For this study, a total of 48 aged (18 months old, weighing 45–50 g) male C57BL/6 mice were ordered from Sun Yat-sen University (Guangzhou, China). Five mice per cage were group-housed (at 22–25^°C) with food and water available ad libitum under a 12-h light/12-h dark cycle for 2 weeks to adapt to the environment. The animals were randomly assigned to four groups, namely, the CON group (no intervention), the CON + TNIL group (light therapy), the PND group (anesthesia, surgery, and sham light therapy), and the PND + TNIL group (anesthesia, surgery, and light therapy) (n = 12). Mice in the PND group were treated for sham light therapy with the device turned off; mice in the CON + TNIL and PND + TNIL groups received light therapy (18 J/cm², 3 days). All mice heads were shaved to eliminate hair interference. In the experiment, mice received isoflurane anesthesia and surgery on day 1; TNIL treatment was performed during days 1–3; behavioral testing was performed during days 7–8 (n = 12); the brain was harvested to perform biochemical (n = 6) and histological (n = 6) analyses on day 9. The schematic timeline of the experimental process is shown in Figure 1.

The right carotid artery exposure procedure was performed under isoflurane anesthesia to establish the PND model, as described in other studies (Min et al., 2022). Mice were anesthetized by 2% isoflurane and kept on spontaneous respiration during the procedure. Rectal temperature was monitored and maintained at 37^°C with the aid of a heating blanket (69020, RWD, China). After the mouse was anesthetized by isoflurane for at least 20 min, a 1.5-cm midline neck incision and soft tissue dissection with 1-cm long right common carotid artery exposure were carefully performed without any damage to the vagus nerve. Subsequently, the wound was irrigated and closed by using a 4–0 surgical suture. The surgical procedure was performed under sterile conditions and lasted around 12 min. The total duration of general anesthesia was 2 h, and an anesthesia monitor (B450, GE, USA) was used to dynamically monitor the depth of anesthesia and maintain the anesthesia level. No response to toe pinching was observed during the whole process. After the surgery, the incision was supplied with 2.5% lidocaine cream to alleviate the postoperative pain.

Aged mice in the CON + TNIL and PND + TNIL groups received laser treatment 4 h after surgery for 3 consecutive days. We used an 810 nm wavelength and 1 W maximum power output diode laser (HW810AD2000-34F, Shenzhen Infrared Laser Technology Co., Ltd., China), which emitted 810 nm CW wavelength near-infrared radiation. Mice were manually restrained by holding the body and shaved head, and a laser was fixed above the midline on the back of the head in the area between the eyes and ears. The spot size was 1 cm² and the distal tip of the fiber optic provided a power density of 120 mW/cm². The laser irradiation was continued for 150 s, and the total energy fluence was 18 J/cm². We also made a measurement of TNIL penetration across the skull of the mice, where the skull was excised from the head of the mice and placed on a foil-coated vessel with a calibrated light sensor at the bottom. Then, TNIL was irradiated to the skull, and the penetration power was recorded by the sensor (the laser was about 4 cm away from the skull, and the skull was about 3 cm away from the sensor), and the measured transmittance was about 8%. All complete parameters including device information, irradiation parameters, and treatment parameters used in TNIL treatment are reported in Table 1.

All behavioral experiments were conducted during the light phase between 8 a.m. and 6 p.m. in a sound-isolated room and an unbiased double-blind manner.

After behavioral tests, mice were deeply anesthetized by isoflurane and transcardially perfused with normal saline, after which the brain tissue was removed (n = 6). The hippocampus was isolated for subsequent experiments on genes and protein levels. All dissection procedures were performed on ice and stored at −80^°C before use. As for others, the transcardial perfusion was given with normal saline and 4% paraformaldehyde (n = 6). Brains were fixed in 4% paraformaldehyde solution for 24 h and then transferred to 10, 20, and 30% sucrose solution for 1 day each to dehydration. The brain tissue was embedded with optimum cutting temperature (OCT) (4583, SAKURA, JP) and then stored for subsequent immunofluorescent staining.

