Abstract

   Developmental neurotoxicity (DNT) induced by sevoflurane exposure poses
   significant risks to pediatric anesthesia, yet effective protective
   strategies remain limited. Here, we developed self-assembling
   Angiopep-2/SIRT1 nanoparticles (Ang/SIRT1-NPs) with favorable
   biocompatibility and brain-targeting properties. Through in vitro and
   in vivo studies, we demonstrate that Ang/SIRT1-NPs effectively
   alleviate sevoflurane-induced neuronal apoptosis, neuroinflammation,
   and dendritic spine loss. Multi-omics analyses identified
   SIRT1-mediated suppression of necroptosis and oxidative stress pathways
   as key mechanisms underlying neuroprotection. Behavioral assays further
   confirmed improved cognitive and motor function in nanoparticle-treated
   mice. Our findings highlight the potential of Ang/SIRT1-NPs as a
   promising neuroprotective strategy for preventing anesthesia-related
   DNT and support their translational application in pediatric
   neuroprotection.

Graphical abstract

   [30]graphic file with name 12951_2025_3639_Figa_HTML.jpg

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s12951-025-03639-w.

   Keywords: Sevoflurane, Developmental neurotoxicity, Angiopep-2/SIRT1
   nanoparticles, Necroptosis, Multi-omics

Introduction

   Sevoflurane is a volatile anesthetic commonly used in clinical
   anesthesia due to its ability to induce a rapid and controllable
   anesthetic state [[31]1, [32]2]. However, increasing research indicates
   that sevoflurane may have adverse effects on the developing brain,
   particularly in infants and children [[33]3, [34]4]. The issue of
   sevoflurane-induced developmental neurotoxicity (DNT) has garnered
   significant attention because it can lead to long-term cognitive
   impairment and behavioral abnormalities [[35]5–[36]7]. The developing
   brain is extremely vulnerable, and external factors such as the use of
   anesthetics can profoundly affect neuronal growth and differentiation
   [[37]8]. Therefore, studying the mechanisms of sevoflurane-induced DNT
   is crucial for protecting children’s brain health [[38]5, [39]9,
   [40]10].

   Current research on sevoflurane-induced DNT primarily focuses on its
   induction of neuronal apoptosis and inflammatory responses [[41]3,
   [42]11]. Necroptosis, a form of programmed cell death, plays a crucial
   role in sevoflurane-induced neurotoxicity [[43]12–[44]14]. Multi-omics
   approaches, such as high-throughput sequencing, proteomics, and
   metabolomics, have been widely employed to elucidate the molecular
   mechanisms underlying sevoflurane-induced DNT [[45]15, [46]16].
   However, despite numerous studies revealing the direct impact of
   sevoflurane on neural cells, its precise molecular mechanisms remain
   incompletely understood [[47]5, [48]9, [49]17]. Particularly in
   clinical settings, effective interventions are still lacking [[50]5,
   [51]18, [52]19]. Therefore, exploring new therapeutic strategies and
   molecular targets is of paramount importance for mitigating
   sevoflurane-induced DNT.

   Sirtuin 1 (SIRT1) is a crucial deacetylase that plays a key role in
   neuroprotection and anti-apoptosis [[53]20]. Studies have shown that
   SIRT1 regulates cell survival and apoptosis by deacetylating specific
   target proteins [[54]21, [55]22]. SIRT1 is particularly recognized for
   its protective effects in inflammatory responses and necroptosis
   [[56]23–[57]25]. Its neuroprotective role has been widely acknowledged
   in various neurotoxic and neurodegenerative disease studies [[58]26,
   [59]27]. For instance, research on Alzheimer’s and Parkinson’s diseases
   indicates that SIRT1 can reduce neuronal damage and improve cognitive
   function [[60]28–[61]30]. Therefore, SIRT1 presents a promising
   therapeutic target with broad application potential [[62]28, [63]31,
   [64]32].

   In this context, self-assembled Angiopep-2/SIRT1 nanoparticles have
   been introduced as a novel drug delivery system in the study of
   sevoflurane-induced DNT. Angiopep-2 is a peptide capable of crossing
   the blood-brain barrier (BBB), and when combined with SIRT1, it forms
   nanoparticles with excellent stability and biocompatibility
   [[65]33–[66]35]. The application of nanoparticle technology in drug
   delivery, particularly in crossing the BBB, shows immense potential
   [[67]36–[68]38]. Through targeted delivery using nanoparticles, drugs
   can more efficiently reach damaged brain regions and exert their
   therapeutic effects [[69]39]. Previous studies have demonstrated the
   significant efficacy of nanoparticles in treating neurological diseases
   such as stroke and brain tumors [[70]40–[71]42]. Therefore,
   Angiopep-2/SIRT1 nanoparticles exhibit innovative and important
   application value in the research of sevoflurane-induced DNT.

   This study aims to investigate the molecular mechanisms by which
   self-assembled Angiopep-2/SIRT1 nanoparticles regulate necroptosis in
   sevoflurane-induced DNT. Through both in vivo and in vitro experiments,
   we evaluate the stability, biocompatibility, and intervention effects
   of Angiopep-2/SIRT1 nanoparticles on sevoflurane-induced DNT.
   Multi-omics analysis is employed to uncover the key genes, proteins,
   and related metabolic pathways involved in the nanoparticles’
   amelioration of DNT. The scientific significance of this study lies in
   elucidating the molecular mechanisms of sevoflurane-induced DNT,
   proposing a novel therapeutic strategy, and providing new molecular
   targets for clinical application. This research has the potential for
   significant clinical value, offering new therapeutic approaches to
   enhance the safety of pediatric anesthesia.

Materials and methods

Preparation of Angiopep-2/SIRT1 nanoparticles

   Angiopep-2/SIRT1 nanoparticles were prepared using the anti-solvent
   precipitation method. To prepare the SIRT1 solution, 10 mg of SIRT1
   (ab287914, Abcam) was added to 10 ml of distilled water, and the
   mixture was stirred at room temperature for 1 h. The pH was adjusted to
   7.0 using 0.1 M HCl or NaOH. The protein solution was then heated to
   90 °C or higher for 30 min, resulting in a final concentration of
   1 mg/ml SIRT1. Separately, 1 mg of Angiopep-2 and 1 mg of SMCC (22360,
   Thermo, USA) were each dissolved in 1 ml of water, creating Angiopep-2
   (1 mg/ml) and SMCC (1 mg/ml) solutions. Add 1 mL of water to a 2 mL
   centrifuge tube, followed by the addition of 100 µL of Angiopep-2
   solution. Stir vigorously. Next, add 100 µL of SMCC solution to the
   mixture and stir for 3 h at 1500 rpm in an ice bath to prepare the
   Angiopep-2-SMCC solution. Withdraw 100 µL of the Angiopep-2-SMCC
   solution and slowly add it to the SIRT1 solution. Stir vigorously for
   2 h in an ice bath to obtain the Angiopep-2/SIRT1.

   Transmission electron microscopy (TEM) and scanning electron microscopy
   (SEM) were used to analyze the Angiopep-2/SIRT1 nanoparticles. The
   morphology of the Angiopep-2/SIRT1 nanoparticles was examined using a
   TEM (TECNAI G2 F20, FEI, USA) at an acceleration voltage of 30 kV. The
   nanoparticles were dried using a freeze dryer. For SEM analysis, the
   samples were sputter-coated for 45 s at 20 mA and observed with an SEM
   (Quanta 250, FEI, Hillsboro, OR, USA) to assess cross-sectional
   morphology and serum stability.

   For flow cytometry analysis of SIRT1, SH-SY5Y cells were treated with
   0.25% trypsin (25200072, Gibco, USA) and counted. Approximately
   1 × 10^6 cells were resuspended in 200 µL of fluorescence-activated
   cell sorting (FACS) buffer (660585, BD Biosciences, USA). Fluorescent
   antibodies (2 µL each) were added, and the cells were incubated on ice
   for 30 min. Subsequently, the SH-SY5Y cells were washed with FACS
   buffer, fixed with 10% formalin ([72]R04587, Merck, USA), and analyzed
   for antigen positivity using a BD FACSCalibur flow cytometer (BD
   Biosciences, USA). The antibody used was a fluorescently labeled SIRT1
   antibody (PA5-85921, 1:200, Invitrogen, USA).

Stability assessment of Angiopep-2/SIRT1 using dynamic light scattering (DLS)

   DLS was employed to measure the size, polydispersity index (PDI), and
   surface zeta potential of the Angiopep-2/SIRT1 nanoparticles. The
   measurements were conducted using a nanoparticle size and zeta
   potential analyzer (Nano-ZS90, Malvern Instruments, UK) equipped with a
   633 nm He-Ne laser. All experiments were performed at 25 °C and
   repeated three times. To investigate the physiological stability of the
   nanoparticles, they were resuspended in water, saline,
   phosphate-buffered saline (PBS), and RPMI 1640 and incubated for 24 h.
   Additionally, Angiopep-2/SIRT1 was resuspended in solutions with
   different pH values (4.5, 5.5, 6.5, 7.5, and 8.5) and incubated for
   24 h. After incubation, DLS was used to measure the size, distribution,
   and scattering intensity of the nanoparticles. SEM (S-4800, Hitachi,
   purchased from Shanghai Fulai Optical Technology Co., Ltd.) was
   employed to observe the morphology of the nanoparticles.

In vitro biocompatibility assessment

   Live/Dead Cell Staining: SH-SY5Y cells at a density of 1 × 10^6 were
   seeded into 12-well culture plates and co-cultured with PBS, 0.5 µM
   SIRT1, and 0.5 µM Angiopep-2/SIRT1 samples for 1, 2, and 3 days. A
   live/dead cell staining solution (ZY140632, Zeye Biotechnology, China)
   was prepared in a ratio of 1 mL PBS: 3 µL calcein-AM: 5 µL PI, then
   added to each group and incubated at 30 °C for 37 min. The live (green)
   and dead (red) cells in each field of view were then counted using a
   fluorescence microscope (IMT-2, Olympus, Japan).

   CCK-8 Assay: SH-SY5Y cells at a density of 1 × 10^6 were co-cultured
   with CON, SIRT1, and Angiopep-2/SIRT1 samples for 1, 2, and 3 days.
   After the incubation period, 100 µL of CCK-8 solution (C0037, Beyotime)
   was added to each well and incubated for 2 h. Subsequently, 100 µL of
   the supernatant was transferred to a 96-well plate, and the absorbance
   was measured at 450 nm using a microplate reader (BioTek Synergy Neo2
   Hybrid, Agilent, USA).

Hemolysis assay

   To conduct the hemolysis assay of Angiopep-2/SIRT1 nanoparticles,
   prepare solutions with varying concentrations of Angiopep-2/SIRT1
   nanoparticles, ranging from 100 µg/mL to 3.13 µg/mL. Use PBS as the
   negative control and deionized water as the positive control. Fresh
   human whole blood containing an anticoagulant (EDTA) (Human Whole
   Blood, IPHASE, Suzhou, China) is then mixed with the different
   concentrations of nanoparticle solutions and control liquids in
   specific proportions. The mixtures are incubated at 37 °C for 3 h.
   After incubation, the samples are centrifuged at 2000 g for 10 min to
   separate the undissolved cellular components (CC) from the plasma. The
   supernatant is collected, and the absorbance is measured at 540 nm
   using a spectrophotometer to quantify the hemoglobin content dissolved
   in the blood.

Evaluation of Angiopep-2/SIRT1 crossing the BBB

   Cellular Assay: HBMEC cells were cultured in a Transwell system.
   Successful formation of the BBB was confirmed when the transendothelial
   electrical resistance reached 150 ohms×cm². Subsequently, 0.5 µM SIRT1
   and 0.5 µM Angiopep-2/SIRT1 were added to the upper chamber of the
   Transwell. After 4 h, the fluorescence intensity of FITC (green) in
   HBMEC cells (upper chamber) and SH-SY5Y cells (lower chamber) incubated
   with Angiopep-2/SIRT1/FITC was detected using a confocal laser scanning
   microscope (CLSM) (FV3000, Olympus, Japan). The HBMEC cell membranes
   were labeled with Dil fluorescent dye (red), and the SH-SY5Y cells were
   stained with DAPI (blue). After a 4-hour incubation, the fluorescence
   intensity in the upper and lower chambers of each group was measured
   using a microplate reader. When red and green fluorescence were
   simultaneously present in the images, the red fluorescence was
   pseudocolored to purple. The penetration ratio of SIRT1 and
   Angiopep-2/SIRT1 across the BBB was determined by calculating the ratio
   of the fluorescence intensity in the lower chamber to the total
   fluorescence intensity in both chambers using the following formula:
   Penetration Rate (%) = Fluorescence Intensity in the Upper
   Chamber/(Fluorescence Intensity in the Upper Chamber + Fluorescence
   Intensity in the Lower Chamber) × 100%.

