State-Dependent Remodeling of Astrocytic Proteome and Phosphorylation Signaling Networks Across Wake, Sleep, and General Anesthesia

Abstract

Astrocytes critically regulate states of consciousness, yet their molecular profiles across wake, sleep, and general anesthesia remain unclear. This study conducted proteomic and phosphoproteomic analyses of rat cortical astrocytes across these states using sevoflurane. Data quality was validated using principal component analysis (PCA) and Pearson correlation coefficient (PCC). Proteomics showed state-specific signatures: sleep and anesthesia shared similar changes (downregulated structural proteins, upregulated membrane transport complexes) but diverged in molecular expression. Anesthesia specifically suggested potential activation of cellular differentiation/structural plasticity-related pathways but implied potential disruption of metabolism and molecular clearance processes compared to sleep. Phosphoproteomics revealed the unique phosphorylation changes during general anesthesia compared to wake and normal sleep: downregulated phosphorylation of nuclear casein kinase and cyclin-dependent kinase substrate 1 (NUCKS1) at Ser188, suggesting the potential suppression of nuclear transcription and/or cell cycle activity, which may act as a potential molecular signature associated with the anesthetic state. Clustering analysis showed that sleep was associated with upregulated mRNA processing, while anesthesia indicated potential enhancement of synaptic signaling and suggested possible suppression of development-related programs. In summary, astrocytes undergo extensive molecular reprogramming during transitions of consciousness; while they share common features in morphological remodeling, sleep and anesthesia differ fundamentally in astrocytic molecular outcomes, offering new insights into astrocytic roles in unconsciousness.

1. Introduction

Wake and unconsciousness constitute two fundamental conscious states. Physiologically, two widely recognized conditions that reversibly suppress consciousness are normal sleep and general anesthesia, both of which induce a transient loss of awareness [1]. Administered to millions of surgical patients each year, general anesthesia shares many neurophysiological features with normal sleep [2,3]. For example, an increase in slow-delta (0.1–4 Hz) oscillations, typically observed during sleep onset, has also been reported under anesthesia induced by propofol, sevoflurane, and dexmedetomidine [1,4]. Likewise, functional neuroimaging has shown obvious parallels (but also some differences) between the anesthetized brain and the brain during deep non-rapid eye movement (NREM) sleep [5,6,7]. By selectively suppressing neural activity, anesthesia provides a pharmacologically controlled means to study unconsciousness [3]. Despite these parallels, there is growing interest in whether general anesthesia and normal sleep share common neural pathways in driving transitions of consciousness. Addressing this question could not only advance our understanding of how consciousness is regulated but also help identify novel neural targets for developing sleep-mimetic anesthetics.

While changes in neuronal activity have traditionally been considered the primary drivers of consciousness transitions [8,9,10], emerging evidence highlights astrocytes as active, state-dependent regulators of brain state [11,12]. For example, during slow-wave sleep, astrocytes exhibit rhythmic volume changes that facilitate glymphatic clearance of metabolic waste [13]. Their calcium signaling dynamics also vary substantially across sleep–wake cycles [14]. Our previous work revealed that general anesthetics alter astrocytic morphology and calcium activity, while astrocytic modulation can influence consciousness transitions [15]. These findings position astrocytes not merely as supportive cells, but as essential contributors to state-dependent cortical functions.

Despite growing recognition of the role of astrocytes in regulating consciousness states, the underlying molecular mechanisms remain unclear. In contrast to conventional studies focusing on single molecules or phenomenological observations, proteomic and phosphoproteomic analyses can systematically map molecular pathways activated under specific conditions [16]. Research has established that sleep can trigger widespread, physiologically significant alterations in phosphorylation patterns throughout the brain. [17,18]. Prior work indicates that elevated phosphorylation of the K⁺/Cl⁻ cotransporter 2 (KCC2) in the ventral posteromedial nucleus (VPM) promotes recovery from anesthesia [19,20]. Our previous study suggests that the phosphorylation/dephosphorylation cycle is an important Ezrin-mediated morphological plasticity mechanism of astrocytes during sevoflurane- or propofol-induced consciousness regulation [15]. These findings highlight the role of phosphorylation changes in mediating the effects of general anesthesia. However, there is a lack of systematic comparison of astrocytic molecular responses across wake, sleep, and general anesthesia.

Thus, this study aims to comprehensively characterize the proteome and phosphoproteome of cerebral cortical astrocytes across wake, sleep, and general anesthesia. Sevoflurane, a first-line inhaled anesthetic widely used in clinical perioperative practice, was chosen as the anesthetic in this study, which can induce reversible unconsciousness with well-defined sleep-mimicking neurophysiological features [1,4]. As the supreme processing unit for consciousness, the cerebral cortex exhibits significant functional variations across different states of consciousness [15,21,22]. By providing a comprehensive molecular profile in the cerebral cortex, we seek to elucidate the functional plasticity of astrocytes during transitions of consciousness and uncover shared or distinct astrocytic mechanisms in sleep and anesthesia.

