Abstract Background Sleep disorder is widespread and involves a variety of intricate factors in its development. Sleep deprivation is a manifestation of sleep disorder, can lead to energy metabolism disturbances, weakened immune system, and compromised body functions. In extreme situations, sleep deprivation can cause organ failure, presenting significant risks to human health. Purpose This study aimed to investigate the efficacy and mechanisms of Astragalus Radix vesicles-like nanoparticles (AR-VLNs) in counteracting the deleterious effects of sleep deprivation. Methods The ICR mice were divided into control, model, AR-VLNs high dose (equivalent to 20 g/kg crude drug), AR-VLNs low dose (equivalent to 10 g/kg crude drug), AR high dose (equivalent to 20 g/kg crude drug), and AR low dose (equivalent to 10 g/kg crude drug). The REM (rapid eye movement) sleep-deprivation model was established, and evaluations were conducted for motor function, antioxidant capacity, and energy metabolism indices. Moreover, CACO-2 cells damage was induced with lipopolysaccharide to evaluate the repairing ability of AR-VLNs on the intestinal cell mucosa by measuring permeability. Furthermore, metabolomics was employed to elucidate the mechanisms of AR-VLNs action. Results AR-VLNs were demonstrated to enhance the motor efficiency and antioxidant capacity in REM sleep-deprived mice, while also minimized pathological damage and restored the integrity of the intestinal mucosal barrier. In vitro experiments indicated the anti-inflammatory effect of AR-VLNs against LPS-induced cell damage. Additionally, metabolomic analysis linked these effects with regulation of the amino acid metabolic pathways. Further confirmation from molecular biology experiments revealed that the protective effects of AR-VLNs against the deleterious effects of REM sleep deprivation were associated with the restoration of the intestinal mucosal barrier and the enhancement of amino acid metabolism. Conclusion AR-VLNs administration effectively improved energy metabolism disorders in REM sleep deprived mice, by facilitating the repair of the intestinal mucosal barrier and regulating the amino acid metabolism. Graphical Abstract [44]graphic file with name 12951_2024_3034_Figa_HTML.jpg Supplementary Information The online version contains supplementary material available at 10.1186/s12951-024-03034-x. Keywords: Nanoparticles, Sleep deprivation, Amino acid metabolism, Astragali Radix, Metabolomics, Intestinal barrier function Introduction Sleep is a fundamental physiological activity necessary for optimal functioning of the human body. Sleep disorder is a widespread concern, including insomnia and hypersomnia. Sleep deprivation is a manifestation of sleep disorder to simulate insomnia. Lack of sleep can lead to harmful substances accumulated, body constitution decreased, mental stress and fatigue. Nearly half of the global population reporting instances of insomnia at some point in their lives [[45]26]. Approximately 10 percent of people suffer from reduced mobility or illness due to sleep deprivation [[46]19]. Chronic sleep deprivation can lead to organ lesions and increase the risk of various diseases [[47]6]. Sleep deprivation is influenced by multiple factors, including living condition, study environment, depression, anxiety, and diseases [[48]22]. It has been linked to metabolic dysfunction and reduced muscle protein synthesis. Clinical studies indicate that acute sleep deprivation leads to muscle protein breakdown, elevated cortisol and testosterone levels. Even a single night of insufficient sleep can trigger metabolic dysfunction [[49]12]. Moreover, sleep deprivation can be a secondary issue related to immune system imbalance, exacerbating existing diseases. Several adverse effects are associated with sleep deprivation, including cognitive decline, lysosomal damage, altered expression of immune-regulatory genes [[50]20], reduced exercise endurance, lowered inspiration threshold, and diminished nerve tone [[51]38]. Sleep deprivation has extensive pharmacological research value. Various animal models have been employed to study sleep deprivation, including the horizontal table model, forced movement model, and gentle induction method. This suggest that made sleep deprivation model in animals is feasible. The horizontal table model being the most commonly used. Tai et al. utilized a horizontal table model to create a rapid eye movement (REM) sleep deprivation model to explore potential treatments for cognitive deficits in sleep-deprived individuals [[52]29]. Similarly, Xu et al. employed a horizontal model to establish a rat sleep deprivation model to assess the impact of sleep deprivation on nerve damage and cognitive impairment induced by subarachnoid hemorrhage [[53]41]. Additionally, Youri G. Bolsius et al. used gentle induction by tapping cages to create a sleep deprivation model to study the effects of sleep deprivation on neurons [[54]1]. Above literature show that horizontal table model can efficiently cause sleep lack in animals. Metabolomics is a technique used for the qualitative or quantitative analysis of all endogenous metabolites in organisms or cells. Due to its capabilities for high-throughput detection and data analysis, it has been extensively utilized in systems biology research and drug mechanism discovery. Several studies have shown that sleep deprivation can disrupt the endogenous metabolic environment. For example, Kim et al. showed that sleep deprivation disrupted glucose metabolism and altered the expression of the glucose transport enzyme O-GlcNAc in mice [[55]10]. Moreover, it was reported that acute sleep deprivation significantly raised body temperature, as well as the temperature of the hypothalamus and cortex in mice, indicating disturbances in the internal environment [[56]34]. Zhang et al. found that Poria cocos could alleviate anxiety induced by chronic sleep deprivation by modulating sphingolipid, phenylalanine, taurine, and hypotaurine metabolism [[57]46]. Numerous studies have indicated that sleep deprivation can inflict damage on the intestinal tissues. For example, Yi Chung et al. discovered that sleep deprivation decreased intestinal permeability and hastened the damage of tight junction in mice [[58]4]. Similarly, Yao et al. discovered that sleep deprivation disrupted the intestinal microbiota, leading to an imbalance in the microbiota and neuroinflammation, suggesting that this damage might be hereditary [[59]43]. Lamon et al. transplanted gut microbes, which resulted in improved cognitive impairment caused by sleep deprivation [[60]36]. These studies suggested that the treatment of sleep deprivation may be associated with intestinal repairment. However, further research is required to elucidate the specific mechanism. Astragalus is an herb used for tonification in Traditional Chinese Medicine (TCM) and is known to have various functions, including promoting tissue regeneration and wound healing, and inducing diuresis to alleviate edema. Recent research has shown that Astragalus can enhance energy metabolism. Studies on its anti-aging effects have revealed that Astragalus can increase the body’s antioxidant capacity by improving energy and amino acid metabolism [[61]42]. Additionally, Astragalus has been found to alleviate chronic fatigue syndrome by regulating metabolic profiles and intestinal microbiota [[62]37]. Numerous studies have investigated the effects of Astragalus on the intestine. The existing literature indicates that Astragalus has anti-inflammatory effects both in vivo and in vitro, protects intestinal cells, and intervenes in intestinal inflammatory changes [[63]3]. These processes are closely associated with the activation of the immune barrier and the repair of the intestinal mucosa [[64]14, [65]32]. Additionally, studies