Animal experiments were approved by the Ethics Committee of Beijing Stomatological Hospital, Capital Medical University (Approval No. MDKN-2021-055). 18 months old male C57BL/6J mice, weighing 26–34 g, were purchased from the Model Animal Research Center of Nanjing University (Nanjing, China) and housed at 24 $\pm$ 1 °C and 59 $\pm$ 1% humidity with a 12-hour cycle of light/dark. Mice freely accessed to food and water. POCD model establishment was performed as described in the previous study [15]. Briefly, C57BL/6J mice underwent tibial fracture surgery under isoflurane anesthesia (anesthesia induction of 3.0% isoflurane and the maintenance of 1.5% isoflurane). During the operation, the temperature of mice was monitored through the heating pad and kept between 36 and 37 °C. Postoperative pain was treated locally with lidocaine (2%). In control group, mice were received 100% oxygen without any anesthesia/surgery procedure [15]. After 1 week, model mice were intraperitoneal injected with 2.5 mg/kg/day propofol for 8 weeks. Moreover, the overexpression vector of SIRT3 was constructed and packaged into adeno-associated virus (OBiO Technology, Shanghai, China) to upregulate the expression of this gene in the brain. Using brain stereolocation technology, SIRT3 overexpression vector (titers $>$1.0 $\times$ 10${}^{12}$) was injected into CA1 region of the brain (0.5 $\mu$L/side) 28 days before anesthesia/surgery.

The spatial memory of mice was determined by Morris water maze. Morris water maze testing was carried out in a round white pool (diameter: 100 cm; depth: 38 cm), in which a platform was placed. The pool contained 30-cm-depth water at 20 $\pm$ 2 °C and divided in four quadrants. The platform was placed in one quadrant. During the training period, each mouse was placed in a quadrant without the platform for a 2-min free swimming for habit. The 5-day training was performed 4 times every day. Time to found the platform were recorded as escape latency. When each mouse failed to find the platform over 120 s, the experimenter put it onto the platform. Each mouse should stay on the platform for 15 s. At day 6, the platform was removed. Each mouse was placed at the same site to enter the water. The times crossing the previous platform were recorded during 2 min.

After a 7-day habitation in cages, each mouse was placed in a box (25 $\times$25 cm) without the top to habit the experimental environment 24 h before testing. Two stone were placed in both ends of the same side of the box. Each mouse was placed in the opposite side. The behavior of the mouse contacting with the two stone was recorded during 10 min. Then, the mouse was removed from the box and placed in the cage. After 1 h, one of the two stone was replaced with a stick. The mouse was placed in the box and the contacting behavior was recorded. The behavior was recorded by a digital camera (Nikon, Tokyo, Japan). The recognition index was calculated as following: recognition index = new/(new + old).

After the last behavior test, mice were performed with the euthanize via decapitation. Hippocampus tissues were isolated and embedded into paraffins after dehydration using ethanol. Tissues were cut into 3-$\mu$m-thickness sections. Sections was performed with Dewaxing and rehydration for IHC. Sections soaked in 1$\times$ Citrate Unmasking Solution (Cell Signaling Technology, Danvers, MA, USA) to be heated under microwave for 10 min at 95 °C after the solution boiling. After incubated in 3% H${}_2$O${}_2$ for 10 min, sections were incubated with 1$\times$ TBST containing 5% normal goat serum (#5425, Cell Signaling Technology, Boston, MA, USA) at room temperature for 1 h. Sections were incubated with diluted LC3 antibody (#12741, 1:500, Cell Signaling Technology) at 4 °C overnight, followed by the culture with Boost Detection Reagent (Cell Signaling Technology) for 30 min at room temperature. Sections were stained by DAB solution for 5–10 min. Sections soaked in ddH${}_2$O were dehydrated by ethanol and sealed using Mounting Medium (Cell Signaling Technology). Numbers of the positive cells were determined under a microscope (Olympus, Tokyo, Japan).

