Near-infrared fluorescence imaging with fluorescently labeled albumin: A novel method for non-invasive optical imaging of blood–brain barrier impairment after focal cerebral ischemia in mice

Abstract

Impairment of the blood–brain barrier (BBB) after cerebral ischemia leads to extravasation of plasma constituents into the brain parenchyma. We describe a novel method using non-invasive near-infrared fluorescence (NIRF) imaging and bovine serum albumin labeled with a NIRF dye (NIRF–BSA) to detect BBB impairment after middle cerebral artery occlusion (MCAO) in mice. We first explored the time course of BBB impairment after transient MCAO using Evans blue (EB), which binds to plasma albumin in vivo. An initial BBB impairment was observed at 4–8 h and a second impairment at 12–16 h after reperfusion. No EB extravasation was detected at 8–12 h. Non-invasive NIRF imaging with NIRF–BSA confirmed biphasic BBB impairment. Upon co-injection of NIRF–BSA with EB we found a strong correlation between the detected NIRF signal and the amount of extravasated EB (r = 0.857, P = 0.00178). When MCAO mice received NIRF–BSA together with gadolinium–diethylene triamine penta-acetic acid (Gd–DTPA), T1-weighted images showed Gd–DTPA enhancement at all times while NIRF imaging showed biphasic BBB impairment. In conclusion, NIRF–BSA is a suitable marker of plasma albumin extravasation in the mouse brain. Non-invasive NIRF imaging with NIRF–BSA is a useful tool to study BBB integrity in preclinical models of central nervous system pathology.

Introduction

Cerebral ischemia causes dysfunction of the tight and adherens junctions of the cerebral endothelium (Petty and Lo, 2002). This leads to an impairment of blood–brain barrier (BBB) integrity and thereby to extravasation of plasma constituents and cells into the brain parenchyma. BBB impairment has been implicated in edema formation and hemorrhagic transformation (Wang et al., 2004). While an impaired BBB might expose the brain to possibly harmful substances, it might also provide a route for the delivery of putative therapeutic or diagnostic agents (Abbott et al., 1999, Pardridge, 2002). Therefore, knowledge about the temporal pattern of BBB impairment after cerebral ischemia would be highly relevant in experimental and clinical studies.

BBB impairment has been characterized with fluorescent and radiolabeled markers in animal models of cerebral ischemia (Kuroiwa et al., 1985, Preston and Webster, 2002, Knight et al., 2005, Nagaraja et al., 2008a). Those markers do not pass the BBB under physiological conditions, but extravasate when the BBB is impaired. For that purpose, Evans blue (EB), which binds to plasma albumin when administered intravenously, has been extensively employed (Rawson, 1943, Kuroiwa et al., 1985, Belayev et al., 1996, Del Valle et al., 2008). EB extravasation was observed to follow a biphasic pattern (Kuroiwa et al., 1985, Belayev et al., 1996). However, the use of EB requires sacrificing the animal under study.

Contrast-enhanced magnetic resonance imaging (MRI) with gadolinium–diethylene triamine penta-acetic acid (Gd–DTPA) has been employed to assess BBB impairment in vivo (Lo et al., 1994, Neumann-Haefelin et al., 2000, Jiang et al., 2005, Nagaraja et al., 2008b, Strbian et al., 2008). However, differences in the spatial extravasation pattern of Gd–DTPA and albumin-tagged EB were reported (Nagaraja et al., 2008b, Strbian et al., 2008). This might be explained by the lower molecular weight of Gd–DTPA, which has no binding affinity to plasma albumin, compared to the albumin-tagged EB (≈550 Da vs. ≈68,000 Da) (McMurry et al., 2002). It has been implied that due to its low molecular weight, Gd–DTPA might not be a suitable marker to detect extravasation of plasma constituents of larger molecular weight or to predict areas of subsequent hemorrhagic transformation (Nagaraja et al., 2008b).

Therefore, we investigated the use of bovine serum albumin covalently labeled with a NIRF dye (NIRF–BSA; molecular weight of 70,800 Da) for detection of BBB impairment in a mouse model of focal cerebral ischemia. Near-infrared fluorescence (NIRF) imaging techniques are becoming increasingly important tools to study animal models of cerebral ischemia (Klohs et al., 2006, Klohs et al., 2008, Bourayou et al., 2008). NIRF imaging permits the sensitive detection of fluorescent probes in vivo and facilitates simple, inexpensive and high-throughput imaging of animals. In this study, we tested whether non-invasive detection of BBB impairment with planar NIRF imaging using NIRF–BSA is feasible and compared the temporal pattern of NIRF–BSA extravasation with EB histology and contrast-enhanced MRI with Gd–DTPA..