AMPK activation increases postoperative cognitive impairment in intermittent hypoxia rats via direct activating PAK2

Journal Pre-proof AMPK activation increases postoperative cognitive impairment in intermittent hypoxia rats via direct activating PAK2 Xi Mei, Guolin Tan, Wenxiang Qing

PII:

S0166-4328(19)31248-3

DOI:

https://doi.org/10.1016/j.bbr.2019.112344

Reference:

BBR 112344

To appear in:

Behavioural Brain Research

Received Date:

13 August 2019

Revised Date:

18 October 2019

Accepted Date:

4 November 2019

Please cite this article as: Mei X, Tan G, Qing W, AMPK activation increases postoperative cognitive impairment in intermittent hypoxia rats via direct activating PAK2, Behavioural Brain Research (2019), doi: https://doi.org/10.1016/j.bbr.2019.112344

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.

AMPK activation increases postoperative cognitive impairment in intermittent hypoxia rats via direct activating PAK2

＊

Xi Mei1, Guolin Tan2 , Wenxiang Qing1

Changsha, China,

ro of

1. Department of Anesthesiology, The Third Xiangya Hospital, Central South University,

2. Department of Otolaryngology, Head and Neck Surgery, The Third Xiangya Hospital, Central

-p

South University, Changsha, China, [email protected]

Abstract

ur

na

lP

re

The pathogenesis of postoperative cognitive impairment of obstructive sleep apnea-hypopnea syndrome (OSAHS) individuals remains unclear. AMP activated protein kinase (AMPK) is a ubiquitous sensor/effector of cell stresses. Thus we detected the role and underlying mechanisms of AMPK in postoperative cognitive impairment of OSAHS individuals in intermittent hypoxia rats. Cognitive function was evaluated by novel object recognition test and Barnes maze during the first 4 days after laparotomy. We found that laparotomy induced postoperative cognitive impairment and AMPK activation in intermittent hypoxia rats, but not in adult rats. Inhibiting AMPK activation via Compound C during laparotomy improved postoperative cognitive impairment and alleviated surgery-induced upregulation of p-PAK2, AMPK-PAK2 complex, and neuroinflammation (marked by microglial activation and IL-1β level) in intermittent hypoxia rats. These data suggested that AMPK played an important role in postoperative cognitive impairment of OSAHS individuals via directly activating PAK2.

Jo

Key words: postoperative cognitive impairment, AMPK, Obstructive sleep apnea-hypopnea syndrome, PAK2, neuroinflammation.

Introduction

Obstructive sleep apnea-hypopnea syndrome (OSAHS) is a common chronic disorder, characterized by recurrent episodes of hypopnea or apnea with intermittent arousal and sleep disturbances [1]. Accumulating evidences have showed that OSAHS could disturb fundamental 1

biochemical processes and induce oxidative stress and low-grade systemic inflammation. Thus OSAHS increases individual’s vulnerability and is strongly associated with postoperative neurocognitive impairment[2-5]. For example, Flink et al found that the incidence of postoperative neurocognitive impairment in OSAHS patients was more than twice that in non-OSAHS patients [2].

Nadler et al also found 21 (18.4%) of in 114 surgical patients at risk for OSAHS experienced

postoperative neurocognitive impairment[4]. Similarly, in OSAHS model animals, sevoflurane also

ro of

increased cognitive decline[6-7]. Further studies have showed that neuroinflammation is involved in the pathogenesis of postoperative neurocognitive impairment[7]. However, due to the specificity of OSAHS patients, the molecular mechanisms under postoperative neurocognitive impairment of

-p

OSAHS patients remain elusive.

re

Adenosine monophosphate activated protein kinase (AMPK) is a ubiquitous serine/threonine protein kinase and a sensor/effector of cell stress. Metabolic stress [9], ischemia[10], pain，

lP

depression[11]，inflammation[12], exercise[13] and neurodegenerative changes[14-15] all could induce

na

AMPK activation. Inhibiting AMPK activation could alleviate the impairment of post-ischemic hippocampus[16], and the impairment of learning , memory and synaptic plasticity in Alzheimer’s

ur

disease[14-15]. The downstream targets of AMPK include Tau[14-15], eukaryotic elongation factor 2(eEF2)[17], SIRT1 [18] , mammalian target of rapamycin(mTOR)[19] , P21-activated kinase

Jo

2(PAK2)[20-22]. In our previous study, we also found that anesthesia and surgery induced AMPK activation in an age-dependent manner[23]. Inhibiting AMPK activation improved postoperative cognitive impairment in aged rats [23]. These showed an important role of AMPK in postoperative neurocognitive impairment of aged individuals. However， OSAHS individuals are usually adult, and have experienced chronic intermittent hypoxia for a long time. It is still unclear whether and 2

how AMPK plays a role in postoperative cognitive impairment of OSAHS patients. Intermittent hypoxia is commonly used to reproduce the main characters of OSAHS in animals [24].

