Background
Postoperative cognitive dysfunction (POCD) is a relatively common and well-known complication in surgical patients. The occurrence of POCD is associated with increased mortality, risk of withdrawal from the labor market, and dependency on social transfer payments [
1,
2]. However, despite enormous research efforts in recent decades, the pathogenesis of POCD remains obscure. Mounting evidence has revealed that inflammation plays a key role in the disease process [
3‐
5]. Peripheral inflammation due to surgical trauma and the release of accompanying systemic inflammatory mediators have been shown to influence inflammatory processes of the central nervous system, triggering the activation of neurogliocytes and the concurrent endogenous production of pro-inflammatory cytokines [
6‐
9]. Volatile anesthetics, particularly isoflurane and sevoflurane, directly increased the production of pro-inflammatory cytokines in the brains of mice and impaired the acquisition of spatial memory in aged rats [
10‐
12]. In addition to surgical trauma and anesthetics, advanced age is one of the main risk factors for the development of POCD [
13]. Aging has been associated with an exacerbated inflammatory response in the normal aged brain when the immune system is irritated [
14,
15], which may increase susceptibility to POCD in the aged [
16‐
18]. However, the occurrence of POCD exhibits pronounced individual differences in the elderly, and POCD is also observed in younger and adult patients [
2].
Accumulating evidence indicates that early life adversity is associated with an increased risk for various long-term mental and physical diseases, including depression, anxiety, autoimmune disorders, diabetes, hypertension, cardiovascular diseases, cancers, and premature mortality [
19‐
25]. The consequences of early life adversity can last a lifetime, even when individuals’ life situations improve after experiencing early negative life events [
26,
27]. A hypothesis concerning the mechanism of susceptibility to chronic diseases is that early life adversity induces glucocorticoid insensitivity and exaggerates inflammatory responses to injury or infection [
28]. Several studies have shown that early life adversity results in an increased methylation of the glucocorticoid receptor (GR) gene promoter and reduced GR expression in rodents and humans [
29‐
31], which may lead to glucocorticoid feedback resistance and GR desensitization [
28,
32]. It is generally accepted that GR signaling regulates the inflammatory response to noxious stimuli, and evidence has supported this hypothesis. Adult individuals who experienced early life adversity were shown to have elevated levels of C-reactive protein (CRP) and pro-inflammatory cytokines (interleukin (IL)-1, IL-6, and tumor necrosis factor (TNF)-α) compared with participants without early life stress [
33‐
36]. Furthermore, an animal study showed that early life stress enhanced the production of pro-inflammatory cytokines in response to viral infection [
37]. Thus, will early life adversity amplify the neuroinflammation induced by volatile anesthetics or surgeries and increase the risk of POCD?
The aim of this study was to examine whether early life adversity induces cognitive decline and exaggerates neuroinflammation after sevoflurane anesthesia in adult rats. Moreover, we examined the effects of early life adversity on the expression of GR and the methylation of GR gene promoters and whether behavioral changes and neuroinflammation after anesthesia were reversible if GR expression was increased by altering DNA methylation.
Methods
Maternal separation model
All experimental procedures involving animals were approved by the Animal Care and Use Committee of Xuzhou Medical University (Xuzhou, Jiangsu Province, China). Wistar rats were obtained from the Experimental Animal Center of the Xuzhou Medical University. Male and female rats were housed together in standard rat cages, each containing one male and two females. Each female rat was housed separately after pregnancy. Animals were maintained under standard laboratory conditions, with a 12-h light/dark cycle (lights on at 6:00 a.m.), a room temperature of 23 ± 1
°C, and food and water provided ad libitum. The pups remained undisturbed at birth (postnatal day 0, P0). Half of the male pups from each litter were randomly assigned to undergo maternal separation (MS). The other half of the male pups and all female pups were kept with their dam the entire time. The separated pups were removed from the home cage and placed in a warm standard laboratory cage for 6 h per day (7:00–10:00 and 13:00–16:00) from P1 to P21 [
38]. The pups were returned to the home cage and remained with their dam the rest of the time. Following weaning, all pups were group-housed at P22 (6–7 per cage) according to gender and whether they had been subjected to MS or not. We did not observe any differences in body weight between MS rats and normal rats (data not shown). Subsequent experiments were performed when the animals reached adult age (85–90 days old).
