Materials and Methods
Animal
Adult Sprague Dawley rats (MGI Cat# 5651135, RRID: MGI:5651135; sex: male; weight: 250-300 g) (Samkako, Osan, Korea) were used in this study. Animals were housed under controlled conditions (20-24°C, 55-65% humidity, 12 h light/dark cycle), and have been given free access to water and standard laboratory chow (5L79, PMI Nutrition International, St Louis, MO). All animal experiments were performed in accordance with the Guidelines and Policies for Rodent Survival Surgery provided by the Animal Care Committee of the Kyung Hee University (Permission number: KHUASP(SE)-17-059). Animal studies are reported in compliance with the ARRIVE guidelines and with the recommendations made by the British Journal of Pharmacology (Kilkenny et al., 2010).
SCI surgery
SCI was performed using New York University (NYU) impactor in accordance with previously described methods (Lee et al., 2014a). Experimental induction of a contusive SCI in a rat model using the NYU impactor device has been validated as an analog to human SCI. A comparison between the rat model of SCI and human SCI shows functional electrophysiological and morphological evidence of similar patterns recorded in motor evoked potentials and somatosensory evoked potentials as well as high-resolution magnetic resonance imaging (Basso et al., 1995). In brief, rats were anesthetized with chloral hydrate (500 mg∙kg-1) by intraperitoneal (i.p.) injection and laminectomy was performed at the T9-T10 level, exposing the spinal cord beneath without disrupting the dura. The spinous processes of T8 and T11 were then clamped to stabilize the spine, and the exposed dorsal surface of the cord was subjected to moderate contusion injury (10 g X 25 mm) at the T9-T10 level using a NYU impactor. Throughout the surgical procedure, body temperature was maintained at 37 ± 0.5°C with a heating pad (Biomed S.L., Alicante, Spain). After the injury, the muscles and skin were closed in layers, and rats were placed in a temperature and humidity-controlled chamber overnight. Postoperatively, rats were received subcutaneously supplemental fluids (5 ml, lactated ringer) and antibiotics (gentamicin, 5 mg∙kg-1, intramuscular injection) once daily for 5 d. The bladder was emptied manually three times per day until reflexive bladder emptying was established.
Drug treatment
Animals were randomly divided into three experimental groups (Sham, Vehicle, and GA-treated group) (Fig. 1). GA (Cayman Chemical, Ann Arbor, MI) was dissolved in 0.9% saline and rats were given GA (50 mg∙kg-1) immediately after SCI by i.p. injection and then received the same dose of GA at 6 h and 12 h, and then further treated once a day for 7 d. Based on our preliminary study, we found that a dose of 50 mg/kg of GA was an optimal dose for the reduction of BSCB disruption after SCI (Fig. 4D) and thus we used 50 mg/kg of GA throughout this study. Sham-operated animals received no pharmacological treatment and vehicle group received equivolumetric administration of 0.9% saline.
Tissue preparation
At indicated time points, rats were anesthetized with chloral hydrate (500 mg∙kg-1, i.p.) and perfused via cardiac puncture initially with 0.1 M PBS and subsequently with 4% paraformaldehyde in 0.1 M PBS. A 20-mm spinal cord, centered at the lesion site, was dissected out, post-fixed by immersion in the same fixative (4% paraformaldehyde) for 5 h and placed in 30% sucrose in 0.1 M PBS. The segment was embedded in OCT for frozen sections, and longitudinal or transverse sections were then cut at 10 or 20 µm on a cryostat (CM1850; Leica, Wetzlar, Germany). For molecular work, rats were perfused with 0.1 M PBS and the segments of the spinal cord (10 mm) including the lesion site were isolated and frozen at -80°C until use.
Molecular docking study
The structure used as a template for the structure-based docking study was prepared by using the X-ray structure (PDB code 4ASK) (Kruidenier et al., 2012). All the water molecules and ligand were removed, and the hydrogen atoms were added. The coordinate file for the structure of inhibitor compounds was constructed in Sybyl X-2.1.1 and energetically minimized using a Tripos force field with Gasteiger-Huckel charges. Docking was carried using Surflex-Dock GeomX module interfaced in Sybyl X-2.1.1 by generating protocol using the co-crystallized ligand in the binding site (Jain, 2003). The ligand was docked to the receptor and analyzed with the original ligand bound to the receptor to validate the method. The generated compounds were also docked into the binding site and the results were analyzed and ranked by Total Score (-logKd).