The immunofluorescent labeling and quantification of the staining were similar to previous studies (Zheng et al., 2017; Lai et al., 2021). Notably, 20-μm thick sections of the coronal brain were sequentially cut from bregma -2 to -4 mm in the hippocampus of mice. The sections were washed three times with phosphate-buffered solution (PBS) to remove OCT from the surface. Subsequently, permeable liquid (PBS plus 0.3% TritonX-100) was added and the sections were allowed to stand at room temperature (RT) for 10 min. Then, the sections were washed three times with PBS. After blocking with 5% goat serum (16210072, Gibco, USA) for 30 min at RT, the sections were incubated with the primary Iba-1 antibody (1:100, 10904-1-AP, Proteintech, China) at 4^°C overnight. Tissue sections were washed three times with PBS and then incubated with Cy3-labeled goat anti-rabbit IgG (H + L) (1:200, A0516, Beyotime, China) at RT for 2 h and then incubated with DAPI (G1012, Servicebio, China) at RT for 10 h. The tissue sections were sealed with an anti-fluorescence quenching reagent (P0128M, Beyotime, China). The images of each section were acquired by an inverted fluorescence microscope (Olympus IX73, JP). Image J (National Institutes of Health, Bethesda, MD, USA) was used to quantify the mean value of the immunofluorescence and numbers of Iba1⁺ cells in each section. For each mouse brain, six consecutive slices of the hippocampus were used for positive staining and cell counting and averaged to reflect the level of fluorescence intensity and Iba1⁺ cells. The positively stained area for the Iba1 marker was presented as a ratio of average fluorescence intensity. The numbers of Iba1- and DAPI-co-staining positive cells in the CA1 region were counted. The quantitative analyses were performed by blind method.

Hippocampus was dissolved in RIPA lysis buffer (P0013B, Beyotime, China) and protein concentration was measured by a BCA protein quantification kit (P0010, Beyotime, China). The sample was electrophoresed on SDS-PAGE gels to separate protein and transferred to PVDF membranes. The membranes were blocked with 5% skim milk (A600669, Sangon Biotech, China) in Tris-buffered saline with Tween (TBST) for 2 h at RT and then incubated with primary antibodies overnight at 4^°C. After the membranes were washed three times in TBST, they were incubated with secondary antibodies for 2 h at RT. The protein bands were detected by ECL reagent (WBKLS0100, Merck Millipore, USA), exposed to a chemiluminescence imager (SmartChemi™ 910, China), and quantitated with Image J. Each sample was subjected to Western blotting analysis using the following primary antibodies, as reported in Table 2. Horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (1:1,000, A0208, Beyotime, China) was used as the secondary antibody.

Total RNA was extracted by an RNA Quick Purification kit (RN001, ES Science, China). The concentrations of the RNA samples were determined by NanoDrop ND-2000 (Thermo, USA) instrument. Reverse transcription was finished by using All-in-One First-Strand Synthesis MasterMix (with dsDNase) (F0202, LABLEAD, China). Taq SYBR^® Green qPCR Premix (R0202, LABLEAD, China) and Roche Light Cycler 480 II Real-Time PCR System (Roche, USA) were used for qPCR quantification. GAPDH was used as the internal reference for normalizing target gene expression. Data were obtained using the 2^–ΔΔCt method. The sequences of primers are presented in Table 3.

Malondialdehyde (MDA) was measured by a Lipid Peroxidation MDA Assay Kit (S0131S, Beyotime, China) according to the manufacturer’s instructions. A microplate reader (TECAN Spark10M, China) was used to determine MDA activity by measuring the absorbance at 532 nm (U/mg of protein).

