TRNPV-Dependent Failure of Thalamocortical Timing Delays the Return of Conscious Content during Early Emergence from Propofol Anaesthesia

TRNPV-Dependent Failure of Thalamocortical Timing Delays the Return of Conscious Content during Early Emergence from Propofol Anaesthesia | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article TRN PV -Dependent Failure of Thalamocortical Timing Delays the Return of Conscious Content during Early Emergence from Propofol Anaesthesia Yu Zhang, Hui Liu, Shiyu Chen, Ying You, Yuxuan Chen, Meimei Zhao, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8937051/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 10 You are reading this latest preprint version Abstract Backround: Recovery from general anaesthesia is usually defined by return of behavioural responsiveness, but the capacity to encode new information may recover later. We hypothesised that early emergence from propofol reflects incomplete recovery of TRN gated thalamocortical timing needed for learning. Methods: PV-Cre mice received intravenous propofol until cortical EEG reached burst suppression, then infusion was stopped for spontaneous emergence. LORR and RORR indexed behavioural state. Mice were trained after RORR in inhibitory avoidance and auditory fear conditioning, with memory tested 24 h later. Thalamocortical dynamics were measured with electrophysiology using a single spindle-detection pipeline across NREM sleep, propofol anaesthesia, and post-emergence NREM. TRN PV activity was recorded with fibre photometry, suppressed chemogenetically, and examined ex vivo in the prefrontal, TRN, and thalamic circuit. Results: Training 5 to 15 min after RORR impaired learning despite restored arousal. Spindle-band dynamics remained abnormal early after emergence. TRN neurons were hyper-recruited around spindles, showed more burst-like firing, and exhibited altered spike to spindle phase coupling. Photometry showed elevated baseline but blunted stimulus- and cue-evoked TRN PV responses at 5 and 15 min, with partial recovery by 30 min. Chemogenetic TRN PV suppression worsened learning deficits. Ex vivo recordings indicated increased TRN PV intrinsic excitability, strengthened inhibition onto mediodorsal thalamus, and increased gain to prefrontal inputs without increased prefrontal excitatory drive, consistent with impaired top-down timing. Conclusion: Early emergence is a distinct vulnerable state in which learning lags behind arousal due to persistent TRN PV dependent disruption of spindle-associated thalamocortical timing. Biological sciences/Neuroscience/Cognitive neuroscience Biological sciences/Neuroscience/Cellular neuroscience General anaesthesia Early emergence Thalamic reticular nucleus Thalamocortical timing Conscious content Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction General anaesthesia is typically evaluated—and often operationally defined—by the loss and recovery of behavioral responsiveness. However, the return of overt arousal does not necessarily imply restoration of the neural computations that support conscious experience and flexible cognition [ 1 ] . Clinically, patients frequently display a transient period of confusion, amnesia or impaired learning immediately after emergence, despite appearing awake and interactive [ 2 ] . This dissociation during emergence from general anaesthesia suggests that the “content” of consciousness—operationally defined here as the brain’s capacity to acquire, integrate and store new information—may recover on a slower timescale than arousal “state” [ 3 , 4 ] . Identifying the circuit-level bottlenecks that delay the recovery of content-related processing after anesthesia remains challenging. A major obstacle is that global behavioral metrics provide limited insight into whether post-emergence brain dynamics have returned to physiological modes of thalamocortical communication [ 5 ] . Thalamocortical timing is a compelling candidate mechanism because it constrains effective information transfer, sensory gating and the coordination of distributed cortical ensembles [ 3 , 6 , 7 ] . In natural sleep, these functions are closely tied to spindle-associated dynamics—brief, rhythmic events that reflect structured interactions between cortex, thalamus and the thalamic reticular nucleus (TRN) [ 8 , 9 ] . Spindles can support memory-related processing in sleep, yet spindle activity can also accompany reduced environmental responsiveness [ 10 ] , indicating that spindle activity may correspond to distinct computational states depending on their timing, coupling and circuit implementation. The TRN occupies a strategic position to regulate these state-dependent thalamocortical dynamics [ 11 , 12 ] . As a GABAergic shell around the dorsal thalamus, the TRN gates thalamic relay activity and shapes thalamocortical synchrony, placing it in a key position to regulate both the global state of consciousness (arousal/behavioral responsiveness) and the content-related computations that support perception, integration and memory [ 3 ] . Parvalbumin-positive (PV) TRN neurons, in particular, are well suited to impose temporally precise inhibition and thereby influence both the transitions into and out of general-anaesthetic unconsciousness and the fine structure of spindle-associated rhythms that are thought to support information processing. [ 13 , 14 ] . Disrupted TRN timing control may leave a post-emergence state that is behaviorally aroused yet less capable of encoding new information, consistent with content-related processing lagging behind arousal. Propofol provides a powerful experimental framework to test this idea. Beyond producing reversible unconsciousness, propofol strongly reorganizes thalamocortical activity and promotes rhythmic patterns that can resemble spindle-range (alpha range) oscillations [ 12 ] . Our prior work [ 15 ] shows that spindle-range oscillations do not immediately return to physiological sleep-like dynamics after propofol emergence. What remains unresolved is whether these altered rhythms causally interfere with information acquisition during early recovery, and whether PV-defined TRN circuitry is a principal substrate for the persistent disruption of thalamocortical timing. Accordingly, we focus on the minutes-to-tens-of-minutes period immediately after emergence from general anaesthesia, a transient state distinct from postoperative delirium [ 16 ] or longer-lasting perioperative neurocognitive disorders [ 17 ] . We combined behavior, systems electrophysiology, cell-type-specific calcium photometry, chemogenetic perturbation, and ex vivo circuit physiology to test the hypothesis that recovery of behavioral arousal from propofol anaesthesia precedes recovery of thalamocortical timing dynamics that support learning. We find that learning is impaired early after emergence despite restored gross motor performance, coincident with persistent spindle-like thalamocortical timing abnormalities and disrupted TRN PV dynamics. Causal and circuit-level analyses further implicate TRN PV gated top-down corticothalamic pathway as a substrate for delayed recovery of learning-relevant processing after anesthetic emergence. Methods 1. Animals and ethics All experiments were performed in accordance with institutional guidelines for the care and use of laboratory animals and complied with relevant animal welfare regulations. Experimental protocols were reviewed and approved by the Experimental Animal Ethics Committee of Zunyi Medical University (Appl. No. zyfy-an-2023-0284). Healthy adult male specific-pathogen-free (SPF) mice were used. Wild-type C57BL/6J mice (10–12 weeks old; 20–25 g) and PV-IRES-Cre mice (10–12 weeks old; 20–25 g) were housed under controlled environmental conditions (25 ± 2 °C; 50 ± 5% humidity; noise < 50 dB) on a 12 h light/12 h dark cycle (lights on 08:00–20:00; lights off 20:00–08:00), with ad libitum access to food and water and free movement in their home cages. 2. Anaesthesia and peri-anaesthetic procedure 2.1 Experimental groups and overall workflow Healthy adult SPF C57BL/6J mice (male, 10–12 weeks old, 20–25 g) were randomly assigned to one of four experimental conditions: Control (no anaesthesia), and post-emergence groups trained at 5 min, 15 min, or 30 min after recovery of righting reflex (RORR) following propofol anaesthesia. The core experimental workflow was: handling habituation → tail-vein propofol infusion under EEG monitoring → termination of infusion at burst suppression (BS) → monitoring until RORR → behavioral testing at the assigned post-emergence time point. 2.2 Pre-experimental habituation and stress minimization To minimize stress-related confounds in behavioral performance, mice were habituated to experimenter handling by gentle daily holding and stroking for 3 consecutive days prior to experiments. Animals were considered habituated when they could remain calmly in the experimenter’s hands without struggling or attempting to escape, while maintaining spontaneous movement. 2.3 Propofol administration and BS endpoint Propofol was administered by continuous tail-vein infusion using a syringe pump at a constant rate of 10 mg kg/min. Anaesthetic depth was titrated to an electrophysiological endpoint: burst suppression (BS) on EEG. Once BS was detected, propofol infusion was immediately discontinued. BS was used as an operational marker of deep anaesthesia. Based on our experimental premise, a brief period of BS per se does not measurably impair learning; therefore, the post-anaesthetic behavioral phenotype was interrogated as a function of time after RORR. 2.4 Emergence monitoring and definition of RORR After cessation of propofol infusion, animals were continuously monitored until recovery of righting reflex (RORR), defined as the ability to self-right from a supine to a prone position. RORR time served as time zero for post-emergence grouping. Mice were then tested at the designated interval after RORR (+5 min, +15 min, or +30 min) for behavioral assays. 3. Behaviour: learning/memory assays and motor control All behavioral experiments were conducted during the light phase. Mice were transferred to the testing room and allowed to acclimate for ≥30 min before each assay. Apparatuses were cleaned with 75% ethanol between animals to minimize olfactory cues. 3.1 Behavioral timeline and task order To reduce potential carry-over effects across paradigms, the same cohort of mice completed behavioral testing in the following order: motor assessment first, then contextual and auditory-cued fear conditioning after a 7-day interval, followed by inhibitory avoidance after an additional 7-day interval. 3.2 Rotarod test (motor coordination) The rotarod test was used to assess motor coordination and balance, quantified as latency to fall. Habituation/training. Mice underwent two adaptation sessions 24 h before testing. During each session, rod speed was ramped from 4 to 10 rpm over 30 s, and mice were allowed to remain on the rod for 10 min. Two sessions were conducted with an inter-session interval of ≥1 h. If a mouse fell during training, it was promptly returned to the rod to continue training until the session ended. Testing. Depending on group assignment, mice were tested at 5 min, 15 min, or 30 min after recovery of righting reflex (RORR), or in a non-anaesthetized control condition. During testing, rod speed was ramped from 4 to 40 rpm over 5 min, and latency to fall was recorded. 3.3 Open field test (locomotor activity) The open field test was used to assess general locomotor activity after anaesthesia. Each mouse was placed in the centre of the arena and allowed to freely explore for 5 min while its trajectory was recorded. The primary outcome measure was total distance travelled during the session. 3.4 Contextual and auditory-cued fear conditioning (learning and memory) Contextual and auditory-cued fear conditioning were used to quantify associative learning by pairing an auditory conditioned stimulus (CS) with an aversive unconditioned stimulus (US; footshock). Freezing was used as the behavioral readout of learned fear and was defined as the absence of movement other than respiration. Training (Day 1). Mice were placed in the conditioning chamber and allowed to explore for 2 min, followed by three CS–US pairings within a total training duration of 5 min. Each pairing consisted of an auditory CS (2,000 Hz, 80 dB, 28 s) immediately followed by a footshock US (0.6 mA, 2 s), with a 30 s interval after each shock before the next pairing. Testing (Day 2): (i) Contextual test: Mice were returned to the same chamber for 3 min with no CS and no footshock. (ii) Cued test: Mice were placed in a novel context and presented with the auditory CS (2,000 Hz, 80 dB, 28 s) without footshock for three cycles during a 3-min session. Behaviour quantification. Freezing was automatically scored using Labmaze V3.0 (Beijing ZS Dichuang Technology Development Co., Ltd.). The freezing detection threshold was set to 4%. The primary outcome was percentage freezing (time spent freezing/total observation time × 100%). 3.5 Inhibitory avoidance (IA) Inhibitory avoidance was used to assess aversive associative learning based on rodents’ preference for dark environments paired with a footshock. The apparatus consisted of connected light and dark compartments; the dark compartment contained a grid floor for shock delivery. Training. Mice were placed in the dark compartment, and entry was paired with a footshock (0.6 mA, 2 s). During the 5-min training session, each subsequent entry into the dark compartment was paired with the same footshock. Testing. Mice were placed in the light compartment, and the latency to first enter the dark compartment was recorded as the memory index (maximum 300 s; mice not entering within 300 s were assigned a latency of 300 s). Longer latency indicates stronger avoidance learning/memory. 4. EEG Implantation, Acquisition, and Pre-processing EEG/EMG electrodes were purchased from Kedou (Suzhou) Brain-Computer Technology Co., Ltd. (KD-EEG/EMG-AG) and implanted using standard aseptic stereotaxic procedures. Briefly, animals were anesthetized and secured in a stereotaxic apparatus, the scalp was incised to expose the skull, and the electrode assembly was positioned according to the experimental design and fixed with dental cement; animals were allowed to fully recover before recording. Continuous EEG signals were then acquired using the Apollo portable neural signal data acquisition system (Yige Biotechnology (Nanjing) Co., Ltd.) at a sampling rate of 1000 Hz. Raw data were exported for offline pre-processing, during which recordings were visually inspected and segments with gross movement, poor electrode contact, or other artifacts were excluded; where applicable, signals were re-referenced and baseline-corrected prior to subsequent analyses (e.g., spectral or event-related quantification). 5. Sleep scoring, spindle detection and spectral analysis Sleep–wake state was scored prior to spindle quantification using the Lunion Stage automatic sleep analysis system (https://stage.luniondata.com/). Continuous EEG and accompanying EMG channels were segmented and automatically classified into vigilance states; only epochs scored as NREM sleep were included for spindle analyses. For the post-emergence window (10–20 min after recovery of righting reflex, RORR), animals were left undisturbed and frequently re-entered NREM-like sleep/drowsiness, consistent with strong post-propofol sleep pressure; spindles were therefore quantified exclusively from epochs meeting the same NREM criteria used for baseline sleep, allowing direct comparison across conditions (baseline NREM, propofol anaesthesia, and post-emergence NREM at RORR 10-20 min). Spindle events were detected from mouse cortical EEG recordings acquired with the Apollo portable neural signal data acquisition system (Yige Biotechnology (Nanjing) Co., Ltd.) using a batch-detection strategy adapted from prior work on oscillation detection and plasticity mechanisms [18, 19] . All offline signal processing, filtering, event identification, feature measurement and statistical summaries were implemented with custom scripts in MATLAB (R2016b, MathWorks). For spindle detection, continuous EEG was band-pass filtered in the sigma range (7–15 Hz), full-wave rectified, and the amplitude envelope was extracted to capture time-varying changes in oscillatory amplitude. A spindle was defined when the envelope exceeded a preset threshold of 1.5× the standard deviation (computed from the corresponding recording segment) and remained above threshold for 0.5–3.0 s. Event onset and offset were determined using threshold-crossing rules, with onset marked by the first positive threshold crossing and termination marked by the last negative threshold crossing. For each detected spindle, the root mean square (RMS) of the band-limited signal was computed, and the mean RMS across events was used as the primary amplitude metric for subsequent analyses. Spectral quantification was performed in MATLAB on the same preprocessed EEG signals, focusing on sigma-band (7–15 Hz) activity. 