Pharmacokinetics and Amyloid Plaque Targeting Ability of a Novel Peptide-Based Magnetic Resonance Contrast Agent in Wild-Type and Alzheimer's Disease Transgenic Mice

Development of amyloid plaques in the extracellular space of the brain parenchyma is considered a primary event in the pathogenesis of Alzheimer's disease (AD) (Selkoe, 2001). Amyloid plaques consist predominantly of Aβ 40 and Aβ 42, which are produced continuously by cells in the nervous system and peripheral tissues. Currently, there is no definitive diagnosis for AD except clinically by elimination of other neurodegenerative disorders and histologically by post mortem observation of plaques and tangles. Early diagnosis of AD is difficult at present because of the inability to visualize plaques in vivo. In recent years, substantial effort has focused on the development of a pre-mortem diagnosis of AD, which involves detection of the plaques using various imaging techniques such as magnetic resonance imaging (MRI) (Poduslo et al., 2002) and positron emission tomography (Klunk et al., 2003, 2004). MRI used in conjunction with a contrast agent has a theoretical capability of resolving individual plaques and also differentiating plaques from other interfering structures such as blood vessels, myelinated fibers, iron-enriched glial cells, and neuronal cell populations whereas positron emission tomography has lower limits of resolution and can only detect bulk tissue enhancement (Poduslo et al., 2002).

Utilizing the ability of Aβ 40 to cross the blood-brain barrier (BBB) and accumulate in the brain, our laboratory and others have developed Aβ 40 and its derivatives as MRI probes for imaging amyloid plaques (Wengenack et al., 2000a; Lee et al., 2002; Poduslo et al., 2002, 2004; Wadghiri et al., 2003). These probes carry covalently attached gadolinium (Gd) diethylenetriaminepentaacetic acid (DTPA), which provides contrast for MRI. Previously, we reported the efficacy of putrescine (PUT)-modified Aβ 40 (PUT-Gd-Aβ 40) to cross the BBB compared with Gd-Aβ 40 without the putrescine modification and to provide contrast enhancement of plaques during MRI of APP/PS1 mouse brains ex vivo after i.v. injection (Poduslo et al., 2002). However, the utility of PUT-Gd-Aβ 40 for diagnostic use in animal models and patients is limited, because the carbodiimide-mediated modification of Aβ 40 with putrescine is associated with problems such as cross-linking, aggregate formation, and insolubility (Kandimalla et al., 2006). To avoid this carbodiimide modification of the peptide and these inherent issues, the complete chemical synthesis of a new probe, Gd[N-4ab/Q-4ab]Aβ 30, was achieved for its further development as a putative MRI contrast agent (Poduslo et al., 2004). Apart from having chemical purity, the putative neurotoxic domain found in Aβ 40 was truncated to minimize potential cellular toxicity. Furthermore, it is not amyloidogenic like Aβ 40 (Giles et al., 2005). Autoradiographic studies conducted on APP/PS1 mouse brain have demonstrated that ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 labels >90% plaques throughout the cortex and hippocampus after i.v. injection (Poduslo et al., 2004). The objective of the present study was to evaluate the pharmacokinetics and plaque targeting ability of Gd[N-4ab/Q-4ab]Aβ 30 in both wild-type and AD transgenic mice (APP/PS1).

Materials and Methods

Subjects. These studies were performed using wild-type mice (B6/SJL) and transgenic mice of the same background strain that express two mutant human proteins associated with familial AD. Wild-type (WT) mice were obtained from The Jackson Laboratory (Bar Harbor, ME) at 6 to 8 weeks of age. Hemizygous transgenic mice (Tg2576) expressing mutant human amyloid precursor protein (APP695) (Hsiao et al., 1996) were mated with a strain of homozygous transgenic mice (M146L6.2) expressing mutant human PS1 (Holcomb et al., 1998). The animals were genotyped for the expression of both transgenes by a polymerase chain reaction method using a sample of mouse tail DNA. These double transgenic mice have been shown to exhibit an accelerated phenotype with amyloid deposits and behavioral deficits by 12 weeks of age (Holcomb et al., 1998; Wengenack et al., 2000a). The mice were housed in a virus-free barrier facility under a 12-h light/dark cycle with ad libitum access to food and water. All procedures performed were in accordance with the Guidelines for the Care and Use of Laboratory Animals (National Institutes of Health Publication 85-23, revised 1996) using protocols approved by the Mayo Institutional Animal Care and Use Committee.

