Dexmedetomidine alleviates LPS-induced neuronal dysfunction via AKT/GSK-3β/CRMP-2 pathway

Dexmedetomidine (Dex), an α2-adrenergic receptor agonist, shows intervention effect on cognitive dysfunction in elderly patients after general anesthesia. However, the underlying pathological mechanism remains to be further elucidated. The aim of the study was to investigate effects of Dex on LPS-induced neuronal damages in cultured hippocampal neurons. We applied LPS to mimic the post-surgical inammation in cultured hippocampal neuron. Cultured hippocampal neurons were treated with LPS in a dose- and time- dependent manner, then were administrated with or without Dex. The neuronal morphology including neurite outgrowth and synaptic transmission was observed, and mEPSCs was recorded by electrophysiological patch clamp. Cell lysates were subjected for western blot to assess the explore the potential underlying mechanism.

even loss of self-care ability, prolongs the length of hospital stay, increases medical costs, and causes a series of medical, social and economic problems. POCD is closely related to neurodegenerative diseases in the elderly such as Alzheimer's disease (AD), Parkinson's disease (PD), Huntington's disease, multiple sclerosis [6,7]. Therefore, how to reduce the occurrence of cognitive dysfunction after general anesthesia in the elderly has become a research hotspot in recent years.
The in ammatory response can affect the function of the central nervous system, especially the in ammation of the central nervous system (CNS) can lead to changes in cognitive function [8], which leads to the occurrence of POCD. Current research suggests that both anesthesia and surgery may induce in ammatory responses in the central nervous system [9,10]. Thus, the occurrence and development of POCD is closely related to in ammatory response. Lipopolysaccharide (LPS; cell wall immunostimulatory component of Gram-negative bacteria) was originally identi ed as a Toll-like receptor 4 (TLR-4) ligand [11]. Once microglia activated by LPS, it produces pro-in ammatory cytokines such as TNF-α, IL-1β, prostaglandin E2 (PGE2), and NO [12,13]. These cytokines are key mediators that mediate the neuroin ammatory process. Inhibition of TLR-4 may abolish LPS induced in ammatory response [14]. The administration of LPS in animals induces cognitive impairment [15,16] and other complex dysfunctions, including anorexia, decreased exercise, weight loss, exploratory behavior, increased anxiety, lethargy and behavioral depression. These symptoms are very similar to the clinical symptoms of human neurodegenerative diseases. However, the speci c mechanism of LPS-induced cognitive impairment remains to be elucidated. Dexmedetomidine (Dex) is a highly selective α2 adrenergic receptor agonist, similar to remifentanil during anesthesia [17]. Dex also shows anti-in ammatory activity [18][19][20]. In recent years, Dex has been reported to have a certain intervention effect on cognitive dysfunction in elderly patients after general anesthesia. Data show that the administration of Dex extends the survival time and only 6% of patients with general anesthesia have postoperative delirium [21,22], comparing to 45% higher rate of postoperative delirium with the application of general drugs such as propofol or midazolam [23].
However, the detailed mechanism of Dex in reducing POCD remain unknown.
Here, the main purpose of this study is to determine whether Dex regulates LPS-induced neurological damage and trying to elucidate the underlying mechanisms, trying to provide new approaches and intervention targets for the clinical prevention and treatment of POCD.

Methods
Neuron culture and transfection Newborn 1-day (male and female) and 7 days (male) of Sprague-Dawley (SD) rats were provided by Zhongshan Animal Medical Center, Zhongshan University. The experiments were performed on 1-day-old newborn Sprague-Dawley rats, and were prepared as previously described (Zhang et al., 2007). Brie y, rats were sacri ced using CO 2 anesthetization, hippocampi from brains were dissected and gently minced brain tissue in a 60 mm tissue culture dish. The hippocampal tissue was incubated with 0.25% trypsin at 37°C for 5 min. Repeat a total of 3 times. Neurons were counted and cultured in Dulbecco's modi ed Eagle's medium (Gibco, USA), supplemented with 10% fetal bovine serum (Gibco), and plated in a 24 well plate with coverslips at a density of 6 × 10 4 cells/well. Neurons were placed in a 37℃, 5% CO2 incubator overnight and the medium was replaced with neurobasal feeding media (Neurobasal medium containing 2% B27 supplement and 0.5 mM glutamine solution). Half the volume of media was replaced with same volume of fresh neurobasal feeding media every 3 days. Transient transfections of neurons were performed with Lipofectamine ® 2000 according to the manufacturer's instructions. The protocol was approved by the Institutional Animal Care and Use Committee at Jinan University.