Thus we detected the role and underlying mechanisms of AMPK in postoperative cognitive

impairment of OSAHS individuals in intermittent hypoxia rats. We found that postoperative activated AMPK impaired the postoperative cognition of intermittent hypoxia rats and PAK2 played an important role in the pathogenesis as a direct downstream target of AMPK.

ro of

Materials and methods Animals

The experiment was conducted in accordance with the guidelines for the use of laboratory

-p

animals. The protocol (LLSC (LA)2017-011) was approved by the ethics committee of Central

re

South University Third Xiangya Hospital. The experiment was divided into two parts.

lP

Part 1: To explore the effects of inhalation anesthesia and surgery on hippocampal AMPK, the male Sprague Dawley rats (3–4 months, 280–300 g) were randomly divided into control (C) group

na

(n = 13 for behavioral study, n = 6 for mechanism study), surgery (S)group (n = 14 for behavioral study, n = 6 for mechanism study)，intermittent hypoxia(IH) group(n = 15 for behavioral study, n

ur

= 6 for mechanism study) and intermittent hypoxia+ surgery (IHS) group(n = 17 for behavioral study, n = 8 for mechanism study). Rats in the group S and IHS underwent inhalation anesthesia

Jo

and surgery[25]. Rats in group C and IH didn’t receive anesthesia or laparotomy. Rats in group C and S didn’t receive intermittent hypoxia. Part 2: To detect the effects of inhibiting AMPK activation on postoperative cognitive impairment, rats were assigned randomly into three groups: (1) intermittent hypoxia (IH) group: After two weeks intermittent hypoxia, animals didn’t receive anesthesia or laparotomy (n = 12 for 3

behavioral study, n = 7 for mechanism study); (2) intermittent hypoxia+surgery+ DMSO (IHS+DMSO) group: After two weeks intermittent hypoxia, rats underwent anesthesia, hepatolobectomy and the intraoperative intraperitoneal injection of dimethylsulfoxide (DMSO) (n = 16 for behavioral study, n = 7 for mechanism study); (3) intermittent hypoxia+ surgery+ Compound C(IHS+ Compound C) group: After two weeks intermittent hypoxia, rats received anesthesia, hepatolobectomy and the intraoperative intraperitoneal injection of Compound C (1

ro of

mg/kg)[23] (n = 20 for behavioral study, n = 8 for mechanism study). All of rats were sacrificed at the 1st or 3rd days after surgery.

-p

Intermittent hypoxia

re

In the IH，IHS，IHS+DMSO and IHS+ Compound C groups, rats received 20 cycles per hour

lP

intermittent hypoxia for 2 weeks, as previous report [26]. During each cycle of intermittent hypoxia, the oxygen concentration decreased from 21% to 8%, maintained around 8% and then returned to

na

21%, respectively 60s.

Anesthesia and laparotomy

Rats inhaled 5% sevoflurane with 100% oxygen (3 L/min); 3% sevoflurane mixed with 100%

ur

oxygen was used to maintain the anesthesia. Partial hepatolobectomy was implemented according to reported procedures [25]. Rats received the pain control by postoperative subcutaneous injection

Jo

of 0.25% bupivacaine. Blood pressure, body temperature, respiratory rate, PET CO2, FiO2 and FiSev were monitored and remained within normal physiological range.

Drug administration

Compound C is a kind of AMPK inhibitor. According to the reported method and dose [23], after opening the rats’ abdomen cavity, Compound C (Calbiochem, Germany, 171260) dissolved with

4

DMSO was intraperitoneally injected(1mg/kg) [23].

Behavioural test Open field test Loco-motor activity was detected in this test. The rat was gently placed into the center of the action observing box(100cm×100cm×40cm). The total crossing number and the time in the central zone were recorded by the computerized video track system (Logitech, China).

ro of

The novel object recognition test

The hippocampus-mediated declarative memory ability of all rats was evaluated by the NORT protocol. Rats were adapted to empty box before surgery once a day, every 30 minutes for 2

-p

consecutive days. On 1st day before operation, rats were placed into the box containing 2 identical

re

objects, for 10 minutes each time. At 6h and 1st day after surgery, one of the objects was

lP

replaced by a new one. When animals touched or sniffed the objects, it was considered as an effective exploration. The exploration time was recorded by two blinded to treatment allocation

na

observers. The “recognition index” was figured out according to a formula: (exploring novel object time- exploring familiar object time) / (exploring familiar object time + exploring novel

ur

object time).

Barnes maze test

Jo

The hippocampus-mediated spatial memory ability was evaluated by this test. Firstly, animals

were gently placed in the safety cage for 1 minute, then moved to an opaque box at the central of platform with 20 holes, only one of which allowed animal to exit the platform into the safety cage, for 1 minute and then the opaque box was taken away. On the postoperative 1 to 4 days, animals were trained to locate the correct hole three trials/day (allowing for at least 15 minutes in between 5

trials). The number of incorrect hole and time finding the correct hole were recorded by the observers.

