NUCLEUS PREEMINENTIALIS OF MORMYRID FISH, A CENTER FOR RECURRENT ELECTROSENSORY FEEDBACK .1. ELECTROSENSORY AND COROLLARY DISCHARGE RESPONSES

1. The nucleus preeminentialis (PE) is a large central structure that projects both directly and indirectly to the electrosensory lobe (ELL) where the primary afferents from electroreceptors terminate. PE recei ves electrosensory input directly from ELL and also from higher stages of the electrosensory pathway. PE is thus an important part of a cent ral feedback loop that returns electrosensory information from higher stages of the system to the initial stage in ELL. 2. This study descri bes the field potentials and single-unit activity that are evoked in P E by electrosensory stimuli and by corollary discharge signals associa ted with the motor command that drives the electric organ to discharge . All recordings were extracellular in this study. 3. Two types of neg ative-going corollary discharge-evoked field potentials were found in PE: 1) a shallow, long-lasting negative wave with a latency at the pea k of similar to 11 ms, and 2) a more sharply falling and larger negati ve wave with a shorter latency at the peak of similar to 9 ms. The lon g-latency wave was predominant in the dorsolateral and posterior parts of PE, whereas the short-latency wave was predominant in the medial a nd rostral regions. Both waves were only found in PE and thus can serv e for its identification. 4. Electrosensory stimuli given either local ly to a restricted skin region or symmetrically to the entire body evo ked characteristic held potentials in both regions of PE. The mean lat ency between the stimulus and the peak of the response was 6.9 ms in t he early negativity region and 12.2 ms in the late negativity region. The responses to such stimuli were strongly facilitated by the electri c organ corollary discharge. 5. Field potential responses to the elect ric organ corollary discharge were markedly plastic. Responses to the corollary discharge plus a paired electrosensory stimulus decreased ov er time and the response to the corollary discharge alone was markedly enhanced after a period of such pairing. 6. Local electrosensory stim ulation of the skin showed that the caudal-rostral body axis is mapped from dorsal-medial to ventral-lateral in PE. The same somatotopy was found in the regions of the early and late negativities. The ventral a nd dorsal body appeared not to be separately mapped in PE. The areas r epresenting the head and chin appendage (''Schnauzenorgan'') are espec ially large in PE, due presumably to the high density of electrorecept ors in these areas. 7. Two main types of units were recorded in PE: 1) inhibitory (I) cells with a corollary discharge response that was inh ibited by an electrosensory stimulus to the center of their receptive fields; and 2) excitatory (E) cells with an excitatory response to ele ctrosensory stimuli that was facilitated by the corollary discharge. S ome of the E cells responded to the corollary discharge alone and some did not. Most cells appeared to be responding to input from mormyroma st electroreceptors, but a few cells were driven by ampullary electror eceptors and a few by Knollenorgan electroreceptors. 8. The corollary discharge effects on I cells and E cells were plastic and depended on previous pairing with a sensory stimulus. The corollary discharge faci litation of E cells and inhibition of I cells decreased during pairing with a sensory stimulus, and the corollary discharge-driven excitatio n of I cells was much larger after pairing than before. 9. The results provide an initial overview of a major component in the control of el ectrosensory information processing by recurrent feedback from higher stages of the system.