Western blotting
Protein samples were extracted from rat hippocampal neurons, and the protein concentrations were quanti ed using the BCA assay (Sigma, MO, USA). The extracted protein samples were separated by 10% SDS-PAGE with ~30 µg protein loaded per lane and transferred onto a PVDF membrane (EMD Millipore, MA, USA). The membrane was blocked with 5% non-fat milk in TBS , 0.1% tween-20 (TBST) at room temperature for 1 h and then was incubated with antibodies against p-AKT, AKT, p-GSK-3β, GSK-3β, p-CRMP2 and CRMP2 (all from Abcam, USA) in TBS buffer with 3% BSA at 4°C overnight, respectively. GAPDH was used as a loading control. After incubation with the secondary antibodies at room temperature for 1 h, the blots were visualized using enhanced chemiluminescence reagents. The intensity of the bands was analyzed quanti ed by densitometric analysis using Image-Pro Plus 7.0 (Media Cybernetics, Inc.). The Western blotting results are representative of three independent experiments.

Neuronal morphology analysis
All images were collected by ordinary uorescence microscope or confocal microscope. Image-Pro Plus software for neurite length or ImageJ software with the Neuronal Tracer plug-in and Sholl plug-in for the Sholl analysis were used to analysis the morphometric of protrusion as previously described [25]. Brie y, each image was processed with the Neuronal Tracer plug-in, and a mask of all neurite on a confocal image was drawn manually. By transfection GFP and staining GFP and Tau-1 to reveal all the neurite by immunocytochemistry. The total length of neurite and sholl analysis were performed to reveal the neuronal development.

Statistical analysis
The experimental data were presented as mean ± SEM from at least three experiments, and SPSS 19.0 software (SPSS Software, Chicago, IL, USA) was used to make statistics. Statistics were performed by ttest or One-way ANOVA method, p < 0.05 was considered to be signi cantly different, * or # donates p < 0.05, ** donates p < 0.01, *** donates p < 0.001.

LPS administration impairs hippocampal neurite outgrowth
In order to clarify the effect of LPS on the morphology of hippocampal neurons, neurons were transfected with GFP encoding plasmids and treated with different concentrations of LPS for 12 h (0, 0.1, 0.5,1.0, 2.0 μg/ml). LPS was found to induce signi cant changes in the growth of neuronal processes in a concentration-dependent manner (Fig. 1A). The protrusion length of LPS groups was signi cantly lower than that of the control group (p < 0.01) (Fig. 1B). Sholl analysis counted the number of branch points in different radius ranges showed that the complexity of neuronal morphology was also decreased in a concentration-dependent manner (Fig. 1C). We selected the dose of 1.0 μg/ml LPS to further explore the effect. Neurons were treated with LPS for different time range (0, 3, 6, 12 and 24 h; Fig. 1D) and the total length (Fig. 1E) and the sholl analysis (Fig. 1F) of protrusions were calculated. The data showed that LPS induced neuronal morphological changes in a time-dependent manner. In general, these data suggest that LPS impairs neurite outgrowth.

Dex antagonizes LPS-induced impairment of neurite outgrowth
To determine the effect of Dex on the development of hippocampal neuron development, the neuronal development was observed under the treatment of LPS with or without the addition of Dex. As shown in Fig. 2A, LPS signi cantly induced neuronal morphological changes and these damages were rescued by the addition of Dex. The length of total neurite (Fig. 2B) and number of intersection (Fig. 2C) in each group were calculated. The data suggest that Dex treatment would antagonize LPS induced damage on neuronal morphology.