Rats anesthesia and treatment
Rats received 3% sevoflurane mixed with pure oxygen for 2 h in anesthetizing chambers. Control groups received pure oxygen for 2 h in identical chambers. The anesthesia with 3% sevoflurane mixed with pure oxygen for 2 h was clinically relevant and did not cause significant changes in blood pressure and blood gas compared with the control group (data not shown). Anesthetizing chambers were placed on a heating pad, and the rat’s body temperature was maintained at 37 ± 0.5 °C. The concentrations of sevoflurane and oxygen were measured continuously (Drager, Lubeck, Schleswig-Holstein, Germany). Anesthesia was terminated by discontinuing sevoflurane and by administering pure oxygen until the animal regained its righting reflex. To evaluate the role of the nuclear factor-kappa B (NF-κB) signaling pathway, pyrrolidine dithiocarbamate (PDTC; 100 mg/kg) [
39], an NF-κB inhibitor, was given to rats intraperitoneally 30 min before exposure to sevoflurane. For intervention studies, the histone deacetylase inhibitor trichostatin A (TSA) was used for the epigenetic regulation of the glucocorticoid receptor. Rats were implanted with a stainless steel guide cannula (22 gauge, RWD, China) directed toward the left lateral ventricle (1.6 mm lateral to the midline, 1.0 mm posterior to the bregma, and 4.0 mm below the surface of the dura). Other procedures were performed as described in previous studies [
29]. A volume of 2 μl of TSA (100 ng/ml in DMSO) was infused using a micro-syringe through the infusion cannula for at least 1 min. Rats received a single infusion daily for seven consecutive days before exposure to sevoflurane.
Morris water maze tests
The Morris water maze (MWM) was a cylindrical, black-painted pool (1.5 m in diameter, 0.6 m in height) filled with water (0.3 m deep, 24 ± 1 °C) and divided into four virtual quadrants, with one starting point in each quadrant. A black-painted platform (10 cm diameter, 1 cm below water surface) was placed in the determinate quadrant. The experimental room contained cues that remained unchanged throughout the study. The movements of the rats were recorded by a video-tracking/computer-digitizing system (Shanghai Jiliang Software Technology Co., Ltd., Shanghai, China). The P86 rats were tested in the MWM four times per day for 4 days (acquisition tests). Each rat was gently placed in the water at one of the four starting points (in a random order) along the water maze perimeter with its face toward the wall of the pool. Rats were given 120 s to find the platform and were then left on the platform for 30 s. If the rat did not find the hidden platform within 120 s, the researcher would gently guide it to the platform, and its escape latency to find the platform was then marked as 120 s. At the end of the MWM training (P90), the platform was removed and the animals were allowed to search in the pool for 120 s (probe test); the number of crossings over the target area and the search time in the target quadrant were recorded [
40]. Animals that performed MWM tests were not used in other tests.
Context fear conditioning tests
Fear conditioning experiments were performed in an acrylic conditioning chamber with a grid floor composed of 19 stainless steel bars. A ventilation fan supplied background noise (65 db), and overhead infrared lights were left on. An infrared camera was located on the walls to monitor the rats’ freezing behavior. P86 rats were given 5 min to acclimate to the chamber and were then presented with one tone (2.2 kHz and 96 dB for 30 s). During the last 2 s of the tone, the rats were given a single shock (2.0 mA, 2 s). After the shock, the rats stayed in the chamber for 30 s and were then placed back into their housing area. Memory retention was assessed at 48 h post-conditioning. P88 rats were returned to the original chamber and allowed to stay there for 3 min. The chamber was maintained in the same context as when the rats were conditioned but without a footshock or tone. Freezing time during this period was recorded as a measure of contextual fear memory [
41]. Animals that underwent context fear conditioning tests were not used in other tests.
Hippocampus tissue preparation
Immediately after undergoing anesthesia with 10% chloral hydrate (0.3 ml/100 g, i.p.), all animals were sacrificed by decapitation. The hippocampus was rapidly removed, frozen in liquid nitrogen, and then stored at −80 °C until further processing. For enzyme-linked immunosorbent assay and Western blot analysis, the harvested hippocampus tissues were homogenized on ice using radioimmunoprecipitation assay (RIPA) lysis buffer supplemented with protease inhibitors (Beyotime Institute of Biotechnology, Haimen, Jiangsu, China). The lysates were collected and centrifuged at 12,000 rpm for 15 min to extract total proteins. The EpiQuik™ Nuclear Extraction Kit (Epigentek, Brooklyn, NY, USA) was used for the preparation of nucleoproteins from the hippocampal tissues according to the manufacturer’s instructions. The supernatant was quantified for total protein using the bicinchoninic acid (BCA) protein assay kit (Beyotime Institute of Biotechnology, Haimen, Jiangsu, China).
