Synthesis, Fluorine-18 Radiolabeling, and Biological Evaluation of N-((E)-4-Fluorobut-2-en-1-yl)-2β-carbomethoxy-3β-(4′-halophenyl)nortropanes: Candidate Radioligands for In Vivo Imaging of the Brain Dopamine Transporter with Positron Emission Tomography

Abstract

The N-(E)-fluorobutenyl-3β-(para-halo-phenyl)nortropanes 9-12 were synthesized as ligands of the dopamine transporter (DAT) for use as ¹⁸F-labeled positron emission tomography (PET) imaging agents. In vitro competition binding assays demonstrated that compounds 9-12 have a high affinity for the DAT and are selective for the DAT compared to the serotonin and norepinephrine transporters. MicroPET imaging with [¹⁸F]9-[¹⁸F]11 in anesthetized cynomolgus monkeys showed high uptake in the putamen with lesser uptake in the caudate, but significant washout of the radiotracer was only observed for [¹⁸F]9. PET imaging with [¹⁸F]9 in an awake rhesus monkey showed high and nearly equal uptake in both the putamen and caudate with peak uptake achieved after 20 min followed by a leveling-off for about 10 min and then a steady washout and attainment of a quasi-equilibrium. During the time period 40-80 min post-injection of [¹⁸F]9 the ratio of uptake in the putamen and caudate vs. cerebellum uptake was ≥ 4.

Introduction

The human dopamine transporter (DAT)^a is a 620-amino acid transmembrane protein which belongs to the family of Na⁺/Cl^- dependent transporters.^{3, 4} In the central nervous system (CNS) the DAT is located on presynaptic neurons and functions to remove the neurotransmitter dopamine from the synapse thereby terminating the signaling action of dopamine.^{5, 6} The DAT is found in high densities in certain brain regions which include the putamen, caudate, nucleus accumbens, and olfactory tubercle whereas lower densities are found in the substantia nigra, amygdala, and hypothalamus.^{7, 8} Several neuropsychiatric disorders have been associated with the DAT including Parkinson's Disease (PD),^9-11 attention-deficit hyperactivity disorder (ADHD),¹² supranuclear palsy,¹⁰ and Tourette's Syndrome.¹³ The ability to image the DAT with positron emission tomography (PET) may aid in the diagnosis, monitoring, and treatment of these diseases by providing a means to study the DAT in vivo and by allowing for the in vivo measurement of DAT density in specific brain regions.¹⁴ Furthermore, a suitable DAT PET tracer can be used to measure the occupancy of DAT therapeutics¹⁵ and may facilitate the development of new DAT therapeutics.¹⁶

The DAT is also the target of several drugs of abuse including cocaine,^{17, 18} amphetamines, and MDMA (ecstasy) and this has led to the search for compounds that can be employed as potential cocaine addiction therapeutics.^{19, 20} From this research evolved the 3β-phenyl tropane class of DAT ligands of which compounds 1-6 were the first to be prepared.^21-24 These compounds have since been exploited for PET imaging due to the ability to radiolabel with carbon-11 on either the N-methyl group or the O-methyl ester.^25-28 Carbon-11 has a half-life of 20.4 min which limits the use of ¹¹C-labeled tracers to the location where they are prepared and to imaging sessions of about 2 h. Fluorine-18 has a half-life of 109.8 min which allows for longer radiosynthesis times and imaging sessions, and also for the transport of the ¹⁸F radiotracer to PET imaging facilities that do not have onsite cyclotrons. Additionally, ¹⁸F positrons have a lower maximum energy (0.64 MeV)²⁹ than ¹¹C positrons (0.97 MeV) which therefore deposits less energy into tissue and also results in a positron with a shorter linear range which allows for higher spatial resolution.^{30, 31} These physical properties of ¹⁸F are serendipitous due to the increasingly valuable role that ¹⁹F is playing in medicinal chemistry^32-35 and a variety of synthetic methods have now been developed to incorporate ¹⁸F or ¹⁹F into molecules.^36-38

Out of compounds 1-6 only compound 2 has the potential to be radiolabeled with ¹⁸F. Thus, [¹⁸F]2 was prepared and evaluated in rats³⁹ and humans.⁴⁰ In humans the peak uptake of [¹⁸F]2 in the putamen and caudate occurred at 225 min followed by a slow washout. Although [¹⁸F]2 showed high uptake in the striatum and the aryl-fluorine bond was found to be metabolically stable to defluorination, the time to peak uptake was considered to take too long and so [¹⁸F]2 was not an optimal DAT PET tracer. Numerous other fluorinated derivatives of 1-6 which contain the ¹⁸F-radiolabel as an N-fluoroalkyl group or an O-fluoroalkyl ester have also been prepared and evaluated.⁴¹ Among the numerous derivatives reported, compounds 7 (FECNT)⁴² and 8 (FPCIT)⁴³ emerged as viable DAT PET tracers and have found use in human PET imaging. ^44-50 Both compounds [¹⁸F]7 and [¹⁸F]8 achieve a high uptake and specific binding in the putamen and caudate but neither compound washes out significantly during the course of the study which is needed to enable kinetic modeling of the tracer behavior.^{51, 52} Compound 8 also has a higher binding affinity at the serotonin transporter (SERT) than the DAT⁵³ and so [¹⁸F]8 is not a DAT-specific PET tracer. Additionally, [¹⁸F]8 defluorinates, which is similar to other tracers containing an [¹⁸F]fluoropropyl group,^54-56 whereas [¹⁸F]7 is metabolized to a polar radiometabolite⁵⁷ which can cross the blood-brain barrier (a similar result was recently reported for the amyloid imaging agent [¹⁸F]FDDNP⁵⁸). Thus, further improvements are still needed in order to obtain a DAT-selective ¹⁸F-labeled PET tracer that can achieve peak uptake, binding equilibrium, and washout in a relatively short period of time while also not defluorinating or producing radiolabeled metabolites that can cross the blood-brain barrier.

