Asymmetric Modulation of Auditory-Responsive Subpopulations in the Mediodorsal Thalamus by the Prefrontal Cortex

Technical Summary: Asymmetric Modulation of Auditory-Responsive Subpopulations in the Mediodorsal Thalamus by the Prefrontal Cortex

Problem Statement
While the bidirectional circuit between the prefrontal cortex (PFC) and the mediodorsal thalamus (MD) is well-established as critical for cognitive control and executive function, the specific mechanisms by which the PFC modulates sensory processing within the MD remain unclear. Specifically, it is unknown whether the MD processes auditory information through internal functional subpopulations that are subject to selective, spatially distinct cortical modulation, or if the MD acts as a homogeneous relay. Furthermore, the extent to which top-down PFC inputs regulate the timing versus the accuracy of sensory-driven behaviors requires elucidation.

Methodology
The study employed an in vivo experimental paradigm combining high-density electrophysiology, optogenetics, and behavioral testing in mice.

• Subjects: VGAT-ChR2-EYFP transgenic mice (for optogenetic manipulation of GABAergic interneurons in the PFC) and C57BL/6 wild-type mice (as behavioral controls).
• Electrophysiology: Neuropixels 1.0 probes were implanted in the MD to record single-unit activity from 1,188 neurons across three mice. Depth alignment was achieved using theta-band power peaks at the hippocampal fissure as a reference.
• Optogenetics: Local inhibition of the PFC was achieved via blue light (473 nm) stimulation of ChR2-expressing GABAergic interneurons, inducing feedforward inhibition of excitatory projection neurons.
• Stimuli: Rhythmic auditory stimuli (4, 16, and 40 Hz pulse sequences) were presented. Optogenetic inhibition (40 Hz light pulses) was applied 500 ms prior to sound onset to assess top-down modulation.
• Behavioral Task: A head-fixed Go/No-go auditory discrimination task was used, where 40 Hz sounds were targets (Go) and 4 Hz sounds were non-targets (No-go). Performance metrics included hit rate, false alarm rate, and lick response latency.
• Analysis: Neuronal responses were analyzed using Principal Component Analysis (PCA) for clustering, spatial mapping to the Allen Mouse Brain Common Coordinate Framework (CCF), and decoding analysis (SVM) to assess frequency discrimination capabilities.

Key Contributions and Results

1. Identification of Two Functionally Distinct MD Subpopulations:
   PCA clustering of auditory-evoked responses revealed two spatially segregated subpopulations within the MD:

   • Cluster 1 (Anterior-Central): Characterized by long-latency, sustained excitation, poor rhythm tracking, and weak frequency discrimination (decoding accuracy ~60%).
   • Cluster 2 (Posterior): Characterized by short-latency, phase-locked firing to low frequencies, transient responses to high frequencies, and superior frequency discrimination (decoding accuracy ~80.8%).

2. Spatial Correspondence with PFC Projections:
   Spatial mapping demonstrated that Cluster 1 is enriched in the anterior-central MD, a region that significantly overlaps with the high-density terminal field of descending PFC projections. In contrast, Cluster 2 is biased toward the posterior MD, an area with lower PFC projection density.

3. Asymmetric Modulation by PFC Inhibition:
   Local optogenetic inhibition of the PFC exerted differential effects on the two subpopulations:

   • Cluster 1: Experienced profound, sustained firing suppression and a reduction in cumulative excitatory response (+AUC). Interestingly, under 4 Hz stimulation, their firing pattern shifted toward a more phase-locked response.
   • Cluster 2: Showed only brief initial inhibition followed by rapid recovery or rebound excitation. Crucially, their primary firing patterns, peak amplitudes, and frequency coding capabilities remained largely preserved despite PFC inhibition.

4. Behavioral Impact: Latency vs. Accuracy:
   In the Go/No-go task, PFC inhibition significantly prolonged the latency of the lick response to the target sound (from ~595 ms to ~935 ms) in experimental mice, while control mice showed no change. However, the inhibition did not impair discrimination accuracy; hit rates and false alarm rates remained stable and comparable between light-on and light-off conditions.

Significance and Claims
The paper claims to reveal a subpopulation-selective regulatory mechanism within the PFC-MD pathway. The authors conclude that the PFC does not uniformly modulate the entire MD but rather preferentially targets the anterior, non-discriminative subpopulation (Cluster 1). This selective modulation appears to regulate the timing of sensory-driven behaviors (response preparation or motor gating) rather than the accuracy of sensory discrimination. The preservation of frequency coding in the posterior subpopulation (Cluster 2) suggests that the PFC-MD pathway may function to modulate the speed of behavioral output triggered by sound detection, while leaving the fundamental sensory classification to other pathways or distinct thalamic subpopulations.

The study emphasizes that the MD is not a homogeneous structure but possesses an internal functional topology that allows for distinct cognitive control over different aspects of auditory processing. The authors note limitations, including the lack of molecular characterization of the clusters and the potential for non-specific inhibition of parallel PFC pathways, suggesting that future work should employ pathway-specific interventions to fully delineate causal mechanisms.