Western blot Proteins in the hippocampal tissue were isolated using 12% SDS-PAGE gel, then transferred to the polyvinylidene fluoride(PVDF) membranes, followed by incubation in primary antibodies (AMPK: 1:300, CST, 2532; p-AMPK(Thr172): 1:300, CST, 2535; PAK2: 1:500, Abcam，ab76293;

ro of

p-PAK2(ser20): 1:500, Sigma,SAB4504051; β-actin:1:5000, Immunoway,YM3028.) at 4°C

overnight and then incubated with second antibody (HRP-conjugated Goat Anti-Rabbit IgG,

1:2000, CWBIO,CW0103) at room temperature for 2h. The results were analyzed by ImageJ

-p

software.

re

Quatitative real –time analysis

lP

Total RNA was separated by Trizol reagent (Thermo Fisher Scientific), followed by the cDNA synthesis using the HiScript II Q RT SuperMix (Vazyme, China), next, quantitative PCR was

na

implemented with the AceQ qPCR SYBR Green Master Mix（Vazyme, China）according to manufacturer’s instructions. The sequences of the studied gene primers are listed in Table 1. Table 1: The primers used for the gene expression studies of RT-PCR

IL1β GluA2 β-actin

primer Reverse(5,-3, )

Amplification size(bp)

GGCTTCCTTGTGCAAGTGTC

TGTCGAGATGCTGCTGTGAG

202

GGAAAAGACCAGTGCCCTCA

ACATCACTCAAGGTCATCCCC

267

TTCGAGTACCCCTTCAGGCT

TGAAAGCGAACACGGCTGTA

247

CACCAGGGTGTGATGGTG

GTACATGGCTGGGGTGTTG

282

Jo

SYP

(5,-3, )

ur

primer Forward

Co-immunoprecipitation Proteins in the fresh hippocampal tissue were isolated using IP lysis buffer, followed by removal of non-specific binding using Protein A+G Agarose(Santa Cruz Biotechnology, USA). 6

The supernatant was incubated with the mouse Anti-AMPK antibodies (1:100, abcam, ab80039) overnight and then with the Protein A+G Agarose at 4℃for 2h. The immune coprecipitation of AMPK-PAK2 was detected by SDS-PAGE electrophoresis analytic method.

Immunohistochemistry After transcardial infusion with 0.9% saline and then 4% paraformaldehyd, the brain tissues were removed, fixed in 4% paraformaldehyde and sequentially dehydrated in 15% and 30% sucrose in

ro of

0.01 M phosphate buffered saline at 4 ℃. Next, brains were cut at 20 μm thickness in the coronal plane. All specimens received immunohitochemical staining by the Avidin biotin - peroxidase complex (ABC) method. Specimens were incubated with rabbit anti- ionized calcium binding

-p

adaptor molecule 1 (Iba1) -1 (1:1000 dilution, Wako, Japan) overnight at 4°C, and then washed,

re

sequentially exposed to the secondary antibody (1:200 dilution, Vector Laboratories, United States)

lP

for 2 h. The results were obtained through a microscope (Nikon, Japan) and analyzed. The percent of activated microglia in the CA1 were determined. Activated microglia were characterized as

na

cells with pleomorphic bi- or tri- polar cell bodies, or spindle/rod-shaped cell bodies, with branches which were shortened, twisted, or displayed no ramification.

ur

Statistical analysis

Data were presented as mean ± standard deviation (mean ± SD). The data of a single time point

Jo

with more than three groups were analyzed by one-way ANOVA. Repeated Measures ANOVA was used for the data involving multiple time points. Data analysis used SPSS 19.0 for Windows, Two-tailed p < 0.05 was considered as statistically significant.

Results Laparotomy induced postoperative cognitive impairment and AMPK 7

activation in intermittent hypoxia rats, but not in adult rats In adult rats, laparotomy didn’t induce significant changes in travel distance in the open field test 6h and 1d after surgery (p＞0.05) (Fig2A), in preference index in NORT 6h and 1d after surgery (p＞0.05) (Fig2B) and in the latency and errors in Barnes maze from the 1st day to 4th day after surgery (p＞0.05, respectively) (Fig2C and D). However, in intermittent hypoxia rats, laparotomy significantly decreased the preference index in NORT 6h and 1d after surgery (p<0.05

ro of

for 6h, p<0.0001 for 1d ) ( Fig2B), and increased the latency and errors in Barnes maze from the

1st day to 4th day after surgery (latency: F (3, 23) = 103.9， p<0.0001; error: F (3, 23) = 32.62， p<0.0001) (Fig2 C and D). These showed that laparotomy induced postoperative cognitive

-p

impairment in intermittent hypoxia rats, but not in adult rats.

re

Corresponding to the impairment of learning and memory in intermittent hypoxia rats,

lP

laparotomy also significantly increased p-AMPK level of hippocampus of intermittent hypoxia rats 6h after surgery (p＜0.05, respectively), but not the adult rats (p＞0.05, respectively) (Fig2 E

na

and F). These suggested that AMPK activation possibly played a role in postoperative cognitive

ur

impairment of intermittent hypoxia rats.

