Synthesis and characterization of 11 C-labeled benzyl amidine derivatives as PET radioligands for GluN2B subunit of the NMDA receptors.

GluN2B-containing NMDA receptors (NMDARs) play fundamental roles in learning and memory, although they are also associated with various brain disorders. In this study, we synthesized and evaluated three 11 C-labeled N-benzyl amidine derivatives 2-[11 C]methoxybenzyl) cinnamamidine ([11 C]CBA), N-(2-[11 C]methoxybenzyl)-2-naphthamidine ([11 C]NBA), and N-(2-[11 C]methoxybenzyl)quinoline-3-carboxamidine ([11 C]QBA) as PET radioligands for these receptors. The 11 C-benzyl amidines were synthesized via conventional methylation of corresponding des-methyl precursors with [11 C]CH3 I. In vitro binding characteristics were examined in brain sagittal sections using various GluN2B modulators and off-target ligands. Further, in vivo brain distribution studies were performed in normal mice. The 11 C-labeled benzyl amidines showed high-specific binding to the GluN2B subunit at in vitro. In particular, the quinoline derivative [11 C]QBA had the best binding properties in terms of high-brain localization to GluN2B-rich regions and specificity to the GluN2B subunit. Conversely, these 11 C-radioligands showed the brain distributions were inconsistent with GluN2B expression in biodistribution experiments. The majority of the radiolabeled compounds were identified as metabolized forms of which amido derivatives seemed to be the major species. Although these 11 C-ligands had high-specific binding to the GluN2B subunit, significant improvement in metabolic stability is necessary for successful positron emission tomography (PET) imaging of the GluN2B subunit of NMDARs.

Synthesis and characterization of [¹²5I]2-iodo N-[(S)-{(S)-1-methylpiperidin-2-yl}(phenyl)methyl]3-trifluoromethyl-benzamide as novel imaging probe for glycine transporter 1.

In this study, 2-iodo substituted 1-methylpiperidin-2-yl benzamide derivatives were synthesized and evaluated as candidate SPECT imaging agents for glycine transporter 1 (GlyT1). In JAR cells, which predominantly express GlyT1, 2-iodo N-[(S)-{(S)-1-methylpiperidin-2-yl}(phenyl)methyl]3-trifluoromethyl-benzamide (5) showed excellent inhibitory activity of [(3)H]glycine uptake (IC(50)=2.4 nM). Saturation assay in rat cortical membranes revealed that [(125)I]5 had a single high affinity binding site with a K(d) of 1.54 nM and a B(max) of 3.40 pmol/mg protein. In vitro autoradiography demonstrated that [(125)I]5 showed consistent accumulation with GlyT1 expression. The in vitro binding was greatly inhibited by GlyT1 inhibitors but not by other site ligands, which suggested the high specific binding of [(125)I]5 with GlyT1. In the biodistribution and ex vivo autoradiography studies using mice, [(125)I]5 showed high blood-brain barrier permeability (1.68-2.17% dose/g at 15-60 min) and similar regional brain distribution pattern with in vitro results. In addition, pre-treatment of GlyT1 ligands resulted in significant decrease of [(125)I]5 binding in the GlyT1-rich regions. This preliminary study demonstrated that radio-iodinated 5 is a promising SPECT imaging probe for GlyT1.

Development of N-[11C]methylamino 4-hydroxy-2(1H)-quinolone derivatives as PET radioligands for the glycine-binding site of NMDA receptors.

