This study describes the application of prospective in vivo optical imaging protocols to study brain accumulation of systemically injected Aβ peptides in wild-type and animals deficient in specific transporters previously implicated in Aβ transport across the blood-brain barrier.
Radio-labeled [125I]-or [3H]-Aβ peptides have been used to study their BBB transport in animal models. The labelled peptides are either injected intravenously to analyze brain uptake or intra-cerebrally to investigate their clearance from the brain; animals are sacrificed at different time points and the radioactivity is determined in desired compartments. In vivo molecular imaging approaches that ‘track’ Aβ peptides non-invasively are dynamic methods that can be used for assessing Aβ levels in response to treatments. Notably, PET imaging with [C11]-PiB [N-methyl-[11C]2-(4-methylaminophenyl)-6-hydroxybenzothiazole] has been used for quantitative assessment of brain Aβ load in Alzheimer’s patients and in APP/PS1 mouse. Apart from requiring ‘on-site’ radioisotope labeling and access to expensive PET equipment, this approach is not applicable for ‘tracking’ peripheral Aβ peptides. Optical molecular imaging/tracking of Aβ peptides functionalized with the near-infrared imaging tracer is a viable alternative that can provide high sensitivity in experimental setting, although it does not have the quantification capabilities of PET. Among in vivo optical imaging systems, time-domain optical imaging has a clear advantage over Continuous Wavelength (CW) systems in that its pulsed laser source can penetrate skull to excite the fluorescent tracer in deep tissues. In contrast to CW systems where emitted light is collected by a CCD camera that cannot resolve the depth of the signal, with time-resolved imaging platform each collected photon retains time-of-flight distribution (also called a Temporal Point Spread Function or TPSF) from which depth (optical tomography), fluorescence concentration and fluorescence lifetime can be extracted[24–27]. This and other studies[35, 36] have shown that this imaging method is a useful non-invasive approach to investigate Aβ transport, distribution, and clearance from the brain that complements other imaging approaches.
The aberrant transport and clearance of Aβ peptides across the BBB, mediated by a spectrum of receptors and transporters including RAGE, LRP-1, and members of ABC family, contributes to Aβ accumulation in the brain and in the cerebral vasculature[7, 37, 38]. ABC family members MDR-1 P-glycoprotein/ABCB1 and ABCG2/BCRP are two major drug efflux transporters located at the luminal surface of the BBB[39, 40]. In mice, mdr-1a (Abcb1a) is the primary drug efflux transporter expressed at the BBB; while mdr-1b (Abcb1b) is the main isoform detected in the brain parenchyma. Murine mdr-1 P-glycoprotein is encoded by both mdr-1a (Abcb1a) and mdr-1b (Abcb1b), which share 90% sequence homology and have 80% homology to human MDR1 (ABCB1). The mdr-1a/b (Abcb1a/b) double knockout completely eliminates mdr-1-mediated transport activity at the BBB. Several published studies[8, 15–20] presented the evidence that inhibition or deficiency of Abcg2 or mdr-1 P-glycoprotein increases Aβ intake in cell models and reduces brain Aβ clearance in animal models.
To further evaluate the roles of Abcb1 and Abcg2 in Aβ trafficking across the BBB, we developed the non-invasive optical imaging method for ‘tracking’ systemically injected fluorescently-labeled Aβ peptides in Abcb1-KO and Abcg2-KO mice. For the purpose of in vivo tracking Aβ peptides were conjugated to the near-infrared optical fluorescence tracer Cy5.5. Since Aβ degrading proteases including insulin degrading enzyme (IDE), angiotensin converting enzyme (ACE) and neprilysin[42, 43] are active in the blood and can contribute to Aβ degradation, the stability of Cy5.5-Aβ conjugates in serum over 8 hours was confirmed ex-vivo, proving that the optical signal in imaging experiments originated predominantly from intact Cy5.5-Aβ conjugates. Imaging assessment of the whole-body biodistribution and elimination kinetics of Cy5.5-Aβ peptides, demonstrated similar elimination kinetics in wild-type and KO animals; the majority of peripheral tracer was eliminated by 2–4 h after the injection. This is in agreement with previous studies that reported the circulation half-life of injected [125I]-Aβ peptides of about 35–45 min; ~81% of the injected Aβ was cleared from blood by 60 min after administration in adult monkey[32, 33, 44].
