Animal preparation and MIA administration
All animal experiments accorded with UK Home Office Guidelines [Animals (Scientific Procedures) Act 1986]. Eighteen large white male piglets aged < 24 h were anesthetized and surgically prepared as described previously (Lorek et al. 1994). Briefly, piglets were sedated with intramuscular midazolam (0.2 mg/kg) and arterial O2 saturation was monitored (Nonin Medical Inc, Plymouth, Minnesota, USA). Isoflurane anaesthesia [initial concentration 4% (v/v)] via a facemask facilitated tracheostomy and intubation and was maintained (3% during surgery and 2% otherwise) with continuous morphine infusion (0.05 mg/kg/h). Mechanical ventilation settings were adjusted to maintain PaO2 and PaCO2 at 8–13 kPa and 4.5–6.5 kPa respectively, allowing for temperature correction of the arterial blood sample.
An umbilical venous catheter was inserted for infusion of maintenance fluids (10% dextrose, 60 mL/kg/day), morphine (10–40 μg/kg/h) and antibiotics (benzylpenicillin 50 mg/kg and gentamicin 2.5 mg/kg, every 12 h). An umbilical arterial catheter was inserted for continuous heart rate and arterial blood pressure monitoring, and intermittent blood sampling for PaO2, PaCO2, pH, electrolytes, glucose and lactate (Abbot Laboratories, Maidenhead, UK). Bolus infusions of colloid (Gelofusin, B Braun Medical Ltd. Emmenbrucke, Switzerland) and dopamine (5–20 μg/kg/min) maintained mean arterial blood pressure (MABP) > 40 mmHg. All animals received continuous physiological monitoring and intensive life support throughout experimentation. Arterial lines were maintained by infusing 0.9% saline (Baxter, 1 mL/h): heparin sodium was added at a concentration of 1 IU/mL to prevent line blockage.
Both common carotid arteries were surgically isolated at the level of the fourth cervical vertebra and encircled by remotely controlled vascular occluders (OC2A, In Vivo Metric, Healdsburg, CA, USA). After surgery, piglets were positioned prone in a plastic pod; the scalp was firmly immobilized below the MRS surface coil and the pod was inserted into the MRS system. Piglets were maintained normothermic (38.5°C) throughout experimentation using a warmed water mattress and circulating warm air.
Whilst in the MRS system, transient HI was induced by occluding both common carotid arteries, using the inflatable vascular occluders, and reducing fractional inspired (Fi) O2 to 12% (v/v). During HI cerebral-energetic changes were observed every 2 min by 31P MRS and the β-NTP peak height was measured automatically. HI was continued until 10 min after the β-NTP peak height had fallen to 30% of baseline: during this 10 min period, FiO2 was adjusted to maintain this β-NTP level. After this 10 min, the occluders were deflated and FiO2 normalized. 31P spectra were acquired for a further hour to monitor recovery. The time integral of the reduction in β-NTP relative to the exchangeable phosphate pool [ePP = Pi + PCr + (2γ + β)−NTP] during HI and the first 60 min of resuscitation quantified acute energy depletion (insult severity, AED) as described previously (Lorek et al. 1994).
Baseline data were acquired before transient HI following stabilization of the animal in the MRS system. Following resuscitation, piglets were randomized into two groups (each n = 9): (i) saline placebo and normal care; and (ii) MIA 3 mg/kg intravenously commencing 10 min post-insult and 8 hourly thereafter and normal care. Normothermia was maintained throughout.
For intravenous administration, 100 mg MIA dissolved in 1 mL of 350 mmol/L acetic acid, and diluted with saline to give 2 mg/mL MIA. pH was checked to ensure normality. Aliquots of 6 mg in 3 mL MIA were made up and stored at −20°C. Just before use, aliquots were defrosted rapidly in a water bath to avoid crystallization.
