Propofol addiction may cause brain cell aging and memory loss by disrupting cell cleanup and protective proteins

The Cdkn2a^flox/flox (p16^{INK4a flox/flox}) mice were (Cat. NO. NM-CKO-234392) were purchased from Shanghai Model Organisms Center, Inc. CaMKIIa-Cre mice were also provided by Shanghai model organisms. The Cdkn2a^flox/flox (p16^{INK4a flox/flox}) mice were mated with the CaMKIIa-driven Cre recombinase transgenic mice, and a neuronal-specific p16^INK4a-deficient mouse model (p16^INK4a-CKO mice) was generated. Atg7^flox/flox mice (Cat. NO. NM-CKO-220047) were purchased from Shanghai Model Organisms Center, Inc. The ATG7^flox/flox mice were bred with the CaMKIIa-driven Cre recombinase transgenic mice, and a neuronal-specific ATG7-deficient mouse model (ATG7-CKO mice) was generated. All mice were bred in the Experimental Animal Central of Ningbo University, and mice were housed under a 12 h light/dark cycle with controlled humidity (40%–60%) and temperature (23 ± 1 °C). For autophagy inhibition, C57B/L6 mice received intranasal injection of Lys05 (10 mg/kg, Selleck, USA, Catalog S8369) once daily from the day of propofol exposure until the mice were sacrificed. All animals used in this study were 2-month-old male mice.

Wild-type C57BL/6 Mice:

These animals were used to establish the baseline phenotypical and molecular effects of chronic propofol exposure. This model is well-characterized in neuroscience research and provides a standardized genetic background, minimizing variability in behavioral and biochemical readouts related to cognitive function and cellular senescence.

p16^INK4a Knockdown Mice (p16^INK4a-CKO mice):

To directly investigate the causal role of p16^INK4a in propofol-mediated senescence and cognitive decline, we employed a neuron-specific knockdown model. This allowed us to demonstrate that reduction of p16^INK4a expression rescues both synaptic and cognitive deficits, confirming its pivotal role in the pathway.

Autophagy-Deficient Mice (ATG7-CKO mice):

This neuron-specific autophagy knockout model was selected to interrogate the functional contribution of autophagy to ADAR1 degradation and subsequent senescence activation. Its use provides targeted genetic evidence that autophagic flux is necessary for propofol-induced senescence and behavioral impairment.

We have complied with all relevant ethical regulations for animal use. All animal experiments were performed in strict adherence to the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and received full ethical approval from the Institutional Animal Care and Use Committee at Ningbo University (Approval ID: NBU-IACUC-2021-10476).

For molecular and biochemical analyses: Mice were sacrificed by cervical dislocation under deep anesthesia (induced by isoflurane, R510-22-10, RWD life science, China) 24 hours after the last self-administration session to assess the sustained effects of propofol. For histological analyses: Mice were terminally anesthetized and underwent transcardial perfusion with ice-cold PBS followed by 4% paraformaldehyde (PFA, P0099-100ml, Beyotime Biotechnology, China) to fix the brains prior to extraction and sectioning.

The study used propofol (Diprivan®, 10 mg/mL), which was provided by the Department of Anesthesiology at Ningbo University First Affiliated Hospital and dissolved in 0.9% saline. All animal procedures were carried out in full compliance with the Ningbo University-approved protocol (NBU-IACUC-2021-10476), Good Clinical Practice guidelines, and institutional policies for high-alert medications. Mice were anesthetized by intraperitoneal injection of 1.0% sodium pentobarbital (50 mg/kg). A silastic catheter (Dow Corning, 508-001) was surgically implanted into the jugular vein and connected subcutaneously to a cannula, which was attached to a polyethylene assembly mounted on the dorsal side of the mouse. To maintain catheter patency and prevent postoperative infection, the catheters were flushed daily with sterile saline containing 0.4% heparin sodium¹⁷. Mice were trained in operant chambers where a nose-poke in the left hole was designated as an active response, while the right hole served as an inactive control. Each active nose-poke under the fixed ratio 1 (FR1) schedule triggered an infusion of 2 mg/kg/infusion propofol, followed by a 20-second timeout period. Daily FR1 sessions lasted for 4 hours or until a maximum of 200 infusions was reached. Stable propofol SA was defined as achieving at least 25 infusions per session, an active-to-inactive response ratio greater than 2:1, and less than 20% variability in daily infusions over three consecutive sessions.

