Electric vagal nerve stimulation inhibits inflammation and improves postoperation cognitive dysfunction in aged rats

Objective This study aimed to evaluate effects of vagal nerve stimulation on postoperation cognitive dysfunction in aged rats. Material and Methods: A total of 33 male SD rats were ssigned randomly to three groups, the control group, the s group and sv group. Behavior and memory were evaluated by Open Field Test and Morris Water Maze. The cytokines level were measured by ELISA. The TNF-α protein in hippocampal zone were assessed by Western blotting. qPCR was used to detected the mRNA expression of NF-κB in hippocampus. Results: During the anesthesia/operation, the vital life signs of rats were stable. In SV group, vagal nerve stimulation could decrease heart rate lower than 10% of basic level and kept it at a stable range by regulation of stimulation intensity. After stimulation stop, the heart rate would return to the basic level again. This indicated that the model of vagal nerve stimulation was successful. The serum inflammatory cytokines of TNF-α and IL-6 increased by the operation/anesthesia, but these cytokines could be decreased by the vagal nerve stimulation (all P<0.05). The level of TNF-α protein and mRNA expression of NF-κB in hippocampus were also eliminated by the vagal nerve stimulation compared to S group (P<0.05). The Morris Water Maze results showed the escape latency of postoperation in the S group was significantly longer than the C group (P<0.05), and the times of crossing platform in the S group was lower than that of the C group (P<0.05). Although the escape latency of postopration in the SV group was shorter than that of the S group, there was no significant difference between two groups. There were no significant differences in behavior test by Open Field test between three groups.

3 enough and should be studied and improved in the future.

Background
Postoperation cognitive dysfunction (POCD) is a severe complication of surgery and general anesthesia. 14% of patients have been found cognitive decline and confusion after surgery with anesthesia [1,2]. POCD is frequent in patients older than 60-year undergoing major non-cardiac surgery, which increases both morbidity and mortality [3]. These cognitive impairments are related to language comprehension, attention, social integration and short term memory. Also, POCD may extent the recovery process and hospital stay, and diminish the quality of patients' life [4]. There are many risks developing postoperative cognitive decline in elderly patients, such as increased age, longer time in surgery, longer stay in an intensive care unit and mechanical ventilation time [2]. Although the POCD is not rare, the underlying pathogenic mechanisms of POCD have not been known completely.
The systemic inflammation has been identified as an important process for the occurrence and development of POCD [4,5]. Not only surgical trauma, but also inhaled anesthetics could produce systemic inflammatory response, which leads to disruption of the bloodbrain, neuro-inflammation and cognitive dysfunction [6,7]. So inhibiting the inflammation might be a potential strategy preventing and/or treating cognitive dysfunction.
Vagal nerve has been identified to link to inflammatory response, its activity could suppress TNF-α and other pro-inflammatory cytokines. This anti-inflammation arc has been well-known as cholinergic anti-inflammation pathway [8]. Vagal nerve regulates inflammatory response through the hypothalamic-pituitary-adrenal axis, the release of cortisol and vagovagal reflex [9]. The anti-inflammation effects of vagal nerve stimulation (VNS) were firstly researched by Borovikova and colleague [10]. They found that not only acetylcholine, as vagal neurotransmitter, inhibited the release of inflammatory cytokines 4 in lipopolysaccharide stimulated human macrophage cultures, but also direct VNS could reduce the systemic inflammatory response and prevent the development of shock by attenuating inflammatory cytokines synthesis. Following their works, the antiinflammation mechanisms of VNS have been researched systemically. Although the whole mechanisms and the signaling pathway have been unknown completely, wider range of inflammatory disorders might be ameliorated by VNS [11].
The aim of this study, firstly, was to demonstrate whether surgery and general anesthesia could induce the systemic-and neuro-inflammatory response and impair the cognitive function in elder rats. Secondly, it was to evaluate the effects of VNS improving the cognitive dysfunction by inhibiting the inflammatory response.

Experimental Animals
A total of 33 male Sprague Dawley (SD) rats were obtained from Beijing Vital River Laboratory Animal Technology Co. Ltd. (Beijing, 100012, China. SCXK(Jing)2012-0001.) and maintained in suitable rooms with controlled conditions of temperature at 22±1℃, 40±10% relative humidity and light-dark cycles (12h-12h, light onset at 7:00). Five rats were housed in one cage. During he experiment, standard food and drink water were available to the animals ad libitum. This study was approved by the institution ethical committee for animal care and use, Sanbo Brain Hospital, Capital Medical University. All efforts were made to minimize the number of animals used and their suffering.
Experiments were performed during the day, always at the same time, to avoid circadian variations.

