Tropism in nerve regeneration in vivo. Attraction of regenerating axons by diffusible factors derived from cells in distal nerve stumps of transected peripheral nerves.

We re-examined the hypothesis of Cajal3, later refuted by Weiss and Taylor20, that cells in distal stumps of transected peripheral nerves exert an attractive (tropic) effect on regenerating axons. This question was re-assessed in vivo using surgical materials and assay procedures not available to those workers. Proximal stumps of transected rat sciatic or cat peroneal nerves were inserted into the single inlet end of a hollow, Y-shaped Silastic implant. Regenerating axons were provided with alternative targets consisting of a vacant arm vs one occupied by a sciatic nerve graft (rat), or a tibial (Tout) vs peroneal (Pout) distal nerve stump (cat). In some cases Pout was rendered metabolically compromised relative to Tout by exposing the former to dry ice and inhibitors of DNA and RNA synthesis. At 4.5 or 6 weeks postoperatively, the number of regenerating axons in each fork of the implant was assessed by morphometric analysis (total number of non-myelinated and myelinated axons greater than 1 micron in diameter at 4.5 weeks, and total number of myelinated axons at 6 weeks postoperatively), or by quantification of an axonally transported label. Rat sciatic nerve fibers exclusively regenerated toward the nerve graft, suggesting the existence of a neurotropic lure. In cats, morphometric analysis revealed a 10-(4.5 week) and 6-fold (6 week) greater number of axons growing towards untreated Tout vs treated Pout. When both distal stumps were untreated, more axons were seen in forks leading to Pout. Analysis of transported label confirmed the preferential growth of axons towards untreated Tout vs treated Pout for both motor and sensory axons. In separate experiments, Nuclepore filters (0.2 microns, pore size) were inserted between distal nerve stumps and outlet ends of Silastic implants. Preferential regeneration toward untreated stumps was observed if the distance between proximal and distal nerve stumps was equal to but not greater than 4-5 mm. These results suggest that peripheral nerve fiber regeneration in vivo can be directed by cells in distal stumps of transected nerves, and that this effect can be mediated over distances of several millimeters via diffusible factors.