Back to Journals » Journal of Inflammation Research » Volume 14

Class IIa HDAC Downregulation Contributes to Surgery-Induced Cognitive Impairment Through HMGB1-Mediated Inflammatory Response in the Hippocampi of Aged Mice

Authors Huang C, Cai J, Jin S, Lin Q, Fang Q, Nan K, Han Y, Ge W, Liu Y, Tao Y, Cao H, Li J

Received 3 February 2021

Accepted for publication 18 March 2021

Published 31 May 2021 Volume 2021:14 Pages 2301—2315

DOI https://doi.org/10.2147/JIR.S304060

Checked for plagiarism Yes

Review by Single anonymous peer review

Peer reviewer comments 4

Editor who approved publication: Professor Ning Quan



Chen-miao Huang, 1, 2 Jia-jing Cai, 1, 2 Shao-wu Jin, 1, 2 Qi-cheng Lin, 1, 2 Qian-juan Fang, 1, 2 Ke Nan, 1, 2 Yuan Han, 1, 2 Wen-wei Ge, 1, 2 Yu Liu, 1, 2 Yuan-xiang Tao, 1, 3 Hong Cao, 1, 2 Jun Li 1, 2

1Department of Anesthesiology and Perioperative Medicine, The Second Affiliated Hospital and Yuying Children’s Hospital of Wenzhou Medical University, Wenzhou, Zhejiang, 325027, People’s Republic of China; 2Zhejiang Province Key Laboratory of Anesthesiology, The Second Affiliated Hospital and Yuying Children’s Hospital of Wenzhou Medical University, Wenzhou, Zhejiang, 325027, People’s Republic of China; 3Department of Anesthesiology, New Jersey Medical School, Rutgers, The State University of New Jersey, Newark, NJ, 07103, USA

Correspondence: Jun Li; Hong Cao
Department of Anesthesiology and Perioperative Medicine, The Second Affiliated Hospital and Yuying Children’s Hospital of Wenzhou Medical University, Wenzhou, Zhejiang, 325027, People’s Republic of China
Tel/Fax +86 577-88002925 Email [email protected]; [email protected]

Objective: Perioperative neurocognitive disorders (PND) are a common complication in the elderly. Histone deacetylases (HDACs) are a class of enzymes that control the acetylation status of intracellular proteins. Thus, we explored whether HDACs trigger the release of high mobility group box 1 (HMGB1) through altering the acetylation status in the hippocampi of aged mice.
Materials and Methods: The effect of the Class IIa HDAC in PND was explored using an in vivo form of splenectomy. Sixteen-month-old healthy male C57BL/6J mice were randomly divided into five groups: control, anesthesia plus sham surgery, anesthesia plus splenectomy, LMK235 treatment, and PBS treatment. The hippocampi were harvested on either first, third, or seventh postoperative day. Cognitive function was assessed via a Morris water maze (MWM) test. Quantitative RT-PCR, Western blots and ELISAs were carried out to assess the targeted gene expression at transcriptional and translational levels.
Results: Splenectomy led to a significant deficiency in spatial memory acquisition, marked decreases in mRNA and protein levels of HDAC4 and HDAC5 in the hippocampus, and increases in the levels of total HMGB1 and acetylated HMGB1. In a similar fashion to splenectomy, treatment with the HDAC4/5 inhibitor LMK235 produced impaired spatial memory and an increase in the expression of HMGB1 and its acetylated counterpart in the hippocampus.
Conclusion: These results suggest that surgery leads to PND through class IIa HDAC downregulation-triggered HMGB1 release in hippocampus of aged mice. HDACs may be a potential therapeutic target for postoperative cognitive dysfunction.

Keywords: postoperative cognitive dysfunction, hippocampus, HDAC4, HDAC5, HMGB1, neuroinflammation

Corrigendum for this paper has been published

Introduction

Postoperative neurocognitive dysfunction (PND) is characterized by a deficit in memory and cognitive function following surgery or anesthesia.1 PND can manifest through the impairment of memory, attention, and information processing and can last for several days to several years.41 Risk factors for PND include advancing age, low level of education, duration of anesthesia, multiple surgeries, and postoperative infections. Additionally, age is also a vital risk factor for PND.42 The incidence of PND varies widely in previous reports, possibly due to differences in methodology. The International Study of PND (ISPND) reported that 25.8% of the patients older than 60 years suffered PND at 7 days after surgery and 9.9% still had PND at 3 months.2

Until recently, researchers have attempted to investigate the pathophysiology of PND, but the mechanisms underlying pathogenesis remain unclear. Neuroinflammation triggered by surgical trauma may be a key component in the pathogenesis of PND. Reports have demonstrated that peripheral surgery could impact processes in the central nervous system.3,4 An aseptic injury (such as sterile surgery, where the damage of tissue was caused by mechanical injury) or a non-aseptic injury (such as bacterial or viral invasion) can cause peripheral inflammation.5 Tumor necrosis factor alpha (TNF-α), Interleukin-1β (IL-1β) and other cytokines derived from peripheral inflammation have adverse effects on the integrity of the blood-brain barrier, allowing it to become permeable and permitting the entry of proinflammatory factors into the brain and cause neuroinflammation.1

High mobility group box 1 (HMGB1), a highly conserved protein found in eukaryotic nuclei, is described as a nuclear DNA-binding protein and a pathogenic inflammatory mediator associated with sepsis, stroke, and other disease syndromes.6 A previous study revealed that the level of HMGB1 was increased significantly after cytoreductive surgery and hyperthermic intraperitoneal chemotherapy (CRS-HIPEC) in adults, accompanied by the development of perioperative short-term cognitive dysfunction.7 Administration of neutralizing anti-HMGB1 monoclonal antibodies or glycyrrhizin significantly reduced the severity of neuroinflammation and improved postoperative cognitive decline.8,9 HMGB1 can shuttle between the nucleus and cytoplasm and be released into the extracellular space through a passive process (where HMGB1 is released from damaged cells following sterile insult) or an active process (after acetylation of its lysine residues).10,11 These mechanisms can cause the release of a large amount of extracellular HMGB1, with the active process contributing to the pathogenesis of both infection and injury, and can activate the pattern-recognition receptors for Toll-like receptor 4 (TLR4) and the receptor for advanced glycation end products (RAGE).12,13

