Combination of mild hypothermia with neuroprotectants has greater neuroprotective effects during oxygen-glucose deprivation and reoxygenation-mediated neuronal injury

Co-treatment of neuroprotective reagents may improve the therapeutic efficacy of hypothermia in protecting neurons during ischemic stroke. This study aimed to find promising drugs that enhance the neuroprotective effect of mild hypothermia (MH). 26 candidate drugs were selected based on different targets. Primary cultured cortical neurons were exposed to oxygen-glucose deprivation and reoxygenation (OGD/R) to induce neuronal damage, followed by either single treatment (a drug or MH) or a combination of a drug and MH. Results showed that, compared with single treatment, combination of MH with brain derived neurotrophic factor, glibenclamide, dizocilpine, human urinary kallidinogenase or neuroglobin displayed higher proportion of neuronal cell viability. The latter three drugs also caused less apoptosis rate in combined treatment. Furthermore, co-treatment of those three drugs and MH decreased the level of reactive oxygen species (ROS) and intracellular calcium accumulation, as well as stabilized mitochondrial membrane potential (MMP), indicating the combined neuroprotective effects are probably via inhibiting mitochondrial apoptosis pathway. Taken together, the study suggests that combined treatment with hypothermia and certain neuroprotective reagents provide a better protection against OGD/R-induced neuronal injury.

I schemic stroke is one of the most common diseases that cause death and disability worldwide, which brings a hard burden to families and society. It starts with sudden cessation of blood flow, oxygen, glucose and energy in the lesion area, followed by series of pathologic cascading events including exitotoxicity, calcium influx, free radicals accumulation, inflammation response, blood-brain barrier breakdown, edema, cell death and so on 1,2 .
Several neuroprotectants have synergistic neuroprotection with MH in animal model of cerebral ischemia, including BDNF 37 , magnesium sulfate 26 and albumin 18 . Whether other drugs have similar synergistic role is unclear. In our study, 26 drugs were selected based on the different targets,most of which have undergone phase I or II clinic trials. Oxygen-glucose deprivation and reoxygenation (OGD/ R) was used to induce the neuronal injury model in primary cultured cortical neurons to mimic the brain ischemia in vitro. With this model, we compared the neuroprotective effects of the 26 candidate medicines with or without MH. Among them, compared with single treatment, HUK, MK-801 or Ngb were shown to have better protective effects in combination with hypothermia against OGD/Rinduced neuronal damage.

Results
Screening neuroprotective reagents having better protective effects in combination with MH. To search drugs that have synergistic neuroprotective effects with MH, 26 drugs were firstly selected in combination with MH to treat primary cultured cortical neurons challenged with 3 hrs OGD and 24 hrs reoxygenation. Working concentrations of each drug were indicated in Table 1. The purity of mature neurons was more than 97% as assayed by staining of antibody against Neuronal Class III b-tubulin, a specific marker of neurons (data not shown). As shown in figure 1, OGD/R induced a dramatic reduction of cell viability proportion (CVP), while treatment of MH or six drugs, including Alb, BDNF, GBC, HUK MK801 and Ngb, significantly recovered CVP (P,0.05). Furthermore, combination of MH with each of above drug, except Alb, displayed higher CVP compared with each single treatment (P,0.05). We conducted two-way ANOVA and found that MH and four drugs (BDNF, GBC, MK801 or Ngb) have interactions (P,0.05). These results suggest that combination of MH with BDNF, GBC, HUK MK801 and Ngb has greater protective effects during OGD/R-mediated neuronal CVP reduction.
The remaining 21 reagents presented no additional protective effects combined with MH, compared to single treatment (data not shown).
Combined of MH with HUK,MK-801 or Ngb significantly reduced the level of intracellular ROS and calcium, and stabilized MMP after OGD/R. We then investigated the relevant mechanisms involved in the greater anti-apoptosis effects of the combo of drugs and MH. Neurons challenged with OGD/R have higher level of intracellular reactive oxygen species (ROS) and calcium influx, as well as reduced mitochondrial potential (MMP, DYm), all of which are landmarks of neuronal apoptosis and play important roles in cerebral ischemia and reperfusion 2 . Intracellular ROS was stained with specific probe DCF-DA and quantified according to optical densities (ODs). Results showed OGD/R significantly increased ROS generation, whose level was more than 10 times than that of normal cells (P,0.01, Figure 3). Moreover, the combined treatments of HUK, MK-801 or Ngb with MH decreased intracellular ROS after OGD/R, with the reduction extent more than any single treatment (P,0.05, Figure 3). Statistical analysis indicated that there were interaction between MK-801 or    displaying lower green/red signaling than each single treatment (P,0.05, Figure 5). There were interactions between MK-801 or Ngb and MHon MMP in primary neurons after OGD/R, according to two-way ANOVA analysis (P,0.05). These results, when taken together, suggest that the additional neuroprotective effects of the combined treatments of MH with HUK, Ngb or MK-801 are related to the reduction of calcium influx, intracellular level of ROS and stabilization of MMP.