Western blot analysis
Western blotting was used to determine the expression of GR in the total protein extract and NF-κB p65 in the nuclear extract from the hippocampal tissues. Equal amounts of protein were loaded and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes (Millipore Corporation, Billerica, MA). The membranes were blocked in 5% nonfat milk for 2 h at room temperature and then incubated overnight at 4 °C with rabbit anti-GR (1:200, Santa Cruz Biotechnology, CA, USA), rabbit anti-NF-κB p65 (1:500, Abcam, Cambridge, UK), and mouse anti-β-actin (1:1000, Sigma-Aldrich). After the incubation, the membranes were washed three times with phosphate-buffered saline (PBS, pH 7.4) containing 0.3% Triton X-100 (PBS-T) and incubated with corresponding secondary antibodies conjugated with horseradish peroxidase (1:500, ZSGB-BIO, Beijing, China) for 2 h at room temperature. The protein signals were finally visualized using an enhanced BCIP/NBT Alkaline Phosphatase Color Development Kit (Beyotime Biotechnology, Inc., Haimen, Jiangsu, China).
Enzyme-linked immunosorbent assay
TNF-α, IL-1β, and IL-6 levels were quantified using rat-specific enzyme-linked immunosorbent assay (ELISA) kits (Xitang Bio-tech. Co., Ltd, Shanghai, China) according to the manufacturer’s instructions. The optical density of each well was determined using a microplate reader at 450 nm. Cytokine concentrations were calculated using standard curves generated using recombinant TNF-α, IL-1β, and IL-6.
Immunofluorescent staining
Rats were anesthetized with 10% chloral hydrate (0.3 ml/100 g, i.p.) and perfused transcardially with 200 ml of 0.9% saline followed by 300 ml of 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The brain tissues were removed and postfixed in 4% paraformaldehyde overnight and cryoprotected in 30% sucrose. Thirty-micron-thick frozen sections from the rat brains were cut using a freezing microtome and serially collected throughout the hippocampus. Free-floating tissue sections were rinsed three times with PBS-T. The tissue sections were incubated with 10% normal donkey serum in PBS-T for 2 h, followed by incubation with rabbit anti-p-NF-κB p65 (1:500, Cell Signaling Technology, Inc., Beverly, MA) and mouse anti-glial fibrillary acidic protein (GFAP) monoclonal antibody (1:600, Abcam, Cambridge, UK) at 4 °C for 24 h. After three 5-min rinses in PBS, the sections were incubated with donkey anti-rabbit IgG conjugated to Alexa Fluor® 488 and donkey anti-mouse IgG conjugated to Alexa Fluor® 594 (1:500, Life Technologies, Carlsbad, CA, USA) in the dark for 2 h at 37 °C. Fluorescence intensity was visualized under a confocal microscope (FV1000, Olympus Corp., Tokyo, Japan). The intensity on four slides (three to four sections per slide) was averaged for each animal and then normalized by that of the control group.
Quantitative reverse transcriptase polymerase chain reaction
Quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) was used to assess the mRNA levels of GR in rat hippocampal tissue. Total RNA was extracted using a TRIzol reagent kit (Invitrogen, USA) and transcribed into cDNA using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, USA). The PCR primers for GR were 5′-TAGGTGGGCGTCAAGTGATT-3′ (forward) and 5′-GATCAGGAGCAAAGCAGAGC-3′ (reverse). The primers for GAPDH were 5′-CAAGGTCATCCATGACAACTTTG-3′ (forward) and 5′-GTCCACCACCCTGTTGCTGTAG-3′ (reverse). Real-time PCR analysis was performed using the Roche LightCycler® 480 detection system using a SYBR® Select Master Mix Kit (Life Technologies, Carlsbad, CA, USA). The relative expression levels of GR were normalized to GAPDH.
