RAPID COMMUNICATION Oligodendrocyte-myelin glycoprotein (OMgp) is an inhibitor of neurite outgrowth Vicky Kottis,* Pierre Thibault,Daniel Mikol,à Zhi-Cheng Xiao,* ,1 Rulin Zhang,* , Pauline Dergham§ and Peter E. Braun* *Department of Biochemistry, McGill University, Montreal, Quebec, Canada Institute for Biological Sciences, National Research Council, Ottawa, Canada àDepartment of Neurology, University of Michigan, Ann Arbor, Michigan, USA §Department of Pathology, Universite ´ de Montre ´al, Montreal, Quebec, Canada Abstract A protein fraction purified from bovine brain myelin, previously called arretin because of its ability to inhibit neurite outgrowth, has been identified as consisting predominantly of oligodendrocyte-myelin glycoprotein (OMgp). We show that it is a potent inhibitor of neurite outgrowth from rat cerebellar granule and hippocampal cells; from dorsal root ganglion explants in which growth cone collapse was observed; from rat retinal ganglion neurons; and from NG108 and PC12 cells. OMgp purified by a different procedure from both mouse and human myelin behaves identically in all bioassays tested. Keywords: arretin, myelin inhibitors, nerve regeneration, neurite outgrowth, OMgp. J. Neurochem. (2002) 82, 1566–1569. The failure of injured axons to regenerate long distances in the adult mammalian CNS leads to permanent paralysis and other functional deficits such as those seen after spinal cord injuries. Although axons do not regenerate through adult CNS tissue, they retain the ability to regrow for long distances if provided with an appropriate cellular environment, for example a peripheral nerve graft (David and Aguayo 1981). Work by Schwab and colleagues led to the discovery that this failure of axons to regenerate was likely to be caused, in part, by the influence of axon growth inhibitory activity associated with myelin (Caroni and Schwab 1988; Schwab and Bartholdi 1996; Horner and Gage 2000). Nogo-A (Chen et al. 2000; Fournier et al. 2001; GrandPre et al. 2002), myelin- associated glycoprotein (MAG) (McKerracher et al. 1994; Mukhopad- hyay et al. 1994), and several proteoglycans (Niederost et al.1999)have been identified as myelin associated molecules that impede axonal regeneration in the adult mammalian CNS. InourstudiesonMAGasaneuritegrowthinhibitor(McKerracher et al. 1994), we observed non-MAG containing chromatographic fractions eluting in 2–3 M NaCl solution from an anion exchange column that contained very low concentrations of protein but were highly active in bioassays. Further chromatographic purification on a peanut agglutinin lectin, as an affinity support, resulted in column fractions having very potent neurite growth inhibition. We called this potent activity ÔarretinÕ. However, these biologically active fractions still contained several electrophoretically separable components detected on silver-stained gels. Wenowshowthat:(i)thedominantproteinof ÔarretinÕ isoligodendrocyte- myelin glycoprotein (OMgp); (ii) authentic OMgp isolated by a different procedureandrecombinantOMgpbothinhibitneuriteoutgrowth.Aspects ofthisworkwerepresentedattheTwentyThirdInternationalSymposium on Spinal Cord Trauma, Montreal, 2001b. Materials and methods Purification of arretin and OMgp Myelin was prepared from bovine brain, extracted with octylgluco- side salt, and chromatographed on diethylaminoethyl (DEAE)-Sepharose (Pharmacia) as previously described (McKerracher et al. 1994). Column fractions eluting in 1.5–2 M NaCl solution were pooled, diluted three- fold, and loaded on a peanut agglutinin (PNA)-agarose column. Following a wash with 2 M NaCl in 20 mM triethanolamine (pH 7.5) and a cocktail of protease inhibitors arretin was eluted with 0.5 M D-galactose in the wash buffer. The eluate was dialysed against H 2 O and lyophilized. The protein was re-dissolved in a minimum volume of H 2 O. OMgp from either mouse or human brain was purified following the method described by Mikol and Stefansson (1988). The anti-OMgp used for western blots and for immunodepletion experiments was raised in rabbits and affinity purified (Dr Mikol, University of Michigan, MI, USA). All chemicals were purchased from Sigma-Aldrich Canada Ltd. Immunodepletion of OMgp In order to establish that the bioactivity in our sample of OMgp was not caused by minor contaminants, we subjected the sample to four rounds of immunodepletion (McKerracher et al. 1994). Briefly, the mouse OMgp sample in 50 mM Tris buffer, pH 7, was incubated with the affinity purified rabbit anti-OMgp overnight at 4°C with rotation. Protein A beads were added for 3 h with rotation, centrifuged (50 000 g, 10 min) and the supernatant was subjected to three repeat treatments as described above. The final depleted samples, as well as the protein A beads containing the immunosorbed OMgp, were separated by sodium dodecyl sulfate ) polyacrylamide gel electrophoresis (SDS–PAGE) for silver staining and western blotting, and bioassays were performed as described below. Resubmitted manuscript received July 16, 2002; accepted July 18, 2002. Address correspondence and reprint requests to Peter E. Braun, Department of Biochemistry, McGill University, Montreal, Quebec, Canada H3G 176. E-mail: peter.braun@mcgill.ca 1 Present address: Department of Anatomy, National University of Singapore, Singapore, and Dept of Clinical Research, Singapore General Hospital, Singapore. Abbreviations used: DEAE, diethylaminoethyl; MAG, myelin-associated gly- coprotein; OMgp, oligodendrocyte-myelin glycoprotein; PNA, peanut agglutinin; SDS–PAGE; sodium dodecyl sulfate ) polyacrylamide gel electrophoresis. Journal of Neurochemistry , 2002, 82, 1566–1569 1566 Ó 2002 International Society for Neurochemistry, Journal of Neurochemistry , 82, 1566–1569