Effect of gold salts, D-penicillamine and benoxaprofen on type II collagen-induced arthritis in rats

Old and new therapeutics for Rheumatoid Arthritis: in vivo models and drug development

Development of novel drugs for treatment of chronic inflammatory diseases is to a large extent dependent on the availability of good experimental in vivo models in order to perform preclinical tests of new drugs and for the identification of novel drug targets. Here, we review a number of existing rodent models for Rheumatoid Arthritis in the context of how these models have been utilized for developing established therapy in Rheumatoid Arthritis and, furthermore, the present use of animal models for studies of novel drug candidates. We have studied the literature in the field for the use of in vivo models during development of anti-rheumatic drugs; from Methotrexate to various antibody treatments, to novel drugs that are, or have recently been, in clinical trials. For novel drugs, we have explored websites for clinical trials. Although a single Rheumatoid Arthritis in vivo model cannot mirror the complexity of disease development, there exist a number of good animal models for Rheumatoid Arthritis, each defining some parts in disease development, which are useful for studies of drug response. We find that many of the established drugs were not tested in in vivo models before being used in the clinic, but rather animal models have been subsequently used to find mechanisms for efficacy. Finally, we report a number of novel drugs, tested in preclinical in vivo models, presently in clinical trials.

Liposome encapsulated aurothiomalate reduces collagen-induced arthritis in DBA/1J mice

Collagen-induced arthritis (CIA) generated in rats or mice has long been a model system for the study of rheumatoid arthritis in humans. In particular, this system has been used to study the mechanisms and effects of anti-arthritic drugs in the treatment of the disease. Sodium aurothiomalate (ATM) is an agent often used to treat rheumatoid arthritis in humans; however, it possesses inherent toxicities which limits its usefulness. Liposome-encapsulated drugs are currently being developed to minimize the toxicities associated with a variety of potentially beneficial drugs. We have chosen to encapsulate ATM into small unilamellar vesicles (SUVs) to determine whether greater efficacy would be achieved in treating CIA with SUV ATM as compared to using the free drug. SUVs were prepared from hydrogenated egg phosphatidylcholine and cholesterol. These SUVs were very stable. Vesicles stored at 4 degrees C lost only 0.09% of encapsulated ATM (SUV ATM) after 14 days and were able to reduce collagen-induced arthritis in these mice. Animals treated by i.m. injections of SUV ATM exhibited a 50% reduction in symptoms. More importantly, histological examination of knee joints of the affected animals verified that SUV ATM treatment prevented cellular infiltration of lymphocytes into the synovia of the collagen-sensitized mice. Conditioned media from spleen cell cultures was assayed for the presence of inflammatory lymphokines that might be affected by SUV ATM to account for the success in suppressing collagen-induced arthritis.

Induction of nitric oxide synthase by lipopolysaccharide and its inhibition by auranofin in RAW 264.7 cells

In RAW 264.7 cells, a murine macrophage cell line, treatment with lipopolysaccharide (1 to 10 ng/ml) stimulated production of nitric oxide (NO), which was inhibited by L-N(G)-monomethyl-L-arginine acetate, an inhibitor of NO synthase. Auranofin, an orally active chrysotherapeutic agent, also inhibited the lipopolysaccharide-induced NO production in a concentration-dependent manner (0.3 to 3 microM). Other gold salts such as aurothioglucose and aurothiomalate had no effect. Western blot analysis demonstrated that the lipopolysaccharide (10 ng/ml)-induced expression of inducible NO synthase protein was inhibited by auranofin as well as by the protein synthesis inhibitor cycloheximide. In addition, the lipopolysaccharide-induced increase in the level of mRNA for inducible NO synthase was also lowered by auranofin. Furthermore, auranofin showed no direct effect on the conversion of [3H]arginine to [3H]citrulline by the cell lysate. These findings indicate that auranofin inhibits lipopolysaccharide-induced NO production by suppressing the expression of inducible NO synthase.

Dual effects of auranofin on prostaglandin E2 production by rat peritoneal macrophages

Incubation of rat peritoneal macrophages in medium containing various concentrations of auranofin (1, 3 and 10 microM) increased prostaglandin E2 production at 4 h in a concentration-dependent manner, in accordance with the increase in the release of [3H]arachidonic acid from membrane phospholipids. However, at 20 h, no stimulation of prostaglandin E2 production by auranofin was observed. When the peritoneal macrophages were incubated in the presence of 12-O-tetradecanoylphorbol 13-acetate (TPA), thapsigargin or A23187, prostaglandin E2 production at 4 and 20 h was enhanced. The stimulator-induced prostaglandin E2 production at 20 h was suppressed by 10 microM of auranofin. Western blot analysis demonstrated that auranofin inhibited the induction of cyclooxygenase 2 by TPA, thapsigargin or A23187 at 4 and 20 h. The level of cyclooxygenase 1 did not change by treatment with these stimulators in the presence or absence of auranofin. These findings suggest that auranofin has dual effects on prostaglandin E2 production: without stimulation, auranofin increases prostaglandin E2 production at 4 h due to the increased release of arachidonic acid which is converted to prostaglandin E2 mainly by cyclooxygenase 1, but inhibits the stimulator-induced late-phase prostaglandin E2 production by inhibiting the induction of cyclooxygenase 2.

Effects of gold sodium thiomalate, cyclosporin A, cyclophosphamide, and placebo on collagen-induced arthritis in rats

The prophylactic and therapeutic effects of gold sodium thiomalate, cyclosporin A, cyclophosphamide, and placebo on collagen-induced arthritis (CIA) were evaluated in DA rats. Prophylactic treatment with cyclosporin A and cyclophosphamide suppressed the arthritis incidence, clinical inflammation, destructive bone changes, and development of anti-collagen antibody in DA rats subsequently injected with porcine type-II collagen. Therapeutic treatment with cyclosporin A and cyclophosphamide had a definite suppression on established CIA when started 21 days after the initial collagen injection, but the suppression was less marked than that of prophylactic treatment. Gold had no impact on CIA in DA rats when administered either prophylactically or therapeutically.