6. In vivo TRN spiking recordings, Spike–spindle phase coupling and phase histograms Chronic in vivo extracellular recordings were performed in freely moving mice using a nickel–titanium (NiTi) microwire electrode array (Kedou (Suzhou) Brain-Computer Technology Co., Ltd.) implanted stereotaxically to target the thalamic reticular nucleus (TRN; AP = −0.53 mm, ML = ±1.35 mm, DV = −3.65 mm from the skull surface). Electrodes were fixed to the skull with anchor screws and dental acrylic. Neural signals were acquired with the Apollo portable neural signal data acquisition system (Yige Biotechnology (Nanjing) Co., Ltd.) at 3,000 Hz and band-pass filtered to extract a spike band (300–3,000 Hz) and an LFP band (0.5–300 Hz). Spikes were detected using a negative threshold of −4.5× the root mean square (RMS) noise level and sorted offline using Offline Sorter (Plexon) with principal-component features and manual refinement; putative single units were accepted only when the inter-spike-interval (ISI) violation rate was <0.5–1%. Bursts were defined as sequences of spikes with consecutive ISIs <10 ms and containing ≥3 spikes; the average intraburst firing rate (aIBF) was calculated as the mean of 1/ISI across within-burst intervals and then averaged across bursts for each unit and condition. For peri-spindle analyses, we analysed three conditions: baseline NREM sleep, propofol anaesthesia, and post-emergence NREM sleep (RORR 10-20 min). Each condition included 6 mice. To balance contributions across animals, we randomly sampled 50 spindles per mouse (300 total spindles per condition) for spike-alignment and phase-coupling analyses. Spikes were time-locked to spindle events and peri-event time histograms (PETHs) were computed using 10 ms bins. In addition to spindle-onset alignment, we also generated spindle-triggered averages by band-pass filtering thalamic signals in the spindle band (7–15 Hz) and aligning events to the peak of the central spindle cycle (defined as the maximal-amplitude cycle near the spindle midpoint; t=0). Raster plots depict single-unit activity with each row corresponding to one spindle event, and trial-averaged firing rate was estimated with a 10 ms sliding window with 80% overlap [10] . Spike–spindle phase coupling was quantified by band-pass filtering EEG/LFP in the spindle band (7–15 Hz) and extracting the instantaneous phase from the analytic signal (Hilbert transform). For each detected spindle epoch, to reduce potential bias in phase estimates introduced by spindle onset/offset detection, the onset and offset times were pseudorandomly jittered by ±0.5 spindle cycles (uniform in phase) independently for each spindle, and spikes were included only if they occurred within the jittered epoch boundaries. Each spike was assigned the corresponding LFP phase angle, and phase non-uniformity was assessed using the Rayleigh test. Multiple comparisons across units were controlled using the Holm–Bonferroni correction. For each unit we computed the preferred phase (circular mean) and phase concentration/phase-locking value (mean resultant length, PLV). Spike–spindle phase histograms were generated for strongly phase-locked units (Rayleigh test, corrected) by binning spike phases over 0–2π and plotting normalized counts in polar coordinates; the black vector indicates each unit’s mean resultant vector (direction = preferred phase, length = PLV) [19] . Group summaries were obtained by plotting the population mean preferred phase and mean vector length for each cell class/condition. 7. Fiber Photometry Calcium Imaging Under aseptic conditions, PV-IRES-Cre mice were anesthetized with 1.4% isoflurane, shaved, and placed prone in a stereotaxic frame; erythromycin ophthalmic ointment was applied, the scalp was disinfected, and 1% lidocaine was administered subcutaneously. A ~1 cm midline incision was made to expose the skull, hydrogen peroxide was used to remove periosteum, and the head was leveled to within ±0.03 mm (AP/ML). The thalamic reticular nucleus (TRN) injection site was targeted using the following coordinates relative to bregma: AP = −0.53 mm, ML = ±1.35 mm, DV = −3.65 mm; a craniotomy was drilled and the dura was carefully removed. A Cre-dependent calcium indicator virus (rAAV-EF1α-DIO-jGCaMP7b) was loaded into a 1 μL syringe (220 nL; 5 min wait) and infused into the TRN at 40 nL/min for 180 nL, with the needle left in place for 10 min to minimize backflow. An optical fiber was then lowered to DV = 3.65 mm and secured using 454 cyanoacrylate adhesive, supplemented with two skull screws and self-curing dental acrylic; mice received intramuscular penicillin for 3 consecutive days and were recorded ~3 weeks later to allow viral expression. For fiber photometry, mice were handled for 3 days and habituated to the room for ≥30 min before testing; recordings were performed under light-restricted conditions using a Multi-Channel Fiber Photometry Device (410/470; Inper, Hangzhou, China) connected via a ceramic ferrule, with excitation delivered to the TRN and emitted fluorescence collected and converted to an analog voltage signal. A 10 s stimulus-free baseline was used to estimate F0, and fluorescence changes were expressed as ΔF/F = (F − F0)/F0. 8. Chemogenetics Adult male SPF PV-IRES-Cre mice (10–12 weeks old, 20–25 g; total n = 35) received bilateral TRN injections of Cre-dependent DREADD viruses and were assigned to control (AAV-DIO-mCherry, n = 12), inhibition (AAV-DIO-hM4D(Gi)-mCherry, n = 10) or activation (AAV-DIO-hM3D(Gq)-mCherry, n = 13) groups; after surgery mice recovered on a warming pad and received intramuscular penicillin once daily for 3 consecutive days, and experiments were conducted 3 weeks later to allow stable expression, with additional gentle handling for 3 days before testing to minimize stress. For chemogenetic manipulation, clozapine-N-oxide (CNO; 1 mg/kg, i.p.) or an equal volume of saline was administered 30 min before subsequent procedures in a within-subject, counterbalanced design (saline vs CNO sessions separated by a 7-day washout). After pretreatment, mice underwent the propofol anaesthesia paradigm (see above) and were trained 15 min after emergence, with memory tested 24 h later (contextual and auditory-cued fear conditioning and inhibitory avoidance); for EEG experiments, EEG connectors were attached after pretreatment and mice received tail-vein propofol infusion (10 mg/kg/min) until EEG burst suppression, followed by continuous EEG recording for 35 min after recovery of righting reflex. At the end of experiments, brains were collected for immunofluorescence to verify TRN viral expression and its colocalization with PV-positive neurons by fluorescence microscopy. 9. Acute slice preparation and patch-clamp electrophysiology Acute thalamic slices were prepared from PV-tdTomato and SOM-tdTomato mice (4–6 weeks old). Mice were deeply anaesthetized with isoflurane and decapitated, and brains were rapidly removed into ice-cold (4 °C), carbogenated (95% O 2 /5% CO 2 ) sucrose-based cutting solution containing (in mM): 234 sucrose, 3 KCl, 1.25 NaH 2 PO 4 , 10 MgSO 4 ·7H 2 O, 0.5 CaCl 2 , 2H 2 O, 26 NaHCO 3 , 10 D-glucose. Coronal slices containing the TRN were cut at 240 μm using a vibrating microtome (HM 650V) in continuously carbogenated cutting solution. Slices were then recovered at 32 °C for 20 min in carbogenated sucrose-based solution and subsequently held for ≥60 min at room temperature in carbogenated recording ACSF (NaCl-based) before experiments. Whole-cell patch-clamp recordings were performed at room temperature under infrared DIC optics. TRN PV neurons were identified by tdTomato fluorescence, and MD neurons were targeted based on anatomical location. Pipettes (4–8 MΩ) were pulled from borosilicate glass. For intrinsic excitability measurements (current clamp), a K + -based internal solution was used (composition as specified in the internal-solution section). For optogenetically evoked synaptic currents, a Cs + -based internal solution was used. To record optogenetically evoked inhibitory postsynaptic currents (eIPSCs), cells were voltage-clamped at 0 mV to isolate GABA A receptor–mediated currents. For TRN PV →MD connectivity experiments, TRN PV neurons expressed ChR2, and TRN PV axon terminals in the MD were stimulated using a 470-nm LED light source (X-Cite 110LED; Excelitas Technologies, Japan) coupled to an Olympus microscope with 10 ms pulses delivered at 0.5 Hz, evoking light-evoked IPSCs in MD neurons. To assess mPFC→TRN excitatory input, ChR2 was expressed in mPFC projection neurons and their axon terminals in the TRN were stimulated with the same 470-nm LED protocol; TRN neurons were voltage-clamped at −70 mV to record light-evoked excitatory postsynaptic currents (eEPSCs). Propofol was diluted in recording ACSF to a final concentration of 10 μM and bath-applied via perfusion at 3–4 ml/h. Data were excluded if series resistance exceeded 25 MΩ or changed by >20% during the recording. 10. Histology and verification of targeting At the completion of experiments, mice were anaesthetized with 1.4% isoflurane and transcardially perfused with PBS followed by 4% paraformaldehyde (PFA). Brains were removed, post-fixed in 4% PFA for 12 h, cryoprotected in 30% sucrose until sunk, embedded in OCT, and sectioned coronally at 30 μm on a −20 °C cryostat (brains were pre-cooled for ~30 min before cutting). Free-floating sections were stored in PBS and processed for PV immunofluorescence: sections were washed in PBS (3 × 10 min), permeabilized/blocked (2 h, room temperature; PBS containing 10% goat serum, 1% BSA and 0.1% Triton X-100), incubated with rabbit anti-parvalbumin (29312-1-AP, 1:500; overnight, room temperature), washed, and then incubated with goat polyclonal anti-rabbit IgG secondary antibodies (ab150077 or ab150080, 1:1,000; 2 h, room temperature, protected from light). Sections were washed, mounted, and coverslipped with a DAPI-containing anti-fade medium. Viral expression in the TRN and its co-localization with PV-positive neurons, as well as fibre/electrode targeting where applicable, were examined using a BC43 confocal microscope (Oxford Instruments, UK), and animals with mistargeting or insufficient/misplaced expression were excluded. 11. Statistical analysis Statistical analyses were performed using MATLAB (R2016b), GraphPad Prism and Clampfit. All tests were two-sided unless stated otherwise, with P < 0.05 considered statistically significant; multiple-comparison procedures are indicated in the figure legends. Data are presented as mean ± SD. For analyses of EEG spindles and unit activity, the primary observational unit was the detected event (that is, each spindle or spike-derived event, as specified); events were quantified within predefined 10-min epochs for physiological NREM sleep (10 min), propofol anaesthesia (10 min) and early after emergence (10–20 min after RORR). Because events are nested within animals, event-level statistics were performed using mixed-effects models with condition/epoch as a fixed effect and mouse as a random effect (random intercept), accounting for within-mouse dependence and unequal event counts across mice/conditions, with Dunnett-adjusted post hoc comparisons to the indicated reference condition when applicable. For spindle-phase coupling, instantaneous phase was extracted from band-pass-filtered signals using the Hilbert transform; phase non-uniformity was assessed with the Rayleigh test and multiple comparisons were controlled using the Holm–Bonferroni method. Exact P values and the numbers of mice and total events are reported in the figure legends. Results Arousal recovers before learning after propofol anaesthesia To test whether behavioral emergence from propofol anaesthesia coincides with recovery of learning capacity, we infused propofol intravenously until the cortical EEG reached burst suppression and then stopped the infusion. Loss and recovery of the righting reflex (LORR and RORR) were used as behavioral markers for entry into and exit from anaesthesia (Fig. 1A). To isolate effects on memory formation, mice were trained at defined times after RORR (5, 15, or 30 min) and tested 24 h later. In the inhibitory avoidance task, training soon after emergence disrupted later memory expression. Relative to controls, mice trained at RORR +5 min showed shorter step-through latencies in the 24-h test and spent more time in the dark compartment (Fig. 1B,C), indicating weaker avoidance memory. Performance improved as the training-RORR interval increased, and mice trained at RORR +30 min were indistinguishable from controls (Fig. 1B,C). These data identify a transient post-emergence interval in which learning is impaired despite restoration of righting and exploratory behaviour. An independent associative assay yielded the same temporal profile. Mice trained in auditory fear conditioning shortly after RORR displayed reduced freezing at 24 h in both contextual and cued tests, whereas freezing recovered when training was delayed (Fig. 1E–G). Thus, memory acquisition lags behind behavioral emergence from propofol. Residual motor suppression is unlikely to account for this deficit. In the same cohort, open-field activity was reduced at RORR +5 min but returned to control levels by 15 min (Fig. 1D,H), and rotarod performance likewise recovered by 15 min (Fig. 1I). The learning impairment evident when training occurs within the RORR +15 min window therefore cannot be explained by gross locomotor or coordination deficits. Early emergence cortical spindles remain altered during the learning-impaired window We next asked whether this early post-anaesthetic learning deficit (centred at RORR +15 min; Fig. 1) is accompanied by persistent changes in thalamocortical oscillations. Because spindle generation depends on thalamic circuitry, cortical spindles were quantified using a single detection pipeline applied identically across three conditions: physiological NREM sleep, ongoing propofol anaesthesia, and the early post-emergence window (RORR 10-20 min; Fig. 2A). Relative to physiological NREM sleep, propofol markedly changed spindle amplitude and occurrence (Fig. 2B,D), while spindle duration was less affected (Fig. 2C). During post-emergence NREM (RORR 10-20 min), spindle amplitude and event rate did not return to the NREM pattern (Fig. 2B,D), indicating that the mechanisms shaping spindles remain perturbed during early recovery even when the animal is in NREM sleep. Time–frequency maps and power spectra supported a condition-dependent redistribution of spindle-band structure (Fig. 2E–H). Compared with NREM spindles, events detected during propofol and post-emergence NREM showed shifts in sigma-band organisation, with corresponding changes in centre frequency, spindle-band power, and bandwidth (Fig. 2I–K). Spindle dynamics therefore remain state-dependently altered during the same post-emergence interval in which learning is compromised. TRN neurons show persistent firing and spindle-phase dysregulation at RORR 10-20 min To determine whether the spindle phenotype reflects persistent dysfunction at a candidate generator node, we recorded TRN activity in vivo. Recording locations were confirmed histologically (Fig. 3A), and well-isolated TRN units were identified by waveform properties and spike sorting (Fig. 3B,C). TRN spiking was then aligned to detected spindles and compared across physiological NREM sleep, propofol anaesthesia, and early post-emergence (RORR 10-20 min; Fig. 3D–F). During physiological NREM, TRN firing showed stereotyped modulation time-locked to spindle onset (Fig. 3D). Under propofol, TRN neurons exhibited stronger spindle-associated recruitment, with denser rasters and a larger peri-event increase in firing rate (Fig. 3E). Elevated recruitment persisted into the RORR 10-20 min window (Fig. 3F), matching the time window of behavioral learning impairment. At the level of spike timing, inter-spike interval distributions shifted under propofol and remained shifted after emergence (Fig. 3G), consistent with increased high-frequency discharge and enhanced burst-like firing relative to NREM sleep. Mouse-aggregated measures confirmed higher mean firing rates during propofol with incomplete normalization during early recovery (Fig. 3J). Burst propensity and intraburst structure were similarly elevated—including burst fraction, spikes per burst, and average intraburst firing rate (aIBF) (Fig. 3K–M). Because spindle organization depends on phase-specific timing, we further quantified TRN spike timing relative to instantaneous spindle phase derived from TRN LFP (Gardner et al., 2013). Both phase preference and phase-locking strength differed across conditions (Fig. 3H,I), and altered coupling was still detectable at RORR 10-20 min. TRN output thus remains temporally reorganized after behavioral emergence, providing a plausible substrate for the persistent changes in spindle structure observed during early recovery. TRN PV population dynamics are altered during post-anaesthetic training and recall To probe cell-type-specific dynamics, we recorded fibre-photometry calcium signals from TRN PV neurons (Fig. 4A). TRN PV activity increased during propofol and remained elevated after RORR relative to the pre-anaesthetic baseline (Fig. 4B). We then measured the responsiveness of TRN PV neurons to controlled input during recovery. Somatosensory electrical stimulation evoked robust calcium transients in controls, whereas evoked responses were strongly attenuated at RORR +5 and +15 min and partially recovered by RORR +30 min (Fig. 4C–G). Heat maps across mice showed consistent suppression of stimulus-locked responses early after emergence (Fig. 4E), and quantification within the 0–5 s post-stimulus window confirmed reduced response magnitude at these time points (Fig. 4F,G). This pattern indicates elevated ongoing activity coupled to reduced dynamic responsiveness during early recovery. Finally, we recorded TRN PV activity during learning performed at RORR 10-20 min and during recall 24 h later (Fig. 4H). Relative to controls, mice trained during this early recovery window showed altered TRN PV activity during training (Fig. 4I–K; 0–5 s window) and during the subsequent recall test (Fig. 4L; 5–10 s window). Early post-anaesthetic learning is therefore associated with persistent deviations in TRN PV population dynamics across both acquisition and later retrieval. Chemogenetic suppression of TRN PV neurons exacerbates learning deficits during early emergence To test whether TRN PV neurons contribute to recovery of learning during early emergence, we used a within-subject chemogenetic design in PV-Cre mice expressing inhibitory (hM4D(Gi)) or excitatory (hM3D(Gq)) DREADDs in TRN (Fig. 5A). Mice received CNO or saline before propofol, were trained at RORR 10-20 min, and were tested 24 h later in fear conditioning and inhibitory avoidance (Fig. 5A). In TRN PV ::hM4D(Gi) mice, CNO altered post-emergence spindle expression (Fig. 5B–E) and worsened behavioral performance: freezing decreased in contextual and cued tests, and passive-avoidance latency was reduced compared with saline sessions (Fig. 5F–H). Reducing TRN PV activity during the peri-emergence period therefore aggravates the learning deficit. In contrast, TRN PV ::hM3D(Gq) activation did not improve behavioral outcomes. Although CNO produced measurable changes in spindle features (Fig. 5I–L), fear memory and passive avoidance did not differ between CNO and saline sessions (Fig. 5M–O). These results dissociate necessity from sufficiency: intact TRN PV activity appears required to prevent further deterioration of learning during early recovery, whereas globally increasing TRN PV activity alone is not sufficient to restore memory formation at this time point. Propofol directly alters intrinsic membrane properties of TRN PV neurons ex vivo To test whether propofol can directly reconfigure TRN PV excitability in a manner that could support persistent in vivo firing changes, we performed whole-cell current-clamp recordings from identified TRN PV neurons in acute thalamic slices (Fig. 6A,B). Bath application of propofol (10 μM) produced heterogeneous