The synthetic route to the (E)-5-styryl-1H-indole and (E)-6-styrylquinoline derivatives is shown in Scheme 1.

The key step was the base-catalyzed Wittig reaction between substituted triphenyl phosphonium ylides 2a–g and 1H-indole-5-carbaldehyde or quinoline-6-carbaldehyde. The tributyltin derivatives 11, 19 were prepared in yields of 22.3% and 28.6%, respectively, from the bromo-precursors 7, 16using an exchange reaction catalyzed by Pd(0). [¹²⁵I]8 and [¹²⁵I]17 were prepared via a iodo-destannylation reaction using hydrogen peroxide as the oxidant. The products were purified by radio-HPLC using a reverse-phase column and mobile phase consisting of acetonitrile with a flow rate of 1 mL/min. In order to identify the radiotracer, the non-radioactive 8 and 17 were co-injected and co-eluted with the corresponding radioactive product, respectively. Their HPLC profiles using acetonitrile and water (90:10 v/v) as mobile phase at a flow rate of 1 mL/min are present in Figure 3.

From Figure 3, the retention times of non-radioactive 8 and [¹²⁵I]8 were observed to be 6.45 min and 6.89 min, respectively. The retention times of non-radioactive 17 and [¹²⁵I]17 were observed to be 19.56 min and 19.94 min, respectively. The differences in retention time were in good agreement with the time lag which corresponds with the volume and flow rate within the distance between the UV and radioactive detector of our HPLC system. After purification by HPLC, the radiochemical purities of both [¹²⁵I]8 and [¹²⁵I]17 were greater than 98%. The radiochemical yields of [¹²⁵I]8 and [¹²⁵I]17 were 48–67% and 61–78%, respectively. The log D values of [¹²⁵I]8 and [¹²⁵I]17 were 2.52 ± 0.04 and 2.73 ± 0.03, respectively, which are in the appropriate range for brain imaging agents indicative of good permeability through the blood-brain barrier (BBB).

The affinity of (E)-5-styryl-1H-indole and (E)-6-styrylquinoline derivatives for Aβ_1–40 aggregates was determined by competition binding assay using [¹²⁵I]TZDM as radio-ligand. TZDM was also screened using the same competition assay for comparison. The K_i values shown in Table 1 were varied from 4.1 to 288.4 nM suggesting that all these compounds share the same binding site with ThT. The K_i value of TZDM was 4.2 nM, which is comparable to that of previously reported in the literature (K_i = 0.9 nM) [21]. (E)-5-styryl-1H-indole (4) without any substituents showed moderate binding affinity (K_i = 25.1 nM). Introducing a F, Cl or OCH₃ group at the para-position of the phenyl ring decreased the binding affinity (K_i = 89.3, 51.5 and 32.4 nM for compounds 5, 6 and 10, respectively), while introducing a Br or CH₃ group at the same position increased the affinity (K_i = 16.3 and 15.8 nM for compounds 7 and 9, respectively). It is noteworthy that compound 8 with a iodo group showed K_i value of 4.1 nM, which is comparable with that of TZDM. In general, (E)-5-styryl-1H-indole derivatives showed slightly better potency in binding to Aβ_1–40 aggregates than (E)-6-styrylquinoline derivatives. Since derivatives 8 and 17 with iodine at the para-position of the phenyl ring displayed nanomolar affinities for Aβ_1–40 aggregates, we prepared [¹²⁵I]8 and [¹²⁵I]17 for further evaluation as potential ligands for ¹²³I-labeled SPECT imaging agents.

To confirm the binding affinities of these derivatives for Aβ plaques in the brain, in vitro fluorescent staining of brain sections (8 µm) from a transgenic model mouse (APP/PS1, 12 months, male) was carried out with compound 8. As shown in Figure 4, many fluorescence spots were observed in the brain sections of transgenic mice (Figure 4B). The fluorescent labeling pattern was consistent with that observed with thioflavin-S (Figure 4A). These results suggested that 8 show specific binding to Aβplaques in the transgenic model mouse brain.