In this study, we synthesized and evaluated several amino 4-hydroxy-2(1H)-quinolone (4HQ) derivatives as new PET radioligand candidates for the glycine site of the NMDA receptors. Among these ligands, we discovered that 7-chloro-4-hydroxy-3-{3-(4-methylaminobenzyl) phenyl}-2-(1H)-quinolone (12) and 5-ethyl-7-chloro-4-hydroxy-3-(3-methylaminophenyl)-2(1H)-quinolone (32) have high affinity for the glycine site (K(i) values; 11.7 nM for 12 and 11.8 nM for 32). In vitro autoradiography experiments indicated that [(11)C]12 and [(11)C]32 showed high specific binding in the brain slices, which were strongly inhibited by both glycine agonists and antagonists. In vivo brain uptake of these (11)C-labeled 4HQs were examined in normal mice. Cerebellum to blood ratio of accumulation, of both [(11)C]12 and [(11)C]32 at 30 min were 0.058, which were slightly higher than those of cerebrum to blood ratio (0.043 and 0.042, respectively). These results indicated that [(11)C]12 and [(11)C]32 have poor blood brain barrier permeability. Although the plasma protein-binding ratio of [(11)C]32 was much lower than methoxy analogs (71% vs 94-98%, respectively), [(11)C]32 still binds with plasma protein strongly. It is conjectured that still acidic moiety and high affinity with plasma protein of [(11)C]32 may prevent in vivo brain uptake. In conclusion, [(11)C]12 and [(11)C]32 are unsuitable for imaging cerebral NMDA receptors.

Difference in brain distributions of carbon 11-labeled 4-hydroxy-2(1H)-quinolones as PET radioligands for the glycine-binding site of the NMDA ion channel.

High-affinity iodine- and ethyl-C-5 substituted analogs of 4-hydroxy-3-(3-[11C]methoxyphenyl)-2(1H)-quinolone ([11C]4HQ) were synthesized as new positron emission tomography radioligands for the glycine-binding sites of the N-methyl-d-aspartate (NMDA) ion channel. Although both radioligands showed high in vitro specific binding to rat brain slices, their binding characteristics were quite different from each other. 5-Ethyl-[11C]4HQ (5Et-[11C]4HQ) showed higher in vitro binding in the forebrain regions than in the cerebellum, bindings that were strongly inhibited by both glycine-site agonists and antagonists. In contrast, 5-iodo-[11C]4HQ (5I-[11C]4HQ) showed a homogeneous in vitro binding throughout the brain, which was inhibited by antagonists but not by agonists. This difference in in vitro binding between 5Et-[11C]4HQ and 5I-[11C]4HQ was quite similar to that previously observed between [11C]L-703,717 and [11C]4HQ, both glycine-site antagonists. In vivo brain uptakes of these 11C-labeled 4-hydroxyquinolones were examined in mice. Initial brain uptakes of 5Et- and 5I-[11C]4HQ at 1 min after intravenous injections were comparable to that of [11C]4HQ, but they were 1.3-2.1 times higher than that of [11C]L-703,717. The treatment with an anticoagulant, warfarin, only slightly increased the initial uptakes of [11C]4HQ and 5Et-[11C]4HQ in contrast to [11C]L-703,717. The in vivo regional brain distributions were slightly different between the two radioligands. Pretreatment with nonradioactive ligand (2 mg/kg) slightly inhibited the binding of 5Et-[11C]4HQ (16-36% inhibition) but not that of 5I-[11C]4HQ. In this study, it was found that a small structural change in [11C]4HQ resulted in a major change in binding characteristics and distributions, suggesting the existence of two binding sites for [11C]4-hydroxyquinolones on the NMDA ion channel - agonist-sensitive and agonist-insensitive (or antagonist-preferring) sites.

N-(5-Fluoro-2-phenoxyphenyl)-N-(2-[(131)I]iodo-5-me thoxybenzyl)acetamide: a potent iodinated radioligand for the peripheral-type benzodiazepine receptor in brain.

To image the peripheral-type benzodiazepine receptor (PBR) in vivo, we previously developed two positron emission tomography (PET) ligands, N-(2-[11C],5-dimethoxybenzyl)-N-(5-fluoro-2-phenoxyphenyl)acetamide ([11C]1a) and its [18F]fluoroethyl analogue ([18F]1b), for the investigation of PBR in the living human brain. This time, using 1a as a leading compound, we designed two novel iodinated analogues, N-(5-fluoro-2-phenoxyphenyl)-N-(2-iodo-5-methoxybenzyl)acetamide (3a) and N-(2,5-dimethoxybenzyl)-N-(5-iodo-2-phenoxyphenyl)acetamide (3b) for the PBR imaging. Ligands 3 were synthesized by the iodination of tributystannyl precursors 10. Radiolabeling for 3 with 131I was carried out by the reaction of 10 with [131I]NaI using H2O2 as an oxidizing agent. In vitro competition experiments determined that 3a exhibited both high affinity and selectivity for PBR (IC50: 7.8 nM) vs CBR (>1 microM). Biodistribution study in mice determined that [131I]3a had a high radioactivity level (1.69% dose/g) in the brain, and its distribution pattern in the brain was consistent with the known distribution of PBR in rodents. Ex vivo autoradiography of the rat brain gave visual evidence that [131I]3a was a potent and specific radioligand for PBR.

