The new PET imaging agent [11C]AFE is a selective serotonin transporter ligand with fast brain uptake kinetics

Nuclear Medicine and Biology 31 (2004) 983 – 994 www.elsevier.com/locate/nucmedbio

The new PET imaging agent [11C]AFE is a selective serotonin transporter ligand with fast brain uptake kinetics Zhihong Zhua, Ningning Guoa, Raj Narendrana, David Erritzoea, Jesper Ekelunda, Dah-Ren Hwanga,b,c, Sung-A Baec, Marc Laruellea,b,c, Yiyun Huanga,b,c,* a

Department of Psychiatry, Columbia University College of Physicians and Surgeons, New York, NY 10032, USA Department of Radiology, Columbia University College of Physicians and Surgeons, New York, NY 10032, USA c New York State Psychiatric Institute, New York, NY 10032, USA Received 24 April 2004; received in revised form 30 June 2004; accepted 22 July 2004

b

Abstract A new positron emission tomography (PET) radioligand for the serotonin transporter (SERT), [11C]2-[2-[[(dimethylamino)methyl]phenyl]thio]-5-(2-fluoroethyl)phenylamine ([11C]AFE, 12), was synthesized and evaluated in vivo in rats and baboons. [11C]AFE (12) was prepared from its monomethylamino precursor 11 by reaction with high specific activity [11C]methyl triflate. Radiochemical yield was 32F17% based on [11C]methyl triflate (n = 6) and specific activity was 1670F864 Ci/mmol at end of synthesis (EOS, n = 6). Binding assays indicated that AFE displays high affinity for SERT (K i = 1.80 nM for hSERT) and lower affinity for norepinephrine transporter (K i = 946 nM for hNET) or dopamine transporter (K i N 10,000 nM for hDAT). In addition, AFE displays negligible binding affinities for other serotonin and dopamine receptors, indicating an excellent binding selectivity in vitro. Biodistribution studies in rats indicated that [11C]AFE enters the brain readily and localizes in regions known to contain high concentrations of SERT, such as the thalamus, hypothalamus, frontal cortex and striatum. Moreover, such binding in SERT-rich brain regions is reduced significantly by pretreatment with either citalopram or the cold compound itself, but not by nisoxetine or GBR 12935, thus demonstrating that [11C]AFE binding in the rat brain is saturable, specific and selective for the SERT. Imaging experiments in baboons indicated that the uptake pattern of [11C]AFE is consistent with the known distribution of SERT in the baboon brain, with high levels of radioactivity detected in the midbrain and thalamus, moderate levels in the hippocampus and striatum and low levels in the cortical regions. The uptake kinetics of [11C]AFE in the baboon brain is rapid, with activity in the midbrain and thalamus peaking at 15 – 40 min postinjection. Pretreatment of the baboon with citalopram (4 mg/kg) 20 min before radioactivity injection reduced the binding of [11C]AFE in all SERT-containing brain regions to the level in the cerebellum. Kinetic analysis revealed that in all brain regions examined, [11C]AFE specific-to-nonspecific partition coefficients (V 3W) are similar to those of [11C]McN5652 and [11C]2-[2-[[(dimethylamino)methyl]phenyl]thio]-5-fluorophenylamine ([11C]AFA), but lower than those of [11C]2-[2[[(dimethylamino)methyl]phenyl]thio]-5-fluoromethylphenylamine ([11C]AFM) or [11C]DASB. In summary, [11C]AFE appears to be a PET radioligand with fast brain uptake kinetics and can be used for the visualization and quantification of SERT in vivo. D 2004 Elsevier Inc. All rights reserved. Keywords: Serotonin transporter; PET; Radioligand; Synthesis

1. Introduction The neurotransmitter serotonin is believed to play an important role in the regulation of many physiological functions such as mood, appetite, sleep, pain and aggressive behavior [1]. The serotonin transporter (SERT), located on the cell bodies and terminals of the serotonin neurons, is a marker of serotonin innervation [2]. In the human brain, the * Corresponding author. New York State Psychiatric Institute, Box 31, New York, NY 10032, USA. Tel.: +1 212 543 6629; fax: +1 212 568 6171. E-mail address: [email protected] (Y. Huang). 0969-8051/$ – see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.nucmedbio.2004.07.003

regional density of SERT is high in the midbrain, thalamus, hypothalamus and striatum, moderate in the structures of the limbic system (hippocampus, amygdala and anterior cingulate cortex), low in the neocortical areas and negligible in the cerebellum [3–7]. Alterations of SERT densities have been noted in a number of neuropsychiatric conditions, including major depression, anxiety disorders, schizophrenia, drug abuse, alcoholism, eating disorders and Alzheimer’s and Parkinson’s diseases [8–12]. Most frequently prescribed antidepressants are selective serotonin reuptake inhibitors (SSRIs), which exert their antidepressant activity through

984

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994

inhibition of the SERT [13–16]. Many of the psychostimulants with abuse potential, such as cocaine and amphetamine, are blockers of the monoamine transporters including SERT, dopamine transporter (DAT) and norepinephrine transporter (NET) [16,17]. Therefore, imaging the regional brain distribution of SERT in vivo provides an important tool to study the role of the serotonin system in the pathophysiology and treatment of neuropsychiatric disorders. Until recently, the most widely used positron emission tomography (PET) radioligand for the in vivo investigation of SERT has been [11C]McN5652 [18]. However, as an imaging agent, [11C]McN5652 presents several limitations, including high levels of nonspecific binding, low specific-tononspecific binding ratios and slow brain kinetics [19–22]. In the last few years, a new series of SERT radiotracers has emerged from the substituted diarylsulfide class of compounds. [123I]2-[2-[[(dimethylamino)methyl]phenyl]thio]-5iodobenzyl alcohol ([123I]IDAM) was first reported as a new high-affinity SERT ligand in this series, followed by [123I]2[2-[[(dimethylamino)methyl]phenyl]thio]-5-iodophenylamine ([ 123 I]ADAM) [23,24]. Both [ 123 I]IDAM and [123I]ADAM have been shown to be appropriate radioligands for the imaging of SERT using SPECT, and [123I]ADAM has been validated in human studies [25–29]. Another radioligand of the same class, [11C]DASB, was introduced as a new PET ligand for SERT imaging in humans [30–33]. [11C]DASB offered improvements over [11C]McN5652, in that it displays better signal-to-noise ratios and faster brain uptake kinetics [30,34 –36]. In addition, other ligands in the same structural class have also been synthesized and evaluated as potential PET ligands for SERT. These include [ 11 C]2-[2-[[(dimethylamino)methyl]phenyl]thiol]-5methoxyphenylamine ([11C]DAPP), [11C]ADAM, [11C]5bromo-2-[2-[[(dimethylamino)methyl]phenyl]thio]phenylamine ([11C]DAPA), [11C]2-[2-[[(dimethylamino)methyl]phenyl]thio]-5-methylphenylamine ([ 11 C]MADAM), [11C]2-[2-[[(dimethylamino)methyl]phenyl]thio]-5-fluoromethylphenylamine ([11C]AFM) and [11C]2-[2-[[(dimethylamino)methyl]phenyl]thio]-5-fluorophenylamine ([11C]AFA) (Fig. 1) [31,32,34,37–41]. All these new entries have been shown to be selective SERT ligands, with varying specific-to-nonspecific binding ratios in vivo and differing brain uptake kinetics. In our laboratories, we have been interested in the development of radioligands that can be labeled with

either 18F or 11C in the same molecule, such as AFA, AFM and [11C]2-[2-[[(dimethylamino)methyl]phenyl]thio]5-(2-fluoroethyl)phenylamine (AFE) (Fig. 1) [42]. One goal of these efforts was to search for PET ligands that can be used for imaging not only high SERT density brain regions, but also low SERT density regions, which are important in neuropsychiatric disorders. A second goal was to seek appropriate 18F-labeled PET ligands for wider applications, including off-site imaging applications. We have previously reported the evaluation of [11C]AFM and demonstrated that it provides, among the available PET ligands for SERT, the highest signal-to-noise ratios in imaging studies in nonhuman primates [34,40]. In addition, [11C]AFM also distinguishes itself in that it provides sufficient signal-to-noise ratios in various cortical regions in the baboon brain, thus raising the possibility of imaging low SERT density regions with this radioligand [34,40]. However, the kinetics of [11C]AFM in the baboon brain is relatively slow, thus raising concerns about a potentially long scan time in human studies. As a result, there are some incentives for the development of SERT radioligands that combine the feasibility for 18F labeling, an imaging quality comparable to that of [11C]AFM and a faster brain uptake kinetics. We have reported the synthesis and evaluation of [11C]AFA as a PET ligand with fast brain uptake kinetics [41], while Shiue et al. [43,44] has disclosed the preliminary evaluation of [18F]AFA. In this article, we report the synthesis, in vitro characterization and in vivo pharmacological and pharmacokinetic evaluation of [11C]2-[2-[[(dimethylamino)methyl]phenyl]thio]-5-(2-fluoroethyl)phenylamine ([11C]AFE, 12). The imaging properties provided by [11C]AFE are then contrasted with those afforded by [ 11 C]McN5652, [11C]DASB, [11C]AFA and [11C]AFM for comparative evaluation. 2. Materials and methods 2.1. Chemistry 2.1.1. General All reagents were purchased from commercial suppliers and used without further purification unless otherwise stated. When reactions were worked up by extraction with dichloromethane (CH2Cl2), ethyl acetate (EtOAc) or ethyl

