Chemicals, including synthetic (±)-α-tocopherol (cat. no. T-3251) and ascorbic acid (cat. no. A4544), were purchased from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany) and the highest grade available was used. Hank's balanced salt solution (HBSS) was used, comprising 8.0 g/l NaCl, 0.4 g/l KCl, 1.0 g/l glucose, 0.06 g/l KH₂PO₄ and 0.09 g/l Na₂PO₄-7H₂O, pH 7.1; glucose-free HBSS was prepared by omitting glucose and adding an additional half-molar amount of NaCl to adjust the osmolarity (8.162 g/l NaCl). When used in the present study, magnolol (Wako Pure Chemical Industries, Ltd., Osaka, Japan) was dissolved in dimethylsulfoxide (DMSO) or polyethylene glycol 400 (PEG 400; Merck KGaA).

A malondialdehyde (MDA) assay was performed as previously described (19–21). The rat brain tissue was homogenized in 20 mM Tris-HCl buffer, centrifuged at 1,000 x g for 10 min at 4°C. To 30 µl brain homogenate supernatant was added 10 µl vehicle (0.1% DMSO), magnolol (0.01 µM - 1 mM), α-tocopherol (0.1–100 µM), ascorbic acid (1–5 mM) or β-estradiol (1–150 µM) and 5 µl freshly prepared ferric chloride hexahydrate. The absorbance was measured using a plate reader at 532 nm. The 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay was conducted by adding 100 µl freshly prepared DPPH radical solution to 100 µl vehicle (0.1% DMSO) or magnolol (0.01 µM - 1 mM) as previously described (19–21). The absorbance was measured at 517 nm. The 2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS) assay was conducted by adding 100 µl ABTS radical cation solution to 100 µl vehicle (0.1% DMSO) or magnolol (0.01 µM - 1 mM) (19–21). The absorbance was measured at 734 nm.

Mouse macrophage cell line RAW 264.7 and microglial cell line BV2 (American Type Culture Collection, Manassas, VA, USA) cultured in DMEM (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) were stimulated with lipopolysaccharide (LPS; Sigma-Aldrich; Merck KGaA), 10 and 100 ng/ml, respectively, based on preliminary dose-response experiments, and co-cultured with PBS, vehicle (0.1% DMSO) or magnolol (0.1–50 µM). Following 6 h of incubation, the levels of IL-6 and TNF-α in the supernatant were determined using IL-6 (cat. no. DY-406) and TNF-a (cat. no. DY410-05) DuoSet ELISA kits (R&D Systems, Inc., Minneapolis, MN, USA) as previously described (19–21). The levels of nitric oxide (NO) in the supernatant were measured using a nitrate/nitrite fluorometric assay kit (cat. no. 780051; Cayman Chemical Company, Ann Arbor, MI, USA) following 16 h of incubation (19–21).

Organotypic hippocampal slices were obtained by harvesting 300 µm-thick hippocampal slices using a motorized NVSLM1 Vibroslice (Campden Instruments Ltd., Loughborough, UK) from 6–7-day-old, neonatal Sprague-Dawley rats weighing 20–30 g. The animals were allowed free access to food and water and housed at 25°C and 60% humidity with a 12-h light/dark cycle (19). The slices were placed on 0.4-µm Millicell culture inserts (EMD Millipore, Billerica, MA, USA) and transferred to 35 mm Petri dishes. Slices were cultured with organotypic culture medium consisting of 25% horse serum (Gibco; Thermo Fisher Scientific, Inc.), 25% HBSS and 50% minimum essential medium (both MP Biomedicals, LLC, Santa Ana, CA, USA), 5 mg/ml glucose, 1 mM glutamine, 5 mM KCl and 1.5% Fungizone (Gibco; Thermo Fisher Scientific, Inc.) at 37°C and the culture medium was refreshed every 3 days.