shifts in resting membrane potential at the single-cell level, with both depolarizing and hyperpolarizing responses observed (Fig. 6C–E). Across the population, depolarizing shifts predominated (ΔRMP > +2 mV in most cells; Fig. 6F), and the effect was not anatomically segregated within TRN (Fig. 6G). Despite this heterogeneity in membrane potential, propofol consistently increased intrinsic excitability. Input impedance increased, spike output to depolarizing steps rose, and spike threshold shifted in a direction consistent with facilitated initiation (Fig. 6H–K). Propofol also shortened spike latency (Fig. 6M) and shifted the F–I curve upward (Fig. 6N), with representative examples showing enhanced burst discharge (Fig. 6L). Thus, clinically relevant propofol levels can bias TRN PV neurons toward heightened responsiveness and burst propensity, providing a cellular substrate for the altered TRN firing patterns observed in vivo during early recovery. Propofol reshapes TRN PV →MD output and PFC→TRN input coupling We next asked how these intrinsic changes translate into pathway-level signalling. Using ex vivo whole-cell recordings with optogenetic activation, we assayed TRN PV output to mediodorsal thalamus (MD) and prefrontal cortical (PFC) input to TRN (Fig. 7A–C,H–J). To measure TRN PV inhibition of MD, ChR2 was expressed in TRN PV neurons and MD relay neurons were recorded while TRN PV terminals were stimulated. Propofol (10 μM) increased light-evoked inhibitory drive, yielding larger eIPSCs (Fig. 7D,E) and enhanced hyperpolarizing eIPSPs in current clamp (Fig. 7F,G). Propofol therefore strengthens TRN PV -mediated inhibition of MD, a thalamic hub linked to prefrontal-dependent cognition. To probe cortical recruitment, ChR2 was expressed in excitatory PFC neurons (CaMKIIα promoter) and TRN neurons were recorded while PFC axons were stimulated within TRN (Fig. 7H–J). Propofol did not produce a consistent change in the amplitude of light-evoked EPSCs (Fig. 7K,L), yet it increased synaptically driven spike output, with more action potentials per stimulus train under propofol than under ACSF (Fig. 7M,N). This pattern is consistent with enhanced intrinsic excitability amplifying the input–output transformation without requiring larger monosynaptic currents. In combination, stronger TRN PV inhibition of MD and enhanced TRN responsiveness to PFC drive provide a circuit-level mechanism by which propofol could distort thalamocortical timing control, disrupt spindle organization (Figs. 2–3), and degrade learning during early recovery (Fig. 1). Discussion Recovery of consciousness after general anaesthesia is commonly judged by the return of behavioral responsiveness, yet patients may exhibit persistent, transient deficits in forming coherent experiences and new memories during early recovery [20] . Here we show in mice that early after emergence from propofol anaesthesia, behavioral arousal and motor coordination can recover while learning remains transiently impaired, supporting the concept that the recovery of the content of consciousness lags behind the recovery of arousal [21] . Across two aversive-associative learning paradigms (inhibitory avoidance and contextual/auditory-cued fear conditioning), memory formation was markedly disrupted when training occurred at 15 min after RORR, then progressively recovered with longer post-emergence intervals. Importantly, locomotor activity and rotarod performance had returned to control levels by 15 min after RORR, indicating that the learning impairment at this time point cannot be readily explained by gross motor suppression. Our findings reveal an early post-emergence window in which, even after propofol-only general anaesthesia (without surgical insult), the brain is behaviorally awake yet remains computationally suboptimal for stable memory encoding. A prominent systems-level signature of this window was the persistence of aberrant spindle-like thalamocortical dynamics. Using an identical detection pipeline across states, spindles during propofol and at RORR 10-20 min differed from physiological NREM spindles in event rate and amplitude and were spectrally redistributed, with changes in center frequency, spindle-band power and bandwidth. Thus, emergence did not immediately restore canonical NREM-like spindle organization; instead, the early post-anaesthesia brain expressed a distinct oscillatory regime. Given that spindles reflect precisely timed thalamocortical interactions rather than merely “sleepiness” [22, 23] , our findings are consistent with the idea that early emergence is a transitional state in which behavioral responsiveness returns while the timing architecture supporting efficient information processing remains abnormal. Our in vivo recordings nominate the TRN as a key locus for this altered timing. During physiological NREM, TRN spiking showed stereotyped recruitment around spindle onset, whereas under propofol TRN units exhibited stronger spindle-aligned engagement and elevated burst-like discharge. Crucially, these changes persisted at RORR 10-20 min, including higher firing rate, increased burst propensity and altered intra-burst structure, together with reshaped spike–spindle phase coupling. Because spindle expression depends on coordinated TRN–thalamus interactions and precise phase relationships, persistent TRN hyper-recruitment and phase dysregulation provide a mechanistically plausible substrate for the spindle abnormalities observed during early recovery [24] . In this framework, the post-emergence deficit is not simply reduced or delayed spindle occurrence; rather, it reflects a qualitative mismatch in thalamocortical timing that may be poorly suited to gating and consolidating new information [25, 26] . We further refined this model by identifying TRN PV neurons as a major contributing cell population whose dynamics are strongly perturbed early after emergence. Fibre photometry showed that TRN PV activity remained tonically elevated after RORR, yet stimulus-evoked responses were markedly attenuated at RORR +5 min and +15 min, with partial recovery by +30 min. This combination—elevated baseline with blunted evoked responsiveness—suggests a disruption of population gain and dynamic range during early recovery, consistent with reduced capacity to represent incoming sensory or behaviorally relevant signals despite increased ongoing activity [27] . Notably, during training conducted in the early recovery window (RORR +15 min), TRN PV population activity exhibited reduced temporal precision and atypical event-locked modulation relative to baseline, and these deviations were still evident during the recall session 24 h later. Although fibre photometry cannot resolve microcircuit heterogeneity or millisecond spike timing, the observed population-level loss of temporal specificity is consistent with the idea that early emergence from general anaesthesia is marked by a behaviorally relevant disruption of TRN PV timing signals. Given the TRN’s role in providing temporally precise inhibition to coordinate thalamic ensemble activity and thalamocortical synchrony, such imprecision may degrade the thalamic coordination required for effective information encoding and stabilization of memory traces [3, 28] . Our chemogenetic manipulations provide causal support that TRN PV function constrains post-anaesthetic cognitive vulnerability. Suppressing TRN PV neurons with hM4D(Gi) altered cortical spindle expression and exacerbated impairments in contextual and cued fear memory as well as inhibitory avoidance performance. In contrast, global activation of TRN PV neurons with hM3D(Gq) did not improve behavioral outcomes, despite measurable effects on spindle properties. This pattern suggests that intact TRN PV activity is required to prevent further deterioration, yet simply increasing TRN PV activity is not sufficient to restore encoding during this time window. One parsimonious interpretation is that the critical factor is not overall TRN PV tone but the appropriate temporal structure of TRN output—i.e., restoring correct phase relationships and recruitment dynamics may require temporally precise interventions rather than slow, global modulation [29, 30] . To connect these in vivo signatures to a plausible cellular mechanism, our slice recordings demonstrate that propofol can act directly on TRN PV neurons to alter intrinsic excitability. Although propofol shifted resting membrane potential bidirectionally across cells, it consistently increased input impedance, facilitated spike initiation (including threshold and latency measures), increased spike output across depolarizing current steps, and promoted burst-like firing. These effects provide a cellular explanation for the elevated firing and burst propensity observed in vivo during propofol and early after emergence. More broadly, they suggest that propofol can leave TRN PV neurons in a sensitised intrinsic state that amplifies responses to synaptic inputs, potentially destabilising normal thalamocortical timing even after behavioral arousal has returned [29] . Importantly, the persistence of robust spindles after RORR reflects the characteristic post-anaesthetic somnolence—animals can perform motor tasks when engaged, yet rapidly transition back into sleep when unstimulated—highlighting the dissociation between behavioral arousal and thalamocortical timing needed for information encoding [5] . Finally, pathway-specific optogenetic physiology indicates that propofol reshapes both TRN PV output to MD and PFC input to TRN, offering a circuit-level route to impaired cognitive processing early after emergence. Propofol strengthened TRN PV inhibitory control over MD relay neurons, increasing both eIPSC amplitudes and hyperpolarizing eIPSPs. In parallel, propofol enhanced the conversion of optogenetically evoked PFC excitatory input into TRN spiking without a consistent increase in monosynaptic eEPSC amplitude, consistent with an intrinsic excitability-driven gain change. Given the respective roles of the MD thalamus as a representative TRN output target [31] and the mPFC as a representative source of top–down cortical drive to TRN [32] , our slice experiments support a circuit-level account in which propofol shifts the balance of TRN-mediated inhibition onto thalamic relay neurons while also altering cortical recruitment of TRN. This combination provides a mechanistically coherent way to perturb thalamic gating and thalamocortical timing under propofol, in a manner that the early post-emergence brain may be behaviorally responsive yet biased toward a mode of thalamocortical operation that is inefficient for encoding new associations—captured in our data as abnormal spindle-like dynamics and TRN spike–phase disorganization. Two limitations are particularly important. First, our mechanistic insights into TRN dysfunction under propofol come primarily from acute slice experiments, which establish propofol-associated changes in TRN neuronal excitability and in representative TRN input–output pathways under controlled conditions. These results can therefore only indirectly inform cellular mechanisms for the abnormal TRN state observed during emergence from propofol anaesthesia in vivo; a key next step will be to test these synaptic and intrinsic effects in the intact brain (e.g., with in vivo whole-cell or projection-specific recordings/manipulations across induction and emergence). Second, we did not explicitly consider engram-defined TRN ensembles [33] ; incorporating activity-tagging approaches to identify and manipulate TRN engram cells will be important for determining whether memory-relevant subpopulations within TRN are preferentially disrupted during the post-anesthetic period. In summary, our results support a framework in which the recovery of arousal is dissociable from the recovery of content-related processing. Early after emergence from propofol anaesthesia, thalamocortical timing—as indexed by spindle-like dynamics and TRN spike–phase relationships—remains abnormal, and TRN PV circuitry emerges as a major mechanistic contributor. This work highlights early emergence as a distinct and vulnerable brain state and suggests that targeting thalamocortical timing mechanisms, rather than arousal per se, may be critical for improving cognitive recovery after anesthesia. Declarations Acknowledgements This study was supported by the National Natural Science Foundation of China (No. 82430042, No.82560062). Author contributions YZ: Conceptualization; Writing—review & editing; Supervision; Funding acquisition. HL: Methodology; Investigation; Formal analysis; Writing—original draft. SC: Methodology; Investigation; Formal analysis; Writing—original draft. YY: Methodology; Investigation; Formal analysis; Writing—original draft. YC: Investigation. MZ: Investigation. TY: Funding acquisition. All authors reviewed and approved the final manuscript. Competing interests The authors declare no competing interests. References HANSEN S B. Mechanisms of General Anesthesia [J]. Annu Rev Biochem, 2025, 94(1): 503–30. FIELDS A, HUANG J, SCHROEDER D, et al. Agitation in adults in the post-anaesthesia care unit after general anaesthesia [J]. British journal of anaesthesia, 2018, 121(5): 1052–8. WHYTE C J, REDINBAUGH M J, SHINE J M, et al. Thalamic contributions to the state and contents of consciousness [J]. Neuron, 2024, 112(10): 1611–25. SETH A K, BAYNE T. Theories of consciousness [J]. Nat Rev Neurosci, 2022, 23(7): 439–52. LU F, WANG J, QIN X, et al. Differential engagement of thalamic nuclei orchestrates consciousness states across anesthesia, sleep, and disorders of consciousness [J]. Communications biology, 2025, 8(1): 1784. SHINE J M, LEWIS L D, GARRETT D D, et al. The impact of the human thalamus on brain-wide information processing [J]. Nat Rev Neurosci, 2023, 24(7): 416–30. SHERMAN S M. Thalamus plays a central role in ongoing cortical functioning [J]. Nature neuroscience, 2016, 19(4): 533–41. LI S, WANG C, WU S. Spindle oscillations emerge at the critical state of electrically coupled networks in the thalamic reticular nucleus [J]. Cell reports, 2024, 43(10): 114790. IAVARONE E, SIMKO J, SHI Y, et al. Thalamic control of sensory processing and spindles in a biophysical somatosensory thalamoreticular circuit model of wakefulness and sleep [J]. Cell reports, 2023, 42(3): 112200. BANDARABADI M, HERRERA C G, GENT T C, et al. A role for spindles in the onset of rapid eye movement sleep [J]. Nature communications, 2020, 11(1): 5247. LATCHOUMANE C V, NGO H V, BORN J, et al. Thalamic Spindles Promote Memory Formation during Sleep through Triple Phase-Locking of Cortical, Thalamic, and Hippocampal Rhythms [J]. Neuron, 2017, 95(2): 424 – 35 e6. HALASSA M M, CHEN Z, WIMMER R D, et al. State-dependent architecture of thalamic reticular subnetworks [J]. Cell, 2014, 158(4): 808–21. GAO W J. Thalamic cells and pathway for social memory processing and storage [J]. Neuron, 2024, 112(14): 2259–61. CLEMENTE-PEREZ A, MAKINSON S R, HIGASHIKUBO B, et al. Distinct Thalamic Reticular Cell Types Differentially Modulate Normal and Pathological Cortical Rhythms [J]. Cell reports, 2017, 19(10): 2130–42. YOU Y, LIU H, YANG Z, et al. Anesthetic spindles serve as EEG markers of the depth variations in anesthesia induced by multifarious general anesthetics in mouse experiments [J]. Frontiers in pharmacology, 2024, 15(1474923. XU A, ARNAOUT B, SCOTT D A, et al. Intraoperative electroencephalogram-derived measures and their association with postoperative delirium: a systematic review and meta-analysis [J]. British journal of anaesthesia, 2025, 135(6): 1684–703. MAO L, WANG L, HUANG Z, et al. Perioperative neurocognitive disorders: Advances in molecular mechanisms and bioactive molecules [J]. Ageing Res Rev, 2025, 112(102885. NIETHARD N, NGO H V, EHRLICH I, et al. Cortical circuit activity underlying sleep slow oscillations and spindles [J]. Proceedings of the National Academy of Sciences of the United States of America, 2018, 115(39): E9220-E9. GARDNER R J, HUGHES S W, JONES M W. Differential spike timing and phase dynamics of reticular thalamic and prefrontal cortical neuronal populations during sleep spindles [J]. The Journal of neuroscience: the official journal of the Society for Neuroscience, 2013, 33(47): 18469–80. RAO X, WEI C, YI P, et al. Correlation analysis of postoperative cognitive function and event-related potentials in patients undergoing general anesthesia [J]. International journal of surgery, 2025, 111(12): 9321–32. MASHOUR G A. Anesthesia and the neurobiology of consciousness [J]. Neuron, 2024, 112(10): 1553–67. CASTELNOVO A, D'AGOSTINO A, MAYELI A, et al. Sleep Spindle Abnormalities as Neurophysiological Biomarkers of Schizophrenia Spectrum Disorders: From Cellular Mechanisms and Neural Circuits to Clinical Implications [J]. Biological psychiatry, 2025, 98(11): 809–18. CARROLL E E, SHEN Q, KANSARA V, et al. Sleep spindles as a predictor of cognitive motor dissociation and recovery of consciousness after acute brain injury [J]. Nature medicine, 2025, 31(5): 1578–85. FERNANDEZ L M J, LUTHI A. Sleep Spindles: Mechanisms and Functions [J]. Physiol Rev, 2020, 100(2): 805–68. SCOTT D N, MUKHERJEE A, NASSAR M R, et al. Thalamocortical architectures for flexible cognition and efficient learning [J]. Trends in cognitive sciences, 2024, 28(8): 739–56. ARU J, SUZUKI M, LARKUM M E. Cellular Mechanisms of Conscious Processing [J]. Trends in cognitive sciences, 2020, 24(10): 814–25. AHISSAR E, NELINGER G, ASSA E, et al. Thalamocortical loops as temporal demodulators across senses [J]. Communications biology, 2023, 6(1): 562. MIN B K, KIM H S, PINOTSIS D A, et al. Thalamocortical inhibitory dynamics support conscious perception [J]. NeuroImage, 2020, 220(117066. WOLFF M, MORCEAU S, FOLKARD R, et al. A thalamic bridge from sensory perception to cognition [J]. Neuroscience and biobehavioral reviews, 2021, 120(222 – 35. FOGERSON P M, HUGUENARD J R. Tapping the Brakes: Cellular and Synaptic Mechanisms that Regulate Thalamic Oscillations [J]. Neuron, 2016, 92(4): 687–704. WANG S, LI Z, QIAO B, et al. The neural circuit mechanism for auditory responses in the mediodorsal thalamic nucleus of awake mice [J]. Communications biology, 2025, 8(1): 884. HADINGER N, BOSZ E, TOTH B, et al. Region-selective control of the thalamic reticular nucleus via cortical layer 5 pyramidal cells [J]. Nature neuroscience, 2023, 26(1): 116–30. WANG F, SUN H, CHEN M, et al. The thalamic reticular nucleus orchestrates social memory [J]. Neuron, 2024, 112(14): 2368-85 e11. Additional Declarations There is no conflict of interest Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: revise 03 May, 2026 Review # 2 received at journal 28 Apr, 2026 Review # 1 received at journal 07 Apr, 2026 Reviewer # 2 agreed at journal 22 Mar, 2026 Reviewer # 1 agreed at journal 18 Mar, 2026 Reviewers invited by journal 18 Mar, 2026 Submission checks completed at journal 23 Feb, 2026 First submitted to journal 22 Feb, 2026 Unknown event 22 Feb, 2026 Editor assigned by journal 22 Feb, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8937051","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":595929646,"identity":"5ab0b256-f508-4858-9533-25c277df4651","order_by":0,"name":"Yu Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA00lEQVRIiWNgGAWjYJACZih14MCHH0RqgGphSzw4s4c0LTzGhznYiNAgH5F/+HNBzR27DTdyPhxm4GGQ5xc7gF+L4Y1kBuMZx54lb7iRu+FwgQWD4czZCQS0zEhmSOZhO5xsANIyg4chweA2EVoO8/wDacl5cJiHjQgt8hLJjM28bYftgFoYiNNiwPPYmJm373CC5JlnBsBAliDsF/n2xMefeb4dtuc7nvz4w4cfNvL80oRsuQBRkLjgAJiWwK8cbEs/RKm9fANhxaNgFIyCUTBCAQApTUtSD/59VQAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0001-8819-6033","institution":"Affiliated Hospital of Zunyi Medical University","correspondingAuthor":true,"prefix":"","firstName":"Yu","middleName":"","lastName":"Zhang","suffix":""},{"id":595929647,"identity":"e8e6db4c-6ec4-4a3d-9b10-125e5a59c1da","order_by":1,"name":"Hui Liu","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical University","correspondingAuthor":false,"prefix":"","firstName":"Hui","middleName":"","lastName":"Liu","suffix":""},{"id":595929648,"identity":"544729cc-c230-4a0b-8d94-a8f50a098618","order_by":2,"name":"Shiyu Chen","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical University","correspondingAuthor":false,"prefix":"","firstName":"Shiyu","middleName":"","lastName":"Chen","suffix":""},{"id":595929649,"identity":"fca95e5d-ddc7-47b5-9979-622b7a7095fd","order_by":3,"name":"Ying You","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical University","correspondingAuthor":false,"prefix":"","firstName":"Ying","middleName":"","lastName":"You","suffix":""},{"id":595929650,"identity":"b5338f15-6770-469c-b013-73c547e33a90","order_by":4,"name":"Yuxuan Chen","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yuxuan","middleName":"","lastName":"Chen","suffix":""},{"id":595929651,"identity":"ec7535e3-e4cb-41d3-b3be-cb00dee2c89d","order_by":5,"name":"Meimei Zhao","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical University","correspondingAuthor":false,"prefix":"","firstName":"Meimei","middleName":"","lastName":"Zhao","suffix":""},{"id":595929652,"identity":"7ceddad4-4220-4a79-84f4-9d3d4993fe71","order_by":6,"name":"Tian Yu","email":"","orcid":"","institution":"Affiliated Hospital of Zunyi Medical University","correspondingAuthor":false,"prefix":"","firstName":"Tian","middleName":"","lastName":"Yu","suffix":""}],"badges":[],"createdAt":"2026-02-22 05:35:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8937051/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8937051/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105067350,"identity":"14d3aea9-0b9b-48cf-8920-91a54408117e","added_by":"auto","created_at":"2026-03-20 14:16:22","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":9491022,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLearning remains impaired during early emergence from propofol anaesthesia despite recovery of behavioral arousal\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA, Experimental design. C57BL/6 mice received continuous propofol infusion via the tail vein until electroencephalography (EEG) exhibited burst suppression (marker of deep anaesthesia), at which point infusion was stopped. Loss of righting reflex (LORR) and recovery of righting reflex (RORR) were used as behavioral markers of loss and recovery of arousal state, respectively. After RORR, mice underwent inhibitory avoidance (light–dark avoidance) training or fear-conditioning training at 5, 15 or 30 min, followed by memory testing 24 h later.\u003c/p\u003e\n\u003cp\u003eB,C, Inhibitory avoidance memory. B, Step-through latency in the 24-h test. C, Time spent in the dark compartment in the 24-h test. Training initiated shortly after RORR impaired subsequent avoidance performance, with recovery as the training–RORR interval increased.\u003c/p\u003e\n\u003cp\u003eD, Open-field locomotion measured on the training day in the same cohort of mice. Total distance travelled at the indicated time after RORR shows reduced locomotion at early time points but recovery by 10-20 min.\u003c/p\u003e\n\u003cp\u003eE, Fear-conditioning protocol schematic. During training, conditioned stimulus (CS) presentations were paired with unconditioned stimulus (US) footshocks. Contextual and cued memory were assessed 24 h later during a context test (no CS) and a cue test (CS presentations), respectively.\u003c/p\u003e\n\u003cp\u003eF,G, Fear memory measured 24 h after training. F, Contextual freezing during the context test. G, Cued freezing during the cue test. Training performed shortly after RORR reduced subsequent freezing, with recovery as training was delayed.\u003c/p\u003e\n\u003cp\u003eH, Open-field assay (performed on the training day in the same mice) confirming that locomotor activity is restored by 15 min after RORR, supporting that memory-encoding deficits observed at this time are not attributable to gross locomotor impairment.\u003c/p\u003e\n\u003cp\u003eI, Rotarod performance (performed on the training day in the same mice) showing recovery of motor coordination/endurance by 15 min after RORR.\u003c/p\u003e\n\u003cp\u003eDots represent individual mice (n indicated by the number of dots). Centre values are means and error bars are SD Statistical comparisons were performed using one-way ANOVA followed by Dunnett’s multiple-comparisons test (two-sided), comparing each post-RORR group (5, 15 and 30 min) to the control group. Significance is denoted as\u003cem\u003e P \u0026lt; 0.05 (*), P \u0026lt; 0.01 (**), and P \u0026lt; 0.001 (***)\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003eAbbreviations: LORR, loss of righting reflex; RORR, recovery of righting reflex; EEG, electroencephalography; CS, conditioned stimulus; US, unconditioned stimulus.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8937051/v1/07fc916b42bed7797a271cd1.png"},{"id":105067354,"identity":"493683d1-e16e-4533-87c0-20b89814bcf8","added_by":"auto","created_at":"2026-03-20 14:16:22","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":34253991,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCortical spindles 10-20 min after emergence from propofol anaesthesia remain distinct from physiological NREM sleep spindles\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA, Representative cortical EEG recordings during NREM sleep (left), propofol anaesthesia (middle) and 10-20 min after recovery of righting reflex (RORR; right). Top, state annotation. Middle, spindle-band–filtered signal (7–15 Hz) with detected spindles indicated (red) using the same spindle-detection pipeline across all conditions. Bottom, expanded view of example events.\u003c/p\u003e\n\u003cp\u003eB–D, Spindle morphology across conditions. B, Peak-to-peak amplitude. C, Spindle duration. D, Spindle rate. Compared with NREM sleep spindles, spindles detected during propofol anaesthesia and at RORR 10-20 min show altered features and occurrence.\u003c/p\u003e\n\u003cp\u003eE, Spindle time–frequency representations during NREM sleep, propofol anaesthesia and RORR 10-20 min, illustrating condition-dependent differences in spectral structure.\u003c/p\u003e\n\u003cp\u003eF–H, Spindle power spectral density (PSD; 0–100 Hz) for NREM sleep (F), propofol (G) and RORR 10-20 min (H). Thin grey lines indicate individual spindles; coloured line indicates the mean.\u003c/p\u003e\n\u003cp\u003eI–K, Summary spectral features of spindles. I, Spindle centre frequency. J, Spindle-band power. K, Spindle bandwidth. Spindles detected at RORR 10-20 min remain spectrally distinct from physiological NREM sleep spindles. Each dot represents an individual detected spindle.\u003c/p\u003e\n\u003cp\u003eEach dot represents one spindle event (data collected from 6 mice per condition). Event-level comparisons across conditions were performed using mixed-effects models with condition as a fixed effect and mouse as a random effect (random intercept), followed by Dunnett-adjusted post hoc tests versus NREM sleep. Significance is denoted as P \u0026lt; 0.05 (*), P \u0026lt; 0.01 (**), and P \u0026lt; 0.001 (***).\u003c/p\u003e\n\u003cp\u003eAbbreviations: EEG, electroencephalography; NREM, non-rapid eye movement; RORR, recovery of righting reflex; PSD, power spectral density.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8937051/v1/e51577ed138b38989d210d1c.png"},{"id":105562907,"identity":"1091011c-7e31-4c8b-8d70-e7df06870a7b","added_by":"auto","created_at":"2026-03-27 12:45:13","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":6517245,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTRN spiking and spindle-phase dynamics are altered 10-20 min after emergence from propofol anaesthesia\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA, Histological verification of the recording site in the thalamic reticular nucleus (TRN) (coronal section, bregma -0.9 mm; arrow indicates electrode track/recording location).\u003c/p\u003e\n\u003cp\u003eB, Representative spike waveforms from an isolated TRN unit (individual waveforms overlaid; thick trace, mean waveform; scale bar, 1 ms).\u003c/p\u003e\n\u003cp\u003eC, Example feature space used for spike sorting (PC1, waveform amplitude; PC2, waveform width).\u003c/p\u003e\n\u003cp\u003eD–F, Spindle-aligned TRN local field potential (LFP) and spiking during NREM sleep (D), propofol anaesthesia (E) and 10-20 min after recovery of righting reflex (F). Top, spindle-band TRN LFP (grey, individual spindle events; coloured trace, mean). Middle, spike rasters aligned to spindle onset (each row, one spindle event). Bottom, peri-event time histogram (PETH) of mean spike rate aligned to spindle onset.\u003c/p\u003e\n\u003cp\u003eG, Inter-spike interval (ISI) distributions for the three conditions, illustrating condition-dependent changes in spiking patterning.\u003c/p\u003e\n\u003cp\u003eH, Polar histograms of TRN spike timing as a function of instantaneous spindle phase derived from the TRN LFP.\u003c/p\u003e\n\u003cp\u003eI, Mean resultant vectors summarizing preferred spike phase (vector angle) and phase-locking strength (vector length) for each condition.\u003c/p\u003e\n\u003cp\u003eJ–M, Animal-level quantification of TRN firing and bursting across conditions, including mean spike frequency (J), proportion of burst spiking (K), spikes per burst (L) and average intraburst firing rate (aIBF; M) (definitions in Methods). For each mouse, metrics were computed by aggregating across recorded TRN units (Methods). Bars show mean ± SD; points indicate individual mice.\u003c/p\u003e\n\u003cp\u003eStatistical comparisons were performed using one-way ANOVA with Dunnett’s post hoc test. Significance is denoted as \u003cem\u003eP \u0026lt; 0.05 (*), P \u0026lt; 0.01 (**), and P \u0026lt; 0.001 (***)\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003eAbbreviations: TRN, thalamic reticular nucleus; LFP, local field potential; NREM, non-rapid eye movement; RORR, recovery of righting reflex; ISI, inter-spike interval; PC, principal component; PETH, peri-event time histogram; aIBF, average intraburst firing rate.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8937051/v1/f5ac0410a3ae520606598bde.png"},{"id":105067353,"identity":"c20bc38f-7fc6-4759-b71c-16d287a69807","added_by":"auto","created_at":"2026-03-20 14:16:22","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":21619767,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePersistently elevated TRN\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003ePV\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e activity and blunted sensory- and recall-related responses following propofol anaesthesia\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA, Fibre photometry strategy for recording TRN parvalbumin-positive (PV) neurons. A Cre-dependent calcium indicator (AAV-DIO-GCaMP) was expressed in TRN\u003csup\u003ePV\u003c/sup\u003e neurons and an optic fibre was implanted above the TRN. Representative histology confirms viral expression and fibre placement (TRN outlined).\u003cbr\u003e\nB, Representative TRN\u003csup\u003ePV\u003c/sup\u003e photometry trace (ΔF/F) before, during and after propofol anaesthesia, showing a prolonged elevation in TRN\u003csup\u003ePV\u003c/sup\u003e activity after emergence compared with the pre-anaesthetic baseline.\u003cbr\u003e\nC, Experimental timeline for quantifying TRN\u003csup\u003ePV\u003c/sup\u003e responses to somatosensory electrical stimulation across recovery, including RORR +5, +15 and +30 min time points.\u003cbr\u003e\nD, Mean TRN\u003csup\u003ePV\u003c/sup\u003e calcium responses (ΔF/F) aligned to somatosensory stimulation in control (CON) mice and at RORR +5, +15 and +30 min. Shading indicates mean ± SD\u003cbr\u003e\nE, Heat maps of stimulus-evoked TRN\u003csup\u003ePV\u003c/sup\u003e responses for individual mice across conditions (rows, mice).\u003cbr\u003e\nF, Mouse-level quantification of stimulus-evoked TRN\u003csup\u003ePV\u003c/sup\u003e activity, computed as the mean ΔF/F in the 0–5 s post-stimulus window (as indicated).\u003cbr\u003e\nG, Summary of stimulus-evoked response magnitude (mean ΔF/F, 0–5 s post-stimulus) across conditions.\u003c/p\u003e\n\u003cp\u003eH, Experimental design to assess TRN\u003csup\u003ePV\u003c/sup\u003e activity during learning performed at RORR 10-20 min and during a subsequent recall test.\u003cbr\u003e\nI, Mean TRN\u003csup\u003ePV\u003c/sup\u003e calcium signals aligned to the training epoch in CON mice and mice trained at RORR 10-20 min (with or without preceding anaesthesia, as indicated). Shading indicates mean ± SD\u003cbr\u003e\nJ, Heat maps of TRN\u003csup\u003ePV\u003c/sup\u003e activity during training for individual mice in each group.\u003cbr\u003e\nK, Mouse-level quantification of training-evoked TRN\u003csup\u003ePV\u003c/sup\u003e activity, calculated as the mean ΔF/F during the 0–5 s analysis window (as indicated).\u003cbr\u003e\nL, Mouse-level quantification of recall-associated TRN\u003csup\u003ePV\u003c/sup\u003e activity, calculated as the mean ΔF/F during the 5–10 s analysis window (as indicated), showing persistent abnormalities in TRN\u003csup\u003ePV\u003c/sup\u003e dynamics in mice trained at RORR 10-20 min.\u003c/p\u003e\n\u003cp\u003eStatistical comparisons were performed using one-way ANOVA with Dunnett’s post hoc test. Significance is denoted as \u003cem\u003eP \u0026lt; 0.05 (*), P \u0026lt; 0.01 (**), and P \u0026lt; 0.001 (***)\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003eAbbreviations: TRN, thalamic reticular nucleus; PV, parvalbumin; RORR, recovery of righting reflex; CON, control.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8937051/v1/9e72610b9669ab14a1daf60a.png"},{"id":105067355,"identity":"f7dca115-bb96-4858-93b4-a73735f3c79d","added_by":"auto","created_at":"2026-03-20 14:16:22","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":40635433,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChemogenetic inhibition, but not activation, of TRN\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003ePV\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e neurons exacerbates learning deficits 10-20 min after propofol emergence\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA, Experimental design and targeting strategy. PV-Cre mice received bilateral TRN injections of Cre-dependent DREADDs (AAV-DIO-hM4D(Gi)-mCherry or AAV-DIO-hM3D(Gq)-mCherry) followed by an expression period. CNO or saline was administered before propofol anaesthesia; mice were trained at RORR 10-20 min and memory was assessed 24 h later (contextual and cued fear memory and passive avoidance, as indicated). Representative histology confirms viral targeting and fibre/electrode placement (TRN outlined).\u003cbr\u003e\nB, Representative cortical EEG traces (7–15 Hz band) with detected spindles (markers) from TRN\u003csup\u003ePV\u003c/sup\u003e::hM4D(Gi) mice following saline or CNO.\u003cbr\u003e\nC–E, Mouse-level quantification of spindle properties in TRN\u003csup\u003ePV\u003c/sup\u003e::hM4D(Gi) mice under saline versus CNO: spindle density (C), spindle duration (D) and spindle amplitude (E).\u003cbr\u003e\nF–H, Behavioral performance of TRN\u003csup\u003ePV\u003c/sup\u003e::hM4D(Gi) mice trained at RORR 10-20 min under saline versus CNO: contextual fear memory (F), cued fear memory (G) and passive avoidance latency (H).\u003cbr\u003e\nI, Representative cortical EEG traces (7–15 Hz band) with detected spindles (markers) from TRN\u003csup\u003ePV\u003c/sup\u003e::hM3D(Gq) mice following saline or CNO.\u003cbr\u003e\nJ–L, Mouse-level quantification of spindle properties in TRN\u003csup\u003ePV\u003c/sup\u003e::hM3D(Gq) mice under saline versus CNO: spindle density (J), spindle duration (K) and spindle amplitude (L).\u003cbr\u003e\nM–O, Behavioral performance of TRN\u003csup\u003ePV\u003c/sup\u003e::hM3D(Gq) mice trained at RORR 10-20 min under saline versus CNO: contextual fear memory (M), cued fear memory (N) and passive avoidance latency (O).\u003c/p\u003e\n\u003cp\u003eEach point represents one mouse; lines connect paired measurements from the same mouse (saline vs CNO). Bars indicate mean ± SD Statistical comparisons were performed using two-sided paired \u003cem\u003et\u003c/em\u003e-tests. Significance is denoted as \u003cem\u003eP \u0026lt; 0.05 (*), P \u0026lt; 0.01 (**), and P \u0026lt; 0.001 (***)\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003eAbbreviations: TRN, thalamic reticular nucleus; PV, parvalbumin; CNO, clozapine-N-oxide; EEG, electroencephalography; RORR, recovery of righting reflex.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-8937051/v1/af43b2c2b20aba01a808296c.png"},{"id":105067356,"identity":"5a851cc0-d1b9-4d8e-9d71-c00e5914ce81","added_by":"auto","created_at":"2026-03-20 14:16:22","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":6150799,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePropofol differentially shifts membrane potential while increasing intrinsic excitability in TRN\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003ePV\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e neurons ex vivo\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA, Representative images illustrating whole-cell patch-clamp recording in acute thalamic slices targeting the TRN.