Experiment protocol
The rats were randomly divided into three groups: the control group (n=10, C group, 577.00±47.53g) that received no surgery and anesthesia, the splenectomy group (n=10, S 5 group, 577.60±33.73g) that was isolated cervical vagal nerve without stimulation, and the splenectomy+VNS group (n=13, SV group, 571.15±50.63g). Then all rats were conducted with Morris Water Maze (MWM) train (day 1-4) and test on day 7, Open Field Test (OFT) train on day 3 and test on day 7, the operation was carried on the day 4. Following the behavior tests, all animals were decapitated for tissue preparation on day 7 ( Figure 1).

Electric vagal nerve stimulation
Rats were fixed in a cage and anesthetized with 1% propofol (Fresenius Kabi AB.  [12]. During the experiment, the rats were allowed to breathe pure oxygen through a tube filled with oxygen continuously, which fixed in front of their nose in order to prevent hypoxia. After 10 min stabilization, the heart rate was recorded as the basic line. Thus, neck hair shaving and skin cleaning, aseptic technique was used to make a ventral midline incision in the neck. Then the skin and muscles were retracted.
Because right vagal nerve primarily innervates the atria and the sinoatrial node, and the stimulation may induce significant change in cardiac rhythm. At the same time, right VNS produce smaller cardiorespiratory response, so the right vagal nerve was applied [13,14].
Isolating the right cervical vagal nerve and common carotid artery bundle, a 1.5mm diameter silver bipolar cuff electrode was gently wrapped around the nerve bundle and fixed to the sternocleidomastoid muscle. Then the electrode was connected to the stimulator (BL-820 Biological signal acquisition and processing system. Chengdu Techman Software Co. Ltd. Sichuan, Chengdu, 610100, China). The basic stimulation parameters were adapted to the threshold of the individual animals and included 2V, 10Hz and 1ms, but these parameters were regulated continuously to preserve the heart rate lower than 6 the 10 percent of basic line [14][15][16][17].

Splenectomy
After 30min of VNS, the surgery began with a small lateral peritoneal incision. Using 3.0 silk thread to dissociate and ligate the spleen artery and vein at the hilum of spleen, the spleen was removed at the root of the far end of spleen pedicle. The completely removed spleen was examined to ensure that no residual spleen was left. Then the incision was closed, covering it with sterile activated-iodine gauze, and securing it with adhesive tape.
The operation time was within 60min.

Behavioral tests Open Field Test
The OFT was applied in an apparatus, which was made of brown plywood, surface area was 50×50cm, surrounded by 50cm high walls. The floor was divided by black line into 25 rectangles. Rats were allowed to freely move in the apparatus for 5min. The movement of individual animal in the arena was automatically tracked by AVTAS ver5.0 animal video analysis system (AniLab Software & Istruments Co., Ltd. Ningbo, China.) The number of crossing and rearing activities by each rat during the 5min was used to assess the rats' active explore behavior. After every test, the apparatus was cleaned with 5% ethanol.

Morris Water Maze
The MWM was used to evaluate the spatial reference learning and memory. A circle pool (150cm in diameter and 80cm deep) was filled with water and divided into four quadrants.
An escape platform, which 40cm in height and 15cm in diameter, was submerged by 2cm under water surface and conserved to the center of the northwest (NW) quadrant of the pool. Water was maintained at a temperature of 23±2℃. The evaluation consisted of four training days of five consecutive trails per day, the last trail of each day was accepted.
Rats were randomly introduced in the tank from the different quadrants facing the wall to 7 find the escape platform in 60 sec. If the rats did not find the platform within 60s in the first trial, they were gently guided to the platform to remain for 30s. Then the rats were removed from the tank. The procedure ensured the animals to retain the visual-spatial information during the trail. The movement of individual rats was automatically tracked by

TNF-α protein in hippocampus
The proteins were extracted from the hippocampus and their concentration was determined by the Bicinchoninic Acid Kit (Catalog BCA02, Dingguo Changsheng Biology 8 Technology LTD, Beijing, China). 30ug of protein samples were separated by SDS-PAGE, followed by semi-dry transfer onto a PVDF membrane. The membrane was blocked in 5% non-fat dry milk. Further, membranes were incubated with rat monoclonal primary antibody, anti-TNF-α and anti-tubulin, for overnight at 4℃. This was followed by triple washing with 0.1% TBST and incubation with HRP labeled secondary antibodies for 2h at RT. Immunoreactive bands were detected by ECL and Western blot detection system using Quantity one (Bio-Rad Laboratories. Inc. Hercules, CA, USA). The steps were repeated triply in order to calculate the mean. The final expression of each interest protein was calculated after normalizing the interest protein with the endogenous control tubulin in the same sample.