Synthesis of Diamine- and Gd-Substituted Aβ Derivative. Aβ 1–30, with the sequence Ahx (Fmoc-6-aminohexanoic acid)-DAEFRHDSGYEVHHQKLVFFAEDVGSNKGA, was synthesized as described previously by Poduslo et al. (2004) on an ABI 433 peptide synthesizer (Applied Biosystems, Foster City, CA) using 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate activation and the manufacturer's suggested synthesis protocols. The starting resin was Ala-NovaSyn TGA (Calbiochem-Novabiochem, San Diego, CA). Glutamic acid residues 3, 11, and 22 were synthesized with N-α-Fmoc-l-glutamyl-δ-N-(4-aminobutyl)carbamic acid tert-butyl ester, and aspartic acid residues 7 and 23 were synthesized with N-α-Fmoc-l-aspartyl-γ-N-(4-aminobutyl) carbamic acid tert-butyl ester. After completion of the synthesis and final Fmoc deprotection, DTPA anhydride was added to the N-terminal Ahx residue by dissolving 120 mg of the DTPA in 2 ml of dimethyl sulfoxide and 8 ml of dimethylformamide and reacting the DTPA solution with the peptide resin, which had been washed previously with diisopropylethylamine/dichloromethane. The coupling of DTPA was allowed to proceed overnight at RT. Completion of the reaction was verified by a negative ninhydrin reaction. The Aβ 1–30 peptide was then cleaved from the resin support using 5% crystalline phenol, 5% water, 2.5% triisopropylsilane, and 87.5% trifluoroacetic acid for 2 h at RT. The peptide was purified by reverse-phase high-performance liquid chromatography on a Jupiter C₁₈ column (250 mm × 21.2 mm; Phenomenex, Torrance, CA) using a gradient system of 0.1% aqueous trifluoroacetic acid containing 80% acetonitrile. The calculated mass weight of 3390 atomic mass units for Aβ 1–30 and 4231 atomic mass units for DTPA-[N-4ab/Q-4ab]Aβ 30 was confirmed by electrospray ionization mass spectrometry (API 165; Applied Biosystems/MDS Sciex (Foster City, CA).

Gadolinium Chelation and Radioiodinatation of Proteins. The element Gd was chelated at a mole concentration equal to that of the DTPA functional group of the Aβ 1–30 peptide using Gd(III) chloride hexahydrate (Sigma-Aldrich, St. Louis, MO) in water at RT for 1 h (now designated as Gd[N-4ab/Q-4ab]Aβ 30). After radioiodination using the chloramine T method (Poduslo and Curran, 1996), the ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 was dialyzed overnight against PBS to remove unbound ¹²⁵I and found to be 99% trichloroacetic acid-precipitable.

Pharmacokinetic Studies. Before the beginning of experiment, the femoral vein and the femoral artery of each mouse were catheterized under general anesthesia (1.5% isoflurane and 4 l/min oxygen). After i.v. administration of 100 μCi of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 in the femoral vein, the plasma pharmacokinetics of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 were determined in six WT and six APP/PS1 mice by collecting serial blood samples (20 μl) from the femoral artery over a period of 60 min at time points of 0.25, 1, 3, 5, 10, 15, 30, 45, and 60 min. The blood samples were diluted to a volume of 100 μl using normal saline and centrifuged, and the supernatant was obtained. After trichloroacetic acid precipitation, the samples were assayed for ¹²⁵I radioactivity in a gamma counter (Cobra II; PerkinElmer Life and Analytical Sciences, Boston, MA). The linearity of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 disposition was determined by repeating the experiment with coadministration of 1 or 2 mg of cold Gd[N-4ab/Q-4ab]Aβ 30 with 100 μCi of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30.