The euthanize in C57BL/6J was processed by decapitation after the disinfection of 75% Alcohol. Brain tissues were isolated and then placed in D-hank’s solution (Solarbio, Beijing, China) without irons of calcium and magnesium. Hippocampus was separated from brain tissue, followed by the incubation with 0.25% Trypsin (Solarbio) for 15 min at 37 °C in a 15 mL conical tube. During incubation, the tube was inverted every 5 min. Then, supernatant in the tube was carefully removed. Cell precipitate was washed by D hank’s solution for 3 times and then resuspended in D hank’s solution. After the 2-min standing, cell suspension was centrifugated at 4 °C and 1000 rpm for 10 min. Cells resuspended in DMEM-F12 (Gibco, Grand Island, NY, USA) containing 10% fetal bovine serum (FBS; Gibco) were filtered using a 100-$\mu$m cell filter. Finally, 4 $\times$ 10${}^4$ filtered cells at were cultured in 24-well plates containing Poly-L-lysine (Sigma, St. Louis, MO, USA) for 4–6 h at 37 °C and then medium was replaced with FBS-free neuronal medium containing B27 (Thermo Fisher Scientific, Waltham, MA, USA). At 3th day after cultured in 24-well plates, 2.5 $\mu$g/mL Cytarabine (Selleck, Houston, TX, USA) was added to inhibit non-neuronal cells for 24 h. During cell culture, medium was replaced every 3 days.

6 groups were formed: control, LPS, LPS + propofol, LPS + propofol + pcDNA3.1, LPS + propofol + SIRT3 and LPS + propofol + 3-MA. To induce inflammation in neurons, cells were stimulated by 1 $\mu$g/mL LPS (in cell medium) for 24 h. Meanwhile, propofol (MedChem Express, Monmouth Junction, NJ, USA; purity: 99.52%) administration was performed in LPS-stimulated neurons at the dose of 10 $\mu$g/mL. The overexpression vector of SIRT3 was constructed based on pcDNA3.1. Cell transfected was performed in neurons via Lipofectamine 3000 kit (Invitrogen, Carlsbad, CA, USA). pcDNA3.1 was transfected in cells to be as the negative control of SIRT3 overexpression vector. 48 hours after cell transfection, SIRT3 expression was determined and cells were stimulated by LPS. To inhibited autophagy in neurons, cells were treated with 5 mM 3-MA (MedChem Express, Monmouth Junction, NJ, USA; purity: 99.83%) for 24 h and then incubated with LPS and propofol.

The fluorescent LC3 antibody was used to determine autophagy in neurons. First, neurons were incubated in 100% ice-cold methanol (Sigma) for 15 min at –20 °C after cell medium was remove. Cells were washed using PBS solution for 3 times. Then, cells were incubated with PBS solution (#9808, Cell Signaling Technology) containing 5% normal goat serum (#5425, Cell Signaling Technology) and 0.3% Triton X-100 to be blocked for 60 min at room temperature. LC3 antibody (#4108, 1:200, Cell Signaling Technology) diluted in PBS containing 1% bovine serum albumin (BSA; #9998, Cell Signaling Technology) and 0.3% Triton X-100 was added to incubate cells at 4 °C overnight. After that, cells were incubated with diluted secondary antibody (#4412, 1:500, Cell Signaling Technology) for 1 hour at room temperature, protected from light. Finally, washed cells were incubated with ProLong® Gold Antifade Reagent (#9071, Cell Signaling Technology) to avoid light quenching. Fluorescent LC3 was observed under a fluorescence microscope (Olympus, Tokyo, Japan).

We measured pro-inflammatory cytokines in hippocampus and neuron based on ELISA. According to the guideline of kits, pro-inflammatory cytokines including TNF-$\alpha$(EK282/3-01, MULTI SCIENCES, Hangzhou, China), IL-1$\beta$ (EK201B/3-01, MULTI SCIENCES) and IL-6 (EK206/3-01, MULTI SCIENCES) in cell and tissue lysates were measured after all experiments. To calculated the concentration of cytokines, the absorbance of each sample was read by a microplate reader (Thermo Fisher Scientific, LA, CA, USA) at 450 nm.

