Abstract

In our search for potential cocaine-abuse therapeutic agents, novel derivatives of 1-[2-[bis(4-fluorophenyl)methoxy]ethyl]-4-(3-phenylpropyl)piperazine (GBR 12909,1) and 1-[2-(diphenylmethoxy)ethyl]-4-(3-phenylpropyl)piperazine (GBR 12935,2) with various substituents in positions C2 and C3 of the phenylpropyl side chain were designed, synthesized and evaluated for their ability to bind to the dopamine transporter (DAT) and the serotonin transporter (SERT). The affinities of these ligands for DAT and SERT were controlled by the position of the substituents and their electronic and steric effects. This information could help us better understand and predict the influence of structural modification, and to discover new high-affinity DAT ligands possessing a range of affinities for SERT. The C2 series showed a quite different structure-activity relationship from the C3 series. In the C2 series, the substituent in theS configuration with a lone-pair of electrons significantly enhanced the affinity for DAT, whereas the steric effect of the substituent was detrimental to DAT binding affinity. In the C3 series, neither the lone electron pair nor the steric effect of the substituent seemed to affect DAT binding affinity, whilesp² hybridized substituents had a detrimental effect on affinity for DAT. 2-Fluoro-substituted analogs (S)-10 and (R)-10 displayed substantial enantioselectivity in transporter affinity. In the series, (S)-10 had the highest DAT binding affinity and good DAT selectivity. The 2-amino-substituted (R)-8 showed the highest SERT binding affinity in this series and essentially the same affinity for DAT (K_i = 22 and 20 nM, respectively). Thus, (R)-8 might be a lead for the discovery of a dual-action transporter ligand able to simultaneously block DAT and SERT. Compounds16 and18, the oxygenated derivatives of2, possessed the best selectivity for DAT (SERT/DAT ratio = 254 and 250, respectively).

Graphical Abstract

Introduction

Cocaine has caused many public health and societal problems^1,2 because of its extensive abuse in many countries throughout the world, exacerbating the spread of HIV-1^3,4 and hepatitis B and C, as well as drug-resistant tuberculosis.⁵ Cocaine is known to block the reuptake of several neurotransmitters in the central nervous system, such as dopamine (DA), serotonin (5-HT), and norepinephrine (NE), by interacting with the dopamine transporter (DAT), serotonin transporter (SERT), and norepinephrine transporter (NET) sites, respectively. Abundant studies have indicated that the reinforcing, locomotor-stimulating, and dependence-producing properties of cocaine are mainly mediated by the binding of cocaine to the DAT and subsequent blockade of DA reuptake into presynaptic terminals resulting in increased neurotransmission in the mesolimbic dopaminergic system.^6–12 Therefore, DAT has been considered to be the main target to reduce the effects of cocaine and DAT ligands with high potency and selectivity have been sought as potential therapeutic agents for cocaine abuse.¹³

Recently, SERT and NET have also been thought to play prominent roles in cocaine addiction.^14–24 Cocaine binds to DAT and SERT with moderate affinity (K_i = 341 and 129 nM, respectively),¹⁵ and blocks the reuptake of [³H]DA and [³H]5-HT (IC₅₀ = 478 and 304 nM, respectively).^12,18 In behavioral intervention and knock-out mice experiments, the serotonergic system has been found to influence the reinforcing effects of cocaine.^14,25–27 In contrast to the extensive structure-activity relationship (SAR) studies for DAT-selective ligands, fewer SAR studies have been focused on the discovery and development of ligands with a variety of transporter selectivity for both DAT and SERT.^14,28–31 Novel ligands that differ in their transporter affinity and SERT:DAT ratio may help reveal the pharmacological mechanisms relevant to stimulant abuse, and also lead to an high efficacy medication for cocaine abuse, with few adverse reactions. Such a medication may also reduce the incidence of HIV infection.³²