Syntheses and pharmacological evaluation of two potent antagonists for dopamine D4 receptors: [11C]YM-50001 and N-[2-[4-(4-Chlorophenyl)-piperizin-1-yl]ethyl]-3-[11C]methoxybenzamide.

Two benzamide derivatives as dopamine D4 receptor antagonists, YM-50001(4) and N- [2-[4-(4-chlorophenyl]piperizin-1-yl]ethyl]-3-methoxybenzamide (9), were labeled by positron-emitter (11C), and their pharmacological specificities to dopamine D4 receptors were examined by quantitative autoradiography and positron emission tomography (PET). Radiosyntheses were accomplished by O-methylation of corresponding phenol precursors (5 and 10) with [11C]CH3I followed by HPLC purifications. In vitro binding on rat brain slices showed different distribution patterns and pharmacological properties between the two radioligands. The [11C]4 showed the highest binding in the striatum, which was inhibited not only by 10 microM 4 but also by 10 microM raclopride, a selective dopamine D2 receptor antagonist. In contrast, [11C]9 showed the highest binding in the cerebral cortex, which was inhibited by several D4 receptor antagonists (9, RBI-254, L-745,870), but not by any other receptor ligands (D1/D5, D2/D3, 5-HT1A, 5-HT2A, sigma1 and alpha1) tested. In vivo brain distribution of [11C]9 in rat showed the highest uptake in the frontal cortex, a region that has a high density of D4 receptors. These results indicate that the pharmacological property of [11C]9 matches the rat brain D4 receptors, but that of [11C]4 rather appears to match the rat brain D2 receptors. The results for the benzamide [11C]9 prompted us to further evaluate its potential as a PET radioligand for D4 receptors by employing PET on monkey brain. Unfortunately, in contrast to rats, neither specific binding nor differences in regional uptake of radioactivity were observed in monkey brain after intravenous 11C]9 injection. Based on that specific activities of radioligands might be critical in mapping the neurotransmitter receptors if they are only faintly expressed in the brain, 11C]9 with an extremely high specific activity (1810 GBq/micromol) was used for PET study. However, the effort to determine the specific binding for D4 failed. These results indicate that both of the benzamide derivatives would not be suitable radioligands for D4 receptors with PET.

Synthesis and preliminary evaluation of [18F]FEtP4A, a promising PET tracer for mapping acetylcholinesterase in vivo.

N-[18F]Fluoroethyl-4-piperidyl acetate ([18F]FEtP4A), an analog of [11C]MP4A for mapping brain acetylcholineseterase (AchE) activity, was prepared by reacting 4-piperidyl acetate (P4A) with [18F]fluoroethyl bromide ([18F]FEtBr) using a newly developed automated system. Preliminary evaluation showed that the initial uptake of [18F]FEtP4A in the mouse brain was > 8% injected dose/g tissue. The distribution pattern of [18F]FEtP4A in the brain was striatum>cerebral cortex>cerebellum within 10-120 min post-injection, which reflected the distribution rank pattern of AchE activity in the brain. Moreover, chemical analysis of in vivo radioactive metabolites in the mouse brain indicated that 83% of [18F]FEtP4A was hydrolyzed to N-[18F]fluoroethyl-4-piperidinol ([18F]FEtP4OH) after 1 min intravenous injection. From these results, [18F]FEtP4A may become a promising PET tracer for mapping the AchE in vivo.