The results of in vitro autoradiography of [¹²⁵I]8 in the brain sections of a transgenic model mouse (APP/PS1, 12 months, male) are shown in Figure 5. [¹²⁵I]8 showed excellent labeling of Aβ plaques in the cortex region of the brain sections, and no remarkable accumulation of radioactivity were observed in white matter. The same sections were also stained with thioflavin-S and the localizations of Aβ plaques were in accord with the results of autoradiography. These results demonstrated that [¹²⁵I]8 was specific for Aβ plaques, which were consistent with the high binding affinity of compound 8 to Aβ_1–40 aggregates.

In vivo biodistribution studies of [¹²⁵I]8 and [¹²⁵I]17 were performed in normal mice. The uptake of radiotracer in the organs of interest at different time points after intravenous administration of [¹²⁵I]8 and [¹²⁵I]17 is summarized in Table 2. [¹²⁵I]8 showed high initial brain uptake followed by rapid clearance (4.27 and 0.28% ID/g at 2 and 60 min post injection, respectively). On the other hand, [¹²⁵I]17 showed relatively low brain uptake and slow washout (2.05 and 0.55% ID/g at 2 and 60 min post injection, respectively). As compared with previously reported radioiodinated N-benzoylindole derivatives [18], radioiodinated (E)-5-styryl-1H-indole derivative [¹²⁵I]8 showed greatly improved brain uptake. Because there are no plaques in normal brain, potential Aβ-specific probe should possess high brain uptake followed by fast washout in normal mice. The brain_2min/brain_60min ratio has been used as an index to compare the washout rate from normal brain and select candidate tracers with appropriate kinetics in vivo. It was reported that [¹²³I]IMPY showed a high initial brain uptake and fast washout in normal mice (2.88% ID/organ and 0.21% ID/organ at 2 min and 60 min postinjection, respectively) [16]. The brain_2min/brain_60min ratio of [¹²⁵I]8 (15.3) is higher than that of [¹²³I]IMPY (13.7), indicating [¹²⁵I]8 may possess suitable pharmacokinetic properties for imaging Aβ plaques in AD brain. Accordingly, [¹²⁵I]8 may be comparable or even better for detecting Aβ plaques. Therefore, (E)-5-styryl-1H-indole derivative [¹²⁵I]8, with nanomolar affinity to Aβ_1–40 aggregates, excellent BBB permeability as well as fast washout from the normal brain, may be suitable for development as a novel Aβ imaging agent.

Unless otherwise indicated, all chemicals used in synthesis were commercial products and were used without further purification. Na¹²⁵I (2200 Ci/mmol) was obtained from PerkinElmer Life and Analytical Sciences, USA. The double transgenic (APP/PS1) AD model mouse was obtained from Institute of Laboratory Animal Science, Chinese Academy of Medical Sciences and Comparative Medicine Center of Peking Union Medical College (Beijing, China). ¹H-NMR spectra were obtained on Bruker (400 MHz) NMR spectrometer at room temperature with TMS as an internal standard. Chemical shifts are reported as δ values relative to internal TMS. Coupling constants are reported in hertz. The multiplicity is defined by s (singlet), d (doublet), t (triplet), and m (multiplet). Mass spectra were acquired using the Surveyor MSQ Plus (ESI) (Waltham, MA, USA) instrument. HPLC was performed on a Shimadzu SCL-10AVP system (Shimadzu Corporation, Kyoto, Japan) which consisted of a binary pump with on-line degasser, a model SPD-10AVP UV-VIS detector operating at a wavelength of 254 nm, and a Packard 500TR series flow scintillation analyzer (Packard BioScience Co., Wallingford, CT, USA) with a Alltech Alltima RPC-18 column (5 μm, ID = 4.6 mm, length = 250 mm). The samples were analyzed using acetonitrile and water (90:10 v/v) as mobile phase at a flow rate of 1 mL/min. The sample was separated using acetonitrile as mobile phase at a flow rate of 1 mL/min. All key compounds were proven by analytical HPLC analysis to show ≥95% purity (Supporting information).