have reported changes in the regulation of amino acid metabolism [[66]13, [67]28, [68]35, [69]40, [70]44]. Extracellular vesicles are cell-secreted, membrane-bound structures, including exosomes, which are a specific type of extracellular vesicle with a diameter ranging from 40 to 200 nm [[71]27]. Exosomes play a crucial role in transferring mRNA, miRNA, and proteins between cells. While animal-derived exosomes exhibit specific markers such as CD63 and CD81, plant-derived nanoparticles lack these markers [[72]39]. Consequently, plant-derived nanoparticles are often referred to as vesicle-like nanoparticles (VLNs). VLNs usually have a concave hemispherical or saucer-like shape and a lipid bilayer structure that provides high stability. The primary components of VLNs include small RNAs (mRNA, miRNA, and LncRNAs), lipids, and plant active components [[73]47]. VLNs have shown anti-inflammatory, anti-tumor, and anti-fibrosis effects. For instance, VLNs derived from garlic chives demonstrated therapeutic potential in NLRP3-induced inflammatory diseases [[74]18]. Additionally, honey derived VLNs displayed anti-inflammatory activity in mice with acute liver injury [[75]2]. Moreover, VLNs have the ability to regulate intestinal flora, thereby influencing the body's balance [[76]31]. However, research on Astragalus Radix vesicles-like nanoparticles (AR-VLNs) remains limited, and the pharmacodynamics and functional mechanisms of AR-VLNs are not clear. Therefore, the aim of this study was to detect the effects of AR-VLNs on behavioral performance, tissue function and metabolic environment of sleep deprived mice, and investigated the potential mechanism of AR-VLNs in sleep deprivation. Based on the experimental results, this study proved that AR-VLNs could repair intestinal mucosal barrier, promote amino acid metabolic pool homeostasis and improve REM sleep deprivation induced fatigue. The significance of this study was to provide a basis for the biological process of body function decline caused by REM sleep deprivation, proposed a potential direction related clinical treatment. In addition, compared with previous studies on Astragalus, this study innovatively conducted a comprehensive study on vesicles of Astragalus, enriched the research of plant-derived vesicles and contributed to the development of nanoparticles of herbal effective ingredients. Material and methods Materials and Reagents Astragalus decoction pieces were purchased from Hebei Baicao Kangshen Pharmaceutical Co., LTD (2,008,012, Hengshui, China). Mobile phase solvents of LC/MS grade were purchased from Fisher Co. LTD. (Waltham, USA). Rabbit polyclonal antibody to Occludin (AF7644) and ZO-1 (AF8394) were obtained from Beyotime Biotechnology Co., LTD (Shanghai, China). Rabbit monoclonal antibody to SLC6A19 (ab180516) and Rabbit polyclonal antibody to SLC38A5 (ab72717) were obtained from Abcam Plc (Cambridge, US). Rabbit polyclonal antibody to β-actin (AF7018) was obtained from Affinity Biosciences (Cincinnati, USA). HRP Goat Anti-Rabbit IgG (AS014) was obtain from ABclonal Technology Co., LTD (Wuhan, China). DMEM medium, fetal bovine serum (FBS) and Pierce BCA Protein Assay Kit was purchased from Thermo Fisher Scientific (Waltham, USA). Lipopolysaccharide and FD-4 were purchased from Sigma-Aldrich (Santa Clara, USA). ATP, ALP, MDA and SOD kits were purchased from Beyotime Biotechnology Co., LTD (Shanghai, China). Elisa kits of IL6, IL1β and TNF-α were purchased from Elabscience Biotechnology Co., Ltd. (Wuhan, China). UltraSYBR One Step RT-qPCR kit was purchased from Cwbio Biotechnology Co., LTD (Jiangsu, China). PAGE gel preparation kits were purchased from Epizyme Biotech Co., LTD (Shanghai, China). Exosome fluorescent Dyes (DIR, UR21017) was purchased from Umibio Biotechnology Co., LTD (Shanghai, China). Preparation of AR-VLNs The Astragalus decoction pieces were soaked with 12 times volumes of distilled water and subjected to two rounds of heating reflux, each lasting for 90 min. The filtrates from both rounds were combined and reduced to a final volume of 100 mL using a vacuum rotary evaporator. The resulting solution was then centrifuged at 4℃ for 10 min at 300 × g. The supernatant was then subjected to a second centrifugation at 4°C for 30 min at 3,000 × g. The obtained supernatant was transferred to another 50 mL tube, followed by a third centrifugation at 4°C for 30 min at 10,000 × g. Subsequently, the supernatant was carefully transferred to an ultracentrifuge tube and subjected to ultracentrifugation at 4°C for 90 min at 150,000 × g. The resulting supernatant was discarded, while the precipitate was resuspended in 200 μL of PBS and stored at -80°C. The supernatant collected throughout the entire procedure served as the negative control for administration. Morphological characterization of AR-VLNs Transmission electron microscopy analysis A 15 μL sample was placed on a copper grid and left to sit for 1 min before being gently blotted with filter paper. Subsequently, the sample was stained with 15 μL of a 2% UO[2] acetate staining solution for 1 min and then observed using FEI Tecnai Spirit transmission electron microscopy (Hillsboro, USA). Particle size detection After washing the sample cell with deionized water, the NanoSight LM14 instrument (Malvern, UK) was calibrated using polystyrene microspheres (110 nm). The AR-VLNs were diluted with PBS and injected into the instrument for nanoparticle tracking analysis (NTA) detection. Component identification of AR-VLNs UPLC-MS/MS analysis The obtained AR-VLNs homogenate (20 μL) was added in 400 μL methanol, and the mixture was vortexed for 30 s and centrifuged at 14 000 rpm for 10 min. The supernatant was collected and injected to UHPLC-MS/MS system. The chromatographic separation was conducted using a Waters Acquity UPLC HSS T3 column (2.1 × 100 mm, 1.8 µm). Mobile phase A was composed of acetonitrile, while mobile phase B was a 0.1% formic acid in water, utilized for gradient elution as follows: 0–5 min, 5%–20% A; 5–11 min, 20%–30% A; 11–22 min, 30%–78% A; 22–26 min, 78%–95% A; 26.01–30 min, 100% A; 30–32 min, 100%–5% A; 32–40 min, 5% A. The flow rate was maintained at 0.4 mL/min, the column temperature was 40 °C, and the injection volume was 5 µL. Detection was performed using electrospray ionization (ESI) in both positive and negative ion modes. The specific parameters were set as follows: Scan range of 120–1800 m/z; Spray voltage of 3.5 kV (positive) and 3.0 kV (negative); Resolution of 70,000 for MS1 and 17,500 for MS2; AGC target of 3e^6 for MS1 and 1e^5 for MS2; Maximum injection times of 100 ms for MS1 and 50 ms for MS2; Sheath gas flow rate of 30 arbitrary units; Auxiliary gas flow rate of 5 arbitrary units; Sweep gas flow rate of 0 arbitrary units; Capillary temperature of 320 °C; S-Lens RF level of 50%; Auxiliary gas heater temperature of 350 °C; Normalized collision energies (NCE) of 10, 20, and 40. The original data was processed and analyzed using Compound Discoverer 3.2 software (Thermo Fisher Scientific, Waltham, USA). After peak extraction, the mass spectra of compounds were searched and compared with a local database and the available mzCloud online database. Chemical constituents were identified or tentatively characterized based on mzCloud with a m/z tolerance of 5 ppm. Identification of lipid composition To determine the lipid composition, the obtained AR-VLNs homogenate was mixed with methanol and MTBE. After vortexing and centrifuging at 14,000 rpm for 15 min, the supernatant was collected. UPLC-MS/MS was then used for data acquisition and LipidSearch software version 4.1 (Thermo Fisher Scientific, Waltham, USA) was used for data analysis. Relevant parameters could be found in the supplementary materials. MiRNA sequencing Total RNA was extracted from the