Following acquisition, mice underwent propofol SA under either FR1 or progressive ratio (PR) schedules. PR sessions extended for 6 hours daily, with the response requirement for each subsequent infusion increasing according to the formula: (Response ratio = [5e ^{(injection number x0.18)}] – 5). The breakpoint under the PR schedule was defined as the highest number of nose-pokes completed for the final infusion before a 1-hour period without reinforcement. Mice proceeded to subsequent experiments only after the breakpoint stabilized, with less than 15% variability in infusions over three days.

A circular pool (diameter: 120 cm, height: 50 cm) was filled with milk-tinted water maintained at 25 ± 1 °C. A platform was hidden below the water surface in one pool quadrant. Mice were trained in four trials daily for five consecutive days, starting at different positions for each trial, and allowed to swim freely for 90 s. The time mice spent swimming to the platform was recorded as escape latency; however, if a mouse failed to find the platform within 90 s, the researcher directed and kept it on the platform for 15 s. Probe trials were performed after the last trial of the acquisition period, and the platform was removed. The mice were allowed to swim freely for 90 s. The time spent in the target quadrant that previously contained the platform, and the number of platform location crosses were recorded and the EthoVision XT software was used for tracking.

The Y-maze apparatus consisted of three identical opaque acrylic arms (each 40 cm long × 10 cm wide × 15 cm high) positioned at 120° angles. The test was conducted under dim light (∼50 lux) and low-noise conditions. Mice were individually placed at the end of one arm and allowed to freely explore the maze for 8 min. Behavior was recorded and analyzed using an automated video-tracking system (EthoVision XT). An arm entry was defined as the mouse placing all four paws into the arm. Spontaneous alternation was defined as consecutive entries into all three arms without repetition. The percentage alternation was calculated as: Alternation (%) = [(Number of spontaneous alternation)/ (Total number of arm entries - 2)] × 100%. Mice with fewer than 8 total arm entries were excluded from analysis due to insufficient exploratory activity.

The NOR test was performed in an open-field arena (50 × 50 × 40 cm) under consistent lighting (∼60 lux). During the training phase, two identical objects (e.g., glass cubes, 5 × 5 × 5 cm) were placed in two opposite corners, 10 cm from the walls. Each mouse was allowed to explore the objects freely for 10 min. After a 90-min retention interval, one object was replaced with a novel object of similar size but distinct shape/texture. The mouse was reintroduced for a 10-min test session. Exploration was defined as directing the nose toward the object within ≤2 cm; climbing or sitting on the object was not counted. Sessions were video-recorded and analyzed blind to treatment groups. The recognition index (RI) was calculated as: RI = Time spent exploring novel object/(Time spent exploring novel + familiar object) ×100%. A preference Index (PI) during the training phase was also assessed to confirm absence of innate bias: PI = Time exploring one identical object/Time exploring both identical objects× 100%. Objects and positions were counterbalanced across groups to avoid place preference.

Cells were fixed with 3% paraformaldehyde in PBS (pH 6.0) and subsequently incubated in an X-gal staining solution containing 1 mg/ml 5-bromo-4-chloro-indolyl-β-D-galactopyranoside, 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 150 mM NaCl, and 2 mM MgCl2 in PBS (pH 6.0) for 16–20 hours at 37 °C. After staining, images were captured using a dissecting microscope. The results shown are representative of three independent experiments, each demonstrating consistent staining patterns¹⁸.