Spectra were acquired as previously described (Faulkner et al. 2011) using a 6.5 cm × 5.5 cm elliptical transmit-receive surface-coil tunable for both 1H and 31P nuclei in a 4.7 Tesla Biospec Avance (Bruker Medizintechnik, Karlsruhe, Germany). Spectra were acquired prior to HI (after stabilization of the animal in the MRS system), during HI and 1 h recovery (31P only), and thereafter continuously about every ~5 h. During the whole experiment the piglet remained in the bore of the magnet continuously receiving intensive care support and monitoring. 1H PRESS spectra [repetition time (TR) 5 s, 128 summed transients, echo time (TE) 288 ms] were acquired from two voxel positions (Fig. 1a): deep grey matter centred on both lateral thalami and hypothalami [ventromedial forebrain (vmFB); 14 mm × 14 mm × 7 mm voxel]; and dorsal right subcortical white matter at the centrum semiovale level dorsal subcortical white matter (dscWM); 9 mm × 4 mm × 20 mm voxel). 31P MR spectra were acquired from whole brain using single-pulse acquire with TR 10 s.
Figure 1. Methyl isobutyl amiloride (MIA) reduces the cerebral metabolic changes seen by 1H magnetic resonance spectroscopy (MRS) following hypoxia-ischaemia in the neonatal piglet. (a) Coronal image of piglet brain (midparietal plane) showing the 1H MRS voxels in dorsal subcortical (dsc) white matter (WM) and in ventro medial forebrain (vmFB). 31P MRS used a surface coil therefore most signal originated from superficial brain. (b, c) 1H spectra from the WM voxel of a naive piglet (b) and 48 h after transient, global hypoxia-ischaemia (c). Note the increase in lactate (Lac) at 1.3 ppm and the decrease in N-Acetylaspartate (NAA) at 2 ppm. The 3 ppm creatine (Cr) peak remains largely unchanged. (d, e) 31P spectra from a naive piglet (d) and 33 h after hypoxia-ischaemia (e). Note the increased inorganic phosphate (Pi) and decreased phosphocreatine (PCr), phosphoethanolamine (PET) and γ, α and β adenosine triphosphate (ATP) in (e). (f–m) vmFB Lac/Cr (f, j), Cho/Cr (g, k), NAA/Cr (h, l) and Lac/NAA (i, m) log10 peak-area ratios. Individual plots are shown for each animal randomly assigned to placebo or MIA. MIA reduced the overall Lac/Cr and Lac/NAA increases for up to 48 h. (n) AUC summary [mean + standard error (SEM)] for vmFB 1H MRS (left), WM 1H MRS (center) and 31P MRS whole brain (WB; right). Open columns - placebo, filled columns - MIA-treated. Note the consistent lower AUC trend with MIA reaching significance (*p < 0.05, t-test) for vmFB Lac/Cr and Lac/NAA. NTP/ePP and pH (Pi) were borderline significant (p = 0.056 and 0.053, respectively).
Download figure to PowerPoint
1H MR spectra were analysed using LC Model (Provencher 1993); representative MR spectra are shown in Fig. 1b–e. For each spectrum the relative signal amplitudes for choline (Cho), total creatine (Cr), N acetyl aspartate (NAA) and lactate (Lac) were measured. The primary outcome measures nominated a priori were Lac/NAA and Lac/Cr. Metabolite peak area under the curve (AUC) ratios were plotted from baseline to 48 h after HI for each voxel and each subject on a logarithmic scale and analysed according to randomized group (Fig. 1f–m).
31P MR spectra were analysed by AMARES (Vanhamme et al. 1997). Signal amplitudes and chemical shifts were measured for PCr, Pi (three independent components fitted) and α-, β-, and γ-NTP; ePP was calculated. Estimations of pHi via Pi (pH (Pi)) and phosphoethanolamine [pH (PEtA)] used previously published chemical-shift titrations (Petroff and Prichard 1985; Corbett et al. 1987). pHi was also estimated using the NTP chemical shifts [pH (ATP)] using ‘magnesium and pH from ATP calculation’ (MAGPAC) (Williams and Smith 1995). Again, metabolite peak AUC ratios were plotted from baseline to 48 h after HI for each voxel and each subject on a logarithmic scale and analysed according to randomized group.