2. Results

2.1. Quality Control of Global and Phosphoproteomic Data in Rat Cortical Astrocytes Across Wake, Sleep, and Anesthesia

In our initial attempts, we found that astrocytes isolated from adult rats showed poor cell viability, which made protein extraction particularly challenging. To overcome this limitation, we switched to using two-week-old Sprague-Dawley (SD) rats. Cells obtained from juvenile brains at this stage typically exhibit higher viability, lealding to more efficient protein extraction and ultimately enhancing the reliability of detecting state-dependent molecular changes. Two-week-old male SD littermate rats were randomly divided into three groups with 3 biological replicates in each group: the wake (WAKE) group, normal sleep (SLEEP) group, and sevoflurane anesthesia (SEVO) group. Prior to the experiment, the rats were housed under LD conditions with lights on at 7:00 and off at 19:00, and were allowed free activity and sleep. Rats in the SLEEP group were sampled between 10:00 and 14:00 (mid-light phase), and were placed in a quiet induction chamber to maintain normal sleep for 2 h before sample collection. Rats in the SEVO group were exposed to 2.5% sevoflurane (carried by 30% O₂/70% N₂ at a flow rate of 1.0 L/min) for 2 h to induce a stable anesthetic state. Rats in the WAKE group were kept awake for 2 h until sample collection through non-invasive gentle tactile stimulation (light touching of the limbs and trunk every 10 min), and the operation was standardized to avoid inducing a significant stress response in the animals [23]. The consciousness state of the SLEEP and SEVO groups was confirmed by electroencephalogram (EEG) and electromyogram (EMG) validation before sample collection, to ensure the accuracy of the experimental grouping (Figure S1). After collecting cortical tissues from the rats, primary astrocyte single-cell suspensions were prepared. A portion of the suspension was used for proteomic analysis, while another portion was enriched for phosphor-peptides to perform phosphoproteomic studies (Figure 1A).

To rigorously evaluate data quality across proteomic and phosphoproteomic datasets from rat cortical astrocytes under wake, sleep, and sevoflurane anesthesia states, we implemented the principal component analysis (PCA) and Pearson correlation coefficient (PCC) assessments [24,25]. For global proteomics, PCA demonstrated distinct experimental group clustering along the first two principal components (PC1: 49.1%; PC2: 22.8%), collectively explaining 71.9% of total variance with minimal intra-group dispersion among biological replicates, indicating a high degree of technical reproducibility. This was corroborated by PCC analysis revealing near-perfect intra-group correlations (PCC ≥ 0.99) and consistently strong inter-group relationships (PCC ≥ 0.98), suggesting a high level of biological comparability across physiological states (Figure 1B,C). Parallel phosphoproteomic analysis similarly showed clear group separation along PC1 (38.3% variance) and PC2 (21.23% variance), accounting for 59.53% cumulative variance despite the inherent dynamicity of phosphorylation modifications. Corresponding PCC values maintained robust intra-group consistency (PCC ≥ 0.95) with preserved inter-group patterns (PCC ≥ 0.93) (Figure 1B,C). Collectively, these quality control metrics indicate strong technical reliability and support the biological coherence of the datasets for subsequent cross-state comparisons.

2.2. Molecular Divergence of Astrocytic Proteomes Across Consciousness States

To characterize differential protein expression across conscious states in rat cortical astrocytes, we calculated the fold change (FC) as the ratio of mean relative protein abundance between comparative groups (Supplementary Table S1). Significantly altered proteins were identified using thresholds of p < 0.05 and FC > 1.5 for upregulation or FC < 1/1.5 for downregulation. Global proteomic profiling revealed distinct state-specific signatures. The most dramatic changes were observed in the SLEEP group compared to the WAKE group (65 upregulated, 183 downregulated), suggesting extensive molecular remodeling during sleep. Significant differences also characterized the transition from wake to sevoflurane anesthesia (103 upregulated, 121 downregulated), consistent with active remodeling during pharmacologically induced unconsciousness. Crucially, despite both representing unconscious states, the SEVO group versus the SLEEP group still showed a clear divergence (34 upregulated, 39 downregulated), pointing to fundamental molecular distinctions between physiological sleep and anesthesia (Figure 2A). Collectively, wake, sleep, and anesthesia each exhibit unique astrocytic proteomic signatures, which may reflect precise molecular orchestration underlying brain state physiology.

Detailed intergroup differentially expressed proteins are listed in Supplementary Table S2. Analysis of the top differentially expressed proteins through pairwise comparisons among the WAKE, SLEEP and SEVO groups revealed both state-transition-specific characteristics and significant commonalities. Compared to the WAKE group, the reduced-arousal states (SLEEP and SEVO group) shared prominent features, including consistent downregulation of keratin, type II cytoskeletal 1 (KRT1) and keratin, type I cytoskeletal 10 (KRT10) (Figure 2B,C). KRT1 and KRT10 are structural keratins involved in cytoskeletal integrity; their downregulation may reflect a shared morphological remodeling during unconsciousness.

