Medicinal Chemistry of σ₁ Receptor Ligands: Pharmacophore Models, Synthesis, Structure Affinity Relationships, and Pharmacological Applications

Abstract

In the first part of this chapter, we summarize the various pharmacophore models for σ₁ receptor ligands. Common to all of them is a basic amine flanked by two hydrophobic regions, representing the pharmacophoric elements. The development of computer-based models like the 3D homology model is described as well as the first crystal structure of the σ₁ receptor. The second part focuses on the synthesis and biological properties of different σ₁ receptor ligands, identified as 1-9. Monocyclic piperazines 1 and bicyclic piperazines 2 and 3 were developed as cytotoxic compounds, thus the IC₅₀ values of cell growth and survival inhibition studies are given for all derivatives. The mechanism of cell survival inhibition, induction of time-dependent apoptosis, of compound ent-2a is discussed. Experimentally determined σ₁ affinity shows good correlation with the results from molecular dynamics simulations based on a 3D homology model. Spirocyclic compounds 4 and 5 represent well-established σ₁ receptor ligands. The homologous fluoroalkyl derivatives 4 have favorable pharmacological properties for use as fluorinated PET tracers. The (S)-configured fluoroethyl substituted compound (S)-4b is under investigation as PET tracer for imaging of σ₁ receptors in the brain of patients affected by major depression. 1,3-Dioxanes 6c and 6d display a very potent σ₁ antagonist profile and the racemic 1,3-dioxane 6c has high anti-allodynic activity at low doses. The arylpropenylamines 7 are very potent σ₁ receptor ligands with high σ₁/σ₂ selectivity. The top compound 7g acts as an agonist as defined by its ability to potentiate neurite outgrowth at low concentrations. Among the morpholinoethoxypyrazoles 8, 8c (known as S1RA) reveals the most promising pharmacokinetic and physicochemical properties. Due to its good safety profile, 8c is currently being investigated in a phase II clinical trial for the treatment of neuropathic pain. The most potent ligand 9e of 3,4-dihydro-2(1H)-quinolones 9 shows promising anti-nociceptive activity in the formalin test.

1 Introduction

The sigma-1 (σ₁) receptor is a membrane-bound protein distributed in the central nervous system and in peripheral organs like heart, kidney, and liver (Weissman et al. 1988; Samovilova et al. 1988; Ela et al. 1994). The σ₁ receptor is mainly localized at the endoplasmic reticulum and the mitochondria-associated membranes. Ruoho et al. have shown that the σ₁ receptor consists of two transmembrane regions connected by a loop. Both C- and N-terminus are located extracellularly or in the ER lumen (Chu and Ruoho 2016). Two additional hydrophobic regions of the σ₁ receptor were identified by Fontanilla et al. named steroid binding domain-like regions (SBDL I and II). With the help of N-substituted photoaffinity labels it was shown that the SBDL I overlaps with one of the two transmembrane regions of the σ₁ receptor forming the ligand-binding domain together with the SBDL II (Ruoho et al. 2012). σ₁ Receptors were shown to take part in the regulation of ion channels (e.g., K⁺ and Ca²⁺) and in the modulation of neurotransmitter systems (Lupardus et al. 2000; Hong and Werling 2000; Hayashi and Su 2007). In the brain, the σ₁ receptor is particularly well expressed in areas associated with memory and emotion (Mash and Zabetian 1992). Steroids like progesterone (Su et al. 1988; Schwarz et al. 1989) and N,N-dimethyltryptamine (Fontanilla et al. 2009) were previously discussed to be endogenous ligands but their σ₁ receptor binding affinities are low compared with those of sphingosines showing high affinity in the low-nanomolar range (Ruoho et al. 2012). Since many centrally active drugs show high σ₁ affinity, σ₁ receptors represent promising targets for the research and development of drugs to treat several neurological or neuropsychiatric disorders like depression, psychosis, and cocaine abuse (Hascoet et al. 1995; Matsumoto et al. 2001; Sharkey et al. 1988; Bermack and Debonnel 2001; Ishikawa et al. 2007; Skuza and Rogoz 2006). The fact that many human cancer cell lines show up-regulated levels of σ₁ receptors brought them into focus for the development of new antitumor drugs and cancer diagnostics (Hashimoto and Ishiwata 2006). Based on various studies, Chen and Pasternak postulated that the σ₁ receptor functions as an endogenous anti-opioid receptor system (Chien and Pasternak 1993). By investigation of different σ₁ receptor ligands in animal models it was shown that σ₁ agonists inhibit morphine-induced analgesia whereas σ₁ antagonists potentiate opioid induced analgesia (Chien and Pasternak 1994, 1995). The fact that σ₁ knockout mice show reduced pain response in the formalin test but not hypersensitivity after treatment with capsaicin lead to interest in σ₁ receptors as a target in the treatment of neuropathic pain (Entrena et al. 2009).