Our research has focused on the discovery of novel analogs of 1-[2-[bis(4-fluorophenyl)methoxy]ethyl]-4-(3-phenylpropyl)piperazine (GBR 12909,1) and 1-[2-(diphenylmethoxy)ethyl]-4-(3-phenylpropyl)piperazine (GBR12935,2) (Chart 1).³³ In an effort to develop long-acting cocaine abuse therapeutic agents, Lewis et al. prepared a series of oxygenated analogs of1 and2 and found that a racemic 3-hydroxyl-3-phenyl analog ((±)-3) of1 exhibited a pharmacological profile similar to that of1.³⁴ Later, the decanoate ester of (±)-3, prodrug4, was synthesized and formulated as a depot preparation. One dose of4 successfully suppressed cocaine-maintained responding in rhesus monkeys without affecting food intake for nearly thirty days.³⁵ Our continued studies on the SAR of the hydroxyl-containing analogs of1 and2 have shown that theS andR enantiomers of the 2-hydroxylated derivatives (e.g. (S)-5 and (R)-5) possessed quite different pharmacological profiles,^36,37 whereas the 3-hydroxylated enantiomers (e.g. (S)-3 and (R)-3) displayed very similar biological activities.³⁴ In the 2-hydroxyl series, theS isomers demonstrated substantially higher affinity for the DAT and were more potent DA reuptake inhibitors than the correspondingR isomers, whereas theR isomers showed higher affinity for the SERT and were more potent in inhibiting the reuptake of 5-HT than theS isomers.^36,37 The alcohols (S)-5 and (S)-6 are among the most potent and selective DAT ligands yet known (Table 1).

Based on these observations that the pharmacological profiles (including DAT and SERT binding affinities, transporter selectivity, and enantioselectivity) could be dramatically changed by minor structural modifications at the 2- and 3-positions of the 3-phenylpropyl side chain of1 and2, we undertook the design and synthesis of a new series of 2- and 3-substituted analogs of1 and2 in order to more definitively determine the SAR of the 2- and 3-substituted-3-phenylpropyl derivatives of1 and2. This information could help us identify the steric and electronic effects of the 2- and 3-substitutions in analogs of1 and2 involved in the binding interaction with DAT and SERT, and might lead to the discovery of novel high-affinity DAT ligands possessing a range of affinities for SERT.

Herein, we report the synthesis and SAR of a novel series of 2-substituted-3-phenylpropyl derivatives (7–10 and19–22) (Chart 2), 3-substituted-3-phenylpropyl analogs (13–16), and 2-substituted-2-phenethyl derivatives (23 and24) of1 and2 with amino, fluoro, hydroxyl, methoxyl, methyl, methylene, and oxo substituents at the 2- or 3-position of the 3-phenylpropyl moiety. Compounds7–10 were prepared and pharmacologically evaluated in optically pure (> 98% ee) forms. The roles of the 2- and 3-substituents in these new transporter ligands on their affinity and selectivity for DAT and SERT are discussed.

Chemistry

The monosubstituted piperazines, 1-[2-[bis(4-fluorophenyl)methoxy]ethyl]piperazine (25) and 1-[2-(diphenylmethoxy)ethyl]piperazine (26) (Chart 2) were synthesized in two steps according to the literature,³³ with modification.³⁶ The 2-methyl-, 2-amino-, and 2-methoxy-substituted analogs of1 were prepared via similar reaction sequences, in which25 was coupled with an appropriate acid usingN-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDCI) followed by reduction with a suitable aluminum hydride (Scheme 1).

Coupling of 2-methyl-3-phenylpropionic acid and25 provided amide27. (±)-7 was synthesized by reduction of27 with LiAlH₄. Optically pure (+)-7 and (−)-7 were prepared from (±)-7 using semi-preparative chiral high-performance liquid chromatography (HPLC) followed by recrystallization of their maleate salts.

Enantiomeric-pureN-tert-butyloxycarbonyl (Boc)-protected (S)-29 and (R)-29 were prepared starting from commercially availableN-Boc-protectedL- andD-phenylalanines, respectively. EDCI coupling followed by reduction with diisobutylaluminum hydride (DIBALH) yielded (S)-29 and (R)-29 retaining high ee values. Removing the Boc protective groups of (S)-29 and (R)-29 in the presence of hydrogen chloride afforded 2-amino-substituted (S)-8 and (R)-8, respectively.

The optically pure synthons (S)-2-methoxy-3-phenylpropanoic acid ((S)-30) and (R)-2-methoxy-3-phenylpropanoic acid ((R)-30) for the preparation of (S)-31 and (R)-31 were synthesized from L- and D-3-phenyllactic acids, respectively by selective methylation of the aliphatic hydroxyl groups using an excess amount of NaH in tetrahydrofuran (THF). Alane reduction of (S)-31 and (R)-31 gave (S)-9 and (R)-9, respectively.