Inhibiting AMPK activation during laparotomy improved postoperative

Jo

cognitive impairment in intermittent hypoxia rats In the open field test of intermittent hypoxia rats, there was no significant difference of travel

distance among IH , IHS+DMSO, and IHS+ Compound C groups (p＞0.05) (Fig3A). In the novel objective test, the preference index of IHS+DMSO group was significant less than the IH and IHS+ Compound C groups 6h and 1d after surgery (6h: p=0.0048 vs IH, p<0.0001 vs IHS+ 8

Compound C group; 1d: p=0.0022 vs IH, p<0.0001 vs IHS+ compound C group) (Fig3B). Similarly, in Barnes maze, the latency and errors of IHS+DMSO group were significant more than the IH and IHS+ compound C groups from the 1st day to 4th day after surgery (latency: F (2, 21) = 105.8， p<0.0001; error: F (2, 21) = 43.94， p<0.0001) (Fig3 C and D).

In order to confirm the

role of AMPK inhibitor Compound C, we also detected the level of AMPK activation, and found that Compound C significantly limited the surgery –induced AMPK activation 1d after surgery

ro of

(p＜0.05, respectively) (Fig3 E and F). These showed that inhibiting AMPK activation during laparotomy improved postoperative cognitive impairment in intermittent hypoxia rats.

-p

Inhibiting AMPK activation during laparotomy alleviated surgery-induced

re

upregulation of p-PAK2 and AMPK-PAK2 complex in intermittent hypoxia rats PAK2 is closely associated with spinogenesis，synaptic plasticity and cognition，[21,27-28]. AMPK

lP

could directly regulate PAK2 activation[20-22].Thus we detected the influence of AMPK activation

na

on PAK2 activation. We found that p-PAK2 level in the hippocampus IHS+ DMSO group was significantly higher than the IH and IHS+ Compound C groups 1d after surgery (p＜0.05,

ur

respectively)(Fig 4B) , although there was no difference of PAK2 level among the three groups (p ＞0.05, respectively) (Fig 4A). Immunoprecipitation of AMPK also showed that AMPK-PAK2

Jo

complex of IHS+DMSO group was significantly higher than the IH and IHS+ Compound C groups 1d after surgery (p＜0.05, respectively)(Fig 4C). These showed that AMPK increased the PAK2 activation after surgery via direct interaction. 、30]

GluA2 and SYP are closely involved in synaptic plasticity and memory[29

. Previous studies

found that PAK influenced AMPAR subunits [31]. GluA2 mRNA level in hippocampus of IHS+ 9

DMSO group was significantly less than the IH and IHS+ Compound C groups 1d after surgery (p ＜0.05, respectively)(Fig 4D), but SYP did not (p＞0.05, respectively)(Fig 4E).

Inhibiting AMPK activation during laparotomy limited postoperative neuroinflammation in hippocampus of intermittent hypoxia rats Neuroinflammation plays an important role in postoperative cognitive impairment[32]. Activated

ro of

microglias, which have larger cell bodies and shortened or twisted branches, are commonly used

to identify neuroinflammation. Thus we detected microglia activation and the Interleukin-1β (IL1β) relative expression in the hippocampus at the 1st day after surgery. We found that microglia

-p

activation in the hippocampus IHS+ DMSO group was significantly more than the IH and IHS+

re

Compound C groups 1d after surgery (p＜0.05, respectively)(Fig 5A) while there was no obvious

lP

difference between the IH and IHS+Compound C groups (p =0.3820)（Fig 5A）. Through measuring inflammatory factor in the hippocampus tissues by RT-PCR, we found that the expression

na

of IL-1β genes of IHS+DMSO group was also significantly higher than the IH and IHS+ Compound C groups 1d after surgery (p＜0.05, respectively)(Fig 5B). No significant difference

ur

was found between the IH and IHS+Compound C groups (p =0.8633)（Fig 5B） These results demonstrated that the inhibiting AMPK activation via compound C significantly

Jo

limited neuroinflammation in the hippocampus induced by surgery in intermittent hypoxia rats.