samples and ligated RNA 3' and 5' adapters separately. After reverse transcription and amplification, miRNA libraries were created, and the sequences were compared to the database. Detailed protocols could be found in the supplementary materials. MiRNA target prediction was carried out through TarBase 8.0, miRTarBase 2019 and miRecords databases, the corresponding proteins of target mRNA were imported into String database to build PPI network, the species of human was set, and the maximum confidence of 0.9 was selected. The obtained PPI network results were imported into cytoscape 3.7.2 software for visualization. The Degree algorithm in cytohubba plug-in was used to calculate the top 60 targets as key targets. And these targets were imported into KOBAS 3.0 database to perform KEGG pathway enrichment (type was set to Ensembl Gene ID, species was set to human, p value < 0.01). While GO pathway enrichment was performed by metascape database (species was set as H. Sapens, the “P Value Cutoff” was 0.01, and the “Min Enrichment” was 1.5). Animals and model preparation The male ICR mice (24 ± 2 g) were purchased from the Spiff (Beijing) Biotechnology Co., LTD (SYXK (Jing) 2018–0018) and housed in the specific-pathogen free animal facility at Tsinghua University. The mice were maintained on a 12/12 h light/dark cycle at a temperature of 22–26 ℃ with access to sterile pellet food and water ad libitum. All the animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Tsinghua University (23-JYY1). The ICR mice were divided into six groups: control, model, AR-VLNs high dose (AVH, equivalent to 20 g/kg crude drug), AR-VLNs low dose (AVL, equivalent to 10 g/kg crude drug), AR high dose (AH, equivalent to 20 g/kg crude drug), and AR low dose (AL, equivalent to 10 g/kg crude drug). AR was the extraction of Astragalus with AR-VLNs removed to compare the efficacy of VLNs with other components of Astragalus. The dose of crude drug was converted according to the clinical dosage. The AR-VLNs obtained from a certain amount of raw drug, was considered to be equivalent to the corresponding amount of crude drug. Intragastric administration lasted during the whole sleep deprivation period. The chronic REM sleep deprivation model was implemented in all groups except the control group. A custom-made horizontal platform (height 40 mm, standpipe diameter 10 mm, chassis diameter 60mm, platform diameter 25mm) was placed in the cage, with 25 mm water was added to the cage (Fig.S1). The mice would keep awake to avoid falling into the water. To disrupt the circadian rhythm, the started time and duration of modeling were different every day. The modeling time ranged from 17 to 24 h per day, and at least one day per week was subjected to 24 h of continuous sleep deprivation. During the non-deprivation periods, mice were returned to their normal cages and had free access to food and water. The schematic diagram of the animal experimental process was shown in Fig.S2. Infrared imagery of AR-VLNs distribution in vivo The dye solution was added to the extracted VLNs suspension, the final concentration of dye was 20 μM. The mixture was vortexed for 1 min and incubated at 37 ℃ for 30 min, then 10 mL PBS was added. The VLNs were re-extracted according to the extraction method to remove free dye. Re-suspended precipitation was intragastric administrated to mice in high dose. Mice were anesthetized by isoflurane to obtain infrared image by IVIS Spectrum CT vivo imager (Caliper-Perkin Elmer, MA, USA) at 1 h, 3 h, 4 h, and 6 h after administration, respectively. Behavioral indicators measured of sleep-deprived mice Gripping strength The gripping strength of each group of mice was measured by YLS-13A grip strength test meter (Zhongshi Science, Beijing, China). Gently lifted the tail of mice to suspend it, the mice would touch the panel of the instrument with front paws, dragged it and let mice pull the instrument dial, recorded the data of grasping instrument panel. The gripping strength of each mouse was measured 5 times, and the average value was recorded. Pulse The pulse amplitude of each mouse in control, model, AVH and AH group was measured using the MouseOx Plus non-invasive monitor (Starr life sciences, Oakmont, USA) and recorded. Mice were placed in supine positions after anesthesia. The pulse meter probe was placed on the foot of mice. The pulse meter monitoring images were observed, and the relevant data were recorded. Motor ability The Ugo Basile 47,300 mouse treadmill (Gemonio, Italy) was used to evaluate the motor ability and endurance of the mice. The instrument parameters were set according to the methods reported by Zhang et al. [[77]45]. The running distance and the number of electric shocks were recorded. The mice were placed on different tracks, ran the treadmill, and continuously observed the state of the mice. When the endurance of the mice decreased, they would fall into the electric shock zone behind the track, and after being stimulated by electric shock, the mice would generally return to the track until they were exhausted. The treadmill automatically recorded shocked frequency. Open field The open field consisted of a white acrylic box with dimensions of 50 cm × 50 cm × 50 cm. The mice were put in the center position and allowed to move freely in the open field for 10 min. Recorded the movements by camera. The distance traveled was calculated using EthoVision XT 11.5_win10 software. HE staining The mice were anesthetized with Avertin (300 mg/kg) and blood was collected from the apex of heart. Then mice were euthanized, and the proximal cecal tissue of the intestine was collected. The tissue was fixed with 10% formalin for 24 h, embedded with paraffin wax, and sliced for HE staining. Pathological staining images were obtained using the 3DHISTECH slide scanning system Pannoramic SCAN (Budapest, Hungary). Biomarker measured Intestinal tissue of mice was weighed, and 9 times the volume of PBS was added. Homogenized at 4 ℃ to obtain homogenate for the detection of relevant biochemical indexes. The related biomarkers were measured using test kits, strictly following the instructions provided with the kits. Cell culture CACO-2 cells were purchased from Cell Resource Center, Peking Union Medical College (PCRC). The cells were cultured and differentiated in Transwell plates for 21 days [[78]5, [79]9, [80]16]. They were then divided into control group, model group, VLNs-high dose (VH, 1 × 10^10 particles/mL), VLNs-middle dose (VM, 1 × 10^9 particles/mL) and VLNS-low dose (VL, 1 × 10^8 particles/mL). Different doses of VLNs were simultaneously administered with the LPS (lipopolysaccharide) in the treatment group. CACO-2 cell damage model was induced by LPS, with 10 μg/mL LPS added and incubated for 24 h. Samples were collected for ALP (alkaline phosphatase) detection, FD-4 (Fluorescently labeled dextranan macromolecule) permeability detection, and molecular biology experiments. Evaluation of cellular mucosal barrier damage ALP measurement The ALP content of medium in both inside and outside compartments was measured using a test kit, following the instructions provided. And the ratio between the two sides was calculated. FD-4 permeability test FD-4 was added to the side of the Transwell chamber at the concentration of 0.1 mg/mL, while complete medium was added to the outside of the Transwell chamber. After incubation for 1 h, 0.1 mL of supernatant was obtained from both the inside and outside of the Transwell chamber. The absorbance was detected using an enzyme labeling apparatus, and the ratio between the two sides was calculated. Electron microscope observation Intestinal tissues of mice and CACO-2 cell were observed by electron microscope, respectively. The samples were promptly immersed in an electron microscope fixation solution (a mixture