In an article in Arthritis Research & Therapy, Barthel and colleagues have described that cellular mRNA expressions for the neurotrophins nerve growth factor (NGF) and brain-derived neurotrophic factor (BDNF) are detectable both in the synovial tissue and in the synovial fluid of arthritic patients [1]. The expression for NGF was particularly noteworthy in samples of patients with rheumatoid arthritis. NGF was not expressed in fibroblast-like synoviocytes, as seen by ELISA analysis on culture supernatant. The authors concluded that infiltrating T lymphocytes and myeloid cells are the main sources of NGF in the inflamed peripheral joint. Nevertheless, the authors do not rule out the possibility that fibroblast-like synoviocytes can produce NGF under certain circumstances.

These findings give new evidence for the importance of neurotrophins for the inflammatory process in arthritis. The group of neurotrophins, which apart from NGF and BDNF is constituted of neurotrophin 3 and neurotrophin 4, has been considered of importance in various types of inflammatory conditions. In particular, NGF has been detected at high levels in regions with inflammation. Accordingly, there are high levels of NGF in the synovial fluid of rheumatoid arthritis patients. Neurotrophins are also of importance as factors that promote growth and survival of neurons, and they may be involved in the pathogenesis of pain.

In accordance with the findings of Barthel and collaborators [1], immunoreactions for NGF and BDNF have been found in the inflammatory infiltrates of the synovial tissue of arthritic mice but not in the synovial tissue of healthy mice [2]. Furthermore, cells in cultures of human synovial cells can produce NGF [3], and immunoreactions for BDNF have been detected in macrophages as well as in fibroblast-like synovio-cytes in the synovial tissue of rheumatoid arthritis patients [4]. Of further importance is the fact that immunoreactions for neurotrophins are also detectable in nerve structures of the synovial tissue [2, 5]. Immunoreactions for NGF and BDNF are also detectable on chondrocytes, including articular chondrocytes, as seen in studies on mouse joints [2].

Functional effects of neurotrophins are likely to occur within the synovial tissue. Expressions of the low-affinity neuro-trophin receptor p75 as well as the high-affinity neurotrophin receptor TrkA are therefore present in the synovial tissue of patients with spondyloarthritis [6]. Immunoreactions for p75 and the high-affinity receptor TrkB are also present in the synovial tissue of rheumatoid arthritis patients, the receptors being located in relation to the nerve structures [5]. The results of a recent study suggest there is an upregulation of NGF/TrkA in cytokine-activated fibroblast-like synoviocytes, suggesting that there is a cross-talk between NGF and its receptors in inflammatory arthritis [7]. Autocrine/paracrine effects of neurotrophins also appear to occur concerning the articular chondrocytes [2].

The effects of NGF in the inflamed synovium can be either proinflammatory or protective and regenerative. There is thus evidence that points in both directions. Interestingly, topical application of NGF to human corneal and skin ulcers can have healing actions [8] and, as seen in studies on rat injured ligaments, local application of NGF may improve the healing process [9]. Nevertheless, NGF does also have proinflam-matory effects. One possibility is that the effects of NGF in relation to inflammations vary over time, and furthermore that NGF actually is related to the modulation of and not the induction of the inflammation of joints [3].

Interestingly, neurotrophins have a relationship to TNFα. The production of both NGF and BDNF can thus be stimulated by TNFα. An effect on BDNF levels has been shown in response to anti-TNF treatment. The BDNF levels in plasma were found to be decreased in response to this treatment [5], a finding that is supported by the observation of a tendency for a decrease in plasma BDNF levels after anti-TNF treatment in a previous preliminary report [10]. Furthermore, the degrees of TrkA and p75 immunoreactions in spondyloarthritis synovitis were reported to be downregulated by anti-TNF treatment [6]. It cannot be excluded that the effects of anti-TNF treatment on BDNF levels are not only related to effects on local cells in the synovium, but also to effects on circulating cells and/or BDNF-containing neurons [5].

The results in the recent study by Barthel and collaborators and those obtained in other recent studies show collectively that neurotrophins should be further considered as factors of relevance in arthritic inflammation. It might be that the effects of these should be interfered with in arthritis. Further studies are needed in order to clarify whether antagonists or agonists to NGF (or to BDNF) should be applied.