Before the measurement, cells or hippocampal tissues were incubated in RIPA buffer for lysate. Then, MDA and SOD level in lysates were measured by MDA kit (E-BC-K025-M, Elabscience, Wuhan, China) and SOD kit (E-BC-K020-M, Elabscience), respectively. Procedures of the measurement was followed by the product manual. The absorbance of each sample was read by the microplate reader at 532 nm (for MDA detection) and 450 nm (for SOD detection).

After the medium was removed, neurons were incubated with Trizol regnant (Solarbio) to extract total RNA. Based on One Step RT-qPCR Kit (Sangon, Shanghai, China), RNA was reversely transcribed into cDNA and quantified under a real-time PCR system (Applied Biosystem, Foster City, CA, USA). The procedure of reverse-transcription and quantification was as follows: reverse-transcription for 5 min at 50 °C; pre-denaturation for 3 min at 95 °C; denaturation for 10 s at 95 °C and Annealing/extension/acquisition of fluorescent signals for 30 s, 40 cycles. The relative quantitative expression was determined using the 2${}^{-\mathrm{\Delta }\mathrm{\Delta }\text{CT}}$ method. The sequences of primers were as follows: SIRT3-forward, 5′-GCTACATGCACGGTCTGTCGAA-3′; SIRT3-reverse, 5′-CAATGTCGGGTTTCACAACGCC-3′; TNF-$\alpha$-forward, 5′- GGTGCCTATGTCTCAGCCTCTT-3′; TNF-$\alpha$-reverse, 5′-GCC ATA GAA CTG ATG AGA GGG AG-3′; IL-1$\beta$-forward, 5′-TGG ACC TTC CAG GAT GAG GAC A-3′; IL-1$\beta$-reverse, 5′- GTT CAT CTC GGA GCC TGT AGT G-3′; IL-6-forward, 5′-TACCACTTCACAAGTCGGAGGC-3′; IL-6-reverse, 5′-CTGCAAGTGCATCATCGTTGTTC-3′; GAPDH-forward, 5′-CAT CAC TGC CAC CCA GAA GAC TG-3′; GAPDH-reverse, 5′-ATG CCA GTG AGC TTC CCG TTC AG-3′.

Cell and tissue samples were incubated with the lysis buffer (Beyotime, Shanghai, China) consisting of 20 mM Tris, 150 mM NaCl and 1% Triton X-100 to extract proteins. The concentration of protein samples was measured by a BCA Protein Assay Kit (E-BC-K318-M, Elabscience). Protein samples were then separated using SDS-PAGE under an electrophoresis apparatus (Bio-Rad, Hercules, CA, USA), followed by the transport of blots from gels to nitrocellulose membranes (Cell Signaling Technology). The membranes were blocked by 1$\times$ Tris buffered saline + Tween (TBST, Cell Signaling Technology) containing 5% skimmed milk for 1 h at room temperature. Membranes were incubated with diluted primary antibody (in 1$\times$ TBST containing 5% bovine serum protein) at 4 °C, followed by the next incubation with HRP-linked antibody (#7074, 1:2000, Cell Signaling Technology) diluted in the block buffer at room temperature for 1 h. Enhanced chemiluminescence (ECL) kit (#6883, Cell Signaling Technology) was used to color blots. Membranes were incubated with ECL reagent for 1 min. An X-ray imaging system (Bio-Rad, Hercules, CA, USA) was used to observe and determine blots. Primary antibodies in this investigation were as follows: anti-SIRT3 (#5490, 1:1000, Cell Signaling Technology); anti-AMPK (ab207442, 1:1000, Abcam, Cambridge, MA, USA); anti-p-AMPK (ab133448, 1:2000, Abcam); anti-mTOR (ab134903, 1:10000, Abcam); anti-p-mTOR (#5536, 1:1000, Cell Signaling Technology); anti-LC3 (#4108, 1:1000, Cell Signaling Technology); anti-p62 (#23214, 1:1000, Cell Signaling Technology); anti-Beclin-1 (#3495, 1:1000, Cell Signaling Technology); anti-GAPDH (#5174, 1:1000, Cell Signaling Technology). GAPDH was the internal reference protein.

