Intracellular Ca2+ homeostasis and JAK1/STAT3 pathway are involved in the protective effect of propofol on BV2 microglia against hypoxia-induced inflammation and apoptosis

Background Perioperative hypoxia may induce microglial inflammation and apoptosis, resulting in brain injury. The neuroprotective effect of propofol against hypoxia has been reported, but the underlying mechanisms are far from clear. In this study, we explored whether and how propofol could attenuate microglia BV2 cells from CoCl2-induced hypoxic injury. Methods Mouse microglia BV2 cells were pretreated with propofol, and then stimulated with CoCl2. TNF-α level in the culture medium was measured by ELISA kit. Cell apoptosis and intracellular calcium concentration were measured by flow cytometry analysis. The effect of propofol on CoCl2-modulated expression of Ca2+/Calmodulin (CaM)-dependent protein kinase II (CAMKIIα), phosphorylated CAMKIIα (pCAMKIIα), STAT3, pSTAT3Y705, pSTAT3S727, ERK1/2, pERK1/2, pNFκB(p65), pro-caspase3, cleaved caspase 3, JAK1, pJAK1, JAK2, pJAK2 were detected by Western blot. Results In BV2 cell, CoCl2 treatment time-dependently increased TNF-α release and induced apoptosis, which were alleviated by propofol. CoCl2 (500μmol/L, 8h) treatment increased intracellular Ca2+ level, and caused the phosphorylation of CAMKIIα, ERK1/2 and NFκB (p65), as well as the activation of caspase 3. More importantly, these effects could be modulated by 25μmol/L propofol via maintaining intracellular Ca2+ homeostasis and via up-regulating the phosphorylation of JAK1 and STAT3 at Tyr705. Conclusion Propofol could protect BV2 microglia from hypoxia-induced inflammation and apoptosis. The potential mechanisms may involve the maintaining of intracellular Ca2+ homeostasis and the activation of JAK1/STAT3 pathway.


Introduction
Perioperative acute ischemic/hypoxic stroke is one of the most serious complications during many surgical procedures, and it is associated with high morbidity and mortality. In the central nervous system, microglia, acting as the unique resident immune cells, could be activated following brain ischemia/hypoxia [1,2]. Its activation can lead to the production and release of multiple pro-inflammatory cytokines, such as interleukin (IL)-1 and tumor necrosis factoralpha (TNF-α). TNF-α, the major pro-inflammatory cytokine, may serve as a marker for inflammation. In the central nervous system, TNF-α-induced inflammation may damage neural cells and the blood-brain barrier [3]. In addition, ischemia/hypoxia may also cause microglial apoptosis, leading to impaired immune response [4].
Intracellular calcium, one of the most important second messenger, played a pivotal role in multiple physical procedures and noxious stimulation can induce an increase of Intracellular Ca 2+ . Previous studies reported that hypoxia could induce intracellular Ca 2+ overload [5,6]. Microglial cells are the resident immune cells in the central nervous system and many of their physiological functions are linked to Ca 2+ signaling [7]. Previous studies showed that morphological changes, migration, proliferation, and secretion of cytokines and reactive oxygen species were related to intracellular Ca 2+ overload [8,9]. As the major isoform of Ca 2+ /calmodulin-dependent protein kinase (CaMK) in brain, CaMKIIα is highly sensitive to intracellular Ca 2+ levels. Recent data suggested that the activation of CaMKIIα is directly associated with the production of pro-inflammatory cytokines, such as TNF-α and IL-1β and the mechanism may involve phosphorylating and upregulating the expression of several down-stream proteins [10,11]. Previous studies have showed that ERK 1/2, nuclear factor kappa B (NF-κB) are involved in Ca 2+ -mediated TNF-α release [12].
Janus kinase-signal transducers and activators of transcription (JAK-STAT) signal pathway has been reported to be involved in the immune response of numerous cytokines [13]. Activation of STAT3 and JAK has been shown to mediate ischemic and several pharmacological postconditioning [14][15][16]. Phosphorylation of JAKs may regulate STAT3's activity by phosphorylating STAT3 on tyrosine705 and serine727 residues. Reduced phosphorylation of STAT3 at Ser727 is usually correlated with increased phosphorylation at Tyr705 [17,18]. A previous study showed that JAK1 and STAT3 are activated in neurons, astrocytes and microglia after focal cerebral infarction, and may provide neuroprotection in the acute phase of ischemia [14].
Propofol (2,6-diisopro-pylphenol) is a widely used intravenous anesthetic. In addition to sedative property, propofol has been proved to exert beneficial effects in multiple organs and tissues, such as cardiovascular system [19], respiratory system [20] and urinary system [21]. In brain injury model, a previous study showed that propofol could protect central nervous system from hypoxic injury via improving the oxygen supply and maintaining oxygen metabolism [22]. Another study also showed that propofol protects the brain through maintaining intracellular Ca 2+ homeostasis [23]. However, other groups reported propofol-induced neuroinflammation and cell death in young rodents [24]. An animal study showed that propofol protected against focal cerebral ischemia via inhibition of microglia-mediated proinflammatory cytokines [25]. However, the mechanism of propofol inhibiting hypoxia-induced microglial activation is yet to be investigated. In this study, we used cobalt chloride (CoCl 2 ) to build an in vitro hypoxic model and aimed to clarify whether and how propofol attenuates CoCl 2induced BV2 cell injury.