The selection of the top 10 changed proteins based on FC thresholds and p-value highlights key players in state transitions. Compared to the WAKE group, the SLEEP group involved downregulation of folate receptor alpha (FRalpha, a folate transporter linked to metabolic processes), fatty acid binding protein 4 (FABP4, involved in lipid metabolism), and keratin 84 (KRT84, another structural protein), along with exclusive upregulation of mast cell carboxypeptidase A (MC-CPA, a mast cell-specific protease implicated in peptide processing and immune regulation), complexin 3 (CPLX3, a synaptic regulator), mast cell protease 1 (MCP1, associated with immune responses), fatty acid 2-hydroxylase (FA2H, involved in lipid modification) and cyclin-dependent kinase 1 (CDK1, a cell cycle regulator) (Figure 2B).

Specific features of the SEVO group included selective downregulation of hemoglobin alpha, adult chain 3 (HBA-A3, possibly reflecting altered oxygen handling), carbonic anhydrase 1 (CA1, involved in pH balance and CO₂ transport), and anion exchange protein (AE1, a bicarbonate transporter) compared to both wake and sleep, indicating a potential association with anesthetic-induced molecular mechanisms (Figure 2C,D). Unique anesthesia versus wake changes, such as upregulation of syndecan (SDC, a heparan sulfate proteoglycan involved in cell adhesion and signaling) and transient receptor potential cation channel subfamily M member 3 (TRPM3, a calcium-permeable channel linked to sensory perception) (Figure 2C).

Most importantly, the SEVO group versus the SLEEP group comparison uniquely revealed the difference between the two unconsciousness states. The SEVO group was associated with downregulation of ribosyldihydronicotinamide dehydrogenase (NQO2, an enzyme involved in quinone metabolism and oxidative stress response) and complement factor properdin (CFP, a regulator of innate immunity), and upregulation of desmoplakin (DSP, a desmosomal protein for cell–cell adhesion), fatty acid binding protein 4 (FABP4, also altered in sleep, possibly indicating divergent lipid dynamics), proenkephalin-A (PENK-A, an opioid precursor implicated in pain modulation), fibulin-5 (FBLN5, an extracellular matrix protein associated with tissue remodeling) and TATA box-binding protein-like 1 (TBPL1, a transcription factor regulator) (Figure 2D). These alterations highlight potential functional differences between the drug-induced state of anesthesia and normal sleep, despite their shared reduction in consciousness. Collectively, the proteomic comparisons reveal that each conscious state possesses a unique molecular signature. Most importantly, the direct comparison between the two unconscious states (anesthesia vs. sleep) provides evidence that they are molecularly distinct, moving beyond their shared behavioral phenotype of unresponsiveness.

2.3. Proteomic Clustering Reveals Divergent Functional Modules Across Consciousness States

For biological process, proteomic clustering analysis suggested that astrocytes in SLEEP group relative to WAKE group were characterized by a predominant downregulation of processes related to extracellular matrix homeostasis and development (e.g., basement membrane organization, cilium assembly; p < 0.001), while showing upregulation of pathways potentially involved in immune regulation and synaptic suppression, like negative regulation of soluble NSF attachment protein receptor (SNARE) complex assembly (p < 0.001). Similarly, the SEVO group versus the WAKE group showed a pattern of downregulated biological processes but diverged from sleep by the suppression of distinct pathways such as negative regulation of inflammatory/adaptive immune responses and bicarbonate transport (p < 0.001), alongside unique upregulation of functions related to neuromodulation (GABA secretion/transport, arginine metabolism; p < 0.001). Direct clustering comparison between the SEVO group and SLEEP group identified anesthesia-specific enrichment of proteins related to cellular differentiation and neural plasticity pathways, suggesting potential activation of these programs (e.g., keratinocyte differentiation, positive regulation of collateral sprouting; p < 0.001), with further suppression of bicarbonate transport (Figure 3A).

Shifting to cellular components, clustering indicated that both the SLEEP group and SEVO group downregulated expression of proteins associated with extracellular/cell-surface elements relative to the WAKE group (e.g., basement membrane, collagen-containing extracellular matrix; p < 0.0001), while upregulating expression of proteins related to membrane components (intrinsic component of membrane and integral component of membrane; p < 0.001). Direct SEVO/SLEEP clustering comparison revealed anesthesia-specific downregulation of cytoplasmic vesicle lumen (p < 0.001) and significant upregulation of extracellular matrix components (p < 0.01) (Figure 3B).

Further analysis of molecular functions via clustering showed that the SLEEP group was associated with downregulated expression of proteins with structural/adhesion-related activities (extracellular matrix structural constituents, glycolipid binding; p < 0.0001) but an elevation in monooxygenase activity (p < 0.0001) relative to the WAKE group. The SEVO group similarly showed downregulated expression of proteins associated with signaling or transport functions (signaling receptor activity, xenobiotic transmembrane transporter activity; p < 0.001), yet distinctly indicated upregulated expression of proteins related to endocrine pathways (hormone activity, receptor ligand activity; p < 0.001). Compared to the SLEEP group, the SEVO group was associated with suppression of amyloid-beta binding (p < 0.01) with concurrent elevation of proteins associated with hormone activity and extracellular matrix structural constituent functions (p < 0.01) (Figure 3C). In summary, GO clustering demonstrates that sleep and anesthesia engage divergent functional modules. While both states share a common pattern of structural remodeling relative to WAKE group, SEVO group is uniquely associated with pathways involving cellular differentiation and neural plasticity, suggesting a divergent underlying biology from sleep. Detailed GO annotation and enrichment-based clustering of proteome changes are listed in Supplementary Table S3.