Fig. 1. Representative radioligands for the SERT.

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994

ether (Et2O), organic solutions were dried with anhydrous MgSO4 and concentrated with a rotary evaporator under reduced pressure. Reactions requiring anhydrous conditions were carried out in oven-dried glassware under an inert atmosphere of nitrogen. Anhydrous Et2O and THF were prepared by distilling over Na/benzophenone. Melting points were determined on a Thomas Hoover Unimelt apparatus and are uncorrected. Column chromatography was performed using silica gel 60, 230–400 mesh (Aldrich). Unless otherwise noted, 1H spectra were recorded on a Bruker HF400 spectrometer at 400 MHz, with CDCl3 as solvent and tetramethylsilane as the internal standard (0 ppm). Mass spectra were run on an HRX 70S spectrometer. Elemental analyses were performed at Midwest Microlab (Indianapolis, IN). 2.1.2. 2-[4-[(Methoxycarbonyl)methyl]-2-(nitrophenyl)thio]benzoic acid (3) A mixture of compound 1 (4.47 g, 16.3 mmol), thiosalicylic acid (2, 2.51 g, 16.3 mmol), Cu powder (0.31 g, 4.89 mmol) and K2CO3 (5.18 g, 37.5 mmol) in DMF (100 mL) was heated at 858C overnight, cooled to room temperature and poured to ice water. The mixture was filtered through a layer of Celite. The filtrate was made acidic with 6 N HCl and extracted with CH2Cl2 (3100 mL). The combined organic layers were washed with H2O, dried and concentrated. The crude product was purified on silica gel column with MeOH/CH2Cl2 (5:95) to provide compound 3 (3.18 g, 56.2%) as a yellowish solid, mp = 135– 1368C. 1H NMR (d): 3.67 (s, 2H), 3.72 (s, 3H), 7.06 (d, 1H, J = 8.3 Hz), 7.34 (d, 1H, J = 8.3 Hz), 7.42 (d, 1H, J = 7.7 Hz), 7.45–7.54 (m, 2H), 8.06 (s, 1H), 8.08 (d, 1H, J = 7.5 Hz). Anal. calcd. for C16H13NO6S: C, 55.33; H, 3.77; N, 4.03. Found: C, 55.30; H, 3.76; N, 4.06. 2.1.3. Methyl 2-[4-[2-[[(dimethylamino)carbonyl] phenyl]thio]-3-nitrophenyl]acetate (4) A solution of compound 3 (3.85 g, 10.7 mmol) in thionyl chloride (30 mL) was heated at 708C for 3 h. Excess thionyl chloride was removed. The residue was then redissolved in THF (30 mL). To this solution was added N,N-dimethylamine hydrochloride (4.39 g, 53.5 mmol) and K2CO3 (7.34 g, 53.5 mmol). The reaction mixture was stirred overnight at room temperature, diluted with H2O and extracted with CH2Cl2 (430 mL). The combined organic layers were washed with H2O, dried and concentrated. Column chromatography of the crude products on silica gel and elution with EtOAc/hexane (1:2) afforded compound 4 (2.1 g, 52.5%) as a yellowish solid, mp = 124–1258C. 1H NMR (d): 2.85 (s, 3H), 3.05 (s, 3H), 3.62 (s, 2H), 3.69 (s, 3H), 6.89 (d, 1H, J =8.4 Hz), 7.29 (dd, 1H, J =1.9, 8.4 Hz), 7.42 (dd, 1H, J = 1.5, 7.6 Hz), 7.48 (dd, 1H, J =1.5, 7.6 Hz), 7.50– 7.60 (m, 2H), 8.12 (d, 1H, J = 1.9 Hz). Anal. calcd. for C18H18N2O5S: C, 57.74; H, 4.85; N, 7.48. Found: C, 57.56; H, 4.78; N, 7.43.

985

2.1.4. Methyl 2-[4-[2-[[(methylamino)carbonyl] phenyl]thio]-3-nitrophenyl]acetate (5) In an analogous manner, compound 5 was prepared from compound 3 and N-methylamine hydrochloride in 62.2% yield as a yellowish solid, mp = 82–848C. 1H NMR (d): 2.87 (s, 3H), 3.64 (s, 2H), 3.70 (s, 3H), 6.33 (br s, 1H), 6.86 (dd, 1H, J = 1.8, 8.4 Hz), 7.29 (dd, 1H, J = 1.8, 8.4 Hz), 7.45– 7.65 (m, 3H), 7.77 (dd, 1H, J =1.5, 7.5 Hz), 8.13 (d, 1H, J = 1.6 Hz). Anal. calcd. for C17H16N2O5Sd 0.33H2O: C, 55.73; H, 4.58; N, 7.65. Found: C, 55.62; H, 4.52; N, 7.38. 2.1.5. 2-[4-[2-[[(Dimethylamino)methyl]phenyl]thio]3-nitrophenyl]ethanol (6) Compound 4 (1.78 g, 4.76 mmol) was dissolved in THF (25 mL), and the solution cooled to 08C. To the solution was introduced the BH3–THF complex (23.8 mL, 1 M solution in THF, 23.8 mmol) via a syringe. The reaction mixture was refluxed for 2 h, then stirred overnight at room temperature. Concentrated HCl was added, and the solvent was removed. The aqueous phase was diluted with H2O (20 mL) and heated to reflux for 20 min. After cooling down to room temperature, the mixture was adjusted to pH 8 with 10% NaHCO3 and extracted with CH2Cl2 (415 mL). The combined organic layers were dried and concentrated. Column chromatography on silica gel and elution with MeOH/CH2Cl2 (5:95) afforded product 6 (1.40 g, 88.6%) as yellowish thick oil. 1H NMR (d): 2.20 (s, 6H), 2.87 (t, 2H, J = 6.4 Hz), 3.55 (s, 2H), 3.87 (t, 2H, J =6.4 Hz), 6.63 (d, 1H, J =8.3 Hz), 7.20 (dd, 1H, J =1.9, 8.3 Hz), 7.33 (t, 1H, J = 7.6 Hz), 7.48 (t, 1H, J = 7.6 Hz), 7.54 (d, 1H, J =7.6 Hz) 7.66 (d, 1H, J = 7.6 Hz), 8.12 (d, 1H, J = 1.7 Hz). Anal. calcd. for C17H20N2O3Sd 0.33H2O: C, 60.33; H, 6.16; N, 8.28. Found: C, 60.68; H, 6.07; N, 8.28. 2.1.6. 2-[4-[2-[[(Methylamino)methyl]phenyl]thio]-3nitrophenyl]ethanol ( 7) In an analogous manner, compound 7 was prepared from compound 5 in 62.2% yield as yellowish oil. 1H NMR (d): 2.39 (s, 3H), 2.86 (t, 2H, J = 6.4 Hz), 3.82 (s, 2H), 3.86 (t, 2H, J =6.4 Hz), 6.62 (d, 1H, J = 8.3 Hz), 7.20 (dd, 1H, J = 1.9, 8.4 Hz), 7.36 (dt, 1H, J = 1.3, 7.5 Hz), 7.50 (dt, 1H, J = 1.2, 7.5 Hz), 7.55–7.63 (m, 2H), 8.14 (d, 1H, J = 1.8 Hz). Anal. calcd. for C16H18N2O3Sd 0.33H2O: C, 59.24; H, 5.80; N, 8.64. Found: C, 59.47; H, 5.67; N, 8.59. 2.1.7. N,N-Dimethyl-2-[4-(2-fluoroethyl)-2[(nitrophenyl)thio]]benzylamine (8) To a solution of compound 6 (160 mg, 0.53 mmol) in CH2Cl2 (20 mL) was added [bis(2-methoxyethyl)amino]sulfur trifluoride (117 lL, 0.64 mmol). The mixture was stirred at room temperature for 2 h and washed with 10% Na2CO3 (210 mL). The organic layer was dried and concentrated. The residue was purified by silica gel column with MeOH/CH2Cl2 (2:98) to afford compound 8 (92 mg, 52%) as yellowish oil. 1H NMR (d): 2.2 (s, 6H), 3.05 (td,

