Axon regeneration in the absence of growth cones: acceleration by cyclic AMP

Regenerative failure of spinal axons is commonly ascribed to signaling of F-actin depolymerization and growth cone collapse by molecules such as the myelin-associated growth inhibitors. cAMP is thought to promote regeneration at least in part by neutralizing this effect, either by direct action in the growth cone or indirectly by transcriptional mechanisms. In vivo evidence for this is based mainly on partial lesion studies in which it is sometimes difficult to distinguish regeneration of injured axons from collateral sprouting by uninjured axons. Moreover, previous observations on fixed lamprey central nervous system (CNS) suggested that regeneration may not involve growth cones. To distinguish actively growing axons from static or retracting ones, fluorescently labeled large reticulospinal axons were imaged in the living, transected lamprey cord with and without application of cAMP analogs and then studied by 2-photon microscopy. Axon tip movements over 2-48-hour intervals indicated: 1) regeneration was intermittent; 2) cAMP decreased initial axon retraction and increased subsequent regeneration up to 11-fold; 3) the increase in regeneration was due to an increase in velocity of axon growth, but not in the time spent in forward movement; 4) tips of actively regenerating axons were more sharply contoured than static tips but no filopodia or lamellipodia were observed, even in db-cAMP; and 5) during active growth, axon tips contained vesicle-like inclusions and were highly immunoreactive for neurofilaments. Staining for F-actin and microtubules was variable and F-actin was not concentrated at the leading edge. Thus, cAMP accelerates regeneration of lamprey spinal axons without inducing formation of growth cones.