In vitro histone methylation assay
In vitro Jmjd3 inhibition assay was modified from Epigenase Jmjd3/UTX demethylase activity/Inhibition assay kit (P-3084, Epigentek, Farmingdale, NY). As briefly described, acid-extracted histones were incubated with a Jmjd3 enzyme (E24026-1, Epigentek) in an assay buffer (0.05% Tween-20, 0.05% BSA, 10 μM NH4FeSO4, 200 μM ascorbic acid, 2 μM α-ketoglutaric acid, 25 mM HEPES, pH 7.5) with each inhibitor as a proper concentration for 2 h at 37℃. The demethylation reaction was stopped by adding a 5X SDS sample buffer and boiling for 5 min. Western blotting was carried out using anti-H3K27me3 (Millipore, Cat# 07-449, RRID: AB_310624).
Endothelial cell culture and oxygen-glucose deprivation (OGD)/reperfusion
A mouse brain endothelial cell line, bEnd.3 (ATCC, Cat# CRL-2299, RRID: CVCL_0170) was cultured as previously described (Lee et al., 2012b; Lee et al., 2012c). Prior to each experiment, cells were seeded onto 6-well (5 × 105 cells/well) plates. To achieve OGD, cells were transferred to a humidified anaerobic chamber (APM-30D, Astec, Fukuoka, Japan) under an atmosphere of 0.1% O2, 5% CO2 balanced with 95% N2. The culture medium was replaced three times with deoxygenated and glucose-free DMEM. The cells were treated with GA (10 μM) for 30 min before OGD. At the end of the OGD period, cells were placed under normoxic conditions and the media was quickly replaced with 25 mM glucose containing DMEM. Control cells were cultured in DMEM with 25 mM glucose under normoxia. GA (10 μM) was dissolved in 0.9% saline, which had no effect on cell viability.
Chromatin immunoprecipitation (ChIP)
ChIP assay was performed as previously described (Lee et al., 2016; Lee et al., 2012c). The primary antibodies used in ChIP assay are as follows: Jmjd3 (Abgent, Cat# AP1022a, RRID: AB_889281), H3k27me3 (Millipore), normal IgG (Santa Cruz Biotechnology, Cat# sc-2027, RRID: AB_737197). Cells were crosslinked using 1% formaldehyde in PBS for 15 min at room temperature. To stop the cross-linking, 1.25 M glycine was added (1% final concentration). After washing with ice-cold PBS, the cells in SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl, pH 8.1) was sonicated until the DNA fragments were 300-500 bp in size. The extracts were subsequently centrifuged, and the resulting soluble chromatin solutions were diluted 10 fold with ChIP dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 167 mM NaCl, 16.7 mM Tris-HCl, pH 8.1). The specific antibodies or IgG were added into soluble chromatin solution for overnight at 4 °C and followed by protein A-sepharose beads (Sigma-Aldrich, St. Louis, MO) for 2 h. The beads were extensively washed with low salt wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 150 mM NaCl, 20 mM Tris-HCl, pH8.1), high salt wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 500 mM NaCl, 20 mM Tris-HCl, pH 8.1), LiCl wash buffer (0.25 M LiCl, 1% NP-40, 1% deoxycholate, 1 mM EDTA, 10 mM Tris-HCl, pH 8.1), and finally with TE buffer. After elution of the DNA/protein with 1% SDS, crosslinking was reversed for 6 h at 65 °C. The DNA was recovered using a QIAquick spin column (Qiagen, Valencia, CA). Real time-PCR was performed with a Stratagene Mx3000P using primers that cover mouse,Mmp-3 (5′-TTC CGC CTT TTT TGT TCA-3′ and 5′-CCA CTC AAA AAC AGG TCT ATA ATT T-3′), and Mmp-9 (5′-CCC AGG CTC ATC TTT CCT TCC CC-3′ and 5′-CCC ATC CCC ACA CTG TAG GTT C-3′). The relative proportions of immunoprecipitated fragments were determined using the ΔCt comparative method based on the threshold cycle (Ct) value for each PCR reaction and normalized to input genomic DNA.