We have previously reported the N-(E)-fluorobutenyl compounds 9-12 along with preliminary in vitro binding data and rat biodistribution data.^{59, 60} The rationale for the N-(E)-fluorobutenyl group was based upon the characterization and imaging properties of the iodine-123 DAT imaging agent N-((E)-3-[^123,125I]iodopropen-1-yl-2β-carbomethoxy-3β-(4-chlorophenyl)nortropane (MMG-142E/IPT)⁶¹ which showed nanomolar DAT affinity and high striatal to cerebellar ratios with relatively fast washout kinetics, indicating reversible binding, in non-human and human primates.^{62, 63} We hypothesized that the N-(E)-4-[¹⁸F]fluorobut-2-en-1-yl group would serve as a bioisostere for the N-(E)-3-[^{123, 125}I]iodopropen-1-yl group for a new class of DAT PET imaging agents. The preliminary binding data of compounds 9-12 demonstrated that the (E)-configuration of the fluorobutenyl group was the more biologically active configuration and also that halosubstitution of the phenyl ring resulted in higher binding affinity at the DAT when compared to the unsubstituted phenyl group. Subsequently, the para-methyl analog 13 was reported⁶⁴ and, recently, several related derivatives have been reported.^{65, 66} Herein we report the synthesis and binding affinity determination of 9-12 (along with 13 for comparison) in conjunction with the microPET imaging of [¹⁸F]9-[¹⁸F]11 and [¹⁸F]13 in anesthetized cynomolgus monkeys and the high resolution research tomograph (HRRT) imaging of [¹⁸F]9 in an awake rhesus monkey.⁶⁷

Chemistry

The N-(E)-fluorobutenyl nortropanes 9-13 were synthesized by reacting the appropriate nortropane^{24, 25, 68} with (E)-4-fluoro-1-tosyloxy-2-butene (14) (Scheme 1). Compound 14 was prepared from trans-1,4-dibromo-2-butene (15) as shown in Scheme 1. Reaction of 15 with KOAc in AcOH according to the literature procedure^{69, 70} afforded the diacetoxy compound 16 which was purified on silica rather than by the traditional distillation method. Acid-catalyzed ethanolysis of 16 afforded the diol 17 in nearly quantitative yield. Thus, this method provides a simplified synthesis of diol 17 which avoids the previously reported LAH reduction of 2-butyne-1,4-diol⁷¹ or DIBAL reduction of a dialkyl fumarate.^{72, 73} Diol 17 was reacted with tosyl chloride to give ditosylate 18 which was then reacted with tetrabutylammonium fluoride to give 14.

Scheme 1.

Radiochemistry

Compound [¹⁸F]14 was obtained by reacting 18 with K¹⁸F/Kryptofix-222 complex in CH₃CN. Following preparation of [¹⁸F]14, a DMF-solution of the appropriate nortropane^{24, 25, 68, 74} was added, the mixture was heated at 105 °C for 15 min, and then purified by semi-preparative HPLC. The desired HPLC fractions were combined and diluted with H₂O, and the product was then isolated from this solvent mixture by solidphase extraction according to a previously reported procedure.⁷⁵ The radiotracer was then formulated as a 10% EtOH/saline solution. The octanol/water partition coefficients^{76, 77} of [¹⁸F]9-[¹⁸F]11 were measured according to a previously reported procedure^{78, 79} and the results are shown in Table 1. These values are all in the range of log P = 1-3 which allows for passive diffusion of the radiotracer across the blood-brain barrier. As expected, the lipophilicity of each radiotracer increases with increasing lipophilicity of the halogen substituent.

Table 1.

Octanol/Water Partition Coefficients.

compound	log P7.4a	n
[18F]9	1.95 ± 0.01	8
[18F]10	2.21 ± 0.02	7
[18F]11	2.33 ± 0.03	8

^aAverage value of n determinations ± the standard deviation.

In Vitro Competition Binding Assays

The binding affinities of 2 and the N-(E)-fluorobutenyl nortropanes 9-13 to human monoamine transporters (Table 2) were determined using in vitro competition binding assays with membranes prepared from cells transfected with the human SERT, DAT, or norepinephrine transporter (NET) according to our previously reported procedure.^{80, 81} Compounds 2 and 9-13 were screened as the free bases whereas salts were used for the control compounds: cocaine•HCl (DAT), (S)-citalopram•oxalate⁸² (SERT), and desipramine•HCl⁸³ (NET). The competing ligands employed were [³H](R,S)-citalopram•HBr⁸⁴ (SERT), [¹²⁵I]RTI-55²³ ([¹²⁵I]5, DAT), or [³H]nisoxetine⁸⁵ (NET).