986

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994

2H, J = 6.0, 25.7 Hz), 3.55 (s, 2H), 4.63 (td, 2H, J =6.0, 46.9 Hz), 6.63 (d, 1H, J =8.4 Hz), 7.19 (dd, 1H, J =1.8, 8.4 Hz), 7.33 (t, 1H, J = 7.6 Hz), 7.48 (t, 1H, J = 7.5 Hz), 7.54 (d, 1H, J =7.6 Hz), 7.65 (d, 1H, J =7.5 Hz), 8.12 (d, 1H, J =1.7 Hz). Anal. calcd. for C17H19FN2O2S: C, 61.06; H, 5.73; N, 8.38. Found: C, 60.76; H, 5.74; N, 8.10. 2.1.8. 2-[4-(2-Fluoroethyl)-2-[(nitrophenyl)thio]]-Nmethylbenzylamine (9) In an analogous manner, compound 9 was obtained from compound 7 in 64.7% yield as yellowish oil. 1H NMR (d): 2.44 (s, 3H), 3.02 (td, 2H, J =6.5, 25.8 Hz), 3.95 (s, 2H), 4.63 (td, 2H, J = 6.5, 46.9 Hz), 4.64 (br s, 1H), 6.62 (d, 1H, J = 8.4 Hz), 7.23 (dd, 1H, J = 1.8, 8.4), 7.40 (td, 1H, J = 1.3, 7.6 Hz), 7.53 (td, 1H, J = 1.3, 7.6 Hz), 7.59 (dd, 1H, J = 1.1, 7.6), 7.73 (d, 1H, J = 7.6 Hz), 8.15 (d, 1H, J = 1.8 Hz). Anal. calcd. for C16H17FN2O2S: C, 59.98; H, 5.35; N, 8.74. Found: C, 59.69; H, 5.37; N, 8.64. 2.1.9. 2-[2-[[(Dimethylamino)methyl]phenyl]thio]5-(2-fluoroethyl)phenylamine (10) To a solution of compound 8 (52 mg, 0.17 mmol) in MeOH (5 mL) was added one drop of concentrated HCl. The suspension was cooled to 08C, SnCl2 (77 mg, 0.34 mmol) was added and the reaction mixture stirred overnight at room temperature. The mixture was then diluted with H2O (5 mL) and extracted with EtOAc (23 mL). The organic layers were discarded. The aqueous layer was adjusted to pH 10 with 1 N NaOH and extracted with EtOAc (43 mL). The combined organic layers were washed with H2O, dried and concentrated to give product 10 (40 mg, 77.4%) as colorless oil. 1H NMR (d): 2.17 (s, 6H) 2.95 (td, 2H, J = 6.6, 23.4 Hz), 3.57 (s, 2H), 4.50 (br s, 2H), 4.64 (td, 2H, J =6.6, 47.1 Hz), 6.58–6.65 (m, 1H), 6.85–6.92 (m, 2H), 7.06–7.13 (m, 2H), 7.20–7.26 (m, 1H), 7.42 (d, 1H, J =7.6 Hz). An analytical sample was prepared as the HCl salt. Anal. calcd. for C17H21FN2Sd 2HCld H2O: C, 51.64; H, 6.37; N, 7.09. Found: C, 52.09; H, 6.18; N, 6.80. 2.1.10. 5-(2-Fluoroethyl)-2-[2-[[(methylamino)methyl]phenyl]thio] phenylamine (11) In an analogous manner, compound 11 was obtained from compound 9 in 55% yield as colorless oil. 1H NMR (d): 2.52 (s, 3H), 2.99 (td, 2H, J = 6.3, 23.6 Hz), 3.74 (s, 1H), 3.94 (s, 2H), 4.40 (br s, 2H), 4.60 (td, 1H, J =6.3, 47.1 Hz), 6.59–6.72 (m, 2H), 6.80–6.88 (m, 1H), 7.05–7.18 (m, 2H), 7.26–7.38 (m, 2H). HRMS: calculated for C16H20FN2S (MH+): m/z 291.1331. Found: 291.1326. 2.2. In vitro binding assays Compound 10 was assayed for its affinities to the monoamine transporters (SERT, NET and DAT) in displacement experiments in vitro using cloned human receptors (hSERT, hNET and hDAT) transfected on HEK293 cells and the radioligands [3H]paroxetine (SERT),

[3H]nisoxetine (NET) and [3H]GBR 12935 (DAT), in accordance to the published procedures [38]. Affinities of compound 10 for other receptors are determined as previously described [45]. 2.3. Radiochemistry Instruments used for radiochemistry are as follows: a semipreparative HPLC system including a Waters 515 HPLC pump (Waters, Milford, MA), a Rheodyne 7010 injector with a 2-mL loop, a Prodigy C18 ODS Prep column (10 lm, 10250 mm, Phenomenex, Torrance, CA), an Alltech Model 450 UV detector, a custom-made gamma detector and a PC running LookOut HPLC data acquisition software; an analytical HPLC system consisting of a Waters 515 HPLC pump, a Rheodyne 7125 injector, a Phenomenex Prodigy C18 ODS-3 column (5 lm, 4.6250 mm), a Waters PDA 996 detector, a Flow Cell gamma detector (Bioscan, Washington, DC) and a PC with the Empower software used for system control. [11C]CO2 was produced in a CTI RDS112 Cyclotron with the bombardment of an N2 target with a proton source. The radiolabeling precursor 11 was prepared as described in Scheme 1. Radiosynthesis was carried out as depicted in Scheme 2. For radiolabeling, the precursor 11 (0.3–0.5 mg) was dissolved in acetone (0.4 mL) in a 1.0-mL reaction vial. [11C]MeOTf, produced according to the literature procedure [46], was bubbled through the precursor solution at room temperature. When maximum radioactivity was reached in the vial, the bubbling and vent needles were removed and the reaction mixture was heated in a water bath (708C) for 5 min. The crude product was then purified using the semipreparative HPLC system (mobile phase: 40% MeCN/60% 0.1 M ammonium acetate, pH 6.8; flow rate: 10 mL/min). The product fraction, eluted at ~12 min, was collected, diluted with H2O (100 mL) and passed through a Waters classic C18 Sep-Pak. After rinsing with H2O (10 mL), the Sep-Pak was eluted with EtOH (1 mL) to recover the product. The EtOH solution was then mixed with 0.9% sterile saline (9 mL), filtered through a membrane filter (0.22 Am) and the filtered solution collected in a sterile vial. The final product [11C]AFE (12) was obtained in N 95% radiochemical purity, as indicated by HPLC analysis of the ethanol solution (mobile phase: 35% MeCN/65% 0.1 M ammonium formate, pH 6.4; flow rate: 2 mL/min; retention time for the product: ~9 min). Radiochemical yield was 32F17% (decay-corrected, based on [11C]MeOTf, n = 6). Specific activity was 1670F864 Ci/mmol at EOS (n = 6). The identity of the labeled compound was confirmed by coinjection of the labeled compound 12 with the cold compound 10 and detection of a single UV peak by the analytical HPLC system. Alternatively, [11C]AFE (12) can be prepared by reaction of the precursor 11 with [11C]MeI, with trapping of [11C]MeI at 108C and reaction at 808C for 5 min. Using this method, [11C]AFE was synthesized in a radiochemical yield of 26F14% (decay-corrected, based

