were added to wells as indicator cells at 1.5 × 105 cells/well in a total volume of 500 μl, either directly or into the top compartment of the Cell Culture Insert (BD Biosciences). Incorporation of [3H]thymidine into live synoviocytes was measured during the final 24 h of the 48-h culture. For blockade of IL-1, neutralizing Abs against hIL-1α (20 μg/ml; Endogen) and/or mouse IL-1α (20 μg/ml; R&D Systems) were added to cultures during assays.

Synoviocytes were inoculated at 1.5 × 105 cells/well on 24-well, flat-bottom plates (BD Biosciences) in a total volume of 500 μl and incubated for 24, 48, 72, or 96 h, and culture supernatants were collected before fixation in 1% PFA for 144 h. In MA-IL-1 assays, 1.5 × 105 D10 cells were added to PFA-fixed synoviocytes. In soluble IL-1 assays, 1.5 × 105 D10 cells were incubated with a 25% (v/v) final concentration of culture supernatants from the corresponding time points. Incorporation of [3H]thymidine into D10 cells was measured during the final 4 h of the 48-h culture.

Articular chondrocytes were obtained from glenohumeral joints of young Japanese White rabbits. Freshly isolated chondrocytes were seeded at 1 × 105 cells/ml in a 24-well, flat-bottom plate (BD Biosciences). After 1 wk of culture, confluent cells were incubated for 24 h in 500 μl of fresh medium containing [35S]sulfate (Amersham Pharmacia Biotech) at 5 μCi/ml and washed four times with cold fresh medium. Radiolabeled cells were further incubated for 48 h in the presence or the absence of detergent-insoluble membrane fraction isolated from synoviocytes. In some wells, labeled cells were incubated with membrane fraction isolated from trypsin-treated synoviocytes or with 100 μM l-NG-monomethyl arginine (LMMA; Wako Pure Chemical Industries, Osaka, Japan), an NO synthase inhibitor. The amount of 35S-labeled proteoglycan (PG) in cell and matrix layer and in supernatant was determined as previously described (29). Briefly, 35S-labeled cells and supernatants were separated. A total of 25 μl of supernatant was solubilized using 75 μl of 1.33 M guanidine HCl with 0.5% Triton X-100. Twenty-five microliters of 35S-labeled cell and matrix layer was solubilized for 4 h at 4°C with 4 M guanidine HCl and 0.05 M sodium acetate, pH 6.0, containing protease inhibitors, followed by dilution with 75 μl of dilution buffer containing 0.5% Triton X-100. Next, 100 μl of each sample was prepared in a 96-well MultiScreen filtration plate assembly (Millipore), and 150 μl of 0.2% Alcian Blue was added to the well. Well contents were then filtered through the Millipore Durapore membrane (0.45-μm pore size). Unincorporated [35S]sulfate was removed by three passages of vacuum filtration with wash buffer through the membrane. The membrane disc in each well was punched out and applied to the scintillation counter. All samples were analyzed in triplicate. PG release into supernatant was calculated according to the following equation: % PG release = [([35S]PG in supernatant)/([35S]PG in cell and matrix + [35S]PG in supernatant)] × 100%.

Results were expressed as the mean ± SEM. Statistical comparisons were performed using nonparametric Mann-Whitney U tests. Correlation analysis was performed using StatView-J 5.0 statistical software (SAS Institute, Cary, NC). A value of p < 0.05 was considered statistically significant.

Two-color flow cytometric analysis of transgene-derived MA-IL-1 revealed that freshly isolated synoviocytes consisted of ∼80% F4/80+ synovial macrophages and 20% F4/80 synovial fibroblasts (Fig. 2,A, left panel). In histogram analysis, ∼78% of F4/80+ cells and 70% of F4/80 cells expressed MA-IL-1 on their cell surface (Fig. 2,A, right panel). As hIL-1α Tg mice constitutively express transgene under the control of CAG promoter, both types of synoviocytes constitutively produced hIL-1α. The fact that membrane permeabilization was not required for staining synoviocytes with PE-labeled hIL-1α Ab ensured cell surface distribution of hIL-1α (Fig. 2,B, left panel). Identical staining patterns were observed in PFA-fixed synoviocytes (Fig. 2,B, middle panel). Furthermore, this membrane-localized IL-1 in synoviocytes was removed with mild trypsin treatment (Fig. 2 B, right panel), as reported by others (9, 30). This indicates that MA-IL-1 was substantially anchored in the membrane, with tryptic cleavage sites exposed on the cellular surface.

FIGURE 2.

Flow cytometric analysis of transgene-derived MA-IL-1 on the surface of synoviocytes. A, Freshly isolated synoviocytes from 8-wk-old Tg mice were stained with PE-labeled anti-hIL-1α Ab in conjunction with biotinylated anti-F4/80 Ab to identify synovial macrophages (right panel). The levels of MA-IL-1 after gating on F4/80+ cells or F4/80 cells are shown in histograms. The percentage of cells expressing MA-IL-1 is indicated (right panel). The results are representative of three different experiments. B, MA-IL-1 expression in fifth-passage synoviocytes (purity of synovial fibroblasts, >95%). Histograms show effects of PFA-fixation and mild trypsin treatment on the staining pattern of cell surface hIL-1α. MA-IL-1 can be detected on the surface of synoviocytes in both the presence (middle panel) and absence (left panel) of PFA fixation. MA-IL-1 exposed on the surface of synoviocytes was removed by mild trypsin treatment (right panel). Gray lines show the background with isotype-matched control Abs. Results are representative of three different experiments.