Experimental data were shown as mean $\pm$ standard deviation (SD). Unpaired t-test (for two groups) and One-way analysis of variance (for $\ge$three groups) were used to statistical comparison between groups. Among multiple groups ($\ge$three), Tukey’s multiple comparisons test was used for the post-hoc comparison. Significant difference in statistic was recognized when p value 0.05 at 95% confidence interval.

The spatial memory of mice was described by escape latency, time spending on the quadrant and the times crossing the platform based on water maze assay. Anesthesia/surgery increased escape latency (Fig. 1A) but declined spending time (Fig. 1B) and crossing times (Fig. 1C), indicating the poor capability of spatial memory. After propofol treatment, the escape latency, spending time and crossing times of mice with POCD were close to that of mice in sham group (Fig. 1A–C). Learning memory was tested by new object recognition index. Significantly, propofol enhanced the anesthesia/surgery-reduced recognition index (Fig. 1D), showing the improvement of learning memory in mice with POCD after propofol treatment. In the brain of mice, anesthesia/surgery elevated levels of TNF-$\alpha$, IL-1$\beta$ and IL-6 that were inhibited due to propofol treatment (Fig. 1E), indicating the anti-inflammatory role of propofol in POCD mice. Compared with sham, anesthesia/surgery induced oxidative stress with increased MDA level and decreased SOD activity in mice (Fig. 1F). However, levels of MDA and SOD in POCD mice were close to these in mice in sham after propofol treatment (Fig. 1F). Intriguingly, propofol treatment upregulated anesthesia/surgery-downregulated SIRT3 level in the hippocampus of mice (Fig. 1G). Moreover, LC3 and Beclin-1 levels were decreased and p62 level was increased in the hippocampus during POCD (Fig. 1G–I). Nevertheless, autophagy activation was observed in the same site during propofol treatment, accompanied by increases of LC3 and Beclin-1 and the decrease of p62 (Fig. 1G–I).

Fig. 1.

Propofol enhanced SIRT3 and autophagy in mice with POCD. (A) Escape latency in Morris water maze assay. (B) Time spending on the target quadrant based on Morris water maze. (C) Times crossing the platform in Morris water maze assay. (D) Recognition index was tested by novel object recognition test. (E) Levels of TNF-$\alpha$, IL-1$\beta$ and IL-6 in the hippocampus. (F) Levels of MDA and SOD in the hippocampus. n = 10. (G) SIRT3 expression in the hippocampus was measured using qPCR (n = 10) and western blot (n = 3). (H) LC3 level in the hippocampus was detected by IHC and the quantitative result. n = 3. (I) Autophagy-related protein expressions in the hippocampus were measured via western blot, including LC3, p62 and Beclin-1. n = 3. * p 0.05, ** p 0.01, *** p 0.001.

Primary neurons were isolated (Fig. 2A) and incubated with 1 $\mu$g/mL LPS to induce inflammatory condition during POCD in vitro. To determine the role of SIRT3 in POCD during propofol treatment, the overexpression vector of SIRT3 based on pcDNA3.1 was constructed to upregulate this gene in LPS-stimulated neurons (Fig. 2B,C). According to the measurement using ELISA, there was the marked enhance of pro-inflammatory cytokines including TNF-1$\alpha$, IL-6 and IL-1$\beta$ in neurons during LPS stimulation (Fig. 2D,E). Propofol inhibited these cytokines in LPS-stimulated neurons (Fig. 2D,E). Further, SIRT3 upregulation enhanced the anti-inflammatory of propofol in inflammatory neurons (Fig. 2D,E). Moreover, propofol could protect LPS-stimulated neurons from oxidative stress due to declined MDA and increased SOD (Fig. 2F). SIRT3 upregulation further promoted the protective role of propofol (Fig. 2E). Taken together, SIRT3 overexpression enhanced the protection of propofol in neurons with inflammation and oxidative damage.

Fig. 2.