Evans blue assay
BSCB permeability was investigated with Evans Blue dye extravasation as previously described (Lee et al., 2012a). In brief, 5 ml of 2% Evans blue dye (Sigma) was administered via i.p. injection at 24 h after SCI and then perfused with 0.1M PBS at 3 h later. The spinal cords (5 mm) including lesion epicenter were homogenized and the fluorescence intensity was measured on a Gemini™ XPS and EM Microplate Readers (Molecular device, Sunnyvale, CA) at 620 nm excitation/680 nm emission. The Dye in samples was determined as micrograms per gram of tissue from a standard curve plotted using known amounts of dye.
Immunohistochemistry
Frozen sections were processed for immunohistochemistry with antibodies against myeloperoxidase (MPO; 1:100; Agilent, Cat# A0398, RRID: AB_2335676), ED-1 (1:1,000; Bio-Rad, Cat# MCA341R, RRID: AB_2291300), cleaved caspase-3 (1:100; Cell Signaling Technology, Cat# 9661, RRID: AB_2341188), CC1 (1:100; Abcam,, Cat# ab16794, RRID: AB_443473), Jmjd3 (1:100; Abcam, Cat# ab38113, RRID: AB_943898) and RECA1 (1:100; Bio-Rad, Cat# MCA970GA, RRID: AB_567193) as previously described (Lee et al., 2014b). For quantification of MPO or ED-1 intensity, serial transverse sections (20 μm thickness) were collected every 100 μm section rostral and caudal 3,000 μm to the lesion site (total 60 sections). Digital images of MPO- or ED-1-stained tissues were obtained and quantified the entire fluorescent intensity of each transverse section above the threshold by using MetaMorph software (Molecular devices) and averaged. The threshold value was at least three times the background and the backgrounds were quantified and normalized to the primary antibody omitted control. Immunostaining control studies were performed by omission of the primary antibodies, by replacement primary antibodies with non-immune, control antibody, and by pre-absorption with an excess (10 µg/ml) of the respective antigens. For double labeling, fluorescein isothiocyanate - or cyanin 3-conjugated secondary antibodies (Jackson ImmunoResearch Labs, Cat# 111-165-003, RRID: AB_2338000) was used. Also, nuclei were labeled with DAPI according to the protocol of the manufacturer (Thermo Fisher Scientific, Cat# D3571, RRID: AB_2307445). In all controls, reaction to the substrate was absent if the primary antibody was omitted or if the primary antibody was replaced by a nonimmune control antibody. For quantification of cleaved caspase-3-positive oligodendrocytes (cleaved caspase-3/CC1 double positive), serial transverse sections (10 μm thickness) were collected every 200 μm from 4,000 μm rostral to 4,000 μm caudal to the lesion site (total 40 sections). Cleaved caspase-3-positive oligodendrocytes in the white matter (WM) in each section were counted and averaged. Serial sections were also stained with Cresyl violet acetate for histological analysis.
Cell counting of viable ventral motor neurons (VMNs)
One day after injury, the number of VMNs was counted and assessed. The criteria for VMN counting was based on the previous report (Lee et al., 2014b). In brief, we determined the cells located in the lower ventral horn and were larger than half of the sampling square (20 x 20 μm) as a VMN. The cells above the line at 150 μm ventral from the central canal were excluded. For counting of the total number of VMN, serial transverse sections (20 µm thickness) were collected every millimeter section rostral and caudal 8 mm to the lesion site and stained with Cresyl violet acetate. Motor neurons were manually counted from each field and analyzed by MetaMorph software (Molecular Devices).