2.4. Comprehensive Phosphoproteomic Analysis Reveals Distinct and Dynamic Signatures Across Consciousness States

Analysis of phosphoproteomic changes identified distinct signatures across the WAKE group, SLEEP group, and SEVO group (Supplementary Table S4). Pairwise comparisons demonstrated widespread phosphorylation alterations, with differentially phosphorylated sites consistently exceeding differentially regulated proteins in all comparisons, consistent with multi-site phosphorylation dynamics. The SLEEP group versus the WAKE group showed moderate changes (162 upregulated/34 downregulated proteins; 220 upregulated/36 downregulated phosphorylation sites). The SEVO group versus the WAKE group exhibited the most extensive alterations (223 upregulated/40 downregulated proteins; 311 upregulated/48 downregulated phosphorylation sites), while the SLEEP group versus the SEVO group showed intermediate changes (103 upregulated/41 downregulated proteins; 122 upregulated/47 downregulated phosphorylation sites) (Figure 4A). Upregulated phosphorylation dominated all comparisons, suggesting a general bias toward increased phosphorylation during state transitions.

State-specific signatures included Ser188 in nuclear casein kinase and cyclin-dependent kinase substrate 1 (NUCKS1), which were downregulated in both the SEVO-group-versus-WAKE-group (Figure 4C) and the SEVO-group-versus-SLEEP-group comparisons (Figure 4D). This convergent downregulation suggests it may represent a core anesthesia-suppressed event, potentially reflecting inhibited nuclear processes during pharmacologically induced unconsciousness.

The selection of these top 10 altered phosphosites, based on FC and statistical significance, highlights the most prominent molecular changes underlying state transitions. Compared to the WAKE group, the SLEEP group downregulated sites of Ser55 in vimentin (VIM, a key cytoskeletal component involved in maintaining cellular architecture), Ser86 in neuromodulin (GAP43, a synaptic plasticity-associated protein), Ser10 in DnaJ homolog subfamily C member 5 (DNAJC5, implicated in synaptic vesicle cycling), Ser32 in zinc finger CCCH-type containing 18 (ZC3H18, an RNA-binding protein), and Ser212 in serine/arginine-rich splicing factor 2 (SRSF2, a splicing regulator), alongside upregulated sites of Ser495 in F-BAR domain only protein 2 (FCHO2, participating in membrane curvature and endocytosis), Ser154 in eukaryotic translation initiation factor 5B (EIF5B, a translation initiation factor), Ser1786 in microtubule-associated protein 1B (MAP1B, involved in microtubule stabilization and neuronal development), Ser20 in natural resistance-associated macrophage protein 1 (NRAMP1, a metal ion transporter with immune roles), and Ser767 in syntaxin-binding protein 5 (STXBP5, regulating vesicle docking and fusion) (Figure 4B).

Compared to the WAKE group, the SEVO group showed downregulated sites in Ser1176 in Rho GTPase-activating protein 35 (ARHGAP35, a regulator of Rho GTPase signaling and cytoskeletal organization), Ser219 in ABI gene family member 3 (ABI3, involved in cell migration and signaling), Ser2125 in E3 SUMO-protein ligase RanBP2 (RANBP2, a component of the nuclear pore complex involved in SUMOylation), and Ser43 in armadillo repeat-containing protein 10 (ARMC10, with less characterized function); with upregulated sites: Thr406 in supervillin (SVIL, an actin-binding protein regulating cytoskeleton-membrane interactions), Ser826 in apoptotic chromatin condensation inducer in the nucleus (ACIN1, participating in RNA processing and apoptosis), Ser2184 in spectrin beta chain (SPTBN1, a cytoskeletal protein crucial for membrane integrity), Ser324 in ras-interacting protein 1 (RASIP1, involved in vascular signaling), and Ser123 in ceramide transfer protein (CERT, a key mediator of sphingolipid metabolism) (Figure 4C).

Compared to SLEEP group, SEVO group revealed downregulated sites in Ser1385 in kinesin-like protein KIF1A (KIF1A, a neuronal microtubule-based motor protein involved in axonal transport), Ser499 in la ribonucleoprotein 1 (LARP1, an mRNA translation regulator), Ser49 in histone acetyltransferase (KAT7, an epigenetic modifier), Ser994 in non-specific serine/threonine protein kinase (BMP2K, a signaling kinase), and upregulated sites in Ser1505 in SWI/SNF related, matrix associated, actin dependent regulator of chromatin, subfamily a, member 2 (SMARCA2, a chromatin remodeling complex subunit), Ser171 in brain abundant membrane-attached signal protein 1 (BASP1, implicated in neuronal plasticity), Ser1338 in DENN domain-containing 4C (DENND4C, a Rab GTPase activator involved in membrane trafficking), Ser20 in E3 ubiquitin-protein ligase CHIP (CHIP, involved in protein quality control and degradation), and Ser1285 in serine/arginine repetitive matrix 2 (SRRM2, a spliceosome component) (Figure 4D). Therefore, the phosphoproteome not only confirms state-specific regulation but also indicates that anesthesia induces the most extensive molecular remodeling of astrocytes, which may underlie its profound functional effects. Detailed intergroup phosphoproteomic changes are listed in Supplementary Table S5.

