Abstract and Introduction
Abstract
Innervation of the joint with thinly myelinated and unmyelinated sensory nerve fibres is crucial for the occurrence of joint pain. During inflammation in the joint, sensory fibres show changes in the expression of receptors that are important for the activation and sensitization of the neurones and the generation of joint pain. We recently reported that both neurokinin 1 receptors and bradykinin 2 receptors are upregulated in dorsal root ganglion (DRG) neurones (the cell bodies of sensory fibres) in the course of acute and chronic antigen-induced arthritis in the rat. In this study, we begin to address mechanisms of the interaction between fibroblast-like synovial (FLS) cells and sensory neurones by establishing a co-culture system of FLS cells and DRG neurones. The proportion of DRG neurones expressing neurokinin 1 receptor-like immunoreactivity was not altered in the co-culture with FLS cells from normal joints but was significantly upregulated using FLS cells from knee joints of rats with antigen-induced arthritis. The proportion of DRG neurones expressing bradykinin 2 receptors was slightly upregulated in the presence of FLS cells from normal joints but upregulation was more pronounced in DRG neurones co-cultured with FLS cells from acutely inflamed joints. In addition, the expression of the transient receptor potential V1 (TRPV1) receptor, which is involved in inflammation-evoked thermal hyperalgesia, was mainly upregulated by co-culturing DRG neurones with FLS cells from chronically inflamed joints. Upregulation of neurokinin 1 receptors but not of bradykinin 2 and TRPV1 receptors was also observed when only the supernatant of FLS cells from acutely inflamed joint was added to DRG neurones. Addition of indomethacin to co-cultures inhibited the effect of FLS cells from acutely inflamed joints on neurokinin 1 receptor expression, suggesting an important role for prostaglandins. Collectively, these data show that FLS cells are able to induce an upregulation of pain-related receptors in sensory neurones and, thus, they could contribute to the generation of joint pain. Importantly, the influence of FLS cells on DRG neurones is dependent on their state of activity, and soluble factors as well as direct cellular contacts are crucial for their interaction with neurones.
Introduction
The inflammatory response in organs is produced by numerous inflammatory cell types. These cell types communicate with each other in order to develop an appropriate inflammatory reaction. A large amount of information on the mechanisms of interaction of different inflammatory cells has been obtained from co-culture systems of different cell types, such as T cells and monocytes, T cells and endothelial cells, T cells and fibroblasts, monocytes and fibroblasts, and macrophages and fibroblasts. These data have established the importance of both cell-cell contacts and mediators for the production of the inflammatory activity.
Most tissues are innervated, and nerve fibres play an important role in inflammatory diseases. The activation of nociceptive sensory afferent fibres ('pain fibres') evokes pain, a major symptom of inflammatory diseases. Furthermore, there is growing evidence that primary afferent neurones as well as sympathetic nerve fibres influence the inflammatory process through efferent processes. Despite the functional cross-talk between the inflammatory process and neurones, the mechanistic analysis of interactions between non-neuronal inflammatory cells and neurones has not been carried out in great detail. Recently, a first report appeared on the influence of neurones in the central nervous system on T cells and the potential role of neurone-T cell interactions on experimental autoimmune encephalomyelitis.
The somata of primary afferent neurones are located in the dorsal root ganglia (DRG). Similar to the sensory endings, the somata of these neurones express ion channels and receptors that are important for the activation and/or sensitization of these neurones, and they thus seem to represent the total primary afferent neuron in this respect. In addition, the expression of ion channels and receptors in the somata is altered during peripheral inflammation. We recently took DRGs from normal rats and from rats with unilateral acute and chronic antigen-induced arthritis (AIA) in the knee, cultured them for one day and determined which proportion of DRG neurones express receptors for bradykinin (BK) and substance P (neurokinin 1 (NK1) receptors). In lumbar DRGs but not in cervical DRGs from AIA rats we found a pronounced increase in the proportion of neurones exhibiting BK receptors and NK1 receptors. The upregulation of these receptors found in this study and in other studies on inflammation is thought to be involved in the inflammatory pain response because both BK and substance P activate and/or sensitize proportions of primary afferent neurones for mechanical stimuli, which is a mechanism of mechanical hyperalgesia. Indeed AIA rats show limping of the inflamed knee and a lowered pain threshold when pressure is applied to the knee. In addition, the transient receptor potential V1 (TRPV1) receptor is an ion channel that is involved in thermal inflammatory hyperalgesia. Some studies, but not others, have identified an upregulation of TRPV1 receptors in DRGs in inflammatory models.
We have begun to identify mechanisms that cause the upregulation of BK, NK1 and TRPV1 receptors in DRG neurones. In the present study we co-cultured DRG neurones with fibroblast-like synovial (FLS) cells from either normal knee joints or from acutely or chronically inflamed knee joints from AIA rats. FLS cells are key players in the propagation of joint inflammation and joint destruction during rheumatoid arthritis whereas DRG neurones are key players in the development of chronic pain. We addressed three questions. First, is the expression of these receptors in DRG neurones influenced by the presence of FLS cells? Second, do FLS cells from normal and inflamed knee joints exert different effects on receptor expression in DRG neurones? Third, are effects of FLS cells on DRG neurones mediated by soluble mediators (is the supernatant of FLS cells sufficient?) or is the presence of the FLS cells with cellular contacts important? Preliminary results have been reported.