A total of 120 male C57BL/6 mice (age, 14 months; weight, 26–38 g) were acquired from the Guangdong Medical Laboratory Animal Centre (Guangdong, China). The mice were housed in a temperature-controlled (25±2°C) and humidity-controlled (55±5%) room wherein a 12-h light and dark cycle was maintained for ~2 weeks prior to surgery. All mice had free access to food and water. The experimental procedures were carried out in accordance with the National Institutes of Health ‘Guidelines for the Care and Use of Laboratory Animals’, and were approved by the Animal Ethics Committee of the Capital Medical University.

The 14-month-old mice (n=36) were randomly divided into three different groups: Control group (n=12; mice were not subjected to any treatment); surgical group (n=12; mice were subjected to open tibia fracture surgery 30 min after trace fear conditioning training); and the drug group (n=12; mice were subjected to open tibia fracture surgery with RAGE antagonist intervention). On the third postoperative day, cognitive function was assessed in all mice using fear conditioning tests.

The remaining 14-month-old mice (n=84) were randomly divided into seven experimental groups (n=12 per group). In one group, the mice were treated in the same way as the control group (C group); the other six groups of mice underwent open tibia fracture surgery with or without the RAGE antagonist intervention and were sacrificed on postoperative days 1, 3 and 7 (S1, S3, S7, D1, D3 and D7 group, respectively). Half of all mice (n=6/measurement point/group) were utilized for the western blot analysis and RT-qPCR experiments, while the other half were utilized for immunohistochemical analyses.

In the D1, D3 and D7 groups, the RAGE antagonist FPS-ZM1 (EMD Millipore, Billerica, MA, USA) was administered at a dose of 10 mg/kg via intraperitoneal injection (IP) 30 min prior to surgery, and 1.5 mg/kg once per day postoperatively prior to sacrifice (18). The remaining surgical animal groups (S1, S3 and S7) were given normal saline of a similar quantity at the same time as the RAGE antagonist groups in order to control for the effects of injection-related stress.

As previously described (19), the surgical mice were subjected to open tibia fracture surgery under general anesthesia (2.1% isoflurane in 30% FiO₂). The left hind limb of each mouse was shaved, and sterilized using povidone-iodine. Following a central skin cut, a 22 G pin was inserted into the intramedullary canal. Subsequently, the periosteum was exposed and osteotomy was conducted. Finally, the lesion was dampened and the skin was sewn together with 8–0 sutures. The mice were then allowed to recover consciousness following anesthesia spontaneously in a separate polypropylene cage. By use of warming pads, the temperature was checked and sustained at 36.5–37.5°C throughout the process.

Trace fear conditioning was utilized to evaluate the hippocampus-dependent memory of the mice, as described previously (20). The behavioral investigations were performed by use of a dedicated trace fear conditioning chamber. The chamber comprised a grid floor composed of stainless steel bars attached to a shock release system. During training, the mice were allowed to explore this environment for 3 min. Subsequently, three tone-foot shock pairings (tone: 2,000 Hz, 85 dB, 30 sec; foot shock: 0.6 mA, 2 sec) were delivered with an intertribal interval of 60 sec. Animals were returned to their home cage after an additional 30 sec. Three days after training, animals were returned to the same chamber for five min without tone or shock. Freezing behavior was described as a deficiency of any noticeable movement with the exception of respiration. The freezing behavior related to context was expressed as a percentage of the observation period, and was automatically documented using the software (Xeye Fcs; Beijing MacroAmbition S&T Development Co., Ltd., Beijing, China).

The animals were transcardially perfused with saline followed by 4% paraformaldehyde (PFA) under anesthesia. The brains were harvested, and postfixed in 4% PFA for 2 h, then dried in 30% sucrose under 4°C overnight, after which they were embedded in paraffin. The paraffin-embedded hippocampal tissues were cut into 5-µm sections, which were subsequently deparaffinized and rehydrated prior to antigen retrieval. After blocking with 3% H₂O₂ in phosphate-buffered saline for 10 min, the sections were incubated with a rabbit polyclonal antibody against ionized calcium-binding adaptor molecule 1 (IBA-1; 1:500; 019-19741; Wako Chemicals Inc., Richmond, VA, USA) or a rabbit anti-glial fibrillary acidic protein (GFAP) antibody (1:500; Dako, Copenhagen, Denmark) at 4°C overnight. Subsequent to washing, the sections were incubated with secondary antibody for 2 h at room temperature in the dark. Finally, DAB was added to visualize the microglia and astrocyte proportions of the hippocampus. Quantification was conducted through the use of ImageJ 1.47n software. Positive immunoreactivity levels were quantified as the proportion of positively stained divided by the total image area. Quantitative analyses were conducted in a blinded manner.