Treatment of enantiomerically pure 2-hydroxyl-substituted (S)-5 and (R)-5³⁶ with (dimethylamino)sulfur trifluoride (DAST) produced desired 2-fluoro-substituted (R)-10 and (S)-10 and rearranged products (R)-32 and (S)-32, respectively in high ee (>98%) and in a ratio of 5:3 (Scheme 2). The possible mechanism for this unexpected result is shown inChart 3. Reaction of (S)-5 with DAST yielded intermediate (S)-I and hydrogen fluoride (HF). The basic nitrogen atom could be protonated by HF to form (S)-II, which was in equilibrium with (S)-I and could be attacked by fluoride ion to provide (R)-10. The aziridinium intermediate (S)-III was formed by anchimeric assistance of the unprotonated nitrogen atom of (S)-I. Nucleophilic substitution of fluoride ion at the less hindered position of (S)-III yielded (R)-32. If the fluoride ion was able to overcome the steric effect, and attacked the more hindered position of (S)-III, (S)-10 might be formed and would significantly reduce the ee value of (R)-10. Since both the isolated (R)-10 and (R)-32 showed very high optical purity (ee >98%), the nucleophilic substitution of (S)-III by fluoride ion was highly regioselective. The ee values of chiral compounds7–10 and32 were determined by analytical chiral HPLC, and were above 98%.

3-Methylene and 3-methyl-3-hydroxyl derivatives of1 and2 were synthesized from 3-oxo-substituted11 and12³⁴ (Scheme 3). Accelerated Wittig-type olefination of11 and12 with the CH₂I₂-TiCl₄-Zn-PbCl₂ system³⁸ provided13 and14 in moderate yields. Tertiary alcohols15 and16 were prepared via Grignard reaction of11 and12, respectively with methylmagnesium bromide.

Treatment of5 and6³⁶ under Swern oxidation conditions afforded 2-oxo-substituted17 and18 as depicted inScheme 4. Initial attempts to synthesize the 2-methylene analogs19 and20 and the 2-methyl-2-hydroxyl derivatives21 and22, following the same reaction conditions for the preparation of the corresponding 3-substituted analogs, failed. We recognized that the 3-positions of ketones17 and18 were located between the phenyl and carbonyl functionalities and therefore, the increased acidity of these positions was detrimental to both the Wittig-type olefination and Grignard reaction. Thus, alternative routes were sought in which the 2-methylene and 2-methyl-2-hydroxyl substitutions were introduced in earlier stages.

Coupling of 2-benzyl-acrylic acid^39,40 with25 and26 yielded amides33 and34, respectively (Scheme 5). The 2-methylene analogs19 and20 were then produced by reduction of33 and34 with alane.N-Alkylation of25 and26 with 1-chloro-2-methyl-3-phenyl-2-propanol⁴¹ using K₂CO₃ and NaI³⁴ afforded the desired tertiary alcohols21 and22, respectively. Acetophenones23 and24 were prepared in high yields by heating 2-bromoacetophenone with25 and26, respectively in the presence of diisopropylethylamine (DIPEA) inN,N-dimethylformamide (DMF).

Results and Discussion

Previous data suggested that the 2-substituted (S)-5 and (R)-5 demonstrated enantioselectivity in DAT and SERT binding, while the corresponding 3-substituted (S)-3 and (R)-3 showed no difference in transporter binding.^34,36 Therefore, the binding affinities of7–10, the analogs of5, for DAT and SERT were determined to investigate the steric and electronic effects of structural modifications in the 2-position of the phenylpropyl side chain of1.

The 2-methyl-substituted analogs (+)-7 and (−)-7 had less affinity than1, (S)-5, and (R)-5 in DAT binding, while showing similar SERT affinity as1 and (R)-5 (Table 1). Curiously, there was no meaningful difference in transporter binding affinities between the two enantiomers (+)-7 and (−)-7. Furthermore, the pharmacological profiles of (+)-7 and (−)-7 were quite similar to that of 2-hydroxyl-substituted (R)-5. To determine whether the 2-hydroxyl group of (S)-5 played an essential role in affinity and selectivity to DAT, the 2-amino-substituted analogs (S)-8 and (R)-8 and 2-methoxy-substituted analogs (S)-9 and (R)-9 were studied. The DAT binding affinities of (S)-8, (S)-9, (R)-8, and (R)-9 were in the range of 20–39 nM which were lower than those of1, (S)-5 and (R)-5. Interestingly, the 2-amino-substitution significantly enhanced the binding affinity for SERT. (R)-8 was the most potent SERT ligand in the series and showed approximately the same affinity for DAT and SERT (K_i = 20 and 22 nM, respectively). Hence, (R)-8 may be a potential lead for the development of dual-action transporter ligands.

