PET Imaging of VMAT2 with the Novel Radioligand [18F]FE-DTBZ-d4 in Nonhuman Primates: Comparison with [11C]DTBZ and [18F]FE-DTBZ

The vesicular monoamine transporter type 2 (VMAT2) is believed to be responsible for the uptake of monoamines into the vesicles of the synaptic terminals. Two VMAT2 radioligands [11C]DTBZ and [18F]FP-DTBZ have been used to assess the degree of nigrostriatal deficit in Parkinson’s disease (PD) using positron emission tomography (PET). [18F]FE-DTBZ-d4, the nondeuterated analogue of [18F]FE-DTBZ showed similar imaging properties with better stability against defluorination. Therefore, [18F]FE-DTBZ-d4 draws attention to be investigated as an imaging marker for VMAT2 in the brain. The aim of this study was to investigate the brain kinetics and quantification of [18F]FE-DTBZ-d4 in nonhuman primates (NHPs), with comparison to [11C]DTBZ and [18F]FE-DTBZ. Radiolabeling was successfully achieved either by one-step 11C-methylation or by a two-step fluorine-18 nucleophilic substitution reaction. The stability and radiochemical yield were analyzed with high-performance liquid chromatography (HPLC). Three female cynomolgus monkeys were included in the study and underwent a total of 12 positron emission tomography (PET) measurements. Each monkey was examined with each tracer. In addition, two pretreatment and one displacement PET measurements with tetrabenazine (2.0 mg/kg) were performed for [18F]FE-DTBZ-d4. All PET measurements were conducted using a high-resolution research tomograph (HRRT) system. Radiometabolites were measured in monkey plasma using gradient radio-HPLC. [18F]FE-DTBZ-d4 (SUV: 4.28 ± 1.01) displayed higher brain uptake compared to both [18F]FE-DTBZ (SUV: 3.43 ± 0.54) and [11C]DTBZ (SUV: 3.06 ± 0.32) and faster washout. Binding potential (BPND) values of [18F]FE-DTBZ-d4 in different brain regions (putamen: 5.5 ± 1.4; caudate: 4.4 ± 1.1; midbrain: 1.4 ± 0.4) were higher than those of [11C]DTBZ and [18F]FE-DTBZ. [18F]FE-DTBZ showed faster radiometabolism in plasma compared to [11C]DTBZ and [18F]FE-DTBZ-d4. [18F]FE-DTBZ-d4 is a suitable radioligand for quantification of VMAT2 in the nonhuman primate brain, with better imaging properties than [11C]DTBZ and [18F]FE-DTBZ. A preliminary comparison suggests that [18F]FE-DTBZ-d4 has increased stability against defluorination compared to the nondeuterated analogue.


■ INTRODUCTION
The vesicular monoamine transporter 2 (VMAT2) is an integral membrane protein, previously known as the synaptic vesicular monoamine transporter, mainly present in neuronal cells of the central, peripheral, and enteric nervous system. 1 VMAT2 transports monoamines such as dopamine, norepinephrine, serotonin, and histamine from cellular cytosol into synaptic vesicles. 2 VMAT2 is also required for the vesicular release of the neurotransmitter gamma-aminobutyric acid (GABA) in nigrostriatal and mesolimbic dopamine neurons. 3 One of the most common neurodegenerative movement disorder, Parkinson's disease (PD), is clinically characterized by akinesia, resting tremor, and rigidity. The degeneration of the dopaminergic neurons of the substantia nigra pars compacta (SNc) in PD leads to the loss of nigrostriatal terminals and to the reduction of dopamine levels in the striatum. 4,5 There are several evidence that link VMAT2 to dopaminergic cell loss in PD. Previous studies have reported that the dysfunction of VMAT2 can evoke cytoplasmic dopamine accumulation, which leads to dopaminergic neuron death. 6 Other studies have shown that protein expression levels of VMAT2 were significantly reduced in PD patients 7 and that an increased VMAT2 level or function might protect against the development of PD. 8,9 Therefore, striatal VMAT2 is considered as a presynaptic marker of dopamine terminal loss in PD. 10 Several dihydrotetrabenazine (DTBZ) derivatives have been labeled with 11 C/ 18 F and developed as PET radioligands.