To determine the brain uptake of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30, a destructive sampling study design was followed in which cohorts of five WT mice and one APP/PS1 mouse were sacrificed at each time point. After the surgery to catheterize the femoral vein and artery, 100 μl of the probe (1 μCi/ml) was injected into the femoral vein of WT or APP/PS1 mice. At the end of each experiment, which was terminated at 1, 3, 5, 10, 15, 30, 45, 60, 120, 180, 240, or 300 min, an aliquot of ¹³¹I-BSA (100 μCi, 100 μl) was injected in the femoral vein of the animals to serve as a measure of residual plasma volume (V_p). One minute after the ¹³¹I-BSA injection, the final blood sample was collected, and the animal was sacrificed. The brain of the animal was removed from the cranial cavity, dissected into the anatomical regions (cortex, caudate putamen, hippocampus, thalamus, brain stem, and cerebellum), and then assayed for ¹²⁵I and ¹³¹I radioactivity. The measured activity was corrected for the background and crossover of ¹³¹I activity into the ¹²⁵I channel. The brain regions were lyophilized, and dry weights were determined with a microbalance and converted to wet weights using wet weight/dry weight ratios determined previously.

The permeability of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 at the BBB was determined in a similar fashion. However, the length of the experiment for all permeability studies was kept at 15 min. The saturability of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 transport at the BBB was determined by coadministering 0.5, 1, or 2 mg of cold Gd[N-4ab/Q-4ab]Aβ 30 with 100 μCi of ¹²⁵I-labeled compound.

Labeling of Amyloid Plaques in Vivo. APP/PS1 transgenic mice (8 months of age) were catheterized in the femoral vein under general anesthesia (1.5% isoflurane) and injected with ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 to determine the time course and dose response of the radiolabeled peptide binding to amyloid plaques detected by emulsion microautoradiography. For the time course experiment, each animal was injected with 1.0 mg of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 and then sacrificed after 1, 2, 4, or 8 h. A WT mouse was injected with 1.0 mg of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 and sacrificed after 1 h for comparison. Another APP/PS1 mouse was injected with PBS and sacrificed after 1 h as a negative control. For the dose-response experiment, animals were injected with either 1.25, 2.5, or 5 mg of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 and sacrificed after 2 h. At the appropriate time, each animal was given an overdose of sodium pentobarbital (200 mg/kg i.p.) and perfused with PBS, followed by neutral-buffered, 10% formalin, and then 10% sucrose and 0.1 M sodium phosphate (pH 7.2). Frozen sections (15 μm) of each brain were cut with a cryostat and then processed with anti-Aβ immunohistochemistry (IH) and emulsion autoradiography for the presence of radiolabeled amyloid deposits. In brief, the sections were subjected to IH for amyloid using a standard anti-Aβ monoclonal mouse antibody (4G8, 1:1000; Signet Laboratories, Dedham, MA) and standard immunoperoxidase methods (Vectastain Elite ABC and DAB kits; Vector Laboratories, Burlingame, CA). Next, the sections were dipped in an autoradiographic emulsion (type NTB-3; Eastman Kodak, Rochester, NY) for direct comparison of ¹²⁵I-labeled amyloid deposits to anti-Aβ IH. The slides were dipped in emulsion, exposed for various durations, and developed according to the instructions using Kodak Dektol developer and fixer (Eastman Kodak). The sections were dehydrated with successive changes of ethanol and xylene and then coverslipped.

Silver grains from sections exposed for 8 weeks were quantitated using unbiased, stereological techniques. Silver grains were counted over plaques and adjacent parenchyma in three sections from each animal in the retrosplenial cortex and CA1 region of the hippocampus using a 10-μm × 10-μm dissector at 400×. The mean background level of exposed silver grains was also determined for each section. The background was sampled over the emulsion-coated blank slide adjacent to the tissue section in the vicinity of the retrosplenial cortex and hippocampus. The results were expressed as the mean number of silver grains per 100 μm² minus the background. Statistical analysis was then performed using ANOVA followed by Bonferroni multiple comparisons (Prism, version 4.0; GraphPad Software Inc., San Diego, CA).

