Most CNS axons have a low intrinsic ability to regenerate, the reasons for which remain to be solved. Three possible reasons are: Local mRNA translation. Axons in the peripheral nervous system contain many mRNAs, and the machinery to translate these into proteins. This ability is important for axon regeneration, because blocking local translation inhibits axon regeneration. Comparing the mRNAs from embryonic and adult PNS axons there are many changes, including the absence of kinesin mRNAs in adult axons. We find that one of these, kif3C, plays a key role in growth cone regeneration. Integrins and axon regeneration. In order to grow through the extracellular matrix axons must express appropriate integrins. The main matrix glycoprotein in the damaged CNS is tenascin-C, but tenascin-C binding integrins are lacking. We have transfected alpha9 integrin into neurons, giving them the ability to grow long axons on tenascin in vitro. Transduction of DRG neurons in vivo enhances their ability to regenerate their axons, but only modestly. One problem is that integrin transport into axons is blocked at the axon initial segment. Integrin trafficking relies on Rab11 and Rab coupling protein. Another problem is that CNS inhibitory molecules inactivate integrins. Gangliosides and axon regeneration. Axons contain the membrane enzyme PMGS, which desialyates GD1a ganglioside to produce GM1. We find that axotomy in halothane-anaesthetised rats activates PMGS, converting axonal ganglioside to GM1, and that blocking the enzyme inhibits axon regeneration. Retinal axons from the CNS do not convert their surface gangliosides to GM1 after axotomy, but application of an external sialidase causes both ganglioside conversion and promotes axon regeneration. The activation pathway involves Ca2+ and P38.