FIGURE 2.

Flow cytometric analysis of transgene-derived MA-IL-1 on the surface of synoviocytes. A, Freshly isolated synoviocytes from 8-wk-old Tg mice were stained with PE-labeled anti-hIL-1α Ab in conjunction with biotinylated anti-F4/80 Ab to identify synovial macrophages (right panel). The levels of MA-IL-1 after gating on F4/80+ cells or F4/80 cells are shown in histograms. The percentage of cells expressing MA-IL-1 is indicated (right panel). The results are representative of three different experiments. B, MA-IL-1 expression in fifth-passage synoviocytes (purity of synovial fibroblasts, >95%). Histograms show effects of PFA-fixation and mild trypsin treatment on the staining pattern of cell surface hIL-1α. MA-IL-1 can be detected on the surface of synoviocytes in both the presence (middle panel) and absence (left panel) of PFA fixation. MA-IL-1 exposed on the surface of synoviocytes was removed by mild trypsin treatment (right panel). Gray lines show the background with isotype-matched control Abs. Results are representative of three different experiments.

Close modal

To further confirm membrane localization of transgene-derived hIL-1α, a membrane fraction was isolated from synoviocytes, and immunoprecipitation was performed using specific Abs. The results clearly indicated that transgene-derived hIL-1α within the membrane fraction included a 25-kDa protein, slightly heavier than the 23-kDa primary translation product of the transgene (Fig. 3). In fact, culture supernatants and cell lysates of synoviocytes displayed both 23- and 25-kDa hIL-1α proteins (1). However, only the 25-kDa protein was detected in the membrane fraction. This preferential distribution of 25-kDa hIL-1α implies the promotion of post-translational modifications probably related to membrane localization of hIL-1α, such as phosphorylation (5), mannosylation (6), and myristolation (7). To examine whether this band was the truth, competition analysis was performed by adding excess unlabeled recombinant hIL-1α (∼2.0 μg) during immunoprecipitation. As expected, recombinant hIL-1α completely prevented the immunoprecipitation of labeled hIL-1α with specific Ab, whereas neither recombinant hIL-1β nor mouse IL-1α (mIL-1α) demonstrated any effect (data not shown). This indicates that transgene-derived hIL-1α actually localizes within the membrane of synoviocytes.

FIGURE 3.

Immunoprecipitation of hIL-1α isolated from synoviocyte membrane fraction. Cultured synoviocytes were labeled with [35S]methionine/cysteine (40 μCi/ml) for 6 h. Membrane fraction was obtained as indicated in Materials and Methods, followed by immunoprecipitation using anti-hIL-1α polyclonal Ab. Lane 1, Immunoprecipitation in the presence of excess unlabeled recombinant hIL-1α. Lane 2, immunoprecipitation of synoviocyte membrane fraction, showing the 25-kDa precursor form of hIL-1α (arrow).

FIGURE 3.

Immunoprecipitation of hIL-1α isolated from synoviocyte membrane fraction. Cultured synoviocytes were labeled with [35S]methionine/cysteine (40 μCi/ml) for 6 h. Membrane fraction was obtained as indicated in Materials and Methods, followed by immunoprecipitation using anti-hIL-1α polyclonal Ab. Lane 1, Immunoprecipitation in the presence of excess unlabeled recombinant hIL-1α. Lane 2, immunoprecipitation of synoviocyte membrane fraction, showing the 25-kDa precursor form of hIL-1α (arrow).

Close modal

Kaye and co-workers (31, 32) have reported a IL-1-sensitive T cell clone, D10.G4.1, that can be used to detect and titrate IL-1 by adding test molecules together with Con A. Using these characteristics of D10 cells, MA-IL-1 expression on LPS-stimulated macrophages has been elucidated by [3H]thymidine incorporation into D10 cells cultured on PFA-fixed macrophages (14). This procedure was used to determine the MA-IL-1 activity of synoviocytes from Tg mice. Tg mouse-derived synoviocytes significantly stimulated D10 cell proliferation compared with littermate-derived synoviocytes (Fig. 4,A). To exclude the possibility of the mitogenic activity of MA-IL-1 actually being attributable to minor contaminants in preparations, neutralizing Ab against hIL-1α/mIL-1α was added to cultures during the assay. Addition of anti-hIL-1α Ab resulted in significant inhibition of D10 cell proliferation, suggesting that bioactivity of synoviocytes is due to transgene-derived MA-IL-1. Furthermore, anti-mIL-1α Ab inhibited D10 cell proliferation to a similar degree as anti-hIL-1α Ab, with inhibition reaching a maximum with the combination of both Abs. Transgene-derived MA-IL-1 thus induces the production of endogenous mouse MA-IL-1, and both forms of MA-IL-1 may play a role in the development of proliferative synovitis in Tg mice. As IL-1 has been shown to act as a mitogen for synoviocytes (33, 34, 35), the effects of MA-IL-1 on synoviocyte proliferation were examined. In this experiment, live synoviocytes isolated from Tg mice and RA patients were used as indicator cells for IL-1 activity, instead of D10 cells. Notably, putative MA-IL-1 in Tg mouse-derived synoviocytes led to significant stimulation of [3H]thymidine incorporation into indicator cells compared with that in littermate-derived synoviocytes (Fig. 4 B), indicating that MA-IL-1 on Tg mouse-derived synoviocytes stimulates synoviocyte self-proliferation via juxtacrine mechanisms.

