Graphical abstract graphic file with name fx1.jpg [35]Open in a new tab Highlights * • Chronic sleep deprivation impairs hippocampal function and memory in tauopathy models * • tNIR therapy restores hippocampal blood flow, metabolism, and ion homeostasis * • tNIR reorganizes sleep architecture, enhancing slow-wave sleep and neural recovery * • tNIR improves place cell coding, hippocampal synchronization, and neural oscillations __________________________________________________________________ Natural sciences; Biological sciences; Neuroscience; Systems neuroscience Introduction Alzheimer’s disease (AD) is a progressive neurodegenerative disorder characterized by the pathological accumulation of amyloid-beta (Aβ) plaques and neurofibrillary tangles composed of hyperphosphorylated tau, leading to synaptic dysfunction, neuronal loss and cognitive decline.[36]^1 Among the diverse factors contributing to AD pathology, sleep disturbances have garnered significant attention for their dual role as modifiable risk factors and early biomarkers.[37]^2^,[38]^3 These disturbances, which encompass fragmented sleep, altered architecture, and altered sleep-wake cycles, form a bidirectional relationship with AD: they exacerbate neurodegeneration while being progressively worsened by it.[39]^4 Understanding this interplay offers opportunities for targeted therapeutic interventions. Chronic sleep deprivation (SD), a controlled model of persistent sleep disturbance, intensifies AD pathology by accelerating memory impairments and disrupting systemic homeostasis.[40]^5^,[41]^6 Specifically, SD destabilizes metabolic and ionic homeostasis, notably depleting plasma magnesium and calcium levels, which further impairs sleep-wake cycles and rest-activity rhythms.[42]^7^,[43]^8 In tauopathy models, such as P301S mice, SD exacerbates synaptic dysfunction and cognitive deficits, highlighting the pivotal role of sleep in AD progression.[44]^9 Despite its inherent complexity, chronic SD provides a robust framework for investigating how aggravated sleep loss drives hippocampal dysfunction and perpetuates the pathological feedback loop between sleep and cognition. Sleep is essential for memory consolidation, with non-REM (NREM) sleep facilitating hippocampal sharp-wave ripples (SWRs) that stabilize long-term memories, and REM sleep supporting memory integration and emotional regulation through theta and gamma oscillations.[45]^10^,[46]^11^,[47]^12 Awake SWRs are equally critical, governing working memory, decision-making, and spatial planning. Chronic SD disrupts these processes, impairing both awake and sleep SWRs and leading to profound hippocampal-dependent cognitive deficits.[48]^13^,[49]^14^,[50]^15 Hippocampal subregions exhibit differential vulnerability to SD. The CA3 region, pivotal for pattern completion and SWR reactivation, is particularly sensitive to ripple disruptions.[51]^16^,[52]^17 In contrast, the CA1 region, which integrates hippocampal-cortical communication through theta-gamma coupling, experiences the desynchronization of oscillatory networks, further impairing episodic and spatial memory consolidation.[53]^18^,[54]^19 These disruptions underscore the need for therapeutic strategies targeting hippocampal network recovery. Transcranial near-infrared (tNIR) therapy has emerged as a promising non-invasive intervention for neurodegenerative conditions. By delivering light at specific wavelengths, tNIR enhances mitochondrial function, promotes neuroprotection, and improves cognitive outcomes.[55]^20 Light wavelengths in the ranges of 800–820 nm and 1060–1080 nm, at energy densities such as 535 mW/cm^2, have demonstrated efficacy in improving cognitive function in mild-to-moderate AD.[56]^21 Mitochondrial cytochrome c oxidase (CCO) acts as a primary photoacceptor, absorbing NIR light, particularly within the 800–850 nm range, and converting it into biological signals that enhance mitochondrial activity and cellular metabolism.[57]^22 In acute SD models, 810 nm tNIR therapy reduces hippocampal oxidative stress and ameliorates memory deficits.[58]^23 However, its potential in mitigating chronic SD-induced hippocampal dysfunction and interrupting the pathological sleep-cognition feedback loop in AD remains unexplored. This study investigates the therapeutic efficacy of 808 nm transcranial near-infrared (tNIR) therapy in mitigating chronic sleep deprivation-induced hippocampal dysfunction and cognitive deficits in P301S tauopathy mice. Using advanced techniques, including arterial spin labeling (ASL) imaging, metabolomics, ionomics, and place cell analyses, we hypothesize that enhancing cerebral blood flow (CBF) and optimizing metabolic and ionic balance through tNIR therapy restores hippocampal oscillatory synchronization and reactivates place cell activity in CA1 and CA3. By addressing these mechanisms, this research aims to uncover novel insights into the therapeutic modulation of the sleep-cognition feedback loop in AD. Results Transcranial near-infrared therapy restores cerebral blood flow impaired by chronic sleep deprivation To evaluate the effects of tNIR therapy on CBF disrupted by chronic SD, tauP301S AD mice were divided into control (n = 6), SD (n = 6), and tNIR-treated cohorts (n = 6). Mice in the SD and tNIR groups underwent 6 h of daily sleep deprivation for one month.[59]^24 The tNIR group received targeted irradiation over the hippocampal regions following SD sessions ([60]Figure 1A). CBF was assessed using ASL imaging, with region of interest (ROI) analyses focusing on hippocampal subregions: dentate gyrus (DG), CA1, CA3, and the entire hippocampal area (HP) in both hemispheres ([61]Figure 1B). Quantitative analysis of CBF ([62]Figure 1C; [63]Table S1) demonstrated significant reductions in the SD group (n = 6) compared to controls (n = 6), with the most pronounced deficits in the left dentate gyrus (DG_L, p = 0.0006) and CA3 subregions (CA3_L, p = 0.0414; CA3_R, p = 0.0481). Notably, tNIR therapy (n = 5) significantly increased CBF in these affected regions, particularly within the DG (DG_L, p = 0.0013; DG_R, p = 0.0129), effectively restoring CBF to levels comparable to controls. This robust recovery in hippocampal CBF suggests that tNIR therapy mitigates the vascular dysfunction induced by SD, laying a critical foundation for subsequent improvements in hippocampal metabolomic profiles and neuronal activity.