   Animal Experiment: Angiopep-2/SIRT1 nanoparticles were labeled with
   Cy5.5 NHS and injected into mice via the tail vein at a dose of
   10 mg/kg in a 100 µL volume. The distribution of the labeled
   Angiopep-2/SIRT1 was observed at 6, 12, 24, and 48 h post-injection
   using the IVIS Lumina Series III small animal imaging system
   (PerkinElmer, USA). This experimental protocol and animal study were
   approved by the hospital’s Institutional Animal Care and Use Committee
   (IACUC). All animal experiments in this study were conducted in strict
   accordance with institutional ethical guidelines and approved by the
   Institutional Animal Care and Use Committee (IACUC). Mice were housed
   in a standard SPF-grade animal facility under controlled environmental
   conditions (temperature: 22–25 °C; humidity: 60–65%; 12-hour light/dark
   cycle) to ensure their basic physiological needs. All procedures were
   performed by professionally trained personnel using appropriate
   anesthesia and analgesia to minimize pain and stress. Animal health was
   regularly monitored, and humane endpoints were applied when necessary.
   Throughout the study, the principles of Replacement, Reduction, and
   Refinement (3Rs) were followed to ensure ethical use of animals while
   enhancing the scientific validity and reproducibility of the
   experimental data.

Construction and grouping of sevoflurane-induced DNT mouse model

   SPF-grade C57BL/6J mice (8–12 weeks old, weighing 20–30 g) were
   obtained from Beijing Vital River Laboratory Animal Technology Co.,
   Ltd., Beijing, China. All mice were housed in an SPF-grade animal
   laboratory with controlled conditions: a 12-hour light/dark cycle,
   60–65% humidity, and a temperature range of 22 to 25 °C. The mice had
   free access to food and water. After a one-week acclimation period,
   during which their health status was monitored, the experiment
   commenced. To facilitate mating, 1 to 3 female mice (8–12 weeks old)
   were housed with one proven fertile male mouse. Vaginal plugs were
   checked daily between 08:00 and 10:00 as evidence of successful mating.
   The presence of a vaginal plug marked the first day of pregnancy. The
   pregnant female mice were then separated from the male and housed
   individually under standardized conditions until delivery.

   Silencing SIRT1: Lentiviral packaging kits (V48820, Invitrogen, USA)
   were used to transfect HEK-293T cells (CRL-3216, ATCC) with control
   plasmid sh-NC, sh-SIRT1-1, and sh-SIRT1-2, constructed using the pLKO.1
   vector from Shanghai HanHeng Biotechnology Co., Ltd. After 48 h of
   transfection, the supernatant was collected and concentrated using a
   lentivirus concentrator (631231, Takara). The concentrated lentivirus
   was stored at −80 °C. The shRNA sequences are listed in Table S1.

   Establishment of sevoflurane-induced DNT mouse model: Newborn mice with
   SIRT1 gene silencing were subjected to lentivirus intervention via tail
   vein injection (100 µL, 1 × 10^11 PFU/mL) on days 1, 3, and 5. After
   the injections, on postnatal days 6, 8, and 10 (PNDs), mice in the
   sevoflurane group were placed in a sealed chamber. Using the
   Datex-Ohmeda anesthesia system (Madison, WI, USA), the mice were
   anesthetized with a mixture of 3% sevoflurane and 60% oxygen at a flow
   rate of 2 L/min for induction [[73]5], followed by maintenance at
   1 L/min for 2 h. The control group was exposed to 60% oxygen without
   sevoflurane for 2 h. The concentrations of sevoflurane and oxygen were
   continuously monitored (Vamos, Dräger Medical, Germany), and the mice
   were kept on a heating pad at 37 ± 1 °C throughout the procedure. After
   sevoflurane exposure, the mice were immediately returned to their home
   cages and given standard care. Angiopep-2/SIRT1 nanoparticles were
   administered at a dose of 10 mg/kg via tail vein injection (100 µL each
   time) on PND 27 and PND 29. Detailed group information can be found in
   Table S2 [[74]43].

Hematoxylin and eosin (H&E) staining and TEM

   Mouse brain tissue fixed in paraformaldehyde was dehydrated, embedded
   in paraffin, and sectioned into 5 μm thick slices. Continuous sections
   were taken from the internal and external compartments at 200 μm
   intervals. Selected sections were deparaffinized in xylene and
   rehydrated through a series of ethanol gradients. For H&E staining,
   samples were stained in Harris alum hematoxylin solution (ST2067,
   Beyotime) for 5 min. They were then washed with 0.5% hydrochloric acid
   ethanol (C0165S, Beyotime) for 10 s, followed by staining with eosin
   solution (C0109, Beyotime) for 40 s. Finally, the sections were
   dehydrated, cleared, and mounted using neutral resin for observation
   under a light microscope. For TEM analysis, following H&E staining, the
   striatum was cut into 1 mm³ cubes and fixed in 2.5% glutaraldehyde
   (G5882, Sigma, USA). After primary fixation, the samples were
   post-fixed in 1% osmium tetroxide solution (1.60333, Sigma, USA) to
   enhance electron density and contrast. The samples were then dehydrated
   through a graded ethanol series, infiltrated, and embedded in acrylic
   resin. Ultrathin sections of 70 nm thickness were cut using an
   ultramicrotome and collected on grids for TEM analysis.

Nissl staining

   Nissl staining was performed according to standard protocols. After
   deep anesthesia, mice were perfused transcardially, and their brains
   were extracted, fixed in formaldehyde, dehydrated, and embedded in
   paraffin. The brain tissues were then sectioned into 6 μm slices for
   staining. The sections were deparaffinized and stained with Nissl
   staining solution (C0117, Beyotime, China) at 60 °C for 45 min.
   Following staining, the sections were soaked in absolute ethanol for
   2 min and then in xylene for 10 min before being coverslipped. Images
   were captured using the ScanScope scanning system (Olympus, Japan) and
   analyzed with Image-Pro Plus 6.0 software (Media Cybernetics, Silver
   Spring, USA).

Golgi staining

   Golgi staining was performed using the FD Rapid Golgi Stain Kit (PK401,
   FD NeuroTechnologies, United States) according to the manufacturer’s
   instructions. Brain samples were collected and immersed in a mixture of
   Solutions A and B (1:1) at room temperature in the dark for 14 days.
   Subsequently, the samples were transferred to Solution C and kept in
   the dark for 5 days. The samples were then embedded in an optimal
   cutting temperature compound and frozen at −80 °C for 24 h. Using a
   cryostat (Leica, Wetzlar, Germany), the brains were sectioned coronally
   at 100 μm thickness. The staining procedure was carried out as per the
   manufacturer’s instructions. The slides were observed using an optical
   microscope equipped with a 63× oil immersion objective (Olympus,
   FV1200). Pyramidal neurons in the CA1 region that were well-impregnated
   and distinctly separated from other neurons were analyzed. From each
   pyramidal neuron, five basal dendritic segments, each 30 μm or longer,
   were randomly selected. Dendritic spine density was analyzed using
   ImageJ software.

Short-term neurological assessment

   This study employed random group assignment, with mice randomly
   allocated to each experimental group to ensure baseline comparability
   between groups. In addition, all behavioral tests were conducted by
   well-trained observers who were blinded to the group assignments to
   prevent subjective bias. On days 31, 35, and 37, after the construction
   of the mouse model, 6 mice from each group were randomly selected for
   behavioral tests to evaluate short-term neurological function. The
   tests included negative geotaxis and front-limb suspension. Each mouse
   was subjected to three trials for each test, and the average of the
   three scores was recorded.

   For the negative geotaxis test, Each mouse was placed on an inclined
   board (30 cm long, 45° inclination) with its head facing downward.
   After holding the mouse in place for 5 s, it was released. The time
   taken for the mouse to turn its head upwards within 30 s was recorded.
   For the front-limb suspension Test, Each mouse’s forepaws were placed
   on a suspended wire, and the time it took for the mouse to fall from
   the wire within a 60-second period was recorded.

Long-term neurological assessment

   After completing the treatments, 6 mice from each group were randomly
   selected to undergo two behavioral tests to evaluate long-term
   neurological function.

   Novel object recognition test (NORT): Before the test, mice were placed
   in an open field (60 × 40 × 40 cm) for 10 min per day for three
   consecutive days to acclimate them to the testing environment. At the
   beginning of the test, mice were placed in the open field for 10 min.
   After this acclimation period, two identical balls of the same color
   were introduced into the open field. The duration of interaction with
   the balls, including nose or mouth contact, was recorded for 20 min.
   Between trials, the apparatus and objects were cleaned with 50% ethanol
   to eliminate any olfactory cues. One hour after the initial test, a
   second test was conducted by replacing one of the familiar objects with
   a novel object of a different shape and color. The mice were then
   observed for another 20 min. The time spent investigating the novel
   object versus the familiar object was recorded. The discrimination
   ratio was calculated as follows:

   Discrimination Ratio = Time Spent on Novel Object/(Time Spent on Novel
   Object + Time Spent on Familiar Object).

   Morris water maze test (MWMT): The MWMT was conducted in a black
   cylindrical tank (120 cm in diameter, 60 cm deep). An animal tracking
   system was used to monitor the movements of the mice during the test.
   The water maze was divided into four equal quadrants, with a platform
   submerged just below the water surface in the center of the third
   quadrant. For five consecutive days prior to the test, all mice
   underwent training to locate the platform. During the training, mice
   from different quadrants were placed in the maze, and the time taken to
   find the platform was recorded using tracking software. On the sixth
   day, the platform was moved to a new location, and the mice were placed
   in the maze to swim freely for 60 s. The time taken for each mouse to
   find the new platform location was recorded. The SMART video tracking
   system consists of both hardware and analysis software components. It
   measures indices of memory behavior and automatically tracks and
   monitors the mice. Using color image processing algorithms, the system
   tracks the movement trajectories of the mice in real-time. It can
   analyze the paths and time spent in each of the four quadrants, the
   efficiency over six cycles, orientation angle, average swimming speed,
   time spent crossing the virtual platform, and time spent on the virtual
   platform.

High-throughput RNA sequencing

   Brain tissues were collected from the DNT group (n = 3) and the
   Angiopep-2/SIRT1 group (n = 3) for RNA sequencing. The sequencing
   libraries were generated and sequenced by CapitalBio Technology
   (Beijing, China). A total of 5 µg of RNA was used for each sample.
   Briefly, ribosomal RNA (rRNA) was removed from the total RNA using the
   Ribo-Zero Magnetic Kit (MRZG12324, Epicentre, USA). Sequencing
   libraries were constructed using the NEBNext Ultra RNA Library Prep Kit
   for Illumina (E7760S, NEB, USA). RNA was then fragmented into
   approximately 300 bp (bp) fragments in NEBNext First Strand Synthesis
   Reaction Buffer (5x). First-strand cDNA was synthesized using reverse
   transcriptase and random primers. The second strand was synthesized in
   the presence of dUTP Mix (10x) to create strand-specific libraries. The
   cDNA fragments were end-repaired, a polyA tail was added, and
   sequencing adapters were ligated. After ligation of Illumina sequencing
   adapters, the second strand of cDNA was digested with USER Enzyme
   (M5508, NEB, USA) to construct a strand-specific library. The library
   DNA was amplified, purified, and enriched via PCR. The libraries were
   then validated using the Agilent 2100 Bioanalyzer and quantified with
   the KAPA Library Quantification Kit (kk3605, Merck, USA). Finally,
   paired-end sequencing was performed on the Illumina NextSeqCN500
   sequencing platform.

RNA sequencing data analysis

   The quality of the raw paired-end sequencing reads was assessed using
   FastQC software v0.11.8. Preprocessing of the raw data was conducted
   using Cutadapt software v1.18 to remove Illumina sequencing adapters
   and poly(A) tail sequences. Reads with more than 5% N content were
   removed using a Perl script. The FASTX Toolkit v0.0.13 was then used to
   extract reads with at least 70% of bases having a quality score above
   20. Paired-end sequences were repaired using BBMap software. Finally,
   the filtered high-quality reads were aligned to the mouse reference
   genome using Hisat2 software v0.7.12.

   Differential Analysis: The raw counts matrix was normalized using the
   logCPM (Counts Per Million) method provided by the edgeR package.
   Multiple comparisons were adjusted using the Benjamini-Hochberg method
   (i.e., FDR correction). Genes with adjusted p-values (adjP) < 0.05 and
   |log2FC| >2 were considered differentially expressed. A heatmap of
   intersecting gene expression was generated using the heatmap package in
   R.