Synthesis and evaluation of 3-(4-chlorobenzyl)-8-[11C]methoxy-1,2,3,4-tetrahydrochromeno[3,4-c]pyridin-5-one: a PET tracer for imaging sigma(1) receptors.

3-(4-Chlorobenzyl)-8-methoxy-1,2,3,4-tetrahydrochromeno[3,4-c]pyridin-5-one (1), a putative dopamine D(4) receptor antagonist (k(i) = 8.7 nM), was labeled by positron-emitter (11C) and its pharmacological evaluation was carried out with in vitro quantitative autoradiography and positron emission tomography (PET). 11C-Methylation of a corresponding desmethyl precursor (2) with [11C]CH(3)I gave [11C]1 with >or=98% of radiochemical purity after HPLC purification and 67-90 GBq/micromol of specific activity at the end of synthesis. The in vitro autoradiography using rat brain sections demonstrated that [11C]1 shows no specific binding to the D(4) receptors, but a high specific binding to sigma(1) receptors (IC(50) = 105 nM). In the PET study with monkey brain, [11C]1 was highly taken up by the brain and trapped in the brain for at least 90 min. The distribution pattern of radioactivity in the brain was striatum > thalamus > frontal cortex > cerebellum, which was same as the result of in vitro autoradiography. Pre-treatment with non-radioactive 1 (1 mg/kg) produced a significant reduction of radioactivity in all the regions including the cerebellum. Pre-treatment with (+)pentazocine (1 mg/kg), a selective sigma(1) receptor agonist, also reduced the radioactivity in the same regions to a similar extent. These results indicate that [11C]1 may have some specific binding to the sigma(1) receptors, which is consistent with the result of in vitro autoradiography.

Synthesis, in vitro and in vivo pharmacology of a C-11 labeled analog of CP-101,606, (+/-)threo-1-(4-hydroxyphenyl)-2-[4-hydroxy-4-(p-[11C]methoxyphenyl)piperidino]-1-propanol, as a PET tracer for NR2B subunit-containing NMDA receptors.

A carbon-11 labeled methoxyl analog of CP-101,606, (+/-)threo-1-(4-hydroxyphenyl)-2-[4-hydroxy-4-(p-[11C]methoxyphenyl)piperidino]-1-propanol [(+/-)[11C]1], was synthesized as a new subtype-selective PET radioligand for NMDA receptors. The in vitro binding studies using rat brain slices demonstrated that (+/-)[11C]1 shows an extremely high-specific binding to the NR2B subunit of NMDA receptors. In contrast to the in vitro binding, the in vivo binding to mouse and monkey brains showed no apparent specific localization of the radioactivity in any of the brain regions. Metabolism and physicochemical properties such as the lipophilicity of (+/-)[11C]1 seemed unlikely to affect the in vivo (+/-)[11C]1 binding. Among the various endogenous ligands acting at the NMDA receptors, polyamines (spermine and spermidine) and divalent cations (Mg(2+,) Zn(2+,) and Ca(2+)) strongly inhibited the in vitro (+/-)[11C]1 binding. Thus, the present studies point to the possibility that the polyamines and cations behave as endogenous inhibitors for (+/-)[11C]1 binding, leading to the loss of the specific binding in vivo.

Development of an automated system for synthesizing 18F-labeled compounds using [18F]fluoroethyl bromide as a synthetic precursor.

An automated system was developed to synthesize 18F-labeled compounds using [18F]fluoroethyl bromide ([18F]FEtBr) as a synthetic precursor. The apparatus makes possible the following sequence of processes: (1) production of an aqueous solution of [18F]fluoride ([18F]F-), (2) recovery of [18F]F- from target chamber, (3) drying of [18F]F-, (4) formation and distillation of [18F]FEtBr into a trapping vessel, (5) alkylation of target compounds with [18F]FEtBr, (6) High performance liquid chromatography purification and (7) formulation. [18F]FEtBr, the synthetic precursor for fluoroethylation, was labeled via nucleophilic displacement of 2-trifluoromethanesulfonyloxy ethylbromide (BrCH2CH2OTf) with [18F]F- and was purified from the reaction mixture by distillation. After the conditions for forming [18F]FEtBr and drying [18F]F- were optimized, [18F]FEtBr was obtained in a radiochemical yield of 71 +/- 13% (n = 21, based on [18F]F-, corrected for decay) and a radiochemical purity of 98 +/- 1.4% at end of the syntheses (EOS). Using this automated system, [18F]fluoroethylspiperone ([18F]FEtSP) was prepared by reacting spiperone with [18F]FEtBr in a radiochemical yield and purity of 56 +/- 12% (n = 5, based on [18F]FEtBr, corrected for decay) and 97 +/- 1.5% with a specific activity of 310 +/- 120 GBq/mumol at EOS. The total synthesis time was 55 +/- 2.3 min from the end of bombardment and the developed system has proved to be reliable and reproducible.