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994

987

Scheme 1. Synthesis of AFE (10) and the radiolabeling precursor 11.

on [11C]MeI, n = 4). Specific activity was 867F446 Ci mmol at EOS (n =4). 2.4. Measurement of partition coefficient Measurement of partition coefficient followed the published procedure, with some modifications [47]. Briefly, about 0.4 mCi of the radioligand [11C]AFE (12) was added to a separatory funnel containing n-octanol (20 mL) and 0.02 M phosphate buffer (pH 7.4, 20 mL). The mixture was shaken mechanically for 3 min and the layers were separated. The aqueous layer was discarded. The n-octanol layer (15 mL) was transferred to a second separatory funnel containing 15 mL of the phosphate buffer. The mixture was shaken mechanically for 3 min and the layers were separated. The aqueous layer was discarded and the noctanol layer was then partitioned into four test tubes (2 mL of the octanol solution each) containing 2 mL of the phosphate buffer each. The test tubes were vortexed for 10 min and then centrifuged for 10 min at 1000g to separate the layers. The n-octanol and aqueous phases (1.0 mL each) were transferred into counting tubes and counted with a gamma-counter. The radioactivity counts were decay-corrected and the partition coefficient calculated as follows: P =(counts in n-octanol)/(counts in buffer). Reported logP value represents the mean of eight separate calculations. 2.5. Biodistribution studies in rats Biodistribution experiments in rats were performed according to protocols approved by the New York State Psychiatric Institute–Columbia University Institutional Animal Care and Use Committee (IACUC). The labeled

compound [11C]AFE (12) in saline (~100 lCi for each rat) was injected into groups of male Sprague–Dawley rats (three rats for each group) via the tail vein, and the rats were sacrificed by decapitation, following euthanasia with CO2, at 2, 10, 30 and 60 min after radioactivity injection. The brain regions (cerebellum, hippocampus, striatum, prefrontal cortex, thalamus and hypothalamus), along with samples of blood and part of the tail, were removed, weighed and counted in a Packard Cobra II gammacounter (Packard, Meriden, CT). The percent injected dose (% ID) of the decay- and tail-corrected activity in the brain regions and blood were calculated based upon 11C standards prepared from the same radioligand solution and the % ID/g calculated using the tissue weights. In another set of experiment, four groups of rats (three rats in each group) were treated with the cold compound 10, citalopram (an SSRI), nisoxetine (a selective norepinephrine reuptake inhibitor) or GBR 12935 (a selective dopamine reuptake inhibitor) (2 mg/kg each, iv) 10–15 min before the radiotracer injection. In a control group, rats (n = 3) were injected with saline. Each group of rats was sacrificed by decapitation at 45 min after the injection of radioactivity. Blood sample and brain tissues were taken, counted, weighed and % ID/g calculated based on decay-corrected counts. The average % ID/g in the control group was then compared with that from each of the pretreatment groups. 2.6. PET imaging experiments in baboons Results of six PET experiments are presented. Three 12 to 32 kg male baboons (A, B and C) were used. Baboons A

Scheme 2. Radiosynthesis of [11C]AFE (12).

988

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994

and B were studied twice and Baboon C once under control conditions (n =5 control experiments). In addition, Baboon C was studied once following pretreatment with citalopram at a dose of 4 mg/kg, given iv 20 min before radiotracer administration. 2.6.1. Imaging protocol Baboon PET studies were performed according to protocol approved by the Columbia-Presbyterian Medical Center Institutional Animal Care and Use Committee. Fasted animals were immobilized with ketamine (10 mg/kg im) and anesthetized with 1.8% isoflurane via an endotracheal tube. Vital signs were monitored every 10 min, and the temperature was kept constant at 378C with heated water blankets. An intravenous perfusion line was used for hydration and injection of radioactivity and nonradioactive drug. A catheter was inserted in a femoral artery for arterial blood sampling. The head of the baboon was positioned at the center of the field of view as defined by imbedded laser lines. PET scans were performed with the ECAT EXACT HR+ PET camera (Siemens/CTI, Knoxville, TN). A 15-min transmission scan was obtained before radioactivity injection. Radioactivity was injected intravenously in 30 s. For the blocking experiment, citalopram hydrobromide (4 mg/ kg) in saline was injected intravenously 20 min before radiotracer administration. Emission data were collected in 3D mode for 90 or 120 min as 21 (23) successive frames of increasing duration [6  10 s, 2  1 min, 4  2 min, 2  5 min, 7 (10)  10 min]. 2.6.2. Input function measurement Arterial blood samples were collected every 10 s with an automated system for the first 2 min, every 20 s between 2 and 4 min and manually thereafter at various intervals. A total of 28 samples were collected. Following centrifugation (10 min at 1800g), the plasma was collected and activity was measured in a 200 lL aliquot on a gamma-counter (Wallac 1480 Wizard 3M Automatic Gamma-Counter, Perkin-Elmer, Boston, MA). 2.6.3. Metabolite analysis Additional blood samples were taken at 2, 4, 12, 30, 60 and 90 min after radioactivity injection for analysis of radioactive metabolites and the parent compound. After centrifugation, the plasma was separated and transferred to a tube containing 1 mL of MeOH. The mixture was vortexed and centrifuged at 15,000g for 5 min to separate the pellet from the aqueous phase. The supernatant was then taken up in a syringe and injected onto an HPLC column (Phenomenex Prodigy C18 ODS-3 (10 Am, 4.6250 mm) and analyzed (mobile phase: 40% MeCN/60% 0.1 M ammonium formate; flow rate: 2 mL/min; retention time for the parent compound: ~8 min). Fractions were collected, counted and decaycorrected to calculate the percentage of the parent compound in the blood at different time points. Before plasma sample analysis, the retention time of the parent compound was

established by injection of a small amount of the radioligand solution and detection of the radioactivity peak using a Bioscan gamma-detector. The arterial input function was corrected for the presence of radioactive metabolites as previously described [48] and used for kinetic analysis of brain uptake. The clearance of the parent compound (C L, L/h) was calculated as the ratio of the injected dose to the area under the curve of the input function [49,50]. 2.6.4. Free fraction measurement For the determination of the plasma free fraction ( f 1), triplicate 200 AL aliquots of plasma (separated from blood collected before radioactivity injection) were mixed with the radioligand and pipetted into ultrafiltration units (Centrifree, Amicon, Danvers, MA) and centrifuged at room temperature (20 min at 4000 rpm) [51]. Plasma and ultrafiltrate activities were counted, and f 1 was calculated as the ratio of the ultrafiltrate activity to the total plasma activity. Triplicate aliquots of the radiotracer in Tris buffer ( pH 7.4) were also processed to determine the filter retention of free tracer. 2.6.5. Image analysis Image analysis was performed as previously described [48]. Briefly, a magnetic resonance image (MRI) of each baboon’s brain was obtained for the purpose of identifying the regions of interest (ROI; T1-weighted axial MRI sequence, acquired parallel to the anterior–posterior commissure, TR = 34 ms, TE = 5 ms, flip angle = 458, slice thickness = 1.5 mm, zero gap, matrix =1.511 mm voxels). PET emission data were attenuation-corrected using the transmission scan, and frames were reconstructed using a Shepp filter (cutoff = 0.5 cycles/projection ray). Reconstructed image files were then processed by the image analysis software MEDx (Sensor Systems, Sterling, VA). Frames were summed. The summed image was used to define the registration parameters with the MRI, using between-modality automated image registration algorithm [52], and individual frames were registered to the MR data set. The following ROIs were drawn on the MRIs: thalamus, midbrain, striatum, hippocampus, temporal cortex, cingulate cortex, occipital cortex, parietal cortex and cerebellum. Regions were transferred to the registered PET frames, and time–activity curves were measured and decay-corrected. Right and left regions were averaged. 2.6.6. Kinetic analysis Regional distribution volumes were derived with kinetic analysis of the regional time–activity curves, using the arterial plasma concentrations as input function. Two models were evaluated, one- (1TC) and two-tissue (2TC) compartment models. The use of the 2TC model was associated with an unacceptable number of nonconvergence or convergence with negative rate constants and with large error in the estimates of the total regional distribution volume (V T, mL/g). Therefore, a 1TC model was selected as the more appropriate model. Similar observations were