Affinity of dihydrotetrabenazine (DTBZ) toward VMAT2 binding is stereospecific, 11 and the (+)-enantiomer showed 1000-fold better binding affinity (K i = 0.97 ± 0.48 nM) over the (−)-enantiomer (K i = 2.2 ± 0.3 μM). 12,13 Therefore, PET measurements have been performed mostly with the (+)-enantiomer of DTBZ derivatives. Two VMAT2 radioligands, [ 11 C]-(+)-DTBZ 14,15 and [ 18 F]FP-(+)-DTBZ ([ 18 F]AV-133), 16−18 have already been used in several studies in patients with PD. 18 F-fluoroethyl-(+)-dihydrotetrabenazine ( 18 F-FE-DTBZ) has not been evaluated in humans, but a previous study in rats showed a relatively poor striatum-tocerebellum ratio. 19 VMAT2 is a protein that is also expressed in insulinproducing beta cells in pancreatic islets and might serve as a potential marker for the assessment of the beta cell mass. 20 VMAT2-specific [ 11 C]-(+)-DTBZ has been shown to be a beta cell mass (BCM) biomarker with potential to distinguish between healthy and diabetic subjects longitudinally, however, with some difficulties such as less affinity for VMAT2 and higher fat solubility, which translates into high nonspecific binding. 21−23 In addition, 18 F has a longer half-life than 11 C, making it suitable for a wider range of applications. In an effort to develop a 18 F-labeled VMAT2 radioligand for imaging the beta cell mass, our group investigated [ 18 F]FE-(+)-DTBZ in a large piglet model. 24 The radioligand was found to be metabolized extensively by in vivo defluorination. To improve the in vivo stability, a deuterated analogue [ 18 F]FE-(+)-DTBZ-d4 of [ 18 F]FE-DTBZ 25 has been recently developed and investigated in pigs for in vivo imaging of VMAT2 in insulinsecreting beta cells. PET/CT studies showed that [ 18 F]FE-(+)-DTBZ-d4 had lower bone uptake than [ 18 F]FE-(+)-DTBZ. This property is of advantage for brain imaging, considering that bone uptake might interfere with the measurement of specific binding of 18 F-radioligands in the central nervous system (CNS). 26 For this reason, [ 18 F]FE-(+)-DTBZ-d4 was considered a potential imaging marker of VMAT2 in the brain. The in vivo properties of [ 18 F]FE-(+)-DTBZ-d4 remain, however, to be investigated.
The current study was, therefore, designed to investigate the brain uptake, the kinetic properties, the radiometabolism, and the noninvasive quantification of [ 18  ■ RESULTS AND DISCUSSION [ 11 C]DTBZ was synthesized in a one-step via O-methylation reaction where [ 11 C]CH 3 OTf was used as a methylating agent. The incorporation yields of [ 11 C]CH 3 OTf into [ 11 C]-DTBZ were high (80−90%) for all productions. The total time of the radiosynthesis was 28−30 min, and the range of the molar activity (MA) was >160 GBq/μmol at the end of the synthesis (EOS). The radiochemical purity was more than 98% for all productions at the EOS. The formulated solution of all three radioligands was found to be radiochemically stable for up to 60 min and the radiochemical purity was >98%.
