Endogenous NGF and nerve impulses regulate the collateral sprouting of sensory axons in the skin of the adult rat

We have investigated the co-involvement of endogenous NGF and impulses in the collateral sprouting of cutaneous sensory nerves in adult rats, specifically the A delta-axons involved in mechanonociception and the C-fibers that mediate heat nociception. Their collateral sprouting was measured by the progressive expansion, respectively, of the behaviorally defined "pinch" and "heat" fields into surrounding denervated skin (the light-touch A alpha-fibers do not sprout in adult mammals). The expansions of such "isolated" fields were totally prevented in animals injected daily with anti-NGF serum, but developed normally after treatment was discontinued. Light microscopic and EM examination of the skin confirmed that the effect of the anti-NGF treatment was attributable to its prevention of collateral sprouting. Initiation of treatment would also rapidly halt sprouting already in progress. Finally, intradermal injections of purified NGF protein would not only increase the rate of nociceptive fiber sprouting, but also evoke sprouting de novo within normally innervated skin (again, A alpha-axons were unaffected). We conclude that the collateral sprouting of intact nociceptive nerves following partial denervation of skin is entirely dependent on endogenous NGF. The observed latency of this sprouting was 10-12 d; we estimate, however, that at least 2 d of field expansion is required for its reliable detection. Thus, about 8-10 d are required for NGF levels in the skin to rise to effective levels, and for the neurons to respond and initiate sprouting. From indirect findings, the NGF component of this sprouting latency appears to be about 2 d. In accord with earlier findings, the remaining "initiation time" was reduced by 5-6 d if the neurons were briefly excited, even 2 d prior to the isolation of their fields. Unexpectedly, this phenomenon of "precocious sprouting" requires that endogenous NGF be available; the sprouting latency reverted to normal values when the conditioning impulses were evoked during a 2 d anti-NGF "umbrella." In contrast to the impulse-sensitive neuronal mechanisms involved in the initiation of sprouting, those underlying the sprouting rate were unaffected by nerve activity and were entirely dependent on the level of endogenous NGF. We suggest that interactions like that revealed in these studies between a sprouting agent and impulses that seem to prime the neuron's response to it contribute to plasticity within the nervous system.