Poster (nur Text)
Intravital fluorescence microscopy has been used widely by us (Wahl M et al.: Int. J. Microcirc. Clin. Exp. 4, 3-18, 1985) and others (Mayhan WG and Heistad DD: Amer. J. Physiol. 248, H712-H718, 1985) to study blood-brain barrier (BBB) permeability. Quantification of tracer extravasation, however, is difficult because counting leaky spots provides only a semiquantitative estimation, and measuring the tracer clearance rate may involve significant error from contamination from non-cerebral vessels (e.g. dura or skull bone). We have therefore developed a new method using computerised image analysis for quantification of tracer extravasation from cerebral blood vessels in vivo.
MATERIALS AND METHODS
- Experiments were performed on anaesthetized (a-chloralose, 40-50 mg/kg iv) and immobilised (pancuronium, 0.7 mg/kg/h iv) cats artificially ventilated with controlled blood pressure, body temperature, endtidal CO2- content and arterial acid-base status.
- A left-sided craniotomy (approximately 1.5 x 2.5 cm) was performed and the dura slit under a layer of paraffin oil to expose the cerebral cortex and the superficial extraparenchymal (pial) vasculature.
- FITC-dextran (MW 70,000) was used as fluorescent tracer. Bolus injection of 1.3 ml/kg of a 5% solution was followed by continuous infusion (0.85 ml/kg/h of a 1.25% solution).
- Predetermined fields (usually 4-6) of the exposed cerebral cortex were illuminated intermittently (triggered by the respirator) during inspiratory plateaus. The duration of the plateau was 30% of the ventilation cycle.
- The experimental setup and hardware configuration are shown schematically in Fig. 1.
DESCRIPTION OF IMAGE ANALYSIS PROCEDURE (FIG. 2)
Subtraction images were generated to eliminate the pattern of the pial vasculature and to correct for variations of intensity between individual fields. Generation of subtraction images required the following procedures:
- Digitisation and averaging of consecutive frames (normally 10) of each recording period (Fig. 3, top row)
- Shading correction of the averaged images for inhomogeneity of excitation light
- Correction for displacement artefacts in x- and y-directions of all images of an experimental series relative to the image taken from the first recording period. Subtraction images taken from two fields of observation without (left column) and with tracer extravasation (right column) are shown in the middle row of Fig. 3.
Reduction of structural artefacts in the subtraction images due to dilatation or movement of pial blood vessels was achieved by the following procedures:
- Median filtering (kernel size: 3 x 3) to reduce noise
- Pixels with an intensity lower or eqal 5 were set to 0 to further reduce the number of areas within the subtraction image
- The lengths of the remaining areas were measured in the vertical, horizontal and both diagonals, the median length obtained from a histogram and all lengths smaller than the median length deleted. Images after the reduction of structural artifacts are shown in the bottom row of Fig. 3.
After performing the steps described above the mean intensity of the remaining pixels was calculated for each subtraction image to give the degree of extravasation.
DETERMINATION OF THE DEGREE OF EXTRAVASATION UNDER DIFFERENT CONDITIONS
- Marked dilatation of pial arteries induced by 1 h cortical superfusion with histamine was found not to be accompanied by an increase in the degree of extravasation (Fig. 4). Similar results were found during mild hypercapnia by increasing FI (results not shown).
- Cortical superfusion with high concentrations of nicotine (0.1 and 1.0 mM) did not induce any significant vasomotor responses but increased the degree of extravasation (Schilling L et al.: Clin. Investig. 70, 210-217, 1992). Results obtained in one experiment are shown in Fig. 5. The results of a solvent control experiment are shown in Fig. 6 and demonstrate that spontaneous extravasation did not occur during the observation period.
- We have developed a new method combining intravital fluorescence microscopy with computerised image anlysis to follow an increase in BBB permeability. This approach allows quantification of tracer leakage by taking into account both the size and the intensity of leaky spots.
- Dilatation of blood vessels alone does not alter the degree of extravasation as calculated in the present study.
- Our new method may offer advantages when the mechanism(s) of increased BBB permeability are studied under different experimental conditions or by employing specific receptor or second messenger antagonists.
Zurück zum Abstract
Axel Findling, 1995-09-16
letzte Änderung: 1995-09-16