2.5. Phosphoproteome Clustering Reveals Distinct Molecular Landscapes in Wake, Sleep, and Anesthesia

Phosphoproteomic analysis identified state-selective phosphorylation patterns across the WAKE group, SLEEP group, and SEVO group, suggesting potential shifts in biological processes. The SLEEP group was associated with upregulated phosphorylation of proteins involved in mRNA regulatory functions, including mRNA processing, RNA splicing, and mRNA metabolic processes (p < 0.001). The WAKE group showed suppressed phosphorylation of proteins related to DNA-templated transcription elongation, cholesterol transport, and chemical synaptic transmission (p < 0.01). The SEVO group exhibited a unique profile: phosphorylation of proteins involved in nuclear mRNA export was increased (p < 0.001), but phosphorylation of proteins associated with transcription-related activities like DNA-templated transcription and protein depolymerization was broadly reduced, suggesting potential inhibition of these processes (p < 0.01). Compared to the SLEEP group, the SEVO group further amplified this divergence, with synaptic transmission-related pathways showing elevated phosphorylation of associated proteins, alongside suppressed phosphorylation of proteins involved in developmental processes such as metanephric mesenchyme formation, suggesting potential inhibition of these developmental programs (both p < 0.001) (Figure 5A).

Cellular components clustering indicated that, relative to the WAKE group, the SLEEP group was characterized by concentrated phosphorylation within nuclear subdomains, including nuclear speckles, the nuclear matrix, and nuclear bodies (p < 0.001), concurrent with a depletion in peroxisomal compartments. The SEVO group, conversely, showed phosphorylation enrichment toward dendritic structures like neuron spines (p < 0.001) away from non-membrane-bounded organelles. The SEVO group versus the SLEEP group was associated with a redistribution of phosphorylation foci toward synaptic elements (postsynapses and synapses) and nuclear architecture, contrasting with diminished phosphorylation in axonal components (p < 0.001) (Figure 5B).

Molecular functional changes further differentiated each state’s signature. The SLEEP group exhibited augmented phosphorylation in receptor binding (beta-2 adrenergic receptor), structural integrity (cytoskeleton, actin filaments), and specialized binding (SUMO, syntaxin-1; p < 0.01), while dampening lipid binding relative to the WAKE group. The SEVO group showed promoted phosphorylation of proteins related to tubulin binding and protein folding (p < 0.01) but reduced phosphorylation of proteins associated with protein kinase C binding, suggesting potential inhibition of related functions. The SEVO group versus the SLEEP group comparison showed simultaneously upregulated phosphorylation of proteins associated with the structural constituent of cytoskeleton function (p < 0.001) and downregulated phosphorylation of proteins related to calcium-dependent protein kinase activity, suggesting potential suppression of this kinase activity (p < 0.001) (Figure 5C). These findings paint a comprehensive picture of state-dependent astrocytic phosphorylation signaling: sleep is associated with enhanced phosphorylation of proteins involved in nuclear mRNA processes, whereas anesthesia exhibits a distinct phosphorylation signature involving synaptic signaling-related proteins, spatial redistribution of phosphorylation events, and altered kinase-related protein phosphorylation, suggesting a profound functional divergence at the post-translational level. Detailed GO annotation and enrichment-based clustering of phosphoproteome changes are listed in Supplementary Table S6.

3. Discussion

This study presents the first integrated proteomic and phosphoproteomic atlas of cortical astrocytes across wake, sleep, and sevoflurane anesthesia. We demonstrate that although both sleep and anesthesia share features in morphological remodeling, anesthesia and sleep are fundamentally distinct in functional pathways, phospho-signaling networks, and molecular outcomes, providing new insights into astrocytic involvement in unconscious states.