Brain slices or cells were fixed with 4% paraformaldehyde for 20 minutes at room temperature and subsequently washed three times with phosphate-buffered saline (PBS). The samples were blocked with a solution of 10 mg/ml bovine serum albumin (BSA) in PBS for 1 hour. Primary antibody was applied to the brain slices and incubated at 4°C overnight. After incubation, the slices were washed three times with PBS, each wash lasting 10 minutes. The slices were then incubated with a fluorescent secondary antibody (Proteintech, Chicago, USA) in the dark for 2 hours. Following this, the slices were stained with 4′,6-diamidino-2-phenylindole (DAPI; Solarbio, Beijing, China) for 10 minutes to visualize nuclei. To preserve fluorescence, an anti-fade mounting medium was applied, and the slices were coverslipped. Imaging was conducted using a confocal laser scanning microscope (Leica TCS SP8, Germany)¹⁹. The primary antibodies used in this study included LC3 (14600-1-AP, 1:500), ADAR1 (14330-1-AP, 1:200), PML (21041-1-AP, 1:200), HIRA (83194-5-RR, 1:200) and macroH2A1.2 (61427, 1:50) from Proteintech.

HT22 cells (Cellverse Co., Ltd, China) were cultured in DMEM (Hyclone) with 10% FBS (Gibco) and 100 units/mL penicillin/streptomycin. Hippocampi were dissected from newborn mice brains in ice-cold Hanks' Balanced Salt Solution (HBSS, Gibco). The tissues were enzymatically digested with 0.25% trypsin-EDTA for 15 minutes at 37 °C, followed by termination of digestion using DMEM (Gibco) supplemented with 10% FBS (Gibco). The digested tissues were gently triturated with a fire-polished glass pipette to dissociate neurons. The cell suspension was centrifuged at 800 × g for 5 minutes, and resuspended in neuronal culture medium consisting of Neurobasal-A medium, 2% B-27 supplement, 0.5 mM glutamine, and 1% penicillin-streptomycin. Cells were plated onto poly-D-lysine-coated culture plates or coverslips at a density of 5 × 10⁴ cells/cm² and maintained in a humidified incubator at 37 °C with 5% CO₂. Half of the culture medium was replaced every 3 days. Neuronal cultures were typically used for experiments after 7–10 days in vitro²⁰.

We purchased pADV-U6-shCdkn2a-CMV-EGFP, pADV-U6-blank-CMV-EGFP, AAV-CMV-Cdkn2a-EGFP and AAV-blank-CMV-EGFP from Shanghai Genechem Co.,Ltd. The AAV vectors were stereotaxically injected into hippocampus with a glass pipette at a slow flow rate of 60 nl/min. After the viral injection, the pipette was withdrawn 10 min. The stereotaxic coordinates used for hippocampus injections are defined as follows: Anteroposterior (AP): −2.0 mm to −2.3 mm, Mediolateral (ML): ±1.4 mm to ±1.8 mm, Dorsoventral (DV): −1.4 mm to −1.8 mm below dura mater.

Protein samples (50 μg) were separated by SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE) and subsequently transferred onto a nitrocellulose membrane. The membrane was blocked with 5% non-fat milk in Tris-buffered saline containing 0.1% Tween-20 (TBST) for 1 hour at room temperature, followed by incubation with the appropriate primary antibody diluted in blocking buffer overnight at 4 °C. After washing three times with TBST, the membrane was incubated with horseradish peroxidase (HRP)-conjugated secondary antibody (anti-rabbit or anti-mouse) for 1 hour at room temperature. Immunoreactive bands were visualized using enhanced chemiluminescence (ECL) reagent (GE Healthcare, Little Chalfont, UK) and imaged with a ChemiDoc™ Imager (Bio-Rad). Image acquisition was performed under non-saturating exposure conditions, and no post-capture alterations were applied¹⁸.

Cells were lysed in ice-cold lysis buffer composed of 50 mM Tris (pH 8.0), 150 mM NaCl, 1% NP-40, 10 mM EDTA, 5% glycerol, 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 µg/mL aprotinin, and 10 µg/mL leupeptin. For experiments performed under denaturing conditions, 1% SDS was included in the lysis buffer. Lysates were clarified by centrifugation at 16,000 × g for 25 minutes at 4 °C. To reduce non-specific binding, the clarified lysates were precleared by incubation with normal mouse or rabbit IgG bound to Protein A/G Sepharose beads for 2 hours at 4 °C with gentle rotation, followed by centrifugation at 18,000 × g for 10 minutes. One-tenth of each lysate was reserved as input control for subsequent Western blot analysis. Immunoprecipitation was performed by incubating the precleared lysates overnight at 4 °C with anti-Myc or anti-Flag antibody pre-conjugated to Protein A/G agarose beads. The immunoprecipitates were then collected and washed three times with ice-cold lysis buffer.

