Androgen metabolism and inhibition of interleukin-1 synthesis in primary cultured human synovial macrophages

The presence of androgen receptors on synovial macrophages in human normal and rheumatoid synovial tissues has been described previously. It is now reported that primary cultured human macrophages obtained from normal and rheumatoid synovia express functional androgen receptors. We have investigated the capacity of cultured macrophages to metabolize androgens and have found that these cells were capable of metabolizing testosterone to the bioactive metabolite dihydrotestosterone. Therefore, macrophages contain the key enzymes of steroidogenesis, in particular the 5α-treductase. Furthermore, interleukin-1β production by primary cultured rheumatoid macrophages was analysed, following exposure to physiological concentrations of testosterone (10−8 M). A significant decrease of IL-1β levels in conditioned media after 24 h (p < 0.05) was observed. It is concluded that androgens may act directly on human macrophages and may interfere with some of their functions via receptor-dependent mechanisms.


Introduction
A sex-related dimorphism in immune capabilities is thought to be related to the physiological effects of sex hormones on the immune system. [1][2][3][4][5] Generally, at least at physiological concentrations, oestrogens exert immunostimulatory effects, whereas androgens are immunosuppressive and a preponderance of autoimmune diseases in females compared with males is well recognized. In rheumatoid arthritis (RA), the synovial tissue is widely infiltrated by mononuclear cells and is considered the target tissue for a 'sexual dimorphism' in the immune response. [2][3][4][5][6][7] Sex hormones seem to play a role in immunomodulation via receptor-mediated mechanisms.
High affinity androgen receptors (AR) have already been detected in human maturing thymocytes. -11 However, it is only recently that androgen and oestrogen receptors have been detected in synovial macrophages (mCs) of normal and rheumatoid synovial tissues, by immunostaining of cryosections and by biochemical characterization in homogenized tissues. 12,13 The presence of androgen receptors on mCs induced us to analyse the metabolic pathway for androgens.
MCs were found to be capable of effectively metabolizing testosterone (Tes) to the bioactive dihydrotestosterone (DHT). In addition, since interleukin-l[3 (IL-113) is mainly synthesized by activated macrophages and mediates the Tand 138 Mediators of Inflammation Vol 4 1995 B-lymphocyte stimulation during the immune response, we investigated the influence of androgens on IL-I production by using cultured RA mCs. It was found that physiological concentrations of testosterone (10 -8 M) significantly decreased IL-l[3 content in conditioned media.
In conclusion we proved that human mCs express functional androgen receptors and should be modulated at least in some of their immune-related functions (i.e. IL-I]3 production) by the sex hormones that they actively metabolize.