Conversion of Silver Grain Densities to Radioactivity. Brain uptake data obtained in APP/PS1 animals at 60 and 120 min were compared with the microautoradiography data obtained at the same time points. A proportionality factor relating the number of silver grains (SGN) to radioactivity (RA) was obtained from the following expression: A factor to convert SGN to RA was derived by averaging the proportionality factors obtained at various time points.

Data Analysis. The Gd[N-4ab/Q-4ab]Aβ 30 plasma concentration profile after a single i.v. bolus dose of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 was best described by a biexponential disposition function C(t) = Ae^–αt + Be^–βt, where C(t) = ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 (microcuries) per milliliter of plasma, A and B are the intercepts, and α and β are the slopes of the biexponential curve. Pharmacokinetic parameters were estimated by nonlinear curve fitting using the Gauss-Newton (Levenberg and Hartley) algorithm and weighted least-squares, (WinNonlin Professional, version 4.1; Pharsight; Mountain View, CA). Secondary parameters such as the C_max (maximum plasma concentration), the first-phase (t_1/2(α)) and second-phase (t_1/2(β)) half-lives, the plasma clearance (CL), the steady-state volume of distribution (V_ss), and the area under the plasma concentration curve (AUC) were also calculated using WinNonlin. The mean values of controls and treatments were compared by Student's t test using GraphPad Prism version 4.0.

The amount of extravascular¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 present in the brain (X_B) at various time points was calculated using the following equation system. The residual brain region plasma volume (V_p in microliters per gram) was determined as described previously by Poduslo and Curran (1996): where q_p is the ¹³¹I-BSA content (counts per minute) of tissue, C_v is the ¹³¹I-BSA concentration (counts per minute per milliliter) in plasma, W is the dry weight (g) of the brain region, and R is the wet weight/dry weight ratio for mice of a defined age group. From the total ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 content (X_TB in counts per minute) of the brain region, the amount of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 that enters the brain region extravascular space (X_B in counts per minute per gram) is calculated as where C_a is the final ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 concentration (counts per minute per milliliter) in plasma.

A pharmacokinetic model with one brain compartment (Fig. 3) was investigated to describe ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 uptake in the brain. The differential equations associated with the model are as follows: where X_P is the amount of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 in plasma, X_T is the amount of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 present in the tissue compartment; X_B is the amount of extravascular ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 present in the brain;. K_PB is the rate constant for the transfer from plasma to brain; K_TP is the rate constant for the transfer from the tissue compartment to plasma; K_BP is the rate constant for the transfer from brain to plasma; and K_EL is the rate for the elimination from plasma; all of the rates were assumed to be first-order.

The rate constants K_PB, K_TP, K_BP, and K_EL were obtained by fitting the above equations simultaneously to the ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 plasma concentration-time and the extravascular ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 in the brain (X_B)-time data.

Results

¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 Plasma Pharmacokinetics and Metabolism. The results obtained from these studies are summarized in Figs. 1 and 2 and Table 1. After i.v. administration, the ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 concentration in the plasma of WT as well as APP/PS1 mice declined rapidly, exhibiting a three-exponential disposition (Table 1). Even though, a two-compartment pharmacokinetic model fitted the plasma data reasonably well (Fig. 1), adding another exponential term significantly improved the goodness-of-fit as indicated by the F test, Akaike information criterion, and Schwartz criterion. The three-compartment model parameters such as C_max and A are significantly higher in WT mice than in APP/PS1 mice, whereas C and γ are significantly higher in APP/PS1 than in WT mice (Table 1). However, no significant differences in the plasma pharmacokinetic parameters were observed between WT and APP/PS1 mice when a two-compartment model was fitted to the data (Table 1).