[64]^25 Figure 1. [65]Figure 1 [66]Open in a new tab Cerebral blood perfusion in SD, tNIR treatment, and control groups across hippocampal subregions (A) Schematic diagram of the experimental paradigm, illustrating the daily SD protocol and subsequent tNIR treatment. (B) Representative ASL images displaying CBF in the hippocampal subregions for the SD, tNIR treatment, and control groups. ROI analyses included hippocampal subregions of the DG, CA1,CA3, the whole hippocampus (HP) from both hemispheres. (C) Quantitative comparisons of CBF levels across distinct hippocampal subregions. Statistical significance is indicated as follows: p < 0.05 (∗), p < 0.01 (∗∗), for SD (n = 6) vs. tNIR treatment (n = 5). p < 0.05 (#), p < 0.001 (###), for SD (n = 6) vs. Control (n = 6). p = 0.07 (+) for SD (n = 6) vs. Control (n = 6), indicating a trend toward significance. Data were analyzed using one-way ANOVA followed by Tukey’s HSD post hoc test for multiple comparisons. Sample size: n = 5–6 mice per group. All values are presented as mean ± standard deviation (S.D.). Cerebral blood flow data are provided in [67]Table S1. Transcranial near-infrared therapy modulates metabolic reprogramming to enhance antioxidant defense and mitochondrial efficiency Metabolomic analysis revealed a diverse distribution of metabolites, with amino acids, nucleotides, lipids, and organic acids, with amino acids, peptides, and their analogs comprising the largest proportion (24.7%) ([68]Figure 2A; [69]Table S2). A Venn diagram ([70]Figure 2B) illustrated the broad corrective effects of tNIR therapy, showing the restoration of 58 impaired metabolites and normalization of 57 elevated metabolites in the SD group. Figure 2. [71]Figure 2 [72]Open in a new tab Metabolomic profiling and KEGG pathway analysis across experimental groups (A) Class distribution of detected metabolites across sleep deprivation (SD), tNIR treatment, and control groups (n = 6 mice per group). (B) Venn diagram illustrating the number of metabolites modulated by tNIR treatment following SD (n = 6 mice per group). (C) KEGG pathway analysis of metabolites significantly reduced in SD compared to controls (p < 0.05, n = 6 mice per group). (D) KEGG pathway analysis of metabolites restored by tNIR treatment compared to SD (p < 0.05, n = 6 mice per group). Yellow boxes indicate metabolites decreased in SD that were rescued by tNIR, while the red box highlights upregulation in the mitochondrial energy pathway. (E) KEGG pathway analysis of metabolites significantly elevated in SD compared to controls (p < 0.05, n = 6 mice per group). (F) KEGG pathway analysis of metabolites normalized by tNIR treatment compared to SD (p < 0.05, n = 6 mice per group). The green box highlights a reduction in glycolytic pathway metabolites, indicating a shift in energy supply mechanisms. ∗ indicates p < 0.05, † denotes a trend toward significance. The enrichment significance was calculated using the hypergeometric test and corrected for multiple comparisons using the Benjamini-Hochberg method. Sample size: n = 6 mice per group. Differential metabolite data are provided in [73]Table S2; KEGG pathway enrichment results are available in [74]Table S3. Chronic SD-induced disruptions were observed across multiple metabolic pathways, as highlighted by KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis ([75]Figures 2C and 2E; [76]Table S3). Downregulated pathways ([77]Figure 2C) included biotin metabolism, fatty acid biosynthesis, glycerolipid metabolism, and steroid biosynthesis (p < 0.05), impairing structural integrity and neuroprotective capacity. The suppression of cysteine and methionine metabolism, glycine, serine, and threonine metabolism, and the pentose phosphate pathway weakened antioxidant defenses, DNA repair mechanisms, and neurotransmitter homeostasis, increasing susceptibility to oxidative stress. Impaired glyoxylate and dicarboxylate metabolism further exacerbated metabolic inflexibility and cellular stress by hindering fatty acid-to-glucose conversion and detoxification.[78]^26 Despite these disruptions, SD elicited the adaptive upregulation of several pathways ([79]Figure 2E), including purine and pyrimidine metabolism and glycine, serine, and threonine metabolism, reflecting increased nucleotide synthesis and amino acid demands for cellular repair.[80]^27 Enhanced beta-alanine metabolism, the one-carbon pool by folate, and pentose and glucuronate interconversions indicated heightened DNA repair and antioxidative efforts.[81]^28 However, the upregulation of cysteine and methionine metabolism also underscored the oxidative stress burden imposed by SD.[82]^29 tNIR therapy reprogrammed energy metabolism by shifting reliance from glycolysis to mitochondrial TCA cycle activation ([83]Figures 2D and 2F). This metabolic shift boosted ATP production, enhanced mitochondrial efficiency, and improved antioxidant defenses. Pathways such as alanine, aspartate, and glutamate metabolism, along with histidine and tyrosine metabolism, were upregulated, supporting neurotransmission and cognitive function.[84]^30 Furthermore, enhanced purine, cysteine, and methionine metabolism strengthened cellular defenses against oxidative damage, while elevated arginine biosynthesis and Vitamin B6 metabolism promoted vascular health and immune function.[85]^31^,[86]^32 In addition to restoring deficient pathways, tNIR therapy recalibrated overactivated pathways, including valine, leucine, and isoleucine biosynthesis and pyrimidine metabolism, reducing reliance on glycolytic energy production and enhancing oxidative phosphorylation.[87]^33 This recalibration underscores tNIR’s ability to restore metabolic homeostasis, mitigate oxidative stress, and foster cellular recovery and resilience in the context of SD-induced metabolic dysregulation. Transcranial near-infrared therapy restores metabolic and ionic homeostasis impaired by sleep deprivation to support neuronal excitability To further examine the effects of tNIR therapy on chronic SD-induced metabolic alterations, K-means clustering of metabolites was performed, revealing nine distinct clusters of metabolic changes among SD, tNIR-treated, and control groups ([88]Figure 3A; [89]Table S4). Cluster 2 (n = 24 metabolites) and Cluster 3 (n = 124 metabolites) exhibited the most significant recovery following tNIR intervention. Cluster 2 metabolites, including nucleotide derivatives (AMP, GMP), amino acids (L-Tyrosine, L-Methionine), and antioxidative compounds (L-Carnosine and L-Cystine), were significantly depleted under SD but restored to control levels after tNIR treatment. These metabolites are essential for nucleotide synthesis, protein biosynthesis, enzymatic activity, and antioxidative defenses.