The radiosynthesis of [ 18 F]FE-DTBZ-d4 and [ 18 F]FE-DTBZ was accomplished via the two steps. 18 F-fluoroethylation reaction: The alkylating agents, [ 18 F]FEtBr-d4 and [ 18 F]FEtBr, were generated from 2-bromo-1,2-tetra-2 H-ethyl tosylate and bromoethyl tosylate, respectively, by a one-step nucleophilic 18 F-fluorination reaction as described previously. 27 The purified [ 18 Figure 3. All three radioligands cross the blood−brain barrier and bind rapidly with a time to peak on  3.06 ± 0.32%) and also the highest peak-to-120 min ratio (3.4 ± 0.5) in comparison to the other two radioligands ( Table 2). The highest radioactivity concentration was observed in putamen followed by caudate and midbrain; the lowest uptake was observed in the cerebellum for all three radioligands, which is in accordance with the literature (Figure 4) showed higher uptake in all brain regions of NHPs, further investigations were accomplished to observe specific VMAT2 binding uptake by performing two pretreatment and one displacement experiments using the VMAT2-specific compound tetrabenazine. After pretreatment with tetrabenazine (2 mg/kg, 15 min prior to radioligand injection), the uptake of [ 18 F]FE-DTBZ-d4 in caudate, putamen, and midbrain was decreased almost to the level of the cerebellum and occipital cortex ( Figure 5A). During the displacement experiment, the administration of tetrabenazine (2 mg/kg, 25 min. after radioligand injection) determined a clear reduction of [ 18 F]FE-DTBZ-d4 uptake in caudate, putamen, and midbrain, almost to the level of the cerebellum and occipital cortex ( Figure 5B), demonstrating the reversibility of the binding.
The radioactivity in venous blood samples, plasma, and remaining protein pellet after deproteinization of plasma with acetonitrile were measured by a well counter. The recovery of radioactivity from plasma into acetonitrile was more than 90%. The plasma obtained from venous blood samples taken at various time points following the injection of radioligands was analyzed by reverse-phase radio-HPLC. All of the detected radiometabolites were less lipophilic than the parent radioligands. For all three radioligands, two detected unidentified radiometabolites were eluted with an Rt of 2.3 and 4.7 min, whereas the parent radioligands were eluted with an Rt over 7 min. In the case of [ 18 Figure 6).
This study was designed specifically to evaluate the in vivo properties of [ 18 F]FE-DTBZ-d4 as a potential radioligand for imaging and quantification of VMAT2 in the nonhuman primate brain. The kinetic properties and in vivo radiometabolism of [ 18 F]FE-DTBZ-d4 were compared to the wellestablished VMAT2 radioligand [ 11 C]DTBZ and to the nondeuterated analogue [ 18 F]FE-DTBZ.
The results of the study suggest that [ 18 F]FE-DTBZ-d4 has better in vivo properties than [ 11 C]DTBZ and [ 18 F]FE-DTBZ, in terms of higher brain uptake, earlier peak equilibrium, and a higher peak-to-late uptake ratio. [ 18 F]FE-DTBZ-d4 was also found to be more stable in plasma, likely due to the incorporation of four atoms of deuterium, a procedure known to decrease the in vivo metabolism of radioligands. 26   The main purpose of developing a deuterated analogue was to improve in vivo stability and to decrease defluorination, thereby decreasing bone uptake. The visual interpretation of [ 18 F]FE-DTBZ PET images does not suggest, however, the presence of obvious bone uptake. Therefore, it seems that the main advantages of deuterium incorporation are improved in vivo stability and increased brain uptake of [ 18 F]FE-DTBZ-d4.
The objective of this study was to compare the three radioligands described above and to evaluate the effects of the incorporation of deuterium on the in vivo properties of
Synthesis of [ 11 C]DTBZ. [ 11 C]Methyl iodide ([ 11 C]CH 3 I) was produced according to a previously published method. 29 In short, [ 11 C]CH 4 was produced in the cyclotron and collected in a Porapak Q trap cooled with liquid nitrogen. [ 11 C]CH 4 was released from the trap by heating with pressurized air and subsequently mixed with iodine vapors at 60°C followed by a radical reaction at 720°C. After the reaction, [ 11 C]CH 3 I was collected in a Porapak Q trap at room temperature (RT) and was released by heating at 180°C. [ 11 C]CH 3 OTff was produced by the online transfer of [ 11 C]CH 3 I through a glass column packed with silver triflate at 165°C.
