Tetracyclines have recently been shown to have ''chondroprotective'' effects in inf lammatory arthritides in animal models. Since nitric oxide (NO) is spontaneously released from human cartilage affected by osteoarthritis (OA) or rheumatoid arthritis in quantities sufficient to cause cartilage damage, we evaluated the effect of tetracyclines on the expression and function of human OA-affected nitric oxide synthase (OA-NOS) and rodent inducible NOS (iNOS). Among the tetracycline group of compounds, doxycycline > minocycline blocked and reversed both spontaneous and interleukin 1␤-induced OA-NOS activity in ex vivo conditions. Similarly, minocycline > doxycycline inhibited both lipopolysaccharide-and interferon-␥-stimulated iNOS in RAW 264.7 cells in vitro, as assessed by nitrite accumulation. Although both these enzyme isoforms could be inhibited by doxycycline and minocycline, their susceptibility to each of these drugs was distinct. Unlike acetylating agents or competitive inhibitors of L-arginine that directly inhibit the specific activity of NOS, doxycycline or minocycline has no significant effect on the specific activity of iNOS in cell-free extracts. The mechanism of action of these drugs on murine iNOS expression was found to be, at least in part, at the level of RNA expression and translation of the enzyme, which would account for the decreased iNOS protein and activity of the enzyme. Tetracyclines had no significant effect on the levels of mRNA for ␤-actin and glyceraldehyde-3-phosphate dehydrogenase nor on levels of protein of ␤-actin and cyclooxygenase 2 expression. These studies indicate that a novel mechanism of action of tetracyclines is to inhibit the expression of NOS. Since the overproduction of NO has been implicated in the pathogenesis of arthritis, as well as other inf lammatory diseases, these observations suggest that tetracyclines should be evaluated as potential therapeutic modulators of NO for various pathological conditions.

SummaryClassically, osteoarthritis (OA) has been considered a noninflammatory disease . However, the detection of selected inflammatory mediators in osteoarthritic fluid, in the absence of significant inflammatory cell infiltrate, is increasingly appreciated. We sought to identify the inflammatory component in human OA-affected cartilage that may be involved in cartilage damage/destruction . Using Western blot analysis and an antibody to the conserved region ofnitric oxide synthase (NOS), we have observed up-regulation of NOS, one of the "key players" of inflammation, in chondrocytes of OA-affected patients. Remarkably, none of the cartilage samples examined from normal joints demonstrated detectable amounts of this NOS . Western blot analysis using the same ot-NOS antibody indicated that this NOS from OA-affected cartilage (OA-NOS) was larger in size than (and distinct from) transfected human hepatocyte or murine inducible NOS (iNOS) (150 versus 133 kD) and similar in size to neuronal constitutive NOS (ncNOS) . Antibodies specific for iNOS showed binding to murine and human iNOS but not to OA-NOS, endothelial constitutive NOS, or ncNOS . Antibodies specific for ncNOS bound to ncNOS and also to OA-NOS, but not to murine or human iNOS or endothelial constitutive NOS . Incubation of OA cartilage in serum-free medium resulted in spontaneous release, for up to 72 h, of substantial amounts of nitrite (up to^" 80 RM/100 mg wet tissue), which could be inhibited by at least 80% with various inhibitors of iNOS, including inhibitors ofprotein synthesis and transcription factor NF-KB, but which (unlike murine macrophage iNOS) was not sensitive to hydrocortisone or TGF-(3. Exposure of OA-affected cartilage to interleukin 1 R, tumor necrosis factor--(at, and lipopolysaccharide resulted in -20-50% augmentation of nitrite accumulation, which was also sensitive to cycloheximide and pyrrolidine dithiocarbamate. Hence, our data indicate that OA-NOS (based on immunoreactivity and molecular weight) is similar to ncNOS and that it releases nitric oxide, which may contribute to the inflammation and pathogenesis of cartilage destruction in OA.

These studies characterize differential mechanistic capacity of the immunophilin-binding immunosuppressive drugs (comparable to hydrocortisone) to inhibit both iNOS and COX-2 expression. Inhibition of iNOS and COX-2 mRNA accumulation by cyclosporin and rapamycin seem to be distinct. These studies also highlight potential anti-inflammatory properties of these drugs in addition to their known immunosuppressive activity.