[90]^34^,[91]^35 Conversely, Cluster 3 metabolites, such as lactic acid, ribose-5-phosphate, and oxidative stress markers (6-hydroxymelatonin, nicotinamide), were elevated under SD, reflecting glycolytic stress and compensatory antioxidative responses. tNIR treatment normalized these metabolites, shifting energy production from glycolytic pathways to mitochondrial processes and reducing metabolic stress. Figure 3. [92]Figure 3 [93]Open in a new tab K-means clustering of metabolites and ion level changes across experimental groups (A) K-means clustering analysis of metabolites across sleep deprivation (SD), tNIR treatment, and control groups, revealing distinct patterns of metabolite grouping and functional pathways (metabolite counts per cluster are shown in the figure). (B) Quantitative comparison of 31 distinct ion levels among the three groups (n = 6 mice per group). Statistical significance is denoted as follows: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Data were analyzed using one-way ANOVA followed by Tukey’s HSD post hoc test for multiple comparisons. Sample size: n = 6 mice per group. All values are presented as mean ± S.D. Metabolites grouped by cluster are listed in [94]Table S4, and ion level profiles across groups are provided in [95]Table S5. Given that ionic imbalances are closely tied to SD-induced disruptions in neuronal excitability and mitochondrial function,[96]^7^,[97]^8 we conducted ionomic analysis to assess the therapeutic potential of tNIR therapy in restoring ionic homeostasis. The analysis revealed significant disruptions in elemental balance under chronic SD ([98]Figure 3B; [99]Table S5). Elevated potassium (K^+) (p = 0.0049) and sodium (Na^+) levels in the SD group indicated ionic fluxes associated with cellular stress and altered neuronal excitability. Reduced calcium (Ca^2+) (p = 0.026) and iron (Fe^2+) (p = 0.031) levels pointed to deficiencies in cofactors critical for mitochondrial activity and synaptic signaling. tNIR therapy effectively restored Ca^2+ and Fe^2+ levels to near control values, stabilizing ionic homeostasis and supporting synaptic plasticity. These findings highlight the ability of tNIR therapy to counteract SD-induced metabolic and ionic dysregulation. By restoring key metabolites and rebalancing elemental levels, tNIR enhances cellular resilience, supports mitochondrial function, and promotes neuronal excitability, underscoring its therapeutic potential for mitigating the adverse effects of chronic sleep deprivation. Transcranial near-infrared therapy enhances restorative sleep architecture and mitigates the effects of chronic sleep deprivation To investigate how tNIR therapy-induced metabolic and ionic rebalancing impacts hippocampal function, we assessed sleep architecture and neural oscillatory dynamics in hippocampal subregions CA1 and CA3 under chronic SD and following tNIR treatment. Neural activity was recorded using tetrode electrodes ([100]Figure 4A), and local field potentials (LFPs) were analyzed over a 900-s sleep session. Sleep phases were segmented into active wake (AW), rapid eye movement (REM) sleep, and slow-wave sleep (SWS) within non-rapid eye movement (NREM) sleep ([101]Figures 4B and 4C). Chronic SD resulted in a trend toward increased SWS duration in the CA1 region compared to controls, consistent with compensatory mechanisms reported in response to SD-induced disruptions.[102]^36 tNIR therapy significantly enhanced SWS duration relative to both the SD group and controls (p = 0.0285) ([103]Figure 4D). In the CA3 region, SWS duration was significantly elevated in the SD group compared to controls (p = 0.0089). tNIR therapy further reorganized sleep architecture in CA3, markedly increasing SWS duration (p = 0.0005) and significantly reducing REM sleep duration (p = 0.0136) relative to controls ([104]Figure 4D). These results suggest that chronic SD induces compensatory increases in SWS and reductions in REM sleep, potentially reflecting an adaptive response to disrupted hippocampal function. tNIR therapy amplifies these adaptive mechanisms, promoting a pronounced shift toward restorative SWS while reducing REM sleep. This reorganization of sleep architecture highlights the therapeutic potential of tNIR to enhance reparative neural processes and counteract the adverse effects of chronic SD on hippocampal function. Figure 4. [105]Figure 4 [106]Open in a new tab tNIR treatment increases hippocampal SWS duration, enhances power across multiple frequencies, and restores the ripple event ratios impaired by chronic sleep deprivation (A) Schematic representation of multi-channel electrode implantation in the mouse brain. A 32-channel electrode array, divided into two 16-channel bundles, was implanted in the hippocampal CA1 (left) and CA3 (right) regions. The lower panel shows the histological verification of electrode implantation sites. (B) Representative 900-s local field potential (LFP) recording during a sleep period in the mouse home cage. (C) Power spectral density (PSD) analysis of the 900-s sleep phase, with the classification of sleep states: active wake (AW, red), REM sleep (orange), and slow-wave sleep (SWS, blue). (D) Comparison of sleep stage durations (AW, REM, and SWS) within CA1 and CA3 regions across SD, tNIR treatment and control groups (n = 5 mice/group). (E) LFP traces displaying raw and bandpass-filtered signals for delta (0.5–4 Hz), theta (4–12 Hz), low gamma (30–50 Hz), high gamma (50–100 Hz), and ripple (100–250 Hz) frequency bands. (F) Mean PSD comparison across frequency bands in CA1 (n = 5 mice/group). (G) Peak PSD comparison across frequency bands in CA1 (n = 5 mice/group). (H) Mean PSD comparison across frequency bands in CA3 (n = 5 mice/group). (I) Peak PSD comparison across frequency bands in CA3 (n = 5 mice/group). (J) Representative ripple event ratios and amplitudes across experimental groups during the 900-s sleep phase. (K) Quantitative analysis of ripple event ratios in CA1 and CA3 regions, with group comparisons (n = 5 mice/group). Statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 for SD vs. tNIR; #p < 0.05, ##p < 0.01 for SD vs. Control; +p < 0.05 for tNIR vs. Control. Stage Data were analyzed using two-way ANOVA followed by