F-18 Polyethyleneglycol stilbenes as PET imaging agents targeting Aβ aggregates in the brain

Abstract

This paper describes a novel series of ¹⁸F-labeled polyethyleneglycol (PEG)-stilbene derivatives as potential β-amyloid (Aβ) plaque-specific imaging agents for positron emission tomography (PET). In these series of compounds, ¹⁸F is linked to the stilbene through a PEG chain, of which the number of ethoxy groups ranges from 2 to 5. The purpose of adding PEG groups is to lower the lipophilicity and improve bioavailability. The syntheses of the “cold” compounds and the ¹⁸F-labeled PEG stilbene derivatives are successfully achieved. All of the fluorinated stilbenes displayed high binding affinities in an assay using postmortem AD brain homogenates (K_i=2.9–6.7 nM). Labeling was successfully performed by a substitution of the mesylate group of 10a–d by [¹⁸F]fluoride giving the target compounds [¹⁸F]12a–d (EOS, specific activity, 900–1500 Ci/mmol; radiochemical purity >99%). In vivo biodistribution of these novel ¹⁸F ligands in normal mice exhibited excellent brain penetrations and rapid washouts after an intravenous injection (6.6–8.1 and 1.2–2.6 %dose/g at 2 and 60 min, respectively). Autoradiography of postmortem AD brain sections of [¹⁸F]12a–d confirmed the specific binding related to the presence of Aβ plaques. In addition, in vivo plaque labeling can be clearly demonstrated with these ¹⁸F-labeled agents in transgenic mice (Tg2576), a useful animal model for Alzheimer's disease. In conclusion, the preliminary results strongly suggest these fluorinated PEG stilbene derivatives are suitable candidates as Aβ plaque imaging agents for studying patients with Alzheimer's disease.

Introduction

Accumulation of β-amyloid (Aβ) plaques in the brain is believed to be one of the most significant factors associated with the development of Alzheimer's disease (AD) [1]. Therefore, developing Aβ plaque-specific probe for in vivo imaging studies of Aβ plaques may be important for diagnosis and monitoring of AD patients [2], [3], [4], [5]. There are several potential benefits of imaging Aβ aggregates in the brain. The imaging technique will improve diagnosis by identifying potential patients with excess Aβ plaques in the brain; therefore, they may be likely to develop Alzheimer's disease. It will also be useful to monitor the progression of the disease. When antiplaque drug treatments become available, imaging Aβ plaques in the brain may provide an essential tool for monitoring treatment.

A highly lipophilic tracer, [¹⁸F]FDDNP (Fig. 1), for binding both tangles (mainly composed of hyperphosphorylated tau protein) and plaques (containing Aβ protein aggregates) has been reported [6]. Using positron emission tomography (PET), it was reported that this tracer specifically labeled deposits of plaques and tangles in nine AD patients and seven comparison subjects [5], [7]. This probe is lipophilic, and it is relatively difficult to measure, specifically, the signal contributing from binding of the plaque and tangles in the brain. Using a novel pharmacokinetic analysis procedure called the relative residence time of the brain region of interest vs. the pons, differences between AD patients and comparison subjects were demonstrated [8]. The relative residence time was significantly higher in AD patients. This is further complicated by an intriguing finding that FDDNP competes with some NSAIDs for binding to Aβ fibrils in vitro and to Aβ plaques ex vivo [6], [9].