FIGURE 4.

A, Tg mouse-derived synoviocytes express biologically active MA-IL-1. Fifth-passage synoviocytes isolated from Tg mouse (▪) or littermates (▨) were inoculated on 96-well plates and fixed with 1% PFA in PBS for 144 h. D10 cells were distributed to wells and incubated with or without neutralizing Abs against hIL-1α/mIL-1α. Isotype-matched control IgGs were used to exclude a possibility that these neutralizations include nonspecific reactions. [3H]thymidine incorporation into D10 cells was measured during the final 4 h of a 48-h incubation. Data represent the mean counts per minute ± SEM of four separate experiments. B, MA-IL-1 contributes to self-proliferation of synoviocytes. Tg mouse- or littermate-derived synoviocytes cultured on 96-well plates were fixed with 1% PFA for 144 h. Live synoviocytes from Tg mouse were overlaid on the fixed cells with or without neutralizing Abs against hIL-1α/mIL-1α. [3H]thymidine incorporation into overlaid synoviocytes was measured during the final 4 h of a 48-h incubation. Control comprised [3H]thymidine incorporation into synoviocytes incubated with medium alone. Data represent the mean counts per minute ± SEM of four separate experiments. ∗, p < 0.05; ∗∗, p < 0.01

FIGURE 4.

A, Tg mouse-derived synoviocytes express biologically active MA-IL-1. Fifth-passage synoviocytes isolated from Tg mouse (▪) or littermates (▨) were inoculated on 96-well plates and fixed with 1% PFA in PBS for 144 h. D10 cells were distributed to wells and incubated with or without neutralizing Abs against hIL-1α/mIL-1α. Isotype-matched control IgGs were used to exclude a possibility that these neutralizations include nonspecific reactions. [3H]thymidine incorporation into D10 cells was measured during the final 4 h of a 48-h incubation. Data represent the mean counts per minute ± SEM of four separate experiments. B, MA-IL-1 contributes to self-proliferation of synoviocytes. Tg mouse- or littermate-derived synoviocytes cultured on 96-well plates were fixed with 1% PFA for 144 h. Live synoviocytes from Tg mouse were overlaid on the fixed cells with or without neutralizing Abs against hIL-1α/mIL-1α. [3H]thymidine incorporation into overlaid synoviocytes was measured during the final 4 h of a 48-h incubation. Control comprised [3H]thymidine incorporation into synoviocytes incubated with medium alone. Data represent the mean counts per minute ± SEM of four separate experiments. ∗, p < 0.05; ∗∗, p < 0.01

Close modal

To elucidate the contribution of MA-IL-1 to the development of arthritis, the severity of arthritis was evaluated according to a scoring system. Clinical symptoms of arthritis in all four paws and histology of bilateral knee joints were scored, and these macroscopic and histological scores were compared between the two Tg mouse lines, Tg1706 and Tg101, which overexpress pro-IL-1α and mature IL-1α, respectively. Interestingly, these scores of Tg1706 were significantly higher than those of Tg101, indicating relatively severe arthritic phenotype in Tg1706 compared with Tg101 (Fig. 5,A). In the next experiment, the relationship between MA-IL-1 activity of synoviocytes and severity of arthritis was examined in 10 6-wk-old Tg mice. Correlations between these scores and levels of MA-IL-1, soluble IL-1, and serum hIL-1α were determined. Linear analyses revealed that MA-IL-1 activity displayed significant correlations with both macroscopic and histological scores (Fig. 5 B). However, soluble IL-1 activity and serum concentrations of hIL-1α displayed no correlation with either score. MA-IL-1 expression in synovial tissue may therefore represent a key element in the development of synovitis and subsequent joint destruction in Tg mice.

FIGURE 5.

A, Comparison of severity of arthritis between Tg1706 and Tg101. Macroscopic and histological findings were scored at 6-wk-old Tg mice. Data are presented as the mean ± SEM of four mice. B, Correlation between macroscopic score, histological score, relative MA-IL-1 activity, relative soluble IL-1 activity, and serum hIL-1α level in 10 6-wk-old Tg mice. Data are presented as the coefficient (R) and p value derived from linear regression analysis. ∗, p < 0.05.

FIGURE 5.

A, Comparison of severity of arthritis between Tg1706 and Tg101. Macroscopic and histological findings were scored at 6-wk-old Tg mice. Data are presented as the mean ± SEM of four mice. B, Correlation between macroscopic score, histological score, relative MA-IL-1 activity, relative soluble IL-1 activity, and serum hIL-1α level in 10 6-wk-old Tg mice. Data are presented as the coefficient (R) and p value derived from linear regression analysis. ∗, p < 0.05.