   Lasso regression modeling was conducted with the “glmnet” package in R.
   By setting the penalty coefficient λ, Lasso regression enables
   compression estimation and variable selection of metabolites, thereby
   identifying genes significantly associated with osteoarthritis (OA). In
   the Lasso regression model, the optimal value of λ was determined
   through cross-validation using the “cv.glmnet” function in R, ensuring
   the model’s best predictive ability. After analysis, a coefficient path
   plot was generated to illustrate the changes in gene variable
   coefficients across different λ values using the “plot.glmnet”
   function, and significant differential genes were extracted as key
   features.

   First, a Support Vector Machine (SVM) model was implemented using the
   “e1071” package in R. In the SVM model, the Radial Basis Function (RBF)
   was used as the kernel function. Model parameters, such as the cost
   parameter C and the kernel parameter γ, were optimized using the grid
   search method via the “tune.svm” function. Next, the Recursive Feature
   Elimination (RFE) strategy was applied to optimize the feature subset
   by recursively removing the least contributing features. In each
   iteration, the importance of each feature was assessed using the weight
   coefficients of the SVM model, and the least important features were
   removed until a predefined number of features was reached.

   Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG)
   analyses were conducted using the “clusterProfiler”, “org.Hs.eg.db”,
   “enrichplot”, “DOSE”, and “ggplot2” packages in R.

Quantitative proteomics detection

   Brain tissues from DNT group (n = 3) and Angiopep-2/SIRT1 group (n = 3)
   mice were collected. Tumor tissues were extracted using RIPA buffer
   containing protease inhibitors (87786, Thermo, USA). During extraction,
   samples were sonicated for 30 s every 5 minutes, repeating this process
   three times to ensure complete cell lysis and protein release. The
   concentration of the extracted protein samples was determined using a
   BCA Protein Assay Kit (23225, Thermo Fisher, USA), ensuring each
   sample’s concentration fell within the acceptable range based on the
   standard curve. After adjusting the pH to 8.0, trypsin was added at a
   ratio of 1:50 (enzyme to protein) and incubated at 37 °C for 16 h for
   protein digestion. After enzymatic digestion, the samples were cleaned
   using ZipTip C18, loaded onto a high-performance liquid chromatography
   (HPLC) system, and connected to a mass spectrometer for mass
   spectrometry (MS)/MS analysis. Data processing, including protein
   identification and quantification, was performed using MaxQuant
   software. Desalted peptides were labeled with iTRAQ reagent containing
   0.1% formic acid (FA) for MS analysis. Each sample was analyzed three
   times using the QSTAR Elite Hybrid MS (Applied Biosystems/MDS-SCIEX)
   and an online HPLC system (Shimadzu, Japan). For each analysis, 30 µl
   of peptide solution was injected and separated on a homemade nanopore
   C18 column with a micro-nano spray (75 μm ID × 15 cm, 5 μm particles)
   (New Objectives, Wubrun, MA). A 90-minute HPLC gradient was established
   using mobile phase A (0.1% FA/2% ACN) and mobile phase B (0.1% FA/100%
   ACN) with an effective flow rate of 0.2 µl/min. A constant flow rate of
   30 µl/min was achieved using a splitter. The mass spectrometer was set
   to operate in positive ion mode for data acquisition. The mass range
   was set from 300 to 2000 m/z, and precursor ions with charge states of
   + 2 to + 4 were selected for fragmentation. For each MS/MS spectrum,
   the three most abundant peptide ions with a count threshold exceeding 5
   were selected. The selected precursor ions were dynamically excluded
   for 30 s, with a mass tolerance of 30 mDa. Intelligent
   information-dependent acquisition was activated using automated
   collision energy and automatic MS/MS accumulation. The fragment
   intensity multiplier was set to 20, with a maximum accumulation time of
   2 s. Three LC-MS/MS injections (technical replicates = 3) were
   performed to ensure better coverage of the target proteome and to
   achieve better statistical consistency. Proteomics data validation was
   conducted on the same samples using Western Blot (WB) analysis. The MS
   parameters were: scan range (m/z) = 350–1500, resolution = 120,000, AGC
   target = 4e^5, and maximum injection time = 50 ms. The HCD-MS/MS
   parameters were: resolution = 30,000, AGC target = 1e^5, and collision
   energy = 33. For data-independent acquisition (DIA), each window
   overlapped by 1 m/z, with a total of 47 windows. The iRT kit (Ki3002,
   Biognosys AG, Switzerland) was added to calibrate the retention time of
   the extracted peptide peaks. The DIA dataset was generated using
   Spectronaut V13 (Biognosys, Switzerland), including data normalization
   and relative protein quantification. Welch’s ANOVA test was used for
   statistical analysis, and multiple comparisons were corrected using the
   Benjamini-Hochberg method to control the false discovery rate.
   Differentially expressed proteins were filtered based on an adjusted
   p-value (adjP) < 0.05 and |log2FC| >1.

Metabolomics analysis

   Brain tissues from the DNT group (n = 6) and the Angiopep-2/SIRT1 group
   (n = 6) were collected. Each sample was transferred to a 1.5 mL
   polypropylene tube containing 300 µL of tissue, mixed with 900 µL of
   80% methanol and 0.1% FA, and vortexed for 2 min. The mixture was then
   centrifuged at 12,000 g for 10 min. The supernatant was transferred to
   autosampler vials.

   Plasma metabolomics analysis was performed using an LC20
   ultra-high-performance liquid chromatography (UHPLC) system (Shimadzu,
   Japan) coupled with a Triple TOF-6600 mass spectrometer (AB Sciex).
   Chromatographic separation was achieved using a Waters ACQUITY UPLC HSS
   T3 C18 column (100 × 2.1 mm, 1.8 μm). The column temperature was
   maintained at 40 °C, and the flow rate was set at 0.4 mL/min. The
   mobile phase consisted of acetonitrile with 0.1% FA in water. The
   gradient elution program for mobile phase B was as follows: 5% from 0.0
   to 11.0 min, 90% from 11.0 to 12.0 min, and 5% from 12.1 to 14.0 min.
   The eluent was directly introduced into the mass spectrometer without
   splitting.

   The mass spectrometry conditions were as follows: ionization voltage,
   5500 V; capillary temperature, 550 °C; nebulizer gas flow rate, 50 psi;
   and auxiliary heating gas flow rate, 60 psi. Orthogonal partial least
   squares-discriminant analysis (OPLS-DA) and 100-permutation tests were
   performed on the preprocessed data to prevent overfitting. In the
   OPLS-DA model, variables with a variable importance in projection (VIP)
   value > 1 were considered significant. Subsequently, univariate
   statistical analysis was conducted using Welch’s t-test, which is
   appropriate for samples with unequal variances. Multiple comparisons
   were corrected using the Benjamini-Hochberg (BH) method to control the
   false discovery rate (FDR). Metabolites were considered differentially
   expressed (DMs) if they met the criteria of adjusted p-value
   (adjP) < 0.05 and |log₂FC| >1. Metabolites that satisfied both VIP > 1
   and the statistical significance criteria were defined as final
   differential metabolites. MetaboAnalyst (Version 5.0) was used to
   identify the associated metabolic pathways.

Cell transfection and grouping

   Human neuroblastoma SH-SY5Y cells were obtained from the American Type
   Culture Collection (SCSP-5014, Cell Bank of the Chinese Academy of
   Sciences) and routinely cultured in Dulbecco’s Modified Eagle
   Medium/Ham’s F-12 (31765035, Gibco, USA) supplemented with 10% fetal
   bovine serum (FBS) (12483020, Gibco, USA) at 37 °C in a humidified
   atmosphere with 5% CO[2]. Appropriate cell densities were used for
   various assays according to our experimental requirements.

   Silencing SIRT1: Lentivirus packaging was performed using a packaging
   kit (V48820, Invitrogen, USA). Control plasmid sh-NC, sh-SIRT1-1, and
   sh-SIRT1-2 (constructed in pLKO.1 vector) with shRNA sequences listed
   in Table S3 were transfected into HEK-293T cells (CRL-3216, ATCC).
   After 48 h of transfection, the supernatant was collected and
   concentrated using a lentivirus concentration solution (631231,
   Takara). The concentrated lentivirus was stored at −80 °C. For
   lentivirus-mediated cell transfection, 5 × 10⁵ SH-SY5Y cells were
   seeded into 6-well plates. When cell confluence reached 70–90%, the
   medium containing lentivirus (MOI = 10, working titer approximately
   5 × 10⁶ TU/mL) and 5 µg/mL polybrene (TR-1003, Sigma-Aldrich, UK) were
   added. After 4 h, the medium was diluted with an equal volume of fresh
   medium to reduce polybrene concentration. After 24 h, the medium was
   replaced with a fresh culture medium. After 48 h, transfection
   efficiency was observed using a luciferase reporter gene. Stable cell
   lines were selected using an appropriate concentration of puromycin
   (E607054, Sangon Biotech, Shanghai, China). Once the cells no longer
   died in the puromycin-containing medium, they were collected, and
   knockdown efficiency was confirmed by RT-qPCR.

   Establishment of the sevoflurane-induced DNT cell model: Cells in the
   exponential growth phase, between passages 35 and 60, were exposed to
   different concentrations of sevoflurane for 24 h. Freshly prepared
   sevoflurane stock solution (4 g/L in distilled water) was used to
   create exposure concentrations of 20, 40, and 60 mg/L (equivalent to
   0.476, 0.952, and 1.429 mM) [[75]44]. Based on the results of the CCK-8
   assay, 60 mg/L was selected for modeling. On days 1, 9, and 11, general
   cytotoxicity, neuronal precursor cell migration, and neurite outgrowth
   were measured in 96-well plates. Grouping details are provided in Table
   S4.

   General cytotoxicity: Cells were thawed and grown to 90% confluence in
   T175 flasks (Greiner, Hamburg, Germany). After trypsinization, the
   cells were resuspended in DMEM/F12 medium without RA and seeded at
   10,000 cells per well in 200 µL medium in uncoated 96-well plates
   (Corning Costar, Kaiserslautern, Germany). After 1 h of adhesion, cells
   were exposed to Angiopep-2/SIRT1. Following a 24-hour incubation, the
   cells underwent the Resazurin reduction assay (V23110, Invitrogen, USA)
   and were then fixed in a 4% paraformaldehyde phosphate-buffered
   solution (PFA) for 15 min. Cell analysis and imaging were performed
   using a Zeiss Axiovert 200 inverted microscope equipped with a Colibri
   LED light source, Zeiss Axiocam50 monochrome digital camera, and Zeiss
   ZENlite 2.6 blue edition software.

   Cell migration assay: Cells were seeded at a density of 4 × 10⁶ to
   5 × 10⁶ cells per dish in 96 mm bacteriological-grade culture dishes
   (Greiner, Hamburg, Germany). Within 24 h, cells formed free-floating
   spheroid aggregates with diameters ranging from 300 to 800 μm. On day
   1, 10 mL of medium was added to each dish. Over the following days, 10
   µM retinoic acid was added, and the medium was changed every 2–3 days.
   Cell suspensions were transferred to centrifuge tubes and centrifuged
   at 200 × g for 7 min. After 9 days of culture, aggregates were gently
   resuspended using a Pasteur pipette and seeded at 60,000 cells per well
   in black 96-well plates with flat, clear bottoms (Nunc). The plates
   were coated with poly-D-lysine (A3890401, Invitrogen, USA) and laminin
   ([76]A29249, Invitrogen, USA) (each at 10 µg/mL) and equipped with
   silicone stoppers from the Oris Cell Migration Assay (AMS
   Biotechnology, Abingdon, UK). Allow the cells to adhere overnight. On
   the next day, remove the stoppers, leaving a 2 mm diameter circular
   hole in the monolayer of cells. Incubate the cells with seven different
   concentrations of copper sulfide nanoparticles in DMEM/F12/RA medium
   for 44 h, with six technical replicates for each concentration. Perform
   the Resazurin reduction assay for 2 h. Wash the cells once with PBS,
   fix them with PFA for 15 min, wash twice with PBS-T, stain with DAPI
   (0.5 µg/mL) for 5 min, and wash twice with PBS. Attach a black plastic
   mask with 96 holes to the bottom of the plate, leaving a 2 mm diameter
   central circular area for each well. This setup allows observation of
   cells that migrated into the free area left by the silicone stopper
   during seeding (Fig. [77]1B). Quantify cell migration by taking two
   photographs per well and measuring the distance from the migration
   front to the black edge of the mask using ImageJ. To compensate for
   occasional asymmetry due to slight inaccuracies in the placement of
   silicone stoppers, average measurements from four quadrants per well.
   Use 100 nM cytochalasin D (PHZ1063, Invitrogen, USA) in six wells per
   plate as a reference for no migration, which completely inhibits cell
   movement over 44 h without affecting cell viability. Subtract the
   average zero-migration distance from the measured migration distances
   before evaluation. As an endpoint-specific control, add the Rho kinase
   inhibitor Y-27,632 (HY-10071, MedChemExpress, USA) at 50 µM to cells in
   six wells, which increases precursor cell migration rates to over 125%
   without affecting overall cell survival. Discard cell cultures that do
   not meet these standards.