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994 11

11

reported for [ C]McN5652 [20 – 22], [ C]DASB [30], [11C]AFM [34] and [11C]AFA [41]. In the 1TC model, V T is derived as the K 1/k 2 ratio, where K 1 (mL g 1 min 1) and k 2 (min 1) are the unidirectional fractional rate constant for the transfer of the tracer in and out of the brain, respectively. The cerebellum V T (V T CER) was used as an estimate of the nonspecific distribution volume in all brain regions because of the negligible levels of SERT specific binding in the cerebellum [3–5]. In the ROIs, the specific binding was estimated by the parameter V 3W, the specific-to-nonspecific partition coefficient at equilibrium [53]. V 3W was derived as (V T ROI V T CER)/V T CER. Under these conditions, V 3W is equal to f 2d B max/K D, where f 2 is the free fraction in the nonspecific distribution volume ( f 2 = 1/V T CER), B max is the regional concentration of SERT (nM) and K D is the affinity of the tracer for SERT (nM). Kinetics parameters were derived by nonlinear regression using a Levenberg– Marquart least-square minimization procedure implemented in MATLAB (Math Works, South Natick, MA), as previously described [54]. 3. Results and discussion 3.1. Chemistry Synthesis of compound 10 (AFE) and its radiolabeling precursor 11 is depicted in Scheme 1. Briefly, methyl 2-(4bromo-3-nitrophenyl)acetate (1) was reacted with thiosalicylic acid (2) in DMF to give the benzoic acid 3, which was taken to amide 4 or 5 by conversion of 3 to the acid chloride, then reaction of the acid chloride with either N,Ndimethylamine hydrochloride or N-methylamine hydrochloride. Concurrent reduction of the amide and ester functionalities in 4 or 5 with borane–THF complex led to compounds 6 and 7. Deoxofluorination of compounds 6 and 7 with [bis(methoxyethyl)amino]sulfur trifluoride afforded compounds 8 and 9. Finally, reduction of the nitro group in compounds 8 and 9 with tin (II) chloride under acidic conditions provided compounds 10 and 11. 3.2. In vitro binding affinities Compound 10 was assayed for its affinities for the monoamine transporters (SERT, NET and DAT) in displacement experiments in vitro using cloned human monoamine transporters and the radioligands [3H]paroxetine (SERT), [3H]nisoxetine (NET) and [3H]GBR 12935 (DAT). Results from binding assays demonstrate that compound 10 displays a high affinity for SERT (K i =1.80F0.07 nM), while its affinity for NET (K i = 946.2F222.5 nM) and DAT (K i N10,000 nM) are much lower, indicating an excellent selectivity for SERT over NET and DAT. In a broad screening, compound 10 was found to lack any appreciable affinity (K i N 10,000 nM) for serotonin and dopamine receptors, such as 5-HT1A, 5-HT2A, 5-HT2C, 5-HT3, 5HT6, 5-HT7 and D1–D5, as well as opioid (A, n and y), benzodiazepine and NMDA (PCP site) receptors. These

989

results further underline the highly selective nature of AFE (10) binding to the SERT. 3.3. Radiochemistry The radiolabeled compound [11C]AFE (12) was synthesized by N-methylation of the precursor 11 with [11C]methyl triflate in acetone or [11C]methyl iodide in DMF (Scheme 2). Using [11C]methyl triflate, [11C]AFE was prepared in 32F17% radiochemical yield (decay-corrected, based upon [11C]methyl triflate, n =6) and with specific activity of 1670F864 Ci/mmol at EOS (n = 6). Radiochemical purity of the final formulated product was N 95%. Total synthesis time was 35F6 min. The variability in radiochemical yield is due to a combination of factors such as precursor amount and trapping efficiency of [11C]methyl triflate as well as differences in the fractional collection of the radioactive product from the semipreparative HPLC system. When [11C]methyl iodide was used as the methylating agent, [11C]AFE was prepared in 26F14% radiochemical yield (decay-corrected, based upon [11C]methyl iodide, n = 4) and with specific activity of 867F446 Ci/mmol at EOS (n =4). Total synthesis time was about 35 min. 3.4. Lipophilicity Lipophilicity of the radioligand was measured using a modified procedure of Wilson et al. [47]. The logP of the labeled compound [11C]AFE (12) was determined to be 2.38F0.10 (n = 8 measurements), comparable to that of [ 11 C]AFA (2.53F0.21), [ 11 C]AFM (2.44F0.03) or [11C]DASB (2.38F0.03) [34,41]. 3.5. Biodistribution in rats Results from ex vivo studies in rats are presented in Table 1. In biodistribution studies, [11C]AFE (12) displayed an overall brain uptake of 0.8% ID/g at 10 min after the tail vein injection of the radiotracer, indicating a good penetration of the radioligand into the brain. At later time points (30 and 60 min), when nonspecific binding was cleared, the radiotracer displayed an uptake pattern consistent with the distribution of SERT in the rat brain, with high levels of radioactivity in the hypothalamus, thalamus, frontal cortex and striatum, moderate levels in the hippocampus and low levels in the cerebellum. There also appeared to be an excellent localization of specific binding in SERT-rich areas, with the ratio of specific-to-nonspecific binding (expressed as the ratio of % ID/g in the ROI to that in the cerebellum) reaching 5.5 in the hypothalamus and 4.1 in the thalamus at 60 min after radiotracer injection (Table 1). Moreover, such binding in SERT-rich regions appears to be saturable and specific. When rats were treated with either citalopram, an SSRI, or the cold compound 10 (2 mg/kg for each compound, given iv 10–15 min before radiotracer injection), 44–62% of the specific binding in the hypothalamus, thalamus, striatum, hippocampus and frontal cortex was displaced at 45 min after injection of the radiotracer (Fig. 2). On the other hand, when rats were treated with the selective

990

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994

Table 1 Regional brain uptake and specific binding of the radiotracer [11C]AFE (12) in male Sprague–Dawley rats Time (min)

Blood

Cerebellum

Frontal cortex

Striatum

Hippocampus

Hypothalamus

Thalamus

2

0.29F0.13 – 0.15F0.03 – 0.07F0.01 – 0.05F0.01 –

0.27F0.20 – 0.31F0.03 – 0.12F0.00 – 0.05F0.00 –

0.51F0.39 (1.82F0.11) 0.61F0.07 (2.01F0.33) 0.39F0.03 (3.28F0.18) 0.19F0.02 (3.58F0.33)

0.41F0.28 (1.57F0.12) 0.58F0.08 (1.87F0.26) 0.35F0.05 (2.97F0.36) 0.19F0.02 (3.63F0.24)

0.30F0.20 (1.15F0.08) 0.44F0.04 (1.43F0.14) 0.31F0.03 (2.61F0.21) 0.18F0.02 (3.40F0.26)

0.46F0.27 (1.98F0.71) 0.61F0.11 (2.01F0.51) 0.44F0.04 (3.68F0.26) 0.29F0.03 (5.51F0.46)

0.41F0.26 (1.59F0.17) 0.63F0.08 (2.05F0.33) 0.44F0.05 (3.71F0.33) 0.22F0.01 (4.08F0.33)

10 30 60

Regional uptake is represented as % ID/g (meanFS.D., three animals per group); specific binding in the brain regions (in bracket) is expressed as the ratio of activity in the region of interest to that in the cerebellum (meanFS.D.).

norepinephrine uptake inhibitor nisoxetine or the selective dopamine reuptake inhibitor GBR 12935 (2 mg/kg each, given iv 10–15 min before radiotracer injection), there were no significant changes in the specific binding of [11C]AFE in any of the SERT-rich brain regions (Fig. 2). Taken together, biological studies in rats indicated that the radioligand [11C]AFE (12) displayed high binding selectivity and specificity for the SERT in the brain. 3.6. PET imaging experiments in baboons Injected doses for baboons were 5.0F0.1 mCi (n = 6). Specific activity at the time of injection was 990F338 Ci/ mmol. Injected mass was 1.66F0.47 lg (n =6). Analysis of blood samples indicated that the radioligand metabolized rapidly into more polar compounds. No radioactive lipophilic metabolites were detected by HPLC analysis of the blood samples. Percent of activity corresponding to the parent compound was 84F11%, 80F11%, 47F20%, 23F13%,

