Optical Ministroke Reveals Dual-Role Frontal Thalamocortical Networks in Sleep Quality and Memory

Nonrapid eye movement sleep (NREMS) is a critical physiological state supporting neural plasticity, memory consolidation, and functional recovery after brain injury. It is characterized by thalamocortical slow-wave (SW) activity (0.5-4 Hz) and spindles (10-16 Hz) that synchronize cortical activity and regulate synaptic strength. Thalamic strokes, though rare, often disrupt sleep-wake cycle architecture, NREMS oscillations, and cognitive and sensory processing. Conventional rodent stroke models lack access to deep brain structures and require anesthesia, restricting investigation of thalamic circuit dysfunction after injury. To address these limitations, we developed an optically guided photothrombotic stroke model that enables focal, anesthesia-free lesions of the mediodorsal thalamus (MD) in freely behaving mice. We tested whether MD stroke impairs thalamocortical oscillations and cognition, and whether sleep-targeted auditory stimulation rescues these deficits.

This was a randomized, controlled, interventional study with a within-species design in male C57BL/6JRj mice enrolled at 10 to 16 weeks of age across 8 cohorts (n=8-12/group). Chronic electroencephalogram/electromyogram electrodes and optical fibers targeting the MD were implanted. An optically guided photothrombotic stroke model was induced with intraperitoneal Rose Bengal (10 mg/mL) followed by 532-nm light (10 mW, 6 min) in awake animals; sham controls received no light. Animals with poor electroencephalogram/electromyogram signals, artifacts, or incomplete testing were excluded. Primary outcomes were longitudinal (20 days) sleep-wake features and oscillations, working memory (Y-maze), and pain sensitivity. A subset received daily 1-Hz auditory stimulation (≈1 h/session for 10 days during NREMS-rich periods). Group differences were analyzed using 2-way ANOVA with the Bonferroni post hoc tests, unpaired t tests, and Pearson correlations (99% CIs).

Optically guided photothrombotic stroke model induced stable focal MD lesions that increased wake-NREMS-wake transitions, elevated SW activity during wakefulness, and persistently reduced frontal individual SWs and spindles during NREMS compared with shams (P=0.03-P<0.001). These alterations were accompanied by impaired working memory and pain hypersensitivity (P<0.001), recapitulating hallmark features of paramedian thalamic infarcts in humans. Notably, auditory stimulation normalized sleep continuity, restored SW-spindle coupling, and working memory to sham levels (working memory errors were negatively correlated with spindle rate [r=-0.88] and SW-spindle coupling [r=-0.81]) and rescued impaired MD-anterior cingulate cortex connectivity to parvalbumin-positive interneurons.

Focal MD lesions disrupt sleep-wake stability, thalamocortical oscillations, and working memory, whereas noninvasive auditory stimulation restores sleep dynamics, cognitive performance, and MD-anterior cingulate cortex synaptic connectivity. Together, our findings establish the optically guided photothrombotic stroke model as a versatile model for dissecting stroke recovery mechanisms and highlight noninvasive stimulations as a promising approach to restore sleep and cognitive function.