Materials and methods
Cell culture and reagents BV2 microglia cells were obtained from GuangZhou Jennio Bio-tech. The cells were cultured in DMEM (Sigma-Aldrich, Shanghai, China) with 10% fetal bovine serum (Gibco, Life technologies, USA), 100 units/ml penicillin (sigma) and 100 μg/ml streptomycin (sigma) in an incubator containing 5% CO 2 at 37˚C. BV2 cells were sub-cultured when reaching 90% confluence. The eighth passage was used in the present study. Propofol

Experiment design
To determine the appropriate treatment condition, BV2 cells were treated with 500 μM CoCl 2 (sigma) for 0, 1, 2, 4, 8 and 12 h respectively. Cells culture in DMEM, without any treatment, served as the control group. The time with maximal effects on TNF-α production was used as the appropriate treatment condition. Then cells were pre-treated with different concentrations of propofol (sigma) (5 μmol/L, 25 μmol/L, 50 μmol/L and 100 μmol/L), followed by CoCl 2 (500 μM, 8 h) treatment, and the concentration with maximal protective effects was determined. In the following experiments, the optimal concentration of CoCl 2 and propofol were used to investigate the potential mechanisms.

Enzyme-linked immunosorbent assay (ELISA)
To determine TNF-α concentration in the culture medium, a TNF-α sandwich ELISA kit from BioSource International Inc was used according to the manufacturer's instructions. The absorbance at 450 nm was measured with a microplate reader. The range of detection was from 0 to 1000 pg/ml.

Measurement of intracellular free Ca 2+ concentration
Intracellular free Ca 2+ concentration was detected by the fluorescent dye Fluo-3 AM (Beyotime biotechnology, shanghai, China). The Fluo-3 could specifically bind to the Ca 2+ and has a strong fluorescence with an excitation wavelength of 488 nm. After designed treatment, BV2 cells were harvested and washed twice with PBS, then resuspended with fluo-3 AM (5 μM) for 30 min in the dark. Intracellular Ca 2+ was detected by Flow cytometer at 488 nm excitation wavelength.

Cell apoptosis analysis
Cell apoptosis was detected by propidium iodide (PI) and annexin V staining according to the manufacture's instructions. Briefly, after designed treatment, cells were harvested and stained with annexin V and PI. Then stained cells were analyzed by flow cytometry (Cytomics FC 500 MPL, Beckman Coulter).

Statistical analysis
Results were expressed as mean ± SD, and data were obtained from at least 5 separately performed experiments. Differences between groups was determined by one-way ANOVA followed by the Newman-Keuls test using the InStat statistical program (GraphPad Software, San Diego, CA, USA). All results were considered statistically significant at a value of p< 0.05.

Effects of propofol on CoCl 2 induced TNF-α generation
In BV2 cells, 500μM CoCl 2 treatment induced TNF-α production in a time-dependent manner. Compared with control, we found that treatment of cells with 500μM CoCl 2 for 8 hours induced a notably generation of TNF-α ( Fig 1A, p< 0.05). To observe the effect of propofol on CoCl 2 induced TNF-α generation, BV2 cells were pretreated with propofol for 2 h with different concentration (5, 25, 50, 100μM) after CoCl 2 treatment. As shown in Fig 1B, 25μM propofol significantly reduced the CoCl 2 -induced generation of TNF-α. We found that propofol alone had no effect on TNF-α generation ( Fig 1C). Thereafter, 8 h treatment of 500μM CoCl 2 and 25μM of propofol pretreatment for 2h were used in the following experiments to study the signaling pathway involved in the protective effects of propofol.