15F12% and 10F7%, respectively, at 2, 4, 12, 30, 60 and 90 min postinjection (n =5 per time point). Plasma clearance rate (C L) of the parent compound was 63F9 L/h under control conditions (n = 5). Citalopram pretreatment accelerated the plasma clearance of [11C]AFE (C L =95 L/h, n =1). The free fraction of [11C]AFE in the baboon blood was 7.9F2.6% (n =5). Pretreatment with citalopram had no effect on the free fraction of [11C]AFE in the plasma. Over time, activity became highly concentrated in brain regions with high SERT densities (i.e., in the thalamus, midbrain and striatum). Moderate levels were found in the hippocampus, temporal and cingulate cortices, and lower levels in other cortical regions. The lowest radioactivity concentrations were observed in the cerebellum (Fig. 3, middle row). When the baboon was treated with citalopram 20 min before the injection of radioactivity, uptakes of [11C]AFE in the midbrain, thalamus, striatum, hippocampus and cortex were all reduced to the level in the cerebellum,

Fig. 2. Specific binding in the rat brain (expressed as the activity ratio of ROI to cerebellum) at 45 min following injection of [11C]AFE (12) under control conditions and after pretreatment with the cold compound AFE (10), citalopram, nisoxetine, or GBR 12935 (2 mg/kg each, given iv 10 –15 min before [11C]AFE administration). Data represent the meanFS.D. of each group (three rats per group). Regions are the prefrontal cortex (PFC), striatum (STR), hippocampus (HIP), hypothalamus (HYP) and thalamus (THA).

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994

991

Fig. 3. Coronal (left column), sagittal (middle column) and transaxial (right column) views of the brain MRI images of Baboon C (top row) and [11C]AFE (12) distribution in the brain under control conditions (middle row) and following pretreatment with citalopram (4 mg/kg, given iv 20 min before [11C]AFE injection) (bottom row). Under control conditions, high accumulation of activity is detected in the midbrain, thalamus, striatum and hippocampus. The activity distribution is homogenous across the brain following citalopram pretreatment.

indicating blocking of the [11C]AFE binding sites by citalopram (Fig. 3, bottom row). Typical time–activity curves under control and blocking conditions are shown in Fig. 4. Under control conditions,

activities in regions rich in SERT increased for the first 15 min, peaked between 15 and 40 min and decreased rapidly thereafter. Under blocking conditions, activities peaked early in all regions, and these peaks were followed by a

Fig. 4. Time–activity curves of the radiotracer [11C]AFE (12) in the brain of Baboon C. The left panel represents time–activity curves of [11C]AFE under control conditions. The right panel represents time–activity curves of [11C]AFE after pretreatment of the baboon with citalopram (4 mg/kg, given iv 20 min before radioactivity administration). Points are activities measured in the cerebellum (open circles), temporal cortex (closed triangles), striatum (closed circles) and thalamus (closed diamonds). Lines are values fitted by kinetic analysis.

992

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994

Table 2 Regional distribution volume (V T) and equilibrium specific-to-nonspecific partition coefficient (V 3W) of [11C]AFE, [11C]AFA, [11C]DASB, [11C]McN5652 and [11C]AFM in baboon brain regions Region

Cerebellum Midbrain Thalamus Striatum Hippocampus Temporal cortex Cingulate cortex Occipital cortex a b c

[11C]AFEa

[11C]AFAb

[11C]DASBc

[11C]McN5652c

[11C]AFMc

V T (mL/g)

V 3W

V T (mL/g)

V 3W

V T (mL/g)

V 3W

V T (mL/g)

V 3W

V T (mL/g)

V 3W

16.2F2.6 29.8F6.9 33.8F8.1 25.9F5.8 22.8F4.3 19.4F3.9 18.8F3.2 17.3F3.1

– 0.83F0.17 1.07F0.17 0.59F0.12 0.40F0.05 0.19F0.06 0.16F0.03 0.07F0.03

18.8F5.0 36.8F10.4 39.6F11.3 32.7F10.6 27.7F8.1 23.5F6.8 22.8F6.2 20.5F5.9

– 0.95F0.07 1.10F0.08 0.71F0.13 0.46F0.05 0.24F0.07 0.21F0.06 0.08F0.04

17.3F0.5 46.4F4.3 47.5F5.0 31.7F1.7 27.3F1.4 22.5F2.1 21.1F0.9 19.3F1.5

– 1.68F0.21 1.74F0.27 0.83F0.08 0.58F0.08 0.29F0.10 0.22F0.04 0.11F0.06

27.7F4.0 53.6F9.3 56.5F10.3 45.3F7.0 37.2F5.9 33.9F4.2 33.1F4.7 30.6F3.9

– 0.94F0.18 1.04F0.16 0.64F0.13 0.34F0.06 0.23F0.08 0.19F0.02 0.11F0.06

30.6F3.9 96.3F15.3 101.9F20.4 61.9F10.4 51.5F6.1 42.4F4.5 39.0F4.7 37.9F3.6

– 2.14F0.16 2.31F0.25 1.01F0.08 0.69F0.06 0.39F0.05 0.28F0.02 0.24F0.06

Values are meanFS.D. of five experiments. Values are taken from [41]. Values are taken from [34].

rapid clearance. The fast kinetics of [11C]AFE in the baboon brain indicated a desirable property for a 11C-labeled PET radioligand. Listed in Table 2 are values of regional total distribution volume (V T) and equilibrium specific-to-nonspecific partition coefficients (V 3W) derived from kinetic analysis of the time–activity curves under control conditions. The V T and V 3W values of [11C]AFA, [11C]DASB, [11C]McN5652 and [11C]AFM are also listed in Table 2 for comparison. In terms of brain uptake kinetics, [11C]AFE is similar to [11C]AFA, with peak uptake occurring at 15–40 min after tracer injection. [11C]DASB displays a faster uptake kinetics, with peak uptake at 10–25 min after injection, while the uptake of [11C]McN5652 and [11C]AFM is slower, with peak activity at 25–60 min after tracer administration. Nonspecific binding, as represented by the distribution volume in the cerebellum, is low for [11C]AFE and similar to those of

[11C]AFA and [11C]DASB (cerebellum V T of 16.2 mL/g for [11C]AFE vs. 18.8 mL/g for [11C]AFA and 17.3 mL/g for [11C]DASB). Nonspecific binding of [11C]McN5652 and [11C]AFM is higher (cerebellum V T of 27.7 and 30.6 mL/g, respectively). The regional V 3W values of [11C]AFE are comparable to those observed in baboons under similar conditions with [11C]McN5652 or [11C]AFA, but lower than those of [11C]DASB or [11C]AFM (Table 2) [34,40,41]. Under blocking conditions, V T values for [11C]AFE were 11.82, 13.75, 12.79, 13.49, 11.53 and 12.30 mL/g in the midbrain, thalamus, striatum, hippocampus, temporal cortex and cingulate cortex, respectively. Following citalopram pretreatment, [11C]AFE V 3W were reduced by 91% in the midbrain, 73% in the thalamus and striatum, 38% in the hippocampus and 80% in the temporal cortex, with values of 0.06, 0.23, 0.15, 0.21 and 0.03 mL/g in the midbrain, thalamus, striatum, hippocampus and temporal cortex, respectively (Fig. 5). Data from the blocking experiment demonstrated the binding specificity of [11C]AFE to SERT in the baboon brain. 4. Conclusion

Fig. 5. Specific-to-nonspecific partition coefficient at equilibrium (V 3W) of [11C]AFE (12) in baboon brain regions under control conditions (n = 5) and after pretreatment with citalopram (n = 1, 4 mg/kg, given iv 20 min before [11C]AFE administration). Regions are the midbrain (MID), thalamus (THA), striatum (STR), hippocampus (HIP) and temporal cortex (TEM).