Close modal

To investigate whether direct cell-to-cell interactions are required for MA-IL-1 activity, a coculture system using the cell culture insert with 1-μm pores was employed, allowing the infiltration of macromolecules, but not direct cell-to-cell contact. Similar to the experiment in Fig. 5,B, live synoviocytes were used as indicator cells for IL-1 activity and cocultured with PFA-fixed synoviocytes with or without cell culture inserts. Significant differences in live synoviocyte proliferation were observed between the two different cultures. Live synoviocytes displayed obvious proliferation when directly cultured with PFA-fixed synoviocytes without separation (Fig. 6). However, once cells were separated from each other using a cell culture insert, the proliferative activity of PFA-fixed synoviocytes was abrogated. When neutralizing Abs against hIL-1α/mIL-1α were added to cultures during the assay, synoviocyte proliferation was significantly diminished in culture without cell culture insert, indicating that this proliferative activity was attributable to MA-IL-1 in PFA-fixed synoviocytes. Weak, but nonsignificant, neutralization was observed in culture with the cell culture insert; in contrast to D10 cells, Tg mouse-derived synoviocytes spontaneously produce soluble IL-1 and MA-IL-1, and endogenous IL-1-dependent proliferation of these cells was blocked by the specific Abs. These results indicate that direct cell-to-cell contact is indispensable in the promotion of proliferative activity by MA-IL-1.

FIGURE 6.

MA-IL-1 promotes synoviocyte proliferation in a cell-to-cell contact-dependent manner. Tg mouse-derived synoviocytes cultured on 24-well plates were fixed with 1% PFA for 144 h. Live synoviocytes were added directly to the fixed synoviocytes (▪) or the top compartment of the cell culture insert (▨) and incubated for 48 h. For blockade of IL-1 activity, neutralizing Abs against hIL-1α/mIL-1α were added during incubation. [3H]Thymidine incorporation into live synoviocytes was determined during the final 4 h of a 48-h incubation. Data are presented as the mean counts per minute ± SEM for four separate experiments. ∗, p < 0.05; ∗∗, p < 0.01

FIGURE 6.

MA-IL-1 promotes synoviocyte proliferation in a cell-to-cell contact-dependent manner. Tg mouse-derived synoviocytes cultured on 24-well plates were fixed with 1% PFA for 144 h. Live synoviocytes were added directly to the fixed synoviocytes (▪) or the top compartment of the cell culture insert (▨) and incubated for 48 h. For blockade of IL-1 activity, neutralizing Abs against hIL-1α/mIL-1α were added during incubation. [3H]Thymidine incorporation into live synoviocytes was determined during the final 4 h of a 48-h incubation. Data are presented as the mean counts per minute ± SEM for four separate experiments. ∗, p < 0.05; ∗∗, p < 0.01

Close modal

To investigate the kinetics of synthesis for MA-IL-1 and soluble IL-1, incorporation of [3H]thymidine into synoviocytes was determined when cells were overlaid on PFA-fixed synoviocytes as a feeder layer of MA-IL-1. Soluble IL-1 secreted into culture supernatant by overlaid synoviocytes was demonstrable from 24 h after inoculation and plateaued between 72 and 96 h (Fig. 7,A), whereas the corresponding MA-IL-1 activity reached a plateau by 24 h after inoculation, remaining stable until at least 96 h (Fig. 7 B). In addition, the proliferative activity of soluble IL-1 was ∼5-fold higher than that of MA-IL-1.

FIGURE 7.

Synthetic kinetics of MA-IL-1 and soluble IL-1α. Fifth-passage synoviocytes derived from Tg mice were fixed on the indicated day of culture after inoculation. Corresponding culture supernatants were collected immediately before fixation of synoviocytes. D10 cells were incubated on fixed cells or with a 25% (v/v) final concentration of culture supernatants. The IL-1 activity of supernatants (A) and fixed cells (B) was determined by measuring [3H]thymidine incorporation into D10 cells during the last 4 h of a 48-h incubation (▪). Control data for synoviocytes were derived from littermates (□). Data are presented as the mean counts per minute ± SEM for four separate experiments.

FIGURE 7.

Synthetic kinetics of MA-IL-1 and soluble IL-1α. Fifth-passage synoviocytes derived from Tg mice were fixed on the indicated day of culture after inoculation. Corresponding culture supernatants were collected immediately before fixation of synoviocytes. D10 cells were incubated on fixed cells or with a 25% (v/v) final concentration of culture supernatants. The IL-1 activity of supernatants (A) and fixed cells (B) was determined by measuring [3H]thymidine incorporation into D10 cells during the last 4 h of a 48-h incubation (▪). Control data for synoviocytes were derived from littermates (□). Data are presented as the mean counts per minute ± SEM for four separate experiments.

Close modal

Monolayer-cultured articular chondrocytes derived from Japanese White rabbits were labeled with [35S]sulfate for 24 h, then incubated with synoviocyte membrane fraction in the presence or the absence of anti-hIL-1α Ab for 48 h. Release of 35S-labeled PG from cell and matrix layer was examined. The synoviocyte membrane fraction significantly stimulated PG release into culture supernatant compared with control (Fig. 8), and stimulation was decreased almost to control levels by the addition of anti-IL-1α Ab. In contrast, the membrane fraction isolated from synoviocytes treated with mild trypsin did not affect PG release, compatible with the flow cytometric data in Fig. 3 C showing that MA-IL-1 has a tryptic cleavage site and can be removed by mild trypsin treatment. Furthermore, the NO synthase inhibitor, LMMA, for the most part inhibited membrane fraction-stimulated PG release, indicating the involvement of NO in this process. These data suggest that MA-IL-1 induces PG release from the cell and matrix through generating NO in chondrocyte monolayer culture, further indicating that MA-IL-1 may play a role in cartilage destruction in vivo.