Effects of propofol on CoCl 2 modulated JAK/STAT3 pathway
Previous studies have shown that after focal cerebral infarction, the activation of JAK1 and STAT3 in microglia may provide neuroprotection in the acute phase of ischemia [14]. Here, we found that CoCl 2 treatment decreased the expression and phosphorylation of JAK1 ( Fig  4A, 4B and 4C, p<0.05), but increased the expression and phosphorylation of JAK2 (Fig 4D,  4E and 4F, p<0.05). Besides, CoCl 2 treatment decreased the phosphorylation of STAT3 at Tyr-705, but had no effect on the phosphorylation of STAT3 at Ser-727 (Fig 4G, 4H and 4I, p<0.05). More importantly, we found propofol attenuated these CoCl 2 -modulated effects (Fig 4, p<0.05).

Effects of propofol, BAPTA-AM, KN93, U0126, INCB039110 on CoCl 2 induced TNF-α generation
To confirm the role of Ca 2+ homeostasis and the phosphorylation of CAMKIIα, ERK and NF-κB on the production of TNF-α, BV2 cells were pretreated with calcium chelator BAPTA-AM, CAMKIIα inhibitor KN93, or ERK inhibitor U0126 followed by CoCl 2 treatment. We found that compared with CoCl 2 treatment, BAPTA-AM, KN93 and U0126 significantly decreased the generation of TNF-α, which was similar to propofol treatment. Further, to confirm the role of JAK1/STAT3 pathway in the anti-inflammatory property of propofol, BV2 cells were pretreated with propofol and selective JAK1 inhibitor INCB039110 followed by CoCl 2 treatment. We found INCB039110 alleviated the effect of propofol, resulting in increased production of TNF-α (Fig 5, p<0.05).

Effects of propofol, BAPTA-AM, KN93, U0126 on CoCl 2 induced the phosphorylation of CAMKIIα, ERK, NF-κB
To confirm the role of the phosphorylation of CAMKIIα, ERK and NF-κB in the protective effects of propofol against CoCl 2 treatment, BV2 cells were pretreated with calcium chelator BAPTA-AM, CAMKIIα inhibitor KN93, or ERK inhibitor U0126 followed by CoCl 2 treatment. We found that compared with CoCl 2 treatment, BAPTA-AM and KN93 significantly decreased the phosphorylation of CAMKIIα, which was similar with propofol treatment. However, U0126 pretreatment had no effect on the phosphorylation of CAMKIIα (Fig 6A and  6B, p<0.05). Compared to CoCl 2 treatment, BAPTA-AM, KN93 and U0126 significantly decreased the phosphorylation of ERK (Fig 6C and 6D, p<0.05) and NF-κB (Fig 6E and 6F, p<0.05), which was similar to propofol treatment.