Histone deacetylases (HDACs) are a family of enzymes that act synergistically with histone acetyltransferases to remove acetyl groups from lysine residues. Human HDAC enzymes are divided into four major classes: Class I HDAC (including HDAC1, 2, 3 and 8); Class II HDAC (including HDAC4, 5, 6, 7, 9 and 10); Class III HDAC (alternatively named Sirtuins; including SIRT1 to SIRT7); and Class IV HDAC (the only member is HDAC11).14 Class III HDAC also have a nicotinamide adenine dinucleotide (NAD+) dependence. Accumulating evidence shows that pan-HDAC inhibitors exhibit complicated properties in animal models of various neurological diseases.15 Recent results demonstrated that HDAC inhibition promoted HMGB1 hyperacetylation and subsequently its extracellular release in liver ischemia/reperfusion (I/R) injury.16,17 However, the role of individual Class IIa HDACs (HDAC4 and HDAC5) in PND remains unclear. The aim of the present study was to examine whether surgery contributed to neuroinflammation-induced PND through Class IIa HDAC downregulation-triggered HMGB1 in the hippocampi of aged mice.

Materials and Methods

Animals

In view of the neuroprotective effect of estrogen, we chose male mice as the experimental subjects to avoid the interference of estrogen to the experiment.43 Sixteen-month C57-BL6 male mice (body weight 20–30 g, the total sample size is about 120) were purchased from the Animal Use Committee of Wenzhou Medical University and were grouped housed in a standard environment (temperature 24 ± 1 °C, a 12 h light-dark cycle, and 50% ± 10% humidity) with access to food and water ad libitum. Animals were acclimated to the environment for 7 days prior to the experiment. All experimental operations were approved by the Institutional Animal Care and Use Committee of Wenzhou Medical University (protocol number: SYXK2015-0009), which operates according to the recommendations for the care and use of experimental mice. The mice were randomly divided into five major groups (n = 24/group): 1) control, in which the mice do not undergo anesthesia and surgery; 2) anesthesia, in which mice were only anesthetized with 1% pentobarbital solution (7 mL/kg, i.p.); 3) anesthesia plus splenectomy; 4) treatment with LMK235 only and (5) treatment with phosphate-buffered saline (PBS, a vehicle solution). Mice were sacrificed on days 1, 3, and 7 after intervention. Each major group contains three subgroups (n=8/subgroup) according to the intervention time (1 day after intervention; 3 days after intervention; 7 days after intervention).

Surgical Procedures

For splenectomy, a midline incision was performed, and the two primary sources of blood flow (splenic artery and gastric artery) were tied off with sterile suture proximal to the spleen. The spleen was then isolated and removed. Lincomycin lidocaine gel was applied to the wound every 8 hours within 24 hours after surgery.

Drugs Application

Ten milligrams/kilogram of LMK235 (an HDAC4/HDAC5 inhibitor, dissolved in PBS) was administered intraperitoneally. PBS served as a control. The dosage and route of LMK235 administration were based on previous studies and slightly adjusted according the experimental condition, this method could increase histone acetylation in hippocampal neurons.44

Morris Water Maze (MWM) Test

The MWM test was used to evaluate spatial learning and memory through a computerized video track system (Xinruan Information Technology Co., Ltd., shanghai). Briefly, a nonvisible round platform (10-cm diameter) was placed 1 cm below the water surface located in the center of the target quadrant (SW quadrant) of a circular pool. The mice were placed on the platform for 30 seconds preceding the start of each trial and was then released into the water facing the tank wall in one of the four present stating locations. Mice were trained for five consecutive days with four trials per day (60 seconds maximum for a trial). The mice were given 60 seconds to locate the platform, otherwise they were gently guided to the platform and remained for 15 seconds. On day 6, animals underwent surgery, general anesthesia, or LMK235 treatment. On post-treatment days 1, 3, and 7, mice were subjected to a probe test where they were released into the water in the quadrant opposite to the target quadrant. The time to locate the platform (latency), the percentage of time spent in the target quadrant and the numbers of platform crossings were recorded.

Open Field Trial

On days 1, 3, and 7 following treatment, the locomotor activity was tested with the open field, prior to this, all mice had been subjected to a probe test in MWM test. Briefly, all mice were adapted to the environment for 1.5 hours prior to the trial. The mouse was placed directly into the center of the open field (dimensions of 50 × 50 cm) and a computerized video track system (Logitech Suzhou, China) recorded total movement distance of the mice in the box, the mice were monitored for 10 minutes. The total traveled distance and mean speed were measured as the parameter of locomotor activity.

Cytokine Analysis

On days 1, 3, and 7 following treatment, the blood of four mice in each subgroup was collected transcardially, and then the hippocampus of the four mice were isolated. We divided the hippocampus of a mice into two parts, which are used to extract protein and RNA, respectively. Other four mice in each subgroup were used for immunofluorescence. Blood samples were centrifuged at 2500 g for 25 minutes and the plasma was collected. After the plasma was stand for 2 hours, the serum is extracted to measure the content of TNF-α, IL-1β, and IL-6 with the use of ELISA kits (Haixi Tang Biotechnology Co., Ltd., Shanghai, China). Blood samples from mice without intervention served as controls, and samples were processed as described above.

Protein Extraction and Western Blot Analysis

Mice hippocampi were homogenized in cold RIPA buffer and centrifuged at 4 °C at 12,000 g for 30 minutes. A BCA protein assay kit (Thermos Fisher, USA) determined the concentration of protein in the supernatants. Protein samples of 35 µg were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene fluoride membranes. A TBS buffer of 10% non-fat dried milk blocked membranes for 2 hours and incubated with primary antibodies: rabbit anti-HMGB1 monoclonal (1:3000; ab-18256; Abcam, Inc., MA, USA), rabbit anti-RAGE polyclonal (1:1000; DF6618; Affinity, UK), rabbit anti-TNF-α polyclonal (1:1000; sc-12744; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), rabbit anti-IL-1β polyclonal (1:1000; sc-7884; Santa Cruz Biotechnology Inc.), rabbit anti-HDAC4 (1:2000; ab-79521; Abcam, Cambridge, UK), rabbit anti-HDAC5 polyclonal (1:2000; sc-133225; Santa Cruz Biotechnology Inc.), rabbit anti-TLR4 polyclonal (1:1000; sc-293072; Santa Cruz Biotechnology Inc.), and mouse anti-MyD88 monoclonal antibody (1:1000; sc-74532; Santa Cruz Biotechnology Inc.) overnight at 4 °C. The anti-β actin (1:3000; TA-09; Beijing Zhongshan Jinqiao Biotechnology) was used as a loading control. Membranes were incubated with goat anti-mouse (1:3000; Biosharp) or goat anti-rabbit (1:3000; Biosharp) secondary antibodies at room temperature for 2 hours. Immune complexes were assessed with peroxidase and an enhanced chemiluminescence system (ECL; Thermo Fisher Scientific); band intensity was quantified with densitometry.