Dex rescues synaptic transmission disorder induced by LPS
To verify whether Dex and LPS were related to neuronal function, electrophysiological patch-clamp assay was applied in cultured hippocampal neurons to determine the micro excitatory currents (mEPSCs) (Fig.   1A). As shown in Fig. 3B, the results showed that the synaptic transmission of neurons was impaired by LPS, and the amplitude (Fig. 3C) and frequency (Fig. 3D) of mEPSCs were signi cantly decreased. However, the suppressed amplitude and frequency of mEPSCs by LPS were restored after the addition of Dex. Dex alone treatment could induce the increase of amplitude and frequency. These data suggest that Dex could resist LPS induced synaptic transmission disorder.
Dex encounters the inhibitory effect of LPS via AKT/GSK-3β/CRMP-2 pathway To further explore the molecular mechanism of Dex resistance to LPS-induced neurodevelopmental impairment and synaptic transmission disorder, neuronal lysates were subjected to western blotting. The results showed that the phosphorylation levels of AKT and GSK-3β were downregulated, while that of CRMP-2 was upregulated under LPS treatment (Fig. 4A). Application of Dex signi cantly increased the phosphorylation levels of AKT and GSK-3β and decreased that of CRMP-2 and encountered the inhibitory effect of LPS (Fig. 4B). These results suggest that Dex may antagonize LPS-induced neuronal damage by activating the AKT/GSK-3β/CRMP-2 pathway.

Inhibition of AKT/GSK-3 abolishes the alternative effect of Dex against LPS
In order to further determine that Dex may interact with LPS-induced neural damage through activation of the AKT/GSK-3β/CRMP-2 pathway, pharmacological approach was applied. IGF was used to promote neuron development and AKT inhibitor LY294002 and GSK-3 inhibitor SB216763 were used to suppress the AKT/ AKT/GSK-3 axis. As shown in Fig. 5A, IGF function to activate AKT pathway, induced the phosphorylation levels of AKT and GSK-3β and inhibited that of CRMP-2, consistent with the pattern of Dex. However, the addition of AKT inhibitor and GSK-3 inhibitor signi cantly abolished the activation effect of Dex on this pathway, the phosphorylation levels of AKT, GSK-3β and CRMP-2 returned to levels equal to LPS treatment group (Fig. 5B). These data con rm that Dex functions via the AKT/GSK-3β/CRMP-2 pathway to antagonize LPS treatment in hippocampal neurons.