The OGD model was generated by hypoxia combined with aglycemia, as previously described (19–21). Briefly, the hippocampal slices were cultured for 8–14 days and incubated with medium (1 ml/well glucose-free HBSS, bubbled with 95% N₂ and 5% CO₂) at 37°C for 20 min prior to OGD. The culture dishes were incubated in an anaerobic chamber at 37°C with 95% N₂ and 5% CO₂ for 120 min. A normal control group was also established, in which the hippocampal slices were incubated in HBSS with 95% O₂ and 5% CO₂ at 37°C. Replenishment of the cells was induced by replacing the medium and incubating the dishes under normoxic conditions for a further 24 h. To evaluate cell damage, 5 µg/ml propidium iodide (PI) was incubated with the culture at 37°C for 20 min, and the absorbance was measured at 630 nm and analyzed using Image Pro Plus software, version 5.1 (Media Cybernetics, Inc., Rockville, MD, USA) with a digital CoolSNAP-Pro_cf camera (Media Cybernetics, Inc.) and an IX71 epi-fluorescence inverted micro-scope (Olympus Corporation, Tokyo, Japan). The fluorescence intensity (Ft) of the slices was obtained from three different fields of the slices. Regions without the slices were used as the background intensity (F0). At 48 h after OGD, culture slices were incubated with 10 mM glutamate to determine the final PI fluorescence (Ffin). The PI uptake, as an indicator of cell death, was subsequently obtained from the following equation: PI (%)=(Ft-F0)/(Ffin-F0) ×100 (19). An OGD duration of 120 min induces 70% of the maximal PI-uptake in CA1 pyramidal neurons (19). Experimental treatment groups were established as follows: i) Pre-treatment group, treated with magnolol (10 or 100 µM) or vehicle (0.1% DMSO) for 1 h prior to the OGD period, with no additions to the medium during the following 24 h; ii) post-treatment group, treated with magnolol (100 µM) or vehicle (0.1% DMSO) at 2, 4 or 6 h post OGD.

Male Sprague-Dawley rats (n=120; 8 weeks old, weighing 260–300 g), were obtained from the National Cheng Kung University Animal Center (Tainan, Taiwan). The animals were allowed free access to food and water and housed at 25°C, 60% humidity on a 12-h light/dark cycle. Rats were anesthetized with halothane, 3–4% for induction and 1–2% for maintenance. The right femoral artery was cannulated for the measurements of arterial blood gases, hematocrit, glucose, heart rate and blood pressure. During surgery the core temperature was maintained at 37.0±0.5°C by the use of a heating blanket (Harvard Apparatus, Holliston, MA, USA).

All procedures within the present study were performed in accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (eighth edition, 2011). The protocol for the present study was approved by the Committee on the Ethics of Animal Experiments of the National Cheng Kung University Hospital (permit no. 102249; Tainan, Taiwan).

A stable ischemia animal model was generated as previously described, by the occlusion of the right proximal middle cerebral artery (MCA) with an intra-arterial suture for 90 min (19–21). The local cortical cerebral perfusion (local cerebral blood flow, LCBF) was measured using Laserflo BMP2 Laser-Doppler flowmetry (Vasamed Inc., Eden Prairie, MN, USA) to ensure the quality of the ischemia and reperfusion induction (19–23).

The investigators were blinded to the treatment paradigms and the rats were randomly separated into Magnolol and vehicle groups. In a first series of experiments, magnolol (0.01, 0.1, 1 or 5 mg/kg; n=10, 10, 13 and 9, respectively) or vehicle (PEG 400, 10 ml/kg; n=12) was administered intravenously (i.v.) 30 min prior to the onset of ischemia. Rats were euthanized 72 h following the onset of ischemia and their brains examined using standard 2,3,5-triphenyltetrazolium chloride (TTC)-stained histological sections. In a second series of experiments, rats were treated with magnolol (1 mg/kg, i.v.; n=10) or vehicle (i.v.; n=10) 2 h after the onset of ischemia. Following ischemia for 72 h, rats were euthanized using histological sections and immunofluorescence staining.

Another set of rats were treated with magnolol (0.01, 0.3 or 1 mg/kg, i.v.; n=8 per group) or vehicle (PEG 400; 10 ml/kg, i.v.; n=8) at 2 h following the onset of ischemia, and euthanized at 24 h post-insult, to assess the levels of myeloperoxidase (MPO), malondialdehyde (MDA), NO_X and the ratio of reduced glutathione/oxidized glutathione (GSH/GSSG) in the brain tissue. The NO_X level was accessed using the aforementioned nitrate/nitrite fluorometric assay kit.