NF-κB in hippocampus expression
Total RNA was isolated from 100mg of hippocampal cortex using 1 ml of TRIzol (Invitrogen Corporation; Carlsbad, CA, USA) and then RNA was treated with RNase-free DNase I and quantified using Q5000 Spectrophotometer (Quawell Technology, Inc. San Jose, CA, USA). cDNA was obtained from 5ug of total RNA using 2ul of ReverTra Ace reverse transcriptase kit (Catalog TRT-101, TOYOBO STC (Shanghai) CO., LTD, Shanghai, China), 2ul of Oligo dT 50um, 2ul of dNTP mix 10mM, and water grade molecular biology to 20ul.
cDNA was used to amplify each gene using Sybr Green I (Catalog GG1301-50, Gen-View Scientific Inc. Florida, USA). The amplification reactions contained 1 ul of respective SybrGreen I, 12.5ul of Mix (Catalog PER012-1, Dingguo Changsheng Biology Technology LTD, Beijing, China), and 1ul of cDNA in a final volume of 25ul. The conditions were for qPCR were 3min for pre-denaturation at 95℃, followed by 35 cycles of amplification of 30s denaturation at 94℃, 30s annealing at 60℃ and 30s extending at 72℃. The last extending 9 step was 10min at 72℃. Rat GAPDH was used as internal control gene for normalization. to compare the groups. When the variance was heterogeneous, Dunnett T3 was used to compare the groups. The two-way repeated-measured ANOVA was use to analysis the data of escape latency. The mean difference was significant at the 0.05 level.

Results
There were no significant different between three groups on the body weight (F=0.073, Till to ten minute after the stopping VNS, the heart rate were no significant difference between three groups (F=0.230, P=0.796), and recovered to the basic level ( Figure 2).

The level of inflammatory cytokines in serum
The TNF-α level was increased by the surgery and general anesthesia in S group (61.028±8.642pg/ml, P<0.001, Dunnett T3) and SV group (41.609±8.249pg/ml, P<0.001, Dunnett T3), compared to the C group (27.180±2.038pg/ml). And it could be reduced by VNS, because TNF-α in SV group was significant lower than that in S group (P<0.001, Dunnett T3). The level of IL-6 was similar to the TNF-α. Compared to the C group (1.278±0.258pg/ml), the IL-6 in the S group (6.789±1.827pg/ml) and the SV group  5±1.3, P=0.543, LSD). The times of crossing platform of the SV group were more than that of the S group, but there was no significant difference between the two groups. After the operation, the duration of time spent in the target quadrant was no significant difference between the three groups (F=1.751, P=0.191). However, the time spending on the platform in the S group (11.49±2.58) was shorter than that of the C group (15.54±6.72) and the SV group (15.73±6.88).