Imaging β-amyloid in the brain of AD patients by using a benzothiazole aniline derivative, [¹¹C]6-OH-BTA-1 (also referred to as [¹¹C]PIB), was recently reported [10]. Contrary to that observed for [¹⁸F]FDDNP, [¹¹C]6-OH-BTA-1 binds specifically to fibrillar Aβ in vivo [11]. Patients with diagnosed mild AD showed marked retention of [¹¹C]6-OH-BTA-1 in the cortex, known to contain large amounts of amyloid deposits in AD. In the AD patient group, [¹¹C]6-OH-BTA-1 retention was increased most prominently in the frontal cortex. Large increases were also observed in parietal, temporal and occipital cortices, and in the striatum. [¹¹C]6-OH-BTA-1 retention was equivalent in AD patients and comparison subjects in areas known to be relatively unaffected by amyloid deposition (such as subcortical white matter, pons and cerebellum). Recently, another ¹¹C-labeled Aβ plaque-targeting probe, a stilbene derivative-[¹¹C]SB-13, has been studied. In vitro binding using the [³H]SB-13 suggests that the compound showed excellent binding affinity, and binding can be clearly measured in the cortical gray matter, but not in the white matter of AD cases [12]. There was a very low specific binding in cortical tissue homogenates of control brains. The K_d values of [³H]SB-13 in AD cortical homogenates were 2.4±0.2 nM. High binding capacity and comparable values were observed (14–45 pmol/mg protein) [12]. As expected, in AD patients [¹¹C]SB-13 displayed a high accumulation in the frontal cortex (presumably an area containing a high density of Aβ plaques) in mild to moderate AD patients, but not in age-matched control subjects [13].

Generally, the ¹¹C-labeled agents (T_1/2=20 min, β+) are likely to be useful in major medical centers with PET scanner and cyclotron facilities. However, to reach a wider patient population, tracers labeled with ¹⁸F (T_1/2=110 min, β+) may be more useful as PET imaging agents for detection of aggregates in the brains of living patients. Compared to ¹¹C-labeled probes, the ¹⁸F imaging agents have several advantages: ¹⁸F-labeled agents may be easier for routine manufacturing and distribution. Due to the limited half-life of ¹¹C (T_1/2=110 vs. 20 min), supplying ¹¹C agents will always be limited to facilities with an on-site cyclotron and a one-batch-one-patient model. However, preparation of one lot of ¹⁸F-labeled imaging agent may allow imaging studies in many patients.

We have demonstrated that a series of 4-amino-4′-hydroxyl-substituted stilbene derivatives, including SB-13, showed high binding affinity to Aβ aggregates [14], [15], [16]. It appears that only one of the electron-donating group, either a nitrogen or oxygen, is needed at each end of the stilbene for preserving the binding affinity, freeing the other end for substitution with different radiolabeling groups. One of the key factors in developing in vivo imaging agents containing fluorine atom is the need for a moderate lipophilicity (log P=1–3). Adding one O-fluoroethyl group on SB-13 appears to increase the lipophilicity significantly, leading to an increase in blood levels as well as nonspecific binding in the brain (unpublished data). Nonspecific binding in the normal brain area will lower the signal-to-noise ratio, and, thus, rendering the imaging agent not selective in targeting Aβ plaques in the brain. To circumvent the problem of having high lipophilicity after adding a fluoroalkyl group, we have prepared stilbene derivatives with an additional hydroxyl group, such as FMAPO (see Fig. 1), which showed a moderate log P value of 2.95 and an excellent brain penetration (brain uptake at 2 min post-intravenous injection in mouse brain was 9.75 %dose/g) [16]. Similarly, we have added polyethylene glycol (PEG) units (n=2–5), as a tether, on the 4′-hydroxyl group of 4-N-methylamino-4′-hydroxyl-stilbene, SB-13, through which the fluorine atom is attached at the end of PEG side chain. The fluorinated PEG group provides a flexible tool to adjust the lipophilicity and maintain a relatively small size. To further reduce the molecular size and lipophilicity in these two successful examples we have also used a N-monomethylamino group instead of a more readily prepared N,N-dimethylamino group as the anchor for binding to the Aβ plaques. While this work was in progress a series of ¹⁸F-labeled styrylbenzoxazole compounds, such as 2-(4-methylaminostyryl)-6-(2-[¹⁸F]fluoroethoxy) benzoxazole, were reported to show promise as potential PET imaging agents targeting amyloid plaques in the brain [17].

Reported herein are the synthesis and binding characterization of a series of fluorinated PEG stilbenes as potential imaging agents for PET imaging of Aβ plaques in the brain.