Table 2.

Results of In Vitro Competition Binding Assays with Transfected Human Monoamine Transporters.

compd	Ki (nM)a			DAT selectivity
	hDAT	hSERT	hNET	SERT/DAT	NET/DAT
9	9.5 ± 2.6d	357.4 ± 193.0g	2607 ± 1432d	∼ 38	∼ 274
10	0.6 ± 0.3c	11.3 ± 6.4h	142 ± 30e	∼ 19	∼ 237
11	0.4 ± 0.0c	8.5 ± 0.8d	87 ± 8	∼ 21	∼ 218
12	0.5 ± 0.2c	2.2 ± 1.1d	145 ± 125d	∼ 4	∼ 290
13	2.7 ± 0.1c	147.0 ± 79.3i	265 ± 146e	∼ 54	∼ 98
2	10.6 ± 0.8b	N/D	N/D	N/A	N/A
Cocaine•HCl	46.3 ± 13.5c	N/D	N/D	N/A	N/A
(S)-Citalopram•oxalate	N/D	1.0 ± 0.6	N/D	N/A	N/A
Desipramine•HCl	N/D	N/D	0.8 ± 0.7b	N/A	N/A

^aGeometric mean of n determinations ± the standard deviation (each determination performed in triplicate).

^bn = 2.

^cn = 3.

^dn = 4.

^en = 5.

^fn = 6.

^gn = 7.

^hn = 8.

ⁱn = 9.

N/A = Not applicable.

N/D = Not determined.

Compound 12, which is the N-(E)-fluorobutenyl analog of 5 (RTI-55/β-CIT), has a high affinity for the DAT as well as for the SERT and thus displays only a four-fold selectivity for the DAT over the SERT (this lack of selectivity is very similar to that of 5^{53, 86}). Compounds 10 and 11 also have a high DAT affinity and display about a 20-fold selectivity over the SERT. This high DAT affinity and moderate SERT affinity is also similar to the N-methyl analogs 3 (RTI-31) and 4 (RTI-51), respectively.⁸⁷ Compound 13 has a DAT affinity similar to the N-methyl analog 6 (RTI-32)⁸⁷ and is about 54-times more selective for the DAT over the SERT. Compound 9 has a reduced DAT affinity relative to compounds 10-13 but the affinity is very similar to the N-methyl analog 2 (the binding affinity determined for 2 is in close agreement with other reported values^{22, 25, 88}) indicating that replacement of the N-methyl group with an N-(E)-fluorobutenyl group does not significantly alter DAT binding. Compound 9 is also more selective for the DAT over the SERT than the other halo-substituted compounds 10-12. All of the N-(E)-fluorobutenyl tropanes tested have a high DAT vs. NET selectivity with the halo-substituted compounds 9-12 displaying a selectivity more than twice that of the methyl-substituted compound 13.

In Vivo Nonhuman Primate PET Imaging

MicroPET imaging was performed in anesthetized cynomolgus monkeys using a Concorde microPET P4 according to our previously reported procedure.⁸⁰ The microPET images for a 145-minute baseline study with [¹⁸F]9 are shown in Figure 1 and the time-activity curves (TACs) are shown in Figure 2. As shown by Figures 1 and 2, high uptake of radioactivity is observed in the regions of the brain known to have high DAT density.^{7, 8} Rapid uptake of [¹⁸F]9 into the putamen is observed (Figure 2) with peak uptake achieved within 20 min followed by a steady washout. Uptake of [¹⁸F]9 in the caudate is less than that observed in the putamen along with a slower rate of uptake and a minimal washout during the first 75 min of the study. Uptake of [¹⁸F]9 in the substantia nigra is rapid with peak uptake achieved within 10 min followed by a steady washout. Uptake in the cerebellum is rapid followed by a rapid washout with little or no retention of the tracer as expected for a brain region devoid of DAT. A small amount of radioactivity is also observed in the skull (Figure 1) indicating a low degree of defluorination. This is not surprising since the ¹⁸F-radiolabel is in an allylic position and it has been previously reported that ¹⁸F-benzyl fluorides can also undergo defluorination.^{89, 90} In order to demonstrate that the uptake of [¹⁸F]9 observed in the baseline study is a result of preferential binding to the DAT, a chase study was performed with the DAT ligand RTI-113•HCl^{87, 91, 92}19. As shown in Figure 3, administration of 19 (0.3 mg/kg) at 90 min post-injection results in a complete displacement of [¹⁸F]9 thus indicating that the observed uptake was a result of binding to the DAT.

Figure 1.

MicroPET images (summed 0-145 min) obtained by injection of [¹⁸F]9 into an anesthetized cynomolgus monkey.

Figure 2.

MicroPET baseline TACs obtained by injection of [¹⁸F]9 into an anesthetized cynomolgus monkey.

Figure 3.

MicroPET TACs showing the result of injection of 19 (0.3 mg/kg) into an anesthetized cynomolgus monkey at 90 min post-injection of [¹⁸F]9.

