Auditory input to neurons in mesencephalic and rostral pontine reticular formation: An electrophysiological and horseradish peroxidase study in the cat

1. A total of 383 neurons were isolated in the midbrain of 21 chloralose-anesthetized cats. The largest samples were histologically localized to the mesencephalic and rostral pontine reticular formation (RF; n=185) and the inferior colliculus (IC; n=98). 2. The majority (90%) of RF neurons encountered were responsive to acoustic stimulation, and many (60%) acoustically responsive neurons also responded to visual and/or somatosensory stimulation. Although these proportions are undoubtedly biased by the search strategy and stimulus repertoire employed, it appears that a large proportion of RF neurons receive auditory input and that the majority of acoustically responsive neurons are polymodal. 3. Most acoustically responsive RF neurons responded to tone- or noise-burst stimulation with simple onset responses; a small proportion (8.5%) were characterized by an onset response followed by some form of later discharge. Only 2% of RF neurons exhibited sustained responses of the type observed in adjacent auditory lemniscal structures. Acoustic response latencies for most RF units were in the range 10-30 ms. 4. All RF neurons were characterizied by broad frequency tuning, responding over 5-6 octaves or more at near-threshold intensities. Some units were relatively insensitive to pure tones but had low noise thresholds; for these neurons, convergent frequency input apparently results in greater sensitivity to spectrally complex sounds. 5. Almost all RF neurons were binaurally influenced, the largest proportion (67%) receiving excitatory input from each ear. The responses of neurons monaurally excited only by stimulation of the contralateral ear were most commonly facilitated by simultaneous ipsilateral stimulation, in contrast to such cells in IC for which the ipsilateral influence was most often inhibitory. 6. In a further series of five cats, small iontophoretic deposits of horseradish peroxidase were made into acoustically responsive regions of RF. Labeled cells were found in a number of auditory lemniscal and acoustically responsive extralemniscal structures ipsilateral and contralateral to the injection site. The most consistent labeling in the auditory lemniscal pathway was seen in the nuclei of the lateral lemniscus, although a few labeled cells were found in the superior olivary complex, the inferior colliculus, and nucleus sagulum. Consistent labeling in a number of midbrain extralemniscal structures that themselves receive auditory input (viz., the deep layers of the superior colliculus, periaqueductal gray, and cuneiform nucleus) suggest that indirect sources are also important in providing auditory input to RF. 7. Both the electrophysiological and anatomical evidence indicate a high degree of within-modality convergence in the auditory input to RF. The consequence of this convergence appears to be that the information conveyed to RF and presumably transmitted over its ascending and descending projections concerns the occurrence of acoustic signals but not the details of their spectral composition or spatial location.