DNA methylation assay
DNA was extracted from the hippocampal tissue using the genome extraction kit (Generay Biotechnology, Shanghai, China). Sodium bisulfite modification of DNA was performed using the EpiTect Bisulfite kit (Qiagen, Venlo, Netherlands) according to the manufacturer’s protocol. PCR amplification was carried out in a 50-μl reaction volume containing 25 μl of PCR master mix (Qiagen, Germany), 2 μl of each primer, 2.5 μl of bisulfite-modified DNA, and H2O. The thermocycler protocol was as follows: a 4-min denaturation at 95 °C; 40 cycles of 30-s denaturation at 95 °C, a 30-s annealing period at 55 °C, and a 40-s extension at 72 °C; and a final 5-min extension at 72 °C. The PCR product (285 bp) was used as a template for subsequent nested PCR reactions. The PCR product (177 bp) was purified and separated on a 2% agarose gel. The targeted DNA fragment was extracted from the gel. Following elution in Tris buffer, the PCR product was then subcloned into the pTG19-T vector (Generay Biotechnology) according to the manufacturer’s protocol. Ten plasmids containing the exon 17 GR promoter DNA fragment were screened and sequenced per animal.
Statistical analysis
All of the data are presented as the mean ± SD of independent experiments. For the escape latency data, repeated-measures two-way ANOVAs between groups were conducted. The post hoc Bonferroni test was used to compare the differences in escape latency between groups on each day of the MWM. One-way ANOVA followed by the Student–Newman–Keuls post hoc test or Student’s two-sample t test was performed for the other behavioral tests, Western blotting, ELISA, and methylation of all CpG sites within exon 17 of GR. Statistical analysis was performed using SPSS 16.0 (SPSS, Chicago IL) or GraphPad Prism 5.0 for Windows (GraphPad Software, Inc., San Diego, CA). P values <0.05 were considered to be statistically significant.
Discussion
Sevoflurane is one of the most commonly used inhalational anesthetics for general anesthesia. We first found that sevoflurane anesthesia in adult (P85) MS rats induced cognitive impairment. Moreover, we found that anesthesia with 3% sevoflurane for 2 h induced cognitive impairment only in MS rats and not in normal rats (Fig.
1), which suggests that sevoflurane anesthesia-induced cognitive impairment has different selectivity in different individuals. These findings provide a possible explanation for the clinical observation that the occurrence of POCD shows distinct individual differences and suggest that individuals who have experienced early life adversity may have a predisposition to cognitive impairment after anesthesia.
Accumulating evidence suggests a pivotal role for neuroinflammation in the POCD process. Pro-inflammatory cytokine release and astrocyte activation were associated with a decline in cognitive performance in humans and animals [
6‐
8]. Pro-inflammatory cytokines, such as TNF-α, IL-1β, and IL-6, can be released by activated astrocytes, triggering neuroinflammation and leading to cognitive dysfunction [
9]. Higher concentrations of pro-inflammatory cytokines inhibit long-term potentiation and impair memory [
47]. We found that 3% sevoflurane anesthesia for 2 h increased the levels of TNF-α, IL-1β, and IL-6 in the hippocampus after anesthesia in both normal and MS rats, but the degree of pro-inflammatory cytokines increase in MS rats was markedly higher than that in normal rats after anesthesia (Fig.
2). These data suggest that sevoflurane anesthesia in adult MS rats may cause cognitive impairment by inducing excessive neuroinflammation. Sevoflurane anesthesia also increased GFAP (the marker of astrocyte activation) immunofluorescence in the hippocampus after anesthesia in both normal and MS rats. The degree of increase in GFAP immunofluorescence in MS rats was higher than that in normal rats, similar to the change in pro-inflammatory cytokine levels (Fig.
3). These results further suggest that rats with different early life experiences may have different neuroinflammatory reactions to sevoflurane anesthesia.
Inhalation anesthetics have been shown to activate NF-κB signaling [
39,
44]. NF-κB signaling plays a pivotal role in immune and inflammatory responses. Inactive NF-κB, a dimer of p50 and p65, remains in the cytosol. In response to diverse internal and external inflammatory stimuli, NF-κB p65 is phosphorylated and rapidly translocates to the nucleus [
48,
49]. In the current study, we found that sevoflurane anesthesia activated hippocampal NF-κB signaling both in MS rats and normal rats, but the levels of nuclear NF-κB p65 protein and the fluorescence intensity of p-NF-κB p65 in the hippocampus in MS rats were significantly higher than those in normal rats (Figs.
2 and
3). These results suggest that NF-κB signaling in MS rats was overactivated under sevoflurane anesthesia. Next, we found that the pre-inhibition of NF-κB signaling by PDTC significantly suppressed excessive pro-inflammatory cytokine release after sevoflurane anesthesia in MS rats (Fig.
