Sterile synovial tissue was obtained during open joint surgery from patients with non-inflammatory joint diseases including osteoarthritis or osteonecrosis. No patient was receiving gold, D-penicillamine or corticosteroid at the time of surgery.

A portion of synovium was placed in OCT compound (Tissue Tek) and snap-frozen in liquid nitrogen for tissue sections. The fresh synovial tissues were repeatedly washed with sterile phosphate-buffered saline (PBS). Then the synovial membrane was dissected away and finely minced in Dulbecco’s modified Eagle’s (DME) media (Gibco). The tissue was mixed with 20 volumes of DMEM containing 1 mg/ml collagenase (Sigma) and 0.15 mg/ml DNase (Sigma) and stirred in a small spinner flask for 2–3 hours at 37°C until completely digested¹²⁾. An equal volume of 0.025% trypsin was added and incubated for a further 30 min under the same conditions. Subsequently, the cell suspension was filtered through a nylon sieve (Tetco) with a pore sized of 80 μm. These cells were then washed three times. The pellet was resuspended in DMEM supplemented with 10% fetal calf serum (Gibco), 100 u/ml penicillin and 100 μg/ml streptomycin (Flow lab). The suspension was transferred to 60 mm plastic Petri dishes (Falcon) and left undisturbed at 37°c in a humidified chamber of 5% CO₂. After overnight incubation the non-adherent floating cell population was removed. The adherent population was maintained in culture and passaged with trypsinization. For study, cells were plated on the third or fourth passage in 24-well trays (Costar) in 1 × 10⁵ cells/well.

Lymphocytes were separated from heparinized peripheral blood by density gradient centrifugation over Ficoll-Hypague (Bio-Rad) and then washed in PBS. The cell pellet was resuspended in RPMI 1640 (Gibco) with 10% fetal calf serum and incubated overnight at 37°C. After overnight incubation the non-adherent cell population was collected and readjusted to a final concentration of 2 × 10⁶ cells/ml. This cell suspension was incubated at 37°C in the presence of 4 μg/ml of Con-A (Flow).

After 48 hour the supernatant was harvested after centrifugation and stored at −70°C until needed¹³⁾.

Affinity purified human gamma interferon (1×10⁶ u/ml) and alpha interferon (1×10⁶ u/ml) were purchased from Interferon Science, USA.

A sterile coverslip was cut to the appropriate size and placed into each well of a 24-well tissue culture plate. A 1 ml aliquots of the synovial cell suspensions containing 2×10⁴ cells was added to each well. The plate was incubated at 37°c overnight until the synovial cells were fully attached to the coverslip. Then 50 μl of gamma, alpha interferon or Con-A CM in the appropriate dilutions or titers were added to each well. After an incubation of varying periods, the coverslip was washed in PBS and fixed in cold acetone¹¹⁾.

For indicrect immunofluorescence the coverslip was incubated sequentially in normal goat serum, monoclonal anti-HLA-DR (DAKO) or 22c6 (anti-la, a gift from Dr. Robert Winchester) antibody¹⁴⁾ and FITC-conjugated goat anti-mouse IgG (DAKO). The coverslip was then examined using fluorescent microscopy (Nikon). For immunoper-oxidase staining an acetone fixed cryostat section was incubated with normal horse serum followed by primary monoclonal antibody and biotinylated horse anti-mouse IgG (Vector) sequentially. Endogenous peroxidase activity was blocked by incubating the slides in 0.3% hydrogen peroxide in PBS for 30 min, followed by an avidin-biotinylated complex incubation for 45 min. The sections were developed by amino-ethyl-carbazol (AEC) or 3, 3′-Diaminobenizidine (DAB) and counterstained⁵⁾.

Non-inflammatory synovial tissue showed scattered areas of Ia antigen expression, both on synovial intima (Fig. 1A) and in subsynovium (Fig. 1B). In subsynovium. Ia expression was observed around blood vessels (Fig. 1C). In contrast rheumatoid synovium, with a broader layer of synoviocytes, expressed Ia antigen in nearly 100% of cells (Fig. 1D).

Cells from synovial tissue were readily dispersed using proteolytic enzymes. In the initial stage of culture, cells of varying morphology were present (Fig. 2A). Many small round cells were observed along with stellate cells and fibroblast like cells. The small round cells considered to be macrophages were positively stained with alpha-naphthyl acetate esterase (ANAE) (Fig. 2B & C). Passage of the primary cultures resulted in loss of the small round cells (Fig. 2D). In contrast spindle shaped fibroblast like cells became dorminate in the population. The large stellate cells persisted throughout the culture (Fig. 2E).

By the time of the third to fourth cell passage, regardless of cell type, la-positive cells were rarely seen. Therefore cells from the third or fouth passage were used for test. Treatment of a monolayer of synovial cells with gamma interferon reuslted in a dose dependent increase of Ia expression (Table 1). Maximal induction was observed with 100 u/ml gamma interferon, Ia expression and fluorescence intensity seemed to form a plateau at this concentration. The effect of gamma interferon was observed within 24 hours but required three to four days to become maximal. Some induction of Ia could be observed with a concentration as little as 1 u/ml of gamma interferon (Fig. 3A–3D).

In contrast, alpha interferon failed to induce Ia at any concentration in the range of 1–1000 μ/ml.

Con-A CM exerted a similar effect on Ia expression to gamma interferon. However, with a 10 vol % of Con-A CM, percentage of Ia inducible cells was not as high as that with gamma interferon.