The primary antibodies used in this study included LC3 (14600-1-AP, 1:1000), ADAR1 (14330-1-AP, 1:1000), and β-actin (66009-1-Ig, 1:5000) from Proteintech. GPX4 (ab125066, 1:1000), SIRT1 (ab110304, 1:1000) were purchased from Abcam. LAMP-2a (51-2200, 1:1000) from Invitrogen. p16^INK4a (ZMS1072, 1:1000) was purchased from sigmaaldrich.

Total RNA was isolated from cells or tissues using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) through chloroform phase separation and isopropanol precipitation. RNA concentration and purity were determined spectrophotometrically. Complementary DNA (cDNA) was synthesized from 1 µg of total RNA using a High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Waltham, MA, USA) in a 20 µL reaction volume under the following conditions: 25 °C for 10 minutes, 37 °C for 120 minutes, and 85 °C for 5 minutes.

Quantitative PCR was carried out using Power SYBR® Green PCR Master Mix (Thermo Fisher Scientific) on an Applied Biosystems 7500 Real-Time PCR System. Each 20 µL reaction contained 1 µL of cDNA template, 10 µL of Master Mix, 200 nM of each forward and reverse primer, and nuclease-free water. The amplification protocol consisted of an initial denaturation at 95 °C for 10 minutes, followed by 35 cycles of denaturation at 94°C for 20 seconds, annealing at 60°C for 20 seconds, and extension at 72°C for 35 seconds. Melting curve analysis was performed to confirm amplification specificity. Gene expression levels were normalized to GAPDH or β-actin as internal controls and analyzed using the 2^−ΔΔCT method.

Total RNA was extracted by TRIzol (Invitrogen, USA) and reverse transcribed using PrimeScript RT reagent Kit (Takara, Japan). qRT-PCR was performed on an ABI StepOne Plus using SYBR Green^® Premix Ex Taq (Takara, Japan) as described previously¹⁸. The relative gene expression level was normalized to GAPDH and then calculated by means of the 2^−ΔΔCT method. Primer sequences are listed in a new Supplementary Table 1.

Validation of ADAR1 and SIRT1 Expression Changes:

Protein Level: Protein expression of ADAR1, SIRT1, and p16^INK4a was analyzed by Western blot. Hippocampal tissues or cultured neurons were lysed in RIPA buffer. Equal amounts of protein (20–30 μg) were separated by SDS-PAGE, transferred to PVDF membranes, and incubated overnight at 4 °C with primary antibodies: anti-ADAR1 (1:1000), anti-SIRT1 (1:1000), anti-p16^INK4a (1:1000), and anti-β-actin (1:5000) as a loading control. Band intensities were quantified using ImageJ software (NIH) and normalized to β-actin.

mRNA Level: Total RNA was extracted using TRIzol reagent. cDNA was synthesized with a Reverse Transcription Kit. mRNA levels were determined by qRT-PCR using SYBR Green Master Mix. The 2^−ΔΔCT method was used for relative quantification, with Gapdh as the reference gene. Primer sequences are listed in a new Supplementary Table 1.

Analysis of Autophagic Activity:

Autophagic flux was assessed by measuring the protein levels of LC3-II via Western blot (1:1000). The interaction between LC3 and ADAR1 was confirmed by Co-IP. Protein lysates were incubated with an anti-ADAR1 antibody or IgG control overnight at 4 °C, followed by protein A/G beads. The precipitated complexes were analyzed by Western blot using anti-LC3 and anti-ADAR1 antibodies.

Functional Validation through Gain- and Loss-of-Function Studies:

To establish a causal link, we employed both pharmacological and genetic approaches.

In vitro: Primary hippocampal neurons were transfected with ADAR1-overexpression plasmid or siRNA targeting Adar1 prior to propofol exposure. The efficiency of manipulation was confirmed by qRT-PCR and Western blot. In vivo: The functional role of autophagy was validated using neuronal-specific Atg7 knockout mice, which we have now described in greater detail above.