On day 1, 3 and 7 postoperatively, mice (n=6) were sacrificed via cervical dislocation under anesthesia. The total RNA was extracted and purified from homogenized hippocampal tissues using TRIzol (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the guidelines of the manufacturer. RNA was reverse transcribed into a cDNA template by use of a PrimeScript™ Reverse Transcription Reagent kit (Takara Bio Inc., Shiga, Japan) in a 20-µl reaction system. Following this, cDNA was amplified and analyzed by qPCR using SYBR-Green qPCR Master mix (Invitrogen Life Technologies) and an ABI 7500 Real-Time PCR system (Applied Biosystems, Foster City, CA, USA). The specific primers utilized were as follows: TNF-α forward, 5′-CATGATCCGAGATGTGGAACTGGC-3′ and reverse, 5′-CTGGCTCAGCCACTCCAGC-3′; IL-1β forward, 5′-CATGATCCGAGATGTGGAACTGGC-3′ and reverse, 5′-CTGGCTCAGCCACTCCAGC-3′; IL-6 forward, 5′-ACTCACCTCTTCAGAACGAATTG-3′ and reverse, 5′-CCATCTTTGGAAGGTTCAGGTTG-3′; and GAPDH forward, 5′-GGAGCGAGATCCCTCCAAAAT-3′ and reverse, 5′-GGCTGTTGTCATACTTCTCATGG-3′. The PCR conditions were set at 95°C for 3 min (initial denaturation), followed by 40 cycles of 95°C for 30 sec and 60°C for 5 sec (amplification). Relative mRNA levels were then normalized with respect to the GAPDH control values, and quantified with the 2^−ΔΔCt technique.

A total of 6 hippocampal tissue samples were homogenized on ice and lysed in cold radioimmunoprecipitation assay buffer with a protease-inhibitor cocktail. The homogenates were then centrifuged at 12,000 × g at a temperature of 4°C for ~30 min. Supernatant was then collected, and the proportion of protein in the supernatants was quantified using a BCA Protein Assay kit, according to the protocol of the manufacturer. After denaturing with SDS sample buffer, proteins (20 µg) were separated via SDS-polyacrylamide gel electrophoresis, and transferred to a polyvinylidene difluoride membrane. The membranes were blocked in 5% non-fat dried milk dissolved in Tris-buffered saline with Tween-20 (TBST) at room temperature for 1 h, then incubated with rabbit anti-p65 (1:1,000) and anti-GAPDH (1:2,000; both from Cell Signaling Technology, Inc., Danvers, MA, USA) primary antibodies overnight at 4°C. After thorough washing, the membranes underwent incubation with a secondary goat anti-rabbit antibody conjugated to horseradish peroxidase (1:5,000; Cell Signaling Technology, Inc.) for 1 h at room temperature. Thereafter, the membranes were treated with an enhanced chemiluminescence detection kit (Pierce; Thermo Scientific, Inc., Shanghai, China), and the concentration of each band was measured by densitometry using Bio-Rad Quantity One software. The relative expression levels were normalized to GAPDH.

All statistical analyses were carried out with SPSS (version 14; SPSS, Inc., Chicago, IL, USA) and Prism software. Data are presented as the mean ± standard deviation. The variations between the means were assessed by one-way ANOVA. Additionally, Bonferroni post hoc tests were used when the ANOVA test indicated significance. P<0.05 was considered to indicate a statistically significant difference.

In order to investigate whether inhibition of RAGE could improve POCD following surgery, the effect of FPS-ZM1 on learning and memory impairment was evaluated using the conditioned fear test. In this study, surgery considerably reduced the freezing time percentage compared with the control group (40.51±9.63 vs. 59.28±11.37%; n=12; P<0.001). Exposure to the RAGE inhibitor FPS-ZM1 attenuated the surgery-induced reduction in freezing behavior; these mice showed no significant difference in freezing time compared with the non-surgical control group (54.38±10.32 vs. 59.28±11.37%; n=12; P>0.05) (Fig. 1).

To examine the surgery-associated activation of inflammatory cells and the role of RAGE antagonism, IBA-1 and GFAP expression levels in the mouse hippocampi were investigated by immunohistochemical staining. Microglial activation was characterized by increased numbers of microglia and a trend towards increased microglial cell body-to-cell size ratio. Astrocyte activation was characterized by an increased astrocyte cell soma size and greater width and length of the stained processes. IBA-1 and GFAP immunoreactivity were enhanced in the mice of the surgical groups on the first and third postoperative days relative to the control mice (those which did not undergo open tibia fracture surgery). These effects were attenuated by treatment with the RAGE antagonist (Figs. 2 and 3). The findings suggest that the RAGE antagonist may inhibit surgery-induced activation of microglia activation and astrocytes.

Compared with the control group, NF-κB p65 protein levels were significantly increased on postoperative day 3 (P<0.01) and decreased to baseline by the seventh postoperative day in the surgical group. This increase was inhibited by FPS-ZM1 treatment (Fig. 4).

The TNF-α, IL-1β and IL-6 mRNA expression levels in hippocampal tissues of 14-month-old mice were determined using RT-qPCR on the first, third and seventh postoperative days, with or without RAGE antagonist administration. As illustrated in Fig. 5, the expression levels of TNF-α, IL-1β and IL-6 were considerably upregulated on the first postoperative day. Upregulation of IL-1β and IL-6, relative to the control group, remained until the third postoperative day. However, IP administration of the RAGE antagonist considerably attenuated the surgery-induced upregulation of these pro-inflammatory cytokines.

To investigate the results of surgery and RAGE antagonism on the expression of synaptic proteins, we measured the levels of postsynaptic density protein-95 (PSD-95, also known as DLG4). PSD-95 was increased markedly on the first postoperative day in the surgical group, reaching a peak, decreased slowly on the third day and then returned to baseline on the seventh day. The RAGE antagonist inhibited the elevation of PSD-95 on the first and third postoperative days, but increased the overexpression of PSD-95 on the seventh day after surgery (Fig. 6).