Discussion
In the current study, we demonstrated that LPS induction impaired hippocampal neuron development and disrupted synaptic transmission. Dex can antagonize LPS induced impairment on hippocampal neuronal developmental and synaptic transmission. In addition, DEX reactivated the AKT/GSK-3β/CRMP-2 pathway to antagonize LPS-induced nerve damage. POCD refers to the absence of mental disorders and abnormal brain function in elderly patients after surgery, referring to such as mental disorder, anxiety, personality changes and memory impairment [26]. General anesthesia combined with surgical shock can change the physiological function of elderly patients [27]. Under the stimulation of stress response, a series of adverse cardiovascular and cerebrovascular events can easily cause the occurrence of cognitive dysfunction in elderly patients [28]. It is generally believed that the pathogenesis of POCD is mainly considered from three aspects: (1) the in uence of surgical factors on the internal environment, such as the size of the surgical wound, the duration of the operation, stress response, arthritis response, micro thromboembolism, blood loss and uid loss [29]; (2) The effects of anesthetic factors on patients, such as the effect of general anesthetics on the central nervous system and the effects of anesthetics on the homeostasis of the body, such as hypotension, hypertension, bradycardia and hypothermia [30]; (3) The pathophysiological status and social factors of the patients undergoing surgery, such as the patient's old age, combined diabetes, hypertension, and cognitive impairment before surgery have a certain correlation with the occurrence of POCD [31]. Therefore, in-depth study of the neuropathic mechanism of POCD shows important scienti c value and signi cance.
Surgeries could trigger neuroin ammation, which would induced POCD [32]. Surgeries induced release of in ammatory factors or immune cells from the peripheral blood enter into the brain and affects the central nervous system [33]. These in ammatory factors could activate microglia cells to exaggerate immune response, leading to the release of large number of in ammatory factors [34], such as IL-1β, TNF-α and et al.. In addition, peripheral immune cells that entered the brain could participate to amplify the in ammatory response [35]. Accumulated in ammatory mediators cause reversible or irreversible damages to brain tissue, leading to the degeneration of neurite and cognitive dysfunction. The memory center, the hippocampus is more sensitive to overloaded in ammatory cytokines because of the widelyexpressed receptors [36]. Here, we applied LPS to induced in ammatory response in cultured primary hippocampal neurons. LPS signi cantly induced damages to neurite and synaptic transmission, which is consistent with previous studies [37].
In clinical practice, Dex is widely used as an α2-agonist [38]. Studies show that Dex performs antiin ammatory effect [39], where the underlying mechanisms has not been fully described. In a rat model of cerebral ischemia, Dex was found to decrease the blood catecholamine content and to decrease sympathetic nerve activity [40]. In septic shock rats, Dex inhibits systemic in ammatory responses and improves survival [41]. In spinal cord injury mode of rats, Dex also showed anti-in ammatory effects [42].
In clinical patients, Dex application markedly reduced the expression of pro-in ammatory cytokines, such as IL-1β, TNF-α and IL-6 [43]. In rat model of POCD, Dex protected aged rats from cognitive dysfunction by decreasing hippocampal in ammation [44]. Here, although we did not detect the levels of in ammatory factors, the data showed that Dex signi cantly alleviated LPS induced neurite outgrowth and synaptic transmission damages.
showed that gap junctional communication dysfunction during ischemia-reperfusion injury could cause cognitive impairment via PI3K/AKT pathway [47]. Rui et al. reported that miR-410 showed neuroprotective effect against sevo urane induced cognitive dysfunction in rats via PI3K/AKT pathway [48]. Here in the current study, LPS administration signi cantly suppressed the phosphorylation activation of AKT/GSK-3β/CRMP-2 pathway and Dex re-activated this pathway way to encounter the inhibitory effect of LPS in cultured hippocampal neurons, rescuing the neurite development and synaptic transmission disorder.
In summary, the mechanism by which Dex interferes with cognitive dysfunction in elderly patients after general anesthesia remains unclear. The main purpose of this study is to determine whether Dex regulates LPS-induced neurological damage, trying to provide new approaches and intervention targets for clinical prevention and treatment of cognitive dysfunction in POCD.

Conclusions
To summarize, we demonstrated that LPS-induced neuronal outgrowth inhibition and synaptic transmission dysfunction could be retarded by the administration of dexmedetomidine, via the regulation of PI3K/ AKT/GSK-3β signalling pathway. This study provided basis for the study of dexmedetomidine effect, assisting the clinical application for the treatment of POCD.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.   Dex reactivates the inhibited AKT/GSK-3β/CRMP-2 pathway by LPS. The hippocampal neurons were treated with LPS together with or without Dex. Cell lysates were subjected to western blotting and immune-stained with phosphorylated and total AKT, GSK-3β and CRMP2 antibodies to observe the activation of this pathway. The typical blots were shown in (A) and the gray intensity measurement was shown in (B). * represents p < 0.05, ** represents p < 0.01, and *** represents p < 0.001, compared to the control group. Pharmacological inhibition of AKT/GSK-3 pathway abolishes the antagonize effect of Dex against LPS. The hippocampal neurons were treated as indicated, with or without AKT inhibitor LY294002 or GSK-3 inhibitor SB216763. Then the cell lysates were subjected to western blotting as Figure 4. * represents p <0.05, ** represents p <0.01, and *** represents p <0.001, compared to the control group.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.