MicroPET baseline studies were also performed with [¹⁸F]10, [¹⁸F]11, and [¹⁸F]13 (Figures 4-6, respectively). All three tracers show high uptake in the putamen followed by minor washout during the course of the 235-min study. The initial uptake into the putamen is rapid with peak uptake achieved within 30 min for [¹⁸F]10 and [¹⁸F]11, and within 55 min for [¹⁸F]13. Uptake in the caudate is less than that observed in the putamen for [¹⁸F]10, [¹⁸F]11, and [¹⁸F]13 and the rate of uptake is also slower. Furthermore, the uptake in the caudate fails to wash out during the entire course of the study for [¹⁸F]10, [¹⁸F]11, and [¹⁸F]13. The different levels of uptake of [¹⁸F]13 in the putamen and caudate and the lack of washout are very similar to the results previously reported for [¹¹C]13 in an anesthetized baboon.⁶⁴

Figure 4.

MicroPET baseline TACs obtained by injection of [¹⁸F]10 into an anesthetized cynomolgus monkey.

Figure 6.

MicroPET baseline TACs obtained by injection of [¹⁸F]13 into an anesthetized cynomolgus monkey.

It has been demonstrated in numerous instances that anesthesia can affect the behavior of PET tracers.^93-98 Therefore, an 85-min PET study was performed on a high resolution research tomograph (HRRT) with [¹⁸F]9 in an awake rhesus monkey to determine whether there is a difference between the behavior of [¹⁸F]9 in awake and anesthetized states (Figures 7 and 8). As shown in Figure 8, the uptake of [¹⁸F]9 is nearly equal in the putamen and caudate in each hemisphere of the awake rhesus monkey brain. This suggests that the different levels of uptake between the putamen and caudate observed in the anesthetized monkey microPET studies shown in Figures 2-6 is due to anesthesia effects. The rate of uptake of [¹⁸F]9 into the putamen and caudate of the awake rhesus monkey is fast with peak uptake achieved after about 20 min followed by a leveling-off for a period of about 10 min and then a slow but steady washout during which time a quasi-equilibrium¹⁴ is achieved. The kinetic behavior of [¹⁸F]9 in the putamen and caudate of the awake monkey is very similar to the kinetic behavior in the putamen of the anesthetized monkey suggesting that the overall kinetics of [¹⁸F]9 in the putamen are not significantly altered by anesthesia whereas the different degree of uptake and lack of washout from the caudate in the anesthetized monkey may be the result of anesthesia. Based on these results, it is believed that the rate of washout from the putamen for [¹⁸F]10, [¹⁸F]11, and [¹⁸F]13 observed in the anesthetized monkey microPET studies would not change significantly in an awake state.

Figure 7.

HRRT PET images (summed 55-75 min) obtained by injection of [¹⁸F]9 into an awake rhesus monkey.

Figure 8.

HRRT baseline TACs obtained by injection of [¹⁸F]9 into an awake rhesus monkey.

The difference in the rate of washout between [¹⁸F]9-[¹⁸F]11 and [¹⁸F]13, aside from any anesthesia effects, is believed to be the result of the different binding affinities shown in Table 2. The DAT is found in high density in the caudate and putamen and, therefore, a very high-affinity ligand is not needed to image the DAT.⁹⁹ Compounds 10 and 11 have a sub-nanomolar affinity for the DAT and compound 13 has an affinity of 2.7 nM, whereas 9 has an affinity of about 9.5 nM. The lower affinity compound [¹⁸F]9 is able to bind to the DAT, then dissociate and washout whereas the higher affinity compounds have a delayed washout due to stronger and/or prolonged binding to the DAT in combination with the high density of available binding sites.

Table 3 shows the ratio of uptake in the putamen and caudate vs. cerebellum uptake in the awake study with [¹⁸F]9 and the uptake ratios are plotted vs. time in Figure 9. From Table 3 and Figure 9 it can be seen that the highest uptake ratios are obtained after 60 min post-injection and that ratios of ≥ 4 are maintained for the period 40-80 min post-injection. From the HRRT PET images in Figure 7 it can be seen that the uptake of [¹⁸F]9 is in agreement with the known distribution of DAT in the brain. But, just as was seen with the microPET images in Figure 1, visualization of the skull indicates some degree of defluorination. Compound [¹⁸F]9, therefore, displays many of the desired properties of an ¹⁸F-labeled DAT PET tracer including selectivity for the DAT over the SERT and NET, and the achievement of peak uptake and binding equilibrium in a short time frame followed by a steady washout. Unfortunately, a minor degree of defluorination at the allylic position occurs. But, it may be possible to block this defluorination by the administration of disulfiram which has already been proven to inhibit defluorination of [¹⁸F]FCWAY.¹⁰⁰

Table 3.

Ratio of Uptake of [¹⁸F]9 in the Caudate and Putamen vs. Cerebellum Uptake at Selected Time Points for the HRRT Awake Rhesus Monkey Study Shown in Figure 7.