MDA and MPO activity in brain tissue were determined as previously described (19,20). Absorbance was measured using a plate reader at 532 and 492 nm, respectively.

Rat brains were sectioned coronally as previously described (20,22,23), into 40-µm sections at 1-mm intervals from the bregma AP (4.2 to −6.8 mm) on a cryostat. Two sets of sections were used for the immunofluorescence staining of 4-HNE and 8-OHdG. The expression of 4-HNE and 8-OHdG was used to identify membranous lipid peroxidation and DNA damage, respectively (5). Sections were washed in PBS (Molecular Probes; Thermo Fisher Scientific, Inc.) and incubated in 3% H₂O₂ with 50% methanol/PBS at room temperature for 30 min, then placed in PBS containing 0.3% Triton X-100 and blocked with 1% normal sheep serum (Millipore, Merck KGaA) at room temperature for 30 min. Sections were incubated with monoclonal antibodies directed against 8-OHdG (1:750; MOG-100P) and 4-HNE (1:1,000; MHN-100P; both JaICA, Fukuroi, Japan) overnight at 4°C. Goat anti-rabbit IgG secondary antibodies conjugated with biotin (1:100; cat. no. 111-065-144; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) were subsequently incubated with the sections at room temperature for 1 h, followed by fluorescein (DTAF)-conjugated streptavidin (1:150; cat. no. 016-010-084; Jackson ImmunoResearch Laboratories, Inc.). Immunopositive cells were traced and measured using a Zeiss Axioskop 2 Mot microscope (Zeiss AG, Oberkochen, Germany) equipped with a digital CoolSNAP-Pro_cf camera and computerized image analyzer (MCID Elite 6.01.4; Imaging Research Inc., St Catharines, ON, Canada). Fluorescence was measured at 450–490 nm and emission >515 nm for DTAF detection.

For the assessment of ischemic damage to the neuronal perikarya, another set of sections was stained with 0.5% cresyl violet. Using light microscopy (Zeiss Axioskop 2 Mot micro-scope), the areas of neuronal perikarya displaying typical morphologic features of ischemic damage were evaluated.

Sections 40 µm in thickness were stained with 5 mM DHE at 37°C for 30 min as previously described (5), and subsequently incubated with DAPI (2 µM) in PBS at room temperature for 15 sec in a dark chamber. Section fluorescence was evaluated at an excitation wavelength of 510–550 nm and emission wavelength of >580 nm for oxidized DHE (HEt) detection, and at excitation and emission wavelengths of 365 and >420 nm for DAPI detection. The density and intensity of the HEt in the ischemic brain tissue was compared between the magnolol-treated animals and the controls.

In the ischemic brain tissues, the total GSH and GSSG levels were determined as previously described by Anderson (24). Briefly, GSH was oxidized by 5,5′-dithio-bis(2-nitrobenzoic acid) to generate GSSG with stoichiometric formation of 5-thio-2-nitrobenzoic acid (TNB). GSSG was subsequently reduced to GSH by the action of the highly specific glutathione reductase and NADPH. The rate of TNB formation was measured at 412 nm and was proportional to the sum of GSH and GSSG. For GSSG determination, 2-vinylpyridine was used to destroy the reduced form of GSH, followed by the same procedure as the GSH assay.

An5imals were euthanized at 72 h following the onset of ischemia. A rat brain matrix (RBM 4000C; ASI Instruments Inc., Warren, MI, USA) was used to cut 2-mm coronal sections. These sections were stained using TTC at room temperature for 30 min in a dark chamber (15). In the delayed treatment paradigm, ischemic brain damage was measured using TTC and cresyl violet-stained brain sections as previously described (20,22,23). Coronal section images were captured using a DXC-390 3-CCD color camera (Sony, Tokyo, Japan) equipped with a Micro-Nikon 55 mm f2.8 lens. Areas of ischemic damage were analyzed using a computerized image analyzer (MCID Elite 6.01.4). The cerebral ischemic infarction volume was expressed as a percentage of the contralateral hemisphere volume as previously described (20,22,23). Additionally, individual striatal and cortical infarct sizes were calculated.

