Focal selectivity for binaural sound pressure level in cat primary auditory cortex: Two-Way Intensity Network tuning

1. The influence of sound pressure level (SPL) at the two ears was studied in single-neuron responses recorded in high-frequency regions of primary auditory cortex (AI) of anesthetized cats. For each unit, many binaural combinations of SPL were tested by using best-frequency tone pips presented to each ear independently via sealed stimulus delivery systems. In the preceding paper, we illustrated the different forms of response observed in our sample of units. Here we explore in more detail the mechanisms underlying the properties of the largest single class of binaural response, characterized by joint nonmonotonic tuning to the SPLs at the two ears. We have described such units as being influenced by a Two-Way Intensity Network (TWIN). 2. Under binaural conditions, 62% of our sample of well documented neurons (81/130) exhibited a nonmonotonic relation between response magnitude and the SPL at one or the other ear. Of these units, 47 displayed clear bilateral nonmonotonicity (TWIN tuning), 17 units displayed only unilateral nonmonotonicity, and an additional 17 units showed intermediate (or transitional) characteristics between unilateral and bilateral nonmonotonicity. These characteristics can also be described in terms of average binaural level (ABL) and interaural level difference (ILD). Thus there is commonly a nonmonotonic relation between response magnitude and ABL and sometimes a TWIN tuning to ABL and ILD. The distribution of best frequencies for TWIN neurons is broad. 3. Under monaural conditions, TWIN neurons exhibit diverse properties. Some are responsive only under binaural conditions [i.e., predominantly binaural (PB)]. Some monaurally responsive TWINs are contralaterally excitable and ipsilaterally unresponsive (EO), some are ipsilaterally excitable and contralaterally unresponsive (OE), and a few are bilaterally excitable (EE). Monaural rate/level functions are monotonic for some of these neurons and nonmonotonic for others. Neurons of the PB class have previously been found to have nonmonotonic selectivity for ILDs near zero. In this study we have found that virtually all PB neurons are also nonmonotonically selective for ABL with different PB neurons having different best ABLs. 4. For TWIN neurons that respond monaurally, it is possible to demonstrate a mixed binaural influence. The optimal stimulus (or best binaural combination) for a TWIN neuron is associated with binaural facilitation. Flanking the most effective combination of ABL and ILD are less effective combinations that generate lower response magnitudes, either through threshold effects (at low SPLs) or through binaural suppression (at higher SPLs). These mixed binaural influences are not apparent in terms of latency. Typically, the latency to the initial discharge falls exponentially with increasing level at either ear, even in the presence of a significant reduction in number of discharges or probability of discharge. 5. The particular stimulus combination associated with the excitatory response focus varied considerably among TWIN neurons in our sample: best binaural combinations spanned the range of ABLs and ILDs presented. Most excitatory foci occurred with SPL combinations favoring the contralateral ear or when SPLs were equivalent at the two ears; some excitatory foci were found to favor the ipsilateral ear. Although the response overlap within the population was considerably less extensive than would occur in the absence of the nonmonotonic influences that shape TWIN tuning, a given TWIN tuning response matrix was never specific to a single combination of ILD and ABL (or of a single SPL combination). Each individual TWIN response delineated a range of SPL combinations, and considerable overlap existed in the foci of different neurons. 6. These results suggest that TWIN tuning might contribute to a representation of binaural stimuli that provides information about spatial position and source amplitude simultaneously within a single neuronal population rather than through the activities of separate populations. Each unique acoustic stimulus would evoke a unique swathe of activity across AI dominated by those neurons whose best binaural combinations and best frequencies represent the spectrum of the stimulus as experienced at the two ears. The pattern and extent of activation across the population could collectively constitute a sensitive representation of the binaural stimulus both in terms of interaural disparity and overall level.