Time (min)	Putamen (left)	Caudate (left)	Putamen (right)	Caudate (right)
30	3.8	3.5	4.2	4.0
40	4.4	4.3	4.8	4.8
50	4.7	4.5	5.2	5.0
60	4.8	4.6	5.5	5.2
70	4.3	4.1	5.0	4.9
82.5	4.0	3.9	4.6	4.6

Figure 9.

Graph of the ratio of uptake of [¹⁸F]9 in the caudate and putamen vs. cerebellum uptake with time for the HRRT awake rhesus monkey study shown in Figure 8.

Summary

The N-(E)-fluorobutenyl-3β-(para-halo/methyl-phenyl)nortropanes 9-13 were synthesized by reacting the respective nortropane with 14, which was synthesized in 4 steps from 15. In vitro competition binding assays demonstrated that 9-13 have binding affinities and selectivities similar to their N-methyl analogs and that the chloro-, bromo-, and iodo-derivatives 10-12, respectively, bind to the DAT with a 15-fold or greater affinity than the fluoro-derivative 9. MicroPET imaging in anesthetized cynomolgus monkeys with [¹⁸F]9-[¹⁸F]11 and [¹⁸F]13 demonstrated that this very high binding affinity of the chloro-, bromo-, and methyl-derivatives prevented the radiotracer from significantly washing out of the DAT-rich brain regions during the course of the study. This apparent irreversible binding is believed to result from the combination of high DAT affinity of the radiotracer and high DAT density in the putamen and caudate. The lower affinity radiotracer [¹⁸F]9 achieved rapid peak uptake in the putamen followed by a steady washout whereas uptake in the caudate barely washed out during the course of the study. HRRT PET imaging with [¹⁸F]9 in an awake rhesus monkey showed rapid and high uptake in both the putamen and caudate with peak uptake achieved after 20 min post-injection followed by a steady washout from both the putamen and caudate thus indicating that the lack of washout in the microPET study was most likely due to an anesthesia effect. Uptake ratios of ≥ 4 were obtained in the putamen and caudate during the period 40-80 min post-injection of [¹⁸F]9 in the awake study. Thus, in an awake state, [¹⁸F]9 is able to achieve rapid peak uptake in the putamen and caudate with high uptake ratios relative to cerebellum uptake. A minor drawback of the [¹⁸F]-fluorobutenyl group is that some degree of defluorination is observed but it may be possible to prevent this defluorination by administering disulfiram prior to the imaging session.

Experimental Section

General

Trans-1,4-dibromo-2-butene (15) was purchased from both Acrös and Aldrich. Crystallization of tropanes was performed by stirring a refluxing hexanes solution of the tropane for 10 min, decanting the hot solution into a pre-heated 25-mL Erlenmeyer flask, capping the flask with a rubber septum, and storing the flask in a freezer (-15 °C). NMR spectra were obtained on Varian Unity and Mercury spectrometers at the specified frequencies. ¹H chemical shifts are referenced to internal TMS or residual CHCl₃ (7.26 ppm) and ¹³C chemical shifts are referenced to CDCl₃ (77.23 ppm). Silica gel used was EMD Silica Gel 60, 40-63 μm. Vacuum flash chromatography was performed by placing silica in a medium-fritted filter (31 cm length × 4 cm i.d. – Kontes Glassware #956250-0044 with adapter #205000-2440), eluting under vacuum, and collecting fractions in 125-mL flat-bottomed boiling flasks. Radial chromatography was performed with a Harrison Research Chromatotron. Semipreparative HPLC: Waters XTerra Prep RP₁₈, 5 μm, 19 × 100 mm + guard cartridge (19 × 18 10 mm), 60:40:0.1 v/v/v MeOH/H₂O/NEt₃. Analytical HPLC: Waters NovaPak 3.9 × 150 mm, 75:25:0.1 v/v/v MeOH/H₂O/NEt₃. HRMS was performed by the Emory University Mass Spectrometry Center. Compound purity was determined by elemental analysis and was found to be >95%. Elemental analysis was performed by Atlantic Microlab, Inc. (www.atlanticmicrolab.com).

N-((E)-4-fluorobut-2-en-1-yl)-2β-carbomethoxy-3β-(4′-fluorophenyl)nortropane (9)