Materials and Methods
Reagents: Anti-androgen receptor mAbs that are known to react to human androgen and that do not cross-react with other receptors or with glucocorticoid receptors, were purchased from BioReagents (Neshanic Station, NJ, USA were used as controls. 15 All the RA patients were found to be affected by Steinbrocker class II RA. At the time of surgery all the patients were being treated exclusively with nonsteroidal anti-inflammatory drugs. None had received any oral or intraarticular corticosteroid therapy and slow-acting antirheumatic drugs during the 4 months preceding the investigation, nor was the use of the oral contraceptive pill, 1,25-dihydroxyvitamin D3 or any hormone replacement therapy referred by the patients. [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18] All subjects had normal liver, renal, prostate and thyroid functions and were within 20% of ideal body weight. Freshly obtained tissue samples from the synovial lining were placed on ice, and stored at-20C in 30% sucrose-50% glycerol buffer (v/v) (0.25 M sucrose, 10 mM HEPES and 1.5 mM MgCl2, pH 7.4) until analysis for sex hormone receptors. Under these conditions, the concentration and molecular form of the receptor have been shown to be stable for at least 3 months. 19 At least four further samples collected from each subject and from different areas of the synovial tissue, were immediately snap-frozen in liquid nitrogen, cryoprotected with OCT (Miles, Noperville, IL, USA) and stored at -90C until used for immunohistopathological studies. Another series of synovial tissue samples was used on the same day for cell cultures. The synovial tissue was dissected away from fatty, capsular components, cut into 3-5 mm pieces, washed in Hanks' balanced salt solution and incubated in collagenase (1 mg/ml) for 3 h at 37C with constant agitation. The digest was then passed through a wire mesh with a pore size of 200 l.tm to separate dissociated cells from tissue debris. The cells were washed twice in collagenase and counted. The mean cell yield was 4.9 x 106 (range 0.53-16.8 x 106) and the mean viability was 88% (range: 69-98%). The cells were then cultured with RPMI supplemented with 10% FCS at 37C in humidified 5% CO2 in air. After 3 h nonadherent cells were removed and the adherent cells were cultured with RPMI supplemented with 10% FCS at 37C in humidified 5% CO2 in air for 72 h. The adherent population was subsequently stained for macrophage-specific antigens (see immunostaining section) and was found to be >80% ms. Every 24 h cytocentrifuge preparations (800 x g for 3 min) of suspended cells (gentle scraping) were obtained and stored at-80C until immunostaining.
Immunostaining of synovial cryosections and cultured synovial macrophages: Five m thick cryostatic sections (Cryocut 1800, Reichter-Jung, Leika, Nussloch bei Heidelberg, Germany) of RA and control synovia as well as cytocentrifuge preparations of cultured adherent synoviocytes, were mounted on glass slides, air dried, and fixed in cold 95% acetone for 10 min. A number of slides were incubated for 1 h with a rat anti-androgen receptor mAb (IgG) that also recognizes the DNA-binding sequence of the receptor (1:100). 2 After two washes with phosphate buffered saline, the second step was performed using the improved biotin-streptavidin-amplified detection system. Briefly, staining of anti-androgen receptor mAb, was carried out using the second-step concentrated link antibody (biotinylated anti-Ig specific respectively for mouse and rat primary antibodies) and the concentrated enzyme label (horseradish peroxidase-conjugated streptavidin). The same slides were treated a second time, with mAbs directed towards single different macrophage-associated antigens Ber-MAC3 (1:70) and MAC387 (1:70), then with the second-step concentrated link antibody and finally with the concentrated enzyme label (alkaline phosphatase). Control slides were treated in an identical manner except that the first step was replaced by nonspecific mouse or rat antibodies. The samples were therefore examined by using a Leitz microscope equipped with an epifluorescence system.
Steroid metabolism experiments: Methodological approaches and procedures used to measure metabolic pathways of steroids have been previously established and optimized. 2 Adherent synoviocytes (see synoviocyte culture) were harvested by mild trypsinization, counted in a haemocytometer and plated onto 6-well tissue culture plates at a density of 0.5-1 x 106 cells/dish. After 24 h, cells were washed twice with PBS-A and the medium substituted with FCS-free, phenol red free RPMI medium, containing 10 -8 M tritiated Tes as precursor. Controls were treated with FCS-free and phenol red free RPMI medium alone. Following 24 h incubation, medium was transferred to sterile plastic tubes (Costar) and stored at-80C until analysis; cells were washed three times using PBS-A and solubilized in 3 ml of 0.1% SDS at 37C for 15-30 min. Aliquots (100 t.tl) of the cell lysates were therefore used to estimate DNA content, as described elsewhere. 22 Steroid extraction was carried out using 1 ml of medium with 10 ml of dimethyl ether for androgens. After mixing at 4C for 30 min, the resulting aqueous phase was freeze-dried in a Speed Vac evaporatorconcentrator (Savant Instruments Inc., Farmingdale, NY, USA) and then resuspended using 970 l.tl of acetate buffer (0.75 M, pH 5.0) and 30 l-tl of Glusulase enzyme mixture (DuPont Co., Wilmington, DE). The solution was incubated at 37C for 18 h to hydrolyse steroid conjugates (sulphates and glucuronides). The ether phase was evaporated to dryness under nitrogen and then stored at-20C until required for analysis using reverse phase (RP)-HPLC. The HPLC system consisted of a model 324 system, equipped with a model 160 UV detector set at 280 nm (Beckman Instruments Inc., Berkeley, CA, USA) and with 'on-line' Flo-One/beta (model lC) three channel radiometric detector (Radiomatic Instruments, High Wycombe, Bucks, UK). Steroids were separated under isocratic conditions using a Spherisorb OS-11 (Aldrich Chimica, Milan, Italy) column (250 x 4.6 mmi.d.) and either 40% or 45% acetonitrile in citric acid (0.05 M) as optimized mobile phases for oestrogen and androgen separation respectively, with a flow rate of i ml/min. Routine data integration was automatically achieved and computed in net cpm by a Flo-One/beta F1B program (Radiomatic, Tampa, FL, USA). The amount of substrate turnover, which is proportional to the cytokine concentration, was determined colorimetrically by measuring the absorbance using a computerized reader (Vmax EASIA reader, Medgenix). Concentrations as low as 2 pg/ml of IL-1 are detectable with this assay. Cross-reaction with other cytokines including IL-10t, TNF and interferons have been reported to be insignificant. The intraassay coefficient of variation was 7% for IL-I. When double-immunostaining was performed using cytocentrifuge preparations of cultured ms, and the same mAbs, an intense staining was confirmed for the AR (Fig. I(B)). After staining for macrophagespecific antigens (Ber-MAC3 and MAC-387) over 90% of the cells from centrifuge preparations were found to be ms.