Our proteomic analysis indicates that the transition to unconsciousness—whether physiological (sleep) or pharmacological (anesthesia)—elicits a common astrocytic response characterized by suggested potential structural simplification and enhanced transport-related functions. This is reflected in the downregulation of structural proteins such as keratins, consistent with previous reports of cytoskeletal alterations in astrocytes under sleep deprivation [26], and with our earlier work showing that sevoflurane modulates astrocytic morphology via the cytoskeletal regulator Ezrin [15]. These changes align with the known retraction of perisynaptic astrocytic processes during sleep [14,27], suggesting a potential generalized reduction in structural complexity during states of low arousal, which may reflect attenuated interaction with the extracellular milieu or an internal reorganization. Existing studies have shown that cytoskeletal alterations, including keratins, are closely associated with neurocognitive outcomes [28,29,30]. As anesthesia may trigger such cytoskeletal changes during consciousness transitions, exploring the temporal dynamics of KRT1 and KRT10 under different durations of anesthesia may help reveal their roles in neurocognitive regulation associated with astrocyte morphological remodeling. On the other hand, both states exhibit upregulation of membrane transport complexes, supporting the hypothesis of an enhanced capacity for ion and metabolic homeostasis in these states. This is consistent with previous observations that astrocytic aquaporin-4 (AQP4) expression peaks during sleep, facilitating waste clearance [13,31], and that adenosine release from astrocytes helps regulate sleep homeostasis [32,33]. Taken together, these findings suggest that reduced arousal employs a conserved molecular strategy: attenuating structural complexity while reinforcing fundamental transport and homeostatic functions. The functional inferences of this signature are hypotheses that need to be verified by subsequent functional experiments.

Despite these shared features, direct comparison reveals that anesthesia and sleep are fundamentally distinct. Sevoflurane anesthesia specifically downregulated proteins implicated in immunomodulation (e.g., complement factor properdin) and cellular metabolism (NQO2), contrasting with sleep’s generally immune-supportive role [34] and aligning with the known transient immunosuppressive effects of anesthetics [35]. Conversely, anesthesia upregulated PENK-A, an opioid precursor, and TBPL1, a regulator of transcription and apoptosis. Bioinformatic analysis further confirmed that anesthesia activates pathways related to cellular differentiation and structural plasticity, while impairing processes like bicarbonate transport and β-amyloid binding, the latter being implicated in anesthetic-associated side effects [36,37]. Collectively, these findings illustrate that anesthetic unconsciousness is not a mere mimicry of sleep. Existing studies have confirmed that normal sleep is a homeostatic physiological process, in which astrocytes play a key role in metabolic waste clearance and brain homeostasis maintenance [13,14,38]. In contrast, the proteomic changes induced by sevoflurane anesthesia in astrocytes are characterized by potential perturbation of immune and metabolic functions, which is distinct from the homeostatic molecular remodeling of sleep.

Phosphoproteomic analysis revealed distinct signaling dynamics during anesthesia. A key finding was the downregulation of NUCKS1 (Ser188) in anesthesia compared to both wake and sleep, a nuclear protein involved in transcription and cell cycle regulation [39], suggesting it may serve as a potential molecular signature associated with the anesthetic state. Further comparison identified the distinctions of sevoflurane anesthesia relative to sleep, characterized by: (1) elevated phosphorylation of synaptic signaling-related proteins coupled with reduced phosphorylation of developmental pathway-associated proteins, suggesting potential enhancement of synaptic signaling and possible suppression of developmental programs, which may indicate altered plasticity regulation; (2) redistribution of phosphorylation events from axonal compartments to synaptic and nuclear regions, implying a shift in signaling and transcriptional activity; and (3) functional changes such as increased tubulin/actin binding and protein folding, but reduced protein kinase C activity. These cytoskeletal alterations are consistent with reported effects of anesthetics like propofol on actin phosphorylation in neurons [40], while the decreased calcium-dependent kinase activity aligns with our previous observation of sevoflurane-induced suppression of astrocytic calcium signaling [15]. In conclusion, downregulation of NUCKS1 phosphorylation at Ser188 may be a specific alteration induced by general anesthesia. And general anesthesia is associated with an extensive and unique reconfiguration of the astrocytic phosphoproteome. This signature, characterized by suggested potential aberrant synaptic signaling, possible suppression of developmental programs, spatial redistribution of phosphorylation events, and altered cytoskeletal/kinase-related protein phosphorylation, provides a novel post-translational perspective on the mechanisms of pharmacologically induced unconsciousness.

These findings provide a novel molecular framework for understanding the astrocytic contribution to unconscious states, revealing specific targets for future research. However, this study has several limitations. First, the omics data reveal associations, not causal relationships, necessitating future functional validation. Second, considering that prolonged anesthesia may exert toxic effects on developing rats, we selected a 2 h anesthesia duration. The single-time-point design limits insights into temporal dynamics and subregional heterogeneity. Follow-up studies will verify the mRNA expression levels of key differential proteins (including KRT1, KRT10 and NUCKS1) via qRT-PCR and explore the temporal expression changes in these molecules under different durations of sleep and anesthesia. Third, sevoflurane was chosen for this study because of its widespread clinical use and well-established experimental models, which provided an optimal window for mechanistic exploration. Therefore, our conclusions are primarily confined to this specific agent. While n = 3 per group is standard for exploratory proteomic studies, the statistical power is still limited. Potential false positive risk remains for low-abundance proteins, which will be further verified by expanded sample size and orthogonal experiments in follow-up studies. Most importantly, the exclusive use of astrocytes derived from two-week-old juvenile rats is a major limitation of this study. Astrocytes at this developmental stage are not fully mature, and differ substantially from adult astrocytes in metabolic programs, synaptic regulation functions, and plasticity-related signaling pathways. These developmental differences may lead to distinct responses of astrocytes to consciousness state transitions between juvenile and adult rats, which limits the direct extrapolation of our conclusions to adult physiological and pathological states. Subsequent studies should focus on establishing causal relationships between these identified molecular changes and functional outcomes and validate these mechanisms in mature animal models to assess their translational relevance.