Nor-β-CFT²⁵ (145 mg, 0.55 mmol), 14 (129 mg, 0.53 mmol), i-Pr₂NEt (0.12 mL, 0.69 mmol), and CHCl₃ (25 mL) were stirred at reflux under Ar_(g) for 17 h and then cooled to room temperature. The solution was poured onto dry silica (43 mm h × 43 mmi.d.) and eluted under vacuum: CH₂Cl₂ (25 mL), hexanes (50 mL), hexanes/EtOAc/Net₃ v/v/v 90:8:2 (50 mL), 75:20:5 (200 mL). The desired fractions were combined and the solvent was removed to give a yellow syrup that was further purified by radial chromatography (2 mm silica, 98:1:1 v/v/v hexanes/EtOAc/NEt₃ (300 mL)) to afford a colorless syrup (152 mg). Crystallization from refluxing hexanes (3 mL) afforded white crystals (122 mg, 66%): ¹H NMR (600 MHz, CDCl₃) δ 7.22 (dd, 2 H, J = 8.4 Hz, J = 5.4 Hz), 6.95 (apparent t, 2 H, J = 8.4 Hz), 5.78 (m, 2 H), 4.83 (dd, 2 H, J_HF = 47.4 Hz, J = 4.8 Hz), 3.66 (partially resolved dd, 1 H, J = 3.3 Hz, J = 6.9 Hz), 3.50 (s, 3 H), 3.43 (m, 1 H), 3.02 (m, 1 H + 1 H overlapping resonances), 2.88 (m, 1 H + 1 H overlapping resonances), 2.60 (td, 1 H, J = 12.8 Hz, J = 2.6 Hz), 2.10 (m, 1 H), 2.01 (m, 1 H), 1.75 (m, 1 H), 1.66 (m, 2 H); ¹³C/APT (even +, odd -) NMR (75 MHz, CDCl₃) δ 172.07 (+), 161.30 (+, d, ¹Jcf = 243.6 Hz), 138.75 (+, d, ⁴Jcf = 3.2 Hz), 134.36 (-, d, J = 11.8 Hz), 129.00 (-, d, ³Jcf = 7.7 Hz), 126.45 (-, d, J = 16.6 Hz), 114.85 (-, d, ²Jcf = 20.9 Hz), 83.36 (+, d, ¹Jcf = 161.5 Hz), 62.47 (-), 61.46 (-), 55.09 (+, d, ⁴Jcf = 1.4 Hz), 52.99 (-), 51.21 (-), 34.41 (+), 33.86 (-), 26.24 (+), 26.07 (+); HRMS (APCI) [MH]⁺ Calcd for C₁₉H₂₄O₂NF₂: 336.1770, found: 336.1769; semi-preparative HPLC: t_R = 18.9 min (9 mL/min); analytical HPLC: t_R = 3.9 min (0.95 mL/min); Anal. Calcd for C₁₉H₂₃F₂NO₂: C, 68.04; H, 6.91; N, 4.18; found: C, 68.09; H, 6.91; N, 4.26.

A-((E)-4-[¹⁸F]fluorobut-2-en-1-yl)-2β-carbomethoxy-3β-(4′-fluorophenyl)nortropane ([¹⁸F]9)

Nor-β-CFT²⁵ (∼1.9 mg) was dissolved in DMF (0.3 mL) and added to the V-tube containing [¹⁸F]14. The mixture was heated at 105 °C for 15 min and then cooled in a 0 °C ice bath for 1 min. The mixture was diluted with HPLC solvent (∼0.5 mL) and purified by semi-preparative HPLC (9.1 mL/min; t_R = 17-21 min (range)). The desired fractions were combined, diluted 1:1.5 v/v with H₂O, and collected on a Waters C₁₈ Sep-Pak. The Sep-Pak was rinsed with 0.9% saline (35 mL) and then EtOH (0.5 mL). The product was eluted from the Sep-Pak with EtOH (1.5 mL) and collected in a sealed sterile vial containing 0.9% NaCl_(aq) (3.5 mL). This solution was passed successively through a 1 μm filter and then a 0.2 μm filter (Acrodisc PTFE) under Ar-pressure and collected in a sealed sterile dose vial containing 0.9% NaCl_(aq) (10 mL). The total synthesis time was ∼1 h from addition of the nortropane to [¹⁸F]14 with a 24% radiochemical yield (decay corrected). The product was then analyzed by analytical HPLC (t_r = 3.8 min, 1 mL/min) to determine the radiochemical purity (99%) and specific activity (SA = 851 mCi/μmol).

Compounds 10-13 were prepared in a similar manner as compound 9. Compounds [¹⁸F]10 (SA = 3859 mCi/μmol), [¹⁸F]11 (SA = 1440 mCi/μmol), and [¹⁸F]13 (SA: not determined) were prepared in a similar manner as [¹⁸F]9.

N-((E)-4-fluorobut-2-en-1-yl)-2β-carbomethoxy-3β-(4′-chlorophenyl)nortropane (10)

White needle crystals (22 mg, 21%): ¹H NMR (600 MHz, CDCl₃) δ 7.23 (d, 2 H, J = 9.0 Hz), 7.19 (d, 2 H, J = 9.0 Hz), 5.78 (m, 2 H), 4.83 (dd, 2 H, ²J_HF = 47.4 Hz, J = 5.4 Hz), 3.67 (partially resolved dd, 1 H, J = 6.9 Hz, J = 2.7 Hz), 3.50 (s, 3 H), 3.42 (m, 1 H), 3.01 (m, 1 H + 1 H, overlapping resonances), 2.88 (m, 1 H + 1 H, overlapping resonances), 2.58 (td, 1 H, J = 12.6 Hz, J =3.0 Hz), 2.10 (m, 1 H), 2.01 (m, 1 H), 1.74 (m, 1 H), 1.66 (m, 2 H); ¹³C NMR (75 MHz, CDCl₃) δ 172.01, 141.74, 134.33 (d, J = 11.5 Hz), 131.71, 128.95, 128.23, 126.48 (d, J = 16.6 Hz), 83.36 (d, J = 161.7 Hz), 62.49, 61.39, 55.09 (d, J = 1.4 Hz), 52.86, 51.26, 34.20, 33.97, 26.24, 26.06; HRMS (APCI) [MH]⁺ Calcd for C₁₉H₂₄O₂N³⁵ ClF: 352.1474, found: 352.1473; Anal. Calcd for C₁₉H₂₃ClFNO₂: C, 64.86; H 6.59; N, 3.98; found: C, 64.81; H, 6.63; N, 3.97; semi-preparative HPLC: t_R = 28.5 min (9 mL/min).

