The activation of the ventral tegmental area (VTA) can rebuild the tonotopic representation in the primary auditory cortex (A1), but the cellular mechanisms remain largely unknown. Here, we investigated the firing patterns and membrane potential dynamics of neurons in A1 under the influence of VTA activation using in vivo intracellular recording. We found that VTA activation can significantly reduce the variability of sound evoked responses and promote the firing precision and strength of A1 neurons. Furthermore, the compressed response window was caused by an early hyperpolarization as a result of enhanced circuit inhibition. Our study suggested a possible mechanism of how the reward system affects information processing in sensory cortex: VTA activation strengthens cortical inhibition, which shortens the response window of post-synaptic cortical neurons and further promotes the precision and strength of neuronal activity.

Valence detection and processing are essential for the survival of animals and their life quality in complex environments. Neural circuits underlying the transformation of external sensory signals into positive valence coding to generate appropriate behavioral responses remain not well-studied. Here, we report that somatostatin (SOM) subtype of GABAergic neurons in the mouse medial septum complex (MS), but not parvalbumin subtype or glutamatergic neurons, specifically encode reward signals and positive valence. Through an ascending pathway from the nucleus of solitary tract and then parabrachial nucleus, the MS SOM neurons receive rewarding taste signals and suppress the lateral habenula. They contribute essentially to appetitive associative learning via their projections to the lateral habenula: learning enhances their responses to reward-predictive sensory cues, and suppressing their responses to either conditioned or unconditioned stimulus impairs acquisition of reward learning. Thus, MS serves as a critical hub for transforming bottom-up sensory signals to mediate appetitive behaviors.

Frequency and intensity are two independent attributes of sound stimuli. Psychoacoustic studies have found that the sound intensity can affect the perception of frequency; however, the underlying neuronal mechanism remains largely unknown. To investigate if and how the sound level affects the frequency coding for auditory cortical neurons, we recorded the activities of neuronal ensembles and single neurons, as well as the synaptic input evoked by pure tones of different frequency and intensity combinations, in layer 4 of the rat primary auditory cortex. We found that the best frequency (BF) shifted bidirectionally with the increases in intensity. Specifically, the BF of neurons with a low characteristic frequency (CF) shifted lower, whereas the BF of neurons with a higher CF shifted higher. Meanwhile, we found that these shifts in the BF can lead to the expansion of high- and low-frequency areas in the tonotopic map, increasing the evenness of the BF distribution at high intensities. Our results revealed that the frequency tuning can bidirectionally shift with an increase in the sound intensity at both the cellular and population level. This finding is consistent with the perceptual illusions observed in humans and could provide a potential mechanism for this psychoacoustic effect.

No abstract

By measuring the frequency selectivity at different intensities in the primary auditory cortex of adult rats, we found that a small group of cortical neurons can exhibit relatively weak but robust selectivity at multiple frequencies that are different from the most preferred frequency. Both in vivo multi-unit recordings (26/93 recordings) and single-unit recordings (16/137 neurons) confirmed that the preferred frequencies are periodic and have an averaged bandwidth (BW) of 0.3-0.4 octaves, which leads to multi-peaked frequency selectivity. Interestingly, the averaged bandwidth of the ripple in the frequency response tuning curve was invariant with the sound intensity. An investigation of the synaptic currents in vivo also revealed similar multi-peaked frequency selectivity for both excitation and inhibition. While the excitatory and inhibitory inputs were relatively balanced for most frequencies, the ratio between excitation and inhibition at the peak and valley of each ripple was highly unbalanced. Since this multi-peaked frequency selectivity can be observed at the synaptic, single-cell, and population levels, our results reveal a potential mechanism underlying the multi-peaked pattern of frequency selectivity in the primary auditory cortex.