4. Materials and Methods

4.1. Animals and Experimental Grouping

All experimental animals were treated in compliance with the guidelines of the Animal Research Committee at West China Hospital of Sichuan. Two-week-old SD littermate male rats were bought from Dossy Experimental Animals Co., Ltd. (Chengdu, China). The rats were housed under a 12 h light/dark cycle (lights on at 07:00, lights off at 19:00) with free access to food and water, and ad libitum activity and sleep for acclimatization prior to experiments. Rats were randomly allocated to three experimental groups (n = 3 biological replicates per group): wake (WAKE) group, normal sleep (SLEEP) group, and sevoflurane anesthesia (SEVO) group.

4.2. Validation of Consciousness States

For the SLEEP group, rats were placed in a quiet induction chamber and allowed to maintain undisturbed normal sleep for 2 h between 10:00 and 14:00 (mid-light phase) before sample collection. For the SEVO group, stable general anesthesia was induced and maintained for 2 h using 2.5% sevoflurane delivered in a gas mixture of 30% O₂/70% N₂ at a flow rate of 1.0 L/min. For the WAKE group, rats were kept awake for 2 h until sample collection via standardized non-invasive gentle tactile stimulation (light touching of the limbs and trunk every 10 min), a protocol designed to avoid inducing significant stress responses [23]. Throughout the procedure, a heating pad was used to maintain the induction chamber temperature at 36–37 °C. To ensure the accuracy of experimental grouping, EEG and EMG signals were recorded before sample collection to objectively verify their consciousness states. For EEG recordings, stainless steel screw electrodes were stereotaxically implanted into the frontal and parietal cortices of rats. For EMG recordings, paired insulated wire electrodes were implanted into the dorsal neck muscles. The sampling frequency was set as 400 Hz, and EEG/EMG was recorded by the Pinnacle EEG/EMG recording system (Part #8200-SL; Pinnacle Technology, Seattle, DC, USA). Consciousness states were defined based on combined EEG, EMG, and behavioral criteria: (1) normal sleep state: defined by high-amplitude, low-frequency slow-delta oscillations (0.1–4 Hz) on EEG and a marked reduction in EMG activity; (2) sevoflurane anesthesia state: identified by synchronized high-amplitude slow-delta EEG activity, complete suppression of EMG activity, and the absence of the righting reflex.

4.3. Tissue Dissociation and Astrocyte Sorting

The cortical hemispheres from the control and SEVO group rats at P14 were dissociated following published guidelines [41] with slight modifications. Briefly, the cortex was dissected and digested using gentleMACS Octo Dissociator with Heaters (Miltenyi Biotec, Shanghai, China) for 30 min at 37 °C with 2 mL of enzyme solution (Dulbecco’s minimum essential medium [DMEM], 0.1% collagenase, and 0.01% trypsin). After digestion, the tissue was filtered with a 70 µm mesh to obtain a single-cell suspension. Then, the debris was removed from cell suspension using a debris-removing solution (Miltenyi Biotec).

Astrocytes were separated by a MACS method according to the manufacturer’s protocol (Miltenyi Biotec). Cell suspensions were labeled with superparamagnetic MicroBeads coupled to antibodies specific for the astrocyte marker GLAST (Anti-GLAST, MicroBead Kit, Miltenyi Biotec). Cells were suspended in PBS with 0.5% BSA and the cell suspension was loaded onto an MS Column (Miltenyi Biotec), which was placed in the magnetic field of a MiniMACS Separator (Miltenyi Biotec). The magnetically labeled GLAST-positive cells were retained within the column and eluted as the positively selected cell fraction after removing the column from the magnet. The astrocyte labeling and sorting procedures strictly followed the standardized GLAST-based protocol previously established and validated by our research team [42]. As rigorously demonstrated in our prior work, this GLAST-mediated sorting strategy exhibits high specificity for astrocytes in developing rodents.

4.4. Peptide Sample Preparation

The following experiment was performed with the support of Jingjie PTM Biolabs (Hangzhou, China) Co., Ltd. The re-suspended GLAST-positive cell suspension was grinded with liquid nitrogen into cell powder. After that, four volumes of lysis buffer (1% SDS, 1% protease inhibitor cocktail) were added to the cell powder, followed by sonication for 3 min on ice and centrifugation at 12,000× g at 4 °C for 10 min. The supernatant was collected and added with pre-cooled acetone, precipitated at −20 °C for 2 h. The precipitate was washed 2~3 times with the pre-cooled acetone and then redissolved in 200 mM TEAB and ultrasonically dispersed. Trypsin was added at 1:50 trypsin-to-protein mass ratio for the first digestion overnight. The sample was reduced with 5 mM dithiothreitol for 30 min at 56 °C and alkylated with 11 mM iodoacetamide for 15 min at room temperature in darkness. The peptides were desalted by a Strata X SPE column and incubated with IMAC microsphere suspension with vibration in loading buffer (50% acetonitrile/0.5% acetic acid). To remove the non-specifically adsorbed peptides, the IMAC microspheres were washed with 50% acetonitrile/0.5% acetic acid and 30% acetonitrile/0.1% trifluoroacetic acid, sequentially. The elution buffer containing 10% NH₄OH was added to elute the enriched phosphopeptides. The supernatant was collected and lyophilized for LC-MS/MS analysis.