Sections were obtained between the Bregma AP-0.22 and −0.78 mm. Six non-overlapping regions (500×400 µm²) were randomly selected for cell counting of the surviving neurons in the ischemic brain, as previously described (20,22,23). Surviving neurons were expressed as the mean number of viable neurons per mm².

Neurologic evaluation was conducted 1 and 3 days following stroke induction by a qualified observer unaware of the treatment protocol. Two different neurologic grading systems were used in the present study. The first was a sensorimotor grading scale with five categories (0–4) for forward and sideways visual placing tests of the affected forelimb, and five categories (0–4) for motor outcome, which is a modification of previously described methods (20,22,23,25). The second was a neurobehavioral grading scale of 0 to 28 for rodents, as previously described (26).

Data are expressed as the mean ± standard deviation and differences between groups were evaluated using an unpaired Student's t-test or one-way analysis of variance with Fisher's protected least significant difference post hoc test. To evaluate the response in changing conditions a paired Student's t-test was used. The data for neurobehavioral testing are expressed as the median (95% confidence interval) and were evaluated using a nonparametric test for independent groups, including the Kruskal-Wallis and Mann-Whitney U tests. SPSS software, version 17.0 (SPSS, Inc., Chicago, IL, USA) was used to conduct the analysis. P<0.05 was considered to indicate a statistically significant difference.

Magnolol attenuated Fe³⁺-induced lipid peroxidation in brain tissue homogenate with an IC₅₀ value of 0.8±2.8 µM (Table I), which was lower than the IC₅₀ values for α-tocopherol, ascorbic acid and β-estradiol (45.9±7.4, 2,770.0±110.0 and 65.6±7.9 µM, respectively). However, in the DPPH radical scavenging assay, the IC₅₀ value of magnolol was 531.9±1.0 µM, which was greater than the IC₅₀ value for a-tocopherol and ascorbic acid (84.4±2.8 and 47.5±2.4 µM, respectively), but lower than that for β-estradiol (202.4±18.0 mM). In the ABTS radical cation scavenging assay, the IC₅₀ value of magnolol was 22.3±0.5 µM, which was similar to the IC₅₀ values of the antioxidants α-tocopherol, β-estradiol and ascorbic acid (35.3±7.3, 16.9±3.2 and 31.7±6.8 µM, respectively).

Magnolol significantly reduced NO production when applied to LPS-stimulated BV2 and RAW 264.7 cells at concentrations of 10 and 50 µM, respectively (P<0.05; Fig. 1A and B). Similarly, magnolol significantly reduced TNF-α production in LPS-stimulated BV2 and RAW 264.7 cells when administered at concentrations of 10 and 50 µM, respectively (P<0.05; Fig. 1C and D). Additionally, magnolol significantly inhibited IL-6 production when administered at 1–10 µM in LPS-stimulated BV2 cells and 50 µM in LPS-stimulated RAW 264.7 cells (P<0.05; Fig. 1E and F).

No significant differences were identified in the PI uptake in cultures treated with 1–100 µM magnolol compared with the control cultures (data not shown). Pre-treatment with magnolol at 100 µM 1 h prior to OGD led to a significant reduction in cell death at 24 h compared with the control (P<0.05; Fig. 2A and B). The PI uptake was reduced by 25.0, 41.6 and 56.6% in the CA1, CA3 and DA regions, respectively compared with the OGD control. Treatment with magnolol (100 µM) at 2 or 4 h following the onset of OGD significantly reduced the OGD-induced increases in PI uptake in the CA1, CA3 and DA sub-regions of the brain compared with the control (P<0.05; Fig. 2C and D). The PI uptake was reduced by 23.8–44.1% in CA1, 17.7–33.0% in CA3 and 40.0–62.9% in DA.

A total of 7 animals (6.0%) died during the course of the experiment and were excluded. Of these animals, 3 (7.9%) were in the vehicle-injected group and 4 (5.1%) were in the magnolol-treated group. As described previously (20,21), animals subjected to transient MCA occlusion invariably exhibit spontaneous hyperthermia. The post-insult core temperatures and LCBF were not notably changed by the intravenous administration of magnolol at 0.01–5 mg/kg (data not shown). The physiological parameters blood hematocrit, arterial blood pressure, heart rate and glucose of the experimental animals were not notably different from those of the control animals over the course of the experiment (data not shown).

