It was shown recently that synovial fibroblast transformation into adipocytes reduced the expression of interleukin-6 (IL-6) and IL-8. However, the synovial fibroblast adipogenesis was inhibited in inflammatory conditions induced by the tumor necrosis factor-a (TNF-a). Furthermore, adipogenesis is often accompanied by leptin production, a proinflammatory adipokine in rheumatic diseases. In this study, we tested the phytohormone genistein for adipogenic and anti-inflammatory properties on human synovial fibroblasts. Results showed that genistein was able to transform synovial fibroblasts into adipocytes that expressed perilipin-A and produced adiponectin, but not leptin. Furthermore, genistein enhanced glucocorticoid-mediated synovial fibroblast adipogenesis and, in parallel, downregulated glucocorticoid-induced leptin and leptin receptor. Endogenous and TNF-a-induced expressions of IL-6, IL-8, p38, p65 and C/EBP-b were also downregulated by genistein, showing its anti-inflammatory properties. Peroxisome proliferatoractivated receptor-g (PPAR-g) agonist, rosiglitazone, had a synergic effect on genistein-induced adipogenesis, whereas the non-active tyrosine kinase inhibitor, daidzein, had a significantly inferior adipogenic activity than genistein. The Janus kinase-2 tyrosine kinase inhibitor, AG 490, mimicked the anti-leptin effect of genistein. These results showed that genistein-induced adipogenesis involves PPAR-g induction and tyrosine kinase inhibition. In conclusion, genistein, alone or coupled with glucocorticoids, have both adipogenic and anti-inflammatory effects on synovial fibroblasts.

Leptin plays a central role in maintaining energy balance, with multiple other systemic effects. Despite leptin importance in peripheral regulation of mesenchymal stem cells (MSC) differentiation, little is known about its expression mechanism. Leptin is often described as adipokine, while it is expressed by other cell types. We have recently shown an in vitro leptin expression, enhanced by glucocorticoids in synovial fibroblasts (SVF). Here, we investigated leptin expression in MSC from bone marrow (BM-MSC) and umbilical cord matrix (UMSC). Results showed that BM-MSC, but not UMSC, expressed leptin that was strongly enhanced by glucocorticoids. Transforming growth factor b1 (TGF-b1) markedly inhibited the endogenous-and glucocorticoid-induced leptin expression in BM-MSC. Since TGF-b1 was shown to signal via ALK-5-Smad2/3 and/or ALK-1-Smad1/5 pathways, we analyzed the expression of proteins from both pathways. In BM-MSC, TGF-b1 increased phosphorylated Smad2 (p-Smad2) expression, while ALK-5 inhibitor (SB431542) induced leptin expression and significantly restored TGF-b1-induced leptin inhibition. In addition, both prednisolone and SB431542 increased p-Smad1/5 expression. These results suggested the ALK-5-Smad2 pathway as an inhibitor of leptin expression, while ALK-1-Smad1/5 as an activator. Indeed, Smad1 expression silencing induced leptin expression inhibition. Furthermore, prednisolone enhanced the expression of TGF-bRII while decreasing p-Smad2 in BM-MSC and SVF but not in UMSC. In vitro differentiation revealed differential osteogenic potential in SVF, BM-MSC, and UMSC that was correlated to their leptin expression potential. Our results suggest that ALK-1/ALK-5 balance regulates leptin expression in MSC. It also underlines UMSC as leptin nonproducer MSC for cell therapy protocols where leptin expression is not suitable.

Inhibition of proapoptotic pathways in synovial fibroblasts is one of the major causes of synovial proliferation and hyperplasia in rheumatic diseases. We have shown previously that NF-κB inhibitor BAY 11-7085, through inactivation of PPAR-γ, induces apoptosis in human synovial fibroblasts. In this work we showed that BAY 11-7085 induced autophagy that preceded BAY 11-7085-induced apoptosis. Of interest, BAY 11-7085 induced Serine 211 phosphorylation and degradation of glucocorticoid receptor (GR). Glucocorticoid prednisolone induced both activation and degradation of GR, as well as autophagy in synovial fibroblasts. BAY 11-7085-induced cell death was significantly decreased with glucocorticoid inhibitor mifepristone and with inhibitors of autophagy. Both BAY 11-7085-induced autophagy and GR activation were down regulated with PPAR-γ agonist, 15d-PGJ2 and MEK/ERK inhibitor UO126. Inhibition of autophagy markedly decreased endogenous and BAY 11-7085-induced ERK phosphorylation, suggesting a positive feed back loop between ERK activation and autophagy in synovial fibroblasts. Co-transfection of MEK1 with PPAR-γ1 in HEK293 cells caused known inhibitory phosphorylation of PPAR-γ1 (Serine 112) and enhanced GR degradation, in the absence or presence of prednisolone. Furthermore, GR was both phosphorylated on Serine 211 and down regulated in synovial fibroblasts during serum starvation induced autophagy. These results showed that GR activation and PPAR-γ inactivation mediated BAY 11-7085-induced autophagy.

A solid phase radioimmunoassay (RIA) has been developed for detection of protein A in human serum or plasma using plates coated with specific goat anti-protein A IgG and affinity-purified anti-protein A F (ab')2 fragments labelled with 125Iodine as a tracer. This RIA detected 25 ng of protein A per ml of serum or plasma. The 'intra-plate' coefficient of variation (CV) was 4.26%, whereas the 'inter-plate' CV was 5.86%, even in the presence of aggregated IgG. The presence of protein A was searched for by this RIA in samples collected during ex vivo perfusion of the plasma of patients suffering from a metastatic mammary adenocarcinoma either over columns of 'native' protein A (seven patients; group A) or over columns of protein A in which the Fc-binding capacity was destroyed (five patients; group B). A protein A leaking occurred with the same incidence (about 20% of the sessions) in both groups. However, major acute side effects (hypotension, chills, mild fever, etc.) and necrolytic responses were seen only in some patients from group A. In these cases, the incidence of side effects, but not of tumoricidal responses, increased with the protein A serum concentrations reached 1 h after the end of the sessions (25-150 ng ml-1). The data suggest that the therapeutic efficacy of ex vivo protein A immunoadsorption in breast cancer is related to the Fc-binding capacity of the columns rather than to a leakage of 'native' protein A into the blood stream.