A substantial amount of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 was found in the liver, kidney, and spleen of both WT and APP/PS1 animals perfused with PBS at the termination of the experiment. The accumulation of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 was higher in the kidney than in the liver or spleen. However, no significant differences in the accumulation of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 in these organs was observed between WT and APP/PS1 animals (data not shown). The kinetics of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 degradation was further elucidated in vitro in slices of liver, kidney, and spleen of WT and APP/PS1 mice. Substantial ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 degradation was observed in these tissue slices compared with that in plasma (Fig. 2). However, no significant differences in ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 degradation were observed between APP/PS1 and WT mouse tissues.

To determine whether the disposition of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 in the peripheral circulation is saturable, the plasma kinetics of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 was studied by coadministering various amounts of unlabeled Gd[N-4ab/Q-4ab]Aβ 30 (1 and 2 mg). A three-compartment pharmacokinetic model fitted to the plasma data resulted in poor precision in the parameter estimates, most probably due to the saturation of kinetic events described by one or more of the exponential terms in the three-compartment model. Therefore, a simpler two-compartment pharmacokinetic model was used to evaluate the saturability of Gd[N-4ab/Q-4ab]Aβ 30 plasma disposition, which gave highly precise parameter estimates (Fig. 1; Table 1). A close examination of the plasma pharmacokinetic parameter values indicated that the coadministration of 1 mg of Gd[N-4ab/Q-4ab]Aβ 30 with 100 μCi of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 resulted in no significant changes in the plasma pharmacokinetics (Fig. 1; Table 1). However, upon the coadministration of 2 mg of Gd[N-4ab/Q-4ab]Aβ 30, the clearance (CL) of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 decreased significantly (p < 0.01), whereas the AUC (p < 0.05) and C_max (p < 0.01) increased significantly (Fig. 1; Table 1) compared with that in the WT mice given labeled reagent alone.

¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 Brain Uptake. The results from brain uptake studies are presented in Figs. 3, 4, 5 and Table 2. The uptake of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 into cortex, hippocampus, and total brain tissue after i.v. administration in 24-week-old WT and APP/PS1 mice was adequately described by the three-compartment open model depicted in Fig. 3. The parameter values that adequately describe ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 brain kinetics were obtained by fitting the differential equations to the plasma and brain kinetic data simultaneously. The ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 pharmacokinetic profile in the brain of WT mice is characterized by a rapid uptake (t_max ∼15 min) from plasma (K_PB, 1/min), which was estimated as 0.026 ± 0.002 (Fig. 4, Table 2) and a significantly lower rate of elimination from brain to plasma (K_BP, 1/min), which was estimated to be 0.006 ± 0.001. The K_PB and K_BP values in APP/PS1 mice were similar to those seen in WT mice. Elimination of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 from the plasma compartment of WT (K_EL = 0.22 ± 0.04), 1/min) as well as APP/PS1 mice (Table 2: K_EL = 0.14 ± 0.02) estimated by this method coincided very well with the value estimated by fitting either a two-compartment (Table 1: K_EL,WT = 0.21 ± 0.04; K_EL,APP/PS1 = 0.14 ± 0.01) or a three-compartment open model (Table 1: K_EL,WT = 0.20 ± 0.02; K_EL,APP/PS1 = 0.15 ± 0.02) to the plasma data alone.

To verify whether the uptake of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 at the BBB is receptor-mediated, various amounts of unlabeled Gd[N-4ab/Q-4ab]Aβ 30 were coadministered along with ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 in 24-week-old WT mice. Upon coadministration of 0.5 or 2 mg of Gd[N-4ab/Q-4ab]Aβ 30, the extravascular accumulation of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 in various brain regions remained unaffected (Fig. 5).

Verification of Amyloid Plaque Targeting of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 by Emulsion Autoradiography. Direct verification of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 brain targeting was made by performing emulsion autoradiography on sections of APP/PS1 mouse brain obtained at various time points after an i.v. bolus injection. Similar studies were also performed by administering various amounts of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 to determine the effect of dose on the extent of plaque targeting.