All animals that underwent successful surgery and were confirmed to carry the intended genotype were included in the study. Animals were excluded from the analysis a priori if they met any of the following criteria¹: post-surgical weight loss exceeded 20% of pre-surgical body weight²; exhibited signs of severe infection or poor health unrelated to the experimental manipulation³; catheter patency failure occurred prior to the completion of the behavioral experiment (verified by a rapid loss of righting reflex following a brief intravenous infusion of propofol); or⁴ technical failure during tissue processing or data acquisition. No data points were excluded during statistical analysis unless they were identified as significant outliers using the Grubbs' test (alpha = 0.05), which was established as the sole statistical exclusion criterion a priori.

A total of 21 mice were excluded from the final dataset for the reasons stated above. Specifically, 8 mice were excluded due to catheter patency failure, 9 mice due to post-surgical complications, and 4 data points due to technical failures. The final n for each group reported in the results represents only those animals and data points that passed all inclusion criteria.

The sample size (n) for each experimental group is reported in the corresponding figure legend. All n values represent biological replicates, not technical replicates. Sample sizes were chosen based on our previous experience with these models and behavioral tests to ensure sufficient statistical power, and are consistent with those commonly employed in the field.

Randomisation. Randomisation was used to allocate animals to treatment groupswithin each genotype. The randomisation sequence was generated using the =RAND() function in Microsoft Excel for each cohort of mice. For example, within a litter containing both p16^INK4a-CKO and control littermates, each mouse was assigned a random number and then allocated to the Propofol or Saline treatment group based on this sequence, ensuring an even distribution of littermates across groups where possible.

Minimisation of Confounders. To minimise potential confounding factors, the following strategies were employed:

Cage Location: All cages within the animal facility were systematically rotated on different shelves of the rack every three days throughout the experiment to avoid location-specific effects.

Order of Treatments and Measurements: The order in which mice underwent behavioral testing and tissue collection was randomized daily using a lottery system. This was done to ensure that any time-of-day effects or experimenter fatigue were distributed evenly across all experimental groups.

Batch Effects: The entire experiment was conducted over multiple replicated cohorts. Animals from all genotypes and treatment groups were represented in each cohort to prevent the confounding of experimental results with batch-to-batch variability.

A detailed study protocol outlining the primary research question, hypothesis, key experimental design, and planned analyses was prepared internally and approved by our institutional animal ethics committee prior to the commencement of the study (Approval ID: NBU-IACUC-2021-10476).

All schematic diagrams and graphical elements in Figs. 3L, 5M, and the Graphical Abstract were created from scratch by the authors using basic drawing software (Adobe Illustrator). No pre-existing templates, elements from databases, or third-party toolswere used. All content is original and owned by the authors.

All statistical analyses were performed using SPSS (version 20.0; IBM, USA). Data are presented as mean ± standard deviation (SD). Comparisons between two groups were made using an unpaired two-tailed Student's t-test. For comparisons among three or more groups, one-way or two-way analysis of variance (ANOVA) was applied. A p-value of less than 0.05 was considered statistically significant. The specific test used in each experiment is indicated in the corresponding figure legend. All experiments were independently repeated at least three times with consistent results. The sample size (n) for each experiment represents biological replicates, defined as cells or animals derived from distinct cultures or individuals. No statistical methods were used to predetermine sample size; however, sample numbers were chosen based on common practices in the field and our previous experience to ensure adequate power. Data inclusion and exclusion criteria were pre-established; no outliers were excluded unless clearly resulting from technical error, and such instances were noted during analysis. Efforts were made to ensure reproducibility through rigorous experimental standardization, including the use of aliquoted reagents, controlled passage numbers for cell lines, and blinded assessment during data collection and analysis where feasible.