Tukey’s honestly significant difference (HSD) post hoc test for multiple comparisons. n = 5 mice per group. CA1 region: treatment group: F(2, 8) = 3.624, p = 0.0758; sleep stage: F(2, 8) = 86.38, p < 0.0001; interaction: F(4, 16) = 4.335, p = 0.0145. Tukey’s test revealed a significant difference between the tNIR-treated and control groups in the SWS stage (p = 0.0285, mean difference = 64.12, 95% CI = 6.43 to 121.8). CA3 region: Treatment group: F(2, 8) = 0.063, p = 0.9395; Sleep stage: F(2, 8) = 127.7, p < 0.0001; Interaction (group x stage): F(4, 16) = 9.650, p = 0.0004. Post hoc Tukey’s test showed significant differences in REM and SWS stages: REM: tNIR vs. Control (p = 0.0136, mean difference = 60.49, 95% CI = 12.20 to 108.8). SWS: SD vs. Control (p = 0.0089, mean difference = 64.35, 95% CI = 16.06 to 112.6); tNIR vs. Control (p = 0.0005, mean difference = 90.90, 95% CI = 42.60 to 139.2). PSD Data were analyzed using one-way ANOVA followed by Tukey’s HSD post hoc test for multiple comparisons. n = 5 mice per group. Transcranial near-infrared treatment enhances hippocampal oscillatory power in chronic sleep deprivation To evaluate the effects of tNIR therapy on hippocampal oscillatory activity disrupted by chronic SD, LFP recordings from sleep sessions were bandpass-filtered into distinct frequency bands for power spectral analysis ([107]Figure 4E). Significant alterations were observed in both CA1 and CA3 regions, reflecting the effects of SD and subsequent tNIR intervention. In the CA1 region, mean power spectral density (PSD) analysis ([108]Figure 4F) showed that tNIR therapy significantly increased PSD in the delta (p = 0.0375), theta (p = 0.0152), low gamma (p = 0.0128), and high gamma (p = 0.004) frequency bands compared to the SD group. Notably, theta power under tNIR therapy exceeded control levels (p = 0.0218). Peak PSD analysis further revealed significant increases in theta (p = 0.0478), low gamma (p = 0.0079), high gamma (p = 0.0065), and sharp-wave ripple (SWR) (p = 0.0077) frequency bands ([109]Figure 4G), with theta power activity surpassing control values (p = 0.0243). In the CA3 region, SD caused reductions in mean PSD across all frequency bands, with significant deficits in low gamma (p = 0.0079) and high gamma (p = 0.0372) bands ([110]Figure 4H). tNIR therapy restored oscillatory power in CA3, with significant improvements in peak PSD for low gamma (p = 0.0119) and high gamma (p = 0.009) bands compared to the SD group ([111]Figure 4I). These results highlight the efficacy of tNIR therapy in restoring hippocampal oscillatory dynamics disrupted by SD. Enhanced theta, gamma, and ripple activity suggests improved neural synchronization and interregional communication, underscoring the therapeutic potential of tNIR therapy in mitigating SD-induced impairments in hippocampal function. Transcranial near-infrared therapy boosts sharp-wave ripple activity under sleep deprivation Sharp-wave ripples (SWRs) are essential for hippocampal-cortical communication during memory consolidation.[112]^37 To evaluate the baseline readiness of hippocampal circuits for memory encoding, we analyzed ripple events during pre-task sleep sessions ([113]Figures 4J and 4K). Pre-task SWR activity serves as a measure of the preparatory state of neural networks, influencing subsequent task performance and memory-related processes. During the 900-s pre-task sleep session, the SD group showed an upward trend in ripple event ratios compared to controls in both CA1 and CA3 regions. Notably, tNIR therapy further elevated ripple event ratios beyond both SD and control groups, with a significant increase observed in CA3 compared to controls (p = 0.049) ([114]Figures 4J and 4K). The average ripple event ratios between CA1 and CA3 remained consistent across experimental conditions, indicating stable network-level coordination between these subregions ([115]Figure 4K). These findings demonstrate that tNIR therapy enhances ripple-associated neural activity, potentially priming the hippocampus for improved memory encoding and cognitive performance.[116]^38 Transcranial near-infrared therapy restores firing rates, exceeding its role in place cell recruitment during chronic sleep deprivation To evaluate the effects of tNIR therapy on hippocampal functionality disrupted by chronic SD, mice from three experimental groups (SD, tNIR-treated, and control; n = 5 mice/group) were tested using two spatial exploration paradigms: an open field (OP) arena and a linear track, each designed with distinct environmental cues. In the OP paradigm, mice completed three sessions: initial exploration (OP01), exploration following a 90° counterclockwise arena rotation with cue modification (OP02), and a final session with a clockwise 90° rotation restoring the original cues (OP03) ([117]Figure 5A). The linear track paradigm consisted of three phases: exploration of a white-background track (Track01), a black-background track with altered cues (Track02), and a return to the original white track (Track03) ([118]Figure 5B). Single-neuron spike waveforms recorded via tetrodes (Tr1–Tr4) were clustered based on spike width and firing rate to differentiate pyramidal neurons from interneurons ([119]Figures 5C and 5D). SD disrupted the spatial tuning of place cells in both paradigms, resulting in diffuse and disorganized firing patterns. tNIR therapy restored spatial organization, as reflected in spatial distribution and firing rate heatmaps ([120]Figures 5E and 5F). Figure 5. [121]Figure 5 [122]Open in a new tab Restoration of place cell (PC) function by tNIR treatment during spatial learning tasks (A) Open field paradigm: Mice explored a 60 × 60 × 40 cm arena with distinct visual cues. The experiment included three 15-min sessions: OP01 (baseline exploration), OP02 (cue modification with 90°arena rotation), and OP03 (reset to the original configuration), each separated by a 5-min rest. (B) Track paradigm: Mice explored a 1.5 m × 8 cm linear track. Sessions included Track 01 (white background with specific cues), Track 02 (black background with altered cues), and Track 03 (return to the original white background), with 5-min rest intervals. (C) Single-neuron spike waveforms: Spike waveforms from a single neuron recorded by four electrodes in a tetrode (Tr1-Tr4). Pyramidal neuron spikes are shown in orange, interneuron spikes in blue. Vertical scale bar: 100μV. (D) Neuron classification: Clustering analysis based on