The representative autoradiographs obtained from the hippocampal region of 8-month-old APP/PS1 transgenic mice injected with ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 qualitatively depict accumulation of the probe in the brain tissue (Fig. 6, A–D). The autoradiographs clearly demonstrated preferential accumulation of the radiolabeled probe on the plaques compared with the brain parenchyma. The amount of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 accumulated on the plaques was estimated from similar autoradiographs by counting the number of exposed silver grains (Fig. 6E). These data indicate that the maximum number of silver grains was associated with the plaques 1 h after the injection of the probe and decreased significantly in the following 3 h. The elimination rate of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 from the brain parenchyma obtained from the pharmacokinetic experiments coincided well with the rate obtained via the silver grain count resulting from the emulsion autoradiography studies (Fig. 7). Furthermore, accumulation of silver grains on the plaque surface is directly proportional to the administered dose (Fig. 8). It is obvious from both time course and dose-response studies that the accumulation of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 is higher on the hippocampal plaques than on in the cortex.

Even though it is clear from the autoradiography studies that ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 preferentially targets amyloid plaques in APP/PS1 transgenic mouse brain, they do not present direct evidence that gadolinium, which provides contrast for MRI, remained chelated to the DTPA arm of the protein accumulated on the plaque. Therefore, identical targeting studies were performed using ¹⁵³Gd chelated to the unlabeled protein (¹⁵³Gd[N-4ab/Q-4ab]Aβ 30) without the addition of ¹²⁵I. The results of the targeting studies showed plaques densely populated with silver grains (Fig. 9), which reflect the presence of the probe carrying ¹⁵³Gd.

Discussion

Various Aβ derivatives have been developed in our laboratory as carriers of the MRI contrast agent gadolinium to the plaque surface. If successful, these carriers could help identify amyloid plaques present in APP/PS1 mouse brain and eventually in humans, which is critical for the early detection of Alzheimer's disease. Gd[N-4ab/Q-4ab]Aβ 30 is such a novel MRI contrast agent developed in our laboratory, which was previously shown to cross the BBB and bind to amyloid plaques in the APP/PS1 mouse brain (Poduslo et al., 2004). Therefore, extensive plasma and brain pharmacokinetic studies and emulsion autoradiography studies were conducted on this contrast agent in WT and APP/PS1 mice. The results from these studies provide a comprehensive quantitative estimate of the capabilities of the contrast agent to target amyloid plaques and also outline physiological conditions under which the potential of this contrast agent could be realized.

The amount of Gd[N-4ab/Q-4ab]Aβ 30 available for plaque binding is dependent on the C_max, t_max, and the residence time of the contrast agent in the brain, which in turn is dependent on the plasma concentration of the contrast agent. Based on the plasma pharmacokinetic profile presented in Fig. 1, Gd[N-4ab/Q-4ab]Aβ 30 exhibits rapid peripheral elimination, which is saturable only at very high doses. Although it is believed that a lower plasma clearance resulting in high sustained plasma levels of the probe will enhance BBB permeability, many Aβ 40 derivatives developed in our laboratory as contrast agents exhibit rapid peripheral elimination. It may be that the hepatic metabolism and renal elimination caused by cationic charge density, a common structural feature shared by many of these proteins, is responsible for the rapid systemic elimination. Previous studies have demonstrated that imaging of amyloid plaques in AD transgenic mouse brain was possible even with contrast agents with low plasma residence time, such as PUT-Aβ 40 (Wengenack et al., 2000a), suggesting that low plasma residence time might be offset by favorable brain kinetics.