FIGURE 8.

MA-IL-1 stimulates PG release from cartilage matrix through generation of NO. Cultured articular chondrocytes were radiolabeled with [35S]sulfate for 24 h, then incubated for 48 h with synoviocyte membrane fraction alone, membrane fraction plus anti hIL-1α Ab, membrane fraction plus 100 mM LMMA, or membrane fraction isolated from trypsin-treated synoviocytes. Control data were derived from chondrocytes without membrane fraction. The release of 35S-labeled PG from the cell and matrix layer to culture supernatant was examined. The percent PG release was calculated according to the following equation: % release = [35S]PG in supernatant/[35S]PG in cell and matrix + [35S]PG in supernatant. Data are presented as the mean counts per minute ± SEM for four separate experiments. ∗, p < 0.05 compared with the control.

FIGURE 8.

MA-IL-1 stimulates PG release from cartilage matrix through generation of NO. Cultured articular chondrocytes were radiolabeled with [35S]sulfate for 24 h, then incubated for 48 h with synoviocyte membrane fraction alone, membrane fraction plus anti hIL-1α Ab, membrane fraction plus 100 mM LMMA, or membrane fraction isolated from trypsin-treated synoviocytes. Control data were derived from chondrocytes without membrane fraction. The release of 35S-labeled PG from the cell and matrix layer to culture supernatant was examined. The percent PG release was calculated according to the following equation: % release = [35S]PG in supernatant/[35S]PG in cell and matrix + [35S]PG in supernatant. Data are presented as the mean counts per minute ± SEM for four separate experiments. ∗, p < 0.05 compared with the control.

Close modal

MA-IL-1 was found to play key roles in the development of arthritic phenotypes in Tg mice. Of interest is the fact that both macroscopic and histological scores were correlated with activity of MA-IL-1, but not with activity of soluble IL-1 produced by synoviocytes. Moreover, cartilage destruction of Tg101 line overexpressing 17-kDa mature IL-1α was relatively mild even at 12 wk after birth, although the Tg1706 line overexpressing pro-IL-1α demonstrated complete loss of cartilage at 8 wk after birth, which reflected low macroscopic and histological scores in the Tg101 line. This observation was not attributable to the difference in levels of transgene expression between the two lines, because the levels of serum IL-1α were almost similar (∼100 pg/ml). Thus, as in Tg mouse studies on membrane-associated TNF (36, 37), the arthritogenic properties of MA-IL-1 may be sufficient to cause severe arthritis even in conditions without processing of proteins to mature form.

However, we cannot neglect the fact that, in general, transgene expression can be affected by copy number and integration site of the transgene, and a simple comparative study of phenotypic characteristics among Tg mouse lines is unlikely to provide informative data. In actual fact, we established two transgenic founders for each Tg mouse for pro- and active IL-1α. As assessed by tail Southern blot analysis, copy number of transgenes was similar among the four transgenic founders (three or four copies), and differences in integration site were confirmed by fluorescence in situ hybridization analysis. Northern blot analysis revealed quite similar levels and patterns of mRNA expression in all four transgenic founders, and of course in offspring of Tg101 and Tg1706. All four founders exhibited arthritic phenotypes, and a more severe arthritic phenotype in pro-IL-1α Tg mice than in active IL-1α Tg mice was noted as a universal trend, even in offspring. This indicates that in our study the effects of copy number and integration site of transgenes can be neglected, allowing direct comparison of the two Tg mouse lines. We therefore believe that Tg mice for pro-IL-1α exhibited a more progressive arthritic phenotype than mice for active IL-1α, and membrane IL-1 plays an important role in the evolution of inflammatory arthritis.

To date, MA-IL-1 has been shown to be more potent than soluble IL-1 in a variety of situations, such as neutrophil extravasation via endothelial cells, T cell activation during Ag presentation, and osteoclast formation through up-regulation of receptor activator NF-κB ligand expression on osteoblasts, all of which play crucial roles in the development of inflammatory joint diseases. Of the pleiotropic activities of MA-IL-1, the present study focused on the effects on synoviocytes and chondrocytes, as IL-1 has been shown to act as a mitogen for rheumatoid synovial fibroblasts, and abnormal IL-1 production contributes to synovial proliferation and degradation of the cartilage matrix in RA and collagen-induced arthritis in mice. As MA-IL-1 synthesis is spontaneously promoted in hIL-1α Tg mice and persists due to the characteristics of the promoter, synoviocytes cultured on PFA-fixed synoviocytes displayed marked proliferation in the absence of stimuli. Moreover, transgene-derived hIL-1α further up-regulated endogenous mouse MA-IL-1 synthesis via autocrine mechanisms, and this may also be involved in the joint pathology of hIL-1α Tg mice.

Kurt-Jones et al. (12) provided the first evidence that PFA-fixed macrophages stimulate IL-1-sensitive T cell clone, D10 G4.1 proliferation due to IL-1 activity on the external plasma membrane of macrophage. In the present study MA-IL-1 expression on the surface of synoviocytes isolated from arthritic joints was directly identified using flow cytometry. Cellular staining regardless of membrane permeabilization and dissociation of hIL-1α from the cell surface by mild trypsin treatment indicated that IL-1α is undoubtedly associated with the exterior plasma membrane surface of synoviocytes. Matsushima et al. (30) also documented the release of biologically active IL-1 from plasma membrane, when LPS-stimulated macrophages are treated with mild trypsin or plasmin-like proteases.

