We read the article by Barthel and colleagues  in this issue of Arthritis Research & Therapy with great interest. They reported that neurotrophin receptors/ligands and specifically nerve growth factor (NGF) and NGF receptor (NGF-R) (TrkA and p75) are expressed in synovial fluid (SF) cells and synovial tissue (ST) of patients with rheumatoid arthritis (RA) or spondyloarthritis (SpA). The authors also looked for the cellular source of NGF and demonstrated that the T cells and monocytes/macrophages derived from SF of patients with RA/SpA are enriched with NGF. Furthermore, they reported that ST-derived fibroblast-like synoviocytes (FLSs) did not produce NGF in vitro. In a recent publication, we observed similarly that NGF levels in SF were significantly higher in patients with psoriatic arthritis (PsA) (365.5 ± 85.2 pg/mL) or RA (120 ± 35 pg/mL) than in patients with osteoarthritis (OA) (30 ± 6 pg/mL) . However, in regard to the source of NGF, we observed that FLSs produced a significant amount of NGF. Here, we would like to share our observations about the NGF/TrkA system in human FLSs and its function.
Synovial biopsies from patients with meniscal injury without any other joint diseases or PsA, OA, or RA were collected. FLSs were isolated and examined for NGF/NGF-R expression in accordance with our standardized protocols . Using a highly sensitive enzyme-linked immunosorbent assay (ELISA) for human NGF (NGFEmax assay; Promega Corporation, Madison, WI, USA) and Hi-D fluorescence-activated cell sorting analyses, we observed that under basal conditions FLSs from healthy individuals express low levels of NGF/TrkA. However, there was a marked upregulation of NGF and TrkA in these FLSs following stimulation with tumor necrosis factor-alpha and interleukin-1-beta (Table (Table1).1). A critical observation was that FLSs from patients with PsA or RA produced spontaneously higher levels of NGF and had increased expression of TrkA compared with FLSs of patients with OA (Table (Table1).1). Furthermore, we observed that NGF significantly stimulated the proliferation of FLSs derived from PsA synovial tissue (Figure (Figure1).1). Barthel and colleagues  cultured FLSs from only one patient with RA and one patient with SpA and noticed that FLSs did not produce NGF, but the authors do mention that they can't rule out the production of NGF by FLS. We cultured FLSs from 15 subjects (Table (Table1),1), and it is also possible that the NGF ELISA kit that we used is more sensitive.
A fully formed pannus is characterized by proliferation of FLSs, inflammatory infiltrates, and a marked angiogenesis.NGF and its receptor system are known to influence angiogenesis and cell trafficking . In patients with RA or PsA, pannus tissue adheres to the surface of articular cartilage; proliferating FLSs produce proteinases that degrade cartilage and underlying cortical bone . We noticed that proinflammatory cytokines upregulate NGF/TrkA in FLSs, NGF/TrkA is upregulated in FLSs of inflammatory arthritis (Table (Table1),1), and NGF is mitogenic to FLSs (Figure (Figure1).1). These observations suggest an autocrine loop of NGF for FLS proliferation and suggest that dysregulated production of NGF has the potential to influence the inflammatory and proliferative cascades of PsA and RA.