of 2% PFA (paraformaldehyde) and 2.5% glutaraldehyde) for 24 h, followed by 1% acid fixation for 1 h. Subsequently, the samples were dehydrated using gradient alcohol and embedded in resin. They were then sliced using a Leica EM UC6 ultra-thin microslicer, attached to copper wire, and observed under a Hitachi 7650B transmission electron microscope. Untargeted metabolomics analysis Sample preparation A total of 200 μL serum samples from mice were taken and mixed with 800 μL of 50% methanol acetonitrile solution (1:1) using vortex for 5 min. The mixture was centrifuged at 14,000 rpm for 15 min. The supernatant was collected, dried with nitrogen, and redissolved in 100 μL of a 50% acetonitrile water solution (including an internal standard of 5 μg/mL nimodipine). UPLC-QE-MS/MS conditions The chromatographic column used was a Waters BEH Amide Column (2.1 × 100 mm, 1.7 μm). The mobile phase A consisted of a 0.1% formic acid and 10 mM ammonium acetate solution, while phase B was acetonitrile containing 0.1% formic acid with gradient elution (0–5 min, 0% A; 5–6 min, 0%-5% A; 6–15 min, 25% A; 15–16 min, 25%-50% A; 16–25 min, 50% A; 25–26 min, 50%-0% A; 26–27 min, 0% A). The flow rate was set at 0.3 mL/min, the column temperature at 35 °C, and the injection volume at 5 µL. Data was obtained using the UHPLC-Q-Exactive mass spectrometry system (Thermo Fisher Scientific, Waltham, USA). The detection was performed using electrospray ionization (ESI) in positive and negative ion mode, respectively. The detailed parameters are as follows: Scan type Full ms-ddMS2; Scan range m/z 100–1200; Spray voltage 3.5 kV; Resolution 70,000 (MS1), 17,500 (MS2); AGC target 1e6; TopN 5; Sheath gas flow rate 40; Aux gas flow rate 15; Capillary temp 320 ℃; Aux gas heater temp 350 ℃; Collision energy (NCE) 30, 40, 50. Multivariate statistical analysis The original data was preprocessed and analyzed using Compound Discoverer 3.2 software. This involved several steps including peak extraction, peak alignment, retention time correction, and peak area extraction. After preprocessing, the structures of the metabolites were identified by comparing the mass spectra with a local database and the mzCloud online database. Subsequently, statistical analysis was performed using the MetaboAnalyst platform. This involved Principal Component Analysis (PCA), Partial Least Squares Discriminant Analysis (PLS-DA), and the screening of differential metabolites. These analyses help in understanding the variance in the dataset, identifying the metabolites that contribute most to the variance, and finding the metabolites that are significantly different between groups. The criteria for screening differential metabolites were as follows: 1. A p-value less than 0.05 for the T-test, which indicates that the difference in the mean value of the metabolite between groups is statistically significant. 2. A fold change in peak area greater than 2 between groups, which indicates that the metabolite is significantly more or less abundant in one group compared to another. 3. A Variable Importance in Projection (VIP) value greater than 1 for PLS-DA analysis, which indicates that the metabolite is important for the model and contributes significantly to the separation between groups. The combination of these criteria helps in identifying the most important and relevant metabolites that are affected by sleep deprivation and the treatments administered. GO and KEGG pathway enrichment analysis Metabolites showing significant differences were selected for subsequent bioinformatics analysis, including cluster analysis, correlation analysis, and pathway analysis. The GO pathway and KEGG pathway analyses were performed using the ClueGO plug-in in Cytoscape software. Targeted metabolomics analysis Sample preparation The intestinal tissue and fecal were accurately weighted, 80% methanol were added and mixed via vortex, followed ultrasound treated for 20 min at 4 ℃. Mixed 50 μL serum sample with 200 μL 50% methanol acetonitrile solution (1:1) by vortex for 5 min. The mixture was centrifuged at 14,000 rpm for 15 min. The supernatant was collected, dried with nitrogen and redissolved. UPLC-QTOF-MS/MS conditions Data was obtained by Waters ACQUITY UPLC I-Class AB UPLC system and SCIEX QTRAP 4500 MS system. The chromatographic column used was a Waters BEH Amide Column (2.1 × 100 mm, 1.7 μm). The mobile phase A consisted of acetonitrile solution containing 0.1% formic acid, while phase B consisted of aqueous solution containing 0.1% formic acid (0–4 min, 0%–10% B; 4–11 min, 10%–30% B; 11–14 min, 30%–55% B; 14–16 min, 55% B; 16–16.1 min, 55%–0% B; 16.1–20 min, 0% B). The flow rate was set at 0.2 mL/min, the column temperature at 40 °C, and the injection volume at 5 µL. The detection was performed using ESI in positive ion mode. The detailed parameters are as follows: The ion source temperature was 500℃, the gas curtain gas was 25 psi, the impact gas was medium, the ion voltage was 5500V, the spray gas and the auxiliary heating gas were 50 psi, respectively. Specific parameter was shown in Table [81]1. Table 1. Specific parameter of targeted metabolomics No Compound name Parent ion Fragment ion Inlet voltage /V Collision energy /V 1 L-Lysine 147 84 70 22 2 L-Histidine 156 110 75 22 3 L-Arginine 175 70.1 90 31 4 L-Serine 166 60.1 60 17 5 L-Glutamic acid 148 84.1 65 22 6 L-proline 116.1 70 65 22 7 L-isoleucine 132.1 86.1 65 16 8 L-leucine 132 86 75 19 9 L-phenylalanine 160 120 75 27 10 L-methionine 150 104 65 28 11 Hypoxanthine 137 110 80 30 12 Trytophan 205.1 188 70 16 [82]Open in a new tab Data analysis Compound information was collected by Analyst Software 1.6.4, and compound quantitative analysis was performed by MultiQuant^tm 3.0.3 Software. The compound content of the sample was calculated according to the standard peak area/concentration ratio. Quantitative reverse transcriptase (RT)-PCR analysis Total RNA was extracted from intestinal tissues and CACO-2 cells by Trizol reagent. A one-step SYBR dye fluorescence quantitative PCR kit (Jiangsu Cowin Biotech Co., Ltd, CW0659) was used, following the manufacturer's instructions. The Roche LightCycler 480 instrument (Basel, Switzerland) was used for the detection and the details was shown in supplementary materials. The primers were synthesized by Genewiz Biotechnology Co., LTD (Suzhou, China) and the sequences were listed in Table S1. Western blot analysis Proteins were extracted from intestinal tissues or CACO-2 cells by RIPA buffer and quantified by BCA kits. 10% PAGE gel was prepared for protein separation. The proteins were separated using electrophoresis and transferred to a PDVF membrane. The membrane was blocked with 5% BSA for 2 h, incubated with a primary antibody for 14 h, followed by a secondary antibody for 1 h. ECL chemiluminescence was performed and bands were visualized by iBright1500 gel imaging system (Thermo Fisher Scientific, Waltham, USA). Grayscale calculation was conducted using Image-J software. Statistical analysis Statistical analysis was performed by SPSS Statistics 24 software. The results were expressed as means ± SEM. Shapiro–Wilk method was used to test the data of each group. If the data met the normal distribution, one-way ANOVA was performed, P < 0.05 was considered to have no homogeneity of variance, and the Dunnett T3 method was used for comparison, otherwise the LSD was used for comparison, P < 0.05 was considered as significant difference between the two groups of data. If the data of each group did not meet the normal distribution, the non-parametric rank sum test was performed, and the Kruskal-Walli’s test was performed first. If there were significant differences between the data of each group, the Mann–Whitney U test was further used for comparison. If P < 0.05, it was considered to have significant