IL-1α precursor propeptide lacks a classical signal sequence (4), which is known to regulate the processing of secreted and integral plasma membrane-associated proteins. To date, a number of post-translational modifications within the NH2-terminal domain have been proposed to affect the intracellular distribution of IL-1α, including phosphorylation (5), mannosylation (6), and myristolation (7). However, the details of these processes have remained unknown. Several speculations have been proposed regarding such post-translational modifications and their impact on intracellular distribution of IL-1α. One investigator has demonstrated that phosphorylation of newly synthesized IL-1α signifies intracellular routing of IL-1α precursor, and ∼10% of phosphorylated IL-1α precursor is committed to the membrane-associated form. Another study revealed that glycosylation of IL-1α precursor allows association with membrane-bound lectins and membrane-localization of IL-1α (6). Alternatively, striking evidence has been proposed that physical injury or programmed cell death (i.e., apoptosis) play a role in IL-1α secretion through membrane disruption (38). Although certain post-translational modifications are likely to reflect the difference between transgene-predicted (25 kDa) and observed (23 kDa) masses of IL-1α precursor in immunoprecipitation of the synoviocyte membrane fraction in IL-1α Tg mice, the mechanisms affecting membrane localization of IL-1α remain unknown.

MA-IL-1 expression on the surface of synoviocytes was clarified from another perspective. Synoviocytes were plated onto 24-well plates and fixed using 1% PFA. Live synoviocytes were directly added to fixed synoviocytes or the top compartment of the cell culture insert, allowing soluble IL-1α, but not MA-IL-1, to migrate between the top and bottom compartments. This experiment indicated that synoviocytes without separation engaged in direct cell-to-cell interactions, resulting in higher proliferation attributable to the activities of soluble IL-1 plus MA-IL-1. However, the true magnitude of [3H]thymidine incorporation into indicator synoviocytes cultured on the PFA-fixed synoviocytes actually appeared higher than that cultured on nonfixed live synoviocytes. This can be explained by our unpublished observations that live synoviocytes spontaneously produce IL-1 receptor antagonist in vitro, which may block IL-1 activity during the experiment.

As reported by van de Loo et al. (39, 40), IL-1 inhibits synthesis of PG by chondrocytes through generation of NO in zymosan-induced arthritis. The present study demonstrated that membrane fraction isolated from synoviocytes induces PG release from the cartilage matrix in chondrocyte monolayer culture, and that this phenomenon is mediated by NO synthesis. This indicates that MA-IL-1 within the membrane is essentially implicated in chondrocyte PG loss, suggesting the possibility that MA-IL-1 contributes to cartilage destruction during the course of arthritis in IL-1α Tg mice. However, PG loss was not detected when chondrocytes were cultured in agarose gels (data not shown). The absence of chondrocyte PG loss is probably attributable to the prevention of direct cell-to-cell contact by the surrounding agarose gel. Chondrocyte PG loss caused by the synoviocyte membrane fraction may thus, for the most part, be due to MA-IL-1 within the membrane.

Finally, the importance of membrane-associated molecules proposed in the current experimental study is that cell-cell interactions between macrophage-like synoviocytes and T lymphocytes activate the production of proinflammatory cytokines at the inflamed synovium (41, 42, 43). These include membrane-associated IL-1 and TNF, which induce fibroblast-like synoviocytes to produce large amounts of matrix metalloproteinases that degrade cartilage and bone. In the present study using IL-1α Tg mice, MA-IL-1 expressed on synoviocytes may trigger synoviocyte self-proliferation and induce cartilage degradation, mechanisms that may operate in the cartilage-pannus junction through cell-cell interactions in vivo. Moreover, a correlation between MA-IL-1 activity and severity of arthritis indicates that MA-IL-1 is a potent effector of joint inflammation. As the present results were obtained purely from animal studies, the importance and extent of MA-IL-1 contribution to the pathogenesis of human inflammatory joint diseases such as RA warrant investigation.

We are grateful to the late Prof. Masayuki Shinmei (Department of Orthopedic Surgery, National Defense Medical College) for the planning of this investigation. We also thank Prof. Takushi Tadakuma (Department of Parasitology, National Defense Medical College) for providing the D10.G4.1 cells used in this study.

3

Abbreviations used in this paper: RA, rheumatoid arthritis; hIL-1α, human IL-1α; LMMA, l-NG-monomethyl arginine; MA-IL-1, membrane-associated IL-1; mIL-1α, mouse IL-1α; PFA, paraformaldehyde; PG, proteoglycan; Tg, transgenic.