Discussion In the present study, our aim was to investigate the role and underlying mechanisms of AMPK in postoperative cognitive impairment of OSAHS individuals in intermittent hypoxia rats. We found that laparotomy induced postoperative cognitive impairment and AMPK activation in 10

intermittent hypoxia rats, but not in adult rats. Inhibiting AMPK activation during laparotomy improved postoperative cognitive impairment and alleviated surgery-induced upregulation of p-PAK2, AMPK-PAK2 complex, and neuroinflammation in intermittent hypoxia rats. These data suggested that AMPK played an important role in postoperative cognitive impairment of OSAHS individuals via directly activating PAK2. Previous studies have showed that neuroinflammation, oxidative stress, synapse loss, and Tau’s abnormality all are involved in the pathogenesis of postoperative neurocognitive impairment The key molecule orchestrating these pathological changes remains unclear. In addition,

ro of

[25,33-34].

etiological factors are always different among different individuals with postoperative

neurocognitive impairment[35-36,6,31]. Reported etiological factors include inflammation, pain, sleep

-p

disturbance, hypoperfusion, and so on [35-37]. Thus it is very important to find a shared molecule

re

which orchestrates the pathological changes in the pathogenesis of postoperative neurocognitive impairment and plays key roles in different etiological factor-induced postoperative Interestingly, AMPK are involved in many kinds of pathological

lP

neurocognitive impairment.

conditions such as metabolic stress, ischemia, pain，depression，inflammation, exercise and

na

neurodegenerative diseases. We also found that AMPK activation contributed to postoperative

ur

cognitive impairment in an age-dependent manner[23]. In the study, we further found that laparotomy induced postoperative cognitive impairment and AMPK activation in intermittent

Jo

hypoxia rats, but not in adult rats. Inhibiting AMPK activation during laparotomy improved postoperative cognitive impairment, neuroinflammation, and synaptic changes of intermittent hypoxia rats. These data showed that consistent with our previous reports[23], AMPK was also a key molecule in postoperative cognitive impairment of OSAHS individuals. At the same time, these data suggest that AMPK is a possible molecule shared by different etiological factor-induced

11

postoperative neurocognitive impairment. PAK2 is a member of PAKs family and reported to play a key role in regulation of synaptic function, as well as in morphology of spine growth[21,27-28]. Xing J et al reported that inhibition of AMPK prevented melatonin-mediated PAK2 upregulation in hypoxia-reoxygenation –treated N2a cells[19]. In order to investigate the downstream molecule of AMPK during postoperative neurocognitive impairment, we detected the effect of AMPK activation on p-PAK2 and

ro of

AMPK-PAK2 complex. Consistent with the report of Xing J, inhibiting AMPK activation during laparotomy alleviated surgery-induced upregulation of p-PAK2 and AMPK-PAK2 complex in

intermittent hypoxia rats.Inhibiting AMPK activation also improved the postoperative cognitive

-p

impairment. These showed that PAK2 was a direct downstream target of AMPK in postoperative

re

cognitive impairment of intermittent hypoxia rats. Previous study showed that overexpression of PAKs decreased the stabilization of glutamate receptors and down-regulated mEPSC amplitudes

lP

in hippocampal neurons [38]. Thus we detected the expressions of GluA2 and synaptophysin of

na

intermittent hypoxia rats after surgery. Corresponding to the increase of p-PAK2, GluA2 mRNA, but not synaptophysin, decreased in the hippocampus of intermittent hypoxia rats 1 day after

ur

surgery. These suggested a close relationship between p-PAK2 and GluA2 in postoperative cognitive impairment of intermittent hypoxia rats. In addition, previous studies have showed that

Jo

ca2+/calmodulin-dependent protein kinase kinase β (CaMKKβ), liver kinase B (LKB) and transforming growth factor-β-activated kinase 1 (TAK1) all could directly modulate the expression and activation of AMPK[39-40]. In the pathogenesis of postoperative cognitive impairment of intermittent hypoxia rats, it is unclear that what molecules activate AMPK.

12

Reference 1. Isono S. Obstructive sleep apnea of obese adults: Pathophysiology and perioperative airway management. Anesthesiology. 2009; 110:908–21. 2. Flink BJ, Rivelli SK, Cox EA, et al. Obstructive sleep apnea and incidence of postoperative delirium after elective knee replacement in the nondemented elderly. Anesthesiology2012;116:788–96.

ro of

3. Roggenbach J, Klamann M, von Haken R, et al. Sleep-disordered breathing is a risk factor for delirium after cardiac surgery: a prospective cohort study. Crit Care. 2014 Sep 5;18(5):477. 4. Nadler JW, Evans JL, Fang E, et al. A randomised trial of peri-operative positive airway

-p

pressure for postoperative delirium in patients at risk for obstructive sleep apnoea after

re

regional anaesthesia with sedation or general anaesthesia for joint arthroplasty. Anaesthesia.

lP

201-7Jun;72(6):729-736

5. Aibek E. Mirrakhimov Carey L. Brewbaker Andrew D. Krystal Madan M. Kwatr

na

Obstructive sleep apnea and delirium: exploring possible mechanisms. 6. Dong P, Zhao J, Li N, et al. Sevoflurane exaggerates cognitive decline in a rat model of chronic

ur

intermittent hypoxia by aggravating microglia-mediated neuroinflammation via downregulation of PPAR-γ in the hippocampus. Behav Brain Res. 2018 Jul 16;347:325-331.

Jo

7.Zhang X, Li N, Lu L, Lin Q, Li L, Dong P, Yang B, Li D, Fei J. Pioglitazone prevents sevoflurane induced neuroinflammation and cognitive decline in a rat model of chronic intermittent hypoxia by upregulating hippocampal PPAR-γ. Mol Med Rep. 2019 May;19(5):3815-3822.