Effects of indomethacin, cyclosporin, cyclophosphamide, and placebo on collagen-induced arthritis of mice

The long term effects of indomethacin, cyclosporin, cyclophosphamide, and placebo on collagen-induced arthritis in mice were tested under two different treatment protocols. A prophylactic experiment examined the effects of the daily drug administration for 180 days beginning one day before the first collagen injection. Under this dosage schedule, cyclophosphamide and cyclosporin decreased the severity of arthritis, while indomethacin did not. A therapeutic protocol examined the effects of these same drugs when daily administration was delayed until the animals had active disease at 78 days after the first collagen injection. Under this protocol, all three drugs reduced the progression of disease. In both protocols, the most significant suppression of arthritis was seen in animals receiving cyclophosphamide which was associated with a decrease in anti-collagen antibody levels. Collagen-induced arthritis in mice should be further investigated as a model to study the long term effects of "slow-acting" anti-rheumatic drugs.

A modified mouse air pouch model for evaluating the effects of compounds on granuloma induced cartilage degradation

1. Employing rat femoral head cartilage implanted in a 6 day old mouse air pouch, the effects of inflammatory stimuli (i.e. cotton pellets, carrageenan, zymosan) on the loss of proteoglycan and collagen and granuloma formation have been studied. 2. Wrapping of the cartilage in cotton resulted in granuloma formation with accelerated loss of proteoglycan and collagen over the 14 day implantation period. The amount of loss increased with increasing weight of cotton. 3. The effects of different classes of anti-rheumatic drugs on granuloma formation and proteoglycan and collagen loss from cotton wrapped femoral head cartilage in the mouse air pouch have been studied. 4. Non-steroidal anti-inflammatory drugs (NSAIDs) had no influence on granuloma formation, but in general accelerated the rates of proteoglycan and collagen loss. 5. Dexamethasone and prednisolone significantly reduced granuloma formation and had a marked protective effect on cartilage breakdown. 6. Of the slow acting anti-rheumatic drugs examined, only gold sodium thiomalate (GSTM) and dapsone significantly decreased cartilage loss, with an accompanying modest decrease in granuloma formation. 7. The immunosuppressants cyclophosphamide and methotrexate, but not azathioprine, reduced cartilage degradation, but had no effect on granuloma formation. 8. The results for the different classes of anti-inflammatory and anti-rheumatic drugs are discussed in relation to their effects in other animal models and their reported therapeutic activities in man. It is concluded that the mouse air pouch method as described offers advantages as an animal model over existing procedures to predict therapeutic efficacy in man.

Immunological abnormalities in rats with adjuvant-induced arthritis--II. Effect of antiarthritic therapy on immune function in relation to disease development

The effects of the experimental immunomodulatory agent tilomisole (Wy-18,251; (3-(p-chlorophenyl) thiazolo [3,2-a]benzimidazole-2-acetic acid) on disease development and immune function in rats with adjuvant-induced arthritis was assessed in comparison with indomethacin and levamisole. Daily p.o. administration of tilomisole (100-200 mg/kg/day) to M. butyricum-injected rats significantly reduced both edema and bone erosion in the uninjected paw. Moreover, tilomisole treatment restored to normal the diminished Con A-induced proliferative response and IL 2 synthesis observed in spleen cells from arthritic rats, but had no effect on macrophage IL 1 production. In contrast, levamisole treatment (25 mg/kg/day) of arthritic rats improved splenic immune function but did not influence paw edema or bone erosion. Conversely, indomethacin (1 mg/kg/day) significantly reduced paw edema and bone erosion but did not improve the deficient proliferative response or IL 2 synthesis by "arthritic" spleen cells. These results indicate that tilomisole possesses combined antiinflammatory and immunomodulatory activity in adjuvant-arthritic rats which is distinctly different from the effects of either indomethacin or levamisole. Moreover, these data suggest that tilomisole has potential disease-modifying activity in arthritis, which is currently being more closely examined in clinical trials.

Comparison of inflammatory changes in established type II collagen- and adjuvant-induced arthritis using outbred Wistar rats

Type II collagen- and adjuvant-induced arthritis in outbred Wistar rats were compared using parameters that measured the inflammatory response, cellular and humoral immunity, blood protein changes, drug metabolism and histopathological and bony changes of the inflamed paws. There was a lesser incidence (40-70%) and severity of collagen disease than the adjuvant model (incidence approximately 100%). The use of MDP increased the incidence and severity of collagen arthritis. The acute phase protein response (plasma fibrinogen) was similar in both models during the peak of inflammatory response. Drug metabolism was inhibited in both type II collagen boosted with MDP or M. butyricum sensitized rats with arthritis; however, arthritic rats sensitized with collagen alone produced no inhibition. Only collagen arthritic rats produced type II collagen antibody and exhibited delayed hypersensitivity to type II collagen. Bony changes as assessed by radiographic evaluation were more severe in adjuvant arthritic rats than in the collagen arthritic model; histopathological findings from these animals confirmed this observation. The primary lesions in both models were periosteal reaction of the bone and ankylosis. Several classes of antiarthritic drugs were compared in both models using paw edema measurements and bony changes by radiographic evaluation. Drugs with inhibitory activity in both models were indomethacin, methylprednisolone, D-penicillamine and gold sodium thiomalate. Levamisole, chloroquine and auranofin were inactive in both models.