\u003cbr\u003e\nB, Post hoc confirmation of a recorded TRN\u003csup\u003ePV\u003c/sup\u003e neuron. A biocytin-filled neuron (green) recorded from PV-Cre::tdTomato tissue (PV cells, red).\u003cbr\u003e\nC, Representative current-clamp recordings from spontaneously firing TRN\u003csup\u003ePV\u003c/sup\u003e neurons showing that bath-applied propofol (10 μM; blue bar) can produce either a depolarizing (top) or hyperpolarizing (bottom) shift in membrane potential across cells.\u003cbr\u003e\nD, Representative current-clamp recordings from quiescent TRN\u003csup\u003ePV\u003c/sup\u003e neurons showing depolarizing (top) or hyperpolarizing (bottom) membrane-potential shifts during propofol application (10 μM; blue bar).\u003cbr\u003e\nE, Summary of propofol-evoked changes in resting membrane potential (RMP). Neurons were classified as depolarizing (ΔRMP \u0026gt; +2 mV) or hyperpolarizing (ΔRMP \u0026lt; −2 mV). Each dot represents one neuron; lines connect ACSF and propofol measurements from the same neuron.\u003cbr\u003e\nF, Fraction of recorded TRN\u003csup\u003ePV\u003c/sup\u003e neurons exhibiting depolarization or hyperpolarization in response to propofol (10 μM), based on the criteria in E.\u003cbr\u003e\nG, Anatomical distribution of recorded neurons within the TRN, colour-coded by the direction of the propofol-evoked membrane-potential shift.\u003cbr\u003e\nH, Representative responses of a TRN\u003csup\u003ePV\u003c/sup\u003e neuron to step current injections in ACSF and during propofol application (10 μM), illustrating increased spiking output.\u003cbr\u003e\nI–K, Population summaries of intrinsic properties in ACSF versus propofol (10 μM): input impedance (Ri) (I), number of evoked spikes (J) and action potential threshold (K). Each dot represents one neuron; lines indicate paired measurements from the same neuron.\u003cbr\u003e\nL, Representative responses of a TRN\u003csup\u003ePV\u003c/sup\u003e neuron exhibiting burst firing in ACSF and during propofol application (10 μM).\u003cbr\u003e\nM, Summary of spike latency (from current-step onset to the first spike) in ACSF versus propofol (10 μM). Each dot represents one neuron; lines indicate paired measurements.\u003cbr\u003e\nN, Spike output as a function of injected current (F–I relationship) in ACSF and during propofol (10 μM). Data are mean ± SD; repeated measurements were obtained from the same set of neurons across current steps.\u003c/p\u003e\n\u003cp\u003eStatistical significance is denoted as \u003cem\u003eP \u0026lt; 0.05 (*), P \u0026lt; 0.01 (**), and P \u0026lt; 0.001 (***)\u003c/em\u003e. Paired comparisons were assessed using two-sided paired \u003cem\u003et\u003c/em\u003e-tests.\u003c/p\u003e\n\u003cp\u003eAbbreviations: TRN, thalamic reticular nucleus; PV, parvalbumin; RMP, resting membrane potential; ACSF, artificial cerebrospinal fluid.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-8937051/v1/9c1458201affbcf0f0ec7830.png"},{"id":105067352,"identity":"d6bcd09e-8dc4-45b9-abd6-a6661397e8fd","added_by":"auto","created_at":"2026-03-20 14:16:22","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":13988372,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePropofol enhances TRN\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003ePV\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e inhibition of MD and increases TRN responsiveness to prefrontal input ex vivo\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA, Experimental schematic for optogenetic activation of TRN\u003csup\u003ePV\u003c/sup\u003e neurons and recording of their synaptic output onto mediodorsal thalamus (MD) neurons. PV-Cre mice received TRN injections of a Cre-dependent ChR2 virus; in acute thalamic slices, MD neurons were whole-cell recorded while TRN\u003csup\u003ePV\u003c/sup\u003e axons were stimulated with brief blue-light pulses.\u003cbr\u003e\nB, Representative ChR2-dependent expression in TRN\u003csup\u003ePV\u003c/sup\u003e cell bodies (left) and corresponding TRN\u003csup\u003ePV\u003c/sup\u003e axon terminals/fibres in MD (right).\u003cbr\u003e\nC, Example MD neuron filled with biocytin during recording.\u003cbr\u003e\nD, Representative light-evoked inhibitory postsynaptic currents (eIPSCs) recorded in an MD neuron in ACSF and during bath application of propofol (10 μM).\u003cbr\u003e\nE, Summary of eIPSC amplitude in ACSF versus propofol. Each dot represents one recorded MD neuron; lines indicate paired measurements from the same neuron.\u003cbr\u003e\nF, Representative light-evoked inhibitory postsynaptic potentials (eIPSPs) recorded in current clamp in ACSF and during propofol (10 μM).\u003cbr\u003e\nG, Summary of eIPSP amplitude in ACSF versus propofol. Each dot represents one recorded MD neuron; lines indicate paired measurements from the same neuron.\u003c/p\u003e\n\u003cp\u003eH, Experimental schematic for optogenetic activation of prefrontal cortical (PFC) inputs to TRN. Mice received PFC injections of a CaMKIIα-driven ChR2 virus to target excitatory cortical neurons; in acute thalamic slices, TRN neurons were whole-cell recorded while PFC axons in TRN were stimulated with brief blue-light pulses.\u003cbr\u003e\nI, Representative EYFP-labelled PFC axons in TRN.\u003cbr\u003e\nJ, Immunohistochemical confirmation of CaMKIIα-driven ChR2 expression.\u003cbr\u003e\nK, Representative light-evoked excitatory postsynaptic currents (eEPSCs) recorded in a TRN neuron in ACSF and during propofol (10 μM).\u003cbr\u003e\nL, Summary of eEPSC amplitude in ACSF versus propofol. Each dot represents one recorded TRN neuron; lines indicate paired measurements from the same neuron.\u003cbr\u003e\nM, Representative synaptically evoked spiking responses of a TRN neuron to PFC axon stimulation in ACSF and during propofol (10 μM).\u003cbr\u003e\nN, Summary of the number of light-evoked action potentials per stimulus train in ACSF versus propofol. Each dot represents one recorded TRN neuron; lines indicate paired measurements from the same neuron; error bars indicate mean ± SD where shown.\u003c/p\u003e\n\u003cp\u003ePropofol was bath applied at 10 μM as indicated. Statistical significance is denoted as \u003cem\u003eP \u0026lt; 0.05 (*), P \u0026lt; 0.01 (**), and P \u0026lt; 0.001 (***)\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAbbreviations:\u003c/strong\u003e TRN, thalamic reticular nucleus; PV, parvalbumin; MD, mediodorsal thalamus; PFC, prefrontal cortex; ACSF, artificial cerebrospinal fluid; eIPSC, evoked inhibitory postsynaptic current; eIPSP, evoked inhibitory postsynaptic potential; eEPSC, evoked excitatory postsynaptic current.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-8937051/v1/daa57520252f3b8edb0f6e9d.png"},{"id":105067326,"identity":"c84b5ffb-83f7-4a33-9b59-6e891a58ac47","added_by":"auto","created_at":"2026-03-20 14:16:16","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":867392,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8937051/v1/bcf37063-0751-46fa-b497-c6dd6d69dec7.pdf"}],"financialInterests":"There is no conflict of interest","formattedTitle":"\u003cp\u003eTRN\u003csup\u003ePV\u003c/sup\u003e-Dependent Failure of Thalamocortical Timing Delays the Return of Conscious Content during Early Emergence from Propofol Anaesthesia\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGeneral anaesthesia is typically evaluated\u0026mdash;and often operationally defined\u0026mdash;by the loss and recovery of behavioral responsiveness. However, the return of overt arousal does not necessarily imply restoration of the neural computations that support conscious experience and flexible cognition\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]\u003c/sup\u003e. Clinically, patients frequently display a transient period of confusion, amnesia or impaired learning immediately after emergence, despite appearing awake and interactive\u003csup\u003e[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e. This dissociation during emergence from general anaesthesia suggests that the \u0026ldquo;content\u0026rdquo; of consciousness\u0026mdash;operationally defined here as the brain\u0026rsquo;s capacity to acquire, integrate and store new information\u0026mdash;may recover on a slower timescale than arousal \u0026ldquo;state\u0026rdquo; \u003csup\u003e[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIdentifying the circuit-level bottlenecks that delay the recovery of content-related processing after anesthesia remains challenging. A major obstacle is that global behavioral metrics provide limited insight into whether post-emergence brain dynamics have returned to physiological modes of thalamocortical communication\u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/sup\u003e. Thalamocortical timing is a compelling candidate mechanism because it constrains effective information transfer, sensory gating and the coordination of distributed cortical ensembles\u003csup\u003e[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/sup\u003e. In natural sleep, these functions are closely tied to spindle-associated dynamics\u0026mdash;brief, rhythmic events that reflect structured interactions between cortex, thalamus and the thalamic reticular nucleus (TRN)\u003csup\u003e[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]\u003c/sup\u003e. Spindles can support memory-related processing in sleep, yet spindle activity can also accompany reduced environmental responsiveness\u003csup\u003e[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/sup\u003e, indicating that spindle activity may correspond to distinct computational states depending on their timing, coupling and circuit implementation.\u003c/p\u003e \u003cp\u003eThe TRN occupies a strategic position to regulate these state-dependent thalamocortical dynamics\u003csup\u003e[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/sup\u003e. As a GABAergic shell around the dorsal thalamus, the TRN gates thalamic relay activity and shapes thalamocortical synchrony, placing it in a key position to regulate both the global state of consciousness (arousal/behavioral responsiveness) and the content-related computations that support perception, integration and memory\u003csup\u003e[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/sup\u003e. Parvalbumin-positive (PV) TRN neurons, in particular, are well suited to impose temporally precise inhibition and thereby influence both the transitions into and out of general-anaesthetic unconsciousness and the fine structure of spindle-associated rhythms that are thought to support information processing.\u003csup\u003e[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]\u003c/sup\u003e. Disrupted TRN timing control may leave a post-emergence state that is behaviorally aroused yet less capable of encoding new information, consistent with content-related processing lagging behind arousal.\u003c/p\u003e \u003cp\u003ePropofol provides a powerful experimental framework to test this idea. Beyond producing reversible unconsciousness, propofol strongly reorganizes thalamocortical activity and promotes rhythmic patterns that can resemble spindle-range (alpha range) oscillations\u003csup\u003e[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/sup\u003e. Our prior work\u003csup\u003e[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]\u003c/sup\u003e shows that spindle-range oscillations do not immediately return to physiological sleep-like dynamics after propofol emergence. What remains unresolved is whether these altered rhythms causally interfere with information acquisition during early recovery, and whether PV-defined TRN circuitry is a principal substrate for the persistent disruption of thalamocortical timing.\u003c/p\u003e \u003cp\u003eAccordingly, we focus on the minutes-to-tens-of-minutes period immediately after emergence from general anaesthesia, a transient state distinct from postoperative delirium\u003csup\u003e[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]\u003c/sup\u003e or longer-lasting perioperative neurocognitive disorders\u003csup\u003e[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]\u003c/sup\u003e. We combined behavior, systems electrophysiology, cell-type-specific calcium photometry, chemogenetic perturbation, and ex vivo circuit physiology to test the hypothesis that recovery of behavioral arousal from propofol anaesthesia precedes recovery of thalamocortical timing dynamics that support learning. We find that learning is impaired early after emergence despite restored gross motor performance, coincident with persistent spindle-like thalamocortical timing abnormalities and disrupted TRN\u003csup\u003ePV\u003c/sup\u003e dynamics. Causal and circuit-level analyses further implicate TRN\u003csup\u003ePV\u003c/sup\u003e gated top-down corticothalamic pathway as a substrate for delayed recovery of learning-relevant processing after anesthetic emergence.\u003c/p\u003e "},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003e1. Animals and ethics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll experiments were performed in accordance with institutional guidelines for the care and use of laboratory animals and complied with relevant animal welfare regulations. Experimental protocols were reviewed and approved by the Experimental Animal Ethics Committee of Zunyi Medical University (Appl. No. zyfy-an-2023-0284).\u003c/p\u003e\n\u003cp\u003eHealthy adult male specific-pathogen-free (SPF) mice were used. Wild-type C57BL/6J mice (10–12 weeks old; 20–25 g) and PV-IRES-Cre mice (10–12 weeks old; 20–25 g) were housed under controlled environmental conditions (25 ± 2 °C; 50 ± 5% humidity; noise \u0026lt; 50 dB) on a 12 h light/12 h dark cycle (lights on 08:00–20:00; lights off 20:00–08:00), with ad libitum access to food and water and free movement in their home cages.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2. Anaesthesia and peri-anaesthetic procedure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e2.1 Experimental groups and overall workflow\u003c/p\u003e\n\u003cp\u003eHealthy adult SPF C57BL/6J mice (male, 10–12 weeks old, 20–25 g) were randomly assigned to one of four experimental conditions: Control (no anaesthesia), and post-emergence groups trained at 5 min, 15 min, or 30 min after recovery of righting reflex (RORR) following propofol anaesthesia. The core experimental workflow was: handling habituation → tail-vein propofol infusion under EEG monitoring → termination of infusion at burst suppression (BS) → monitoring until RORR → behavioral testing at the assigned post-emergence time point.\u003c/p\u003e\n\u003cp\u003e2.2 Pre-experimental habituation and stress minimization\u003c/p\u003e\n\u003cp\u003eTo minimize stress-related confounds in behavioral performance, mice were habituated to experimenter handling by gentle daily holding and stroking for 3 consecutive days prior to experiments. Animals were considered habituated when they could remain calmly in the experimenter’s hands without struggling or attempting to escape, while maintaining spontaneous movement.\u003c/p\u003e\n\u003cp\u003e2.3 Propofol administration and BS endpoint\u003c/p\u003e\n\u003cp\u003ePropofol was administered by continuous tail-vein infusion using a syringe pump at a constant rate of 10 mg kg/min. Anaesthetic depth was titrated to an electrophysiological endpoint: burst suppression (BS) on EEG. Once BS was detected, propofol infusion was immediately discontinued. BS was used as an operational marker of deep anaesthesia. Based on our experimental premise, a brief period of BS per se does not measurably impair learning; therefore, the post-anaesthetic behavioral phenotype was interrogated as a function of time after RORR.\u003c/p\u003e\n\u003cp\u003e2.4 Emergence monitoring and definition of RORR\u003c/p\u003e\n\u003cp\u003eAfter cessation of propofol infusion, animals were continuously monitored until recovery of righting reflex (RORR), defined as the ability to self-right from a supine to a prone position. RORR time served as time zero for post-emergence grouping. Mice were then tested at the designated interval after RORR (+5 min, +15 min, or +30 min) for behavioral assays.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3. Behaviour: learning/memory assays and motor control\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll behavioral experiments were conducted during the light phase. Mice were transferred to the testing room and allowed to acclimate for ≥30 min before each assay. Apparatuses were cleaned with 75% ethanol between animals to minimize olfactory cues.\u003c/p\u003e\n\u003cp\u003e3.1 Behavioral timeline and task order\u003c/p\u003e\n\u003cp\u003eTo reduce potential carry-over effects across paradigms, the same cohort of mice completed behavioral testing in the following order: motor assessment first, then contextual and auditory-cued fear conditioning after a 7-day interval, followed by inhibitory avoidance after an additional 7-day interval.\u003c/p\u003e\n\u003cp\u003e3.2 Rotarod test (motor coordination)\u003c/p\u003e\n\u003cp\u003eThe rotarod test was used to assess motor coordination and balance, quantified as latency to fall.\u003c/p\u003e\n\u003cp\u003eHabituation/training. Mice underwent two adaptation sessions 24 h before testing. During each session, rod speed was ramped from 4 to 10 rpm over 30 s, and mice were allowed to remain on the rod for 10 min. Two sessions were conducted with an inter-session interval of ≥1 h. If a mouse fell during training, it was promptly returned to the rod to continue training until the session ended.\u003c/p\u003e\n\u003cp\u003eTesting. Depending on group assignment, mice were tested at 5 min, 15 min, or 30 min after recovery of righting reflex (RORR), or in a non-anaesthetized control condition. During testing, rod speed was ramped from 4 to 40 rpm over 5 min, and latency to fall was recorded.\u003c/p\u003e\n\u003cp\u003e3.3 Open field test (locomotor activity)\u003c/p\u003e\n\u003cp\u003eThe open field test was used to assess general locomotor activity after anaesthesia. Each mouse was placed in the centre of the arena and allowed to freely explore for 5 min while its trajectory was recorded. The primary outcome measure was total distance travelled during the session.\u003c/p\u003e\n\u003cp\u003e3.4 Contextual and auditory-cued fear conditioning (learning and memory)\u003c/p\u003e\n\u003cp\u003eContextual and auditory-cued fear conditioning were used to quantify associative learning by pairing an auditory conditioned stimulus (CS) with an aversive unconditioned stimulus (US; footshock). Freezing was used as the behavioral readout of learned fear and was defined as the absence of movement other than respiration.