Reorganization of connectivity in amphibian central auditory system following VIIIth nerve regeneration: time course

Abstract

In the frog, most neurons in the primary (dorsal medullary nucleus, DMN) and secondary (superior olivary nucleus, SO) auditory nuclei have V-shaped tuning curves, almost as narrowly tuned as those recorded in the nerve. Thus, the innervation pattern is such that if more than one excitatory afferent innervates a postsynaptic cell, they must all possess similar best frequencies (BFs). Similarly, binaural cells in these nuclei display matched frequency selectivities when acoustically stimulated via either ear. The VIIIth nerve was unilaterally severed and allowed to regenerate back into the DMN. At various postoperative intervals, extracellular single-unit recordings were made in the SO contralateral to the regenerated nerve, as that nucleus receives its dominant excitatory input (approximately 80%) from the contralateral side. Recordings were also made in the SO of a number of unoperated control animals. Functional reinnervation commenced between 4 and 5 wk postoperatively and by 6 wk, a normal innervation density, as judged by physiological criteria, was achieved. Single units of any best frequency represented within the frog's two auditory papillae could be recorded during earliest reinnervation. In general, the tuning curves of both monaural and binaural cells were V shaped in the 6 wk regenerates. Although many tuning curves were narrowly tuned (Q10dB greater than 1.0) as in unoperated animals, some were very broadly tuned (Q10dB less than 0.5). The mean Q10dB value for all contralaterally excited cells was 1.45 +/- 0.77 (SD), which was significantly lower than that of SO units in unoperated frogs (Q10dB = 1.66 +/- 0.52 (SD)). Binaural cells often had mismatched BFs and tuning curves. By 8 wk after nerve transection, tuning curves were as narrow as in unoperated animals (Q10dB = 1.64 +/- 0.68 (SD)), and the BFs of binaural cells evinced a greater match than at 6 wk. By 12 wk postoperatively, V-shaped tuning curves were still as narrow as in controls (Q10dB = 1.71 +/- 0.69 (SD)), and the tuning curves and BFs of binaural cells were well matched again. At all postoperative intervals, about 10% of the tuning curves in the SO of regenerates were W shaped. This was never seen in normal animals. The return of narrow V-shaped tuning curves in the majority of neurons and the recurrence of matched binaural cells in the SO are interpreted as evidence of specificity for potential postsynaptic targets in the DMN by regenerating auditory afferents.