13

8. Li D, Luo L, Xu M, Wu J, Chen L, Li J, Liu Z, Lu G, Wang Y, Qiao L. AMPK activates FOXO3a and promotes neuronal apoptosis in the developing rat brain during the early phase after hypoxia-ischemia. Brain Res Bull. 2017 Jun;132:1-9. 9. Kim J, Yang G, Kim Y, Kim J, Ha J. AMPK activators: mechanisms of action and physiological activities. Exp Mol Med. 2016 Apr 1;48:e224. 10. Luo X, Zou H, Xu P, Wen L, Stanley JL, Jiang X, Han TL, Olson D, Peng C, Zhang C, Xiao

ro of

X, Tong C, Qi H, Baker PN. Blunted Cardiac AMPK Response is Associated with Susceptibility to Ischemia/Reperfusion in Male Offspring of Gestational Diabetic Rats. Cell Physiol Biochem. 2019;52(5):1103-1116.

-p

11.Odaira T, Nakagawasai O, Takahashi K, Nemoto W, Sakuma W, Lin JR, Tan-No K.

re

Mechanisms underpinning AMP-activated protein kinase-related effects on behavior and

lP

hippocampal neurogenesis in an animal model of depression. Neuropharmacology. 2019 May 15;150:121-133.

na

12. Campbell NK, Fitzgerald HK, Fletcher JM, Dunne A. Plant-Derived Polyphenols Modulate Human Dendritic Cell Metabolism and Immune Function via AMPK-Dependent Induction of

ur

Heme Oxygenase-1. Front Immunol. 2019 Mar 1;10:345. 13. Steinberg GR, Carling D. AMP-activated protein kinase: the current landscape for drug

Jo

development. Nat Rev Drug Discov. 2019 Mar 13. 14. Cai, Z., Yan, L.J., Li, K., Quazi, S.H., Zhao, B. Roles of AMP-activated protein kinase in Alz-heimer’s disease. Neuromol. Med. 2012.14, 1–14. 15. Ma, T., Chen, Y., Vingtdeux, V., Zhao, H., Viollet, B., Marambaud, P., Klann, E., Inhibition of AMP-activated protein kinase signaling alleviates impairments in hippocampal synaptic plasticity 14

induced by amyloid beta. J. Neurosci. 2014.34,12230–12238. 16. Li, J., Zeng, Z., Viollet, B., Ronnett, G.V., McCullough, L.D.Neuroprotective effects of adenosine monophosphate-activated protein kinase inhibition and gene deletion in stroke. Stroke 2007.38, 2992–2999. 17. Deng Z, Luo P, Lai W, Song T, Peng J, Wei HK. Myostatin inhibits eEF2K-eEF2 by

9;494(1-2):278-284.

ro of

regulating AMPK to suppress protein synthesis. Biochem Biophys Res Commun. 2017 Dec

18. Zhang Y, Sun C, Xiao G, Shan H, Tang L, Yi Y, Yu W, Gu Y. S-nitrosylation of the

Peroxiredoxin-2 promotes S-nitrosoglutathione-mediated lung cancer cells apoptosis via

-p

AMPK-SIRT1 pathway. Cell Death Dis. 2019 Apr 15;10(5):329.

re

19. Wang Y, Yang Z, Zheng G, Yu L, Yin Y, Mu N, Ma H. Metformin promotes autophagy in

Sci. 2019 May 15;225:64-71.

lP

ischemia/reperfusion myocardium via cytoplasmic AMPKα1 and nuclear AMPKα2 pathways. Life

na

20. Banko MR, Allen JJ, et al. Chemical genetic screen for AMPKα2 substrates uncovers a network of proteins involved in mitosis. Mol Cell. 2011 Dec 23; 44(6):878-92.

ur

21. Kong D, Dagon Y, et al. A Postsynaptic AMPK→p21-Activated Kinase Pathway Drives Fasting-Induced Synaptic Plasticity in AgRP Neurons. Neuron. 2016 Jul 6; 91(1):25-33.

Jo

22. Varshney P, Dey CS. Resveratrol regulates neuronal glucose uptake and insulin sensitivity via P21-activated kinase 2(PAK2). Biochem Biophys Res Commun. 2017 Apr 1; 485(2):372-378. 23. Cao M, Fang J, et al. Activation of AMP-activated protein kinase (AMPK) aggravated postoperative cognitive dysfunction and pathogenesis in aged rats. Brain Res. 2018 Apr 1; 1684:21-29. 15

24.Moreno-Indias I, Torres M, Montserrat JM, Sanchez-Alcoholado L, Cardona F, Tinahones FJ, Gozal D, Poroyko VA, Navajas D, Queipo-Ortuño MI, Farré R. Intermittent hypoxia alters gut microbiota diversity in a mouse model of sleep apnoea. Eur Respir J. 2015 Apr;45(4):1055-65. 25. Le, Y., Liu, S., Peng, M., Tan, C., Liao, Q., Duan, K., Ouyang, W., Tong, J., 2014. Aging differentially affects the loss of neuronal dendritic spine, neuroinflammation and memory impairment at rats after surgery. PLoS One 9, e106837.

ro of

26.Xue J, Zhou D, Poulsen O, et al. Intermittent Hypoxia and Hypercapnia Accelerate

Atherosclerosis, Partially via Trimethylamine-Oxide. Am J Respir Cell Mol Biol. 2017 Nov; 57(5): 581–588.