A new PET radioligand for the SERT, [11C]AFE, was synthesized in high radiochemical yield and high specific activity. In biodistribution studies in rats and imaging studies in baboons, [11C]AFE was demonstrated to enter the brain easily and accumulate in brain regions known to be rich in SERT. Blocking experiments in rats and baboon confirmed the binding specificity and selectivity of [11C]AFE in vivo. In the baboons, [11C]AFE displays a fast brain uptake kinetics, with peak uptake in the thalamus and midbrain at 15– 40 min after tracer injection. Taken together, results from animal studies indicate that [11C]AFE is an appropriate PET radioligand for the visualization and quantification of SERT in vivo. Comparison of [11C]AFE with other available PET radioligands for SERT offers some interesting similarities and differences. In the baboons, the brain uptake kinetics of [11C]AFE is fast and similar to that of [11C]AFA, but a

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994 11

little slower than that of [ C]DASB. The signal-to-noise ratios of [11C]AFE are similar to those of [11C]McN5652 or [11C]AFA, but lower than those of [11C]DASB or [11C]AFM. From these data, it appears that [11C]AFE does not present any advantages over [11C]DASB, either in terms of uptake kinetics or signal-to-noise ratio. Therefore, it is believed that [11C]DASB remains the 11C-labeled ligand of choice for imaging the SERT in vivo, and there is no clear incentive in the further development of [11C]AFE for human applications. However, in contrast to DASB, the SERT ligands AFA, AFM and AFE can all be labeled with the longer-lived isotope 18F, and thus, the 18 F-labeled versions of these ligands will be useful in imaging applications off-site. Data presented in this article indicate that [11C]AFA and [11C]AFE hold advantage over [11C]AFM in terms of brain uptake kinetics and nonspecific binding, while [11C]AFM is superior to [11C]AFA and [11C]AFE in that it displays a much higher specific binding signal. From these results, it can be reasonably predicted that [18F]AFM will be able to provide better imaging quality than either [18F]AFA or [18F]AFE, and as a result, could potentially become the 18F-labeled ligand of choice for SERT imaging. However, a potential drawback could come from the very structure of [18F]AFM, as the presence of 18F fluorine at the benzylic position might induce some defluorination. On the other hand, the fluorine atom is directly attached to the aromatic ring in [18F]AFA and at the h-position of the alkyl chain in [18F]AFE. Consequently [18F]AFA and [18F]AFE might be less susceptible to defluorination in vivo, and thus could become alternative 18F-labeled ligand for SERT imaging in case [18F]AFM is proved unstable in humans. From these analyses, it appears that [ 18F]AFA, [ 18F]AFM and [18F]AFE represent an interesting series of 18F-labeled PET radioligands for SERT. A comprehensive in vivo comparative study of these three ligands in baboons is ongoing in our laboratories to determine their respective brain uptake kinetics, signal-to-noise ratios and in vivo defluorination potentials.

Acknowledgments The authors thank the expert technical assistance of Van Phan, Luydmilla Savenkova, Nurat Quadri and Chaka Peters. They would also like to thank Dr. Bryan Roth and the NIMH Psychoactive Drug Screening Program for conducting the binding assays of new compounds. This work was supported by the Public Health Service (NIMH/ NIDA R21 MH66624-01) and the Lieber Center for Schizophrenia Research at Columbia University.

References [1] Lucki I. The spectrum of behaviors influenced by serotonin. Biol Psychiatry 1998;44:151 – 62.

993

[2] Blakely RD, Bauman AL. Biogenic amine transporters: regulation in flux. Curr Opin Neurobiol 2000;10:328 – 36. [3] Laruelle M, Vanisberg MA, Maloteaux JM. Regional and subcellular localization in human brain of [3H]paroxetine binding, a marker of serotonin uptake sites. Biol Psychiatry 1988;24:299 – 309. [4] Backstrom I, Bergstrom M, Marcusson J. High affinity [3H]paroxetine binding to serotonin uptake sites in human brain tissue. Brain Res 1989;486:261 – 8. [5] Cortes R, Soriano E, Pazos A, Probst A, Palacios JM. Autoradiography of antidepressant binding sites in the human brain: localization using [3H]imipramine and [3H]paroxetine. Neuroscience 1988;27:473 – 96. [6] Rosel P, Menchon JM, Oros M, Vallejo J, Cortadellas T, Arranz B, et al. Regional distribution of specific high affinity binding sites for 3Himipramine and 3H-paroxetine in human brain. J Neural Transm 1997; 104:89 – 96. [7] Plenge P, Mellerup ET, Laursen H. Regional distribution of the serotonin transport complex in human brain, identified with 3Hparoxetine, 3H-citalopram and 3H-imipramine. Prog Neuropsychopharmacol Biol Psychiatry 1990;14:61 – 72. [8] Chen CP, Alder JT, Bowen DM, Esiri MM, McDonald B, Hope T, et al. Presynaptic serotonergic markers in community-acquired cases of Alzheimer’s disease: correlations with depression and neuroleptic medication. J Neurochem 1996;66:1592 – 8. [9] Deakin JF. The role of serotonin in panic, anxiety and depression. Int Clin Psychopharmacol 1998;13(Suppl. 4):S1 – S5. [10] Mann JJ. Role of the serotonergic system in the pathogenesis of major depression and suicidal behavior. Neuropsychopharmacology 1999;21:99S – 105S. [11] Mann JJ, Huang YY, Underwood MD, Kassir SA, Oppenheim S, Kelly TM, et al. A serotonin transporter gene promoter polymorphism (5-HTTLPR) and prefrontal cortical binding in major depression and suicide. Arch Gen Psychiatry 2000;57:729 – 38. [12] Abi-Dargham A, Laruelle M, Aghajanian GK, Charney D, Krystal J. The role of serotonin in the pathophysiology and treatment of schizophrenia. J Neuropsychiatry Clin Neurosci 1997;9:1 – 17. [13] Hyttel J. Pharmacological characterization of selective serotonin reuptake inhibitors (SSRIs). Int Clin Psychopharmacol 1994;1:19 – 26. [14] Charney DS. Monoamine dysfunction and the pathophysiology and treatment of depression. J Clin Psychiatry 1998;59:11 – 4. [15] Schloss P, Williams DC. The serotonin transporter: a primary target for antidepressant drugs. J Psychopharmacol 1998;12:115 – 21. [16] Iversen L. Neurotransmitter transporters: fruitful targets for CNS drug discovery. Mol Psychiatry 2000;5:357 – 62. [17] Koob GF. Drug abuse and alcoholism. Overview. Adv Pharmacol 1998;42:969 – 77. [18] Suehiro M, Scheffel U, Ravert HT, Dannals RF, Wagner Jr HN. [11C](+)McN5652 as a radiotracer for imaging serotonin uptake sites with PET. Life Sci 1993;53:883 – 992. [19] Szabo Z, Scheffel U, Suehiro M, Dannals RF, Kim SE, Ravert HT, et al. Positron emission tomography of 5-HT transporter sites in the baboon brain with [11C]McN5652. J Cereb Blood Flow Metab 1995;15:798 – 805. [20] Szabo Z, Scheffel U, Mathews WB, Ravert HT, Szabo K, Kraut M, et al. Kinetic analysis of [11C]McN5652: a serotonin transporter radioligand. J Cereb Blood Flow Metab 1999;19:967 – 81. [21] Parsey RV, Kegeles LS, Hwang DR, Simpson N, Abi-Dargham A, Mawlawi O, et al. In vivo quantification of brain serotonin transporters in humans using [11C]McN 5652. J Nucl Med 2000; 41:1465 – 77. [22] Buck A, Gucker PM, Schonbachler RD, Arigoni M, Kneifel S, Vollenweider FX, et al. Evaluation of serotonergic transporters using PET and [11C](+)McN-5652: assessment of methods. J Cereb Blood Flow Metab 2000;20:253 – 62. [23] Oya S, Kung MP, Acton PD, Mu M, Hou C, Kung HF. A new singlephoton emission computed tomography imaging agent for serotonin transporters: [123I]IDAM, 5-iodo-2-((2-((dimethylamino)methyl)phenyl)thio)benzyl alcohol. J Med Chem 1999;42:333 – 5.