1
Niki, Y., H. Yamada, S. Seki, T. Kikuchi, H. Takaishi, Y. Toyama, K. Fujikawa, N. Tada.
2001
. Macrophage- and neutrophil-dominant arthritis in human IL-1α transgenic mice.
J. Clin. Invest.
107
:
1127
.
2
March, C. J., B. Mosley, A. Larsen, D. P. Cerretti, G. Braedt, V. Price, S. Gillis, C. S. Henney, S. R. Kronheim, K. Grabstein.
1985
. Cloning, sequence and expression of two distinct human interleukin-1.
Nature
315
:
641
.
3
Mosley, B., D. L. Urdal, K. S. Prickett, A. Larsen, D. Cosman, P. J. Conlon, S. Gillis, S. K. Dower.
1987
. The interleukin-1 receptor binds to the human interleukin-1α precursor but not the interleukin-1β precursor.
J. Biol. Chem.
262
:
2941
.
4
Dinarello, C. A..
1996
. Biologic basis for interleukin-1 in disease.
Blood
87
:
2095
.
5
Beuscher, H. U., M. W. Nickells, H. R. Colten.
1988
. The precursor of interleukin-1α is phosphorylated at residue serine 90.
J. Biol. Chem.
263
:
4023
.
6
Brody, D. T., S. K. Durum.
1989
. Membrane IL-1: IL-1α precursor binds to the plasma membrane via a lectin-like interaction.
J. Immunol.
143
:
1183
.
7
Stevenson, F. T., S. L. Bursten, C. Fanton, R. M. Locksley, D. H. Lovett.
1993
. The 31-kDa precursor of interleukin-1α is myristolated on specific lysines within the 16-kDa N-terminal propiece.
Proc. Natl. Acad. Sci. USA
90
:
7245
.
8
Carruth, L., S. Demczuk, S. Mizel.
1991
. Involvement of a calpain-like protease in the processing of the murine interleukin-1α precursor.
J. Biol. Chem.
266
:
12162
.
9
Kobayashi, Y., K. Yamamoto, T. Saido, H. Kawasaki, J. J. Oppenheim.
1990
. Identification of calcium-activated neutral protease as a processing enzyme of human interleukin 1α.
Proc. Natl. Acad. Sci. USA
87
:
5548
.
10
Lee, R. T., W. H. Briggs, G. C. Cheng, H. B. Rossiter, P. Libby, T. Kupper.
1997
. Mechanical deformation promotes secretion of IL-1α and IL-1 receptor antagonist.
J. Immunol.
159
:
5084
.
11
Conlon, P. J., K. H. Grabstein, A. Alpert, K. S. Prickett, T. P. Hopp, S. Gillis.
1987
. Localization of human mononuclear cell interleukin 1.
J. Immunol.
139
:
98
.
12
Kurt-Jones, E. A., D. I. Beller, S. B. Mizel, E. R. Unanue.
1985
. Identification of a membrane-associated interleukin-1 in macrophages.
Proc. Natl. Acad. Sci. USA
82
:
1204
.
13
Junming, L. E., D. Weinstein, U. Gubler, J. Vilcek.
1987
. Induction of membrane-associated interleukin 1 by tumor necrosis factor in human fibroblasts.
J. Immunol.
138
:
2137
.
14
Kurt-Jones, E. A., W. Fiers, J. S. Pober.
1987
. Membrane interleukin 1 induction on human endothelial cells and dermal fibroblasts.
J. Immunol.
139
:
2317
.
15
Zola, H., L. Flego, Y. T. Wong, P. J. Macardle, J. S. Kenney.
1993
. Direct demonstration membrane IL-1α on the surface on circulating B lymphocytes and monocytes.
J. Immunol.
150
:
1755
.
16
Yamashita, U., F. Shirakawa, H. Nakamura.
1987
. Production of interleukin 1 by adult T cell leukemia (ATL) cell lines.
J. Immunol.
138
:
3284
.
17
Acres, R. B., A. L. F. Larsen, P. J. Conlon.
1987
. IL 1 expression in a clone of human T cells.
J. Immunol.
138
:
2132
.
18
Nishimura, T., Y. Ishihara, T. Noguchi, T. Koga.
1989
. Membrane IL-1 induces bone resorption in organ culture.
J. Immunol.
143
:
1881
.
19
Kaplanski, G., R. Porat, K. Aiura, J. K. Erban, C. A. Dinarello.
1993
. Activated platelets induce endothelial secretion of interleukin-8 in vitro via an interleukin-1-mediated event.
Blood
81
:
2492
.
20
Beasley, D., A. L. Cooper.
1999
. Constitutive expression of interleukin-1α precursor promotes human vascular smooth muscle cell proliferation.
Am. J. Physiol.
276
:
H901
.
21
Weaver, C. T., E. R. Unanue.
1986
. T cell induction of membrane IL-1 on macrophages.
J. Immunol.
137
:
3868
.
22
Suresh, A., A. Sodhi.
1991
. Production of interleukin-1 and tumor necrosis factor by bone marrow-derived macrophages: effect of cisplatin and lipopolysaccharide.
Immunol. Lett.
30
:
93
.
23
Nishihara, T., T. Takahashi, Y. Ishihara, H. Senpuku, T. Koga.
1994
. Membrane-associated interleukin-1 promotes osteoclast-like cell formation in vitro.
Bone Miner.
25
:
15
.
24
McMahon, G. A., S. Garfinkel, I. Prudovsky, X. Hu, T. Maciag.
1997
. Intracellular precursor interleukin (IL)-1α, but not mature IL-1α, is able to regulate human endothelial cell migration in vitro.
J. Biol. Chem.
272
:
28202
.
25
van den Berg, W. B., L. A. B. Joosten, M. Helsen, F. A. J. van de Loo.
1994
. Amelioration of established murine collagen-induced arthritis with anti-IL-1 treatment.