Measurement of glycine binding site of N-methyl-D-asparate receptors in living human brain using 4-acetoxy derivative of L-703,717, 4-acetoxy-7-chloro-3-[3-(4-[11c] methoxybenzyl) phenyl]-2(1H)-quinolone (AcL703) with positron emission tomography.

N-methyl-D-aspartate (NMDA) receptors are of major interest in brain functions and neuropsychiatric disorders. However, at present there are few suitable radioligands for in vivo imaging of NMDA receptors. 7-choloro-4-hydroxy-3-[3-(4-methoxybenzyl) phenyl]-2(1H)-quinolone (L-703,717) is one of the potent ligands for the glycine-binding site of NMDA receptors. 4-Acetoxy derivative of L-703,717 (AcL703) is a candidate, as a positron emission tomography (PET) ligand for NMDA receptors, because of its better permeability at the blood-brain barrier compared with L-703,717. After intravenous injection of 624-851 MBq of [11C]AcL703, dynamic PET scan was performed on six healthy males for 90 min. Regions-of-interest were located on the cerebral cortices, cerebellar cortex, and cerebral white matter. The binding potential (BP) was calculated from the ratio of the area under the curve (AUC) of radioactivities from 40 to 90 min in the target region to that in white matter. Regional radioactivities reached close to equilibrium in all regions after about 40 min postinjection. Regional brain uptake of [11C]AcL703 at 40 min after injection was 0.00028-0.00065% of the injected dose/milliliter. Radioactivity concentration of [11C]AcL703 was highest in the cerebellar cortex and lowest in white matter. AUC in the cerebellar cortex was higher than those of cerebral cortices, thalamus, striatum, and white matter. BP in the cerebellar cortex was twofold higher than in the cerebral cortices (cerebellar cortex: BP=2.20+/-0.72; cerebral cortices: BP=1.05+/-0.45). Despite the low brain uptake of [11C]AcL703, regional distributions were in good agreement with our previous studies of rodents. This indicates the possibility of in vivo evaluation of NMDA receptors using PET with [11C]AcL703 in living human brain.

Synthesis of 11C-labelled bis(phenylalkyl)amines and their in vitro and in vivo binding properties in rodent and monkey brains.

Two new (11)C-labelled ligands, N-(3-(4-hydroxyphenyl)propyl)-3-(4-methoxyphenyl)propylamine ([(11)C]2) and N-(3-(4-hydroxyphenyl)butyl)-3-(4-methoxyphenyl)butylamine ([(11)C]3) were designed based on bis(phenylalkyl)amines (1) which have been reported as polyamine site antagonists with high-selectivity for NR1A/2B NMDA receptors, and radiolabelling of the corresponding phenol precursors with [(11)C]methyl iodide was readily accomplished. The in vitro inhibition experiments using rat brain slices showed that [(11)C]2 and [(11)C]3 share the binding sites with spermine and/or ifenprodil but not with CP-101,606, a highly potent NR2B-selective NMDA antagonist, and that divalent cations such as Zn(2+) produced significant inhibition of both [(11)C]2 and [(11)C]3 bindings. Intravenous injection of [(11)C]3 in mice showed almost homogeneous distribution throughout the brain. Attempts to block the tracer uptake of [(11)C]3 by pre-injection with the unlabelled 3 or spermine in rats were unsuccessful, but a small decrease in the cerebral uptake of [(11)C]3 by co-treatment with the unlabelled 3 was observed in a monkey PET study. The present findings indicate that none of these (11)C-labelled analogues have potential for PET study of binding sites on the N-methyl-D-aspartate (NMDA) receptors.