The Cdkn2a^flox/flox (p16^{INK4a flox/flox}) mice were bred with the CaMKIIa-driven Cre recombinase transgenic mice, and a neuronal-specific p16^INK4a-deficient mouse model (p16^INK4a-CKO mice) was generated. The p16^INK4a-CKO mice prevented senescence markers despite propofol exposure (Fig. 3J, K). Conversely, AAV-mediated p16^INK4a overexpression in naive mice recapitulated propofol-induced pathology, increasing SASP factors 3.0-3.1-5.8-fold (TNF-α: 88.57 ± 2.3 pg/mg, IL-1β: 104.0 ± 1.2 pg/mg, IL-6: 152.5 ± 2.7 pg/mg) and impairing spatial memory (water maze latency: 51.3 ± 4.1 s vs 23.8 ± 2.7 s, p < 0.001) (Fig. 3L–N). These data establish p16^INK4a upregulation as both necessary and sufficient for propofol-induced neurosenescence.

Emerging evidence indicates that ADAR1 interacts with human antigen R (HuR), an RNA-binding protein regulating post-transcriptional stability whose binding to target RNAs is ADAR1-dependent. To investigate whether propofol modulates the ADAR1/SIRT1 axis through HuR-mediated regulation of SIRT1 mRNA, we conducted RNA immunoprecipitation (RIP) assays with anti-HuR antibodies. Our data demonstrated that propofol exposure markedly reduced HuR-SIRT1 mRNA complex formation (Fig. ). Notably, ADAR1 overexpression abolished the propofol-induced reduction in HuR/SIRT1 mRNA interactions (Fig. ), establishing that propofol-mediated ADAR1 downregulation drives SIRT1 mRNA destabilization through impaired HuR binding capacity. S1A S1B

Additionally, our data revealed that propofol exposure markedly enhanced the binding affinity between eukaryotic translation initiation factor 3 subunit A (EIF3a) and p16^INK4a mRNA (Fig. S1C), which correlated with concomitant increases in p16^INK4a translational efficiency as evidenced by polysome profiling assays (Fig. S1D). Notably, SIRT1 overexpression abolished the propofol-induced EIF3a-p16^INK4a mRNA interaction (Fig. S1E). Mechanistically, propofol-induced ADAR1 downregulation drives cellular senescence through a multistep regulatory cascade: (1) Reduced ADAR1 expression impairs HuR-mediated stabilization of SIRT1 mRNA, leading to accelerated SIRT1 transcript degradation; (2) Subsequent SIRT1 protein depletion activates EIF3a-dependent translational machinery; (3) Enhanced EIF3a-p16^INK4a mRNA interaction promotes selective translation of senescence-associated markers. These results suggest that ADAR1/SIRT1 pathway involved in propofol induced p16INK4a-dependent senescence.

To determine whether ADAR1 is an autophagy substrate in propofol induced neuronal senescence, we examined the interaction between LC3 and ADAR1. Firstly, LC3 and ADAR1 co-localization was observed in the cytoplasm of propofol-treated cells (Fig. 7G, H). The co-immunoprecipitation analysis results showed enhanced interaction between LC3 and ADAR1 in propofol-treated cells (Fig. 7I). Indeed, ADAR1 harbored four LC3-interacting region (LIR) motifs that could be potentially recognized by LC3 (Fig. 7J). We generated a line of HT22 cells stably expressing FlagADAR1 (HT22-Flag-ADAR1) and found that LIR3 and LIR4 ADAR1 mutants, which were marked as F632A/L635A (mut3) and F744A/V747A (mut4), showed significant resistance to propofol-induced degradation (Fig. 7K–N), and these two mutants failed to bind to LC3 (Fig. 7O, P). These results disclosed that propofol induces ADAR1 degradation by activating the autophagy pathway. These results support the autophagy-lysosome pathway of ADAR1 degradation during propofol-induced senescence.

Furthermore, under the FR1 reinforcement schedule, the ATG7-CKO-propofol mice group significantly reduced propofol active nose-poke responses (Fig. 8H) and propofol infusions (Fig. 8I) compared with the ATG7^flox/flox -propofol mice group. Under the PR reinforcement schedule, the ATG7-CKO-propofol mice group significantly reduced breakpoints compared with the ATG7^flox/flox -propofol mice group (Fig. 8J). These results suggested that propofol consumption and motivation were decreased by autophagy inhibition. These findings position pharmacological autophagy modulation as a promising therapeutic axis, capable of reversing established neurosenescence while disrupting the self-perpetuating cycle of propofol addiction through proteostasis restoration.