Methylation of the 2-hydroxyl group not only decreased DAT affinity, but also abolished the enantioselectivity of (S)-5 and (R)-5. The transporter binding patterns of 2-methoxy-substituted analogs (S)-9 and (R)-9 were very similar to those of 2-methyl-substituted analogs (S)-7 and (R)-7, which implied that the methyl group has an adverse steric effect in DAT binding. To study the role of the lone electron pairs of oxygen in the hydroxyl group of (S)-5 without the interference of steric effects, as in the case of9, 2-fluoro-substituted analogs (S)-10 and (R)-10, were examined. In comparison with1, (S)-10 showed increased DAT binding affinity and decreased SERT binding affinity and therefore, much higher DAT selectivity (SERT/DAT ratio = 109). (S)-10 was the most potent DAT ligand (K_i = 2.7 nM)) in this study. On the other hand, (R)-10 showed significantly lower DAT binding affinity and selectivity versus SERT than (S)-10. Thus, the 2-fluoro-substituted analogs (S)-10 and (R)-10 demonstrated the highest enantioselectivity among this novel series of 2-substituted analogs of1 and possessed similar pharmacological profiles as those of 2-hydroxyl-substituted analogs (S)-5 and (R)-5.

The diverse transporter binding affinities, SERT/DAT ratios, and enantioselectivity of these 2-substituted analogs of1 allowed us to identify steric and electronic effects of the 2-substituent on the binding interaction to transporters. In previous SAR studies,36, a conformational-constrained analog of1 (Chart 4), demonstrated higher DAT affinity than1.³⁴ Under physiological conditions, the N4-nitrogen atom in piperazine should be protonated. Therefore, the conformerC-1 with an extended zigzag phenylpropyl side chain might be the active conformation of1 for DAT binding (Chart 4). A similar conformationally-constrained analog of1,37, in which the potential high-affinity conformerC-37 was unlikely to be as stable asC-1 due to the additional steric repulsion, did demonstrate slightly lower DAT and SERT affinities than1.³²

The hydroxyl groups in (S)-5 and (R)-5 could act as a hydrogen bond donor or a hydrogen bond acceptor. The pharmacological similarity between the fluoro- and hydroxyl-substituted analogs led us to hypothesize that its action as a hydrogen bond acceptor played a critical role in the enantioselectivity. As shown inChart 5, when the oxygen lone pairs in (S)-5 and (R)-5 formed hydrogen bonds with the N4-ammonium groups, conformationally-constrained conformersC-S5 andC-R5 would be the favorable conformers of (S)-5 and (R)-5, respectively. As a result, the phenyl ring in (S)-5 and (R)-5 would be forced into opposite directions in space.C-S5 bound to DAT with higher affinity than didC-1 andC-R5, whereasC-S5 showed lower SERT binding affinity than didC-1 andC-R5. The position of the phenyl group inC-1 was between the likely positions of the phenyl groups inC-S5 andC-R5, and this could explain why the DAT affinity of1 was between those of (S)-5 and (R)-5. On the other hand,C-S3-1 could be the corresponding cyclic conformer of (S)-3. However, the six-membered ring conformation ofC-S3-1 might not be as stable as the five-membered ring conformation ofC-S5 and therefore, other conformers, such asC-S3-2, with an extended zigzag side chain similar toC-1 might be the major conformer of (S)-3. This hypothesis could explain why the transporter binding profiles of1, (S)-3, and (R)-3 were approximately the same. Thus, a proper configuration of substituents with an available lone-pair of electrons, such as 2-(S)-OH and 2-(S)-F, which would allow the DAT-binding favorable conformer to be stabilized by their electronic effects, was the most critical factor causing the observed increase in DAT affinity. However, these steric and electronic effects did not promote high SERT affinity. Therefore, (S)-5 and (S)-10 are among the most DAT selective ligands that have been found.