Fig. 1.

   [78]Fig. 1
   [79]Open in a new tab

   Preparation and Characterization of Angiopep-2/SIRT1 Nanoparticles. A
   Schematic diagram of the preparation process of Angiopep-2/SIRT1
   nanoparticles; (B) Transmission electron microscopy (TEM) image of
   Angiopep-2/SIRT1 nanoparticles, scale bar: 50 nm; (C) Particle size
   distribution of Angiopep-2/SIRT1 nanoparticles; (D) Zeta potential of
   Angiopep-2/SIRT1 nanoparticles; (E) Scanning electron microscopy (SEM)
   image of Angiopep-2/SIRT1 nanoparticles, scale bar:200 nm; (F) Serum
   stability of Angiopep-2/SIRT1 nanoparticles; (G) Flow cytometry
   analysis of protein expression in SH-SY5Y cells.All experiments were
   performed in triplicate, and data are presented as mean ± standard
   deviation

Ca²^+ concentration detection

   To detect intracellular free Ca²⁺ levels, use Fluo-3 AM (S1056,
   Beyotime, China). Briefly incubate the cells with Fluo-3 AM at 37 °C
   for 1 h. Then, wash the cells three times with PBS. Measure the
   absorbance of all wells using a microplate reader (Tecan Infinite F50,
   Switzerland) with excitation at 488 nm and emission at 530 nm. Convert
   the fluorescence values to Ca²⁺ concentrations using a standard curve.

Propidium iodide (PI) and Acridine orange (AO) staining

   Cell experiments: Seed the cells in 6-well plates prior to the
   designated treatments. Incubate the cells with 2 µg/mL PI (#P4170,
   Sigma Aldrich) for 10 min or with 200 µM AO (#A6014, Sigma Aldrich) for
   15 min. Wash all PI-stained wells three times with PBS, and wash all
   AO-stained wells three times with PBS. After fixation with 4% PFA for
   20 min, directly capture images using an Olympus IX73 microscope. Stain
   primary neurons with anti-TUJ1 antibody (PA5-80198, 5 µg/mL,
   Invitrogen).

   Animal experiment: For each group, 5 mice were intraperitoneally
   injected with PI 3 h before euthanasia. Thirty minutes after the last
   administration, the mice were rapidly perfused through the heart with
   physiological saline followed by 4% paraformaldehyde. The brains were
   extracted, fixed, dehydrated in a gradient series, embedded in an OCT
   compound, and sectioned (protected from light). The sections were
   incubated overnight with an anti-NEUN antibody (PA5-78499, 1:100,
   Invitrogen). The next day, the sections were washed three times with
   PBS, each wash lasting 5 min. The sections were then incubated in the
   dark at room temperature for 2 h with goat anti-rabbit IgG (H + L)
   secondary antibody (1:500 dilution). After incubation with the
   secondary antibody, the sections were washed again three times with
   PBS, each wash lasting 5 min. The specimens were then stained with DAPI
   and imaged using a fluorescence microscope. The images were analyzed
   using Image LabTM software.

TUNEL staining

   The TUNEL Kit (C1086, Beyotime, China) was used to detect apoptosis in
   mouse brain tissue. Sections were treated with 3% H₂O₂ and then
   incubated with 50 µL of TUNEL reaction mixture for 60 min at 37 °C,
   avoiding light exposure. The cell nuclei were stained with DAPI for
   30 min, followed by three PBS washes. Images of the tumor tissues were
   captured at 400× magnification using a fluorescence microscope equipped
   with a camera (BX53, Olympus, Japan). The TUNEL fluorescence area and
   DAPI fluorescence area were calculated using ImageJ software. Each
   group consisted of 5 mice, and for each mouse, 5 sections were
   analyzed. Randomly selected fields of view (6–10 per section) were
   observed, and the number of positive cells was recorded. The total
   number of cells and the number of apoptotic cells were counted to
   calculate the apoptosis rate.

Enzyme-linked immunosorbent assay (ELISA)

   Collected mouse serum and used a total SOD activity assay kit (S0101S,
   Beyotime, China) to measure SOD levels, and a lipid peroxidation (MDA)
   assay kit (S0131S, Beyotime, China) to measure MDA levels. First, the
   antigen was diluted with coating buffer to the appropriate
   concentration. The wells of the microplate were blocked with 5% calf
   serum (F8318, MSK, Wuhan, China) at 37 °C for 40 min. Next, the diluted
   samples were added to the wells, followed by the addition of the
   enzyme-labeled antibody, and then the substrate solution. Finally, 50
   µL of stop solution was added to each well to terminate the reaction,
   and the results were measured within 20 min. The absorbance was read at
   450 nm using a microplate reader (Bio-Rad, USA), the standard curve was
   plotted, and the data were analyzed.

Immunofluorescence staining for protein expression detection

   SH-SY5Y cells or mouse brain tissues were fixed with 4%
   paraformaldehyde at room temperature for 15 min, followed by two washes
   with PBS and permeabilization with 0.5% Triton X-100 (P0096, Beyotime
   Biotechnology, China) for 10 min. The cells and tissues were then
   incubated overnight at 4 °C with the following primary antibodies:
   rabbit anti-β-III-tubulin (PA5-80198, 5 µg/mL), rabbit anti-PSD95
   (MA1-046, 1:500), p-RIPK1 (PA5-104645, 1:100), p-RIPK3-Human (93654,
   1:800), p-RIPK3-Mouse (91702, 1:100), p-MLKL (#AF3903, 1:100), SIRT1
   (MA5-27217, 1:100), Synaptophysin (MA5-14532, 1:250), MAP2 (PA1-10005,
   1:5000), GFP (A-11122, 1:200), and nNOS (PA1-033, 1:200). After washing
   three times with PBS, the samples were stained with FITC-labeled
   phalloidin (F432, Invitrogen) or rhodamine-labeled phalloidin (R415,
   Invitrogen) to label the cytoskeleton. Subsequently, they were
   incubated with goat anti-rabbit Alexa Fluor 647 (ab150083, 1:200) or
   goat anti-rabbit Alexa Fluor 488 (ab150077, 1:200) secondary antibodies
   for 1 h. Afterward, the samples were washed three times with PBS and
   stained with DAPI (D3571, 10 µg/mL) at room temperature for 10 min. The
   sections were stored at 4 °C and then observed under a fluorescence
   microscope (IMT-2, Olympus, Japan). The fluorescence area of the
   relevant proteins relative to the nucleus fluorescence area was
   calculated using ImageJ software. For cell experiments, five sections
   per sample were randomly observed in 6–10 fields of view, with each
   experiment repeated three times per group. Except for the p-RIPK3-Human
   and p-RIPK3-Mouse antibodies, which were purchased from Cell Signaling,
   and Alexa Fluor 647 and Alexa Fluor 488 from Abcam, the p-MLKL antibody
   from Affinity, all other antibodies were purchased from Invitrogen.

WB

   Cells and brain tissues were lysed in RIPA lysis buffer (P0013B,
   Beyotime Biotechnology, China). Protein concentrations were quantified
   using a BCA assay kit ([80]A53226, Thermo Fisher Scientific, USA).
   After separation by polyacrylamide gel electrophoresis, proteins were
   transferred onto PVDF membranes (PVH85R, Millipore, Germany) using the
   wet transfer method. The membranes were blocked at room temperature for
   1 h with 10% BSA (37520, Thermo, USA) and then incubated overnight at
   4 °C with the following primary antibodies: p-RIPK1 (PA5-104645,
   1:1000), RIPK1 (PA5-20811, 1 µg/mL), p-RIPK3-Human (93654, 1:8000),
   p-RIPK3-Mouse (91702, 1:1000), RIPK3 (PA5-141114, 1:1000), p-MLKL
   (PA5-105678, 1:1000), MLKL (PA5-115578, 1:1000), SIRT1 (MA5-27217,
   1:1000), and GAPDH (MA5-15738, 1:1000). After washing, the membranes
   were incubated with HRP-conjugated secondary IgG antibody
   (AN-170-250UG, 1:5000) for 2 h. The membranes were then washed three
   times with TBST for 5 min each. Visualization, imaging, and analysis of
   all protein bands were performed using the Syngene G: BOX F3 imaging
   system (Antpedia, China) and ImageJ software. The relative protein
   levels were expressed as the grayscale value of the target protein
   bands relative to the grayscale value of the GAPDH protein bands. Each
   cell experiment was repeated three times, and the mouse experiments
   included six mice per group. All antibodies mentioned were purchased
   from Invitrogen.

RT-qPCR

   Total RNA was extracted from cells and tumor tissues using TRIzol
   (15596026, ThermoFisher, USA). The concentration and purity of the
   extracted RNA were measured using a Nanodrop 2000 spectrophotometer
   (ThermoFisher, USA). RNA was reverse-transcribed into cDNA according to
   the PrimeScript RT reagent Kit instructions (RR047A, Takara, Japan).
   The synthesized cDNA was subjected to RT-PCR using the Fast SYBR Green
   PCR kit (11736059, ThermoFisher, USA), with each reaction set up in
   triplicate. GAPDH was used as the internal control, and relative
   expression levels were calculated using the 2^−ΔΔCt method. The
   experiment was repeated three times. The primer sequences used for
   RT-qPCR are listed in Table S5.

Statistical analysis

   Statistical analyses in this study were performed using SPSS software
   (version 21.0, IBM, USA). Data are expressed as mean ± standard
   deviation. Normality and homogeneity of variances were first tested.
   For data that were normally distributed and had homogeneous variances,
   an unpaired t-test was used for comparisons between two groups, while
   one-way ANOVA or repeated measures ANOVA was used for comparisons among
   multiple groups. A p-value of less than 0.05 was considered
   statistically significant.

Results

Preparation and characterization of Angiopep-2/SIRT1 nanoparticles

   Sevoflurane is a widely used inhalation anesthetic in pediatric
   anesthesia due to its rapid onset and favorable tolerance. It is often
   employed in short-duration surgeries for children [[81]5]. However, an
   increasing number of studies indicate that sevoflurane may have
   long-term adverse effects on brain development in infants, leading to
   DNT [[82]45]. This neurotoxicity is primarily manifested as cognitive
   dysfunction and behavioral abnormalities. Epidemiological data show
   that infants frequently or prolongedly exposed to sevoflurane have a
   significantly increased risk of these issues [[83]46].

   SIRT1, through its antioxidant and anti-apoptotic functions, can
   significantly mitigate neurocyte damage induced by sevoflurane,
   offering a novel therapeutic strategy for DNT [[84]47]. However,
   effectively delivering SIRT1 to specific neural cells remains a
   challenge. Here, the self-assembling peptide nanoparticle technology
   provides a promising solution. By encapsulating SIRT1 activators or
   SIRT1 itself, these nanoparticles can more effectively penetrate the
   BBB and directly target brain cells. Angiopep-2, a peptide that can
   specifically recognize and penetrate the BBB, is used to modify SIRT1
   nanoparticles, further enhancing the neuro-specificity and therapeutic
   efficiency of the treatment [[85]48]. In this context, self-assembling
   peptide nanoparticles serve as a unique drug delivery system platform,
   enhancing the stability and targeting of therapeutics [[86]49].
   Specifically, Angiopep-2/SIRT1 nanoparticles combine the
   brain-targeting capability of Angiopep-2 with the antioxidant and
   anti-apoptotic properties of SIRT1. This study demonstrates that these
   nanoparticles can effectively cross the BBB and directly target the
   damaged nervous system. By activating SIRT1, they effectively inhibit
   sevoflurane-induced necroptosis and other cellular damage pathways,
   significantly improving the symptoms of DNT.