For the development of new potent σ₁ receptor ligands with high affinity several pharmacophore models have been developed and optimized. Herein the pharmacophore models reported so far are summarized and compared with respect to existing ligands.

2 Pharmacophore Models

In 1994, Glennon et al. reported a two-dimensional pharmacophore model based on deconstruction–reconstruction analysis of different flexible σ₁ receptor ligands. In this model, two hydrophobic regions flanking a basic amine represent the pharmacophoric elements required for high σ₁ affinity. A distance of 6–10 Å between the amine moiety and the primary hydrophobic region and of 2.5–3.9 Å between the amino group and the secondary hydrophobic region provides optimal binding conditions. The amine could be of primary, secondary, or tertiary nature. In case of a tertiary amine, only small substituents are allowed, whereas the amine could also be part of a ring system (e.g., piperazine ring). The primary hydrophobic region tolerates sterically demanding residues whereas the secondary region favors smaller substituents like a three-carbon chain. As the two hydrophobic binding pockets of the σ₁ receptor tolerate bulky groups, the size of substituents can vary slightly without decreasing binding affinity (Fig. 1) (Glennon et al. 1994; Glennon 2005).

Fig. 1

Pharmacophore model of Glennon (2005)

Laggner et al. presented in 2005 the first computer-aided three-dimensional pharmacophore model (Fig. 2) based on 23 structurally very different σ₁ ligands. The pharmacophoric elements consist of a positive ionizable group like an amine and four hydrophobic features. The calculated distances between the pharmacophoric elements are in good agreement with the results obtained by Glennon et al. (1994).

Fig. 2

3D-σ₁-Pharmacophore model of Laggner et al. (2005) red: positive ionizable group; blue: hydrophobic regions

In 2009, Zampieri et al. designed another computer-based model containing five pharmacophoric elements (Fig. 3) (Zampieri et al. 2009). The model included three hydrophobic areas in total (depicted in blue and pink), whereby two of them should have aromatic character (blue). A basic center (red) is located at a distance of 7.01 and 8.50 Å from the aromatic moieties and at a distance of 3.58 Å from the further hydrophobic elements. These distances are comparable to those postulated in the models of Glennon and Laggner. Additionally, the Zampieri model established an H-bond acceptor function, which was already defined in a pharmacophore model by Gilligan et al. This model was published in the early 1990s and did not differentiate between σ₁ and σ₂ ligands (Gilligan et al. 1992).

Fig. 3

Pharmacophore model of Zampieri et al. (2009) red: basic center; green: H-bond acceptor; blue: aromatic hydrophobic area; pink: hydrophobic area

In 2011, Laurini et al. published the first computer-based 3-dimensional (3D) homology model of the σ₁ receptor. For the identification of a reliable ligand-binding domain, results of docking studies, mutagenesis studies, structure–affinity-relationship studies, and pharmacophore models were combined. The validation of the homology model was implemented by docking studies of well-known σ₁ ligands at the postulated binding site of the receptor, calculation of free binding energy, and comparison with the experimentally determined σ₁ affinities of these ligands (Laurini et al. 2011, 2012).

Schmidt et al. have just published a crystal structure of the σ₁ receptor for the first time (Schmidt et al. 2016). This structure was determined in complex with two different σ₁ receptor ligands, PD144418 and 4-IBP. Contrary to the early findings of Fontanilla and Ruoho (Ruoho et al. 2012) as well as Aydar et al. only one transmembrane domain of the σ₁ receptor was found in the crystal structure (Schmidt et al. 2016). This contradicts also the solution nuclear magnetic resonance (NMR) results of Ortega-Roldan et al. (2015) which closely match the findings of the 3D homology model of Laurini et al. (2011).

