Brain Handles "Unexpected" Neural Circuits

The experience of different sensory stimuli enables humans to make associations between them, especially when they are presented as part of a series of sensory stimuli. For example, we learn to anticipate the appearance of thunder after seeing lightning. If no anticipated stimulus appears, we are surprised because there is a mismatch between the expectation and the actual experience.

May 7, 2025: Shankar Sachidhanandam and his team at the Department of Physiology, University of Bern, Switzerland, published a paper in Nature Communications entitled "Top-down modulation of sensory processing and mismatch in the mouse posterior parietal cortex". This study establishes an association paradigm for auditory prediction of tactile stimuli in head-fixed mice, and finds that posterior parietal cortex (PPC) L2/3 neurons dynamically encode stimulus sequence match/mismatch signals by two-photon imaging and present anticipatory correlated activity. Experiments confirmed that projections from the secondary motor cortex (M2) to the PPC influence these neural representations through top-down modulation: inhibition of M2 neurons specifically attenuates anticipatory and population mismatch responses in the PPC. This suggests that M2 may act as a high-level center that transmits learned sensory predictive signals to the PPC to modulate sensory information processing, revealing the neural circuitry underlying predictive coding between cortical layers.

Products

At Creative Biolabs, we offer a wide range of products and services to fuel your neuroscience research. Please contact us for more information.

Overview

The authors of the present study revealed the neural mechanisms underlying the dynamic encoding of sensory associations and anticipatory updating in L2/3 neurons of the mouse anterior cingulate cortex (PPC) through auditory-tactile sequences. It was found that:

1) the PPC characterizes matching and mismatching signals through different neuronal populations, respectively;
2) sensory learning inhibits anticipatory bottom-up inputs (as evidenced by a significant attenuation of the tactile response from the matching phase to the staggering phase);
3) feedback from higher cortex (e.g., M2) both inhibits and enhances primary sensory processing, and this bi-directional modulation may be realized through specific inhibitory neurons (PV/VIP/ SST-positive) to achieve this.

These findings provide new evidence for understanding the microloop mechanisms of predictive coding between cortical layers, suggesting a critical role for inhibitory interneurons in integrating top-down anticipation with bottom-up sensory signals.

Findings of the Study

• Create sensory associations at PPC and report mismatches

To investigate the establishment of sensory stimulus associations, the authors sequentially performed four experimental phases on head-fixed mice: 1) a "sound" phase (auditory stimuli of varying intensity); 2) a "whisker" phase (whisker stimulation only); 3) a "paired" phase (strong acoustic stimuli followed immediately by tactile stimuli to establish a predictive relationship); and 4) a 'staggered' phase (80% paired trials + 20% randomly omitted tactile "mismatch " trials). L2/3 neuronal activity in the PPC area of mice was recorded by two-photon calcium imaging (virally transfected with GECI), combined with IOS imaging to localize the PPC area.

The authors screened neurons that responded to either stimulus and retained significantly responding neurons for analysis after correction for upset data. The results showed that 37.3% (329/834) of the neurons responded to the whisker stimulus during the "whisker" phase. In the 'paired' phase, whisker stimulation evoked stronger responses (p=7.9×10^-4) and activated a new population of neurons (only 33 out of 154 paired-response neurons overlapped with whisker-responsive neurons), suggesting that sensory associations recruited new neural circuits. Supplementary data also analyzed the activity patterns of sound-responsive neurons.

• Neurological basis of anticipation in the anterior cingulate cortex

The authors further explored whether the PPC region could encode mismatch signals of stimulus intensity. By designing three sequences of interleaved experiments (complete omission of tactile stimuli, attenuation or enhancement of intensity), it was found that the PPC can report intensity mismatches in both directions: enhanced mismatch trials activate positively mismatched neurons, and attenuated trials activate negatively mismatched neurons, and the two types of neuronal populations are relatively independent. Notably, positive mismatch neurons overlap with response neurons in the early pairing phase and may constitute a subset thereof, whereas pairing response neurons remain active on augmented mismatch trials but are unresponsive to the attenuated type. This segregated neural coding pattern (enhancement/attenuation processed by separate populations) is consistent with the classical microloop model within the framework of predictive processing theory, suggesting that PPC dynamically integrates anticipatory and sensory input differences through specific neuronal populations to enable bidirectional mismatch detection. Thus, PPC can report positive and negative mismatches through different neuronal populations whose response dynamics can be matched to the typical microloops of predictive processing.

• M2 modulates perceptual processing in the anterior cingulate region

The authors further investigated the source of feedback that drives the pre-stimulus anticipatory response in the PPC area. The pathway was inhibited by retroviral expression of inhibitory DREADD (hM4Di) in PPC→M2 projection neurons and intraperitoneal injection of CNO. It was found that inhibition of the feedback input from M2 to PPC resulted in unaffected tactile stimulus responses in paired-response neurons and staggered-match neurons, but significantly reduced pre-stimulus anticipatory responses in staggered-match neurons (compared to CNO controls). This suggests that M2 cortical projections to the PPC specifically modulate anticipatory signaling (pre-stimulus activity) without affecting the processing of the sensory stimulus itself, suggesting that the M2 may be a key brain region mediating top-down transmission of anticipatory signals.

For staggered mismatch trials, the authors found that after inhibiting M2 feedback to the PPC, neurons were still able to produce mismatch responses to tentacle stimulus omissions (response strength comparable to controls), but the number of responding neurons was significantly reduced. Population analyses showed that M2 inhibition impaired both the strength of neuronal responses to predictive sounds and subsequent omission mismatch responses. Taken together, these findings suggest that M2 is able to modulate sensory processing in precortical areas, driving pre-stimulus responses/anticipation in different neuronal groups. In addition, it is possible that M2 contributes to the top-down prediction of reaching precortical areas that modulates responses to predictive sounds as well as missing mismatch responses.

Disclaimer: Please note that we do not provide the content above, nor do we hold copyright to it. This article is for informational and knowledge-sharing purposes only and does not constitute an offer of commercial services related to its subject matter.

Resources

Primary Cell Culture and Isolation Services

Human iPSC derived Cell Culture and Isolation Services

Reference

1. Raltschev, Constanze, et al. "Top-down modulation of sensory processing and mismatch in the mouse posterior parietal cortex." Nature Communications 16.1 (2025): 1-10. Distributed under Open Access license CC BY 4.0, without modification.