4). These results further suggest that the excessive activation of NF-κB signaling was involved in the increase of neuroinflammation after sevoflurane anesthesia in MS rats.
Given that glucocorticoid receptors (GRs) exert essential immunoregulatory and anti-inflammatory actions, the interaction between GRs and NF-κB signaling is of particular importance. In principle, GR signaling and NF-κB signaling are mutually antagonistic. Dysregulated GR signaling may enable exaggerated NF-κB signaling, ultimately leading to greater inflammatory responses. In the current study, a decreased expression of GR in MS rats was observed, and MS rats had a greater percentage of DNA methylation in the exon 1
7 GR promoter region (Fig.
5). These findings are in agreement with those of other studies that showed that early life adversity results in greater methylation of the GR gene promoter and reduced GR expression in rodents and humans, with consequent increases in pro-inflammatory gene expression [
29‐
34]. Therefore, we speculate that the MS phenotype was associated with the downregulation of GR and the increased production of pro-inflammatory cytokines controlled by the pro-inflammatory transcription factor NF-κB, which may worsen sevoflurane anesthesia-induced neuroinflammation and ultimately lead to cognitive impairment. A key question is whether the impact of early experiences on epigenetic programming is reversible in adults. It has been reported that DNA methylation produced by insufficient maternal care is reversible in the hippocampus of adult offspring through the modulation of the chromatin structure using the histone deacetylase inhibitor TSA [
50]. Histone deacetylase prevents histone acetylation and ensures that histones bind tightly to DNA. The activation of chromatin through histone deacetylase inhibition might trigger DNA demethylation by increasing the accessibility of methylated DNA to demethylase activity [
51]. We found that TSA decreased the percentage of DNA methylation in the exon 1
7 GR promoter region. Then, marked increases in the hippocampal GR expression were accompanied by the reversal of DNA methylation (Fig.
6).
Finally, we assessed the effects of the reversal of GR expression in adult MS rats on the activation of NF-κB signaling and neuroinflammatory responses to sevoflurane anesthesia. TSA pretreatment significantly suppressed overactivated NF-κB signaling and excessive neuroinflammation (Fig.
7). Moreover, TSA pretreatment ameliorated sevoflurane anesthesia-induced cognitive impairment in adult MS rats. These results suggested that enhanced GR expression may suppress the inflammatory response and prevent cognitive impairment induced by sevoflurane anesthesia. Taken together, our findings support the conclusion that early life adversity downregulates the expression of GR, enhances NF-κB signaling, and worsens sevoflurane anesthesia-induced neuroinflammation, ultimately leading to cognitive impairment.
There are several limitations and caveats that should be considered in this study. First, we did not systematically investigate the effects of MS on the function of the hypothalamic–pituitary–adrenal (HPA) axis and instead focused on the expression of GR in the hippocampus, which is closely linked to HPA axis function [
52]. It is worth noting that higher cortisol levels and ineffective dexamethasone treatments have been reported in POCD patients [
53,
54]. However, the current results suggested that the MS-induced downregulation of GR expression in the hippocampus may amplify neuroinflammation and cause cognitive impairment, which will allow us to further investigate the association between early life adversity, the HPA axis, and POCD. Second, MS can induce DNA hypermethylation in other genes (e.g., brain-derived neurotrophic factor) in addition to the GR gene, and the demethylation function of TSA is not specific to the GR gene [
55,
56]. More research is needed to determine whether other mechanisms of early life adversity are involved in cognitive impairment after anesthesia. However, given the inflammatory mechanisms of POCD, the epigenetic regulation of GR is most likely the chief reason for MS-induced cognitive impairment after anesthesia. Finally, volatile anesthetics seem to have a confounding effect on the brain. Some beneficial effects have been reported for isoflurane on the outcome of cerebral ischemia-reperfusion injury, traumatic brain injury, and lipopolysaccharide-induced cerebral damage [
57‐
59]. Perhaps volatile anesthetics have a neuroprotective effect on brain injury. However, volatile anesthetics themselves may have adverse effects on the brain. Recent studies show that volatile anesthetics can cause neuroinflammation and impairment in a normal physiological state [
10,
11].
Acknowledgements
We thank Nan Sun and Peng Wu (Jiangsu Province Key Laboratory of Anesthesiology, Xuzhou Medical University, Xuzhou, China) for the assistance with the experimental instruction and Ke Wang (Department of Statistics, Xuzhou Medical University, Xuzhou, China) for the assistance with the statistical analysis.