spike width and firing rate separated neurons into pyramidal neurons (orange) and interneurons (blue). (E) Place cell activity in the open field: Top row shows spatial firing paths overlaid with spike locations; bottom row shows firing rate maps. (F) Place cell activity in the track paradigm: Representative spatial firing paths and firing rate maps. (G) Place cell proportions across three open field sessions for the three experimental groups (n = 5 mice/group for each session). (H) Place cell proportions across three track sessions for the three experimental groups (n = 5 mice/group for each session). (I) Mean firing rates of place cells in the open field paradigm across sessions (n = 5 mice/group for each session). (J) Mean firing rates of place cells in the track paradigm across sessions (n = 5 mice/group for each session). Statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns denotes no significant difference. Data were analyzed using the Kolmogorov-Smirnov (K-S) test. n = 5 mice per group. All data are presented as mean ± S.D. SD significantly impaired hippocampal spatial encoding by reducing place cell recruitment and firing rates across all sessions in both paradigms. In the OP paradigm ([123]Figure 5G), SD reduced place cell proportions across sessions (OP01–OP03), indicating diminished spatial representation and active place cell recruitment. Although tNIR therapy did not fully restore place cell proportions to control levels, it showed a trend toward increased recruitment compared to the SD group, suggesting partial recovery of spatial coding. Similar results were observed in the linear track paradigm ([124]Figure 5H), where SD lowered place cell proportions across all track phases (Track01–Track03). tNIR therapy consistently increased place cell proportions in both CA1 and CA3 regions ([125]Figures S2A and S2B), demonstrating its restorative effects on hippocampal network integrity. Firing rate analysis further revealed SD-induced spatial encoding deficits. SD significantly reduced firing rates in CA1 ([126]Figure 5I; OP01: p < 0.0001, OP03: p < 0.0001) and CA3 ([127]Figure S2C; OP01: p < 0.0001, OP02: p < 0.0001, OP03: p < 0.0001), reflecting disrupted hippocampal activity and weakened spatial tuning. tNIR therapy effectively restored firing rates relative to SD across OP sessions in both CA1 ([128]Figure 5I; OP02: p = 0.0065, OP03: p = 0.0060) and CA3 ([129]Figure S2C; OP01: p = 0.0027, OP02: p = 0.0015, OP03: p = 0.0070), approaching control levels in most cases. Similarly, in the linear track paradigm, SD caused pronounced reductions in firing rates across all track phases in CA1 ([130]Figure 5J; Track01–03: p < 0.0001) and CA3 ([131]Figure S2D; Track01–03: p < 0.0001). tNIR therapy rescued firing rates in CA1 across track phases ([132]Figure 5J), with particularly robust effects in CA3 ([133]Figure S2D; Track01: p < 0.0001, Track02: p = 0.0043, Track03: p < 0.0001). These findings indicate that while tNIR therapy partially restores place cell recruitment, its primary restorative effect lies in enhancing firing rates. This improvement likely surpasses its influence on recruitment by enhancing neuronal excitability and network dynamics, supporting hippocampal functionality under dynamic and spatially demanding conditions. Transcranial near-infrared therapy enhances place cell spatial encoding and precision Chronic SD significantly impaired spatial encoding precision, as evidenced by larger place fields in both the open field (OP) and linear track paradigms for CA1 ([134]Figures 6A and 6D) and CA3 ([135]Figures S3A and S3D). In the OP paradigm, place field sizes in SD mice were markedly larger during the initial session (OP01, 80.33 ± 5.6 cm^2) compared to controls (57.47 ± 11.34 cm^2, p < 0.0001, [136]Figure 6A). These impairments persisted across subsequent sessions, including OP03 (p < 0.0001) and all phases of the linear track paradigm (Track01–Track03, p < 0.0001, [137]Figure 6D). Similar trends were observed in CA3, where SD mice consistently displayed enlarged place fields in both paradigms ([138]Figures S3A and S3D, p < 0.0001), reflecting cumulative deficits in spatial learning and encoding. Figure 6. [139]Figure 6 [140]Open in a new tab tNIR treatment restores spatial encoding properties of place cells and enhances theta and ripple phase-locking during spatial learning tasks (A) Place field size of place cells during open field (OP) spatial learning. (B) Spatial information encoded by place cells during OP learning. (C) Firing sparsity of place cells during OP learning. (D) Place field size of place cells during track exploration. (E) Spatial information encoded by place cells during track exploration. (F) Firing sparsity of place cells during track exploration. (G) Theta phase-locking schematic: Example of phase-locking between place cell firing and theta rhythm (4–12 Hz) during OP03 learning. The representative unit shows firing predominantly locked to the trough of the theta cycle, with high-frequency firing concentrated at this phase. (H) Theta phase preference of place cell firing relative to the theta rhythm in CA1, analyzed from local field potential (LFP) data during OP03 learning. (I) Mean vector length of theta phase-locking in CA1 during OP03, indicating the strength of phase concentration for a cell’s firing. (J) Correlation between place cell firing and the theta rhythm at the preferred phase. (K) Ripple phase-locking schematic: Example of phase-locking between place cell firing and ripple oscillations (100–250 Hz). The representative unit demonstrates firing predominantly during ripple events, indicating strong phase-locking. (L) Ripple phase preference of place cell firing relative to ripple oscillations in CA1 during OP03. (M) Mean vector length of ripple phase-locking in CA1 during OP03, reflecting the concentration of a cell’s firing at its preferred ripple phase. (N) Correlation between place cell firing and ripple oscillations at the preferred phase. Statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns denotes no significant difference. Data were analyzed using the Kolmogorov-Smirnov (K-S) test. n = 5 mice per group. All data are presented as mean ± S.D. tNIR therapy significantly reduced place field sizes in both CA1 and CA3, restoring spatial encoding precision impaired by SD. In OP01, tNIR-treated mice displayed smaller place fields (66.79 ± 13.53 cm^2, p < 0.0001) compared to SD mice, with values approaching