Discussion
Our results showed that the surgery and general anesthesia could produce damage to memory and study capability, and aggravate POCD in elderly rats, because the surgery and general anesthesia induce systemic-and neuro-inflammation [19]. Although the improvement of cognitive function was not significantly, electrical VNS did reduce the development of POCD in some degree. The underlying mechanism may be related to the inhibition of the inflammatory response caused by the surgery and general anesthesia.
The inflammation caused by surgery and general anesthesia is the one of most important reason of POCD. There is strong evidence to suggest that acute inflammation affects and exacerbates cognitive function or cause delirium, one kind of clinical important postoperative complication. And the ongoing inflammation might constantly impair cognitive function after the surgery and anesthesia. Thus, inhibition or resolution of the inflammation is the important prerequisites for improvement of cognition [20]. In the present study, the surgery and general anesthesia increased pro-inflammatory cytokines, TNF-α and IL-6. Meanwhile, IL-10, a kind of anti-inflammatory cytokine, was also upgraded.
These changes demonstrated the acute systemic inflammation was induced via activation of the innate immune system.
As the initiating medium of systemic inflammation, TNF-α activates and amplifies the inflammation cascade. In addition, TNF-α can do damage directly to blood brain barrier and induce inflammatory cells infiltration in the hippocampus [21]. Besides, NF-κB is an 13 essential transcription factor to regulate inflammation and innate immunity genes expression. Its activation eventually increases the level of pro-inflammatory cytokines in the brain, such as TNF-α and IL-6, which subsequently cause cognitive impairment [22]. So NF-κB has been known as a potential therapeutic target, inhibition of NF-κB might improve cognitive dysfunction caused by sevoflurane anesthesia [23]. In our study, increase of proinflammation cytokines in serum and higher expression of TNF-α and NF-κB in hippocampus provided the directly evidence that surgery and general anesthesia might induce acute inflammation, which took place systemically and locally, and lead to cognitive damage eventually.
As to why right vagal nerve was stimulated in our study, this was because heart rate changed obviously when it was stimulated [13]. In other words, the change of heart rate is not only the evidence that the stimulation is successful, but also is the baseline to regulate the parameter of stimulation, such as the stimulation intensity and frequency. In the present study, 3min after the onset of stimulation, the heart rate reduced to about 85% of basic heart rate, and 10min after the stop of stimulation, the heart rate increased to the basic rate. During the operation, continuous VNS kept the heart rate at this level, till the operation was finished. In addition, right VNS might produce relatively slighter interruption to circulation and respiration than left VNS [14].
The more important is that vagal nerve plays key sole in inflammation regulation, which is known as cholinergic anti-inflammation pathway. The vagal nerve efferent fibers release acetylcholine, which binding with special receptor on macrophages and inhibit the production and release of TNF-α [24]. In the present study, this effects of VNS were certified that VNS could decrease not only the level of pro-inflammatory cytokines in serum, such as TNF-α and IL-6, also TNF-α protein and transcription factor NF-κB in hippocampus were downgraded by VNS. As to the level of IL-10, although VNS could 14 increase it slightly, there was no significant difference between the S and SV groups. We inferred that the anti-inflammation effects of vagal nerve focus on inhibition of proinflammatory cytokines, not upgrade of anti-inflammatory cytokines. So there was no significant change of IL-10.
Recently, Huffman and colleague reported that VNS could ameliorate cognitive response and decrease systemic and brain inflammation induced by lipopolysaccharide endotoxemia. In their study TNF-α was significantly inhibited by VNS [25], which was similar to the results of our research. Besides the effects of inhibiting inflammatory response, cholinergic system might regulate hippocampus function and memory, so it is not impossible to prevent or ameliorate POCD by VNS [26].
The OFT is used to evaluate spontaneous activity and anxiety-like behaviors. In the present study, the results of the OFT demonstrated that spontaneous movement in three groups were similar after the operation. Although VNS could ameliorated the crossing and rearing movement, a kind of actively explore behaviors [27], there were no significant difference between three group. VNS was reported to increase the score of OFT in the depression model of rat and antagonized the depressive status [28]. The results in the present study were not contradictory with the previous study, however this kind effects of VNS in our study was not significant statistically.
After the operation and general anesthesia, the learning and memory function of rats were damaged, as indicated by the increase of escape latency, and decrease of times of crossing platform and time spending on target region. VNS, as a kind of treatment, in the present study, it did decrease the escape latency, lengthen the time spending on target quarter, and increase the times of crossing platform, when compared to the S group.
These results were consistent with previous study which reported VNS ameliorated the cognitive function [29]. However, in our study, these effects of VNS were not enough to 15 produce protection against POCD completely.
In fact, the detail mechanisms of POCD have not completely elucidated. For exemple, systemic inflammation, neuroinflammatin, cerebral microemboli and hypotension, all of these may cause POCD. In other words, any pathogenies could cause POCD as long as they interrupt the central nerve system metabolic status and its homeostasis [30]. As well known, inflammation is the relatively important pathogeny for POCD, but it is not the only.
Thus, we could explain the results of our study. Because VNS could inhibit the inflammation induced by the surgery and general anesthesia, which demonstrated by lower level of pro-inflammation cytokines in the serum and pro-inflammation protein and gene expression in the hippocampus. However, there might be other pathogenies to advance the development of POCD. So VNS could ameliorate learning, memory function and actively explore movement in some degree, but its protection was not sufficient.
Our study indicated that surgery and general anesthesia could induce systemic and local inflammation. At the same time, the cognitive function of rats was damaged. VNS might inhibit the inflammation to produce protective effect against POCD in some degree.
However, this protection of VNS was insufficient. Maybe the combination of VNS with other therapies might provide better clinical effects, this need to be researched in the future. Ethics approval and consent to participate: We informed the ethics committee of Sanbo Brain Hospital, Capital Medical University, and got its approval. Consent for publication: There is no personal information in this article, so it is not applicable.
Availability of data and materials: The data of this article is available from the corresponding author. The email address of the corresponding author is B2008194@126.com.
Competing interests: Not applicable.
Funding: Not applicable.
Authors' contributions: JX is the first author who was responsible for the experiment. HJW is responsible for data analysis. JX and HJW made the equal contribution to the article. YXS is the corresponding author. Table   Table 1 The body weight and basic heart rate of three groups  Figure 1 The experiment protocol. MWM: Morris water maze. OFT: Open field test Figure 2 The basic HR was no significant difference between three groups. At 3min after begin of VNS, the HR in SV group was lower than the other groups. And at 3min after stop of VNS, the HR of SV group was slightly increased, compared to the other groups. 10min after stopping VNS, the HR of SV group return to the basic level, no difference of HR between three groups. The NF-κB gene expression in hippocampus. The NF-κB expression of the S group was increased significantly compared to the C and SV groups. There was similar expression between the C and SV groups. That showed VNS might reduce the proinflammation gene transcription.
25 Figure 6 The