Co-Immunoprecipitation

Co-immunoprecipitation was performed as previously described,45 Immunoprecipitation was performed to investigate the status of acetylated HMGB1 in hippocampus tissue. A total of 2.5 mg rabbit monoclonal anti-HMGB1 (ab18256; Abcam, Cambridge, UK) or rabbit IgG (Beyotime, China) was added to 500 mg of hippocampi lysates with gentle rotation for 1.5h at 4 °C. After the addition of 20 μL of Protein A/G Plus beads (Santa Cruz Biotechnology, SC-2003), the samples were incubated at 4 °C overnight with gentle rotation. Pellet beads were centrifuged at 2500 rpm for 5 minutes and the pellets were washed four times with PBS. Finally, the samples were resolved on SDS-PAGE and analyzed by Western Blot with anti-acetyl-lysine antibody (1:500; sc-32268; Santa Cruz Biotechnology Inc.).

Quantitative PCR (RT-qPCR)

The hippocampus tissue samples were homogenized with TRIzol reagent, the total RNA was extracted, and cDNA was synthesized with RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific, USA). HDAC1, 2, 4, 5, 11 were amplified with qPCR using SYBR‐Green qPCR Master mix (TOYOBO, Japan). The PCR amplification consisted of a 1 min denaturation step at 95 °C followed by 40 cycles of 15 s at 95 °C and 20 s at 60 °C on an iCycler real-time PCR detection system (Bio-Rad Laboratories, CA, USA). Primer sequences and amplification sizes are presented in Table 1. β-Actin was used as an internal reference to determine the relative expression levels of mRNA using the 2−ΔΔCt method. All samples were analyzed in triplicate.

Table 1 The Primers Used for qPCR Analysis

Brain Tissue Preparation and Double Immunofluorescent Staining

Mice were perfused with normal saline followed by a fixative reagent containing 4% paraformaldehyde in 0.1 M PBS; brain samples were harvested and protected in sucrose solution (30%) overnight and embedded in optimal cutting temperature compound (Solarbio, China) at −80°C. Then, 4-μm-thick coronal sections from of the brains were made in a cryostat, we sliced the entire brain but just focused on the hippocampus in this experiment. Sections were blocked in 5% BSA for 1.5 hours at room temperature and incubated in primary antibodies at 4 °C overnight. For double immunofluorescent staining, rabbit anti-NeuN (1:200; ab177487; Abcam) and mouse anti-NeuN (1:200; Millipore; MAB377) for neurons, rabbit anti-IBA1 (1:50; 10904-1-AP; Proteintech) and mouse anti-IBA1 (1:200; GT10312; GeneTex) for microglia, mouse anti-GFAP (1:200; SC-33637; Santa Cruz Biotechnology) and rabbit anti-GFAP (1:200; ab33922; Abcam) for astrocytes were used as specific cellular markers. The sections were then incubated with Alexa Fluor 488 (1:200) and Alexa Fluor 594 (1:200; Invitrogen) for 1.5 hours at 37 °C. The staining was examined with a fluorescence microscope (OLYMPUS, BX53 MicroPublisherTM 5.0 RTV, Japan).

Statistical Analysis

All data were presented as the means ± SEM. The Western blot band densities were quantified by Quantity one software (Bio-Rad, CA, USA). The results were analyzed by one-way ANOVA, after ANOVA showed a significant difference, pairwise comparisons between mean were tested by Bonferroni multiple comparison. A value of p < 0.05 was considered to manifest a statistically significant difference. All statistical tests were performed using SPSS software 19.0 (IBM, Armonk, NJ, USA). The results were presented using GraphPad Prism software 7.0 (GraphPad, San Diego, CA, USA).

Results

Splenectomy Induced Spatial Memory Impairment without Alteration of Locomotor Activity

To investigate the effects of splenectomy on cognitive function, we tested mice in the Morris Water Maze, which assesses spatial working memory. Comparisons by one-way ANOVA showed no significant difference either in latency or in swimming speeds among the groups on every training day before the surgery (Figure 1B and C, p > 0.05). The result indicated that all mice showed the same improvement in spatial learning and memory over time. The spatial probe test was conducted postoperatively. Compared with the control group, the latency for the first time to the platform increased in the splenectomy group (Figure 1D, 1d aft: F(4,35)=6.191, P < 0.05; 3d aft: F(4,35)=8,731, P < 0.05). The time spent in the target quadrant (the third quadrant) (Figure 1F, 1d aft: F(4,35)=7.154, P < 0.05; 3d aft: F(4,35)= 6.927, P < 0.05) and the number of platform crossings (Figure 1E, 1d aft: F(4,35)=9.161, P < 0.05; 3d aft: F(4,35)= 7.026, P < 0.05) decreased in the splenectomy group at days 1 and 3 after surgery. The results of the MWM revealed no differences between the control group and the anesthesia group in the aforementioned indices (Figure 1B–F, p > 0.05).

Figure 1 Effect of surgery and LMK235 treatment on spatial learning in the Morris Water Maze. (A) Experimental design. (B) Mean escape latency of the five major groups in the acquisition phase, n = 24 per group. (C) Average swimming speed of the major groups in the acquisition phase, n = 24 per group. (D) Latency for the first time get to the platform of each group at 1 d before surgery and 1, 3, and 7 d after surgery in the probe trials, n=8 per group. (E) The number of platform crossings with a 60 s limit of each group at 1 d before surgery and 1, 3, and 7 d after surgery in the probe trials, n = 8 per group. (F) Time spent in the target quadrant of each group at 1 d before surgery and 1, 3, and 7 d after surgery in the probe trials, n = 8 per group. (G) Average swimming speed of each group at 1 d before surgery and 1, 3, and 7 d after intervention in the probe trials, n=8 per group. All data are expressed as the means ± SEM. #p < 0.05 versus the day-matched control group. *p < 0.05 versus the day-matched anesthesia group.