Poduslo et al. (2004) demonstrated that the permeability surface area product of Gd[N-4ab/Q-4ab]Aβ 30 at the BBB in WT and APP/PS1 mice is as high as that of Aβ 40, which is reportedly transported across the BBB via receptor-mediated endocytosis (Poduslo et al., 1999; Deane et al., 2003, 2004; Kandimalla et al., 2005). The permeability surface area product value of a molecule is calculated by dividing the amount in the extravascular compartment of the brain after an i.v. bolus injection by the integral of the amount in the plasma (AUC). Although the plasma AUC of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 in mice increased in the presence of unlabeled Gd[N-4ab/Q-4ab]Aβ 30 because of saturable peripheral elimination, the amount of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 reaching the extravascular brain tissue remained unchanged, which indicates that its plasma and brain kinetics could be different. Hence, the detailed kinetics of extravascular ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 after an i.v. bolus injection were elucidated in WT and APP/PS1 mice.

The kinetics of extravascular ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 in the brain tissue is characterized by a rapid absorption phase and an extended elimination phase. This is an ideal kinetic profile for MRI, which usually requires longer scan times. Although not amenable to statistical tests because of the small sample size, higher accumulation of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 was observed in hippocampus than in cortex despite the presence of greater cortical tissue mass compared with that of hippocampal tissue. Based on the facts that this difference was observed in both WT and APP/PS1 mice and that there was no significant difference in the plaque burden between these two regions in 6-month-old APP/PS1 animals (Wengenack et al., 2000b), the observed differences in ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 uptake could be mostly due to differences in the blood flow or capillary density between the hippocampus and the cortex regions. Despite some differences in the plasma pharmacokinetic parameters, no major differences were observed in the brain kinetics between WT and APP/PS1 mice, which justifies the use of WT mice as cheaper alternatives to APP/PS1 mice in the further development of Gd[N-4ab/Q-4ab]Aβ 30 as an MRI contrast agent.

In the above brain pharmacokinetic studies, extravascular accumulation of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 was determined indirectly by subtracting the amount of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 present in the brain vasculature from the amount present in the total brain tissue. Although very convenient, this method could yield misleading results, particularly if the probe has significant accumulation in the BBB endothelial cells. Therefore, direct verification of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 brain targeting using a semiquantitative technique such as emulsion autoradiography is a necessary prerequisite to conducting expensive and time-consuming in vivo MRI in WT and APP/PS1 mice. In addition to verifying the extent of amyloid plaque targeting of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30, emulsion autoradiography can also provide an alternative means for obtaining the elimination kinetics of the probe from the brain. In the present study, emulsion autoradiography was used to determine 1) elimination kinetics of the probe from brain parenchyma as well as from plaques and 2) the effect of probe concentration on the extent of plaque targeting.

The results obtained from these studies clearly demonstrated plaque-specific targeting of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30. The extent of targeting was similar to that of our previously successful MRI contrast agent (Wengenack et al., 2000a), putrescine-modified Aβ 40, which, when administered i.v., was shown to provide contrast for imaging plaques in APP/PS1 animals using a 7-T MRI system (Poduslo et al., 2002). In addition, 50 times greater accumulation of the probe on plaques than in the parenchyma was observed, which could significantly aid in the detection of plaques against the background. Slow removal of the probe from plaques, which is even more evident on the hippocampal plaques, allows for longer MRI scan times. Confirming the pharmacokinetic observations that the uptake of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30 at the BBB is nonsaturable, the silver grain density on the plaques present in both cortical and hippocampal regions increased linearly with the administered dose. This important observation will provide increased rationale for changing the mode of administration to continuous i.v. infusion, which can counteract the rapid peripheral elimination of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30.

In summary, the current study systematically describes plasma and brain pharmacokinetics of ¹²⁵I-Gd[N-4ab/Q-4ab]Aβ 30, a novel MRI contrast agent to detect amyloid plaques in AD transgenic mouse brain. Emulsion autoradiography studies conducted on the AD mouse brain after i.v. bolus injection of the contrast agent clearly showed plaque specific accumulation of the contrast agent in the brain, thereby demonstrating the potential for achieving an excellent signal/noise ratio on the MRI scans. Both pharmacokinetic studies and autoradiography studies coincided very well in describing the rapid nonsaturable uptake into and slow elimination of the probe from the brain. This information will be immensely helpful in determining the dose, mode of administration, and scan times for future in vivo MRI of amyloid plaques in AD transgenic mice.