Initially, we thought the 2-amino group would be a strong electron-donating group and thus, may result in analogs with high DAT affinity and selectivity. In contrast to the 2-fluoro and 2-hydroxyl analogs, the 2-amino-substituted analogs (S)-8 and (R)-8 had only moderate affinity for DAT. Since the nitrogen lone-pair of electrons in the 2-amino groups of (S)-8 and (R)-8 were protonated under physiological conditions, there was no available electron lone-pair to be able to act as a hydrogen bond acceptor. Actually, the protonated 2-ammonium group would be more likely to act as a hydrogen bond donor. Furthermore, the potential electrostatic and steric repulsion between the two cationic N4- and 2-ammonium groups (Chart 5) could substantially destabilize thesyn-conformersC-S8-1 andC-R8-1 and therefore, force (S)-8 and (R)-8 to adopt other more stable conformations, such asC-S8-2 andC-R8-2, which were significantly different from our postulated conformers with high DAT affinity (e.g.C-1 andC-S5) in this series. Thus, the quite different conformation ofC-R8-2 might be responsible for the high affinity of (R)-8 for SERT. Interestingly, the phenyl groups inC-S5 andC-R8-2 were in the same spatial areas, while the phenyl groups in bothC-R5 andC-S8-2 were in different spatial areas. This observation could help explain why the (R)-8 had higher DAT binding affinity than (S)-8 and the 2-amino-substituted analogs (S)-8 and (R)-8 demonstrated enantioselectivity in both DAT and SERT binding affinities.

Since the 2-fluoro-substituent was not a hydrogen bond donor and 2-fluoro-substituted analogs possessed similar pharmacological profiles to 2-hydroxy-substituted analogs, the ability of the substituent to act as a hydrogen bond donor did not appear to be essential to enantioselectivity. Thus, the lack of enantioselectivity of the 2-methoxy-substituent (S)-9 and (R)-9 could not be ascribed to its inability to act as a hydrogen bond donor, but might be caused by the steric effect of the additional methyl group. Hence, the presence of a small substituent with an available electron lone-pair in a compound with the 2S configuration is likely to be promote DAT binding affinity. Future work should focus on the influence of steric and electronic effects on transporter binding affinity by the introduction of relatively less hinderedsp² hybridized oxo and methylene functionalities into1 and2.

Previously, Lewis et al. reported the synthesis and monoamine transporter affinities of 3-oxo-substituted11 and12.³⁴ In contrast to the unsubstituted1 and2, and 3-hydroxylated (S)-3 and (R)-3,11 and12 demonstrated much lower DAT binding affinity and selectivity. This might be explained by the tendency of the 3-carbonyl group to conjugate with the aromatic phenyl ring, restricting the conformational flexibility of the phenyl group that might be essential for interaction with DAT.

Therefore, the 3-methylene-substituted13 and14 and 3-methyl-3-hydroxyl-substituted15 and16 were prepared to explore the “conjugation hypothesis”. In terms of interaction with DAT,13 and14 had higher affinity than11 and12, but less affinity than1 and2. Although thesp² hybridized 3-methylene group in13 and14 might also be able to conjugate with the phenyl group, the steric interaction between the hydrogen atoms at the vinyl position andortho position of the phenyl ring would tend to reduce the activation energy for rotation of the phenyl group (Chart 6). The resultant weaker conjugation resulted in higher DAT binding affinity. The 3-methyl-3-hydroxyl-substituted analogs,15 and16, with quaternary carbon atoms at C3, had almost the same binding affinity as1 and2 for DAT, and demonstrated greater DAT:SERT selectivity than1 and2. These data indicate that asp³ hybridized 3-position is important for high DAT binding affinity, and the DAT and SERT binding affinities of1 and2 are not affected very much by the substituent at the 3-position.

In contrast to the 3-oxo-substituted compounds11 and12, which had weak affinity and low selectivity for DAT, the 2-oxo-substituted17 and18 showed high DAT binding affinity, and demonstrated higher SERT:DAT ratios than1 and2. The DAT affinity and selectivity of17 and18 were higher than those of (R)-5 and (R)-6 and lower than those of (S)-5 and (S)-6. Based on our hypothesis, the potential hydrogen bonding between the carbonyl lone-pair of electrons and ammonium groups of17 and18 would favor the formation of conformerC-17, which was very similar toC-1 but likely to be more stable (Chart 6). As described earlier,C-1-like conformers were poorer SERT ligands and in agreement with the hypothesis, the SERT binding affinities for17 and18 were lower than those of1 and2.

