Role of Inhibitory Neurons in Stimulus-Specific Adaptation
Tohar S. Yarden, Adi Mizrahi, and Israel Nelken
(see pages 4629–4651)
The nervous system has evolved to minimize processing of repetitive or otherwise uninformative stimuli to accentuate representations of novel or salient stimuli. Neural adaptation to common stimuli occurs at multiple levels, from the periphery to the cortex. In primary auditory cortex (A1), for example, responses to repeated sounds are suppressed within a few trials, while responses to rare sounds are not. Synaptic depression and spike-rate adaptation have been proposed to contribute to this stimulus-specific adaptation (SSA). Yarden et al. have investigated the role of inhibitory interneurons in the superficial layers of A1 in SSA.
The authors recorded parvalbumin- (PV), somatostatin- (SST), vasoactive-intestinal-polypeptide- (VIP), or serotonin receptor- (HTR) expressing neurons while presenting a given tone either frequently or occasionally in different auditory contexts. In all neuron types, neural responses during tone presentation were larger when the tone was presented occasionally (i.e., as a deviant) than when it was presented frequently (i.e., as a standard). Thus, the neurons showed SSA. In addition to these responses during tone presentation, many PV, VIP, and SST neurons showed responses starting ∼90 ms after tone offset. Intriguingly, unlike early responses, the average late response was larger for standards than for deviants. Moreover, the late response to standards often increased over the first few trials.
To examine how inhibitory interneurons contribute to SSA in other neurons, the authors used optogenetic techniques to suppress neural activity. As expected, suppressing PV activity increased tone-evoked responses in pyramidal cells. Notably, pyramidal cell responses to deviants increased more than responses to standards, although the proportional change was similar. In contrast, suppressing VIP activity reduced responses of fast-spiking (presumably PV) neurons to standards, but it did not affect PV responses to deviants or alter pyramidal cell responses to tones in any context.
These results indicate that most classes of inhibitory neurons in the upper layers of auditory cortex undergo SSA. They also show that interneurons can modulate the auditory responses of other local neurons, with the amount of modulation depending on whether the stimulus is common or rare. Varying the amount of inhibition may therefore influence the ability of animals to detect unusual sounds in a steady stream of background noise.
Most neurons in chicken FRLx and optic tectum have annular (top) or round (bottom) SRFs. See Maldarelli et al. for details.
Auditory Spatial Receptive Fields in Chickens
Gianmarco Maldarelli, Uwe Firzlaff, Lutz Kettler, Janie M. Ondracek, and Harald Luksch
(see pages 4669–4680)
Much of what we know about sound localization was learned in barn owls—nocturnal animals that can localize prey in complete darkness relying solely on auditory cues. Owl brainstem nuclei compute differences in sound level and time of arrival of a sound at each ear [the interaural level difference (ILD) and interaural time difference (ITD), respectively]. This information converges on neurons in the inferior colliculus (IC) and its external nucleus (ICx) to produce spatial receptive fields (SRFs). Neurons in these areas are arranged to form a topographic map of auditory space, and they project to the optic tectum, where auditory and visual spatial maps converge. Although similar mechanisms of sound localization are used in other birds and mammals, differences exist. For example, in chickens, the ICx does not send direct projections to the optic tectum, but instead projects to the external portion of the formatio reticularis lateralis (FRLx), a recently discovered area that projects to the tectum. To elucidate the role of FRLx in sound localization, Maldarelli et al. compared auditory spatial tuning in the FRLx and optic tectum of chickens.
Sounds simulating 429 points in space were delivered to chickens via earphones. Most recorded neurons in the FRLx and optic tectum were tuned to ITD and/or ILD and had round or ring-shaped SRFs. Most SRFs were centered near 90° on the horizontal plane (azimuth), that is, in line with the ears. But SRF widths varied considerably, so that sounds coming from progressively more frontal (toward 0° azimuth) or rearward directions activated neurons with progressively wider SRFs. Notably, most neurons with annular SRFs were found in the FRLx, where neurons were arranged topographically, whereas most round SRFs were in the optic tectum, which, surprisingly, lacked a topographic map of auditory space.
These results suggest the optic tectum of chickens is specialized to represent sounds emanating from around 90° azimuth: the same space that is sampled by the central retina. In contrast, the FRLx contains a topographic map in which most neurons have annular receptive fields that exclude the area around 90° azimuth. Thus, FRLx might connect to motor areas that turn the head toward sounds emerging from locations outside the visual field.
Footnotes
This Week in The Journal was written by Teresa Esch, Ph.D.