difference between the two groups. Results Characterization of AR-VLNs The TEM results provided visual evidence of the morphology of the VLNs isolated from Astragalus Radix (AR). The VLNs had a typical 'tea holder' or concave spheroid shape, and the images showed a clean background, which suggests that the sample was relatively pure and free from debris or other contaminants (Fig. [83]1A). Additionally, the nanoparticle tracking analysis provided quantitative data on the concentration and size distribution of the AR-VLNs (Fig. [84]1B–C). The concentration was found to be approximately 1.23e^10 particles/mL with a standard deviation of 5.05e^8 particles/mL, indicating a high concentration of VLNs in the sample. The average particle size was determined to be 141.0 nm, with a size distribution that ranged approximately from 62.8 nm to 258.5 nm (Fig. [85]1D). Fig. 1. [86]Fig. 1 [87]Open in a new tab The characterization and distribution of AR-VLNs. A The representative TEM images of AR-VLNs. B–D The concentration and intensity distribution of AR-VLNs. E–H Infrared imagery of in vivo distribution (1h, 3h, 4h, 6h after administration). I The lipid composition of AR-VLNs. J The lipid contents of AR-VLNs The results from the infrared imagery suggest that after administration, the AR-VLNs were mainly found in the gastrointestinal tract, with a significant concentration in the intestinal region (Fig. [88]1E–H). This localization is crucial as it indicates the potential site of action of the AR-VLNs and suggests a possible protective role in the intestinal mucosa, which was one of the key objectives of the study. Lipidomic analysis provided insights into the lipid composition of the VLNs. The predominant lipid components were found to be glycerides, sphingolipids, and glycerophospholipids. Specifically, triglycerides (TG) constituted the majority of the lipid content (45%), followed by ceramides (Cer) at 23%, and phosphatidylcholines (PC) at 8% (Fig. [89]1I–J). This information is essential for understanding the structure and potential function of the AR-VLNs as different lipids can influence the stability, bioactivity, and interaction of the VLNs with biological membranes. Additionally, the further characterization procedures about chemical constituents of AR-VLNs was performed on UPLC-MS/MS system, and organic acids including azelaic acid, citric acid were identified (detailed information were shown in Table S2 and Table S3 in supplementary information). In summary, the above comprehensive analysis provided a detailed overview of the composition of AR-VLNs, which was crucial for potential biological effects and mechanisms study. MiRNA sequencing and enrichment of AR-VLNs MiRNA sequencing was conducted to analyze the miRNA profile in AR-VLNs. Nucleotide bias was illustrated in Fig. [90]2A. The results from three parallel samples demonstrated the stability of miRNA sequencing within the 20–24 length range (Fig. [91]2B). There were numerous miRNAs existed in the AR-VLNs. Notably, PU-MIR2916-P3_2SS18Tc19Ga was the most highly expressed, accounting for 44% of the miRNA content, followed by PU-MIR2916-P5_1SS3Ag at 29% (Fig. [92]2C). MiRNAs play a critical role in the regulation of gene expression and can influence various biological processes. Therefore, the high expression of specific miRNAs suggested they might have a significant role in the biological activity of the AR-VLNs. Thus, by employing database screening and other techniques, the target genes of the identified miRNAs were forecasted and pathway enrichment analysis was conducted. Fig. 2. [93]Fig. 2 [94]Open in a new tab MiRNA sequencing and enrichment of AR-VLNs. A Total miRNA nucleotide bias of AR-VLNs. The horizontal coordinate represented the miRNA length distribution, and the vertical coordinate represented the percentage of bases. B Length distribution of miRNA from three parallel samples sequenced. Results showed well parallelism. The horizontal coordinate represented the miRNA length distribution, and the vertical coordinate represented the proportion of different length. C The pie chart of miRNA composition percentage of AR-VLNs. D The illustrated bar graphs of biological process, cellular component, and molecular function in GO pathway enrichment analysis of miRNA potential targets. E The bubble graph of biological process in GO pathway enrichment analysis of miRNA potential targets. The horizontal coordinate represented the enrich factor. The size of bubble represented the gene number. The color of bubble represented the P value. F The bubble graph of KEGG pathway enrichment analysis of miRNA potential targets. G The pathway enrichment of miRNA related targets via MetaboAnalyst 6.0 platform Gene Ontology (GO) enrichment analysis illustrated bar graphs of biological process, cellular component, and molecular function were showed in Fig. [95]2D. Additionally, bubble graphs were generated utilizing the KOBAS 3.0 database to visualize the top 20 pathways in both biological process of GO process and KEGG categories (Fig. [96]2E–F). The enriched biological processes in GO analysis primarily encompass signal transduction and transcriptional regulation, while the pathways identified in KEGG enrichment, such as Pathways in cancer, Ras signaling pathway, PI3K-Akt signaling pathway, etc., were linked to cell growth and differentiation, protein transport and secretion. Lastly, the MetaboAnalyst 6.0 platform was employed for pathway enrichment of screened miRNA targets, revealed that the mainly regulated pathway was concentrated on amino acid metabolism in vivo (Fig. [97]2G). AR-VLNs enhanced the energy metabolism and behavior in REM sleep-deprived mice The study presents compelling evidence of the beneficial effects of AR-VLNs on energy metabolism and physical performance in REM sleep-deprived mice. Physical strength and exercise capacity were significantly compromised in REM sleep-deprived mice, as evidenced by reduced grasping strength, decreased motion path in the open-field tests, increased number of electric shocks of treadmill, reduced running mileage and decreased pulse. Administration of AR-VLNs and AR notably improved these parameters, indicating that AR-VLNs could reverse the adverse effects of REM sleep deprivation on muscle strength and exercise capacity, and the efficacy was equivalent to other active components in Astragalus (Fig. [98]3A–E). Fig. 3. [99]Fig. 3 [100]Open in a new tab AR-VLNs enhanced the energy metabolism and behavior in REM sleep-deprived mice. A Grasping strength of different groups of mice (n = 8). B Displacement of open field experiment of different groups of mice (n = 8). C Electric shock of treadmill experiment of different groups of mice (n = 8). D Displacement of treadmill experiment of different groups of mice (n = 8). Above results showed that the motor ability was significantly decreased in REM sleep deprivation mice, and increased after AR-VLNs treatment. E The pulse was significantly decreased in the REM sleep-deprived mice and enhanced by AR-VLNs treated (n = 8). F The lipid oxidation index MDA levels of different groups of mice (n = 8). G The antioxidant index SOD levels of different groups of mice (n = 8). Above results showed that REM sleep deprivation would decrease the antioxidant stress ability of mice, and AR-VLNs significantly improved it. H The ATP contents were significantly reduced in REM sleep-deprived mice and raised in AR-VLNs treated mice (n = 8). AVH, AR-VLNs high dose; AVL, AR-VLNs low dose; AH, AR high dose; AL, AR low dose. Error bars were presented as mean ± SEM. * P < 0.05, ** P < 0.01 vs the model group. # P < 0.05, ## P < 0. 