In the open field trial, there was no difference in the total distance traveled by the groups in the arena (Figure 2, p > 0.05), indicating that the observed cognitive deficits in the MWM were not caused by reduced spontaneous locomotor activity. Comparisons by one-way ANOVA showed no significant difference in swimming speeds among the groups after the surgery on the corresponding date (Figure 1G, p > 0.05).

Figure 2 Motor dysfunctions was not evident in each group in open field. Total distance travelled of each group during monitoring in an open field box, n = 8 per group.

Early Inflammatory Response Following Surgery

Evidence for peripheral inflammation and neuroinflammation was seen at postoperative day 1 and day 3, as demonstrated by the increase in the levels of proinflammatory cytokines in the plasma and the hippocampal tissues of the splenectomy groups compared with the control and anesthesia groups (Figure 3A: 1d aft: F(4,15)=10.552, P < 0.05; 3d aft: F(4,15)= 16.900, P < 0.05; Figure 3B 1d aft: F(4,15)=8.679, P < 0.05; 3d aft: F(4,15)= 11.695, P < 0.05 Figure 3C, 1d aft: F(4,15)=23.226, P < 0.05; 3d aft: F(4,15)= 17.382, P < 0.05). Representative Western blots of TNF-α and IL-1β on postoperative days 1, 3, and 7, n = 4 per group. Each sample repeated the experiment 3 times. (Figure 3D–F) Analysis of Western blots showed that the expression levels of IL-1β and TNF-α were increased in the hippocampi of the splenectomy groups. Expression of TNF only increased significantly on postoperative day 1, while expression of IL-1β decreased on day 7 (Figure 3G 1d aft: F(4,15)=16.549, P < 0.05, Figure 3H 1d aft: F(4,15)=75.519, P < 0.05; 3d aft: F(4,15)= 61.744, P < 0.05).

Figure 3 Early peripheral-inflammatory response and neuroinflammation following surgery or LMK235 treatment. (A-C) The serum level of IL-1β, IL-6, and TNF-α on postoperative day 1, 3, and 7 were measured by ELISA, n = 4 per group. Each sample repeated the experiment 3 times. (D–F) Representative Western blots of TNF-α and IL-1β on postoperative days 1, 3, and 7, n = 4 per group. Each sample repeated the experiment 3 times. (G, H) Densitometric analysis of Western blot data from D–F, n = 4 per group. #p < 0.05 versus the day-matched control group. *p < 0.05 versus the day-matched anesthesia group.

Expression Patterns of HDACs in the Hippocampi of Mice That Developed Postoperative Cognitive Disorders

To determine the expression patterns of the HDAC family in brains with postoperative cognitive dysfunction, RT-qPCR and Western blot analyses were performed. Among Zn2+ dependent HDACs, the expression of Class IIa HDACs were significantly decreased at postoperative day 1 and day 3 in surgery group compared with the anesthesia group and naive control (Figure 4AE, C: F(6,21)=8.885, P < 0.05 Figure 4D: F(6,21)=12.109, P < 0.05). The levels of HDAC1, 2, and 11 had no significant difference at this experimental condition. The protein expression level of HDAC4 and HDAC5 were also significantly decreased at postoperative day 1 and day 3 in the hippocampus (Figure 4FJ, I: 1d aft: F(4,15)=10.582, P < 0.05; 3d aft: F(4,15)= 4.556, P < 0.05 Figure 4J: 1d aft: F(4,15)=30.298, P < 0.05; 3d aft: F(4,15)= 20.684, P < 0.05).

Figure 4 Expression patterns of HDACs in hippocampi from mice that developed postoperative cognitive disorders. (A–E) RT-qPCR analysis of mRNA levels of HDAC1, 2, 4, 5, and 11 from total RNA extracted from the hippocampi of each group, n = 4 per group. Each sample repeated the experiment 3 times. (F–H) Representative Western blots of HDAC4 and HDAC5 on postoperative days 1, 3, and 7, n = 4 per group. Each sample repeated the experiment 3 times. (I, J) Densitometric analysis of Western blot data from (FH), n=4 per group. #p < 0.05 versus the day-matched control group. *p < 0.05 versus the day-matched anesthesia group.

We performed dual immunofluorescent staining in the hippocampus from group C to identify the cell types in which HDAC4 and HDAC5 were expressed. Surprisingly, HDAC4 and HDAC5 were predominantly expressed in neurons and there was no colocalization of HDAC4 (Figure 5) or HDAC5 (Figure 5) in astrocytes or microglia. In conclusion, these results suggest that the neuron is a major source of HDAC4 and HDAC5 expression in the hippocampus.

Figure 5 Neurons are a major source of HDAC4 and HDAC5 expression in the hippocampus. Colocalization of HDAC4 with NeuN (part A), IBA (part B), and GFAP (part C). Colocalization of HDAC5 with NeuN (part D), IBA (part E), and GFAP (part F). The colocalization is in yellow in the image. The arrowhead indicates double-labelled cells (part A'; part D'), the arrow shows the protein of interest that is not colocalized with the cell (part B'; part C'; part E'; part F'). The magnified image of the regions is marked with boxes. Scale bar = 100 mm.

Splenectomy Induced Acetylation of HMGB1 in the Hippocampus

Anesthesia did not lead to the upregulation of HMGB1 expression at any time point compared with the control group (Figure 6AC). However, compared with the control group, the expression of HMGB1 increased significantly in the surgery group at postoperative days 1 and 3, reached a peak after day 1, and largely decreased by 7 days after surgery (Figure 6D, 1d aft: F(4,15)=12.789, P < 0.05; 3d aft: F(4,15)= 22.248, P < 0.05). We performed fluorescent immunostaining on the hippocampus of group S on the first day after surgery to assess the cellular location of HMGB1. As expected, the surgery induced robust release of HMGB1 (Figure 6F). Recently, research revealed that active secretion of HMGB1 from the cytoplasm into the extracellular space requires post-translational modification such as acetylation.18 In order to assess the acetylation status of released HMGB1, the immunoprecipitation was performed in group S on the first day after surgery. We found that the majority of released HMGB1 was acetylated (Figure 6E), which indicates that the HMGB1 liberation may be HDAC related.

