We have previously demonstrated that administration of antisera against NGF (anti-NGF) can have profound effects on developing primary afferents (Ritter et al., 1991). Chronic administration of anti-NGF to rats beginning on the day of birth results in a severe depletion of cutaneous A delta high-threshold mechanoreceptors (HTMRs) from the sural nerve. Here we have carried out further experiments in order to define the period of time over which this change in the cutaneous afferent population can be produced, and to investigate a possible mechanism for the change. Treatment with anti-NGF from postnatal day (PND) 0–14 resulted in a depletion of cutaneous A delta HTMRs from the sural nerve and also a 20% loss of sensory neurons. However, treatment from PND 2–14 produced an identical deficit of HTMRs without any accompanying cell death. Thus, the depletion of cutaneous A delta HTMRs can be achieved in the absence of cell death induced by anti-NGF treatment. It was also found that a 7 d treatment from PND 4–11 was sufficient to reproduce this effect, but that 7 d treatments earlier (PND 2–9) or later (PND 7–14) within the first 2 weeks were much less effective. This critical period, PND 4–11, corresponds to a period of anatomical change in the innervation of the skin, from epidermal innervation to primarily dermal innervation (Fitzgerald, 1967; Reynolds et al., 1991). In every case where anti-NGF treatment reduced the proportion of HTMRs, there was a reciprocal increase in the proportion of sensitive A delta hair follicle (D-hair) afferents. We hypothesize that in the absence of NGF, developing cutaneous A delta HTMRs do not die but innervate novel targets in the dermis and become D-hair afferents instead.

Nerve regeneration is associated with plasticity of sensory neurons, so that even muscle afferents directed to skin form mechanosensitive receptive fields appropriate for the new target. STOML3 is an essential mechanotransduction component in many cutaneous mechanoreceptors. Here we asked whether STOML3 is required for functional and anatomical plasticity following peripheral nerve regeneration. We used a cross-anastomosis model adapted to the mouse in which the medial gastrocnemius nerve was redirected to innervate hairy skin previously occupied by the sural nerve. We recorded from muscle afferents innervating the skin and found that in wild-type mice their receptive properties were largely identical to normal skin mechanoreceptors. However, in mice lacking STOML3, muscle afferents largely failed to form functional mechanosensitive receptive fields, despite making anatomically appropriate endings in the skin. Our tracing experiments demonstrated that muscle afferents from both wild-type and stoml3 mutant mice display remarkable anatomically plasticity, forming new somatotopically appropriate synaptic terminals in the region of the dorsal horn representing the sural nerve territory. The dramatic reduction in stimulus evoked activity from the cross-anastomosed gastrocnemius nerve in stoml3 mutant mice did not prevent central anatomical plasticity. Our results have identified a molecular factor that is required for functional plasticity following peripheral nerve injury.