\u003c/p\u003e\n\u003cp\u003eTraining (Day 1). Mice were placed in the conditioning chamber and allowed to explore for 2 min, followed by three CS–US pairings within a total training duration of 5 min. Each pairing consisted of an auditory CS (2,000 Hz, 80 dB, 28 s) immediately followed by a footshock US (0.6 mA, 2 s), with a 30 s interval after each shock before the next pairing.\u003c/p\u003e\n\u003cp\u003eTesting (Day 2): (i) Contextual test: Mice were returned to the same chamber for 3 min with no CS and no footshock. (ii) Cued test: Mice were placed in a novel context and presented with the auditory CS (2,000 Hz, 80 dB, 28 s) without footshock for three cycles during a 3-min session.\u003c/p\u003e\n\u003cp\u003eBehaviour quantification. Freezing was automatically scored using Labmaze V3.0 (Beijing ZS Dichuang Technology Development Co., Ltd.). The freezing detection threshold was set to 4%. The primary outcome was percentage freezing (time spent freezing/total observation time × 100%).\u003c/p\u003e\n\u003cp\u003e3.5 Inhibitory avoidance (IA)\u003c/p\u003e\n\u003cp\u003eInhibitory avoidance was used to assess aversive associative learning based on rodents’ preference for dark environments paired with a footshock. The apparatus consisted of connected light and dark compartments; the dark compartment contained a grid floor for shock delivery.\u003c/p\u003e\n\u003cp\u003eTraining. Mice were placed in the dark compartment, and entry was paired with a footshock (0.6 mA, 2 s). During the 5-min training session, each subsequent entry into the dark compartment was paired with the same footshock.\u003c/p\u003e\n\u003cp\u003eTesting. Mice were placed in the light compartment, and the latency to first enter the dark compartment was recorded as the memory index (maximum 300 s; mice not entering within 300 s were assigned a latency of 300 s). Longer latency indicates stronger avoidance learning/memory.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4. EEG Implantation, Acquisition, and Pre-processing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEEG/EMG electrodes were purchased from Kedou (Suzhou) Brain-Computer Technology Co., Ltd. (KD-EEG/EMG-AG) and implanted using standard aseptic stereotaxic procedures. Briefly, animals were anesthetized and secured in a stereotaxic apparatus, the scalp was incised to expose the skull, and the electrode assembly was positioned according to the experimental design and fixed with dental cement; animals were allowed to fully recover before recording. Continuous EEG signals were then acquired using the Apollo portable neural signal data acquisition system (Yige Biotechnology (Nanjing) Co., Ltd.) at a sampling rate of 1000 Hz. Raw data were exported for offline pre-processing, during which recordings were visually inspected and segments with gross movement, poor electrode contact, or other artifacts were excluded; where applicable, signals were re-referenced and baseline-corrected prior to subsequent analyses (e.g., spectral or event-related quantification).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e5. Sleep scoring, spindle detection and spectral analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSleep–wake state was scored prior to spindle quantification using the Lunion Stage automatic sleep analysis system (https://stage.luniondata.com/). Continuous EEG and accompanying EMG channels were segmented and automatically classified into vigilance states; only epochs scored as NREM sleep were included for spindle analyses. For the post-emergence window (10–20 min after recovery of righting reflex, RORR), animals were left undisturbed and frequently re-entered NREM-like sleep/drowsiness, consistent with strong post-propofol sleep pressure; spindles were therefore quantified exclusively from epochs meeting the same NREM criteria used for baseline sleep, allowing direct comparison across conditions (baseline NREM, propofol anaesthesia, and post-emergence NREM at RORR 10-20 min).\u003c/p\u003e\n\u003cp\u003eSpindle events were detected from mouse cortical EEG recordings acquired with the Apollo portable neural signal data acquisition system (Yige Biotechnology (Nanjing) Co., Ltd.) using a batch-detection strategy adapted from prior work on oscillation detection and plasticity mechanisms\u003csup\u003e[18, 19]\u003c/sup\u003e. All offline signal processing, filtering, event identification, feature measurement and statistical summaries were implemented with custom scripts in MATLAB (R2016b, MathWorks). For spindle detection, continuous EEG was band-pass filtered in the sigma range (7–15 Hz), full-wave rectified, and the amplitude envelope was extracted to capture time-varying changes in oscillatory amplitude. A spindle was defined when the envelope exceeded a preset threshold of 1.5× the standard deviation (computed from the corresponding recording segment) and remained above threshold for 0.5–3.0 s. Event onset and offset were determined using threshold-crossing rules, with onset marked by the first positive threshold crossing and termination marked by the last negative threshold crossing. For each detected spindle, the root mean square (RMS) of the band-limited signal was computed, and the mean RMS across events was used as the primary amplitude metric for subsequent analyses. Spectral quantification was performed in MATLAB on the same preprocessed EEG signals, focusing on sigma-band (7–15 Hz) activity.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e6. In vivo TRN spiking recordings, Spike–spindle phase coupling and phase histograms\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eChronic in vivo extracellular recordings were performed in freely moving mice using a nickel–titanium (NiTi) microwire electrode array (Kedou (Suzhou) Brain-Computer Technology Co., Ltd.) implanted stereotaxically to target the thalamic reticular nucleus (TRN; AP = −0.53 mm, ML = ±1.35 mm, DV = −3.65 mm from the skull surface). Electrodes were fixed to the skull with anchor screws and dental acrylic. Neural signals were acquired with the Apollo portable neural signal data acquisition system (Yige Biotechnology (Nanjing) Co., Ltd.) at 3,000 Hz and band-pass filtered to extract a spike band (300–3,000 Hz) and an LFP band (0.5–300 Hz). Spikes were detected using a negative threshold of −4.5× the root mean square (RMS) noise level and sorted offline using Offline Sorter (Plexon) with principal-component features and manual refinement; putative single units were accepted only when the inter-spike-interval (ISI) violation rate was \u0026lt;0.5–1%. Bursts were defined as sequences of spikes with consecutive ISIs \u0026lt;10 ms and containing ≥3 spikes; the average intraburst firing rate (aIBF) was calculated as the mean of 1/ISI across within-burst intervals and then averaged across bursts for each unit and condition.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor peri-spindle analyses, we analysed three conditions: baseline NREM sleep, propofol anaesthesia, and post-emergence NREM sleep (RORR 10-20 min). Each condition included 6 mice. To balance contributions across animals, we randomly sampled 50 spindles per mouse (300 total spindles per condition) for spike-alignment and phase-coupling analyses. Spikes were time-locked to spindle events and peri-event time histograms (PETHs) were computed using 10 ms bins. In addition to spindle-onset alignment, we also generated spindle-triggered averages by band-pass filtering thalamic signals in the spindle band (7–15 Hz) and aligning events to the peak of the central spindle cycle (defined as the maximal-amplitude cycle near the spindle midpoint; t=0). Raster plots depict single-unit activity with each row corresponding to one spindle event, and trial-averaged firing rate was estimated with a 10 ms sliding window with 80% overlap\u003csup\u003e[10]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eSpike–spindle phase coupling was quantified by band-pass filtering EEG/LFP in the spindle band (7–15 Hz) and extracting the instantaneous phase from the analytic signal (Hilbert transform). For each detected spindle epoch, to reduce potential bias in phase estimates introduced by spindle onset/offset detection, the onset and offset times were pseudorandomly jittered by ±0.5 spindle cycles (uniform in phase) independently for each spindle, and spikes were included only if they occurred within the jittered epoch boundaries.\u003c/p\u003e\n\u003cp\u003eEach spike was assigned the corresponding LFP phase angle, and phase non-uniformity was assessed using the Rayleigh test. Multiple comparisons across units were controlled using the Holm–Bonferroni correction. For each unit we computed the preferred phase (circular mean) and phase concentration/phase-locking value (mean resultant length, PLV). Spike–spindle phase histograms were generated for strongly phase-locked units (Rayleigh test, corrected) by binning spike phases over 0–2π and plotting normalized counts in polar coordinates; the black vector indicates each unit’s mean resultant vector (direction = preferred phase, length = PLV) \u003csup\u003e[19]\u003c/sup\u003e. Group summaries were obtained by plotting the population mean preferred phase and mean vector length for each cell class/condition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e7. Fiber Photometry Calcium Imaging\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUnder aseptic conditions, PV-IRES-Cre mice were anesthetized with 1.4% isoflurane, shaved, and placed prone in a stereotaxic frame; erythromycin ophthalmic ointment was applied, the scalp was disinfected, and 1% lidocaine was administered subcutaneously. A ~1 cm midline incision was made to expose the skull, hydrogen peroxide was used to remove periosteum, and the head was leveled to within ±0.03 mm (AP/ML). The thalamic reticular nucleus (TRN) injection site was targeted using the following coordinates relative to bregma: AP = −0.53 mm, ML = ±1.35 mm, DV = −3.65 mm; a craniotomy was drilled and the dura was carefully removed. A Cre-dependent calcium indicator virus (rAAV-EF1α-DIO-jGCaMP7b) was loaded into a 1 μL syringe (220 nL; 5 min wait) and infused into the TRN at 40 nL/min for 180 nL, with the needle left in place for 10 min to minimize backflow. An optical fiber was then lowered to DV = 3.65 mm and secured using 454 cyanoacrylate adhesive, supplemented with two skull screws and self-curing dental acrylic; mice received intramuscular penicillin for 3 consecutive days and were recorded ~3 weeks later to allow viral expression. For fiber photometry, mice were handled for 3 days and habituated to the room for ≥30 min before testing; recordings were performed under light-restricted conditions using a Multi-Channel Fiber Photometry Device (410/470; Inper, Hangzhou, China) connected via a ceramic ferrule, with excitation delivered to the TRN and emitted fluorescence collected and converted to an analog voltage signal. A 10 s stimulus-free baseline was used to estimate F0, and fluorescence changes were expressed as ΔF/F = (F − F0)/F0.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e8. Chemogenetics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAdult male SPF PV-IRES-Cre mice (10–12 weeks old, 20–25 g; total n = 35) received bilateral TRN injections of Cre-dependent DREADD viruses and were assigned to control (AAV-DIO-mCherry, n = 12), inhibition (AAV-DIO-hM4D(Gi)-mCherry, n = 10) or activation (AAV-DIO-hM3D(Gq)-mCherry, n = 13) groups; after surgery mice recovered on a warming pad and received intramuscular penicillin once daily for 3 consecutive days, and experiments were conducted 3 weeks later to allow stable expression, with additional gentle handling for 3 days before testing to minimize stress. For chemogenetic manipulation, clozapine-N-oxide (CNO; 1 mg/kg, i.p.) or an equal volume of saline was administered 30 min before subsequent procedures in a within-subject, counterbalanced design (saline vs CNO sessions separated by a 7-day washout). After pretreatment, mice underwent the propofol anaesthesia paradigm (see above) and were trained 15 min after emergence, with memory tested 24 h later (contextual and auditory-cued fear conditioning and inhibitory avoidance); for EEG experiments, EEG connectors were attached after pretreatment and mice received tail-vein propofol infusion (10 mg/kg/min) until EEG burst suppression, followed by continuous EEG recording for 35 min after recovery of righting reflex. At the end of experiments, brains were collected for immunofluorescence to verify TRN viral expression and its colocalization with PV-positive neurons by fluorescence microscopy.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e9. Acute slice preparation and patch-clamp electrophysiology\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAcute thalamic slices were prepared from PV-tdTomato and SOM-tdTomato mice (4–6 weeks old). Mice were deeply anaesthetized with isoflurane and decapitated, and brains were rapidly removed into ice-cold (4 °C), carbogenated (95% O\u003csub\u003e2\u003c/sub\u003e/5% CO\u003csub\u003e2\u003c/sub\u003e) sucrose-based cutting solution containing (in mM): 234 sucrose, 3 KCl, 1.25 NaH\u003csub\u003e2\u003c/sub\u003ePO\u003csub\u003e4\u003c/sub\u003e, 10 MgSO\u003csub\u003e4\u003c/sub\u003e·7H\u003csub\u003e2\u003c/sub\u003eO, 0.5 CaCl\u003csub\u003e2\u003c/sub\u003e, 2H\u003csub\u003e2\u003c/sub\u003eO, 26 NaHCO\u003csub\u003e3\u003c/sub\u003e, 10 D-glucose. Coronal slices containing the TRN were cut at 240 μm using a vibrating microtome (HM 650V) in continuously carbogenated cutting solution. Slices were then recovered at 32 °C for 20 min in carbogenated sucrose-based solution and subsequently held for ≥60 min at room temperature in carbogenated recording ACSF (NaCl-based) before experiments.\u003c/p\u003e\n\u003cp\u003eWhole-cell patch-clamp recordings were performed at room temperature under infrared DIC optics. TRN\u003csup\u003ePV\u003c/sup\u003e neurons were identified by tdTomato fluorescence, and MD neurons were targeted based on anatomical location. Pipettes (4–8 MΩ) were pulled from borosilicate glass. For intrinsic excitability measurements (current clamp), a K\u003csup\u003e+\u003c/sup\u003e-based internal solution was used (composition as specified in the internal-solution section). For optogenetically evoked synaptic currents, a Cs\u003csup\u003e+\u003c/sup\u003e-based internal solution was used. To record optogenetically evoked inhibitory postsynaptic currents (eIPSCs), cells were voltage-clamped at 0 mV to isolate GABA\u003csub\u003eA\u003c/sub\u003e receptor–mediated currents. For TRN\u003csup\u003ePV\u003c/sup\u003e →MD connectivity experiments, TRN\u003csup\u003ePV\u003c/sup\u003e neurons expressed ChR2, and TRN\u003csup\u003ePV\u003c/sup\u003e axon terminals in the MD were stimulated using a 470-nm LED light source (X-Cite 110LED; Excelitas Technologies, Japan) coupled to an Olympus microscope with 10 ms pulses delivered at 0.5 Hz, evoking light-evoked IPSCs in MD neurons. To assess mPFC→TRN excitatory input, ChR2 was expressed in mPFC projection neurons and their axon terminals in the TRN were stimulated with the same 470-nm LED protocol; TRN neurons were voltage-clamped at −70 mV to record light-evoked excitatory postsynaptic currents (eEPSCs). Propofol was diluted in recording ACSF to a final concentration of 10 μM and bath-applied via perfusion at 3–4 ml/h. Data were excluded if series resistance exceeded 25 MΩ or changed by \u0026gt;20% during the recording.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e10. Histology and verification of targeting\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAt the completion of experiments, mice were anaesthetized with 1.4% isoflurane and transcardially perfused with PBS followed by 4% paraformaldehyde (PFA). Brains were removed, post-fixed in 4% PFA for 12 h, cryoprotected in 30% sucrose until sunk, embedded in OCT, and sectioned coronally at 30 μm on a −20 °C cryostat (brains were pre-cooled for ~30 min before cutting). Free-floating sections were stored in PBS and processed for PV immunofluorescence: sections were washed in PBS (3 × 10 min), permeabilized/blocked (2 h, room temperature; PBS containing 10% goat serum, 1% BSA and 0.1% Triton X-100), incubated with rabbit anti-parvalbumin (29312-1-AP, 1:500; overnight, room temperature), washed, and then incubated with goat polyclonal anti-rabbit IgG secondary antibodies (ab150077 or ab150080, 1:1,000; 2 h, room temperature, protected from light). Sections were washed, mounted, and coverslipped with a DAPI-containing anti-fade medium. Viral expression in the TRN and its co-localization with PV-positive neurons, as well as fibre/electrode targeting where applicable, were examined using a BC43 confocal microscope (Oxford Instruments, UK), and animals with mistargeting or insufficient/misplaced expression were excluded.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e11. Statistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eStatistical analyses were performed using MATLAB (R2016b), GraphPad Prism and Clampfit. All tests were two-sided unless stated otherwise, with P \u0026lt; 0.05 considered statistically significant; multiple-comparison procedures are indicated in the figure legends. Data are presented as mean ± SD. For analyses of EEG spindles and unit activity, the primary observational unit was the detected event (that is, each spindle or spike-derived event, as specified); events were quantified within predefined 10-min epochs for physiological NREM sleep (10 min), propofol anaesthesia (10 min) and early after emergence (10–20 min after RORR). Because events are nested within animals, event-level statistics were performed using mixed-effects models with condition/epoch as a fixed effect and mouse as a random effect (random intercept), accounting for within-mouse dependence and unequal event counts across mice/conditions, with Dunnett-adjusted post hoc comparisons to the indicated reference condition when applicable. For spindle-phase coupling, instantaneous phase was extracted from band-pass-filtered signals using the Hilbert transform; phase non-uniformity was assessed with the Rayleigh test and multiple comparisons were controlled using the Holm–Bonferroni method. Exact P values and the numbers of mice and total events are reported in the figure legends.