All data were presented as the mean ± SEM. Statistical analyses were performed using GraphPad Prism version 7.0 (GraphPad Software, Inc.). The inter-group comparisons were analyzed by one-way ANOVA with a Tukey post hoc test for multiple comparisons. The data of NOR were analyzed by the Kruskal–Wallis test and post hoc comparisons were conducted by Dunn’s test. A statistically significant difference was defined as p < 0.05.

At 7 days after surgery, we assessed the locomotor activity and exploratory behavior among four groups by OFT. Mice in all groups had no significant difference in total distance and time spent in the center of OFT (Figures 2A–C). The results suggested that locomotor activity and postoperative anxiety behavior were not affected. Based on the nature of exploring new things in rodents, the spatial memory ability of mice was tested by the NOR experiment. Mice were tested for instant and long-term memory, respectively, 30 s and 24 h after the training sessions. The CON group was not significantly different from the CON + TNIL group in DI either 30 s or 24 h after training sessions. Then, we found that the DIs were decreased in the PND group compared with the CON group, and TNIL improved the DIs at 30 s and 24 h (Figures 2D–H). In other words, these results indicated that anesthesia/surgery induced instant and long-term memory dysfunctions that were attenuated by TNIL in aged mice. In addition, these data showed that TNIL had no significant effect on the behavior of mice in the CON + TNIL group. Therefore, we conducted subsequent experiments in the CON, PND, and PND + TNIL groups to investigate the possible mechanism and signaling pathway.

The IL-1β and TNF-α were employed to assess the levels of neuroinflammation. Compared with the CON group, the expressions of IL-1β and TNF-α were significantly upregulated in the PND group. Furthermore, the expressions were downregulated in the PND + TNIL group compared with the PND group (Figures 3A–C). Given the proinflammatory factors are mainly produced by activated microglia, we studied the activated microglial marker Iba1 by immunofluorescence staining. The results showed the Iba1 expression and the number of Iba1⁺ cell increased in the PND group compared with the CON group, and TNIL attenuated the increase (Figures 3D–F). These results suggested that the neuroinflammation and microglia activation of the hippocampus of aged mice with PND were attenuated by TNIL.

To research the oxidative stress response of aged mice with PND, we first detected the levels of antioxidant enzyme SOD2 and lipid peroxidation product MDA. The results showed a decreased protein expression of SOD2 but an increased content of MDA existed in the PND group, which was relieved in the PND + TNIL group (Figures 4A–C). Then, we tested the expression of antioxidant- and oxidant-related genes in the hippocampus among the three groups. It was found that the mRNA levels of these antioxidant genes (GPX1, TXN2, PRDX3, FOXO1, and HO-1) were decreased in the PND group, and TNIL reversed these changes (Figure 4D). Conversely, compared with the CON group, the mRNA levels of oxidant genes (NOX2, BACH1, ACSL4, and MPO) were significantly increased in the PND group, while TNIL significantly attenuated these upregulated genes (Figure 4E). These results indicated that oxidative stress response may be closely involved in the pathogenesis of PND, and TNIL can significantly reduce oxidative stress imbalance.

The normal synaptic function is considered to be the biological basis of learning and memory (Shivarama Shetty and Sajikumar, 2017). BDNF is the crucial regulator of synaptic function, which promotes synaptogenesis and synaptic transmission (Waterhouse and Xu, 2009). Western blot results showed that the expressions of PSD95, SYP, and BDNF were reduced in the PND group compared with the CON group while increased in the PND + TNIL group (Figures 5A, B). These data indicated that TNIL may have a protective effect against hippocampal synapse impairment in aged mice with PND.

To reveal whether the balance of SIRT3/AMPK/Nrf2 was affected by TNIL, we carried out Western blotting to measure the expression levels of SIRT3, p-AMPK, AMPK, and Nrf2 proteins. The results showed that the expression of SIRT3, p-AMPK/AMPK, and Nrf2 was significantly decreased in the PND group compared with the CON group while increased in the PND + TNIL group (Figures 6A, B). These results suggested that the SIRT3/AMPK/Nrf2 pathway might participate in TNIL treatment of PND-aged mice.