Optical imaging to map blood-brain barrier leakage

Abstract

Vascular leakage in the brain is a major complication associated with brain injuries and certain pathological conditions due to disruption of the blood-brain barrier (BBB). We have developed an optical imaging method, based on excitation and emission spectra of Evans Blue dye, that is >1000-fold more sensitive than conventional ultraviolet spectrophotometry. We used a rat thromboembolic stroke model to validate the usefulness of our method for vascular leakage. Optical imaging data show that vascular leakage varies in different areas of the post-stroke brain and that administering tissue plasminogen activator causes further leakage. The new method is quantitative, simple to use, requires no tissue processing, and can map the degree of vascular leakage in different brain locations. The high sensitivity of our method could potentially provide new opportunities to study BBB leakage in different pathological conditions and to test the efficacy of various therapeutic strategies to protect the BBB.

Figure 1. Plots for Evans Blue dye.

(a) Changes in optical signal and UV absorbance with EB amount. For optical signal, different amounts of EB dye were loaded onto filter discs. (b) Standard plot for EB dye by optical imaging which is the linear range of the plot “a”. Discs exhibited different colors: red indicated the disc loaded with the highest amount of EB dye, blue the disc loaded with the lowest amount. (c) Standard curve for EB dye by UV absorbance, which is the linear range of the plot “a”. Data are shown as mean ± s.e.m., n = 4.

Figure 2. Vascular leakage in stroke.

(a) Photographic image of a stroke brain harvested 2 hrs following EB injection which is 5 hr after induction of stroke. Arrow indicates the infarcted side of the brain. (b) Optical imaging of whole brain, areas of leakage is evident from color codes. (c) Optical imaging of brain slices that show vascular leakage. Areas with more leakage show as red, those with less leakage as blue. A filter paper disc loaded with 20 ng EB dye was used as an internal standard. (d) Photographic images of brain slices showing blue color of EB.

Figure 3. Quantification of optical signals and EB dye in brain slices from a stroke animal.

(a) Standard plot for EB amount in brain tissue homogenate using optical imaging. Total signal was normalized to total pixel area for each disk to obtain signal intensity per pixel which is directly correlated to EB amount of the respective disc. (b) Standard plot for EB amount in brain tissue homogenate using UV absorbance. (c) Correlation between EB amount determined by optical and UV absorbance. (d) Optical image of a brain slice from a stroke animal. Circles marked 1–4 of the above brain slices show different color codes depending upon vascular leakage. (e) Quantification of optical signal intensity corresponding to different color regions.

Figure 4. Effect of t-PA on vascular leakage.

Normal animals received t-PA either by IV injection via tail vein or intraarterially through the carotid artery. (a) Control animals underwent no surgical procedure except that required for tail vein injection of EB dye. (b) Sham control animals underwent a similar surgical procedure as that used for stroke induction, but saline was injected instead of clot. (c) Animals with t-PA (2 mg/kg) administered via tail vein. (d) Animals with t-PA (2 mg/kg) administered via intracarotid injection. (e) Animals with t-PA (4 mg/kg) administered via intracarotid injection; (f) Quantification of total signal intensities of brain slices from above experiments (a to e). Cumulative signal from all the brain slices was normalized to pixel area. A filter paper disc loaded with 20 ng EB dye was used as an internal standard. Data are shown as mean ± s.e.m., n = 4. *p = 0.03, control vs. intracarotid injection of 2 mg/kg t-PA; **p = 0.0001, control vs. intracarotid injection of 4 mg/kg t-PA; ***p, intracarotid injection of 2 mg/kg t-PA vs. 4 mg/kg t-PA. No statistically significant difference between control and sham control (p = 0.82) or control and animal injected with 2 mg/kg t-PA via tail vein (p = 0.54).

Figure 5. Effect of t-PA on vascular leakage in animals with stroke.

(a) Control animals without stroke but received EB. (b) Animals with stroke, imaged 5 hrs after induction of stroke and 2 hrs after EB injection. (c) Animals with stroke that received t-PA (2 mg/kg) administered via intracarotid injection. (d) Quantification of optical signal of all the brain slices from the above groups and normal animal (without stroke, data from Fig. 4d) which received the same dose of t-PA via carotid artery. Data shown are the cumulative signal from all the brain slices normalized to pixel area. A filter paper disc loaded with 20 ng EB dye as an internal standard. Data are shown as mean ± s.e.m., n = 3. *p = 0.036, control vs. stroke; **p = 0.001, control vs. stroke + t-PA (2 mg/kg) administered through intracarotid route; ***p = 0.018, stroke vs. stroke + t-PA (2 mg/kg) administered through intracarotid route; ****p = 0.04 control + t-PA vs. stroke + t-PA (2 mg/kg) administered through intracarotid route.