Discussion
In this study, through screening 26 protective drugs, we are the first to identify HUK, Ngb, or MK-801, in combination with MH, greatly protected neurons against OGD/R-induced apoptotic cell death. Cotreatment of MH with HUK, Ngb or MK-801 induced greater inhibitory effects on ROS formation, intracellular calcium accumulation and mitochondrial depolarization, compared with either single treatment. The combination of MH with HUK, MK-801 or Ngb prevented neurons against OGD/R-induced apoptosis, probably through mitochondrial apoptosis pathway.
HUK, MK-801 and Ngb, presenting additional protection with hypothermia belong to kallikrein, non-competitive NMDAR blocker and nerve peptide, respectively. HUK, a serine protease isolated from human urine, is approved for the treatment of ischemia stroke in China. It can catalyze kininogen into bradykinin which binds to bradykinin B2 receptor 19 . HUK can produce anticoagulation effect 38 , promote vascular dilation in infarcted zone, boost endothelial cell proliferation 39 , inhibit neuronal apoptosis 19 and thereby protecting the ischemic brain. Our study provides additional neuroprotective role of HUK that enhances the effect of hypothermia.
MK-801 was reported to be effective in reversible focal ischemia in rats 40 . Unfortunately, it caused severe side effects in clinical study 41 . It might be better if lower dosage and combined therapy employed. A previous study showed the MK-801/hypothermia combination treatment group displays better neuronal protection in hypoxic-ischemic brain injury in neonatal rats 42 . Another study reported they have no additional protective effect in traumatic brain injury in neonatal rats 43 . We found 10 mM MK-801, in combination with MH, had better neuroprotective effect than any single treatment. This discovery might extend the animal study and clinic usage of MK-801 in the future.
Ngb was shown to regulate oxygen molecule delivery 44 and protected neurons in the model of ischemic stroke 45 . The combined role of Ngb and hypothermia has not been tested.
In summary, our study proved HUK, MK801 and Ngb were good candidates for the combination therapy with hypothermia during OGD/R-mediated neuronal injury.
Animal studies have demonstrated several drugs combined with hypothermia have synergistic neuroprotective effects during cerebral ischemia and reperfusion injury, for example, MgSO 4 46 , BDNF 47 and Alb 48 . Nevertheless, we did not found the better protective effects of these three drugs in cultured neurons compared with single treatment. The inconsistency of the in vitro and in vivo data might due to the cultured neurons lack the specific microenvironment composed of vascular system, stem cells, glia cells and other supporting cells. With the complicated environment in vivo, drugs have additional indirect pathway to protect the brain. For instance, MgSO 4 can dilates blood vessels 17 , while BDNF can promote neurogenesis 47 . We evaluate combined treatments only in cultured neurons which lose interaction with glia cells, endothelial cells and pericytes in neurovascular unit during ischemic stroke. Thus, limitation of the in vitro model used in our study may induce false-negative data.
Another typical example is the combination of GBC with hypothermia. GBC blocks Sur1/trpm4, a Na 1 channel located mainly expresses in endothelial cells, and thereby reducing oncotic endothelial cell death 49,50 . We found the combined effect of GBC with MH on CVP but not on apoptosis rate. Nevertheless, our unpublished data showed that GBC had synergistic protection with hypothermia in a rat model of cerebral ischemia. The neuronal apoptosis rate but not endothelial oncotic cell death was determined in this study might explain this inconsistency. It might also be due to the giant discrepancy of the environment in cell and animal system. Therefore, it is necessary to further evaluate the combined effect of five candidates or even more drugs with hypothermia in animal models. In conclusion, we identify three drugs, HUK, Ngb and MK-801, provides better protection in combination with hypothermia on cortical neurons against OGD/R-induced damage. Our study provides a new cue for understanding the combination role of drugs and hypothermia, and may offer new approach for clinical practice.