those of controls ([141]Figure 6A). This improvement persisted across OP02 and OP03 in CA1 ([142]Figure 6A) and throughout Track01–Track03 in the linear track paradigm ([143]Figure 6D). Similarly, tNIR-treated mice in CA3 exhibited reduced place field sizes compared to SD mice, with results trending toward control levels in both paradigms ([144]Figures S3A and S3D, p < 0.0001). These findings highlight the ability of tNIR therapy to mitigate SD-induced spatial encoding deficits and promote sustained adaptation in hippocampal function. By reducing place field sizes and enhancing spatial precision, tNIR therapy reinforces the hippocampal capacity for precise spatial representation, critical for learning and memory performance over repeated tasks. Transcranial near-infrared therapy restores spatial information content Spatial information content, a key measure of hippocampal spatial coding efficiency,[145]^39 was significantly reduced in SD mice across both the OP and linear track paradigms. In OP01, SD mice exhibited lower spatial information (1.54 ± 0.12 bits/spike) compared to controls (1.97 ± 0.16 bits/spike, p < 0.001) and tNIR-treated mice (1.82 ± 0.26 bits/spike, p < 0.001) ([146]Figure 6B). This deficit persisted through subsequent sessions, with SD mice maintaining reduced spatial information content in OP03 ([147]Figure 6B) and all linear track sessions ([148]Figure 6E). tNIR therapy significantly restored spatial information content impaired by SD. In OP01, tNIR-treated mice showed higher spatial information compared to SD mice (p < 0.001) and maintained these improvements in OP03 (2.1 ± 0.5 bits/spike, p < 0.001), with values approaching controls ([149]Figure 6B). Similar enhancements were observed in the linear track paradigm, where tNIR-treated mice exhibited increased spatial information content across Track01–Track03 compared to SD mice (p < 0.001, [150]Figure 6E). These improvements were consistent in both CA1 ([151]Figures 6B and 6E) and CA3 ([152]Figures S3B and S3E), highlighting the tNIR therapy’s ability to restore spatial coding efficiency. By restoring spatial information content, tNIR therapy supports improved learning and memory performance across dynamic environments. Transcranial near-infrared therapy reduces firing sparsity to improve spatial selectivity Firing sparsity, a neural coding strategy where only a small subset of neurons is active during any given event,[153]^40 plays a critical role in spatial selectivity and efficient neural encoding. Chronic SD significantly elevated firing sparsity, reflecting impaired spatial representation. In the open field (OP) paradigm, SD mice exhibited higher sparsity in CA1 during OP01 (0.75 ± 0.05) compared to controls (0.54 ± 0.05, p < 0.0001) and tNIR-treated mice (0.61 ± 0.1, p < 0.0001) ([154]Figure 6C). Similar patterns were observed in the linear track paradigm ([155]Figure 6F) and in CA3 ([156]Figures S3C and S3F), suggesting reduced spatial selectivity and compromised hippocampal network function. tNIR therapy significantly reduced firing sparsity in both CA1 ([157]Figures 6C and 6F) and CA3 ([158]Figures S3C and S3F), enhancing spatial selectivity and facilitating more efficient spatial encoding. Across all groups, firing sparsity progressively decreased over the course of OP and linear track sessions, indicating improved spatial focus and adaptive learning with repeated exposures. Notably, tNIR-treated mice exhibited firing sparsity levels approaching those of controls. By reducing firing sparsity, tNIR therapy facilitates enhanced spatial selectivity, contributing to improved spatial learning and memory processes. Transcranial near-infrared therapy normalizes theta phase-locking in place cells Theta oscillations are critical for encoding spatial information, with both the firing rate and the phase of spiking relative to local theta rhythms contributing to spatial representation.[159]^41 To evaluate the impact of tNIR therapy on hippocampal network dynamics, we analyzed the phase-locking of place cell firing to theta oscillations during spatial tasks. Theta phase-locking analyses during OP03 and Track03 revealed significant disruptions in the temporal coordination of place cell firing relative to theta oscillations in SD mice. In both CA1 ([160]Figures 6G–6J and [161]S4A–S4C) and CA3 ([162]Figures S3G–S3I and [163]S3M–S3O), SD mice exhibited a pathological firing phase preference around 150° of the theta phase ([164]Figures 6H and [165]S4A), significantly deviating from the patterns observed in controls and tNIR-treated mice (p < 0.0001, [166]Figures 6I and [167]S3I). Mean vector length (MVL) analysis revealed hyper-synchronization in SD mice, which was attenuated by tNIR therapy to levels comparable to controls (p < 0.0001). Furthermore, theta correlation analysis showed reduced synchrony in SD mice during both OP03 and Track03, which was effectively restored by tNIR therapy in CA1 and CA3 (p < 0.0001, [168]Figures 6J and [169]S4C). These findings highlight that tNIR therapy restores theta phase-locking dynamics, re-establishing the temporal precision necessary for efficient spatial representation and hippocampal function during spatial tasks. Transcranial near-infrared reactivates task-related sharp-wave ripple dynamics SWRs are pivotal for reactivating hippocampal memory traces during spatial learning, supporting memory consolidation and memory-guided decision-making.[170]^14 In chronic SD mice, SWR dynamics were significantly disrupted within the hippocampus, with reductions observed in ripple phase-locking strength and network-wide synchronization across both CA1 ([171]Figures 6K–6N and [172]S4D–S4F) and CA3 ([173]Figures S3J–S3L and S3P–S3R). These disruptions impaired coordination between place cell firing and ripple oscillations, hindering the effective reactivation of spatial memory traces. Ripple-phase correlation analysis further revealed markedly reduced correlation values in SD mice compared to controls (p < 0.0001), reflecting compromised SWR-mediated hippocampal network communication. tNIR therapy restored ripple-phase correlation levels in both CA1 and CA3 to near-control values (p < 0.0001), accompanied by strengthened alignment between place cell activity and ripple oscillations. This recovery of SWR dynamics highlights improved hippocampal processing efficiency and network synchronization. These findings demonstrate that tNIR therapy effectively reactivates SWR-associated hippocampal dynamics disrupted by chronic sleep deprivation. By restoring ripple phase-locking strength and enhancing coordination, tNIR therapy supports hippocampal network functionality, facilitating spatial memory integration and task-related cognitive processing. Discussion Chronic sleep disturbances are integral to AD pathology, manifesting as fragmented sleep, altered sleep architecture, and circadian rhythm disruptions.