N-((E)-4-fluorobut-2-en-1-yl)-2β-carbomethoxy-3β-(4′-bromophenyl)nortropane (11)

White needle crystals (20 mg, 19%): ¹H NMR (300 MHz, CDCl₃) δ 7.38 (d, 2 H, J = 8.4 Hz), 7.14 (d, 2 H, J = 8.4 Hz), 5.78 (m, 2 H), 4.83 (partially resolved dd, 2 H, ²J_HF = 47.3 Hz, J = 4.8 Hz), 3.67 (m, 1 H), 3.50 (s, 3 H), 3.42 (m, 1 H), 2.96 (m, 4 H, overlapping resonances), 2.58 (td, 1 H, J = 12.5 Hz, J = 2.8 Hz), 2.06 (m, 2 H), 1.69 (m, 3 H); ¹³C NMR (75 MHz, CDCl₃) δ 171.99, 142.29, 134.33 (d, J = 11.8 Hz), 131.17, 129.38, 126.48 (d, J = 17.2 Hz), 119.85, 83.36 (d, J = 161.7 Hz), 62.49, 61.38, 55.10 (d, J = 1.4 Hz), 51.28, 34.14, 34.05, 26.24, 26.06; HRMS (APCI) [MH]⁺ Calcd for C₁₉H₂₄O₂N⁷⁹BrF: 396.0969, found: 396.0969; Anal. Calcd for C₁₉H₂₃BrFNO₂: C, 57.58; H, 5.85; N, 3.53; found: C, 58.41; H, 5.88; N, 3.62; semi-preparative HPLC: t_R = 33.0 min (9 mL/min).

N-((E)-4-fluorobut-2-en-1-yl)-2β-carbomethoxy-3β-(4′-iodophenyl)nortropane (12)

White solid (117 mg, 69%): ¹H NMR (300 MHz, CDCl₃) δ 7.58 (d, 2 H, J = 8.4 Hz), 7.02 (d, 2 H, J = 8.4 Hz), 5.78 (m, 2 H), 4.83 (partially resolved dd, 2 H, ²J_HF = 47.1 Hz, J = 4.8 Hz), 3.67 (m, 1 H), 3.50 (s, 3 H), 3.42 (m, 1 H), 2.95 (m, 4 H, overlapping resonances), 2.57 (td, 1 H, J = 12.4 Hz, J = 2.8 Hz), 2.05 (m, 2 H), 1.69 (m, 3 H); ¹³C NMR (75 MHz, CDCl₃) δ 171.97, 143.01, 137.11, 134.31 (d, J = 11.8 Hz), 129.72, 126.47 (d, J = 16.6 Hz), 91.34, 83.34 (d, J = 161.7 Hz), 62.46, 61.35, 55.07 (d, J = 1.4 Hz), 52.74, 51.27, 34.10, 34.01, 26.24, 26.03; HRMS (APCI) [MH]⁺ Calcd for C₁₉H₂₄O₂NF¹²⁷I: 444.0830, found: 444.0830; Anal. Calcd for C₁₉H₂₃FINO₂: C, 51.48; H, 5.23; N, 3.16; found: C, 52.07; H, 5.24; N, 3.23; semi-preparative HPLC: t_R = 46.6 min (9 mL/min).

N-((E)-4-fluorobut-2-en-1-yl)-2β-carbomethoxy-3β-(4′-methylphenyl)nortropane (13)

Colorless syrup (would not crystallize from hexanes) (23 mg, 53%): ¹H NMR (300 MHz, CDCl₃) δ 7.16 (d, 2 H, J = 8.1 Hz), 7.08 (d, 2 H, J = 8.1 Hz), 5.79 (m, 2 H), 4.83 (partially resolved dd, 2 H, ²J_HF = 47.3 Hz, J = 4.8 Hz), 3.66 (m, 1 H), 3.49 (s, 3 H), 3.42 (m, 1 H), 3.01 (m, 1 H + 1 H, overlapping resonances), 2.88 (m, 1 H + 1 H, overlapping resonances), 2.61 (td, 1 H, J = 12.6 Hz, J = 3.0 Hz), 2.29 (s, 3 H), 2.04 (m, 2 H), 1.70 (m, 3 H); HRMS (APCI) [MH]⁺ Calcd for C₂₀H₂₇O₂NF: 332.2020, found: 332.2020; Anal. Calcd for C₂₀H₂₆FNO₂: C, 72.48; H, 7.91; N, 4.23; found: C, 72.40; H, 7.86; N, 4.25.