Figure 6 Splenectomy-induced acetylation of HMGB1 expression in the hippocampus. (A–C) Representative Western blots of HMGB1 postoperative day 1, 3, and 7, n = 4 per group. (D) Densitometric analysis of Western blot data from (AC), n = 4 per group. Each sample repeated the experiment 3 times. (E) Co-immunoprecipitation analysis of hippocampus tissue for acetylated HMGB1 level. Anti-rabbit IgG was used as a negative control. (F) Representative images of double immunofluorescent staining with HMGB1 (green) and DAPI (blue); the arrow indicates the nuclear translocation and release of HMGB1. Scale bar=100 mm. #p < 0.05 versus the day-matched control group. *p < 0.05 versus the day-matched anesthesia group.

Predominant Neuronal Expression and Release of HMGB1 in the Hippocampus

To investigate the cell-type expression of HMGB1 in the hippocampus, we performed double immunofluorescent staining in group C. Interestingly, we found that HMGB1 colocalized mainly with neurons and rarely with Iba-1+microglia or GFAP (Figure 7). These results suggest that HMGB1 is highly expressed in hippocampal neurons. This data suggests that the predominant neuronal source of HMGB1 in this model and HMGB1 is the active performers of neuroinflammation in neurons.

Figure 7 Predominant neuronal expression and release of HMGB1 in the hippocampus. Colocalization of HMGB1 with NeuN, IBA, and GFAP. The co-localization is in yellow in the image. The arrowhead indicates double-labelled cells, the arrow shows the HMGB1, which is not colocalized with the cell. The magnified images of the regions are marked with boxes. Scale bar = 100 mm.

The HMGB1 Induces Proinflammatory Cytokine Expression Through TLR4/RAGE Activation

Previous research has found that the oral pretreatment of glycyrrhizin (a natural inhibitor of HMGB1) reduced postoperative neuroinflammation9 and suggested that HMGB1 may function to prime the neuroinflammatory response. However, we do not know if the HMGB1-mediated inflammatory effect in vivo is partly achieved by modulating TLR expression and activation. Therefore, we analyzed the expression of TLR4 in vivo following splenectomy. Our data showed that the expression levels of TLR4 in the hippocampi of the splenectomy groups were increased on postoperative day 1 (Figure 8A) and day 3 (Figure 8B) compared with the control and anesthesia groups (Figure 8D 1d aft: F(4,15)=61.986, P < 0.05; 3d aft: F(4,15)= 32.588, P < 0.05). We next analyzed the expression level of the TLR4 adaptor and signaling protein, MyD88, and found the same degree of increase (Figure 8AC, E 1d aft: F(4,15)=20.455, P < 0.05; 3d aft: F(4,15)= 7.665, P < 0.05). Splenectomy also resulted in increased RAGE levels in hippocampus postoperative day 1 (Figure 8A, B, F 1d aft: F(4,15)=10.157, P < 0.05). Our findings show that HMGB1 causes neuroinflammation in a manner that depends on TLR4 and RAGE.

Figure 8 TLR4 and RAGE play a critical role in HMGB1-mediated neuroimmune activation. (A–C) Representative Western blots of TLR4, MyD88, and RAGE on postoperative day 1, 3, and 7, n = 4 per group. Each sample repeated the experiment 3 times. (D–F). Densitometric analysis of Western blot data from (AC), n = 4 per group. #p < 0.05 versus the day-matched control group. *p < 0.05 versus the day-matched anesthesia group.

Exogenous Administration of LMK235 Triggers PND-Like Features

We evaluated the role of HDAC-HMGB1 signaling in PND by administering LMK235, which is a specific Class II HDAC inhibitor. We investigated whether LMK235 treatment caused spatial memory impairment. First, compared with the naive control, the latency for the first time to the platform increased in LMK235 treatment group (Figure 1D, P < 0.05). The time spent in the target quadrant (Figure 1F, P < 0.05) and the number of platform crossings (Figure 1E, P < 0.05) also decreased on the first and third days after administration. On the seventh day after administration, these parameters returned to normal (Figure 1DF, compared with group C, P < 0.05). Furthermore, LMK235 treatment caused both peripheral inflammation and neuroinflammation. The levels of TNF-α, IL-1β, and HMGB1 were increased in both the serum and hippocampus, and largely decreased at 7 days after administration (Figure 3AH). The administration also led to the upregulation of acetylated HMGB1 expression in the hippocampus (Figure 6E), which resulted in increased expression of TLR4, MyD88, and RAGE (Figure 8). Overall, the results suggest that LMK235 treatment mimics the surgery-induced prolonged inflammation response and cognitive impairment.

Discussion

HDACs are a family of enzymes involved in determining the post-translational acetylation status of non-histone proteins. The Class IIa HDACs (HDAC4, HDAC5) have been found to have a positive effect on the nervous system. Recent data suggest that mice generated with a conditional brain-specific HDAC4 deletion show alterations in motor coordination, accompanied by significant deficits in learning tasks.19 HDAC5 is equally necessary for memory consolidation in mice, and attenuation of HDAC5 does not improve cognition in a mouse model of AD.20 In our study, both the RNA and protein levels of HDAC4 and HDAC5 were significantly reduced in the hippocampi of mice suffering from postoperative cognitive disorders, while the mRNA levels of HDAC1, 2, and 11 had no significant difference at this experimental condition. This suggests that both HDAC4 and HDAC5 play a key role in the development of PND. A recent study found that anesthesia plus surgery increased HDAC2 expression in hippocampus,46 the discrepancy may due to the different species of mice, and on the other hand, the surgical procedures performed are also different. The splenectomy we performed is more traumatic for exploratory laparotomy, which may cause differences in experimental results. Previous studies have shown that the HDAC4/5 can be localized in nucleus and cytoplasm, the cellular localization of HDAC4/5 is mediated by 14-3-3 protein. Similar to HMGB1, HDAC4/5 contains NES to ensure accumulation in the cytoplasm, and contain NLS to direct the nuclear localization, phosphorylation of key residues on HDAC will cause it to bind to 14-3-3, and the combination of the two will mask the NLS site on HDAC4/5, leading to nuclear export. The separation from 14-3-3 will allow HDAC to re-enter the nucleus and perform deacetylation.31 In our study, the reduction in the total amount of HDAC also weakens the activity of the enzyme.