   The Angiopep-2/SIRT1 nanoparticles were prepared according to the
   protocol outlined in Fig. [87]1A. TEM results show that the
   Angiopep-2/SIRT1 nanoparticles have a diameter of 144.86 ± 2.17 nm and
   a zeta potential of −19.06 ± 9.08 mV (Fig. [88]1B-D). SEM analysis
   reveals that the Angiopep-2/SIRT1 nanoparticles are spherical with a
   uniform size distribution (Fig. [89]1E). In the serum stability test,
   both SIRT1 and Angiopep-2/SIRT1 nanoparticles remained stable for 7
   days in PBS containing 10% FBS (Fig. [90]1F). Flow cytometry analysis
   indicated that SH-SY5Y cells had internalized the Angiopep-2/SIRT1
   nanoparticles, as evidenced by the presence of SIRT1 protein within the
   cells (Fig. [91]1G).

   Following the successful preparation of Angiopep-2/SIRT1 nanoparticles,
   we further assessed their stability under various physiological
   conditions (water, PBS, saline, and RPMI 1640) using DLS. The results
   indicated slight variations in the size of the Angiopep-2/SIRT1
   nanoparticles, but overall, they exhibited good stability in all tested
   environments (Fig. S1A). Additionally, in different pH buffer
   solutions, the Angiopep-2/SIRT1 nanoparticles demonstrated a decrease
   in particle size with increasing pH, particularly showing brain pH
   responsiveness at pH 7.4 (Fig. S1B). We evaluated the impact of
   Angiopep-2/SIRT1 nanoparticles on SH-SY5Y cell viability and
   proliferation using LIVE/DEAD staining and the CCK-8 assay. The CCK-8
   assay results showed an increase in cell number over time with no
   significant differences between groups (Fig. S1C). LIVE/DEAD staining
   indicated that nearly all cells remained viable over a 3-day culture
   period, with no differences observed between groups (Fig. S1D).
   Furthermore, the hemolysis assay results revealed negligible hemolysis
   (< 5%) at all concentrations of Angiopep-2/SIRT1 nanoparticles (Fig.
   S1E). These findings demonstrate that Angiopep-2/SIRT1 nanoparticles
   possess good biocompatibility and biosafety.

   SIRT1 and Angiopep-2/SIRT1 were respectively added to the upper chamber
   of a Transwell system, and their uptake by cells in both the upper and
   lower chambers was observed using CLSM. After a 4-hour incubation, the
   fluorescence intensity of FITC was measured using an ELISA reader. In
   SH-SY5Y cells treated with Angiopep-2/SIRT1, the average FITC
   fluorescence intensity was significantly higher compared to cells
   treated with SIRT1 alone (Fig. S2A). To evaluate the targeting
   efficacy, Angiopep-2/SIRT1 nanoparticles conjugated with Cyanine 5.5
   (Cy5.5) were intravenously injected into mice. The Angiopep-2/SIRT1
   nanoparticles group exhibited stronger Cy5.5 fluorescence in the brain
   at 6, 12, 24, and 48 h post-injection, indicating enhanced targeting,
   permeability, and retention effects of Angiopep-2/SIRT1 (Fig. S2B).

   This study demonstrates that Angiopep-2/SIRT1 nanoparticles, due to
   their excellent stability and biocompatibility, successfully achieved
   efficient delivery of SIRT1 protein in both in vitro and in vivo
   models. Notably, in the brain environment, these nanoparticles
   exhibited significant targeting and enhanced cellular uptake, providing
   a potential strategy for the future treatment of related
   neuropathological conditions.

Multi-omics analysis reveals potential pathways through which
Angiopep-2/SIRT1 nanoparticles alleviate sevoflurane-induced DNT

   In this study, we explored the molecular mechanisms underlying
   sevoflurane-induced DNT, focusing on the regulatory role of
   Angiopep-2/SIRT1 nanoparticles. Through a systematic evaluation of
   cellular morphology and functional damage in the DNT model, we analyzed
   how these nanoparticles affect necroptosis-related pathways, providing
   potential therapeutic targets [[92]13, [93]50]. First, we established a
   DNT model by inducing sevoflurane exposure in neonatal mice. H&E
   staining and TEM results showed that, compared to the control group,
   the DNT group exhibited nuclear condensation, vacuolization, and an
   increase in striatal neurons (Fig. S3A). Nissl staining results
   indicated a significant reduction in Nissl-stained cells in the DNT
   group, suggesting neuronal loss in the striatum compared to the control
   group (Fig. S3B). Golgi staining was used to examine dendritic spine
   density, revealing a significant decrease in the number of spines on
   pyramidal neurons in the DNT group compared to the control group (Fig.
   S3C). Immunofluorescence results showed that β-III-tubulin and PSD95
   were significantly reduced in the DNT group compared to the control
   group (Fig. S3D-E).

   After establishing the DNT model, we conducted negative geotaxis and
   front-limb suspension tests on days 31, 35, and 37. The results of the
   negative geotaxis test (Fig. A) showed that the time required for the
   DNT group mice to turn their heads was significantly longer compared to
   the control group (Fig. S4B). In the front-limb suspension test (Fig.
   S4C), the DNT group mice spent significantly less time hanging on the
   wire compared to the control group (Fig. S4D). To further assess the
   long-term neurofunctional damage induced by sevoflurane in DNT mice, we
   performed NORT and MWMT. The NORT results (Fig. S4E) indicated a
   significantly lower discrimination index in the DNT group compared to
   the control group, suggesting a marked decline in their ability to
   differentiate between new and familiar objects, which implies
   impairment in cognitive functions, particularly memory and learning
   abilities (Fig. S4F). The MWMT results showed that the DNT group mice
   spent significantly more time finding the platform and less time in the
   target quadrant compared to the control group, indicating more severe
   cognitive impairment in the DNT group (Fig. S4G). However, there was no
   difference in swimming speed between the two groups (Fig. S4H),
   suggesting that the longer escape latency was primarily due to impaired
   spatial memory rather than slower swimming speed (Fig. S4I-J). After 5
   days of training, the platform was removed from the water maze, and the
   mice’s movement trajectories were recorded for 1 min. The target
   quadrant, where the platform was previously located, was considered the
   escape zone. The control group spent approximately 50% of their time in
   the target quadrant, while the DNT group spent similar amounts of time
   in all four quadrants, indicating significant declines in cognitive
   function and memory in the DNT group (Fig. S4K-L). These results
   confirm the successful establishment of the sevoflurane-induced DNT
   model in neonatal mice, demonstrating both short-term and long-term
   neurofunctional deficits.

   The processing and differential analysis of high-throughput sequencing
   data revealed that, using a threshold of |log2FC|>2 and adjP < 0.05,
   compared to the DNT group, the Angiopep-2/SIRT1 group had 441
   significantly upregulated and 593 significantly downregulated genes in
   the brain tissue (Fig. [94]2A-B). Subsequently, two machine learning
   algorithms were used to screen the differential genes. A Lasso
   regression model and the SVM-RFE algorithm were employed to identify
   feature genes. The Lasso regression algorithm identified nine key
   genes: Ptx3, Gadd45a, Ripk1, Lims2, Txnip, Polr3gl, Nedd9, Rpa2, and
   Cers1 (Fig. [95]2C). The SVM-RFE algorithm identified ten key genes:
   Igf1, Aven, Ift70a1, Dtwd2, Ripk1, Psmd6, Tg, Gm32856, Nfyc, and Mfsd12
   (Fig. [96]2D). By intersecting the genes identified by both algorithms,
   the common feature gene Ripk1 (RIPK1) was obtained, as illustrated in
   the Venn diagram (Fig. [97]2E). Following the identification of
   differential genes, enrichment analysis was conducted. GO analysis
   results indicated significant enrichment of differential genes in
   various categories. In the biological process (BP) category, the genes
   were primarily enriched in programmed necrotic cell death, necroptotic
   process, regulation of programmed necrotic cell death, and necrotic
   cell death. In the CC category, enrichment was observed in receptor
   complex, membrane microdomain, membrane raft, and collagen-containing
   extracellular matrix. In the molecular function (MF) category, the
   genes were mainly enriched in tumor necrosis factor receptor
   superfamily binding, ATP hydrolysis activity, endodeoxyribonuclease
   activity producing 5’-phosphomonoesters, and lyase activity (Fig.
   [98]2F). KEGG pathway analysis showed that the differential genes were
   predominantly enriched in the TNF signaling pathway, NF-kappa B
   signaling pathway, apoptosis, and IL-17 signaling pathway (Fig.
   [99]2G). Using machine learning algorithms, RIPK1 was identified as a
   key gene involved in necroptosis. The results of GO and KEGG, pathway
   enrichment analyses, indicate that the differential genes are closely
   related to necroptosis.

Fig. 2.

   [100]Fig. 2
   [101]Open in a new tab

   Identification of Potential Pathways for Angiopep-2/SIRT1 Nanoparticles
   in Alleviating Sevoflurane-Induced DNT Using Multi-Omics Data. A
   Volcano plot of differentially expressed genes between DNT group
   (n = 3) and Angiopep-2/SIRT1 group (n = 3) based on high-throughput
   sequencing data. Green dots indicate significantly upregulated genes,
   blue dots indicate significantly downregulated genes. Screening
   criteria: |log₂FC| >2 and adjP < 0.05; (B) Heatmap of differentially
   expressed genes between DNT and Angiopep-2/SIRT1 groups; green
   indicates high expression and blue indicates low expression; (C) LASSO
   algorithm identified 9 feature genes; (D) SVM-RFE algorithm identified
   10 feature genes; (E) Venn diagram showing the overlap of the two
   machine learning results, yielding 1 shared gene; (F) Bubble plot of GO
   enrichment analysis of differentially expressed genes; (G) Bubble plot
   of KEGG pathway enrichment analysis of differentially expressed genes;
   (H) Venn diagram of proteomics data showing 1 overlapping protein
   identified by both machine learning algorithms; (I) KEGG pathway
   analysis of 21 differential metabolites (VIP > 1, adjP < 0.05, |log₂FC|
   >1)

   To further investigate the pathways through which Angiopep-2/SIRT1
   nanoparticles mitigate sevoflurane-induced DNT, we collected brain
   tissues from three mice in both the DNT and Angiopep-2/SIRT1 groups.
   Principal component analysis (PCA) results and the loading plot showed
   a clear separation between the DNT group and the Angiopep-2/SIRT1 group
   (Fig. S5A-C). The OPLS-DA results also demonstrated a clear separation
   between the DNT and Angiopep-2/SIRT1 groups in the OPLS-DA score plot
   (Fig. S5D-F). Differential analysis identified, using a threshold of
   |log2FC|>1 and adjP < 0.05, a total of 278 differential proteins, with
   24 upregulated and 20 downregulated proteins (Fig. S6A-B).
   Subsequently, two machine learning algorithms were used to screen the
   differential proteins. The Lasso regression model identified two key
   proteins: Ripk1 and Aarsd1 (Fig. S6C). The SVM-RFE algorithm also
   identified two key proteins: Mospd1 and Ripk1 (Fig. S6D). The
   intersection of the proteins identified by the two algorithms resulted
   in the key protein, RIPK1, as illustrated in the Venn diagram (Fig.
   [102]2H). Next, we performed an enrichment analysis on the differential
   proteins. According to GO analysis, the differential proteins in the BP
   category were mainly enriched in the regulation of necrotic cell death,
   regulation of the necroptotic process, regulation of programmed
   necrotic cell death, and necrotic cell death. In the CC category, they
   were primarily enriched in NuRD complex, CHD-type complex, histone
   deacetylase complex, SWI/SNF superfamily-type complex, and U5 snRNP. In
   the MF category, they were mainly enriched in aminoacyl-tRNA editing
   activity, death receptor binding, transcription corepressor activity,
   and aminoacyl-tRNA ligase activity (Fig. S6E). KEGG pathway analysis
   showed that the differential proteins were predominantly enriched in
   necroptosis, NOD-like receptor signaling pathway, TNF signaling
   pathway, and cytosolic DNA-sensing pathway (Fig. S6F). These
   experimental results indicate that the differential proteins are mainly
   enriched in the necroptosis pathway, consistent with the
   high-throughput sequencing results.