01 vs the control group. ns indicated no significant difference Moreover, REM sleep deprivation led to an oxidative stress imbalance, as indicated by a significant increase in the malondialdehyde (MDA) index and a decrease trend in superoxide dismutase (SOD) content. Notably, AR-VLNs effectively regulated MDA back to normal levels, which is equivalent to other active components of Astragalus. AR-VLNs could also exhibit a tendency of SOD regulation, which demonstrated the antioxidant potential of AR-VLNs (Fig. [101]3F–G). Furthermore, REM sleep deprivation resulted in decreased ATP levels, reflecting a slowdown in energy metabolism. Remarkably, AR-VLNs and AR restored the normal ATP levels, suggesting that AR-VLNs can accelerate energy metabolism efficiency and promote ATP production. Compared with AR, AR-VLNs showed a better regulatory trend (Fig. [102]3H). Overall, these findings suggest that AR-VLNs can enhance energy metabolism, improve physical performance, and regulate oxidative stress in REM sleep-deprived mice, highlighting their therapeutic potential for REM sleep deprivation-induced disorders. AR-VLNs repaired intestinal barrier of REM sleep deprived mice Intestinal images revealed a significant decrease in colonic length in REM sleep-deprived mice, indicative of intestinal damage. However, administration of AR-VLNs significantly increased colonic length compared to the model group (Fig. [103]4A–B). Fig. 4. [104]Fig. 4 [105]Open in a new tab AR-VLNs repaired the intestinal barrier of REM sleep deprived mice. A Representative images of mice colonic tissues in different groups. B Colonic length statistics. C Representative images of HE stains of mice colonic tissues in different groups. D Representative images of TEM images of mice colonic tissues in different groups. E Inflammatory factor levels of intestinal (IL1β, IL6 and TNF-α). n = 3 in each group of (A–D), n = 8 in each group of (E). Data were presented as mean ± SEM. *P < 0.05, **P < 0.01 vs the model group. #P < 0.05, ##P < 0. 01 vs the control group HE staining pathological analysis showed ciliary fusion, crypt disappearance, and inflammation in the intestines of REM sleep-deprived mice, suggesting pathological damage caused by REM sleep deprivation. After AR-VLNs administration, the integrity of ciliary and crypt structures improved, and inflammatory damage disappeared (Fig. [106]4C). TEM observation further confirmed that REM sleep-deprived mice exhibited a disappearance of intestinal microvilli, while AR-VLNs administration effectively repaired the microvilli, protected the intestinal barrier, and restored microvilli morphology (Fig. [107]4D). Elisa kits were used to determine the inflammatory factor levels of intestinal. In the model mice, levels of IL1β, IL6 and TNF-α were significantly increased, which suggested the occurrence of inflammatory pathological injury. These expressions of inflammatory factors were significantly reduced in AVH mice, indicated the therapeutic efficacy of AR-VLNs (Fig. [108]4E). AR-VLNs repaired intestinal tight junction of CACO-2 cell Pathological sections and electron microscopy results from sleep-deprived mice demonstrated that sleep deprivation caused damage to the intestinal barrier. To investigate the reparative effect of AR-VLNs on intestinal cells, CACO-2 cells were cultured and differentiated, and cell damage was induced using LPS. Changes in cell permeability and ALP content were measured. The results showed a significant increase in macromolecular permeability in LPS-induced injured CACO-2 cells, indicating damage to the mucosal barrier. The ratio of bilateral ALP index also significantly increased, consistent with the previous findings. Administration of AR-VLNs led to a significant reduction in the relevant indices, showing a dose-dependent response (Fig. [109]5A–B). Fig. 5. [110]Fig. 5 [111]Open in a new tab AR-VLNs repaired intestinal tight junction of CACO-2 cell. A Ratio of bilateral ALP index in CACO-2 cell. B FD-4 permeability index in CACO-2 cell. C Representative images of TEM images of CACO-2 cell in different groups. n = 5 in each group of (A–B). n = 3 in each group of (C). Data were presented as mean ± SEM. *P < 0.05, ***P < 0.001 vs the model group. ##P < 0. 01, ###P < 0. 001vs the control group. ns indicated no significant difference vs the model group Electron microscope images revealed structural damage in LPS-treated CACO-2 cells, characterized by reduced microvilli and a scattered arrangement. AR-VLNs were able to restore the damaged microvilli and cell barrier, promoting the recovery of cell structure and preventing the infiltration and damage of macromolecular substances (Fig. [112]5C). AR-VLNs modulated amino acid metabolism in REM sleep deprivation mice Mass Spectrometry analysis was performed in both positive and negative modes (Fig. [113]6A–B), revealed a total of 35 differential metabolites identified under the two modes (Table [114]2). These metabolites exhibited significant differences between REM sleep-deprived mice and control mice, and their concentrations were effectively normalized by AR-VLNs. Volcanic maps of different metabolites of each group were exhibited in Fig. [115]6C–D. Fig. 6. [116]Fig. 6 [117]Open in a new tab Untargeted metabolomics analysis indicated that amino acid metabolism was regulated by AR-VLNs in sleep-deprived mice. (A) Total ion flow diagram of mass spectrum in positive mode. B Total ion flow diagram of mass spectrum in negative mode. C Volcano map between control and model group. D Volcano map between model and AVH group. E PCA analysis of control, model and AVH mice. F PLS-DA analysis of control, model and AVH mice. G OPLS-DA analysis between control and model group. H OPLS-DA analysis between model and AVH group. I 200 times iterative test of (G). J 200 times iterative test of (H). C represented the control group, M represented the model group, G represented the AVH group Table 2. Differential metabolites among the three groups No Compound name HMDB ID RT/min Formula m/z Iron Mode Error/ppm 1 L-Histidine HMDB0000177 13.406 C[6] H[9] N[3] O[2] 156.07693 [M + H]^+ 1.1 2 L-Glutamine HMDB0000641 10.806 C[5] H[10] N[2] O[3] 145.06177 [M−H]^− −0.6 3 5-Oxo-L-proline HMDB0000267 10.802 C[5] H[7] N O[3] 130.04999 [M + H] 0.83 4 Choline HMDB0000097 8.715 C[5] H[13] N O 104.107 [M + H]^+ 0.04 5 L-Proline HMDB0000162 14.581 C[5] H[9] N O[2] 116.07068 [M + H]^+ 0.8 6 4-Hydroxybutyric acid HMDB0000710 1.653 C[4] H[8] O[3] 103.03996 [M−H]^− −0.99 7 Glycolate HMDB0000115 4.037 C[2] H[4] O[3] 75.00867 [M−H]^− −1.24 8 L-Glutamate HMDB0000148 10.494 C[5] H[9] N O[4] 146.0458 [M−H]^− −0.29 9 Xanthine HMDB0000292 8.709 C[5] H[4] N[4] O[2] 151.02605 [M−H]^− −0.56 10 L-Arginine HMDB0000517 13.502 C[6] H[14] N[4] O[2] 173.10422 [M−H]^− 0.64 11 L-Phenylalanine HMDB0000159 8.92 C[9] H[11] N O[2] 166.08642 [M + H]^+ 0.81 12 Propionyl carnitine HMDB0000824 8.451 C[10] H[19] N O[4] 218.13885 [M + H]^+ 0.76 13 Pyrrolidine-2-carboximidate HMDB0253910 14.584 C[5] H[10] N[2] O 115.0867 [M + H]^+ 1 14 4-Oxo-proline HMDB0000267 8.398 C[5] H[7] N O[3] 128.03524 [M−H]^− −0.53 15 Asparagine HMDB0000168 11.028 C[4] H[8] N[2] O[3] 131.04613 [M−H]^− −0.5 16 Hypoxanthine HMDB0000157 8.577 C[5] H[4] N[4] O 135.03121 [M−H]^− 0.97 17 Carnosine HMDB0000033 14.348 C[9] H[14] N[4] O[3] 227.11402 [M + H]^+ 0.23 18 L-Lysine HMDB0000182 14.203 C[6] H[14] N[2] O[2] 145.09816 [M−H]^− 0.28 19 L-Serine HMDB0000187 10.918 C[3] H[7] N O[3] 106.04992 [M + H]^+ −0.3 20 Citric acid HMDB0000094 16.51 C[6] H[8] O[7] 191.01969 [M−H]^− −0.21 21 5-Oxo-D-proline HMDB0304793 10.494 C[5] H[7] N O[3] 128.03524 [M−H]^− 0.03 22 D-Ribulose 5-phosphate HMDB0000618 11.577 C[5] H[11] O[8] P 229.01167 [M−H]^− −0.84 23 Indolelactate HMDB0000671 3.206 C[11] H[11] N O[3] 204.06642 [M−H]^− −0.97 24 N-Acetyl-L-phenylalanine HMDB0000512 1.475 C[11] H[13] N O[3] 206.08213 [M−H]^− −0.62 25 Histamine HMDB0000870 13.425 C[5] H[9] N[3] 110.07223 [M−H]^− −1.36 26 Pilocarpine HMDB0015217 12.175 