Discussion
In the present study, we found that CoCl 2 treatment could disrupt the intracellular Ca 2+ hemostasis, and phosphorylate CAMKIIα, ERK and NF-κB, resulting in the production of TNF-α and apoptosis. More importantly, we found propofol could protect BV2 cells from CoCl 2induced injury via maintaining the intracellular Ca 2+ hemostasis and via inhibiting Ca 2+ -related signaling pathway, as well as via up-regulating JAK1/STAT3 pathway.
Emerging evidences have suggested that transient cerebral ischemia/hypoxia may lead to microglial malfunction and microglial activation may lead to neurodegeneration via inflammation or cell injury [26][27][28]. Here, we used CoCl 2 -treated BV2 cell line to mimic the cerebral hypoxia in vivo. We found CoCl 2 treatment induced the release of TNF-α (Fig 1). TNF-α is a main pro-inflammatory cytokine, and is mainly produced by activated microglia during the neuroinflammation. Growing evidence showed that the overproduction of TNF-α by microglia contributes to pathophysiological changes observed in various neurologic diseases and brain injury [3,29]. In addition to inflammation caused by ischemia/hypoxia, we also found CoCl 2 treatment induced cell apoptosis (Fig 2).
As mentioned above, intracellular calcium plays a key role in maintain cell function and noxious stimulations can induce an increase of intracellular Ca 2+ concentration. Besides, intracellular Ca 2+ overload could upregulate the expression of several down-stream proteins. In the situation of ischemic myocardial stress response, CAMKIIα could activate ERK and NF-κB, thus regulating inflammation and injury [12]. In our study, as shown in Fig 3, CoCl 2 treatment disrupted the intracellular Ca 2+ hemostasis and inducing the phosphorylation of CAMKIIα, ERK and NF-κB. JAK/STAT signaling plays an essential role in promoting and modulating immune and inflammatory processes [30]. Activation of the JAK/STAT pathway has been associated with Propofol protected BV2 microglia against hypoxia injury pathological conditions such as cerebral ischemia, traumatic brain injury and brain inflammation [31][32][33]. In the central nervous system, STAT1, STAT3 and STAT6 play an important role during brain development. It has been demonstrated that STAT3 plays critical roles in promoting the survival of neurons under pathological conditions [34]. It is known that STAT3 is activated by JAK1 and JAK2 in neuron cell lines [35]. As shown in Fig 4, we found CoCl 2 and U0126 had no effect on the expression of pro-caspase 3. (* p < 0.05 vs. control, # p < 0.05 vs. CoCl 2 treatement, n = 5, Data were shown as mean ± SD).
affected the phosphorylation of both JAK1 and JAK2. In addition, STAT3 function is controlled by phosphorylation at two sites: phosphorylation of Tyr705 by JAK leads to STAT3 transcriptional activation whereas the role of the second phosphorylation site (Ser727) varies depending on cell type and pathological conditions [36]. Besides, Ser727 phosphorylation was considered a negative regulatory mechanism of STAT3 activity [37]. A previous study showed that JAK1 and STAT3 are activated in neurons, astrocytes and microglia after focal cerebral infarction, and may provide neuroprotection in the acute phase of ischemia [14]. As shown in Fig 4, our study was in line with previous theory.
The neuroprotective effects of propofol have been extensively studied in previous investigations [38]. In this study, our data suggested that 25μM propofol inhibited microglial activation and suppresses cytokine release as well as apoptosis, which were induced by hypoxia injury. And our data is consistent with the previous study, which found that propofol suppressed hypoxia/reoxygenation-induced apoptosis in HBVSMC [39]. In patients, the effect site (brain) concentrations of propofol have been reported to range from about 2 to 6 μg/mL, which are equivalent to 11 to 33μM. In BV2 microglial cells, 30μM propofol was considered a clinically relevant concentration [40]. Therefore, in our study, the effective concentration of propofol was clinically relevant. Further, the underlying mechanism may involve maintaining intracellular Ca 2+ hemostasis. Because we used BAPTA-AM, a calcium chelator, and found the effect of BAPTA-AM was similar to that of propofol. Consistently, animal study has showed that propofol exert neuroprotective effect via preventing calcium-induced mitochondrial swelling [41]. Besides, the JAK1/STAT3 pathway was also involved in the protective mechanism. In cervical cancer cells, previous study showed that propofol could enhance cisplatin-induced apoptosis via EGFR/JAK2/STAT3 pathway [42]. However, in our study, we found propofol could exert protective effects via activating JAK1, and inducing the phosphorylation of STAT3 at Tyr-705. Consistently, we used INCB039110, a selective JAK1 inhibitor, and found it could reverse the effect of propofol.
There are some limitations in this study. Firstly, the study was carried out in cultured cell line. It is noted that the data obtained from cell line may differ from those from primary culture and animal studies, and we are planning to perform the experiment in the mice model to validate our findings. Secondly, we found the beneficial effects of propofol were mediated via activating JAK1/STAT3 pathway, but we didn't exam how JAK1/STAT3 pathway was activated. Future experiments are required to address this issue.
In summary, our study identified that propofol could exhibit anti-neuroinflammatory activity in vitro by suppressing the pro-inflammatory mediators from CoCl 2 -induced BV2 microglial cells and in turn prevents neuronal from further damage. Additionally, these beneficial effects were mediated via inhibiting cellular Ca 2+ overload, modulating the phosphorylation of CaMKII, ERK and NF-κB, and activating JAK1/STAT3 pathway.

Author Contributions
Conceptualization: JC CM.