Methods
Ethics statement. All experiments in our study were approved and carried out in accordance with the Institutional Animal Care and Use Committee of the Laboratory Animals Center, Nanfang Hospital, Southern Medical University.
Primary culture of cortical neurons and immunocytochemistry. Primary cortical neurons were cultured as previously described 51 . Briefly, brains from 2 to 3-days-old postnatal Sprague-Dawley (SD) rats were isolated, minced, dissolved, filtered, and then suspended in the DMEM medium (Life Tech, Grand Island, NY) containing 10% fetal bovine serum (Life Tech). The suspended cells were then seeded on plates (3-5 3 10 5 /ml) coated poly-D-lysine (Sigma, St. Louis, MO) and cultured at 37uC in humidified air with 5% CO 2 . After 2-4 hrs incubation, the medium was replaced by neuronal cultured medium containing Neurobasal A(Life Tech) supplemented with B27 (Life Tech) and L-glutamine (Life Tech). The cells were cultured for additional 6 days (DIV 6). The purity of mature neurons was determined by immunostaining with the antibody against Neuronal Class III b-tubulin (151000, Beyotime, Shanghai, China), a specific marker of mature neurons.
OGD/R. The DIV 6 cortical neurons were exposed to OGD/R. Briefly, cultured medium was replaced by glucose-free Neurobasal A medium (Life Tech) and the cultured neurons were put in a hypoxic chamber at 37uC (Thermo Fisher, Waltham, MA) with a mix gas containing 5% CO 2 , 1% O 2 and 94% N 2 to reach final 2% O 2 , which was monitored with O 2 analyzer (GODEE, China). After 3 hrs OGD, neurons were returned to normal cultured conditions for 24 or 48 hrs reoxygenation. The time of management can be found as Supplementary Chart 1 online.
Treatments of MH and drugs. MH treatment was achieved by placing cultured cells in the cell culture incubator at 34uC environment with 5% CO 2 . After 3 hrs OGD, cells were placed at 34uC for 4.5 hrs and then returned to regular cell incubator for additional 19.5 hrs or 43.5 hrs. After OGD completed, neurons were immediately treated with drugs for 24 hrs or 48 hrs. The drugs used in our study were listed in Table 1. Three or more concentrations were selected around the concentration reported by literature.
Measurement of cell viability with Cell Counting Kit-8 (CCK-8). Cell viability was measured with CCK-8 (Dojindo Laboratories, Tokyo, Japan) according to the instructions of the manufacturers. In brief, DIV 6 cortical neurons were seeded in 96well plates with 150,000 cells/well in the proper medium described above. Six wells were prepared for each treatment or control. 10 ml of the CCK-8 mixture containing 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2Htetrazolium,monosodium salt (WST-8) solution was added to each well containing 100 ml medium. Cells were incubated at 37uC for 3.5 hrs. The ODs were measured at 450 nm with Spectra Max M5 Multi-Detection Microplate Readers (Molecular Devices, Sunnyvale, CA). 24 hrs after OGD or treatments, the ODs were measured again. Cell viability proportion (CVP) was determined by dividing ODs after OGD or treatments by ODs before OGD timed by 100%.
Measurement of apoptosis with Annexin V/PI staining. Flow Cytometery was used to count apoptotic cells with Annexin V-FITC Apoptosis Detection Kit (Sigma, St. Louis, MO). In brief, 48 hrs after OGD or treatment, neurons were harvested, washed and resuspended in binding buffer at a cell concentration of 1 3 10 6 /ml. For each 500 ml cell suspension, 5 ml Annexin V-FITC and 10 ml of propidium iodide were added, followed by detection with FACS (canto II, Becton, Dickinson and Company, Franklin Lakes, NJ).
Detection of ROS. 24 hrs after OGD, cells were incubated with 29,79dichlorofluorescin diacetate (DCF-DA, 20 mM/ml, 100 ml/well, Sigma, St. Louis, MO) in phenol red-free DMEM for 20 min at 37uC in the dark. DCF-DA, a nonfluorescent compound, can be oxidized by ROS into 29, 79-dichlorofluorescin (DCF), a highly fluorescent compound, whose signaling was observed by the Multi-Detection Microplate Readers with excitation wavelength at 488 nm and emission wavelength at 525 nm.
Statistical Analysis. Data were expressed as means 6 standard deprivation. For multiple groups designed experiments, comparisons were made by one-way ANOVA and followed by LSD test if variance was homogeneity, otherwise by Games-Howell test. Two-way ANOVA was applied to analyze interaction of two factors. The single effects and the main effects were analyzed by General Linear Model. All the tests were two-tailed. Statistical analysis was performed in the SPSS 16.0 statistical program (SPSS, Chicago, IL, USA). P,0.05 was considered to be statistically significant in the compared group.