Effects of endogenous agonists, glycine and D-serine, on in vivo specific binding of [11C]L-703,717, a PET radioligand for the glycine-binding site of NMDA receptors.

A positron-emitter (carbon-11) labeled antagonist for the glycine-binding site of NMDA receptors, [(11)C]L-703,717, has a unique in vivo binding characteristic, in which it preferentially binds to cerebellar-specific NMDA receptors consisting of a GluRepsilon3 subunit and eventually accumulates in rodent cerebellum under in vivo conditions, but not under in vitro conditions. In order to understand the in vivo-specific site and subunit localization of this radioligand, we examined the effect of the endogenous glycine site agonists, glycine and D-serine, on in vivo [(11)C]L-703,717 binding. An increase in extracellular glycine concentration by treatment with a glycine transporter 1 (GlyT1)-selective inhibitor, NFPS ethyl ester, significantly decreased the cerebellar localization of [(11)C]L-703,717 in rats. D-serine is known to be concentrated in mammalian forebrain regions. The lack of D-serine detection in the cerebellum may be due to the fact that it has the highest enzymatic activity of D-amino acid oxidase (DAO). It was found that the cerebellar localization of [(11)C]L-703,717 is greatly diminished in mutant mice lacking DAO, in which D-serine content in the cerebellum is drastically increased from a nondetectable level in normal mice. These studies indicate that [(11)C]L-703,717 is susceptible to inhibition by glycine site agonists in its in vivo binding, and suggest that regional differences in inhibitions by endogenous agonists may be a crucial factor in the site- and subunit-specific binding of this glycine-site antagonist.

A positron-emitter labeled glycine(B) site antagonist, [(11)C]L-703,717, preferentially binds to a cerebellar NMDA receptor subtype consisting of GluR epsilon3 subunit in vivo, but not in vitro.

In previous studies, we have found that [(11)C]L-703,717, a positron-emitter labeled antagonist for the glycine-binding site of NMDA receptors, only localizes in rodent cerebellum under in vivo conditions. In order to understand the unusual cerebellar localization, we have examined the binding of [(11)C]L-703,717 to a cerebellar-specific NMDA receptor subtype consisting of GLuRepsilon3 subunit, by comparing its autoradiographic distributions between GluRepsilon3-deficient and wild-type mice. Ex vivo [(11)C]L-703,717 binding to wild-type mice showed a highly specific localization of radioactivity in the cerebellum, whereas that to the GluRepsilon3-deficient mice showed no specific localization of radioactivity in any of the brain regions. In contrast to the ex vivo binding, in vitro [(11)C]L-703,717 binding displayed a similar binding characteristic between GluRepsilon3-deficient and wild-type mice with highly specific localizations in the hippocampus and cerebral cortex. Therefore, the present study clearly demonstrated that [(11)C]L-703,717 preferentially binds to a cerebellar NMDA receptor subtype consisting of GluRepsilon3 subunit in vivo, but not in vitro.

Synthesis and brain regional distribution of [11C]NPS 1506 in mice and rat: an N-methyl-D-aspartate (NMDA) receptor antagonist.

NPS 1506 [3-fluoro-gamma-(3-fluorophenyl)-N-methylbenzenepropamine] is representative of a non-psychotomimetic class of N-methyl-D-aspartate (NMDA) receptor antagonists. [11C]NPS 1506 was prepared at high radiochemical purity (>98%) with a specific activity of around 50 GBq/micromol at the end of synthesis by methylation of the desmethyl precursor with [11C]methyl iodide in the presence of NaH. Biodistribution of [11C]NPS 1506 in mice and rat demonstrated that uptake into the brain was rapid and occurred at high levels. [11C]NPS 1506 showed no appreciable specific binding in rodent brains under in vivo conditions, possibly because of both a large non-specific bound fraction and low in vitro binding affinity for NMDA receptors.