The radiosynthesis of [ 11 C]DTBZ was obtained by trapping [ 11 C]CH 3 OTf at RT in a reaction vessel containing precursor 9-Odesmethyl-(+)-DTBZ (0.4−0.6 mg) and sodium hydroxide (NaOH) (0.5 M, 6 μL) in acetone (300 μL). After the end of trapping, the reaction mixture was diluted with 500 μL of sterile water before injecting into the built-in high-performance liquid chromatography (HPLC) system to purify the labeled compound. The HPLC system was equipped with a semipreparative reverse-phase μ-Bondapak HPLC column (C18, 7.8 Ø × 300 mm, 10 μm, Waters) for purification. The column outlet was coupled with a UV absorbance detector (λ = 254 nm) followed by a GM tube for radioactivity detection. HPLC mobile-phase CH 3 CN/10 mM H 3 PO 4 (15/85) with a flow rate of 6 mL/min led to a complete separation of [ 11 C]DTBZ (retention time = 9 min) from impurities. The radioactive fraction corresponding to pure [ 11 C]DTBZ was collected from HPLC and evaporated to dryness followed by formulation in phosphate-buffered saline (6 mL was purified by distillation at 90°C under nitrogen flow followed by passing through a tube filled with P 2 O 5 and trapped in a second reaction vessel (1 mL) at −5°C containing the precursor 9-Odesmethyl-(+)-DTBZ (2.0−2.5 mg, 6.55−8.19 μmol) and NaOH (15 μL, 5 M in water) in N,N-dimethylformamide (DMF, 500 μL).
The reaction mixture was heated at 110°C for 5 min to synthesize [ 18 F]FE-DTBZ/[ 18 F]FE-DTBZ-d4, followed by cooling to RT and was diluted with water to a total volume of 1 mL before injecting into a built-in HPLC system, which was equipped with a semipreparative reverse-phase μ-Bondapak HPLC column (C18, 7.8 Ø × 300 mm, 10 μm, Waters) for purification. The column outlet was coupled with a UV absorbance detector (λ = 254 nm) followed by a GM tube for radioactivity detection. HPLC mobile-phase CH 3 CN/10 mM H 3 PO 4 (15/85) with a flow rate of 5 mL/min led to a complete separation of pure . The radiochemical purity, identity, stability, and molar activity (MA) were analyzed using an HPLC system, which included an ACE RP column (C18, 3.9 Ø × 250 mm, 5 μm particle size), a Merck-Hitatchi L-7100 Pump, and an L-7400 UV absorbance detector (λ = 254 nm) coupled to a radioactive detector (b-flow, Beckman, Fullerton, CA). CH 3 CN/50 mM H 3 PO 4 (15/85) was used as the HPLC mobile phase with a flow rate of 3 mL/min to elute the product. The effluent was monitored with an UV absorbance detector (ƛ = 254 nm) coupled to a radioactive detector (b-flow, Beckman, Fullerton, CA) and the product was eluted with a retention time (Rt) of 4−5 min for The identity of all radioligands was confirmed by coinjection with the authentic nonradioactive reference standards.
MA was calibrated for UV absorbance (λ = 254 nm) response per mass of ligand and calculated as the radioactivity of the radioligand (GBq) divided by the amount of the associated carrier substance (μmol). Each sample was analyzed three times and compared to a reference standard.