Pre-treatment with magnolol at doses of 0.01, 0.1 or 1 mg/kg 30 min prior to ischemia, significantly reduced the infarction size in the cortex compared with the control (P<0.05; Fig. 3A). Treatment with magnolol at 0.01, 0.1 and 1 mg/kg reduced the infarction volume in the cortex by 30.2, 24.3 and 22.7%, respectively, compared with the control. Treatment with magnolol (0.01–5 mg/kg) also significantly improved the sensory neurologic scores at 72 h post ischemic injury compared with those in the vehicle group (P<0.05; Table II).

When magnolol (1 mg/kg) was administered at 2 h post ischemic onset, the infarction volume in the striatum as a percentage of the volume of the contralateral hemisphere was significantly reduced by 43.3% at 72 h post ischemia compared with the control (P<0.05; Fig. 3B). The infarction volume in the striatum was reduced by 31.2%, and a significant improvement in the number of surviving neurons in the penumbral cortical area was also observed (both P<0.05; Fig. 3B). In addition, treatment with magnolol 2 h post ischemic onset significantly increased the neurologic and sensorimotor scores determined 3 days following the onset of ischemia, compared with the vehicle group (P<0.05; Table III).

In the control group, notably increased levels of MPO and NO_X activity were measure in the brain at 24 h following the onset of ischemia, which is indicative of neutrophil infiltration (Fig. 4A and B). Treatment with magnolol (0.3–1 mg/kg) 2 h following the onset of ischemia resulted in a significant reduction in the levels of MPO and NO_X at 24 h compared with the control group (P<0.05). MPO and NO_X were reduced by 50.4–72.5% and 40.0–56.7% respectively. No significant differences were identified in MPO and NO_X activity following treatment with magnolol at a dosage of 0.01 mg/kg.

Ischemic-reperfusion insult induced a decrease in the reduced glutathione level which was accompanied by an increase in brain malondialdehyde (MDA), indicative of lipid peroxidation, at 24 h following the onset of ischemia (Fig. 4C and D). Treatment with magnolol at 0.3 and 1 mg/kg at 2 h after the onset of ischemia significantly increased the GSH level, which provided protection for the ischemic brain when measured at 24 h following ischemia onset, compared with the control group (P<0.05; Fig. 4D); however, GSSG exhibited no significant difference between the magnolol and control groups. GSH was increased by 109.7–229.8% and MDA accumulation was reduced by 60.6–80.8% in the magnolol treatment groups compared with the vehicle-treated controls. Treatment with magnolol at 0.3 and 1 mg/kg led to significant reductions in the level of MDA 24 h following ischemia (P<0.05; Fig. 4C).

In the fluorescent images, intracellular O₂⁻ production in the perinuclear area and cytosol was detected via oxidized DHE (HEt) which appeared red and the nuclei were indicated by blue DAPI staining (Fig. 5A). The HEt signals were measured in the cytosol. They were primarily localized to the nuclei and also observed in the cytoplasm of the cells, which had a neuronal morphology. HEt signals were observed in the axon- and dendrite-like processes in the vehicle-treated control group. Treatment with magnolol (1 mg/kg) at 2 h after the onset of ischemia significantly reduced the relative intensity of the HEt signals in the axon- and dendrite-like processes and the cytosol (P<0.05). Data are expressed as the percentage of HEt-positive cells relative to the number of DAPI-stained cells. Treatment with magnolol (1 mg/kg) significantly reduced the percentage of DHE-expressing cells compared with the vehicle-treated control cells. Data are expressed as the ratio percentage of the cytosolic HEt fluorescence intensity relative to the back-ground intensity.

In the control group, the immunoreactivity of 8-OHdG was localized to the nuclei in neurons (Fig. 5B). Clear immuno-reactivity of 4-HNE was observed in the region of ischemic damage in neuronal dendrites and axons. Compared with the control, magnolol (1 mg/kg) significantly reduced the number of 4-HNE and 8-OHdG positive cells visible following immunofluorescence staining (P<0.05). The proportion of cells stained with 4-HNE was reduced by 46.5% and the proportion of cells stained with 8-OHdG was reduced by 58.0% in the ischemic brain.