Fluorescence imaging of blood–brain barrier disruption

Pathological alterations of the blood–brain barrier (BBB) can be topographically heterogeneous. The goal of this study was to develop a method to assess rapidly the magnitude and spatial distribution of permeability changes. Rats were perfused via the common carotid arteries with Ringer's solution containing sodium fluorescein (NF) and Evans Blue albumin (EB). Global NF uptake was determined by fluorimetry and EB uptake was determined by absorbance spectroscopy. NF uptake was linear in control animals and at a rate comparable to sucrose, whereas uptake of EB was negligible. Infusion of 1.6 M mannitol immediately prior to perfusion significantly increased uptake of NF while EB uptake was unchanged. BBB disruption was confirmed by confocal microscopy of fresh-frozen sections. In control animals, NF and EB staining were limited to the edges of slices and to the circumventricular organs. In mannitol-treated animals, heavy NF staining was observed throughout the brain, and EB staining was localized around some microvessels. In animals given a ∼500 μl air embolus prior to perfusion, a discrete area of NF and EB staining could be observed near the ventral midline, while the rest of the brain remained unaltered. We find that brain perfusion with NF/EB enables a rapid, reliable, and highly sensitive assessment of global BBB permeability and microscopic visualization of discrete BBB disruptions.