TUNEL staining
One and five days after injury, serial spinal cord sections (20 μm thickness) were collected every 200 μm and processed for TUNEL staining using an ApopTag in situ kit (Millipore, Cat# S7100, RRID: AB_2661855), according to the manufacturer’s instructions. A DAB substrate kit (Vector Laboratories, Burlingame, CA) was used for peroxidase staining. Control sections were treated similarly, but incubated in the absence of TDT enzyme, dUTP-digoxigenin, or anti-digoxigenin Ab, and positive control sections were incubated in DNase 1. TUNEL-positive cells in the gray matter (GM) at 1 d (total 40 sections) and in the WM at 5 d (total 100 sections) after SCI were counted and quantified using a 20 x objective. Only those cells showing morphological features of nuclear condensation and/or compartmentalization in the GM and WM were counted as a TUNEL-positive cell.
RNA isolation and RT-PCR
Total RNA at the indicated time points was isolated from spinal cord segments (10 mm), centered at the lesion site by using TRIZOL reagent (Invitrogen) and RT-PCR were performed as previously described (Lee et al., 2012b; Lee et al., 2003; Yune et al., 2007). The primers used for RT-PCR were synthesized by Genotech (Daejeon, Korea), and the sequences of all primers are presented in Table 1. After amplification, PCR products were subjected to a 1.5 or 3 % agarose gel electrophoresis and visualized by ethidium bromide staining. The relative density of bands (relative to sham value) was analyzed by the AlphaImager software (Alpha Innotech Corporation, San Leandro, CA). The expression of GAPDH was used as an internal control. Experiments were repeated three times and the values obtained for the relative intensity were subjected to statistical analysis. The gels shown in figures are representative of results from three separate experiments.
Western blot
Total protein preparation at the indicated time points and Western blot analysis were performed as previously described (Lee et al., 2012a). Protein sample (30 µg) was separated on SDS-PAGE gel electrophoresis and transferred to nitrocellulose membrane (Millipore). The membranes were blocked in 5% non-fat skim milk or 5% bovine serum albumin in Tris-buffered saline containing tween-20 (0.1%), and subsequently incubated with antibodies. The primary antibodies used in Western blot are as follows. Jmjd3 (1:1,000; Abcam), H3K27me3 (1:1,000; Abcam, Cat# ab6002, RRID:AB_305237), Histon H3 (1:1,000; Cell Signaling Technology, Cat# 9715, RRID: AB_331563), ED-1 (1:1,000; Bio-Rad), iNOS (1:1,000; BD Biosciences, Cat# 610333, RRID: AB_397723), COX-2 (1:1,000; Cayman Chemical, Cat# 160107, RRID: AB_10078833), cleaved caspase-3 (1:1,000; Cell Signaling), ZO-1 (1:1,000; Thermo Fisher Scientific, Cat# 40-2200, RRID: AB_2533456), and occludin (1:1,000; Thermo Fisher Scientific, Cat# 40-4700, RRID: AB_2533468). β-tubulin (1:30,000; Sigma-Aldrich, Cat# T4026, RRID: AB_477577) was used as a loading control. The primary antibody was detected with a horseradish peroxidase-conjugated secondary antibody (Jackson ImmunoResearch). Immunoreactive bands were visualized by chemiluminescence using Supersignal (Thermo Scientific, Rockford, IL). The densitometric values of the bands on Western blots were obtained by AlphaImager software (Alpha Innotech Corporation) and subjected to statistical analysis. Background in films was subtracted from the optical density measurements.
Gelatin zymography
The activities of MMP-2 and MMP-9 by gelatin zymography were performed using total protein (50 µg) at 1 d after injury as previously described (Lee et al., 2012b). Total protein was loaded on a Novex 10% zymogram gel (EC61752; Invitrogen) and separated by electrophoresis. The gel was then incubated with renaturing buffer (2.5% Triton X-100) at room temperature for 30 min to restore the gelatinolytic activity of the proteins. After incubation with developing buffer (50 mM Tris-HCl, pH 8.5, 0.2 M NaCl, 5 mM CaCl2, 0.02% Brii35) at 37°C for 24 h, the gel was stained with 0.5% Coomassie and then destained with 40% methanol containing 10% acetic acid until appropriate color contrast was achieved. The clear bands on the zymogram were indicative of gelatinase activity. Quantification of bands was performed by AlphaImager software (Alpha Innotech Corporation). Experiments were repeated three times and the values obtained for the relative intensity were subjected to statistical analysis.