IL-lfJ assay
Metabolism of sex hormones in vitro by short-term cultured human synovial macrophages: Short-term primary cultures of RA ms were exposed to approximately physiological concentrations of Tes (10 -8 M) for 24 h; tritiated Tes was added to the cell cultures in order to evaluate the enzyme activity of these cells, in restrictive culture conditions (i.e., without foetal calf serum and in phenol red-free medium). Results of these three experiments showed an appreciable formation of dihydrotestosterone (DHT), the biologically active metabolite. Delta4-androstenedione (A4-Ad) formation, even at relatively low levels, was also detected (see Table 1). Experiments were carried out in triplicate; the reproducibility was acceptable, as revealed by the narrow standard deviation value, and good extraction efficiency was obtained (> 90% in all cases). Controls were negative for the presence of androgens and their metabolites in the medium.

IL-I} production by cultured synovial macrophages:
To test whether testosterone modulates IL-I production by cultured RA ms, the cells from five separate RA donors were incubated for 24 h in the presence of either medium alone (control cultures) or medium supplemented with physiological concentrations of Tes (10 -8 M) (IL-I[ was assayed in the supernatants by a specific EIA).
As shown in Fig. 2  press functional androgen receptors. The immunostaining with specific mAbs of the cytocentrifuge preparations of primary cultured ms confirmed intense staining for AR. The cells appear also to contain the key enzymes of steroidogenesis, as shown by their capability to form, in the short term, the active metabolites of testosterone. In particular the results of the study indicate that ms are endowed of 5x-reductase enzymes, the latter forming DHT, the more biologically active metabolite, in relatively short-term exposure (24 h).
In cultured ms the rate of conversion of Tes into active metabolic products (i.e., DHT) was found very close to that observed in classical target cells for androgen activity such as human prostate cancer cells (the conversion rate was 1.4% and 1.8%, respectively). 23'24 Furthermore, DHT levels were found in physiologically relevant amounts (0.56 pmol/100 btg DNA or 7.54 x 10 -1 M) as usually observed in classical target cells for peripheral androgen metabolism (i.e., human prostate cancer cells). 24 Finally, since interleukin-1 is mainly synthesized by activated macrophages and mediates lymphocyte stimulation during immune response, it was found that testosterone treatment (10 -8  I mRNA response are in progress. This inhibitory effect by androgens on macrophage ILproduction, was found in cultured RA ms obtained from patients of both sexes, therefore indicating that the effect is not sex-linked. At the same physiological concentration (10 -8 M) employed in this assay, it is known that Tes translocates androgen receptors to cell nuclei without affecting other steroid receptors (i.e., oestrogen, progesterone) or stimulating cell growth. 25 Furthermore, recent studies, have shown that even pharmacological concentrations of Tes (10 -M) inhibited IL-1[ secretion by peripheral blood mononuclear cells obtained from RA patients. 26 In contrast a dose-dependent influence of oestradiol (E2) on IL-I[ production by human peripheral monocytes has been described. 27 In particular, physiological concentrations of E2 (10 -8 or 10 -9 M) seem to increase and pharmacological (10 -5 M) concentrations seem to inhibit IL-I[ secretion. Similar results have been also obtained in primary cultures of RA ms (Cutolo, unpublished observations).
In all these experiments, the effect of androgens on IL-I[ production by monocyte/macrophages was not found to be sex-linked. In fact, physiological serum androgen concentrations, as observed in basal conditions in male subjects, may well decrease/inhibit the synthesis of IL-I when monocyte/macrophages are activated by different stimula. An immunosuppressive activity exerted by androgens is well assessed. Therefore the lower immune response observed in males, can also explain the reduced incidence of autoimmunity such as observed in Systemic lupus erythematosus (SLE) or in RA, in which females show a much higher susceptibility. In addition, as known, male patients affected by autoimmune diseases have been found to present low serum levels of androgens when compared with age-matched controls. 26 A sex hormone relationship is well assessed with the different F:M ratios observed in various ages of women and men affected by RA (a lower RA incidence in men during fertile ages and a higher incidence in elderly men; whereas the opposite situation occurs in women). 28 Interestingly, a higher incidence of immuno-mediated diseases such as SLE, secondary Sj6gren syndrome (SS) or RA is present in men with Klinefelter's syndrome more frequently than in normal men. Klinefelter's men are characterized by high oestrogen and low androgen levels. 29 Furthermore low androgens, at least in male RA patients, may be a genetically determined precursor rather than a consequence of the disease, as shown by its frequent association with the HLA antigen B15 even in healthy male controls. 3,31 All these observations may represent the basis for the effectiveness of androgen replacement therapy in RA as well as in Klinefelter's patients. 29,32 Recently the administration of dehydroepiandrosterone (DHEA), an adrenal hormone with mild intrinsic androgenic activity, has been reported to improve the disease activity in SLE patients. 3 In conclusion, androgens may at least partially exert their immunosuppressive activity by acting on human activated ms through respective receptormediated mechanisms. 4 These findings underscore the importance of sex hormones in immune system regulation and consequently confirm that both androgen and oestrogen imbalance may have significant implications in autoimmunity.