Inhibiting HMGB1 Reduces Cerebral Ischemia Reperfusion Injury in Diabetic Mice

Male C57BL/6 mice, 5–6 weeks of age, were housed in the Neurosurgical Research Center of Beijing Military Animal Center. All mice used in this study were handled according to the Center’s Health Guide for the Care and Use of Laboratory Animals.

The anti-HMGB1 polyclonal antibody (neutralizing antibody) and chicken IgY (isotype negative control antibody) were obtained from Tecan (Shanghai) Trading Co.,Ltd (Shanghai, China) [20]. Mice were injected intraperitoneally with 600 ug per mouse anti-HMGB1 polyclonal antibody or control IgY 1 h before ischemia as previously described [21].

Type 2 diabetes were induced in mice by a 3-week high-fat diet (DIO Rodent Purified Diet; TestDiet, Richmond, IN, USA, containing 61.6 % fat, 3140 Kcal), followed by an intraperitoneal (i.p.) injection with 20 mg/kg body weight streptozotocin (Sigma Aldrich, St. Louis, MO, USA) dissolved in saline [22, 23], then one additional week of high-fat feeding. Vehicle i.p. injections were administered to control mice in combination with a normal diet (LabDiet 5010, 5.5 % fat). Plasma glucose concentrations were measured using a blood glucose test meter in blood samples collected from the tail vein of mice at different time points (0, 2, 4, 6, and 8 weeks) after the induction of diabetes.

To achieve transient focal cerebral ischemia, we performed MCAO according to a modified intraluminal filament method as previously described [24]. In brief, mice were anesthetized by i.p. injection of ketamine and xylene, and a homeothermal blanket was used to maintain a rectal temperature of 36–37 °C. A vertical incision was made in the middle of the neck to expose the left common carotid artery, left internal carotid artery (ICA), left external carotid artery (ECA), and left pterygopalatine artery. The left ECA and the left pterygopalatine artery were ligated with a 5.0 silk suture and a small clip was used to occlude the left ICA at the bifurcation of the ICA and the pterygopalatine artery. A 6.0 nylon monofilament (0.2–0.22 mm) was inserted into the left ECA immediately after cutting with ophthalmic scissors. After removing the clip, the nylon monofilament was slowly pushed into the distal end of the middle cerebral artery. After 1 h of occlusion, the nylon thread was removed to initiate reperfusion. The arteries were surgically exposed but not occluded in the sham group. After 1 h of reperfusion, mice were sacrificed and brains were removed and processed for hematoxylin and eosin (H&E) staining and real-time PCR analysis.

Blood samples were collected from the mice after experiments using the heart puncture procedure (without anticoagulants). Serum samples were collected by centrifugation of blood samples at 3000 rpm for 10 min.

Brain tissue was fixed in formaldehyde and embedded in paraffin wax. Five-micrometer slices were prepared and H&E staining was performed in the pathology lab at St. Michael’s hospital.

Serum HMGB1 levels were measured using an enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems). Briefly, 96-well ELISA plates were coated with HMGB1 mAbs (Cat. NO. ab12029; Abcam, Cambridge, MA, USA) and serum and standard protein samples were diluted 2–3 times in succession. Biotinylated antibodies were added followed by avidin-conjugated horseradish peroxidase. Tetramethylbenzadine substrate solution was added and the plate was read at 450 nm after 30 min. Serum HMGB1 concentrations were calculated from the standard curve.

To determine changes in vascular permeability, a 4 % solution of Evans blue dye (Urchem, Shanghai, China) was injected intravenously into the tail vein of mice after reperfusion. Three hours later, mice were perfused with 150 ml of saline solution into the left ventricle of the heart and the brain was immediately removed and dissected into hemispheres. The weight of the left hemisphere was recorded. Fifty percent trichloroacetic acid was added to centrifuged brain tissues to extract the Evans blue dye. After centrifugation, the supernatant solution was diluted 1:3 in ethanol and the absorbance was determined at 620 nm. The concentration of Evans blue was calculated from a standard curve and the data were expressed as nanograms per gram of Evans blue.

Total RNA was extracted using Trizol reagent (Invitrogen Canada, Burlington, Canada) according to the manufacturer’s instructions. SuperScript II reverse transcriptase (Invitrogen Canada) was used to reverse transcribe the RNA and the PCR reaction was performed using the SYBR® green system (Applied Biosystems, Foster City, CA, USA). Reactions were monitored on an ABI Prism SDS 7000 (Thermo Scientific, Waltham, MA, USA) machine and results were analyzed with SDS 2.0 software.

The housekeeping gene HPRT (hypoxanthine-guanine phosphoribosyl transferase) was used as an internal control. Primers (Sigma-Aldrich) used were as follows: HPRT (accession number: NM_ J00423): left: 5′-caagcttgctggtgaaaagga-3′, right: 5′-tgaagtactcattatagtcaagggcatatc-3′; IL-1β (accession number: NM_ M15131): left: 5′-gtggaacttgaggccacatt-3′, right: 5′-tgtgacaaaaatgcctggaa-3′; iNOS (accession number: NM_ BC062378): left: 5′-caccttggagttcacccagt-3′, right: 5′-accactcgtacttgggatgc-3′, and IL-6 (accession number: NM_ M24221): left: 5′-ccggagaggagacttcacag-3′, right: 5′-tccacgatttcccagagaac-3′, with the following primer cycling conditions: 95 °C for 15 s, 58 °C for 50 s, and 72 °C for 15 s (40 cycles).

Data analysis was performed using SPSS 13.0 (SPSS In, Chicago, IL, USA). Values are presented as mean ± standard deviation (SD). To compare between two different groups, a Student 2-tailed unpaired t test was applied. p values <0.05 were considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0. 001, ****p < 0.0001).