Head ROI imaging protocols were initiated 2 hours after tracer injection, allowing 3–4 circulation half-lives; therefore, measured head fluorescence concentration was primarily indicative of the brain-accumulated/retained tracer, with small contribution of circulating tracer. In both Abcb1-KO and Abcg2-KO animals, brain tracer concentration was higher than in the wild-type animals at 2 hours, suggesting that any of the following processes or their combination might have been altered in knockout animals: a) the rate of Aβ brain influx was increased; b) the rate of Aβ brain elimination was slower; and c) Aβ binding/uptake into brain vessels was increased. Based on the current data, we cannot exclude any of these processes being responsible for the observed tracer concentration differences at 2 hours after injection. However, given the relatively short circulation half-life of Aβ, we can assume that imaging measurements between 2 and 8 hours after injection reflect predominantly brain elimination kinetics of Aβ. Brain-injected [125I]-Aβ1-40 peptide has been shown to clear rapidly via receptor-mediated transport with t1/2 of 25 minutes. A single photon emission computed tomography (SPECT) study in squirrel monkeys, demonstrated a bi-phasic brain clearance of intracerebrally microinfused [123I]-Aβ1-40, with short t1/2 ranging from 1.1 ~ 2.7 hours and accompanying plasma appearance of [123I]-Aβ1-40, suggesting active brain-to-blood transport. Comparisons of Aβ fluorescence decay curves between 2 and 8 h in wild-type and ABC-transporter knock-out animals indicated similar fluorescence decay (elimination) kinetics within the range of clearance rates described by Bading et al. Due to limited number of imaging time-points and the study design, it was not possible to discern whether the observed elimination kinetics of Aβ are due to active reverse transport across the BBB or to the interstitial fluid bulk-flow clearance.
Whereas lack of Abcg2 in this study did not appear to affect the rate of Aβ elimination from the brain, it resulted in higher initial accumulation of injected Aβ, suggesting that it has a role in either limiting brain access of circulating Aβ or mediating fast brain elimination phase of Aβ, or both. In agreement with our observations, a recent study using the in situ brain perfusion technique showed that GF120918, a dual inhibitor of Abcb1 and Abcg2, strongly enhanced the uptake of [3H]-Aβ1-40 in the brains of Abcb1-deficient mice, but not in the brains of Abcb1/Abcg2-deficient mice. ABCG2 is up-regulated in human AD brain with cerebral amyloid angiopathy (CAA) where it modulates Aβ-induced vascular oxidative stress[33, 47].
Similarly, the deficiency of mdr-1/P-glcoprotein significantly increased brain accumulation of systemically injected Aβ but also slightly accelerated its elimination from the brain. This observation is consistent with some previously reported studies. Deposition of Aβ peptides has been found to inversely correlate with MDR-1 P-glycoprotein/ABCB1 expression in the brains of elderly non-demented humans as well as in the brains of Alzheimer’s patients[37, 48, 49]. In addition, Aβ was found to down-regulate BBB mdr-1 P-glycoprotein (Abcb1) expression in mice. Cirrito and colleagues demonstrated that Aβ removal from the brain was partially mdr-1-dependent in mdr-1a/b KO mice. Furthermore, restoration of mdr-1 P-glycoprotein/Abcb1 at the BBB by PXR (Pregnane X Receptor) agonist reduced brain Aβ load in a mouse model of Alzheimer's disease.
The definitive interpretation of data provided in this study is confounded by possible activation of compensatory mechanisms in knock-out animals. For example, the Abcb1/P-glycoprotein-null mice were found to have lower brain expression of LRP-1 compared to wild-type mice. We found no compensatory changes in Abcb1a/mdr-1a and Abcb1b/mdr-1b expression in the brains of Abcg2-KO mice (data not shown); however, we cannot ascertain whether other Aβ transporters (i.e., RAGE, LRPs) were specifically affected in brain endothelial cells in Abcb1- or Abcg2-KO animals.
Pharmacological studies using selective inhibitors of BBB transporters in cell systems[15, 20] provided strong evidence that both ABCB1/MDR-1 P-glycoprotein and ABCG2 have the capacity to interact with and shuttle Aβ across cellular membranes. In vivo imaging studies, including ours presented here, support this notion and provide means for dynamic analyses of integrative influences of BBB transporters on Aβ trafficking in and out of the brain.
In summary, this study protocol describes potential application of time-domain prospective in vivo imaging in assessing BBB trafficking of systemically injected compounds, including Aβ peptides, labeled with near-infrared fluorescent imaging tracers. The protocol is particularly useful in assessing BBB trafficking of such compounds in animals exhibiting modifications of various BBB transporters, such as for example gene knock-out or over-expression of ABC-family of efflux pumps. Similarly, this imaging method can be used to evaluate kinetics of brain elimination of intra-cerebrally-injected compounds as recently described in our study on FcRn-mediated brain elimination of fluorescently-labeled macromolecules.