PET Measurements in Cynomolgus Monkeys. The Animal Ethics Committee of the Swedish Animal Welfare Agency (Dnr N185/14) approved the study protocol. All of the experiments were performed according to the "Guidelines for planning, conducting and documenting experimental research" (Dnr 4820/06-600) of Karolinska Institutet. 30,31 The NHPs were housed in the Astrid Fagraeus Laboratory (AFL) of the Swedish Institute for Infectious Disease Control, Solna, Sweden. Anesthesia was induced by intramuscular injection of ketamine hydrochloride (approximately 10 mg/kg, Ketaminol vet. Intervet) and maintained by the administration of a mixture of sevoflurane, oxygen, and medical air after endotracheal intubation. The head was immobilized using a fixation device. 32 Three female cynomolgus monkeys (NHP1, NHP2, and NHP3) (weight of 5.8−6.95 kg) were studied in 10 different experimental days for a total of 12 PET experiments. All PET experiments were performed using a high-resolution research tomograph (HRRT) PET scanner (Siemens Molecular Imaging, Knoxville, TN). A 6 min transmission scan was performed every time before the injection of the radioligands using a single 137 Cs source. List-mode data were reconstructed with a series of 35 frames (10 s × 4, 20 s × 4, 1 min × 3, 3 min × 7, and 6 min × 16) for 11 C and (10 s × 4, 20 s × 4, 1 min × 4, 3 min × 7, 6 min × 16, and 12 min × 5) for 18 F, using the ordinary Poisson-3D-ordered subset expectation maximization (OP-3D-OSEM) algorithm with 10 iterations and 16 subsets including modeling of the point spread function (PSF), correction for attenuation, random, and scatter. The resolution of the reconstructed images was 1.5 mm in full-width at half maximum. 33 All three monkeys underwent two PET measurements on the same day at once. for the measurement of protein binding, blood, and plasma radioactivity and radiometabolite analysis. Blood samples were collected at different time points: 5 min before the injection of the corresponding radioligand followed by 2. 5,15,30,45,60,90, and 120 min after the injection. In the case of [ 18 F]FE-DTBZ-d4 and [ 18 F]FE-DTBZ, venous blood sampling was also performed at 180 min after the injection due to the longer half-life of fluorine-18.
The regions of interest (ROIs) were delineated manually on the MRI images of each NHP for putamen, caudate, midbrain, occipital cortex, and cerebellum. The MRI of all of the individual NHPs was coregistered to summed PET images of the whole PET measurement. The time−activity curves of NHP brain regions were generated from dynamic PET data with the application of the coregistration parameters to ROIs. Quantification was performed using the Logan graphical analysis, and the cerebellum was used as a reference region. The outcome measure was the binding potential (BP ND ).
Radiometabolite Analysis. Radiometabolite analysis was performed following a method reported elsewhere. 34  for 2 min to obtain the plasma (0.5 mL). The plasma was mixed with 1.4 times volume of acetonitrile followed by centrifugation at 2000g for 4 min. The extract was separated from the pellet and was diluted with water before injecting into the HPLC equipped with a reverse-phase μ-Bondapak HPLC column, (300 mm × 10 mm I.D). An Agilent binary pump (Agilent 1200 series) coupled to a manual injection valve (7725i, Rheodyne), 1−3.0 mL loop, and a radiation detector (Oyokoken, S-2493Z) placed in a shield of 50 mm thick lead was used for radiometabolite measurements. Chromatographic software (ChemStation Rev. B.04.03; Agilent) was used for the data collection and control of the LC system. The accumulation time of the radiation detector was 10 s. To calculate the recovery of radioactivity from the system, an aliquot (2 mL) of the eluate from the HPLC column was measured and divided by the amount of the total injected radioanalyte.
Plasma Protein Binding. Plasma protein binding of all three radioligands was measured in duplicate in baseline conditions. NHP blood plasma (500 μL) and a PBS solution (pH 7.4, KCl 0.2 mg, KH 2 PO 4 0.2 mg, Na 2 HPO 4 1.42 mg, NaCl 8 mg in 1 mL) were mixed with the respective radioligand (50 μL, approx. 20 MBq in PBS). The plasma mixture was incubated for 10 min at room temperature and a small portion (20 μL) from each incubation mixture was measured in a well counter. A portion (200 μL) from all individual incubation mixtures was pipetted out into ultrafiltration tubes (Millipore Centrifree YM-30) and centrifuged at 3000 rpm for 15 min. Samples (20 μL) from each filtrate were counted in a well counter. The following formula was used to calculate the protein-bound fraction. where C pla (total) is the radioactivity concentration of the incubation mixture in plasma, C PBS (total) is the radioactivity concentration of the incubation mixture in PBS, and C pla (filtrate) and C pla (filtrate) are the radioactivity concentrations from filtrate samples.