HMGB1 is now generally described as a damage-associated molecular pattern involved in a sterile inflammatory response. Recent studies affirmed the hypothesis that HMGB1 is linked to surgery-induced neuroinflammation.21 Protein shuttling between the nucleus and cytoplasm depends on nuclear localization sequences (NLS) and nuclear export sequences (NES). HMGB1 contains two NLS sites that direct the nuclear localization of the vast majority of HMGB1 in normal conditions.22 Post-translational modification such as lysine residue acetylation within NLS promotes HMGB1 release towards cytoplasm. Although previous studies demonstrated that abundant HMGB1 was discharged after surgery, our data further indicate that surgery-induced expression of HMGB1 was acetylated, and ultimately, demonstrates a HDAC-related mechanism.23

Previous studies demonstrated that the induction of HDAC activation resulted in suppression of aceytylation of HMGB1 in human intestinal epithelial HT-29 cells.24 Trichostatin A (TSA, a pan-HDAC inhibitor) treatment promoted nucleocytoplasmic mobilization of HDAC1/4, leading to increased total HMGB1 and acetylated HMGB1 release in a model of rat hippocampal-entorhinal cortex brain slices.25 Consistent with these findings, pharmacological inhibition of HDAC4 and HDAC5 regulates HMGB1 release from hippocampal neurons in our in vivo model. Renet al showed that post-insult treatment with the pan-HDAC inhibitor valproic acid improved behaviors in a middle cerebral artery occlusion stroke rat model.26 These mixed results indicate the possible impact of distinct histone modification in different diseased states and that HDAC subtypes may have specific roles.

Together with recent findings that surgery triggers neuroinflammation via HMGB1 release, we propose three critical steps here for the formation of prolonged neuroinflammation. In the first step, abdominal surgery triggers peripheral inflammation. We observed an increase in TNF-α, IL-1β, and IL-6 in the serum on postoperative day 1 and day 3. The content of these proinflammatory factors were similar to those of mice suffering from sepsis.27 The blood–brain barrier (BBB) is known as a highly specialized endothelial layer and was believed to separate the peripheral immune system from the CNS in normal conditions; however, inflammatory insult allows the BBB to become more permeable.1 Recent studies also demonstrated that the peripheral surgery interferes with the integrity of the BBB.28,29 Our data shows that proinflammatory factors in the serum were increased after surgery. It was previously shown that TNF-α and IL-1β are capable of crossing the BBB, resulting in neuroinflammation.30 In the second step, expression levels of HDAC4 and HDAC5 in neurons are remarkably decreased in the inflammatory environment. The total amount of Class IIa HDACs varies in several pathogeneses such as in ischemia/reperfusion injury.17 We found that both HDAC4 and HDAC5 expression levels were decreased in postoperative hippocampal neurons. These effects led to a highly acetylated state of the nuclei and increase the acetylation of HMGB1. In the final step, hyperacetylation of HMGB1 shifted its equilibrium from a predominantly nuclear location towards extracellular release. We provide evidence that the inhibition of HDAC4 and HDAC5 resulted in increased acetylated HMGB1, accompanied by the rise of TLR4 and RAGE levels. A prior study claims that TLR4 and RAGE are highly probable HMGB1 receptors.32 HMGB1 has three cysteines residues at positions 23, 45, and 106. Three different redox isoforms of HMGB1 have been detected in vivo: 1) fully reduced HMGB1 (fr-HMGB1), where thiol groups are present on each cysteine residue; 2) disulfide HMGB1 (ds-HMGB1), where the disulfide bridge is present between C23 and C45, while C106 remains reduced; and 3) sulphonyl HMGB1 (ox-HMGB1), where all cysteine residues are eventually oxidized.34 When a tissue becomes damaged, the fr-HMGB1 is released from the cell. The extracellular environment is more oxidizing than the cytosol and nucleus, which facilitates the formation of the C23-C45 disulfide bridge and conversion of fr-HMGB1 to ds-HMGB1.33 The ds-HMGB1 could bind the MD2-TLR4 complex and facilitates the activation of NF-κB as well as secretions of various inflammatory mediators.35 Additionally, damage-associated factors can fundamentally increase the expression of TLRs.36 A recent study found that the oral pretreatment of glycyrrhizin inhibits postoperative HMGB1 extracellular release and decreases TLR4 and noted that the treatment reduced postoperative neuroinflammation.9 Treatment using a TLR4 antagonist as well as TLR4 siRNA knockdown remarkably reduced the exogenous HMGB1-induced proinflammatory cytokine release in hippocampal slices.25 We further demonstrated that surgery induced the expression of RAGE. HMGB1 could be rapidly endocytosed by RAGE and transferred to lysosomes. HMGB1 damages the lysosomal membrane within the lysosomes and allows HMGB1 and its chaperones to enter the cytoplasm. The end result is inflammatory body activation, pyrophosphorylation, and subsequent release of proinflammatory mediators.37

Many studies have demonstrated that the hippocampal memory function can be damaged by surgery in aged animals compared with their counterpart controls.38 Surgery induces different degrees of neuroinflammation in adult and aged animals and the age is considered basis for PND. A chronic inflammatory status is the typical characteristic of the aging process and emerged as a crucial onset of several age-related diseases, such as Alzheimer’s disease.39 In the aging hippocampus, microglia become sensitized and primed. During peripheral insult, the inflammatory response in the hippocampi of aged mice is exaggerated and prolonged compared with the young mice.40 Our data suggest that the aged hippocampus is susceptible to a change in acetylation status, which leads to neuroinflammation.

Our study does have limitations. We did not perform a treatment group for HDAC agonists because the current technology fails to synthesize specific Class IIa HDAC agonists. Although the cognitive function was impaired by the LMK235 treatment, it remains unclear whether the HDAC agonists attenuate PND.

Taken together, decrease in Class IIa HDAC after splenectomy contributes to postoperative neuroinflammation and cognitive disorders, enhanced expression and activation of Class IIa HDAC may represent a promising target for PND. Our findings could guide the development of better treatment options for PND patients, eventually leading to the development of prevention and treatment strategies for PND.

Ethical Approval

All experimental operations were approved by the Institutional Animal Care and Use Committee of Wenzhou Medical University (protocol number: SYXK2015-0009) and complied with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Consent for Publication

Not applicable.

