ARTHRITIS & RHEUMATISM Vol. 60, No. 10, October 2009, pp 2892–2901 DOI 10.1002/art.24860 © 2009, American College of Rheumatology Soluble Neuropilin-2, a Nerve Repellent Receptor, Is Increased in Rheumatoid Arthritis Synovium and Aggravates Sympathetic Fiber Repulsion and Arthritis Alexander Fassold, 1 Werner Falk, 1 Sven Anders, 1 Thomas Hirsch, 2 Vladimir M. Mirsky, 2 and Rainer H. Straub 1 Objective. In inflammatory lesions, sympathetic nerve fibers disappear soon after the start of inflamma- tion. We identified sympathetic nerve repellents as possible causal agents in rheumatoid arthritis (RA). On nerve terminals, repellent factors bind to neuropilin-2 and its coreceptor. The aim of this study was to inves- tigate the role of neuropilin-2 in the synovial tissue of patients with RA and patients with osteoarthritis (OA) and in experimental arthritis. Methods. The density of neuropilin-2–positive fibers and cells positive for semaphorin 3F (a sympa- thetic repellent) was investigated using immunofluores- cence staining. Enzyme-linked immunosorbent assay was used to detect soluble neuropilin-2 in body fluids from patients with RA and patients with OA. An axon outgrowth assay and a neuropilin-2 Fc fusion construct (neuropilin-2Fc) were used to investigate semaphorin 3F–induced sympathetic nerve repulsion. In an animal model of type II collagen–induced arthritis, soluble neuropilin-2Fc was studied in vivo. Results. The synovial density of neuropilin-2– positive sympathetic nerve fibers was lower in RA than in OA, but the density of cells positive for semaphorin 3F was similar. In synovial fluid, the level of soluble neuropilin-2 was markedly higher in RA compared with OA. Mouse sympathetic ganglia served as an excellent model with which to study semaphorin 3F–induced nerve fiber repulsion. Neuropilin-2 and its coreceptor were present on sympathetic neurons, and semaphorin 3F bound to neuropilin-2Fc (binding constant 96 nmoles/liter). Semaphorin 3F dose-dependently in- creased sympathetic nerve fiber repulsion (at a 50% maximum response concentration of 160–210 nmoles/ liter). In contrast to our expectations, soluble neuropilin-2Fc did not inhibit repulsion but increased the repellent effect of semaphorin 3F. In experimental arthritis, therapy with neuropilin-2Fc aggravated ar- thritis. Conclusion. Soluble neuropilin-2 has no anti- repellent activity but aggravates sympathetic nerve fiber repulsion and arthritis. Increased shedding of neuropilin-2 is probably an unfavorable sign in RA. Neurotransmitters of the sympathetic nervous system are antiinflammatory by inhibiting the secretion of tumor necrosis factor, interleukin-2 (IL-2), IL-12, interferon-, and other proinflammatory signals (1) in rheumatoid arthritis (RA), osteoarthritis (OA), and experimental arthritis. This antiinflammatory effect is mediated via  2 -adrenergic receptors and A 2 adenosine receptors (2). The antiinflammatory effect is to be expected when neurotransmitter levels are high; for example, at low concentrations, norepinephrine and adenosine bind to  1/2 -adrenoceptors and A 1 adenosine receptors, which stimulate proinflammatory signals (for review, see refs. 1 and 3). Thus, the antiinflammatory effects of norepinephrine or adenosine are ultimately coupled to the presence of sympathetic nerve fibers, which are lost immediately after the commencement of inflammation (4,5). The loss of sympathetic nerve fibers is a general Supported by a DFG grant (Research Unit FOR696, STR 511/15-1,2) to Drs. Falk and Straub. 1 Alexander Fassold, MSc, Werner Falk, PhD, Sven Anders, MD, Rainer H. Straub, MD: University Hospital Regensburg, Regens- burg, Germany; 2 Thomas Hirsch, PhD, Vladimir M. Mirsky, PhD: University of Regensburg, Regensburg, Germany. Address correspondence and reprint requests to Rainer H. Straub, MD, Laboratory of Experimental Rheumatology and Neuroendocrino-Immunology, Division of Rheumatology, Depart- ment of Internal Medicine I, University Hospital Regensburg, Franz- Josef-Strauss-Allee 11, 93042 Regensburg, Germany. E-mail: rainer. straub@klinik.uni-regensburg.de. Submitted for publication April 14, 2009; accepted in revised form June 29, 2009. 2892