Clin. Exp. Immunol.
95
:
237
.
26
Bailly, S., B. Ferrua, M. Fay, M. A. Gougerot-Pocidalo.
1990
. Paraformaldehyde fixation of LPS-stimulated human monocytes: technical parameters permitting the study of membrane IL-1 activity.
Eur. Cytokine Network
1
:
47
.
27
Maeda, T., K. Balakrishnan, Q. Mehdi.
1983
. A simple and rapid method for the preparation of plasma membranes.
Biochim. Biophys. Acta
731
:
115
.
28
Helle, M., L. Boeije, L. A. Aarden.
1988
. Functional discrimination between interleukin 6 and interleukin 1.
Eur. J. Immunol.
18
:
1535
.
29
Masuda, K., H. Shirota, E. J. M. A. Thonar.
1994
. Quantification of 35S-labeled proteoglycans complexed to Alcian Blue by rapid filtration in multiwell plates.
Anal. Biochem.
217
:
167
.
30
Matsushima, K., M. Taguchi, E. J. Kovacs, H. A. Young, J. J. Oppenheim.
1986
. Intracellular localization of human monocyte associated interleukin 1 (IL 1) activity and release of biologically active IL 1 from monocytes by trypsin and plasmin.
J. Immunol.
136
:
2883
.
31
Kaye, J., S. Porcelli, J. Tite, B. Jones, C. A. Janeway, Jr.
1983
. Both a monoclonal antibody and antisera specific for determinants unique to individual cloned helper T cell lines can substitute for antigen and antigen-presenting cells in the activation of T cells.
J. Exp. Med.
158
:
836
.
32
Kaye, J., S. Gillis, S. B. Mizel, E. M. Shevach, T. R. Malek, C. A. Dinarello, L. B. Lachman, C. A. Janeway, Jr.
1984
. Growth of a cloned helper T cell line induced by a monoclonal antibody specific for the antigen receptor: interleukin 1 is required for the expression of receptors for interleukin 2.
J. Immunol.
133
:
1339
.
33
Rupp, E. A., P. M. Cameron, C. S. Ranawat, J. A. Schmidt, B. K. Bayne.
1986
. Specific bioactivities of monocyte-derived interleukin-1α and interleukin-1β are similar to each other on cultured murine thymocytes and on cultured human connective tissue cells.
J. Clin. Invest.
78
:
836
.
34
Alvaro-Gracia, J. M., N. J. Zvaifler, G. S. Firestein.
1990
. Cytokines in chronic inflammatory arthritis. V. Mutual antagonism between interferon-gamma and tumor necrosis factor-α on HLA-DR expression, proliferation, collagenase production and granulocyte macrophage colony-stimulating factor production by rheumatoid arthritis synoviocytes.
J. Clin. Invest.
86
:
1790
.
35
Butler, D. M., D. S. Piccoli, P. H. Hart, J. A. Hamilton.
1988
. Stimulation of human synovial fibroblast DNA synthesis by recombinant human cytokines.
J. Rheumatol.
15
:
1463
.
36
Probert, L., K. Akassoglou, L. Alexopoulou, E. Douni, S. Haralambous, S. Hill, G. Kassiotis, D. Kontoyiannis, M. Pasparakis, D. Plows, et al
1996
. Dissection of the pathologies induced by transmembrane and wild-type tumor necrosis factor in transgenic mice.
J. Leukocyte Biol.
59
:
518
.
37
Georgopoulos, S., D. Plows, G. Kollias.
1996
. Transmembrane TNF is sufficient to induce localized tissue toxicity and chronic inflammatory arthritis in transgenic mice.
J. Inflamm.
46
:
86
.
38
Hogquist, K. A., M. A. Nett, E. R. Unanue, D. D. Chaplin.
1991
. Interleukin 1 is processed and released during apoptosis.
Proc. Natl. Acad. Sci. USA
88
:
8485
.
39
van de Loo, A. A. J., W. B. van den Berg.
1990
. Effects of murine recombinant IL-1 on synovial joints in mice: quantification of patellar cartilage metabolism and joint inflammation.
Ann. Rheum. Dis.
49
:
238
.
40
Van de Loo, F. A. J., O. J. Arntz, F. H. J. van Enckevort, P. L. E. M. van Lent, W. B. van den Berg.
1998
. Reduced cartilage proteoglycan loss during zymosan-induced gonarthritis in NOS2-defficient mice and in anti-interleukin-1-treated wild-type mice with unabated joint inflammation.
Arthritis Rheum.
41
:
634
.
41
Seckinger, P., M. T. Kaufmann, J. M. Dayer.
1990
. An Interleukin 1 inhibitor affects both cell-associated interleukin 1-induced T cell proliferation and PGE2/collagenase production by human dermal fibroblasts and synovial cells.
Immunobiology
180
:
316
.
42
Burger, D., R. Rezzonico, H. M. C. Modoux, R. A. Pierce, H. G. Welgus, J. M. Dayer.
1998
. Imbalance between interstitial collagenase and tissue inhibitor of metalloproteinases 1 in synoviocytes and fibroblasts upon direct contact with stimulated T lymphocytes: involvement of membrane-associated cytokines.
Arthritis Rheum.
41
:
1748
.
43
Burger, D., J. M. Dayer.
1999
. Cytokines and direct cell contact in synovitis: relevance to therapeutic intervention.
Arthritis Res.
1
:
17
.