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eArousal recovers before learning after propofol anaesthesia\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo test whether behavioral emergence from propofol anaesthesia coincides with recovery of learning capacity, we infused propofol intravenously until the cortical EEG reached burst suppression and then stopped the infusion. Loss and recovery of the righting reflex (LORR and RORR) were used as behavioral markers for entry into and exit from anaesthesia (Fig. 1A). To isolate effects on memory formation, mice were trained at defined times after RORR (5, 15, or 30 min) and tested 24 h later.\u003c/p\u003e\n\u003cp\u003eIn the inhibitory avoidance task, training soon after emergence disrupted later memory expression. Relative to controls, mice trained at RORR +5 min showed shorter step-through latencies in the 24-h test and spent more time in the dark compartment (Fig. 1B,C), indicating weaker avoidance memory. Performance improved as the training-RORR interval increased, and mice trained at RORR +30 min were indistinguishable from controls (Fig. 1B,C). These data identify a transient post-emergence interval in which learning is impaired despite restoration of righting and exploratory behaviour.\u003c/p\u003e\n\u003cp\u003eAn independent associative assay yielded the same temporal profile. Mice trained in auditory fear conditioning shortly after RORR displayed reduced freezing at 24 h in both contextual and cued tests, whereas freezing recovered when training was delayed (Fig. 1E–G). Thus, memory acquisition lags behind behavioral emergence from propofol.\u003c/p\u003e\n\u003cp\u003eResidual motor suppression is unlikely to account for this deficit. In the same cohort, open-field activity was reduced at RORR +5 min but returned to control levels by 15 min (Fig. 1D,H), and rotarod performance likewise recovered by 15 min (Fig. 1I). The learning impairment evident when training occurs within the RORR +15 min window therefore cannot be explained by gross locomotor or coordination deficits.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEarly emergence cortical spindles remain altered during the learning-impaired window\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next asked whether this early post-anaesthetic learning deficit (centred at RORR +15 min; Fig. 1) is accompanied by persistent changes in thalamocortical oscillations. Because spindle generation depends on thalamic circuitry, cortical spindles were quantified using a single detection pipeline applied identically across three conditions: physiological NREM sleep, ongoing propofol anaesthesia, and the early post-emergence window (RORR 10-20 min; Fig. 2A).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRelative to physiological NREM sleep, propofol markedly changed spindle amplitude and occurrence (Fig. 2B,D), while spindle duration was less affected (Fig. 2C). During post-emergence NREM (RORR 10-20 min), spindle amplitude and event rate did not return to the NREM pattern (Fig. 2B,D), indicating that the mechanisms shaping spindles remain perturbed during early recovery even when the animal is in NREM sleep.\u003c/p\u003e\n\u003cp\u003eTime–frequency maps and power spectra supported a condition-dependent redistribution of spindle-band structure (Fig. 2E–H). Compared with NREM spindles, events detected during propofol and post-emergence NREM showed shifts in sigma-band organisation, with corresponding changes in centre frequency, spindle-band power, and bandwidth (Fig. 2I–K). Spindle dynamics therefore remain state-dependently altered during the same post-emergence interval in which learning is compromised.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTRN neurons show persistent firing and spindle-phase dysregulation at RORR 10-20 min\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo determine whether the spindle phenotype reflects persistent dysfunction at a candidate generator node, we recorded TRN activity in vivo. Recording locations were confirmed histologically (Fig. 3A), and well-isolated TRN units were identified by waveform properties and spike sorting (Fig. 3B,C). TRN spiking was then aligned to detected spindles and compared across physiological NREM sleep, propofol anaesthesia, and early post-emergence (RORR 10-20 min; Fig. 3D–F).\u003c/p\u003e\n\u003cp\u003eDuring physiological NREM, TRN firing showed stereotyped modulation time-locked to spindle onset (Fig. 3D). Under propofol, TRN neurons exhibited stronger spindle-associated recruitment, with denser rasters and a larger peri-event increase in firing rate (Fig. 3E). Elevated recruitment persisted into the RORR 10-20 min window (Fig. 3F), matching the time window of behavioral learning impairment.\u003c/p\u003e\n\u003cp\u003eAt the level of spike timing, inter-spike interval distributions shifted under propofol and remained shifted after emergence (Fig. 3G), consistent with increased high-frequency discharge and enhanced burst-like firing relative to NREM sleep. Mouse-aggregated measures confirmed higher mean firing rates during propofol with incomplete normalization during early recovery (Fig. 3J). Burst propensity and intraburst structure were similarly elevated—including burst fraction, spikes per burst, and average intraburst firing rate (aIBF) (Fig. 3K–M).\u003c/p\u003e\n\u003cp\u003eBecause spindle organization depends on phase-specific timing, we further quantified TRN spike timing relative to instantaneous spindle phase derived from TRN LFP (Gardner et al., 2013). Both phase preference and phase-locking strength differed across conditions (Fig. 3H,I), and altered coupling was still detectable at RORR 10-20 min. TRN output thus remains temporally reorganized after behavioral emergence, providing a plausible substrate for the persistent changes in spindle structure observed during early recovery.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTRN\u003csup\u003ePV\u003c/sup\u003e population dynamics are altered during post-anaesthetic training and recall\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo probe cell-type-specific dynamics, we recorded fibre-photometry calcium signals from\u0026nbsp;TRN\u003csup\u003ePV\u003c/sup\u003e neurons (Fig. 4A). TRN\u003csup\u003ePV\u003c/sup\u003e activity increased during propofol and remained elevated after RORR relative to the pre-anaesthetic baseline (Fig. 4B).\u003c/p\u003e\n\u003cp\u003eWe then measured the responsiveness of TRN\u003csup\u003ePV\u003c/sup\u003e neurons to controlled input during recovery. Somatosensory electrical stimulation evoked robust calcium transients in controls, whereas evoked responses were strongly attenuated at RORR +5 and +15 min and partially recovered by RORR +30 min (Fig. 4C–G). Heat maps across mice showed consistent suppression of stimulus-locked responses early after emergence (Fig. 4E), and quantification within the 0–5 s post-stimulus window confirmed reduced response magnitude at these time points (Fig. 4F,G). This pattern indicates elevated ongoing activity coupled to reduced dynamic responsiveness during early recovery.\u003c/p\u003e\n\u003cp\u003eFinally, we recorded TRN\u003csup\u003ePV\u003c/sup\u003e activity during learning performed at RORR 10-20 min and during recall 24 h later (Fig. 4H). Relative to controls, mice trained during this early recovery window showed altered TRN\u003csup\u003ePV\u003c/sup\u003e activity during training (Fig. 4I–K; 0–5 s window) and during the subsequent recall test (Fig. 4L; 5–10 s window). Early post-anaesthetic learning is therefore associated with persistent deviations in TRN\u003csup\u003ePV\u003c/sup\u003e population dynamics across both acquisition and later retrieval.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eChemogenetic suppression of TRN\u003csup\u003ePV\u003c/sup\u003e neurons exacerbates learning deficits during early emergence\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo test whether TRN\u003csup\u003ePV\u003c/sup\u003e neurons contribute to recovery of learning during early emergence, we used a within-subject chemogenetic design in PV-Cre mice expressing inhibitory (hM4D(Gi)) or excitatory (hM3D(Gq)) DREADDs in TRN (Fig. 5A). Mice received CNO or saline before propofol, were trained at RORR 10-20 min, and were tested 24 h later in fear conditioning and inhibitory avoidance (Fig. 5A).\u003c/p\u003e\n\u003cp\u003eIn TRN\u003csup\u003ePV\u003c/sup\u003e::hM4D(Gi) mice, CNO altered post-emergence spindle expression (Fig. 5B–E) and worsened behavioral performance: freezing decreased in contextual and cued tests, and passive-avoidance latency was reduced compared with saline sessions (Fig. 5F–H). Reducing TRN\u003csup\u003ePV\u003c/sup\u003e activity during the peri-emergence period therefore aggravates the learning deficit.\u003c/p\u003e\n\u003cp\u003eIn contrast, TRN\u003csup\u003ePV\u003c/sup\u003e::hM3D(Gq) activation did not improve behavioral outcomes. Although CNO produced measurable changes in spindle features (Fig. 5I–L), fear memory and passive avoidance did not differ between CNO and saline sessions (Fig. 5M–O). These results dissociate necessity from sufficiency: intact TRN\u003csup\u003ePV\u003c/sup\u003e activity appears required to prevent further deterioration of learning during early recovery, whereas globally increasing TRN\u003csup\u003ePV\u003c/sup\u003e activity alone is not sufficient to restore memory formation at this time point.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePropofol directly alters intrinsic membrane properties of TRN\u003csup\u003ePV\u003c/sup\u003e neurons ex vivo\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo test whether propofol can directly reconfigure TRN\u003csup\u003ePV\u003c/sup\u003e excitability in a manner that could support persistent in vivo firing changes, we performed whole-cell current-clamp recordings from identified TRN\u003csup\u003ePV\u003c/sup\u003e neurons in acute thalamic slices (Fig. 6A,B). Bath application of propofol (10 μM) produced heterogeneous shifts in resting membrane potential at the single-cell level, with both depolarizing and hyperpolarizing responses observed (Fig. 6C–E). Across the population, depolarizing shifts predominated (ΔRMP \u0026gt; +2 mV in most cells; Fig. 6F), and the effect was not anatomically segregated within TRN (Fig. 6G).\u003c/p\u003e\n\u003cp\u003eDespite this heterogeneity in membrane potential, propofol consistently increased intrinsic excitability. Input impedance increased, spike output to depolarizing steps rose, and spike threshold shifted in a direction consistent with facilitated initiation (Fig. 6H–K). Propofol also shortened spike latency (Fig. 6M) and shifted the F–I curve upward (Fig. 6N), with representative examples showing enhanced burst discharge (Fig. 6L). Thus, clinically relevant propofol levels can bias TRN\u003csup\u003ePV\u003c/sup\u003e neurons toward heightened responsiveness and burst propensity, providing a cellular substrate for the altered TRN firing patterns observed in vivo during early recovery.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePropofol reshapes TRN\u003csup\u003ePV\u003c/sup\u003e→MD output and PFC→TRN input coupling\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next asked how these intrinsic changes translate into pathway-level signalling. Using ex vivo whole-cell recordings with optogenetic activation, we assayed TRN\u003csup\u003ePV\u003c/sup\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003eoutput to mediodorsal thalamus (MD) and prefrontal cortical (PFC) input to TRN (Fig. 7A–C,H–J).\u003c/p\u003e\n\u003cp\u003eTo measure TRN\u003csup\u003ePV\u003c/sup\u003e inhibition of MD, ChR2 was expressed in TRN\u003csup\u003ePV\u003c/sup\u003e neurons and MD relay neurons were recorded while TRN\u003csup\u003ePV\u003c/sup\u003e terminals were stimulated. Propofol (10 μM) increased light-evoked inhibitory drive, yielding larger eIPSCs (Fig. 7D,E) and enhanced hyperpolarizing eIPSPs in current clamp (Fig. 7F,G). Propofol therefore strengthens TRN\u003csup\u003ePV\u003c/sup\u003e-mediated inhibition of MD, a thalamic hub linked to prefrontal-dependent cognition.\u003c/p\u003e\n\u003cp\u003eTo probe cortical recruitment, ChR2 was expressed in excitatory PFC neurons (CaMKIIα promoter) and TRN neurons were recorded while PFC axons were stimulated within TRN (Fig. 7H–J). Propofol did not produce a consistent change in the amplitude of light-evoked EPSCs (Fig. 7K,L), yet it increased synaptically driven spike output, with more action potentials per stimulus train under propofol than under ACSF (Fig. 7M,N). This pattern is consistent with enhanced intrinsic excitability amplifying the input–output transformation without requiring larger monosynaptic currents. In combination, stronger TRN\u003csup\u003ePV\u003c/sup\u003e inhibition of MD and enhanced TRN responsiveness to PFC drive provide a circuit-level mechanism by which propofol could distort thalamocortical timing control, disrupt spindle organization (Figs. 2–3), and degrade learning during early recovery (Fig. 1).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eRecovery of consciousness after general anaesthesia is commonly judged by the return of behavioral responsiveness, yet patients may exhibit persistent, transient deficits in forming coherent experiences and new memories during early recovery\u003csup\u003e[20]\u003c/sup\u003e. Here we show in mice that early after emergence from propofol anaesthesia, behavioral arousal and motor coordination can recover while learning remains transiently impaired, supporting the concept that the recovery of the content of consciousness lags behind the recovery of arousal\u003csup\u003e[21]\u003c/sup\u003e. Across two aversive-associative learning paradigms (inhibitory avoidance and contextual/auditory-cued fear conditioning), memory formation was markedly disrupted when training occurred at 15 min after RORR, then progressively recovered with longer post-emergence intervals. Importantly, locomotor activity and rotarod performance had returned to control levels by 15 min after RORR, indicating that the learning impairment at this time point cannot be readily explained by gross motor suppression. Our findings reveal an early post-emergence window in which, even after propofol-only general anaesthesia (without surgical insult), the brain is behaviorally awake yet remains computationally suboptimal for stable memory encoding.\u003c/p\u003e\n\u003cp\u003eA prominent systems-level signature of this window was the persistence of aberrant spindle-like thalamocortical dynamics. Using an identical detection pipeline across states, spindles during propofol and at RORR 10-20 min differed from physiological NREM spindles in event rate and amplitude and were spectrally redistributed, with changes in center frequency, spindle-band power and bandwidth. Thus, emergence did not immediately restore canonical NREM-like spindle organization; instead, the early post-anaesthesia brain expressed a distinct oscillatory regime. Given that spindles reflect precisely timed thalamocortical interactions rather than merely \u0026ldquo;sleepiness\u0026rdquo;\u003csup\u003e[22, 23]\u003c/sup\u003e, our findings are consistent with the idea that early emergence is a transitional state in which behavioral responsiveness returns while the timing architecture supporting efficient information processing remains abnormal.\u003c/p\u003e\n\u003cp\u003eOur in vivo recordings nominate the TRN as a key locus for this altered timing. During physiological NREM, TRN spiking showed stereotyped recruitment around spindle onset, whereas under propofol TRN units exhibited stronger spindle-aligned engagement and elevated burst-like discharge. Crucially, these changes persisted at RORR 10-20 min, including higher firing rate, increased burst propensity and altered intra-burst structure, together with reshaped spike\u0026ndash;spindle phase coupling. Because spindle expression depends on coordinated TRN\u0026ndash;thalamus interactions and precise phase relationships, persistent TRN hyper-recruitment and phase dysregulation provide a mechanistically plausible substrate for the spindle abnormalities observed during early recovery\u003csup\u003e[24]\u003c/sup\u003e. In this framework, the post-emergence deficit is not simply reduced or delayed spindle occurrence; rather, it reflects a qualitative mismatch in thalamocortical timing that may be poorly suited to gating and consolidating new information\u003csup\u003e[25, 26]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWe further refined this model by identifying TRN\u003csup\u003ePV\u003c/sup\u003e neurons as a major contributing cell population whose dynamics are strongly perturbed early after emergence. Fibre photometry showed that TRN\u003csup\u003ePV\u003c/sup\u003e activity remained tonically elevated after RORR, yet stimulus-evoked responses were markedly attenuated at RORR +5 min and +15 min, with partial recovery by +30 min. This combination\u0026mdash;elevated baseline with blunted evoked responsiveness\u0026mdash;suggests a disruption of population gain and dynamic range during early recovery, consistent with reduced capacity to represent incoming sensory or behaviorally relevant signals despite increased ongoing activity\u003csup\u003e[27]\u003c/sup\u003e. Notably, during training conducted in the early recovery window (RORR +15 min), TRN\u003csup\u003ePV\u003c/sup\u003e population activity exhibited reduced temporal precision and atypical event-locked modulation relative to baseline, and these deviations were still evident during the recall session 24 h later. Although fibre photometry cannot resolve microcircuit heterogeneity or millisecond spike timing, the observed population-level loss of temporal specificity is consistent with the idea that early emergence from general anaesthesia is marked by a behaviorally relevant disruption of TRN\u003csup\u003ePV\u003c/sup\u003e timing signals. Given the TRN\u0026rsquo;s role in providing temporally precise inhibition to coordinate thalamic ensemble activity and thalamocortical synchrony, such imprecision may degrade the thalamic coordination required for effective information encoding and stabilization of memory traces\u003csup\u003e[3, 28]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eOur chemogenetic manipulations provide causal support that TRN\u003csup\u003ePV\u003c/sup\u003e function constrains post-anaesthetic cognitive vulnerability. Suppressing TRN\u003csup\u003ePV\u003c/sup\u003e neurons with hM4D(Gi) altered cortical spindle expression and exacerbated impairments in contextual and cued fear memory as well as inhibitory avoidance performance. In contrast, global activation of TRN\u003csup\u003ePV\u003c/sup\u003e neurons with hM3D(Gq) did not improve behavioral outcomes, despite measurable effects on spindle