[174]^42^,[175]^43 These disturbances exacerbate hippocampal dysfunction and cognitive decline, creating a vicious cycle that accelerates neurodegeneration.[176]^44^,[177]^45^,[178]^46 The chronic SD model employed in this study represents an exacerbated but reversible condition, allowing for the precise exploration of therapeutic strategies targeting modifiable contributors to neurodegeneration.[179]^47^,[180]^48 Our findings establish tNIR therapy as a promising intervention to reverse SD-induced hippocampal dysfunction by addressing multiple interconnected mechanisms, including CBF restoration, metabolic reprogramming, and ionic homeostasis. These improvements collectively support hippocampal oscillatory dynamics, enhance place cell functionality, and promote memory-related network synchronization. Clinically, patients with AD frequently experience compounded sleep disturbances due to stress and environmental factors, further amplifying neurodegenerative processes.[181]^49^,[182]^50 The chronic SD model captures this real-world burden, offering a comprehensive framework for testing interventions such as tNIR therapy to mitigate its detrimental effects. These findings emphasize the potential of tNIR as a non-invasive approach to disrupt the cycle of sleep disturbances and neurodegeneration in Alzheimer’s disease. Transcranial near-infrared restores cerebral blood flow to support recovery Reduced CBF is a hallmark of both AD and sleep deprivation, contributing to mitochondrial dysfunction and neuronal vulnerability.[183]^51^,[184]^52^,[185]^53 Previous studies have demonstrated tNIR therapy’s efficacy in enhancing CBF in stroke models,[186]^54 but its effects on chronic SD-induced vascular deficits in the AD context were largely unexplored. Our results reveal that tNIR therapy significantly restores CBF in the DG and CA3 regions, addressing SD-induced deficits. This increase in CBF is primarily mediated by nitric oxide (NO)-dependent vasodilation rather than direct thermal effects.[187]^55^,[188]^56 Given the minimal thermal effects observed under similar irradiation conditions (<0.3 °C at the scalp, <0.04°C in deep gray matter),[189]^57^,[190]^58 the increase in CBF is unlikely to be driven by passive heating. Instead, tNIR-induced CBF enhancement aligns with NO-mediated vasodilation, independent of thermal contributions. While increased perfusion may facilitate nutrient delivery, KEGG pathway analysis indicates structured metabolic shifts favoring oxidative phosphorylation and antioxidant pathway normalization, suggesting an active metabolic response beyond passive perfusion effects. CBF restoration supports mitochondrial function, nutrient delivery, and oxidative stress mitigation, laying the groundwork for improved neuronal health and network stability.[191]^59 This vascular recovery is integral to downstream benefits, including metabolic reprogramming and enhanced hippocampal functionality. Metabolic reprogramming as a core mechanism of transcranial near-infrared therapy Metabolic dysfunction, including a shift from mitochondrial oxidative phosphorylation to glycolysis, is a hallmark of aging, AD, and chronic SD.[192]^60^,[193]^61 Our findings show that tNIR therapy reactivates mitochondrial oxidative phosphorylation, normalizes glycolytic activity, and restores antioxidant pathways, such as cysteine and methionine metabolism. These metabolic improvements enhance ATP production, reduce oxidative stress, and support synaptic transmission, contributing to cognitive resilience.[194]^62 Relative to the SD group, the tNIR-treated group exhibited a predominant downregulation of metabolites, including steroid hormones, neurotransmitters, amino acids, nucleotides, and intermediates of glycolysis and gluconeogenesis. Many of these pathways were elevated under SD conditions, and their decrease with tNIR suggests a restoration of metabolic homeostasis rather than broad suppression. KEGG analysis further indicates that tNIR redirects energy metabolism from glycolysis to mitochondrial oxidative pathways, thereby optimizing bioenergetic efficiency and alleviating metabolic stress. By reprogramming metabolism and recovering ionic homeostasis, tNIR therapy establishes a synergistic framework for memory consolidation and hippocampal network recovery. The observed metabolic changes in the tNIR group exhibited structured, pathway-specific alterations, indicative of active metabolic regulation rather than nonspecific systemic variations. These targeted metabolic adaptations contribute to the restoration of cellular homeostasis and neuronal function, reinforcing tNIR’s therapeutic efficacy. This integrative approach highlights metabolic reprogramming as a cornerstone of tNIR’s therapeutic efficacy, offering novel insights into mitigating SD-induced impairments and AD progression. Metabolic clustering reveals structured transcranial near-infrared modulation K-means clustering identified structured metabolic shifts across SD, tNIR-treated, and control groups. Clusters 2 and 3 exhibited the most recovery, with Cluster 2 metabolites (AMP, GMP, L-Tyrosine, L-Methionine) restored post-tNIR, supporting nucleotide metabolism, protein synthesis, and antioxidative defenses, while Cluster 3 metabolites (lactic acid, ribose-5-phosphate, nicotinamide) normalized, indicating a shift from glycolysis to mitochondrial respiration. Clusters 0, 5, and 8 followed a V-shaped trajectory, where SD-induced elevations in steroid hormones, neurotransmitters, and oxidative markers were reversed by tNIR, restoring neuroendocrine balance and redox homeostasis. Clusters 1 and 7 displayed a progressive shift, with tNIR modulating metabolic pathways toward equilibrium rather than full reversal. Clusters 4 and 6 exhibited a transient peak post-tNIR, possibly indicating an adaptive metabolic response. These findings suggest tNIR selectively reprograms metabolism, reinforcing its neuromodulatory role beyond passive systemic effects. Transcranial near-infrared rebalances ionic homeostasis to stabilize network function Ionic imbalances, particularly involving calcium (Ca^2+) and iron (Fe^2+), are critical contributors to neuronal dysfunction and synaptic instability in AD.