Acknowledgments

We would like to acknowledge the hard and dedicated work of all the staff that implemented the intervention and evaluation components of the study.

Funding

Zhejiang Provincial Natural Science Foundation of China, Grant/Award Number: LY21H090017; National Natural Science Foundation of China, Grant/Award Number: 81271204.

Disclosure

The authors declare that they have no competing interests.

References

1. Lyman M, Lloyd DG, Ji X, et al. Neuroinflammation: the role and consequences. Neurosci Res. 2014;79:1–12. doi:10.1016/j.neures.2013.10.004

2. Moller JT, Cluitmans P, Rasmussen LS, et al. Long-term postoperative cognitive dysfunction in the elderly ISPOCD1 study. ISPOCD Investigators International Study of post-operative cognitive dysfunction. Lancet (London, England). 1998;351(9106):857–861.

3. Xu J, Dong H, Qian Q, et al. Astrocyte-derived CCL2 participates in surgery-induced cognitive dysfunction and neuroinflammation via evoking microglia activation. Behav Brain Res. 2017;332:145–153. doi:10.1016/j.bbr.2017.05.066

4. Schreuder L, Eggen BJ, Biber K, et al. Pathophysiological and behavioral effects of systemic inflammation in aged and diseased rodents with relevance to delirium: a systematic review. Brain Behav Immun. 2017;62:362–381. doi:10.1016/j.bbi.2017.01.010

5. Alam A, Hana Z, Jin Z, et al. Surgery, neuroinflammation and cognitive impairment. EBioMedicine. 2018;37:547–556. doi:10.1016/j.ebiom.2018.10.021

6. Li M, Chen S, Shi X, et al. Cell permeable HMGB1-binding heptamer peptide ameliorates neurovascular complications associated with thrombolytic therapy in rats with transient ischemic stroke. J Neuroinflammation. 2018;15(1):237. doi:10.1186/s12974-018-1267-5

7. Yu H, Dong R, Lu Y, et al. Short-term postoperative cognitive dysfunction and inflammatory response in patients undergoing cytoreductive surgery and hyperthermic intraperitoneal chemotherapy: a pilot study. Mediators Inflamm. 2017;2017:3605350. doi:10.1155/2017/3605350

8. Terrando N, Yang T, Wang X, et al. Systemic HMGB1 neutralization prevents postoperative neurocognitive dysfunction in aged rats. Front Immunol. 2016;7:441. doi:10.3389/fimmu.2016.00441

9. Kong ZH, Chen X, Hua HP, et al. The oral pretreatment of glycyrrhizin prevents surgery-induced cognitive impairment in aged mice by reducing neuroinflammation and Alzheimer’s-related pathology via HMGB1 inhibition. J Mol Neurosci. 2017;63(3–4):385–395. doi:10.1007/s12031-017-0989-7

10. Lotze MT, Tracey KJ. High-mobility group box 1 protein (HMGB1): nuclear weapon in the immune arsenal. Nat Rev Immunol. 2005;5(4):331–342. doi:10.1038/nri1594

11. Lu B, Wang C, Wang M, et al. Molecular mechanism and therapeutic modulation of high mobility group box 1 release and action: an updated review. Expert Rev Clin Immunol. 2014;10(6):713–727. doi:10.1586/1744666X.2014.909730

12. Nan K, Han Y, Fang Q, et al. HMGB1 gene silencing inhibits neuroinflammation via down-regulation of NF-kappaB signaling in primary hippocampal neurons induced by Abeta25-35. Int Immunopharmacol. 2018;67:294–301. doi:10.1016/j.intimp.2018.12.027

13. Matsuura W, Harada S, Liu K, et al. Evidence of a role for spinal HMGB1 in ischemic stress-induced mechanical allodynia in mice. Brain Res. 2018;1687:1–10. doi:10.1016/j.brainres.2018.02.026

14. Wu Y, Hou F, Wang X, et al. Aberrant expression of histone Deacetylases 4 in cognitive disorders: molecular mechanisms and a potential target. Front Mol Neurosci. 2016;9:114. doi:10.3389/fnmol.2016.00114

15. Lin TB, Hsieh MC, Lai CY, et al. Modulation of nerve injury-induced HDAC4 cytoplasmic retention contributes to neuropathic pain in rats. Anesthesiology. 2015;123(1):199–212. doi:10.1097/ALN.0000000000000663

16. Evankovich J, Cho SW, Zhang R, et al. High mobility group box 1 release from hepatocytes during ischemia and reperfusion injury is mediated by decreased histone deacetylase activity. J Biol Chem. 2010;285(51):39888–39897.

17. He M, Zhang B, Wei X, et al. HDAC4/5-HMGB1 signalling mediated by NADPH oxidase activity contributes to cerebral ischaemia/reperfusion injury. J Cell Mol Med. 2013;17(4):531–542. doi:10.1111/jcmm.12040

18. Wei S, Gao Y, Dai X, et al. SIRT1-mediated HMGB1 deacetylation suppresses sepsis-associated acute kidney injury. Am J Physiol Renal Physiol. 2019;316(1):F20–f31. doi:10.1152/ajprenal.00119.2018

19. Kim MS, Akhtar MW, Adachi M, et al. An essential role for histone deacetylase 4 in synaptic plasticity and memory formation. J Neurosci. 2012;32(32):10879–10886. doi:10.1523/JNEUROSCI.2089-12.2012

20. Agis-Balboa RC, Pavelka Z, Kerimoglu C, et al. Loss of HDAC5 impairs memory function: implications for Alzheimer’s disease. J Alzheimer’s Dis. 2013;33(1):35–44. doi:10.3233/JAD-2012-121009

21. Bonaldi T, Langst G, Strohner R, et al. The DNA chaperone HMGB1 facilitates ACF/CHRAC-dependent nucleosome sliding. EMBO J. 2002;21(24):6865–6873. doi:10.1093/emboj/cdf692

22. Bianchi ME, Crippa MP, Manfredi AA, et al. High-mobility group box 1 protein orchestrates responses to tissue damage via inflammation, innate and adaptive immunity, and tissue repair. Immunol Rev. 2017;280(1):74–82. doi:10.1111/imr.12601

23. He HJ, Wang Y, Le Y, et al. Surgery upregulates high mobility group box-1 and disrupts the blood-brain barrier causing cognitive dysfunction in aged rats. CNS Neurosci Ther. 2012;18(12):994–1002. doi:10.1111/cns.12018