4.5. LC-MS/MS Analysis and Database Search

The peptides were dissolved in solvent A (0.1% formic acid, 2% acetonitrile/in water) and directly loaded onto a reversed-phase analytical column (25 cm length, 100 µm i.d.). The mobile phase consisted of solvent A and solvent B (0.1% formic acid in acetonitrile). Peptides were separated with the following gradient: 0–46 min, 2–22%B; 46–52 min, 22–32%B; 52–56 min, 32–90%B; 56–60 min, 90%B, and all at a constant flow rate of 450 nL/minon a NanoElute UHPLC system (Bruker Daltonics, Billerica, MA, USA). The peptides were subjected to a capillary source followed by the timsTOF Pro mass spectrometry. The electrospray voltage applied was 1.7 kV. Precursors and fragments were analyzed at the TOF detector, with a MS/MS scan range from 100 to 1700. The timsTOF Pro was operated in parallel accumulation serial fragmentation (PASEF) mode. Precursors with charge states 0–5 were selected for fragmentation, and 10PASEF-MS/MS scans were acquired per cycle. The dynamic exclusion was set to 24 s.

The resulting MS/MS data were processed using the MaxQuant search engine (v1.6.15.0). Tandem mass spectra were searched against Rattus_norvegicus_10116_PR_20230103.fasta (47,945 entries) concatenated with reverse decoy and contaminants databases. Trypsin/P was specified as the cleavage enzyme, allowing up to 2 missing cleavages. The minimum peptide length was set as 7, and the maximal number of modifications per peptide was set as 5. The mass tolerance for precursor ions was set as 20 ppm in the first search and 20 ppm in the main search, and the mass tolerance for fragment ions was set as 20 ppm. Carbamidomethyl on Cys was specified as a fixed modification. Acetylation on protein N-terminal, oxidation on Met, and phosphorylation (S/T) were specified as variable modifications. The false discovery rate (FDR) of proteins and peptides was adjusted to <1%.

4.6. Repeatability Analysis

For biological or technical replicate samples, it is necessary to examine whether the quantitative results of these replicates conform to statistical consistency. Two statistical methods were employed here to evaluate repeatability: Pearson’s correlation coefficient (PCC) and principal component analysis (PCA).

Pearson’s correlation coefficient was calculated pairwise based on the intensity values of all samples, and a visualized heatmap was generated. This coefficient measures the degree of correlation between two sets of data. A redder color indicates that the Pearson correlation coefficient is closer to 1, representing a stronger correlation between the two samples. Principal component analysis was performed based on the relative quantitative values of all samples, and a visualized PCA plot was generated. The horizontal and vertical axes display the explained variance of PC1 and PC2, respectively; higher values indicate greater explanatory power. The degree of intra-group clustering reflects the quality of replicate samples within each group—replicate samples within the same group tend to cluster together.

4.7. GO Annotation and Enrichment-Based Clustering

The Gene Ontology (GO) database was used to annotate the biological functions of proteins. Fisher’s exact test was used to analyze the significance of functional enrichment of differentially expressed proteins (using the identified protein as the background). Functional terms with fold enrichment > 1.5 and p-value < 0.05 were considered to be significant.

For further hierarchical clustering based on GO, we first collated all the categories obtained after enrichment along with their p-values, and then filtered for those categories that were at least enriched in one of the clusters with p-value < 0.05. This filtered p-value matrix was transformed by the function x = −log10 (p-value). These p-values were then clustered by one-way hierarchical clustering (Euclidean distance, average linkage clustering) in Genesis. Cluster membership was visualized by a heat map using the “Heatmap” function from the “ComplexHeatmap” R package (version 2.14.0)

5. Conclusions

Though both reversible unconscious states—downregulated structural proteins and upregulated membrane transport complexes—induce shared astrocytic responses, reflecting common morphological remodeling and homeostatic enhancement, they differ fundamentally in functional pathways, phospho-signaling networks, and molecular outcomes. Phosphoproteomic data show the most extensive signaling reprogramming under anesthesia, with anesthesia-specific decreased NUCKS1 phosphorylation at Ser188, implying it may serve as a potential molecular signature associated with the anesthetic state, and this hypothesis needs to be verified by subsequent functional experiments. Anesthesia also shows phosphorylation signatures, suggesting potential enhancement of synaptic signaling, possible suppression of development-related programs, and altered subcellular distribution of phosphorylation events, further confirming molecular differences from sleep. Although the findings need to be further verified in adult animal models to clarify their translational relevance, this work provides a novel molecular framework for understanding astrocytes’ role in consciousness transitions and potential targets for exploring anesthesia and sleep mechanisms.