Behavioral tests
The examination of functional deficits after injury was conducted as previously described (Basso et al., 1995; Lee et al., 2003; Rivlin & Tator, 1977; Yune et al., 2007). Behavioral analyses were performed by trained investigators who were blind as to the experimental conditions. Open-field locomotion to test hindlimb locomotor function was evaluated by using the Basso-Beattie-Bresnahan (BBB) locomotion scale. BBB is a 22-point scale (scores 0-21) that systematically and logically follows recovery of hindlimb function from a score of 0, indicative of no observed hindlimb movements, to a score of 21, representative of a normal ambulating rodent. For the inclined plane test, rats were tested in two positions (right side or left side up) on the testing apparatus. The maximum angle at which a rat could maintain its position for 5 s without falling was recorded for each position and averaged to obtain a single score for each animal. The ability to control and place the hindlimb precisely was tested on a horizontal grid and the analysis was performed by counting the number of footfall (mistake) in foot placing. For the footprint analysis, both forepaws and hind paws were dipped in red and blue dye (nontoxic) and then walked across a narrow box (1 m long and 7 cm wide). The footprints were then scanned, and digitized images were analyzed.
Axon counting and myelin staining
After behavioral tests, rats treated with vehicle and GA were perfused at 35 d after injury, and frozen sections were prepared as described above. For quantitative analysis of axonal density, serial coronal sections collected every millimeter rostral and caudal 5 mm to the lesion site were stained with an antibody specific for 200 kDa neurofilament protein (NF200, 1:4,000; Sigma-Aldrich, Cat# N4142, RRID: AB_477272). Some sections were processed for 5-hydroxytryptamine (5-HT, 1:5,000; ImmunoStar, Cat# 20080, RRID: AB_572263 N) staining. The ABC method was used to detect labeled cells using a Vectastain kit (Vector Laboratories). Axonal densities were determined within preselected fields (40 × 40 μm, 1,600 μm2) at specific sites within the ventral and dorsolateral funiculi as previously described (Yune et al., 2007). Axons were manually counted from each field and analyzed by MetaMorph software (Molecular Devices). For myelin staining, selected slides were incubated in 0.1% Luxol fast blue (Solvent Blue 38; Sigma) in acidified 95% ethanol overnight at 60°C. Differentiation was performed with 0.05% lithium carbonate as previous described (Lee et al., 2018). Digital images of Luxol fast blue-stained tissues were obtained by MetaMorph software (Molecular Devices).
Assessment of lesion volume
The measurement of lesion volume using rats tested for behavioral analyses was performed as previously described (Yune et al., 2008). Serial longitudinal sections (10 μm) through the dorsoventral axis of the spinal cord were used to measure the lesion volume. Every 50 μm section was stained with Cresyl violet acetate. The lesion volume was determined by measuring the area of cavitation at the injury epicenter using a low-power (1.25 x) objective and then calculated by mean of a MetaMorph software (Molecular Devices). Areas of each longitudinal level are determined, and the total lesion volume was induced by numerical integration of sequential areas.
Statistical analysis
The biochemical assays and analysis of the results was carried out without knowledge of the treatment groups (blinded). Each experiment involved at least 5 independent samples (equal size) per randomized group, and that statistical analysis was done using these independent values. The statistical analysis was undertaken only for studies where each group size was at least n = 5. The control and test values in Western blot and RT-PCR were normalized to an internal standard (such as β-Tubulin and GAPDH) to reduce variance. Western blot and RT-PCR were normalized to the mean value of the experimental control group to set Y-axis so the control group value was 1. The units for these normalized data in the Y-axis were the fold of the control group’s mean value. All statistical analyses were performed by SPSS 15.0 (SPSS Science, Chicago, IL). Western blot and RT-PCR presented as the mean ± SD values and Evans blue, BBB, grid walk and inclined plane test data are presented as the mean ± SEM. Comparisons between vehicle and GA-treated groups were made by unpaired Student’s t test. Multiple comparisons between groups were performed one-way ANOVA. Behavioral scores from BBB and inclined plane test were analyzed by repeated measures ANOVA (time vs treatment). Tukey’s multiple comparison was used as Post hoc analysis. For each parameter of the data presented, * indicates p < 0.05.