994

Z. Zhu et al. / Nuclear Medicine and Biology 31 (2004) 983–994

[24] Oya S, Choi SR, Hou C, Mu M, Kung MP, Acton PD, et al. 2-((2((Dimethylamino)methyl)phenyl)thio)-5-iodophenylamine (ADAM): an improved serotonin transporter ligand. Nucl Med Biol 2000; 27:249 – 54. [25] Kung MP, Hou C, Oya S, Mu M, Acton PD, Kung HF. Characterization of [123I]IDAM as a novel single-photon emission tomography tracer for serotonin transporters. Eur J Nucl Med 1999; 26:844 – 53. [26] Acton PD, Kung MP, Mu M, Plossl K, Hou C, Siciliano M, et al. Single-photon emission tomography imaging of serotonin transporters in the non-human primate brain with the selective radioligand [123I]IDAM. Eur J Nucl Med 1999;26:854 – 61. [27] Newberg AB, Plossl K, Mozley PD, Stubbs JB, Wintering N, Udeshi M, et al. Biodistribution and imaging with 123I-ADAM: a serotonin transporter imaging agent. J Nucl Med 2004;45:834 – 41. [28] Kauppinen TA, Bergstrom KA, Heikman P, Hiltunen J, Ahonen AK. Biodistribution and radiation dosimetry of [123I]ADAM in healthy human subjects: preliminary results. Eur J Nucl Med Mol Imaging 2003;30:132 – 6. [29] Acton PD, Choi SR, Hou C, Plossl K, Kung HF. Quantification of serotonin transporters in nonhuman primates using [123I]ADAM and SPECT. J Nucl Med 2001;42:1556 – 62. [30] Ginovart N, Wilson AA, Meyer JH, Hussey D, Houle S. Positron emission tomography quantification of [11C]-DASB binding to the human serotonin transporter: modeling strategies. J Cereb Blood Flow Metab 2001;21:1342 – 53. [31] Wilson AA, Ginovart N, Schmidt M, Meyer JH, Threlkeld PG, Houle S. Novel radiotracers for imaging the serotonin transporter by positron emission tomography: synthesis, radiosynthesis, and in vitro and ex vivo evaluation of 11C-labeled 2-(phenylthio)araalkylamines. J Med Chem 2000;43:3103 – 10. [32] Houle S, Ginovart N, Hussey D, Meyer JH, Wilson AA. Imaging the serotonin transporter with positron emission tomography: initial human studies with [11C]DAPP and [11C]DASB. Eur J Nucl Med 2000;27:1719 – 22. [33] Ichise M, Liow JS, Lu JQ, Takano A, Model K, Toyama H, et al. Linearized reference tissue parametric imaging methods: application to [11C]DASB positron emission tomography studies of the serotonin transporter in human brain. J Cereb Blood Flow Metab 2003;23:1096 – 112. [34] Huang Y, Hwang DR, Narendran R, Sudo Y, Chatterjee R, Bae SA, et al. Comparative evaluation in nonhuman primates of five PET radiotracers for imaging the serotonin transporters: [11C]McN 5652, [11C]ADAM, [11C]DASB, [11C]DAPA, and [11C]AFM. J Cereb Blood Flow Metab 2002;22:1377 – 98. [35] Szabo Z, McCann UD, Wilson AA, Scheffel U, Owonikoko T, Mathews WB, et al. Comparison of (+)-11C-McN5652 and 11C-DASB as serotonin transporter radioligands under various experimental conditions. J Nucl Med 2002;43:678 – 92. [36] Frankle WG, Huang Y, Hwang DR, Talbot PS, Slifstein M, Van Heertum R, et al. Comparative evaluation of serotonin transporter radioligands 11C-DASB and 11C-McN 5652 in healthy humans. J Nucl Med 2004;45:682 – 94. [37] Vercouillie J, Tarkiainen J, Halldin C, Emond P, Chalon S, Sandell J, et al. Precursor synthesis and radiolabeling of [11C]ADAM: a potent radioligand for the serotonin transporter exploration by PET. J Labelled Compd Radiopharm 2001;44:113 – 20. [38] Huang Y, Hwang DR, Zhu Z, Bae SA, Guo N, Sudo Y, et al. Synthesis and pharmacological characterization of a new PET ligand for the serotonin transporter: [ 11 C]5-bromo-2-[2-(dimethylaminomethylphenylsulfanyl)]phenylamine ([11C]DAPA). Nucl Med Biol 2002;29:741 – 51.

[39] Chalon S, Tarkiainen J, Garreau L, Hall H, Emond P, Vercouillie J, et al. Pharmacological characterization of N,N-dimethyl-2-(2-amino-4methylphenylthio)benzylamine as a ligand of the serotonin transporter with high affinity and selectivity. J Pharmacol Exp Ther 2003; 304:81 – 7. [40] Huang Y, Hwang DR, Bae SA, Sudo Y, Guo N, Zhu Z, et al. A new positron emission tomography imaging agent for the serotonin transporter: synthesis, pharmacological characterization, and kinetic analysis of [11C]2-[2-(dimethylaminomethyl)phenylthio]-5-fluoromethylphenylamine ([11C]AFM). Nucl Med Biol 2004;31:543 – 56. [41] Huang Y, Narendran R, Bae SA, Erritzoe D, Guo N, Zhu Z, et al. A PET imaging agent with fast kinetics: synthesis and in vivo evaluation of the serotonin transporter ligand [11C]2-[2- (dimethylaminomethyl)phenylthio]-5-fluorophenylamine ([ 11 C]AFA). Nucl Med Biol 2004;31:727 – 38. [42] Huang Y, Bae SA, Zhu Z, Guo N, Hwang DR, Laruelle M. Fluorinated analogues of ADAM as new PET radioligands for the serotonin transporter: synthesis and pharmacological evaluation. J Labelled Compd Radiopharm 2001;44:S18 – S20. [43] Shiue GG, Fang P, Shiue CY. Synthesis of N,N-dimethyl-2-(2-amino4-[18F]fluorophenylthio)benzylamine as a serotonin transporter imaging agent. Appl Radiat Isot 2003;58:183 – 91. [44] Shiue GG, Choi SR, Fang P, Hou C, Acton PD, Cardi C, et al. N,NDimethyl-2-(2-amino-4-18F-fluorophenylthio)-benzylamine (4-18FADAM): an improved PET radioligand for serotonin transporters. J Nucl Med 2003;44:1890 – 7. [45] Glennon RA, Lee M, Rangisetty JB, Dukat M, Roth BL, Savage JE, et al. 2-Substituted tryptamines: agents with selectivity for 5-HT6 serotonin receptors. J Med Chem 2000;43:1011 – 8. [46] Jewett DM. A simple synthesis of [11C]methyl triflate. Int J Rad Appl Instrum [A] 1992;43:1383 – 5. [47] Wilson AA, Jin L, Garcia A, DaSilva JN, Houle S. An admonition when measuring the lipophilicity of radiotracers using counting techniques. Appl Radiat Isot 2001;54:203 – 8. [48] Abi-Dargham A, Simpson N, Kegeles L, Parsey R, Hwang DR, Anjilvel S, et al. PET studies of binding competition between endogenous dopamine and the D1 radiotracer [11C]NNC 756. Synapse 1999;32:93 – 109. [49] Rowland M, Tozer TN. Clinical pharmacokinetics. 2nd ed. Philadelphia (PA)7 Lea & Febiger; 1989. [50] Abi-Dargham A, Laruelle M, Seibyl J, Rattner Z, Baldwin RM, Zoghbi SS, et al. SPECT measurement of benzodiazepine receptors in human brain with [123I]iomazenil: kinetic and equilibrium paradigms. J Nucl Med 1994;35:228 – 38. [51] Gandelman MS, Baldwin RM, Zoghbi SS, Zea-Ponce Y, Innis RB. Evaluation of ultrafiltration for the free fraction determination of single photon emission computerized tomography (SPECT) radiotracers: h-CIT, IBF and iomazenil. J Pharm Sci 1994; 83:1014 – 9. [52] Woods RP, Mazziotta JC, Cherry SR. MRI-PET registration with automated algorithm. J Comput Assist Tomogr 1993;17:536 – 46. [53] Laruelle M, Al-Tikriti MS, Zea-Ponce Y, Zoghbi SS, Baldwin RM, Charney DS, et al. In vivo quantification of dopamine D2 receptors parameters in nonhuman primates with [123I]iodobenzofuran and single photon emission computerized tomography. Eur J Pharmacol 1994;263:39 – 51. [54] Laruelle M, Baldwin RM, Rattner Z, Al-Tikriti MS, Zea-Ponce Y, Zoghbi SS, et al. SPECT quantification of [123I]iomazenil binding to benzodiazepine receptors in nonhuman primates. I. Kinetic modeling of single bolus experiments. J Cereb Blood Flow Metab 1994; 14:439 – 52.