Functional mechanisms of neuronal mismatch in the auditory midbrain and the medial prefrontal cortex

Zusammenfassung

Context in the environment around us highly influences our perceptions and the neural processing of sensory information. This doctoral thesis studies the mechanisms that shape the neural representations of sounds depending on the context in which they occur. Neurons in all sensory systems adapt rapidly, preserving energy while simultaneously enabling stimuli with potential survival or behavioral relevance to use additional processing resources. Stimulus-specific adaptation (SSA) is a special type of neuronal adaptation, specific to repeated and predictable stimuli, while preserving responsiveness to other different, unexpected, and probably more informative input. SSA has been linked to high-order brain processing such as deviance detection and perceptual inference. The nonlemniscal subdivisions of the inferior colliculus (IC) are the first sites in which these multifaceted coding properties emerge in the auditory hierarchy. Nevertheless, the molecular, cellular, and network mechanisms contributing to the generation of SSA are controversial and a matter of debate. SSA has been classically studied at the somatic spiking output, which results from the interaction of the synaptic inputs, its tuning characteristics, and membrane properties. Hence, in Study I, I report the passive properties, intrinsic properties, and auditory postsynaptic potentials under the oddball paradigm stimulation in 10 whole-cell patch-clamp recordings in vivo in the mouse IC (Valdés-Baizabal et al., 2020b). Although passive properties were similar, data suggest that intrinsic properties such as the firing patterns differed between lemniscal nonadapting and nonlemniscal adapting cells. SSA is absent at the synaptic level of the recorded neurons, which further demonstrates that SSA emerges in the nonlemniscal IC. Ascending along the auditory hierarchy, the encoding of the spectral properties of sound is subsequently substituted by more abstract representations allowing the detection of contextual changes in prefrontal regions. These high-order areas have been classically studied for the generation of automatic deviance detection using the scalp-recorded mismatch negativity (MMN) using similar oddball paradigms that also elicit SSA. However, the mechanisms that generate MMN and its neuronal correlate are neither clearly located nor understood in frontal cortices. Thus, Study II analyses the mechanisms governing deviance detection under the oddball paradigm in the rat medial prefrontal cortex (mPFC) within the predictive processing framework (Casado-Román et al., 2020). My results demonstrate in all mPFC fields and cortical layers that unpredictable auditory stimulation elicited stronger responses than the weak or even absent activity driven by predictable sounds. The time course of prefrontal spiking and LFP activity coincides with the large-scale MMN-like signals in the rat providing the missing link at the microscopic, mesoscopic, and macroscopic levels of automatic deviance detection. Hence, mPFC cells could model the possible neuronal correlate of the frontal MMN generators. Mismatch responses in mPFC are almost purely made of prediction error signaling activity and different in nature from those at the IC, auditory thalamus, and auditory cortex with an important effect of repetition suppression (comparisons with a previous study in our lab by Parras et al., 2017).