As some piglets died prematurely, we needed to calculate time-series AUC or area over the curve (AOC) consistently, therefore we used a uniform time window of 0 to 49 h post-insult with the value for the first datum extended horizontally to 0 h and that for the last datum to 49 h. The resulting area was divided by 49 h. If MRS started before 0 h, a 0 h datum was interpolated linearly from those directly before and after 0 h.
In the case of pH measures derived from animals in subgroups without secondary brain acidosis shown in Figure S2c, f and i (5/9 of the placebo group, and 7/9 of the MIA-treated group), the data samples were first tested using repeated measures anova followed by a paired t-test for individual time points against the baseline.
Forty-eight hours after HI, piglets were killed with pentobarbital and the brains fixed via cardiac perfusion with cold 4% paraformaldehyde in phosphate-buffered saline. Brains were then dissected out and post-fixed at 4°C in 2% paraformaldehyde for 7 days. Coronal slices (5 mm thick) of the right hemisphere, starting from anterior to the optic chiasma, were embedded in paraffin wax and sectioned to 5 μm thickness and stained for hematoxylin and eosin (H&E). To assess cell death and glial activation, adjacent sections were stained for nuclear DNA fragmentation using histochemistry with transferase mediated biotinylated d-UTP nick end-labelling (TUNEL), and for the appearance of cleaved caspase 3 and IBA1 immunoreactivity (Ito et al. 2001). Per animal, two sections, placed 5 mm apart, were examined with each stain.
For all histo- and immunohistochemical stains, brain sections were dehydrated in xylene (3 × 10 min) and rehydrated in graded ethanol solutions (100–70%), followed by double-distilled water. For TUNEL, sections were pre-treated for 15 min in 3% H2O2 in methanol to remove endogenous peroxidase, followed by a 15 min peptidase pre-digestion with 20 ug/mL Proteinase K (Promega, Southampton, UK) at 65°C, and then incubated at 37°C for 2 h with the TUNEL solution (Roche, Burgess Hill, UK) containing biotinylated dUTP. For immunohistochemistry, sections were processed for antigen retrieval (800 mW microwave irradiation in 0.1 M citrate buffer, 10 min), followed by overnight incubation with primary rabbit antibody against caspase 3 (1 : 500) or IBA1 (1 : 1000), and then 2 h incubation with a biotinylated, secondary goat-anti-rabbit immunoglobulin antibody (1 : 100, Jackson Labs, West Grove, PA, USA). The biotin residues were detected with the avidin-biotinylated horseradish peroxidise complex (ABC, Vector Laboratories, Peterborough, UK) and visualized with diaminobenzidine/H2O2 (Sigma, Gillingham, UK), with CoCl2 and NiCl2 included to intensify TUNEL histochemistry. The sections were counter-stained with haemotoxylin, dehydrated in graded alcohol and xylene and mounted with Depex (VWR, Leighton Buzzard, UK).
For each animal, section and region, TUNEL+ nuclei were counted in three fields (40x magnification, with an area of 0.76 mm2) and the average converted into counts/mm2. The less numerous caspase immunoreactive cells were also counted in three fields (20x magnification, 3.1 mm2). In the parasagittal cortex, midtemporal cortex (insular region) and hippocampus (dentate gyrus), the three fields were from the superficial, middle and deep cortical layers. IBA1+ microglial cell bodies and branch density were determined at 40x magnification using a 0.24 mm × 0.24 mm square grid, placed in three fields for each brain region by counting the number of cell bodies inside the grid [cell body density (CBD)] and the average number of branches (B) crossing the three horizontal (top, middle and bottom) and the three vertical (left, middle, and right) 0.24 mm long gridlines. This determined the microglial ramification index (B2/CBD).
The threshold for statistical significance was p < 0.05. For statistical analysis in individual brain regions, original counts were normalized by log(x + 1) algorithm conversion (Werner et al. 2001) and the differences between the two groups detected using one-way anova, followed by post hoc TUKEY test. Trend analysis was performed across all seven forebrain regions, using one-way anova followed by Tukey. Statistical significance of correlation R2 was assessed by the F-test.