The 2-methylene-substituted analogs19 and20, were then synthesized to explore the electronic effect of the 2-oxo-substitution and the steric effect of the 2-methyl group. Surprisingly, the DAT binding affinities of19 and20 were significantly reduced (27- and 110-fold, respectively) as compared to those of1 and2. Initially, we thoughtC-19-1, which was one of the possible conformers of19 and20 should possess less steric hindrance thanC-S7 (i.e. a conformer of (S)-7) and therefore, have higher affinity for DAT than the 2-methyl-substituted7 (Chart 6). However, rotation of the C=C bond to form conformerC-19-1 is likely to be energetically unfavorable and thus,C-19-1 might not exist in a significant amount at room temperature. On the other hand, conformerC-19-2 should be the more stable conformer of19 and20. In the case ofC-19-2, steric repulsion might occur between the vinyl hydrogen atoms and the proton of the protonated N4-ammonium group. Moreover, the vinyl hydrogen atoms could also interfere with the free rotation of the phenyl group by interaction with the hydrogen atoms at theortho positions of the phenyl ring. Therefore, high DAT affinity conformations, such asC-1 andC-S5, could not readily be adopted by19 and20. In the 3-phenylpropyl-substituted series,19 and20 had the lowest affinity for DAT in the bis(4-fluorophenyl)- and diphenyl-substituted analogs, respectively.

Our hypothesis was supported by the data obtained from the 2-methyl-2-hydroxyl-substituted compounds21 and22. Compared to the 2-methylene-substituted analogs19 and20, compounds21 and22, with crowded tertiary 2-positions, possessed higher binding affinity for both DAT and SERT. Presumably, the postulated unfavorable steric repulsion inC-19-2 was relieved. Furthermore,21 demonstrated significantly weaker DAT binding affinity and slightly lower SERT binding affinity than 2-hydroxylated (S)-5 and (R)-5 and 2-methylated (S)-7 and (R)-7. This indicates that the steric effect produced by additional substitution at the 2-position is more unfavorable to DAT binding affinity than to SERT binding affinity. However, the binding data of15,16,21, and22 were determined using racemic samples. The possibility that the two enantiomers of these ligands possessed substantially different transporter binding affinities could not be ruled out.

The tremendous differences in DAT binding affinity for 2-oxo-substituted17 and18 and 3-oxo-substituted11 and12 led us to synthesize acetophenones23 and24 in which the 2-oxo-substituents theoretically can conjugate with the phenyl group.23 and24 were similar to11 and12 in their weak DAT binding affinities, while the unsubstituted acetophenone38³⁶ exhibited high binding affinity for DAT (Chart 2). These data further supported the hypothesis that conjugation with the phenyl ring reduces binding affinity for DAT. On the other hand, the 3-oxo-substituted11 and12, 2-oxo-substituted17 and18, and acetophenone23, all demonstrated a similar reduction (1.7~4.0 fold) in SERT binding affinity as compared to their corresponding unsubstituted analogs.

Interestingly, in the preparation of chiral 2-fluoro-substituted analogs (S)-10 and (R)-10, the rearranged products (S)-32 and (R)-32, which could be viewed as 1-fluoromethyl-substituted analogs of38, were formed enantioselectively in significant amounts. In contrast to (S)-10 and (R)-10, in which the chirality was critical to their pharmacological activity, the 1-substituted (S)-32 and (R)-32 didn’t demonstrate meaningful enantioselectivity. Among the bis(4-fluorophenyl)-substituted analogs, (S)-32 and (R)-32 were the weakest DAT ligands. Nevertheless, (S)-32 possessed higher SERT binding affinity than (R)-32 and was the only SERT-selective ligand found in this study (SERT/DAT ratio = 0.8).