The determination of partition coefficients of [¹²⁵I]8 and [¹²⁵I]17 was performed according to the procedure previously reported with some modifications [23]. Ligand [¹²⁵I]8 or [¹²⁵I]17 (~5 μCi) was mixed with 3 mL each of n-octanol and PBS (0.02 M, pH 7.4) in a test tube. The test tube was vortexed for 5 min at room temperature, followed by centrifugation for 10 min at 3,000 rpm. Two weighed samples from the n-octanol (50 µL) and buffer layers (400 µL) were counted in a γ-counter. The partition coefficient was expressed as the logarithm of the ratio of the counts per gram from n-octanol versus that of PBS. Samples from the n-octanol layer were repartitioned until consistent partition coefficient values were obtained. The measurements were done in triplicate and repeated three times.

The lyophilized white powder of β-amyloid(1–40) were purchased from AnaSpec (San Jose, CA, USA). After reconstituted by adding basic buffer (1% NH₄OH. 60–70 μL) to β-amyloid (1–40) (1 mg) aggregation of Aβ_1–40 was carried out by gently dissolving Aβ_1–40 (0.25 mg/mL) in a PBS buffer (pH 7.4). The solution was incubated at 37 °C for 72 h with gentle and constant shaking. Binding studies were carried out according to the procedure described previously with some modifications using [¹²⁵I]TZDM as the radiolabeled standard [24]. Briefly, the competition binding assays were performed by mixing Aβ_1–40 aggregates (100 µL), and [¹²⁵I]TZDM (100 µL) in appropriate concentration (0.02 nM, diluted in 10% EtOH), test ligand (10⁻⁵–10⁻¹⁰ M, 100 µL) and PBS (0.02 M, pH 7.4, 700 µL) in a final volume of 1 mL. The mixture was incubated at 37 °C for 2 h. Then the bound and free radioactivities were separated by vacuum filtration through Whatman GF/B glass filters via a Brandel Mp-48T cell harvester followed by 3 × 4 mL washes with PBS (0.02 M, pH 7.4, 4 °C) containing 10% ethanol at room temperature. Filters containing the bound ¹²⁵I ligand were counted in a γ-counter (WALLAC Wizard 1470, PerkinElmer Life Sciences, Waltham, MA, USA) with 75% counting efficiency. The IC₅₀ values were determined using GraphPad Prism 5.0, and those for the inhibition constant (K_i) were calculated using the Cheng Prusoff equation: K_i = IC₅₀/(1 + [L]/K_d) [25].

Paraffin-embedded brain sections of transgenic model mouse (8 µm, APP/PS1, 12 months, male) were used for the in vitro fluorescent staining of amyloid plaques. The brain sections were deparaffinized with xylene, ethanol and distilled water. After immersion in PBS (0.02 M, pH 7.4) for 30 min, the brain sections were incubated with 20% ethanol solution (1 µM) of compound 8 for 10 min. The localization of plaques was confirmed by staining with thioflavin-S on the adjacent sections. Finally, the sections were washed with 40% ethanol and PBS (0.02 M, pH 7.4). Fluorescent observation was performed by a Stereo Discovery V12 instrument (Zeiss, Oberkochen, Germany) equipped with a LP 505 filter set (excitation, 405 nm).

The brain sections mentioned above were incubated with [¹²⁵I]8 (5 µCi/100µL) for about 20 min at room temperature. Then the sections were washed with saturated Li₂CO₃ in 40% EtOH for 3 min and 40% EtOH for 3 min, followed by rinsing with water for 30 s. After drying, the ¹²⁵I-labeled sections were exposed to phosphorus film for 8 h and then scanned with the phosphor imaging system (Cyclone, Packard) at the resolution of 600 dpi. The presence and localization of plaques were confirmed by the fluorescent staining with thioflavin-S on the same sections using a Stereo Discovery V12 (Zeiss) instrument equipped with a LP 505 filter set (excitation: 405 nm).

In vivo biodistribution studies were performed in KunMing normal mice (female, average weight 18–22 g) and in accordance with the national laws related to the care and experiments on laboratory animal. A saline solution (100 µL) containing [¹²⁵I]8 or [¹²⁵I]17 (1 µCi) was injected directly into the tail vein of mice. The mice (n = 5 for each time point) were sacrificed at designated time points post-injection. The organs of interest were removed and weighed, and the radioactivity was counted with an automatic γ-counter (WALLAC Wizard 1470).