24. Chi JH, Seo GS, Cheon JH, et al. Isoliquiritigenin inhibits TNF-alpha-induced release of high-mobility group box 1 through activation of HDAC in human intestinal epithelial HT-29 cells. Eur J Pharmacol. 2017;796:101–109. doi:10.1016/j.ejphar.2016.12.026

25. Zou JY, Crews FT. Release of neuronal HMGB1 by ethanol through decreased HDAC activity activates brain neuroimmune signaling. PLoS One. 2014;9(2):e87915. doi:10.1371/journal.pone.0087915

26. Ren M, Leng Y, Jeong M, et al. Valproic acid reduces brain damage induced by transient focal cerebral ischemia in rats: potential roles of histone deacetylase inhibition and heat shock protein induction. J Neurochem. 2004;89(6):1358–1367. doi:10.1111/j.1471-4159.2004.02406.x

27. Xiao D, Zhang D, Xiang D, et al. Effects of fentanyl, midazolam and their combination on immune function and mortality in mice with sepsis. Exp Ther Med. 2015;9(4):1494–1500. doi:10.3892/etm.2015.2227

28. Ni P, Dong H, Wang Y, et al. IL-17A contributes to perioperative neurocognitive disorders through blood-brain barrier disruption in aged mice. J Neuroinflammation. 2018;15(1):332. doi:10.1186/s12974-018-1374-3

29. Zhang S, Dong H, Zhang X, et al. Cerebral mast cells contribute to postoperative cognitive dysfunction by promoting blood brain barrier disruption. Behav Brain Res. 2016;298(Pt B):158–166. doi:10.1016/j.bbr.2015.11.003

30. Gutierrez EG, Banks WA, Kastin AJ. Murine tumor necrosis factor alpha is transported from blood to brain in the mouse. J Neuroimmunol. 1993;47(2):169–176. doi:10.1016/0165-5728(93)90027-V

31. McKinsey TA, Zhang CL, Olson EN. Identification of a signal-responsive nuclear export sequence in class II histone deacetylases. Mol Cell Biol. 2001;21(18):6312–6321. doi:10.1128/MCB.21.18.6312-6321.2001

32. Andersson U, Yang H, Extracellular HH. HMGB1 as a therapeutic target in inflammatory diseases. Expert Opin Ther Targets. 2018;22(3):263–277. doi:10.1080/14728222.2018.1439924

33. Aucott H, Lundberg J, Salo H, et al. Neuroinflammation in response to intracerebral injections of different HMGB1 redox isoforms. J Innate Immun. 2018;10(3):215–227. doi:10.1159/000487056

34. Tang Y, Zhao X, Antoine D, et al. Regulation of posttranslational modifications of HMGB1 during immune responses. Antioxid Redox Signal. 2016;24(12):620–634. doi:10.1089/ars.2015.6409

35. Zhang Y, Karki R, Igwe OJ. Toll-like receptor 4 signaling: a common pathway for interactions between prooxidants and extracellular disulfide high mobility group box 1 (HMGB1) protein-coupled activation. Biochem Pharmacol. 2015;98(1):132–143. doi:10.1016/j.bcp.2015.08.109

36. Rahimifard M, Maqbool F, Moeini-Nodeh S, et al. Targeting the TLR4 signaling pathway by polyphenols: a novel therapeutic strategy for neuroinflammation. Ageing Res Rev. 2017;36:11–19. doi:10.1016/j.arr.2017.02.004

37. Andersson U, Yang H, Harris H High-mobility group box 1 protein (HMGB1) operates as an alarmin outside as well as inside cells. Semin Immunol. 2018;38:40–48.

38. Wuri G, Wang DX, Zhou Y, et al. Effects of surgical stress on long-term memory function in mice of different ages. Acta Anaesthesiol Scand. 2011;55(4):474–485. doi:10.1111/j.1399-6576.2011.02402.x

39. Cevenini E, Caruso C, Candore G, et al. Age-related inflammation: the contribution of different organs, tissues and systems. How to face it for therapeutic approaches. Current Pharm Design. 2010;16(6):609–618.

40. Barrientos RM, Kitt MM, Watkins LR, et al. Neuroinflammation in the normal aging hippocampus. Neuroscience. 2015;309:84–99. doi:10.1016/j.neuroscience.2015.03.007

41. Czyż-Szypenbejl K, Mędrzycka-Dąbrowska W, Kwiecień-Jaguś K, et al. The occurrence of Postoperative Cognitive Dysfunction (POCD) - systematic review. Psychiatr Pol. 2019;53(1):145–160. doi:10.12740/PP/90648

42. Needham MJ, Webb CE, Bryden DC. Postoperative cognitive dysfunction and dementia: what we need to know and do. Br J Anaesth. 2017;119(suppl_1):i115–i125. doi:10.1093/bja/aex354

43. Genazzani AR, Pluchino N, Luisi S, et al. Estrogen, cognition and female ageing. Hum Reprod Update. 2007;13(2):175–187. doi:10.1093/humupd/dml042

44. Trazzi S, Fuchs C, Viggiano R, et al. HDAC4: a key factor underlying brain developmental alterations in CDKL5 disorder. Hum Mol Genet. 2016;25(18):3887–3907. doi:10.1093/hmg/ddw231

45. Evankovich J, Cho SW, Zhang R, et al. High mobility group box 1 release from hepatocytes during ischemia and reperfusion injury is mediated by decreased histone deacetylase activity. J Biol Chem. 2010;285(51):39888–39897.

46. Wu Y, Dou J, Wan X, et al. Histone deacetylase inhibitor MS-275 alleviates postoperative cognitive dysfunction in rats by inhibiting hippocampal neuroinflammation. Neuroscience. 2019;417:70–80. doi:10.1016/j.neuroscience.2019.08.020

Creative Commons License © 2021 The Author(s). This work is published and licensed by Dove Medical Press Limited. The full terms of this license are available at https://www.dovepress.com/terms.php and incorporate the Creative Commons Attribution - Non Commercial (unported, v3.0) License. By accessing the work you hereby accept the Terms. Non-commercial uses of the work are permitted without any further permission from Dove Medical Press Limited, provided the work is properly attributed. For permission for commercial use of this work, please see paragraphs 4.2 and 5 of our Terms.