C[11] H[16] N[2] O[2] 209.12862 [M + H]^+ 0.8 27 Urocanic acid HMDB0000301 13.419 C[6] H[6] N[2] O[2] 137.03551 [M−H]^− −1.05 28 Sphingosine 1-phosphate HMDB0000277 8.424 C[18] H[38] N O[5] P 380.2562 [M + H]^+ 0.14 29 Argininosuccinic acid HMDB0000052 17.053 C[10] H[18] N[4] O[6] 291.13005 [M + H]^+ 0.49 30 Glycine HMDB0000123 10.36 C[2] H[5] N O[2] 74.02468 [M−H]^− −0.99 31 5-Methylcytosine HMDB0002894 8.813 C[5] H[7] N[3] O 126.06631 [M + H]^+ 0.94 32 L-Tryptophan HMDB0000929 8.893 C[11] H[12] N[2] O[2] 203.08252 [M−H]− 0.53 33 N6-Acetyl-L-lysine HMDB0000206 10.102 C[8] H[16] N[2] O[3] 189.12358 [M + H]^+ 0.64 34 Carnitine HMDB0000062 27.072 C[7] H[15] N O[3] 162.1126 [M + H]^+ 0.79 35 N-Acetylornithine HMDB0003357 14.135 C[7] H[14] N[2] O[3] 173.09302 [M−H]^− −0.83 36 Shikimic acid HMDB0003070 8.64 C[7] H[10] O[5] 173.04535 [M−H]^− −1.13 37 D-Lysine HMDB0003405 13.299 C[6] H[14] N[2] O[2] 147.1129 [M + H]^+ −0.76 [118]Open in a new tab Multivariate statistical results of PCA, PLS-DA and OPLS-DA analyses showed significant differences between the control and model mice, as well as between the model and AR-VLNs mice (Fig. [119]6E–H). Besides, 200 times iteration tests results were showed in Fig. [120]6E–H to demonstrate the credibility of the OPLS-DA results. Above results suggested that AR-VLNs had regulated effect on the disordered internal environment of sleep-deprived mice. The heatmap in Fig. [121]7A visualized the expression trends of different metabolites in each group. Pathway enrichment analysis identified that amino acid metabolic pathway was the mainly affected pathway (Fig. [122]7B). Fig. 7. [123]Fig. 7 [124]Open in a new tab Heatmap and pathway enrichment. A Heatmap visualization of control, model and AVH mice. Red dots represent the control group. Blue dots represent the model group. Green dots represent the AVH group. B Pathway enrichment analysis of control, model and AVH mice. n = 6 in each group AR-VLNs regulated the metabolic pool profile of REM sleep deprivation mice Targeted metabolomics was employed to detect endogenous metabolites associated with amino acid metabolism pathways. The levels of amino acid in different positions of REM sleep deprived mice (serum, fecal, intestine) had significant changes. The disturbances of amino acid content in REM sleep-deprived mice that could be regulated by AR-VLNs (Fig. [125]8A–L). Notably, the amino acid metabolic profile changed significantly. In REM sleep deprived mice, the proportion of amino acid content in fecal significantly increased, constituting over 50% of the total amino acid content and becoming the highest proportion within the metabolic pool encompassing serum, tissue, and metabolites. Conversely, in both the control and AR-VLNs groups, fecal amino acid content represented a minor fraction, with the majority of amino acids concentrated in the intestine (Fig. [126]9A–L). The above results indicated that the amino acid absorption or utilization efficiency of REM sleep deprived mice was notably decreased, leading to a substantial excretion of amino acids from the body without intestinal processing. This phenomenon might be attributed to intestinal tissue damage and reduced energy metabolic efficiency, collectively influencing the reabsorption process. Fig. 8. [127]Fig. 8 [128]Open in a new tab Amino acid content determination by targeted metabolomics. A L-Arginine. B L-Glutamic acid. C L-Glutamine. D L-Histidine. E L-Isoleucine. F L-Leucine. G L-Lysine. H L-Methionine. I L-Phenylalanine. J L-Proline. K L-Serine. L Trytophan. n = 8 in each group. Data were presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 vs the model group. #P < 0.05, ##P < 0. 01, ###P < 0. 001vs the control group Fig. 9. [129]Fig. 9 [130]Open in a new tab Metabolic pool profile in different groups of mice. A L-Arginine. B L-Glutamic acid. C L-Glutamine. D L-Histidine. E L-Isoleucine. F L-Leucine. G L-Lysine. H L-Methionine. I L-Phenylalanine. J L-Proline K L-Serine. L Trytophan. n = 8 in each group AR-VLNs regulated the expression of proteins in related pathway Previous experimental results demonstrated that AR-VLNs could interfere with the energy metabolism of sleep-deprived mice through the repair of the intestinal mucosal barrier and regulation of amino acid metabolism. To verify these conclusions, additional molecular biology experiments such as PCR and WB were conducted. One-step RT-PCR was used to detect the gene expression of tight junction-related proteins in the ileum tissues of REM sleep-deprived mice. The results showed a significant reduction in the expression of Claudin-1, Occludin, and ZO-1 in the model group, providing evidence of tight junction damage. Following administration of AR-VLNs, the expression of these genes significantly increased, confirming the effect of AR-VLNs on the intestinal tight junctions (Fig. [131]10A-C). WB results demonstrated lower expression of tight junction-related proteins in REM sleep-deprived mice, while AR-VLNs restored the protein content of Occludin and ZO-1 (Fig. [132]10D). In an in vitro validation, total protein was extracted from CACO-2 cells for WB analysis. Consistent with the results obtained from animal tissues, the content of tight junction-related proteins decreased in the model group cells and significantly increased after administration of AR-VLNs (Fig. [133]10E). These results confirmed that AR-VLNs could repair the intestinal mucosal barrier at the molecular biological level. Fig. 10. [134]Fig. 10 [135]Open in a new tab AR-VLNs regulated the expression of tight junction-related mRNA and proteins in REM sleep deprived mice and CACO-2 cell. A The mRNA expression of ZO-1 in sleep deprived mice. B The mRNA expression of Occludin in sleep deprived mice. C The mRNA expression of Claudin-1 in sleep deprived mice. D Western blot analysis of ZO-1 and Occludin in mice. E Western blot analysis of ZO-1 and Occludin in CACO-2 cell. n = 6 in each group of (A-C). n = 3 in each group of (D–E). Data were presented as mean ± SEM. *P < 0.05, **P < 0.01 vs the model group. #P < 0.05, ##P < 0. 01 vs the control group To confirm the effect of AR-VLNs on the amino acid metabolism pathway, RT-PCR and WB were employed to detect the expression of related genes and proteins in the mouse intestine. The PCR results are shown in Fig. [136]11A–F, while the gray value results depicted in Fig. [137]11G illustrated the differential expression of amino acid transporters between the model group and the control group. AR-VLNs were found to regulate the expression trends of the relevant genes and proteins, aligning them with those observed in the control group, respectively. Fig. 11. [138]Fig. 11 [139]Open in a new tab AR-VLNs regulated the expression of amino acid transporter mRNA and proteins in REM sleep deprived mice. A The mRNA expression of LAT1. B The mRNA expression of SLC3A2. C The mRNA expression of SLC6A19. D The mRNA expression of SLC15A1. E The mRNA expression of SLC15A2. F The mRNA expression of SLC38A5. G Western blot analysis of SLC6A19 and SLC38A5 in mice. n = 6 in each group of (A–F). n = 3 in each group of (G). Data were presented as mean ± SEM. *P < 0.05, **P < 0.01 vs the model group. #P < 0.05, ##P < 0. 