(E)-4-Fluoro-1-tosyloxy-2-butene (14)

(E)-1,4-Ditosyloxy-2-butene (18) (256 mg, 0.65 mmol), Bu₄NF (1 M in THF, 0.7 mL), and THF (10 mL) were stirred at reflux under Ar_(g) for 30 min. The solvent was removed and the residue was purified by flash column chromatography on silica (2:1 v/v hexanes/EtOEt) to afford a colorless oil (60 mg, 38 %): ¹H NMR (300 MHz, CDCl₃) δ 7.80 (d, 2 H, J = 8.4 Hz), 7.36 (d, 2 H, J = 8.4 Hz), 5.99-5.75 (m, 2 H), 4.84 (ddd, 2 H, ²J_FH = 46.5 Hz, J = 4.8 Hz, J = 1.2 Hz), 4.57 (m, 2 H), 2.46 (s, 3 H).

(E)-4-[¹⁸F]Fluoro-1-tosyloxy-2-butene ([¹⁸F]14)

H¹⁸F was produced with a Siemens 11-MeV RDS 112 cyclotron by employing the ¹⁸O(p,n)¹⁸F reaction in H₂¹⁸O. The H¹⁸F_(aq) was transferred to a chemical processing control unit (CPCU), collected on a trap/release cartridge, released with K₂CO_3(aq) (0.9 mg in 0.6 mL H₂O), and combined with a CH₃CN solution of Kryptofix-222 (5 mg in 1 mL) in a V-tube. The V-tube was placed in a 110 ^oC oil bath, the solvent was evaporated under a N_2(g) flow, and CH₃CN (3 mL) was added and evaporated in order to azeotropically dry the Kryptofix-222/K¹⁸F complex. (E)-1,4-ditosyloxy-2-butene (18) (4 mg in 1 mL CH₃CN) was added, the reaction mixture was heated at 90 °C for 10 min, and [¹⁸F]14 was trapped on a Waters silica Sep-Pak Classic (WAT051900) (previously prepped with 10 mL EtOEt). Compound [¹⁸F]14 was eluted with EtOEt, the EtOEt solution was transferred to a hot cell under N_2(g) pressure and collected in a V-tube. The V-tube was placed in an 80 °C oil bath and the EtOEt was evaporated with an Ar_(g) flow. The solution of radiolabeling precursor was then added to this V-tube.

(E)-1,4-Diacetoxy-2-butene (16)

Trans-1,4-dibromo-2-butene (15) (1.98 g, 9.26 mmol), KOAc (4.70 g, 47.89 mmol, 5.2 equiv.), and AcOH (25 mL) were stirred at reflux under Ar_(g) for 21 h. The mixture was cooled to room temperature, filtered, the precipitate was rinsed with toluene, and the AcOH was removed from the filtrate azeotropically with toluene to give a wet white solid that was dried under vacuum for ∼ 10 min. The solid was suspended in CH₂Cl₂ and purified by vacuum flash chromatography on silica (13 cm h × 4 cm i.d.): hexanes (100 mL), v/v hexanes/CH₂Cl₂ – 3:1 (200 mL), 1:1 (200 mL), 1:3 (200 mL), CH₂Cl₂ (900 mL), to afford a colorless oil (1.04 g, 65 %): TLC R_f = 0.4 (silica, CH₂Cl₂, I₂ vapor); ¹H NMR (300 MHz, CDCl₃) δ 5.86 (septet, 2 H, J = 1.4 Hz), 4.58 (dd, 4 H, J = 1.4 Hz), 2.08 (s, 6 H); ¹³C NMR (75 MHz, CDCl₃) δ 170.86, 128.23, 64.07, 21.08.

(E)-1,4-Dihydroxy-2-butene (17)

(E)-1,4-diacetoxy-2-butene (16) (2.08 g, 12.08 mmol), HCl (2.0 M EtOEt, 0.9 mL, 1.8 mmol, 0.15 equiv.), and EtOH (75 mL) were stirred at reflux under Ar_(g) for 16 h, cooled, and the solvent was removed to afford a faint yellow oil (1.04 g, 98 %): ¹H NMR (300 MHz, CDCl₃) δ 5.90 (m, 2 H), 4.18 (m, 4 H).

(E)-1,4-Ditosyloxy-2-butene (18)

(E)-1,4-Dihydroxy-2-butene (17) (326 mg, 3.70 mmol), p-toluenesulfonyl chloride (1.76 g, 9.23 mmol, 2.5 equiv.), and THF (30 mL) were combined and cooled to 0 °C under Ar_(g) followed by addition of sodium t-butoxide (1.07 g, 11.13 mmol, 3.0 equiv.) in portions. The reaction mixture was stirred at room temperature overnight followed by addition of H₂O (100 mL) and then extraction with CH₂Cl₂ (25 mL × 3). The combined CH₂Cl₂ extracts were dried over MgSO₄ and the solvent was removed. The crude product was purified by flash column chromatography on silica (3:1:1 v/v/v hexanes/EtOEt/CH₂Cl₂) to afford white crystals (1.21 g, 82 %): ¹H NMR (300 MHz, CDCl₃) δ 7.77 (d, 4 H, J = 8.4 Hz), 7.35 (d, 4 H, J = 8.4 Hz), 5.74 (m, 2 H), 4.48 (m, 4 H), 2.46 (s, 6 H).

Figure 5.

MicroPET baseline TACs obtained by injection of [¹⁸F]11 into an anesthetized cynomolgus monkey.