   We collected brain tissues from six mice in both the DNT group and the
   Angiopep-2/SIRT1 group. PCA results and the loading plot showed a clear
   separation between the DNT and Angiopep-2/SIRT1 groups (Fig. S7A-C).
   The OPLS-DA results also demonstrated a clear separation between the
   DNT and Angiopep-2/SIRT1 groups, with a permutation analysis showing
   R²Y = 0.944 > 0.8, indicating that the model is stable (Fig. S7D-F).
   Using thresholds of VIP > 2, |log₂FC| >1, and adjusted p-value
   (adjP) < 0.05, a total of 21 DMs were identified. Subsequent KEGG
   pathway analysis revealed that these metabolites were primarily
   enriched in the glyoxylate and dicarboxylate metabolism, alanine,
   aspartate and glutamate metabolism, nitrogen metabolism, and pyruvate
   metabolism pathways (Fig. [103]2I, Table S6). Glyoxylate and
   dicarboxylate metabolism involve key energy production pathways, such
   as the tricarboxylic acid cycle, and their dysfunction may lead to
   mitochondrial damage and necroptosis [[104]51]. Alanine, aspartate, and
   glutamate metabolism regulate intracellular amino acid and nitrogen
   balance, affecting cell survival signals. Dysregulation in these
   pathways may trigger cell death mechanisms [[105]52]. Additionally,
   pyruvate, as a key junction between glycolysis and the tricarboxylic
   acid cycle, is directly linked to necroptosis caused by energy
   depletion when its metabolism is disrupted [[106]53]. The disruption of
   these metabolic pathways is closely related to necroptosis in various
   pathological conditions, suggesting that metabolic regulation could
   provide new strategies for preventing and treating diseases associated
   with necroptosis.

   The results indicate that Angiopep-2/SIRT1 nanoparticles significantly
   ameliorate sevoflurane-induced DNT. The underlying mechanisms primarily
   involve the regulation of multiple key metabolic pathways related to
   necroptosis. These findings not only deepen our understanding of the
   pathological mechanisms of DNT but also provide a crucial molecular
   basis for developing new therapeutic strategies. With further research,
   we hope to more comprehensively uncover the potential of these
   nanoparticles in regulating cell death, offering effective treatment
   options for DNT.

Angiopep-2/SIRT1 nanoparticles activate SIRT1 to inhibit necroptosis and
alleviate sevoflurane-induced DNT in SH-SY5Y cells

   In this study, we utilized an SH-SY5Y cell model to explore the
   neurodevelopmental toxicity effects of sevoflurane and the regulatory
   role of Angiopep-2/SIRT1 nanoparticles. Neuronal migration is a
   critical process in brain development, and its disruption is often
   associated with DNT. Such toxicity can alter migration pathways,
   affecting the formation and function of neural networks and potentially
   leading to cognitive and behavioral abnormalities [[107]54, [108]55].
   Therefore, proper neuronal migration is essential for maintaining the
   health and function of the nervous system. Through detailed cellular
   and molecular biological analyses, we observed the effects of
   sevoflurane on apoptosis and migration, as well as its specific impact
   on neuronal morphology and function. These analyses helped to elucidate
   the potential mechanisms by which Angiopep-2/SIRT1 nanoparticles
   regulate intracellular signaling pathways. Our findings demonstrate
   that Angiopep-2/SIRT1 nanoparticles can activate SIRT1, inhibit
   necroptosis, and significantly reduce sevoflurane-induced DNT in
   SH-SY5Y cells. The technical route for the sevoflurane-induced DNT cell
   model is depicted in Fig. S8A. SH-SY5Y cells were seeded at a density
   of 10,000 cells per well in a 96-well plate using a normal culture
   medium. Within 24 h, the cells proliferated to reach 90% confluency.
   Treatment with 60 mg/L sevoflurane caused the culture medium to turn
   green, and this concentration was subsequently used for model induction
   (Fig. S8B). The cell migration assay results showed that cell migration
   significantly increased in the DNT group compared to the PBS group
   (Fig. S8C). Immunofluorescence results indicated a significant
   downregulation of β-III-tubulin expression in the DNT group compared to
   the PBS group (Fig. S8D). These findings confirm the successful
   establishment of the sevoflurane-induced DNT cell model.

   Initially, two shRNA constructs targeting SIRT1 (sh-SIRT1) were
   transfected into SH-SY5Y cells. RT-qPCR results indicated that
   sh-SIRT1-1 exhibited the highest silencing efficiency, and thus,
   sh-SIRT1-1 (sh-SIRT1) was used for subsequent experiments (Fig. S8E).
   WB results showed that, compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group had significantly lower relative
   expression levels of p-RIPK1/RIPK1, p-RIPK3/RIPK3, and p-MLKL/MLKL, and
   significantly higher expression of SIRT1 in SH-SY5Y cells. In contrast,
   the sh-SIRT1 group exhibited significantly higher relative expression
   levels of p-RIPK1/RIPK1, p-RIPK3/RIPK3, and p-MLKL/MLKL, and
   significantly lower expression of SIRT1. Moreover, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   showed significantly higher relative expression levels of
   p-RIPK1/RIPK1, p-RIPK3/RIPK3, and p-MLKL/MLKL, and significantly lower
   expression of SIRT1 in SH-SY5Y cells (Fig. [109]3A). Immunofluorescence
   results demonstrated that, compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group showed significantly lower fluorescence
   intensities of p-RIPK1, p-RIPK3, and p-MLKL and a significantly higher
   fluorescence intensity of SIRT1 in SH-SY5Y cells. Conversely, the
   sh-SIRT1 group exhibited significantly higher fluorescence intensities
   of p-RIPK1, p-RIPK3, and p-MLKL and significantly lower fluorescence
   intensity of SIRT1. Furthermore, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   displayed significantly higher fluorescence intensities of p-RIPK1,
   p-RIPK3, and p-MLKL, and significantly lower fluorescence intensity of
   SIRT1 in SH-SY5Y cells (Fig. [110]3B-E).

Fig. 3.

   [111]Fig. 3
   [112]Open in a new tab

   Effects of Angiopep-2/SIRT1 Nanoparticles on Necroptosis-Related
   Proteins. A WB analysis of the expression levels of p-RIPK1/RIPK1,
   p-RIPK3/RIPK3, p-MLKL/MLKL, and SIRT1 in SH-SY5Y cells; (B)
   Immunofluorescence staining showing the expression of SIRT1 in SH-SY5Y
   cells, scale bar: 25 μm; (C) Immunofluorescence staining showing the
   expression of p-RIPK1 in SH-SY5Y cells, with GFP-labeled neurons used
   for protein quantification, scale bar: 25 μm; (D) Immunofluorescence
   staining showing the expression of p-RIPK3 in SH-SY5Y cells, with
   GFP-labeled neurons used for protein quantification, scale bar: 25 μm;
   (E) Immunofluorescence staining showing the expression of p-MLKL in
   SH-SY5Y cells, with GFP-labeled neurons used for protein
   quantification, scale bar: 25 μm. Each experiment was repeated three
   times, and values are presented as mean ± standard deviation. ***
   indicates p < 0.001

   CCK-8 assays were used to assess cell viability. The results showed
   that, compared to the sh-NC group, the sh-NC + Angiopep-2/SIRT1 group
   had increased cell viability in SH-SY5Y cells, while the sh-SIRT1 group
   had decreased cell viability. Additionally, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   exhibited reduced cell viability in SH-SY5Y cells (Fig. [113]4A). PI
   staining results indicated that, compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group showed significantly reduced PI
   fluorescence in SH-SY5Y cells, whereas the sh-SIRT1 group exhibited
   significantly increased PI fluorescence. Furthermore, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   displayed significantly increased PI fluorescence in SH-SY5Y cells
   (Fig. [114]4B). AO staining results showed that, compared to the sh-NC
   group, the sh-NC + Angiopep-2/SIRT1 group exhibited increased
   cytoplasmic (red fluorescence) and nuclear (green fluorescence)
   integrity in SH-SY5Y cells, whereas the sh-SIRT1 group showed decreased
   cytoplasmic and nuclear integrity. Furthermore, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   exhibited decreased cytoplasmic and nuclear integrity in SH-SY5Y cells
   (Fig. [115]4C). RT-qPCR results indicated that, compared to the sh-NC
   group, the sh-NC + Angiopep-2/SIRT1 group had significantly
   downregulated mRNA levels of Interleukin-6 (IL-6), Tumor Necrosis
   Factor-alpha (TNF-α), and Transforming Growth Factor-beta (TGF-β) in
   SH-SY5Y cells, while the sh-SIRT1 group showed significantly
   upregulated levels of these cytokines. Additionally, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   had significantly upregulated mRNA levels of IL-6, TNF-α, and TGF-β in
   SH-SY5Y cells (Fig. [116]4D).

Fig. 4.

   [117]Fig. 4
   [118]Open in a new tab

   Effects of Angiopep-2/SIRT1 Nanoparticles on Necroptosis. A CCK-8 assay
   to detect the cell viability of SH-SY5Y cells; (B) PI staining to
   detect apoptosis in SH-SY5Y cells, scale bar: 25 μm; (C) AO staining to
   examine DNA, RNA, and neurons in SH-SY5Y cells, scale bar: 25 μm; (D)
   RT-qPCR analysis of IL-6, TNF-α, and TGF-β mRNA levels in SH-SY5Y
   cells. Each experiment was repeated three times, and values are
   presented as mean ± standard deviation. *** indicates p < 0.001

   Fluo-3 AM was used to measure intracellular Ca²⁺ levels. The results
   showed that, compared to the sh-NC group, the sh-NC + Angiopep-2/SIRT1
   group had significantly increased Ca²⁺ levels in SH-SY5Y cells, whereas
   the sh-SIRT1 group exhibited significantly reduced Ca²⁺ levels.
   Additionally, compared to the sh-NC + Angiopep-2/SIRT1 group, the
   Angiopep-2/SIRT1 + sh-SIRT1 group showed significantly reduced Ca²⁺
   levels in SH-SY5Y cells (Fig. [119]5A). Immunofluorescence analysis was
   used to detect the expression of PSD95 and β-III-tubulin in SH-SY5Y
   cells. The results indicated that, compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group had significantly upregulated expression
   of PSD95 and β-III-tubulin, while the sh-SIRT1 group showed
   significantly downregulated expression of these proteins. Furthermore,
   compared to the sh-NC + Angiopep-2/SIRT1 group, the
   Angiopep-2/SIRT1 + sh-SIRT1 group exhibited significantly downregulated
   expression of PSD95 and β-III-tubulin in SH-SY5Y cells (Fig.
   [120]5B-C). The cell migration assay results showed that compared to
   the sh-NC group, the sh-NC + Angiopep-2/SIRT1 group exhibited a
   significant reduction in cell migration, while the sh-SIRT1 group
   demonstrated a significant increase in cell migration. Additionally,
   compared to the sh-NC + Angiopep-2/SIRT1 group, the
   Angiopep-2/SIRT1 + sh-SIRT1 group showed a significant increase in cell
   migration (Fig. [121]5D). Immunofluorescence staining revealed that
   compared to the sh-NC group, the sh-NC + Angiopep-2/SIRT1 group had a
   significant increase in the total length of neurites and the average
   number of dendritic spines in SH-SY5Y cells. In contrast, the sh-SIRT1
   group exhibited a significant decrease in both neurite length and
   dendritic spine density. Moreover, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   also showed a significant reduction in neurite length and dendritic
   spine number (Fig. [122]5E).

Fig. 5.

   [123]Fig. 5
   [124]Open in a new tab

   Angiopep-2/SIRT1 Nanoparticles Inhibit Necroptosis to Alleviate
   Sevoflurane-Induced DNT in SH-SY5Y Cells. A Fluo-3 AM assay to detect
   Ca^2+ levels in SH-SY5Y cells; (B) Immunofluorescence staining to
   detect the expression of PSD95 in SH-SY5Y cells, scale bar: 25 μm; (C)
   Immunofluorescence staining to detect the expression of β-III-tubulin
   in SH-SY5Y cells, scale bar: 25 μm; (D) Cell migration assay to assess
   migration in SH-SY5Y cells, scale bar: 200 μm; (E) Immunofluorescence
   staining of green (Synaptophysin), red (MAP-2), and blue (DAPI) in
   SH-SY5Y cells, scale bar: 25 μm. Each experiment was repeated three
   times, and values are presented as mean ± standard deviation. ***
   indicates p < 0.001

   The above results indicate that Angiopep-2/SIRT1 nanoparticles
   effectively inhibit the necroptosis signaling pathway induced by
   sevoflurane, including the reduction of p-RIPK1, p-RIPK3, and p-MLKL
   expression, thereby promoting neuronal cell survival and structural
   integrity. Furthermore, these nanoparticles significantly improve the
   physiological state of cells by enhancing cell migration ability and
   reducing changes in intracellular calcium ion concentration. This
   provides strong experimental evidence for the development of new
   therapeutic strategies. These findings offer important molecular
   targets and potential drug delivery systems for future DNT treatment
   research.