Conclusions

In this article, we report SAR studies focused on the 2- and 3-substituted analogs of1 and2. Binding affinities for the DAT and SERT varied depending on the nature of the substituents in the C2- and C3-positions of the phenylpropyl side chain. Steric and electronic effects affected their affinity and selectivity for DAT and SERT. The 2-substituted series showed a quite different SAR than the 3-substituted series. In the C2 series, the substituent in theS configuration with a lone-electron pair significantly enhanced ligand binding affinity for DAT, whereas the steric effect of the substituent was detrimental to DAT binding. In the C3 series, neither the lone-electron pair nor the steric effect of the substituent seemed to affect DAT binding affinity, while compounds withsp² hybridized substituents, possibly due to their lessened conformational mobility, had little affinity for DAT. 2-Oxo-substitution retained the DAT binding seen with their parent compound and slightly increased the SERT:DAT ratio, whereas 3-oxo-substitution was detrimental to the affinity and selectivity for DAT. The SERT binding affinity was insensitive to most of the structural modifications in this study, except for 2-amino-substitution that caused substantially increased SERT binding affinity and selectivity over the DAT.

Novel derivatives with different transporter binding affinities and selectivities were discovered. 2-Fluoro-substituted analogs (S)-10 and (R)-10 displayed substantial enantioselectivity in their binding affinity for the DAT and SERT. (S)-10 had higher affinity and selectivity for DAT than (R)-10. In the series, (S)-10 had the highest DAT binding affinity and exhibited excellent DAT selectivity over the SERT. (R)-8 and (S)-32 demonstrated approximately equal affinities for DAT and SERT (SERT/DAT ratio = 1 and 0.8, respectively). Furthermore, (R)-8 demonstrated the highest SERT binding affinity in this series. Thus, the 2-amino-substituted (R)-8 might be a lead compound for the discovery of potent and dual-action ligands able to simultaneously block the DAT and the SERT. As in the previous SAR studies, diphenyl-substituted analogs always showed higher SERT/DAT ratios than the corresponding bis(4-fluorophenyl)-substituted analogs. In the series,16 and18 were both oxygenated derivatives of2 and possessed the highest selectivity for DAT (SERT/DAT ratio = 254 and 250, respectively). Compounds with customized DAT and SERT binding profiles could be useful research tools to help determine the DAT/SERT ratio needed for the blockade of DAT and SERT that would be more beneficial as a medication for stimulant abuse than the blockade of DAT alone.⁴² The affinities of these analogs for the DAT and SERT appeared to be controlled by their electronic and steric effects and by the position of the substituent. Identification of these effects should facilitate the discovery and development of novel ligands with desired monoamine transporter binding patterns, and these may prove to be more effective treatment agents for cocaine abuse with fewer adverse effects.

Experimental Section

Melting points were determined on a MEL-TEMP II apparatus by Laboratory Devices and are uncorrected. NMR spectra were recorded on Bruker DPX-200, Varian XL-300, and AMX-400 FT-NMR spectrometers. Chemical shifts are expressed in parts per million (ppm) on the δ scale relative to a tetramethylsilane (TMS) internal standard. Chemical ionization (CI) mass spectra were obtained using a Finnigan 1015 mass spectrometer. EI mass spectra and high-resolution mass measurements (HRMS) were obtained using a Finnigan MAT 95S mass spectrometer. Elemental analyses were performed with a Heraeus varioIII-NCSH instrument or by Atlantic Microlabs, Atlanta, GA, and were within ±0.4% for the elements indicated. Optical rotations were obtained using a Perkin-Elmer 341 polarimeter and are reported at the sodium D-line (589 nm), unless otherwise noted. Thin-layer chromatography (TLC) was performed on Merck (art. 5715) silica gel plates and visualized under UV light (254 nm), upon treatment with iodine vapor, or upon heating after treatment with 5% phosphomolybdic acid in ethanol. Flash column chromatography was performed with Merck (art. 9385) 40–63 μm silical gel 60. Anhydrous tetrahydrofuran was distilled from sodium-benzophenone prior to use. No attempt was made to optimize yields.

Acknowledgments

We thank Noel Whittaker and Wesley White (NIDDK, NIH) for CI mass spectral data and the expert technical assistance of Robert H. Horel and Mario Ayestas in helping with the transporter binding assays. We also acknowledge the National Science Council of the Republic of China (Grant No. NSC 95-2323-B-002-011. The research of the Drug Design and Synthesis Section, CBRB, NIDA & NIAAA, was supported by the NIH Intramural Research Programs of the National Institute of Diabetes and Digestive and Kidney Diseases, the National Institute on Drug Abuse (NIDA) and the National Institute of Alcohol Abuse and Alcoholism, and NIDA supported the research of the Clinical Psychopharmacology Section.

References