-p

27. Kreis P, Barnier JV. PAK signalling in neuronal physiology. Cell Signal. 2009; 21:384–393.

re

28.Wang Y, Zeng C, Li J, Zhou Z, Ju X, Xia S, Li Y, Liu A, Teng H, Zhang K, Shi L, Bi C, Xie

lP

W, He X, Jia Z, Jiang Y, Cai T, Wu J, Xia K, Sun ZS. PAK2 Haploinsufficiency Results in Synaptic Cytoskeleton Impairment and Autism-Related Behavior. Cell Rep. 2018 Aug

na

21;24(8):2029-2041.

29. Zhou Z, Liu A, Xia S, Leung C, Qi J, Meng Y, Xie W, Park P, Collingridge GL, Jia Z. The

ur

C-terminal tails of endogenous GluA1 and GluA2 differentially contribute to hippocampal synaptic plasticity and learning. Nat Neurosci. 2018 Jan;21(1):50-62.

Jo

30. Liu SJ, Yang C, Zhang Y, Su RY, Chen JL, Jiao MM, Chen HF, Zheng N, Luo S, Chen YB, Quan SJ, Wang Q. Neuroprotective effect of β-asarone against Alzheimer's disease: regulation of synaptic plasticityby increased expression of SYP and GluR1. Drug Des Devel Ther. 2016 Apr 15;10:1461-9. 31.Hussain NK, Thomas GM, Luo J, Huganir RL. Regulation of AMPA receptor 16

subunit GluA1 surface expression by PAK3 phosphorylation. Proc Natl Acad Sci U S A. 2015 Oct 27;112(43):E5883-90. 32. Feng X, Uchida Y, Koch L, Britton S, Hu J, Lutrin D, Maze M. Exercise Prevents Enhanced Postoperative Neuroinflammation and Cognitive Decline and Rectifies the Gut Microbiome in a Rat Model of Metabolic Syndrome. Front Immunol. 2017 Dec 11;8:1768. 33. Dong, Y., Xu, Z., Huang, L., Zhang, Y., Xie, Z., 2016. Peripheral surgical wounding may

ro of

induce cognitive impairment through interlukin-6-dependent mechanisms in aged mice. Med. Gas Res. 6, 180–186.

34. Degos, V., Vacas, S., Han, Z., van Rooijen, N., Gressens, P., Su, H., Young, W.L., Maze,M.,

-p

2013. Depletion of bone marrow-derived macrophages perturbs the innate immune response to

re

surgery and reduces postoperative memory dysfunction. Anesthesiology 118, 527–536

lP

35. Hu Y, Zhang M, Chen Y, Yang Y, Zhang JJ. Postoperative intermittent fasting prevents hippocampal oxidative stress and memory deficits in a rat model of chronic cerebral

na

hypoperfusion. Eur J Nutr. 2019 Feb;58(1):423-432. 36. He Y, Li Z, Zuo YX. Nerve Blockage Attenuates Postoperative Inflammation in Hippocampus

ur

of Young Rat Model with Surgical Trauma. Mediators Inflamm. 2015;2015:460125. 37. Ni P, Dong H, Zhou Q, Wang Y, Sun M, Qian Y, Sun J. Preoperative Sleep

Jo

Disturbance Exaggerates Surgery-Induced Neuroinflammation and Neuronal Damage in Aged Mice. Mediators Inflamm. 2019 Mar 18;2019:8301725. 38. Thévenot E, Moreau AW, Rousseau V, et al. p21-Activated kinase 3 (PAK3) protein regulates synaptic transmission through its interaction with the Nck2/Grb4 protein adaptor. J Biol Chem. 2011 Nov 18; 286(46):40044-59. 17

39. Vaarmann A, Mandel M, Zeb A, Wareski P1, Liiv J, Kuum M, Antsov E, Liiv M, Cagalinec M, Choubey V, Kaasik A. Mitochondrial biogenesis is required for axonal growth. Development. 2016 Jun 1;143(11):1981-92.

40. Kim J, Yang G, Kim Y, Kim J, Ha J.AMPK activators: mechanisms of action and

Jo

ur

na

lP

re

-p

ro of

physio-logical activities. Exp Mol Med. 2016 Apr 1;48:e224.