[195]^63 Chronic SD exacerbates these deficits, disrupting ionic balance and impairing synaptic signaling. Our findings demonstrate that tNIR therapy effectively restores Ca^2+ and Fe^2+ levels, supporting mitochondrial function and neurotransmitter release. Beyond restoring Ca^2+ homeostasis, tNIR treatment was associated with an upregulation of CaMKII, a key regulator of synaptic plasticity, as suggested by preliminary proteomics analyses conducted independently of this study. While stress can alter phosphorylation states, tNIR-induced kinase modulation appears distinct from nonspecific stress responses. This effect may contribute to the observed improvements in hippocampal network stability and synaptic function. Calcium stabilization enhances LTP, a mechanism underlying synaptic plasticity, while iron recovery strengthens energy metabolism via the electron transport chain.[196]^63^,[197]^64 These ionic improvements underpin hippocampal oscillatory dynamics and network synchronization, facilitating cognitive resilience and neuronal stability.[198]^65^,[199]^66 Transcranial near-infrared restores sleep architecture and hippocampal oscillatory dynamics Chronic SD disrupts sleep architecture, modifying SWS and altering REM sleep, both essential for memory consolidation and cognitive function.[200]^15^,[201]^67 Our results show that tNIR therapy reorganizes sleep architecture by significantly increasing SWS and reducing REM duration. Enhanced SWS supports hippocampal-cortical communication, crucial for memory reactivation and consolidation.[202]^68^,[203]^69 Additionally, tNIR therapy restores oscillatory dynamics across delta, theta, and gamma frequency bands, promoting network synchronization and efficient cognitive processing. Enhanced SWR activity in CA1 and CA3 regions further primes hippocampal networks for memory reactivation and consolidation during sleep. These findings align with the critical role of SWRs in integrating spatial and episodic memory traces during SWS.[204]^70 Transcranial near-infrared enhances oscillatory synchronization and spatial encoding Sleep deprivation-induced disruptions in theta and ripple oscillations impair hippocampal network synchronization, compromising spatial memory encoding and retrieval.[205]^71^,[206]^72 Our study demonstrates that tNIR therapy restores these dynamics, significantly improving theta phase-locking and ripple correlation in CA1 and CA3 regions. This recovery is accompanied by enhanced spatial encoding, as evidenced by reduced place field sizes, increased spatial information content, and decreased firing sparsity across hippocampal subregions. These results highlight tNIR’s role in reinstating hippocampal oscillatory precision and spatial memory encoding, offering a targeted approach to mitigate SD-induced cognitive impairments. Region-specific responses of CA1 and CA3 to transcranial near-infrared therapy Our study identifies region-specific responses to tNIR therapy within the hippocampus. Chronic SD disrupts key processes, including theta phase-locking, ripple dynamics, and spatial encoding,[207]^36^,[208]^73^,[209]^74 with both CA1 and CA3 showing significant recovery following tNIR intervention. Ripple correlation showed pronounced improvement in CA3, reflecting its role in pattern completion and memory replay during task-related exploration.[210]^37^,[211]^75 CA1 exhibited significant enhancements in theta phase-locking and reduced firing sparsity, supporting its integrative role in hippocampal-cortical communication.[212]^18 These findings underscore tNIR’s capacity to restore hippocampal oscillatory synchronization and spatial encoding, accommodating the distinct functional roles of CA1 and CA3. By addressing region-specific dysfunctions, tNIR provides a nuanced approach to mitigate SD-induced hippocampal impairments and support memory function in Alzheimer’s disease. Conclusions This study establishes tNIR therapy as a promising, non-invasive intervention for restoring hippocampal function disrupted by chronic SD. By enhancing cerebral blood flow, metabolic reprogramming, and ionic homeostasis, tNIR restores hippocampal theta and ripple oscillations, improves place cell activity, and enhances spatial encoding. Additionally, tNIR normalizes sleep architecture by promoting SWS and reducing excessive REM sleep, supporting memory consolidation and hippocampal-cortical communication. These findings underscore tNIR’s potential to mitigate SD-induced cognitive impairments and neurodegeneration, offering broad clinical applications for sleep-disrupted and neurodegenerative populations. Limitations of the study While this study establishes the therapeutic potential of tNIR therapy, some limitations must be considered. The chronic SD model in mice, while controlled, does not fully replicate the complexity of sleep disturbances in human patients with AD, who typically experience fragmented sleep and circadian disruptions. Translational studies using human-relevant sleep disruption paradigms are needed. Additionally, optimizing tNIR parameters, such as wavelength and treatment duration, poses challenges for personalized clinical applications. Long-term effects of tNIR therapy on hippocampal function and AD progression remain unexplored and warrant longitudinal studies. Despite these limitations, our findings provide strong evidence for tNIR’s efficacy in mitigating SD-induced hippocampal dysfunction and highlight its potential as a scalable, non-invasive therapy for addressing sleep-related neurodegeneration. Resource availability Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Sheng Wang (wangsheng@hebmu.edu.cn). Materials availability This study did not generate new unique reagents. Data and code availability * • Standardized metabolomics data have been deposited in Mendeley Data ([213]https://www.doi.org/10.17632/69shd7b46g.1) and are publicly available * • Standardized ionomics data have been deposited in Mendeley Data ([214]https://www.doi.org/10.17632/69shd7b46g.1) and are publicly available * • The data and information are available within the main text or [215]supplemental information. Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request. * • This article does not report original code. Acknowledgments