01 vs the control group Above results showed that intestinal damage in REM sleep deprivation mice caused amino acid metabolism disorder (Fig. [140]12), which could be reverted by AR-VLNs. Fig. 12. [141]Fig. 12 [142]Open in a new tab Graphical illustration of the drug action mechanism Discussion Sleep deprivation would lead to pathological changes in the body, manifested as lysosome morphological damage, brain-gut axis function weakens, resulted in energy metabolism imbalance [[143]20]. Sleep deprivation also inhibited glucose metabolism, fatty acid metabolism and amino acid metabolism, which led to energy metabolism disruption [[144]7]. This is of particular concern as energy metabolism is fundamental to the maintenance of cellular and physiological functions. In this study, REM sleep-deprived mice exhibited disturbances in redox and energy metabolism, as evidenced by decreased exercise capacity, weakened muscle strength, significantly increased oxidative stress indices, and reduced ATP content. Further observations revealed significant changes in intestinal pathology, including inflammatory infiltration, intestinal villi injury, disruption of the intestinal mucosal barrier, and reduced content of intestinal tight junction proteins in sleep-deprived mice. These results are consistent with the existing literature. AR-VLNs were found to alleviate these symptoms, repair the intestinal mucosal barrier of REM sleep-deprived mice, and enhance energy metabolism and antioxidant capacity. Furthermore, through experiments conducted on CACO-2 cells, this study confirmed that AR-VLNs effectively restored intestinal epithelial barrier function, alleviated the increased cellular permeability and macromolecular substance penetration caused by intestinal cell inflammation. Metabolomics was utilized to investigate the effects of REM sleep deprivation on mice. The metabolic environment of REM sleep-deprived mice underwent alterations, with pathway enrichment indicating a predominant disruption in amino acid metabolism. The metabolic milieu of AR-VLNs mice and REM sleep-deprived mice experienced significant changes, exerting regulatory influence on the aforementioned pathway. Heat map analysis demonstrated that AR-VLNs normalized the levels of differential metabolites affected by sleep deprivation disturbances. Amino acids are essential for various biological processes. The disturbance of energy metabolism and the imbalance of amino acid balance often occur simultaneously [[145]15]. Tricarboxylic acid cycle is a crucial process for energy metabolism and redox homeostasis in the body. This process promoted the convergence of amino acid metabolism and facilitated biological interactions [[146]24]. Lysosomes released amino acids to decompose products of the tricarboxylic acid cycle, thereby replenished organic elements of mitochondrial metabolism and maintained metabolic homeostasis [[147]8]. Exogenous small molecules could significantly improve energy metabolism by enhancing the intensity of the tricarboxylic acid cycle and regulating amino acid levels [[148]17]. Additionally, amino acid metabolism could ameliorate mitochondrial dysfunction by modulating succinyl-CoA levels [[149]30]. Amino acid transferase promoted branched-chain amino acid metabolism, which in turn contributed to mitochondrial metabolism. [[150]21]. Combined with the literature, the regulation of amino acid metabolism can enhance the energy metabolic cycle, which might be the main mechanism that AR-VLNs ameliorated fatigue in REM sleep-deprived mice and enhanced their physical and behavioral performance. Amino acid absorption and metabolism are closely related to the structure and function of intestines. Amino acids can pass through the intestinal barrier for biological metabolism and transmembrane transport in intestinal cells [[151]23]. The integrity of the intestinal mucosal barrier greatly influenced the efficiency of amino acid absorption [[152]33]. It was reported that prolonged parenteral eutrophication had relationship with intestinal atrophy and weakened barrier function, leading to abnormal amino acid availability [[153]11]. Intestinal epithelial cells played a role in maintaining the homeostasis of amino acids in the systemic circulation, and a normal intestinal barrier limited the transport of excess amino acids into the basal layer of cells [[154]25]. In this investigation, we employed targeted metabolomics to assess the amino acid levels in the serum, intestinal tissue and feces of mice across various experimental groups. The results revealed notable alterations in the amino acid metabolic profile of REM sleep-deprived mice, particularly with a significantly elevation in fecal amino acid content, which comprised half of the total metabolic pool. This indicated that amino acid absorption efficiency was significantly decreased among REM sleep-deprived mice, resulted in substantial excretion of amino acids. Based on prior experimental evidence of intestinal mucosal damage by sleep deprivation, we hypothesized a cascade of events within the mouse model group: Sleep deprivation induced weakened metabolic capacity and vitality in mice, led to intestinal mucosal injury, diminished amino acid reabsorption efficiency by intestinal epithelial cells, induced amino acid excretion, thereby disrupted the energy metabolism process and exacerbated mouse frailty. Lipid components and miRNA of AR-VLNs were capable of modulating amino acid metabolism. Additionally, AR-VLNs were located on mouse intestinal tissues and demonstrated reparative properties on the intestinal mucosa. Consequently, AR-VLNs offered an intervention to ameliorate the altered metabolic pool distribution in REM sleep-deprived mice, enhanced intestinal tissue amino acid absorption efficiency, and exerted anti-sleep deprivation effects. In order to further confirm the above conclusions, molecular biology experiments were conducted, and the results showed that the expression of intestinal tight-connection-related genes and proteins decreased in REM sleep deprived mice, while the expression of most amino acid transporter proteins increased. It might be due to the negative feedback regulation mechanism of the body in response to amino acid loss, consistent with the increase of amino acid content in the serum. AR-VLNs could improve the expression of genes and proteins related to intestinal tight junctions, repair intestinal tissue and cell tight junctions, and regulate the amino acid transport of intestinal. These findings were consistent with previous reports in the literature, and provided an explanation for the abnormal amino acid metabolism and damage on intestinal observed in REM sleep deprived mice, as well as the pharmacodynamic mechanism of AR-VLVs. This study also had the following limitations. First, the relationship between abnormal expression of amino acid transporters and specific amino acid has not been accurately verified. Secondly, the specific targets of AR-VLNs remain to be clarified. Finally, whether there is a fusion relationship between the lipid membrane structure of AR-VLNs and mucosal is still to be explored. Therefore, we had the following suggestion for future research. First, accurate studies on the interaction between targets and small molecules could be conducted. Secondly, the role of components encapsulated in AR-VLNs is still an interesting research point. Finally, the research of AR-VLNs membrane structure also has a certain prospect. Conclusion AR-VLNs could repair the intestinal mucosal barrier, maintain the balance of amino acid content, alleviate amino acid metabolism disorders, regulate energy metabolism, and intervene in the decline of bodily functions caused by REM sleep deprivation. This suggests that AR-VLNs could be a promising therapeutic agent for addressing the multifaceted adverse effects induced by REM sleep deprivation. Supplementary Information [155]Supplementary material 1.^ (924.7KB, docx) [156]Supplementary material 2.^ (45.1KB, docx) [157]Supplementary material 3.^ (9.4KB, xlsx) [158]Supplementary material 4.^ (13.1KB, xlsx) [159]Supplementary material 5.^ (12.7KB, xlsx) [160]Supplementary material 6.^ (11.2KB, xlsx) [161]Supplementary material 7.^ (11.6MB, xlsx) [162]Supplementary material 8.^ (41.4MB, xlsx) [163]Supplementary material 9.^ (860KB, xlsx) [164]Supplementary material 10.^ (10.9MB, xlsx) [165]Supplementary material 11.^ (30.1KB, xlsx) Acknowledgements