properties. This pattern suggests that intact TRN\u003csup\u003ePV\u003c/sup\u003e activity is required to prevent further deterioration, yet simply increasing TRN\u003csup\u003ePV\u003c/sup\u003e activity is not sufficient to restore encoding during this time window. One parsimonious interpretation is that the critical factor is not overall TRN\u003csup\u003ePV\u003c/sup\u003e tone but the appropriate temporal structure of TRN output\u0026mdash;i.e., restoring correct phase relationships and recruitment dynamics may require temporally precise interventions rather than slow, global modulation\u003csup\u003e[29, 30]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eTo connect these in vivo signatures to a plausible cellular mechanism, our slice recordings demonstrate that propofol can act directly on TRN\u003csup\u003ePV\u003c/sup\u003e neurons to alter intrinsic excitability. Although propofol shifted resting membrane potential bidirectionally across cells, it consistently increased input impedance, facilitated spike initiation (including threshold and latency measures), increased spike output across depolarizing current steps, and promoted burst-like firing. These effects provide a cellular explanation for the elevated firing and burst propensity observed in vivo during propofol and early after emergence. More broadly, they suggest that propofol can leave TRN\u003csup\u003ePV\u003c/sup\u003e neurons in a sensitised intrinsic state that amplifies responses to synaptic inputs, potentially destabilising normal thalamocortical timing even after behavioral arousal has returned\u003csup\u003e[29]\u003c/sup\u003e. Importantly, the persistence of robust spindles after RORR reflects the characteristic post-anaesthetic somnolence\u0026mdash;animals can perform motor tasks when engaged, yet rapidly transition back into sleep when unstimulated\u0026mdash;highlighting the dissociation between behavioral arousal and thalamocortical timing needed for information encoding\u003csup\u003e[5]\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eFinally, pathway-specific optogenetic physiology indicates that propofol reshapes both TRN\u003csup\u003ePV\u003c/sup\u003e output to MD and PFC input to TRN, offering a circuit-level route to impaired cognitive processing early after emergence. Propofol strengthened TRN\u003csup\u003ePV\u003c/sup\u003e inhibitory control over MD relay neurons, increasing both eIPSC amplitudes and hyperpolarizing eIPSPs. In parallel, propofol enhanced the conversion of optogenetically evoked PFC excitatory input into TRN spiking without a consistent increase in monosynaptic eEPSC amplitude, consistent with an intrinsic excitability-driven gain change. Given the respective roles of the MD thalamus as a representative TRN output target\u003csup\u003e[31]\u003c/sup\u003e and the mPFC as a representative source of top\u0026ndash;down cortical drive to TRN\u003csup\u003e[32]\u003c/sup\u003e, our slice experiments support a circuit-level account in which propofol shifts the balance of TRN-mediated inhibition onto thalamic relay neurons while also altering cortical recruitment of TRN. This combination provides a mechanistically coherent way to perturb thalamic gating and thalamocortical timing under propofol, in a manner that the early post-emergence brain may be behaviorally responsive yet biased toward a mode of thalamocortical operation that is inefficient for encoding new associations\u0026mdash;captured in our data as abnormal spindle-like dynamics and TRN spike\u0026ndash;phase disorganization.\u003c/p\u003e\n\u003cp\u003eTwo limitations are particularly important. First, our mechanistic insights into TRN dysfunction under propofol come primarily from acute slice experiments, which establish propofol-associated changes in TRN neuronal excitability and in representative TRN input\u0026ndash;output pathways under controlled conditions. These results can therefore only indirectly inform cellular mechanisms for the abnormal TRN state observed during emergence from propofol anaesthesia in vivo; a key next step will be to test these synaptic and intrinsic effects in the intact brain (e.g., with in vivo whole-cell or projection-specific recordings/manipulations across induction and emergence). Second, we did not explicitly consider engram-defined TRN ensembles\u003csup\u003e[33]\u003c/sup\u003e; incorporating activity-tagging approaches to identify and manipulate TRN engram cells will be important for determining whether memory-relevant subpopulations within TRN are preferentially disrupted during the post-anesthetic period.\u003c/p\u003e\n\u003cp\u003eIn summary, our results support a framework in which the recovery of arousal is dissociable from the recovery of content-related processing. Early after emergence from propofol anaesthesia, thalamocortical timing\u0026mdash;as indexed by spindle-like dynamics and TRN spike\u0026ndash;phase relationships\u0026mdash;remains abnormal, and TRN\u003csup\u003ePV\u003c/sup\u003e circuitry emerges as a major mechanistic contributor. This work highlights early emergence as a distinct and vulnerable brain state and suggests that targeting thalamocortical timing mechanisms, rather than arousal per se, may be critical for improving cognitive recovery after anesthesia.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by the National Natural Science Foundation of China (No. 82430042, No.82560062).\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYZ: Conceptualization; Writing\u0026mdash;review \u0026amp; editing; Supervision; Funding acquisition.\u003c/p\u003e\n\u003cp\u003eHL: Methodology; Investigation; Formal analysis; Writing\u0026mdash;original draft.\u003c/p\u003e\n\u003cp\u003eSC: Methodology; Investigation; Formal analysis; Writing\u0026mdash;original draft.\u003c/p\u003e\n\u003cp\u003eYY: Methodology; Investigation; Formal analysis; Writing\u0026mdash;original draft.\u003c/p\u003e\n\u003cp\u003eYC: Investigation.\u003c/p\u003e\n\u003cp\u003eMZ: Investigation.\u003c/p\u003e\n\u003cp\u003eTY: Funding acquisition.\u003c/p\u003e\n\u003cp\u003eAll authors reviewed and approved the final manuscript.\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\n"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eHANSEN S B. Mechanisms of General Anesthesia [J]. Annu Rev Biochem, 2025, 94(1): 503\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFIELDS A, HUANG J, SCHROEDER D, et al. Agitation in adults in the post-anaesthesia care unit after general anaesthesia [J]. British journal of anaesthesia, 2018, 121(5): 1052\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWHYTE C J, REDINBAUGH M J, SHINE J M, et al. Thalamic contributions to the state and contents of consciousness [J]. Neuron, 2024, 112(10): 1611\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSETH A K, BAYNE T. Theories of consciousness [J]. Nat Rev Neurosci, 2022, 23(7): 439\u0026ndash;52.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLU F, WANG J, QIN X, et al. Differential engagement of thalamic nuclei orchestrates consciousness states across anesthesia, sleep, and disorders of consciousness [J]. Communications biology, 2025, 8(1): 1784.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSHINE J M, LEWIS L D, GARRETT D D, et al. The impact of the human thalamus on brain-wide information processing [J]. Nat Rev Neurosci, 2023, 24(7): 416\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSHERMAN S M. Thalamus plays a central role in ongoing cortical functioning [J]. Nature neuroscience, 2016, 19(4): 533\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLI S, WANG C, WU S. Spindle oscillations emerge at the critical state of electrically coupled networks in the thalamic reticular nucleus [J]. Cell reports, 2024, 43(10): 114790.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIAVARONE E, SIMKO J, SHI Y, et al. Thalamic control of sensory processing and spindles in a biophysical somatosensory thalamoreticular circuit model of wakefulness and sleep [J]. Cell reports, 2023, 42(3): 112200.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBANDARABADI M, HERRERA C G, GENT T C, et al. A role for spindles in the onset of rapid eye movement sleep [J]. Nature communications, 2020, 11(1): 5247.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLATCHOUMANE C V, NGO H V, BORN J, et al. Thalamic Spindles Promote Memory Formation during Sleep through Triple Phase-Locking of Cortical, Thalamic, and Hippocampal Rhythms [J]. Neuron, 2017, 95(2): 424\u0026thinsp;\u0026ndash;\u0026thinsp;35 e6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHALASSA M M, CHEN Z, WIMMER R D, et al. State-dependent architecture of thalamic reticular subnetworks [J]. Cell, 2014, 158(4): 808\u0026ndash;21.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGAO W J. Thalamic cells and pathway for social memory processing and storage [J]. Neuron, 2024, 112(14): 2259\u0026ndash;61.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCLEMENTE-PEREZ A, MAKINSON S R, HIGASHIKUBO B, et al. Distinct Thalamic Reticular Cell Types Differentially Modulate Normal and Pathological Cortical Rhythms [J]. Cell reports, 2017, 19(10): 2130\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYOU Y, LIU H, YANG Z, et al. Anesthetic spindles serve as EEG markers of the depth variations in anesthesia induced by multifarious general anesthetics in mouse experiments [J]. Frontiers in pharmacology, 2024, 15(1474923.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXU A, ARNAOUT B, SCOTT D A, et al. Intraoperative electroencephalogram-derived measures and their association with postoperative delirium: a systematic review and meta-analysis [J]. British journal of anaesthesia, 2025, 135(6): 1684\u0026ndash;703.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMAO L, WANG L, HUANG Z, et al. Perioperative neurocognitive disorders: Advances in molecular mechanisms and bioactive molecules [J]. Ageing Res Rev, 2025, 112(102885.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNIETHARD N, NGO H V, EHRLICH I, et al. Cortical circuit activity underlying sleep slow oscillations and spindles [J]. Proceedings of the National Academy of Sciences of the United States of America, 2018, 115(39): E9220-E9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGARDNER R J, HUGHES S W, JONES M W. Differential spike timing and phase dynamics of reticular thalamic and prefrontal cortical neuronal populations during sleep spindles [J]. The Journal of neuroscience: the official journal of the Society for Neuroscience, 2013, 33(47): 18469\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRAO X, WEI C, YI P, et al. Correlation analysis of postoperative cognitive function and event-related potentials in patients undergoing general anesthesia [J]. International journal of surgery, 2025, 111(12): 9321\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMASHOUR G A. Anesthesia and the neurobiology of consciousness [J]. Neuron, 2024, 112(10): 1553\u0026ndash;67.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCASTELNOVO A, D'AGOSTINO A, MAYELI A, et al. Sleep Spindle Abnormalities as Neurophysiological Biomarkers of Schizophrenia Spectrum Disorders: From Cellular Mechanisms and Neural Circuits to Clinical Implications [J]. Biological psychiatry, 2025, 98(11): 809\u0026ndash;18.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCARROLL E E, SHEN Q, KANSARA V, et al. Sleep spindles as a predictor of cognitive motor dissociation and recovery of consciousness after acute brain injury [J]. Nature medicine, 2025, 31(5): 1578\u0026ndash;85.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFERNANDEZ L M J, LUTHI A. Sleep Spindles: Mechanisms and Functions [J]. Physiol Rev, 2020, 100(2): 805\u0026ndash;68.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSCOTT D N, MUKHERJEE A, NASSAR M R, et al. Thalamocortical architectures for flexible cognition and efficient learning [J]. Trends in cognitive sciences, 2024, 28(8): 739\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eARU J, SUZUKI M, LARKUM M E. Cellular Mechanisms of Conscious Processing [J]. Trends in cognitive sciences, 2020, 24(10): 814\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAHISSAR E, NELINGER G, ASSA E, et al. Thalamocortical loops as temporal demodulators across senses [J]. Communications biology, 2023, 6(1): 562.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMIN B K, KIM H S, PINOTSIS D A, et al. Thalamocortical inhibitory dynamics support conscious perception [J]. NeuroImage, 2020, 220(117066.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWOLFF M, MORCEAU S, FOLKARD R, et al. A thalamic bridge from sensory perception to cognition [J]. Neuroscience and biobehavioral reviews, 2021, 120(222\u0026thinsp;\u0026ndash;\u0026thinsp;35.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFOGERSON P M, HUGUENARD J R. Tapping the Brakes: Cellular and Synaptic Mechanisms that Regulate Thalamic Oscillations [J]. Neuron, 2016, 92(4): 687\u0026ndash;704.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWANG S, LI Z, QIAO B, et al. The neural circuit mechanism for auditory responses in the mediodorsal thalamic nucleus of awake mice [J]. Communications biology, 2025, 8(1): 884.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHADINGER N, BOSZ E, TOTH B, et al. Region-selective control of the thalamic reticular nucleus via cortical layer 5 pyramidal cells [J]. Nature neuroscience, 2023, 26(1): 116\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWANG F, SUN H, CHEN M, et al. The thalamic reticular nucleus orchestrates social memory [J]. Neuron, 2024, 112(14): 2368-85 e11.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"experimental-and-molecular-medicine","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"emm","sideBox":"Learn more about [Experimental \u0026 Molecular Medicine](http://www.nature.com/emm/)","snPcode":"12276","submissionUrl":"https://mts-emm.nature.com/cgi-bin/main.plex","title":"Experimental \u0026 Molecular Medicine","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"General anaesthesia, Early emergence, Thalamic reticular nucleus, Thalamocortical timing, Conscious content","lastPublishedDoi":"10.21203/rs.3.rs-8937051/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8937051/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eBackround:\u003cbr\u003e\nRecovery from general anaesthesia is usually defined by return of behavioural responsiveness, but the capacity to encode new information may recover later. We hypothesised that early emergence from propofol reflects incomplete recovery of TRN gated thalamocortical timing needed for learning.\u003c/p\u003e\n\u003cp\u003eMethods:\u003cbr\u003e\nPV-Cre mice received intravenous propofol until cortical EEG reached burst suppression, then infusion was stopped for spontaneous emergence. LORR and RORR indexed behavioural state. Mice were trained after RORR in inhibitory avoidance and auditory fear conditioning, with memory tested 24 h later. Thalamocortical dynamics were measured with electrophysiology using a single spindle-detection pipeline across NREM sleep, propofol anaesthesia, and post-emergence NREM. TRN PV activity was recorded with fibre photometry, suppressed chemogenetically, and examined ex vivo in the prefrontal, TRN, and thalamic circuit.\u003c/p\u003e\n\u003cp\u003eResults:\u003cbr\u003e\nTraining 5 to 15 min after RORR impaired learning despite restored arousal. Spindle-band dynamics remained abnormal early after emergence. TRN neurons were hyper-recruited around spindles, showed more burst-like firing, and exhibited altered spike to spindle phase coupling. Photometry showed elevated baseline but blunted stimulus- and cue-evoked TRN\u003csup\u003ePV\u003c/sup\u003e responses at 5 and 15 min, with partial recovery by 30 min. Chemogenetic TRN\u003csup\u003ePV\u003c/sup\u003e suppression worsened learning deficits. Ex vivo recordings indicated increased TRN\u003csup\u003ePV\u003c/sup\u003e intrinsic excitability, strengthened inhibition onto mediodorsal thalamus, and increased gain to prefrontal inputs without increased prefrontal excitatory drive, consistent with impaired top-down timing.\u003c/p\u003e\n\u003cp\u003eConclusion:\u003cbr\u003e\nEarly emergence is a distinct vulnerable state in which learning lags behind arousal due to persistent TRN\u003csup\u003ePV\u003c/sup\u003e dependent disruption of spindle-associated thalamocortical timing.\u003c/p\u003e","manuscriptTitle":"TRNPV-Dependent Failure of Thalamocortical Timing Delays the Return of Conscious Content during Early Emergence from Propofol Anaesthesia","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-20 14:16:04","doi":"10.21203/rs.3.rs-8937051/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"revise","date":"2026-05-03T23:09:21+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-04-28T13:13:40+00:00","index":2,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-04-07T13:43:16+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-03-23T03:35:38+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-03-18T13:12:17+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewersInvited","content":"","date":"2026-03-18T11:27:04+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-02-23T23:47:28+00:00","index":"","fulltext":""},{"type":"submitted","content":"Experimental \u0026 Molecular Medicine","date":"2026-02-23T04:29:48+00:00","index":"","fulltext":""},{"type":"checksFailed","content":"","date":"2026-02-22T22:53:22+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-02-22T05:34:53+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"experimental-and-molecular-medicine","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"emm","sideBox":"Learn more about [Experimental \u0026 Molecular Medicine](http://www.nature.com/emm/)","snPcode":"12276","submissionUrl":"https://mts-emm.nature.com/cgi-bin/main.plex","title":"Experimental \u0026 Molecular Medicine","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d081d1e8-2c46-4130-90b1-d6e32e51a765","owner":[],"postedDate":"March 20th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"revise","date":"2026-05-03T23:09:21+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":64858842,"name":"Biological sciences/Neuroscience/Cognitive neuroscience"},{"id":64858843,"name":"Biological sciences/Neuroscience/Cellular neuroscience"}],"tags":[],"updatedAt":"2026-05-03T23:10:29+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-20 14:16:04","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8937051","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8937051","identity":"rs-8937051","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}