Angiopep-2/SIRT1 nanoparticles activate SIRT1 to inhibit necroptosis and
alleviate sevoflurane-induced DNT in mice

   In this study, we rigorously evaluated the effects of Angiopep-2/SIRT1
   nanoparticles on sevoflurane-induced DNT through comprehensive cellular
   and animal experiments. Utilizing a mouse model, we conducted a series
   of biochemical and histological analyses to explore how these
   nanoparticles modulate the SIRT1-related necroptosis pathway and to
   elucidate their potential neuroprotective mechanisms. We paid
   particular attention to the nanoparticles’ effects on nuclear
   condensation, vacuolization, dendritic spine density, and neurite
   outgrowth, as well as their impact on performance in behavioral tests.
   Using the model of sevoflurane-induced DNT in mice established
   according to Fig. [125]6A results from H&E staining and TEM revealed
   the following: Compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group exhibited significantly reduced nuclear
   condensation and vacuolization, as well as an increased number of
   striatal neurons. Conversely, the sh-SIRT1 group showed significantly
   increased nuclear condensation and vacuolization, along with a
   decreased number of striatal neurons. Additionally, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   showed significantly increased nuclear condensation and vacuolization,
   and a reduced number of striatal neurons (Fig. [126]6B). Nissl staining
   results indicated that, compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group had a significant increase in
   Nissl-stained cells, whereas the sh-SIRT1 group had a significant
   decrease. Furthermore, compared to the sh-NC + Angiopep-2/SIRT1 group,
   the Angiopep-2/SIRT1 + sh-SIRT1 group exhibited a significant decrease
   in Nissl-stained cells (Fig. [127]6C). Immunofluorescence results
   showed that β-III-tubulin and PSD95 were significantly increased in the
   brain tissues of the sh-NC + Angiopep-2/SIRT1 group compared to the
   sh-NC group, whereas they were significantly decreased in the sh-SIRT1
   group. Additionally, β-III-tubulin and PSD95 levels were significantly
   reduced in the Angiopep-2/SIRT1 + sh-SIRT1 group compared to the
   sh-NC + Angiopep-2/SIRT1 group (Fig. [128]6D-E). Golgi staining
   examined the density of dendritic spines, and the results indicated
   that, compared to the sh-NC group, the number of pyramidal neuron
   spines was significantly increased in the sh-NC + Angiopep-2/SIRT1
   group, while significantly decreased in the sh-SIRT1 group.
   Furthermore, the number of pyramidal neuron spines was significantly
   reduced in the Angiopep-2/SIRT1 + sh-SIRT1 group compared to the
   sh-NC + Angiopep-2/SIRT1 group (Fig. [129]6F).

Fig. 6.

   [130]Fig. 6
   [131]Open in a new tab

   Effects of Angiopep-2/SIRT1 Nanoparticles on Sevoflurane-Induced DNT in
   Mice. A Technical roadmap for establishing a sevoflurane-induced DNT
   model in mice; (B) H&E staining and TEM analysis of cellular morphology
   and structure in mouse brains, scale bars: 100 μm (top row), 2 μm
   (second row), 500 nm (bottom row), with blue arrows indicating
   mitochondria and green arrows indicating autophagic vacuoles; (C) Nissl
   staining to detect striatal neurons in mouse brains, black arrows
   indicating neurons, scale bar: 100 μm; (D) Immunofluorescence staining
   to detect the expression of β-III-tubulin in mouse brains, scale bar:
   50 μm; (E) Immunofluorescence staining to detect the expression of
   PSD95 in mouse brains, scale bar: 50 μm; (F) Golgi staining to detect
   dendritic spines in the hippocampus of mice, scale bar: 10 μm. Each
   group consisted of 6 mice, and values are presented as mean ± standard
   deviation. *** indicates p < 0.001

   On days 31, 35, and 37, after establishing the DNT model, we conducted
   negative geotaxis and front-limb suspension tests. The results of the
   negative geotaxis test showed that, compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group required significantly less time to turn
   their heads, while the sh-SIRT1 group required significantly more time.
   Furthermore, compared to the sh-NC + Angiopep-2/SIRT1 group, the
   Angiopep-2/SIRT1 + sh-SIRT1 group took significantly more time to turn
   their heads (Fig. [132]7A). In the front-limb suspension test, the
   sh-NC + Angiopep-2/SIRT1 group spent significantly more time hanging on
   the wire compared to the sh-NC group, whereas the sh-SIRT1 group spent
   significantly less time. Additionally, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   showed a significant decrease in the hanging time (Fig. [133]7B). To
   further assess the long-term neurological damage caused by sevoflurane
   in DNT mice, we conducted NORT and MWMT. The NORT results indicated
   that the discrimination index was significantly higher in the
   sh-NC + Angiopep-2/SIRT1 group compared to the sh-NC group, while it
   was significantly lower in the sh-SIRT1 group. Additionally, compared
   to the sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1
   group showed a significantly lower discrimination index (Fig. [134]7C).
   MWMT results showed that the sh-NC + Angiopep-2/SIRT1 group spent
   significantly less time finding the platform compared to the sh-NC
   group, whereas the sh-SIRT1 group spent significantly more time.
   Furthermore, compared to the sh-NC + Angiopep-2/SIRT1 group, the
   Angiopep-2/SIRT1 + sh-SIRT1 group took significantly more time to find
   the platform. However, there was no significant difference in swimming
   speed between the two groups (Fig. [135]7D-G). In the MWMT, the
   sh-NC + Angiopep-2/SIRT1 group spent significantly more time in the
   escape zone compared to the sh-NC group, while the sh-SIRT1 group spent
   significantly less time. Compared to the sh-NC + Angiopep-2/SIRT1
   group, the Angiopep-2/SIRT1 + sh-SIRT1 group also showed a
   significantly reduced duration in the escape zone (Fig. [136]7H).

Fig. 7.

   [137]Fig. 7
   [138]Open in a new tab

   Effects of Angiopep-2/SIRT1 Nanoparticles on Neurological Reflexes in
   Sevoflurane-Induced DNT in Mice. A Negative geotaxis test results on
   days 31, 35, and 37 post-construction of the sevoflurane-induced DNT
   model in different groups of mice; (B) Front-limb suspension test
   results on days 31, 35, and 37 post-construction of the
   sevoflurane-induced DNT model in different groups of mice; (C) NORT
   test results post-construction of the sevoflurane-induced DNT model in
   different groups of mice, n = 6; (D-G) MWMT results post-construction
   of the sevoflurane-induced DNT model in different groups of mice,
   including swimming speed, escape latency, swimming distance, and time
   spent in the target quadrant; (H) Representative Morris water maze
   swimming trajectory images post-construction of the sevoflurane-induced
   DNT model in different groups of mice. Each group consisted of 6 mice,
   and values are presented as mean ± standard deviation. *** indicates
   p < 0.001

   First, the WB results showed that, compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group exhibited significantly downregulated
   relative expression levels of p-RIPK1/RIPK1, p-RIPK3/RIPK3, and
   p-MLKL/MLKL in the brain tissue, while the expression of SIRT1 was
   significantly upregulated. Conversely, in the sh-SIRT1 group, the
   relative expression levels of p-RIPK1/RIPK1, p-RIPK3/RIPK3, and
   p-MLKL/MLKL were significantly upregulated, and SIRT1 expression was
   significantly downregulated. Additionally, compared to the
   sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1 group
   showed significantly upregulated relative expression levels of
   p-RIPK1/RIPK1, p-RIPK3/RIPK3, and p-MLKL/MLKL, with a significant
   downregulation of SIRT1 expression (Fig. S9A). Immunofluorescence
   results showed that, compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group exhibited significantly reduced
   fluorescence intensity of p-RIPK1, p-RIPK3, and p-MLKL in the brain
   tissue, while the fluorescence intensity of SIRT1 was significantly
   increased. In contrast, the sh-SIRT1 group showed significantly
   increased fluorescence intensity of p-RIPK1, p-RIPK3, and p-MLKL, with
   a significant decrease in the fluorescence intensity of SIRT1.
   Additionally, compared to the sh-NC + Angiopep-2/SIRT1 group, the
   Angiopep-2/SIRT1 + sh-SIRT1 group exhibited significantly increased
   fluorescence intensity of p-RIPK1, p-RIPK3, and p-MLKL, and a
   significantly decreased fluorescence intensity of SIRT1 (Fig. S9B-E).

   TUNEL results showed that, compared to the sh-NC group, neuronal
   apoptosis in the brain tissue was significantly reduced in the
   sh-NC + Angiopep-2/SIRT1 group, while it was significantly increased in
   the sh-SIRT1 group. Furthermore, compared to the
   sh-NC + Angiopep-2/SIRT1 group, neuronal apoptosis was significantly
   increased in the Angiopep-2/SIRT1 + sh-SIRT1 group (Fig. [139]8A).
   ELISA results indicated that, compared to the sh-NC group, the
   sh-NC + Angiopep-2/SIRT1 group exhibited a significant increase in the
   content of SOD and a significant decrease in MDA in brain tissue.
   Conversely, the sh-SIRT1 group showed a significant decrease in SOD
   content and a significant increase in MDA content. Additionally,
   compared to the sh-NC + Angiopep-2/SIRT1 group, the
   Angiopep-2/SIRT1 + sh-SIRT1 group exhibited a significant decrease in
   SOD content and a significant increase in MDA content (Fig. [140]8B-C).
   PI staining results showed that compared to the sh-NC group, PI
   fluorescence in the brain tissue was significantly reduced in the
   sh-NC + Angiopep-2/SIRT1 group, while it was significantly increased in
   the sh-SIRT1 group. Furthermore, compared to the
   sh-NC + Angiopep-2/SIRT1 group, PI fluorescence was significantly
   increased in the Angiopep-2/SIRT1 + sh-SIRT1 group (Fig. [141]8D).
   RT-qPCR results indicated that compared to the sh-NC group, the mRNA
   levels of IL-6, TNF-α, and TGF-β in brain tissue were significantly
   downregulated in the sh-NC + Angiopep-2/SIRT1 group, while they were
   significantly upregulated in the sh-SIRT1 group. Additionally, compared
   to the sh-NC + Angiopep-2/SIRT1 group, the Angiopep-2/SIRT1 + sh-SIRT1
   group exhibited significantly upregulated mRNA levels of IL-6, TNF-α,
   and TGF-β (Fig. [142]8E).

Fig. 8.

   [143]Fig. 8
   [144]Open in a new tab

   Effects of Angiopep-2/SIRT1 Nanoparticles on Necroptosis. A TUNEL
   staining to detect apoptosis in mouse brain tissue, scale bar: 50 μm;
   (B) ELISA to measure SOD content in mouse brain tissue; (C) ELISA to
   measure MDA content in mouse brain tissue; (D) PI staining to detect
   apoptosis in mouse brain tissue, scale bar: 50 μm; (E) RT-qPCR to
   measure mRNA levels of IL-6, TNF-α, and TGF-β in mouse brain tissue.
   Each group consisted of 6 mice, and values are presented as
   mean ± standard deviation. *** indicates p < 0.001

   These experimental results demonstrate that Angiopep-2/SIRT1
   nanoparticles effectively mitigate sevoflurane-induced cellular and
   tissue damage by reducing the expression of necroptosis-related
   proteins and enhancing the cellular antioxidant defense system. These
   nanoparticles not only improved cell survival and structural integrity
   at the cellular level but also enhanced the cognitive and motor
   functions of the treated mice in vivo. This finding highlights the
   potential therapeutic value of Angiopep-2/SIRT1 nanoparticles in the
   prevention and treatment of DNT.

Discussion

   This study utilizes multi-omics approaches to elucidate the mechanisms
   by which self-assembled Angiopep-2/SIRT1 nanoparticles alleviate
   sevoflurane-induced DNT, showcasing significant innovation compared to
   previous research. Earlier studies have primarily focused on
   sevoflurane-induced neuronal apoptosis and inflammatory responses,
   lacking in-depth exploration of the specific molecular mechanisms
   involved [[145]56, [146]57]. Through systematic multi-omics analysis,
   our research reveals key changes in genes and proteins, offering new
   insights into the understanding of sevoflurane-induced DNT. This
   approach not only provides comprehensive molecular-level information
   but also uncovers complex biological networks and signaling pathways,
   laying a solid foundation for future research.

   SIRT1, an important deacetylase, has been confirmed to play a
   neuroprotective role in multiple studies [[147]58]. Our study found
   that SIRT1 significantly alleviates sevoflurane-induced neurotoxicity
   by inhibiting necroptosis, which is consistent with previous research
   findings. However, past studies have primarily focused on the
   regulation of apoptosis by SIRT1, whereas our study is the first to
   reveal its crucial role in necroptosis [[148]59–[149]61]. Compared to
   other studies on neurotoxicity and neurodegenerative diseases, our
   research further expands the protective scope of SIRT1, highlighting
   its broad application potential under different pathological
   conditions.

