Functional Characterization of the Nogo-66 Domain during Growth and Regeneration in the Fish Visual System

Abstract

The mammalian central nervous system (CNS) lacks the regenerative capacity to regrow axons upon injury. RTN-4/Nogo-A, a member of reticulon-family of proteins, is a potent myelin associated inhibitor for axon growth and regeneration in the adult CNS of higher vertebrates. In stark contrast, CNS lesioned axons in the zebrafish (ZF) optic nerve regenerate readily and re-establish functional connections with the brain. This correlates well with the absence of the inhibitory NogoA-specific N-terminal domains from the zf-rtn4/nogo (reticulon-4) gene since Nogo-A blocks axon regeneration in mammals - and also in ZF retinal ganglion cell (RGC) axons in vitro. In addition to the N-terminal NogoA-specific domains (such as NIG∆20) rat rtn-4/nogo carries a second inhibitory domain at the C-terminal, called Nogo-66, which is 67% identical with ZF-Nogo66. Therefore the main question we wanted to answer in this project was Why fish is able to regenerate its lesioned axons despite the presence of the Nogo-66 . Thus, we examined whether ZF-Nogo66 is inhibitory, like its Rat counterpart, and how it may affect ZF axon regeneration, using three classical assays. In the outgrowth assay ZF RGCs extended roughly 1.7 and 2.7 times more axons on ZF-Nogo66 than on Rat-Nogo66 and Rat-NIG∆20, respectively. In the collapse assay 79% of the growth cones elongated unimpaired by ZF-Nogo66 whereas Rat-Nogo66 and NIG∆20 caused collapse in 78 and 82% of ZF growth cones, respectively. Furthermore, in the contact assay ZF RGC growth cones showed collapse (42%) or avoidance (42%) upon contact with transfected HeLa cells expressing Rat-Nogo66 (and 63% and 20%, respectively, with Rat-NIG∆20) as GPI-anchored EGFP-fusion proteins but grew onto cells expressing ZF-Nogo66 (65%). We further analysed the expression and distribution of ZF-Nogo66 and ZF Nogo receprors (NgRs); ZF-Nogo66 was mainly located inside the oligodendrocytes, and probably not exposed on the cell surface to be in contact with regenerating axons, but Nogo-66 on myelin debris may lye in the path of regenerating (or regrowing) axons. NgR was found on the axonal surface, and may mediate the inhibitory activity of the Rat-Nogo66 in tissue culture assays. PIPLC treatment of RGC axons abolished the inhibitory effect of the Rat-Nogo66, suggesting that a GPI-anchored receptor could mediate the Rat-Nogo66 inhibition, which is probably NgR. Our results suggest, surprisingly, that ZF-Nogo66 is functionally different from the mammalian Nogo-66, it has a growth-permissive effect on ZF axons, quite in contrast to its Rat-Nogo66 homolog which inhibits axon growth. Thus, not only the NogoA-specific domain is absent in fish, but Nogo-66, the second inhibitory domain has non-inhibitory properties in fish so that ZF RTN-4/Nogo does not impair axon regeneration.We also tried to render the permissive fish oligodendrocytes to inhibitory cells for axon growth by introducing the Rat-NIG∆20. Surprisingly, cells expressing the NIG∆20 exhibited a morphological change, they were losing their processes.To obtain insights into the function of the zf-rtn4 gene during development, we investigated its spatio-temporal expression pattern in developing ZF embryos and larvae. Zf-rtn4 is expressed early and later during development, showing strong expression in the somites at embryonic stages, and in the brain at larval stages. RTN-4 in fish could play a role during development in neuronal and non-neuronal structures.The present study together with earlier data imply that the fish CNS is optimized for successful neural regeneration and thus could serve as a model for the identification of parameters required for robust CNS axon regeneration in general.