18

Fig.1. showing the description of the behavioral protocol. (A) open filed, the NOCT and Barnes

ro of

maze test of part 1. (B) open filed , the NOCT and Barnes maze test of part 2.

-p

Fig.2. Laparotomy induced postoperative cognitive impairment and AMPK activation in the

re

intermittent hypoxia rats, but not in adult rats. (A) showing the crossing number of the open field

lP

test, the data were analyzed by repeated measures ANOVA (F(6,84)=0.4983, P=0.8079). (B) showing the preference index of the novel object recognition at 6 hours and the 1st day after

na

surgery. The data were analyzed by Repeated Measures ANOVA (IHS vs. IH, p=0.0153, at 6 hours after surgery. IHS vs. IH, p=0.0006, at the 1st day after surgery. ). (C) showing the latency

ur

of Barnes maze test( the 1st to 4th day after surgery), the data were analyzed by Repeated Measures ANOVA(IHS vs. IH, p<0.0001). (D) showing the errors of Barnes maze test( the 1st to

Jo

4th day after surgery), the data were analyzed by Repeated Measures ANOVA(IHS vs. IH, p<0.0001). (E) showing Western blot analysis the level of AMPK at the 1st day and 3rd day after surgery, the data were analyzed by one-way ANOVA(p>0.05,respectively). (F) showing Western blot analysis the level of p-AMPK at the 1st day and 3rd day after surgery, the data were analyzed by one-way ANOVA(IHS vs. IH, p=0.0308, at the 1st day after surgery.). 19

ro of

Fig.3 Inhibiting AMPK activation during laparotomy improved postoperative cognitive impairment in intermittent hypoxia rats. (A) showing the crossing number of the open field test, the data were analyzed by repeated measures ANOVA(F (8, 63) = 1.449，p = 0.1941). (B)

-p

showing the preference index of the novel object recognition at 6 hours and 1st day after surgery. The data were analyzed by Repeated Measures ANOVA (IHS+DMSO vs. IH, p < 0.005;

re

IHS+DMSO vs. IHS+Compound C, p < 0.0001, at 6 hours and the 1st day after surgery.). (C)

lP

showing the errors of Barnes maze test (the 1st and 4th day after surgery), the data were analyzed by Repeated Measures ANOVA (IHS+DMSO vs. IH, p < 0.0001; IHS+DMSO vs.

na

IHS+Compound C, p < 0.0001). (D) showing the latency of Barnes maze test (the 1st to 4th day after surgery), the data were analyzed by Repeated Measures ANOVA (IHS+DMSO vs. IH, p <

ur

0.0001; IHS+DMSO vs. IHS+Compound C, p < 0.0001). (E) showing Western blot analysis the

Jo

level of AMPK at the 1st day and 3rd day after surgery. The datas of the level of AMPK were analyzed by one-way ANOVA (p>0.05, respectively). (F) showing Western blot analysis the level of p-AMPK at the 1st day and 3rd day after surgery, the datas of the level of p-AMPK were analyzed by one-way ANOVA (IHS+DMSO vs. IH, p =0.0131; IHS+DMSO vs. IHS+Compound C, p =0.0129, at the 1st day after surgery. P>0.05, respectively, at the 3rd day after surgery).

20

Fig.4 Inhibiting AMPK activation during laparotomy improved postoperative cognitive

ro of

impairment in intermittent hypoxia rats via directly inhibiting p-PAK2 and limiting GluA2

down--regulation. (A) showing Western blot analysis and corresponding quantification of PAK in hippocampus(p>0.05, respectively). (B) showing Western blot analysis and corresponding

-p

quantification of p-PAK2 in hippocampus（IHS+DMSO vs. IH, p =0.0034; IHS+DMSO vs.

re

IHS+Compound C, p =0.0027）.(C) showing COIP analysis and corresponding quantification of

lP

AMPK-PAK2 complex in hippocampus (IHS +DMSO vs. IH, p =0.0032; IHS+DMSO vs. IHS+Compound C, p =0.0025). (D) showing RT-PCR analysis the level of GluA2 mRNA in

na

hippocampus (IHS+DMSO vs. IH, p =0.0004; IHS+DMSO vs. IHS+Compound C, p =0.0009;IH vs. IHS+Compound C, p>0.05). (E) showing RT-PCR analysis the level of SYP mRNA in

Jo

ur

hippocampus (p>0.05,respectively) .

21

Jo

ur

na

lP

re

-p

ro of

Fig.5 Laparotomy induced neuroinflammation in hippocampus of intermittent hypoxia rats, limiting AMPK activation after surgery inhibited neuroinflammation. (A) Representative images of Iba-1 staining in the CA1. (one-way ANOVA, F (2, 6) = 35.35, p = 0.0005) ∗∗ p < 0.01. (B) showing RT-PCR analysis the level of IL1β mRNA at the 1st day after surgery, the data were analyzed by one-way ANOVA (IHS+DMSO vs. IH, p=0.0022; IHS+DMSO vs. IHS+Compound C, p =0.0032).

22