SIRT3 enhanced the inhibitory effects of propofol on inflammation and oxidative stress in LPS-stimulated hippocampus neurons. (A) Images of isolated neurons. (B) SIRT3 expression in primary neurons was detected using qPCR. (C) SRIT3 expression in primary neurons was detected using western blot. (D) The expressions of pro-inflammation factors including TNF-$\alpha$, IL-1$\beta$ and IL-6 in primary neurons were measured using qPCR. n = 3. (E) Levels of TNF-$\alpha$, IL-1$\beta$ and IL-6 in primary neurons were measured using ELISA. n = 4. (F) Levels of MDA and SOD in primary neurons. n = 4. * p 0.05, ** p 0.01, *** p 0.001.

The phosphorylation of AMPK/mTOR pathway was enhanced in LPS-stimulated neurons during propofol treatment (Fig. 3A), suggesting that propofol could induce the activation of AMPK/mTOR pathway in inflammatory neurons. SIRT3 further elevated the propofol-enhanced phosphorylation level of AMPK and decreased mTOR phosphorylation level in inflammatory neurons (Fig. 3A). Moreover, propofol could reverse LPS-induced altered expressions of autophagy-related proteins including LC3, Beclin-1 and p62. Propofol caused the increases of LC3 and Beclin-1 and the decrease of p62 (Fig. 3B). SIRT3 upregulation enhanced the role of propofol in alterations of autophagy-related proteins (Fig. 3B). Collectively, SIRT3 upregulation enhanced propofol induced autophagy mediated by AMPK/mTOR pathway in LPS-treated neurons.

Fig. 3.

SIRT3 increased the role of propofol in AMPK/mTOR pathway mediated neuronal autophagy. (A) Expressions of AMPK/mTOR pathway-related proteins were measured using western blot. (B) Autophagy-related protein expressions were determined in neurons, including LC3, p62 and Beclin-1 via western blot. n = 3. * p 0.05, ** p 0.01, *** p 0.001.

To demonstrated whether autophagy mediated the anti-inflammatory role of propofol in neurons, 3-MA, an inhibitor of autophagy, was administrated in LPS-stimulated neurons during propofol treatment. For pro-inflammatory cytokines, 3-MA reversed propofol-decreased levels of TNF-$\alpha$, IL-1$\beta$ and IL-6 in LPS-stimulated neurons (Fig. 4A,B). For oxidative stress, 3-MA offset propofol-induced alterations in MDA and SOD in neurons undergoing inflammatory condition (Fig. 4C). These data indicated that autophagy contributed to the anti-inflammatory role of propofol in inflammatory neurons.

Fig. 4.

Autophagy contributed to the protective role of propofol in LPS-stimulated neurons. (A) Expressions of TNF-$\alpha$, IL-1$\beta$ and IL-6 in primary neurons were determined using qPCR. n = 3. (B) Levels of TNF-$\alpha$, IL-1$\beta$ and IL-6 in primary neurons were measured using ELISA. n = 4. (C) Levels of MDA and SOD in primary neurons. n = 4. * p 0.05, ** p 0.01, *** p 0.001.

Compared with propofol group, escape latency (Fig. 5A), spending time (Fig. 5B), crossing times (Fig. 5C) and recognition index (Fig. 5D) in propofol + SIRT3 group were closer to these in sham group. Also, further decreases of inflammation (Fig. 5E) and oxidative stress (Fig. 5F) were observed in propofol-treated mice after SIRT3 upregulation. In addition, SIRT3 upregulation also enhanced the regulatory role of propofol in autophagy-related proteins including LC3, Beclin-1 and p62 in the hippocampus of mice with POCD (Fig. 5G,H). Therefore, SIRT3 promoted the protective role of propofol in mice with POCD via activating autophagy.

Fig. 5.

SIRT3 promoted the protective role of propofol in mice with POCD. (A) Escape latency in Morris water maze assay. (B) Time spending on the target quadrant in Morris water maze assay. (C) Times crossing the platform in Morris water maze assay. (D) Recognition index in novel object recognition test. (E) Levels of TNF-$\alpha$, IL-1$\beta$ and IL-6 in the hippocampus. (F) Levels of MDA and SOD in the hippocampus. n = 10. (G) LC3 expression in the hippocampus were detected using IHC. n = 3. (H) Autophagy-related protein expressions in the hippocampus were measured via western blot, including LC3, p62 and Beclin-1. n = 3. * p 0.05, ** p 0.01, *** p 0.001.