   The development of self-assembled Angiopep-2/SIRT1 nanoparticles
   represents another major innovation in this study. Compared to other
   nano-drug delivery systems, Angiopep-2/SIRT1 nanoparticles offer
   superior stability and biocompatibility, efficiently crossing the BBB
   to achieve targeted delivery [[150]62]. Previous research has shown
   that many nanoparticles suffer from poor stability and targeting
   ability in vivo, limiting their clinical applications. Our study
   employed various characterization methods and biocompatibility tests to
   confirm the superior performance of Angiopep-2/SIRT1 nanoparticles,
   demonstrating their significant potential in the treatment of
   neurological diseases.

   The application of multi-omics approaches in this study further
   enhances the reliability and comprehensiveness of the research
   findings. Through high-throughput sequencing, proteomics, and
   metabolomics analyses, we not only identified key genes and proteins
   involved in sevoflurane-induced DNT but also analyzed related metabolic
   pathways. Compared to single-omics methods, multi-omics approaches
   provide more comprehensive biological information, revealing complex
   biological networks and signaling pathways [[151]63–[152]65]. The
   results of our multi-omics analysis offer new insights and directions
   for understanding the molecular mechanisms underlying
   sevoflurane-induced DNT.

   The importance of in vivo and in vitro experiments in validating the
   intervention effects of nanoparticles cannot be overstated. In this
   study, we systematically evaluated the protective effects of
   Angiopep-2/SIRT1 nanoparticles using both in vivo mouse models and in
   vitro cell experiments. The in vivo experiments demonstrated that the
   nanoparticles significantly alleviated sevoflurane-induced neurological
   deficits and neuronal apoptosis in mice. The in vitro experiments
   further confirmed the protective effects of the nanoparticles at the
   cellular level. The consistency between in vivo and in vitro results
   strengthens the credibility and reproducibility of our findings.

   Necroptosis plays a crucial role in sevoflurane-induced DNT
   [[153]66–[154]68]. Through various experimental techniques, this study
   detailed the regulatory mechanisms of self-assembled Angiopep-2/SIRT1
   nanoparticles on necroptosis (Fig. [155]9). Unlike previous studies
   [[156]12], this research systematically explored the role of
   necroptosis in sevoflurane-induced DNT, filling a significant gap in
   the field. By inhibiting the expression of p-RIPK1, p-RIPK3, and
   p-MLKL, SIRT1 effectively reduced necroptosis, thereby protecting
   neuronal cells.

Fig. 9.

   [157]Fig. 9
   [158]Open in a new tab

   Molecular Mechanisms of Self-Assembled Angiopep-2/SIRT1 Nanoparticles
   in Alleviating Sevoflurane-Induced DNT by Regulating Necroptosis
   Through SIRT1 Activation

   Inflammatory responses are a major pathological process in
   sevoflurane-induced DNT [[159]69–[160]71]. This study found that
   Angiopep-2/SIRT1 nanoparticles significantly attenuate inflammatory
   responses by regulating the expression of IL-6, TNF-α, and TGF-β.
   Compared to other anti-inflammatory strategies, our approach not only
   effectively reduces inflammation but also reveals its precise molecular
   mechanisms through multi-omics analysis. The dual role of nanoparticles
   in modulating inflammatory responses underscores their broad
   application potential in neuroprotection.

   In conclusion, this study successfully developed self-assembled
   Angiopep-2/SIRT1 nanoparticles and explored their effects on inhibiting
   sevoflurane-induced DNT. Compared to existing research, the innovation
   of this study lies in using Angiopep-2 as a carrier, which effectively
   enhances the intracellular delivery efficiency of SIRT1. Although
   previous studies have demonstrated that SIRT1 can protect neurons
   through antioxidant and anti-inflammatory actions, its in vivo
   application has been limited by the lack of an effective delivery
   system. By utilizing Angiopep-2/SIRT1 nanoparticles, this study not
   only improved the stability and bioavailability of SIRT1 in neural
   cells but also significantly enhanced its efficacy against DNT.

   This study, through multi-omics approaches and in vivo and in vitro
   experiments, revealed the mechanisms by which self-assembled
   Angiopep-2/SIRT1 nanoparticles alleviate sevoflurane-induced DNT,
   providing new molecular targets and therapeutic strategies. By
   activating SIRT1 and inhibiting necroptosis and inflammatory responses,
   the nanoparticles significantly improved neural function impairment.
   These findings not only deepen our understanding of the mechanisms
   underlying sevoflurane-induced neurotoxicity but also offer a potential
   clinical intervention method that could enhance the safety of pediatric
   anesthesia and reduce the risk of long-term neurological damage
   [[161]5, [162]43].

   In this study, Angiopep-2/SIRT1 self-assembling nanoparticles were
   successfully prepared using the anti-solvent precipitation method. The
   resulting nanoparticles demonstrated good particle size uniformity and
   biocompatibility, making them suitable for laboratory-scale preparation
   and application. However, the industrial-scale production of
   nanoparticles still faces numerous challenges, including control of
   batch-to-batch consistency, stability of physicochemical properties
   during scale-up, and optimization of production costs. Future studies
   should focus on the standardization and optimization of process
   parameters to ensure the stability and functionality of the
   nanoparticles during large-scale manufacturing, thereby promoting their
   clinical translation potential [[163]72, [164]73].

   Although this study verified the biocompatibility of Angiopep-2/SIRT1
   nanoparticles through in vitro cell viability assays, hemolysis tests,
   and short-term animal experiments, the long-term in vivo metabolism,
   accumulation, and potential toxicity of the nanoparticles have not yet
   been systematically evaluated. Nanomaterials may undergo slow
   degradation in vivo or accumulate in specific organs (such as the
   liver, kidneys, spleen, and brain), potentially triggering chronic
   inflammation or immune responses [[165]74]. Therefore, future work
   should include long-term toxicological studies, covering the
   nanoparticles’ in vivo biodistribution, metabolic pathways, immune
   activation, and potential tissue toxicity, with particular emphasis on
   safety monitoring within brain tissue to comprehensively assess
   clinical application risks.

   Angiopep-2, as a brain-penetrating peptide, primarily mediates
   brain-targeted delivery of nanoparticles through low-density
   lipoprotein receptor-related protein 1 (LRP1). Given that LRP1 is
   expressed in various tissues, there is a risk of off-target
   distribution of nanoparticles in non-target organs. Such off-target
   effects may lead to unintended biological responses, thereby affecting
   therapeutic safety and efficacy. In future studies, we will use
   high-sensitivity imaging techniques and multi-organ distribution
   analyses to systematically evaluate the targeting specificity of the
   nanoparticles. Moreover, strategies such as adjusting Angiopep-2
   density and surface modifications (e.g., PEGylation) will be explored
   to reduce off-target effects and enhance therapeutic precision
   [[166]75]. Additionally, since nanoparticles may activate the immune
   system and trigger immune responses such as complement activation,
   comprehensive immune compatibility testing will be essential for safety
   assessment.

   This study demonstrated that Angiopep-2/SIRT1 nanoparticles alleviate
   sevoflurane-induced DNT by activating SIRT1 and inhibiting necroptosis,
   exhibiting favorable brain-targeting ability and biocompatibility. In
   recent years, lipid nanoparticles (LNPs) delivering mRNA encoding the
   necroptosis effector MLKL have been shown to induce tumor cell
   necroptosis and activate immune responses, exhibiting strong antitumor
   potential [[167]76]. Furthermore, CRISPR-Cas9 systems are being
   developed for gene regulation in central nervous system (CNS) diseases,
   enabling precise modulation of pathological gene expression, reduction
   of off-target effects, and improved therapeutic specificity [[168]77].
   In comparison, the Angiopep-2/SIRT1 nanoparticles in this study
   primarily activate endogenous protective factor SIRT1, inhibit
   necroptosis signaling pathways, and reduce inflammatory factor
   expression. This approach focuses on regulating intracellular signaling
   and metabolic pathways, offering advantages in terms of
   biocompatibility and safety. Unlike RNA therapy and gene editing, this
   strategy avoids the direct introduction of exogenous genes, potentially
   reducing immunogenicity and off-target risks; however, it may be
   relatively limited in its capacity for precise gene-level modulation
   [[169]78]. In the future, integrating this nanoparticle platform with
   RNA or CRISPR technologies may yield synergistic effects, enabling more
   efficient regulation of programmed necrosis and neuroprotection.
   Despite the significant progress made in elucidating the mechanisms of
   sevoflurane-induced DNT, this study has certain limitations. First, the
   relatively small sample size may affect the generalizability of the
   findings. Second, the experimental conditions and selected models may
   not fully replicate clinical scenarios, thus limiting the extrapolation
   of the results. Moreover, the long-term safety and efficacy of the
   nanoparticles remain to be further validated.

   Future research should focus on increasing the sample size and
   optimizing the experimental design to validate the broad applicability
   and consistent efficacy of self-assembled Angiopep-2/SIRT1
   nanoparticles. Additionally, the potential application of these
   nanoparticles in various neurotoxic and neurodegenerative diseases
   should be further explored, and their long-term safety and efficacy in
   clinical settings should be assessed. Through continued research and
   clinical trials, we aim to develop more effective and safer therapeutic
   strategies for sevoflurane-induced DNT, thereby enhancing anesthesia
   safety and protecting children’s brain health.

Supplementary Information

   [170]Supplementary material 1^ (588.3KB, jpg)
   [171]Supplementary material 2^ (402.6KB, jpg)
   [172]Supplementary material 3^ (479.7KB, jpg)
   [173]Supplementary material 4^ (442.4KB, jpg)
   [174]Supplementary material 5^ (480.3KB, jpg)
   [175]Supplementary material 6^ (598KB, jpg)
   [176]Supplementary material 7^ (1.1MB, jpg)
   [177]Supplementary material 8^ (1.2MB, jpg)
   [178]Supplementary material 9^ (2.2MB, jpg)
   [179]Supplementary material 10^ (24.4KB, docx)

Abbreviations

   AO
          Acridine Orange

   BBB
          Blood-Brain Barrier

   BP
          Biological Process

   CC
          Cellular Components

   CLSM
          Confocal Laser Scanning Microscope

   DIA
          Data-Independent Acquisition

   DLS
          Dynamic Light Scattering

   DMs
          Differential Metabolites

   DNT
          Developmental Neurotoxicity

   FACS
          Fluorescence-Activated Cell Sorting

   FA
          Formic Acid

   FBS
          Fetal Bovine Serum

   GO
          Gene Ontology

   H&E
          Hematoxylin and Eosin

   HPLC
          High-Performance Liquid Chromatography

   IL-6
          Interleukin-6

   KEGG
          Kyoto Encyclopedia of Genes and Genomes

   MDA
          Malondialdehyde

   MF
          Molecular Function

   MWMT
          Morris Water Maze Tests

   NORT
          Novel Object Recognition Test

   OPLS-DA
          Orthogonal Partial Least Squares-Discriminant Analysis

   PCA
          Principal Component Analysis

   PBS
          Phosphate-Buffered Saline

   PDI
          Polydispersity Index

   PI
          Propidium Iodide

   rRNA
          Ribosomal RNA

   RBF
          Radial Basis Function

   RFE
          Recursive Feature Elimination

   SEM
          Scanning Electron Microscopy

   SIRT1
          Sirtuin 1

   SOD
          Superoxide Dismutase

   SVM
          Support Vector Machine

   TEM
          Transmission Electron Microscopy

   TGF-β
          Transforming Growth Factor-beta

   TNF-α
          Tumor Necrosis Factor-alpha

   WB
          Western Blot

Author contributions

   Y.C. and X.Z. contributed equally to this work. Y.C. and S.Z. designed
   the research. X.Z. and G.Q. performed the experiments and data
   analysis. Y.C. conducted the behavioral assays and contributed to data
   interpretation. S.Z. and G.Q. supervised the project and provided
   critical revisions to the manuscript. All authors reviewed and approved
   the final version of the manuscript.

Funding

   Not applicable.

Data availability

   All data can be provided as needed.

Declarations

Ethics approval and consent to participate

   All animal experiments were approved by the Animal Ethics Committee of
   the First Hospital of China Medical University.

Consent for publication

   All authors participating in this research gave their full consent to
   publish the findings.

Competing interests

   The authors declare no competing interests.

Footnotes

   Publisher’s note

   Springer Nature remains neutral with regard to jurisdictional claims in
   published maps and institutional affiliations.

   Yi Chang and Xue Zhang are regarded as co-first authors.

Contributor Information

   Shuo Zhang, Email: zhangshuo1005@sina.com.

   Ge Qu, Email: 546653916@qq.com.

References