3 Ligands Introduction

In the literature, a great variety of σ₁ receptor ligand classes are reported. These classes include piperazines 1, bicyclic compounds of type 2 and 3, spirocyclic compounds 4 and 5, 1,3-dioxanes 6, arylalkenylamines 7, morpholinoethoxypyrazoles 8, and 3,4-dihydro-2(1H)-quinolones 9 (Fig. 4). The synthesis and the pharmacological properties of these ligands are presented herein.

Fig. 4

Representative σ₁ receptor ligands of different compound classes. Inhibition of σ₁ receptor radioligand binding at 1 μM concentration of test compound. PMB p-methoxybenzyl; Naph naphthyl

4 Homologous 2-Piperazinealkanols

Piperazines 1 with different hydroxyalkyl side chains represent well-established σ₁ receptor ligands. A broad structure affinity relationship study was performed based on 2-hydroxyethyl substituted piperazines with high σ₁ affinity and their larger and smaller homologs bearing hydroxypropyl or hydroxymethyl side chains. The cytotoxic activity against human cancer cell lines was tested by in vitro cell survival assays (Weber et al. 2014; Holl et al. 2012; Bedurftig and Wunsch 2004).

(S)-Serine, (S)-aspartic acid, and (S)-glutamic acid as enantiomerically pure amino acids of nature’s chiral pool were used for the synthesis of homologous piperazinealkanols 1a-g. The first reaction step includes the esterification of the particular amino acid. The dioxopiperazines 11 were prepared from the aminoesters 10 ^.HCl in a three-step reaction sequence consisting of reductive alkylation, chloroacetylation, and ring closure with different primary amines. Reduction with LiAlH₄ led to the piperazinealkanols 1a-g (Scheme 1). Reduction of the ester moiety of 11d (n = 1, R = CO₂CH₃) followed by Wittig reaction of the resulting aldehyde with methyl(triphenylphosphoranylidene)acetate and subsequent reduction of the α,β-unsaturated ester led to the homologous hydroxybutyl piperazine 1g (Weber 2012).

Scheme 1

Synthesis of homologous 2-piperazinealkanols 1. Reagents and reaction conditions: (a) (H₃C)₃SiCl, CH₃OH, room temperature (rt), 16 h; (b) (1) Ph-CH=O, NEt₃, CH₂Cl₂, rt, 16 h; (2) NaBH₄, CH₃OH, 0 °C, 40 min; (3) ClCH₂COCl, NEt₃, CH₂Cl₂, rt, 2.5 h; (4) R¹-NH₂, NEt₃, CH₃CN, rt, 16 h–3 d; (c) LiAlH₄, THF, reflux, 16 h. PMB p-methoxybenzyl, 2-NaphCH ₂ 2-naphthyl (Weber et al. 2014; Holl et al. 2012; Bedurftig and Wunsch 2004)

The σ₁ and σ₂ affinity of hydroxyalkyl piperazines 1a-g was tested with tissue membrane preparations of animal origin (guinea pig brain, rat liver). Selected ligands (1d-g) were also assayed with membrane preparations bearing human σ₁ receptors to evaluate ligand-binding affinity towards σ₁ receptors from different species (Table 1).

For a high σ₁ affinity, the length of hydroxyalkyl side chain and the size of the residues at both N-atoms are of particular importance. Short side chains like hydroxymethyl and hydroxyethyl are well-tolerated by the σ₁ receptor leading to K _i values in the range of 4–20 nM. The σ₁ affinity of the hydroxypropyl piperazines is more than tenfold lower (e.g., 1f, K _i = 275 nM). The extension of the side chain by another methylene moiety in case of hydroxybutyl piperazine 1f leads to an increased σ₁ affinity, but the K _i value of 52 nM remained higher than the K _i value measured for hydroxymethyl and hydroxyethyl derivatives. In accordance with the pharmacophore model of Glennon postulating two hydrophobic regions, the N-methyl substituted piperazine 1c does not show high σ₁ receptor affinity. The affinity increases by the introduction of a larger residue such as cyclohexylmethyl or p-methoxybenzyl group.

The hydroxyethyl derivatives show almost the same σ₁ affinity as the hydroxymethyl derivatives, but show reduced σ₂ affinity than hydroxyethyl piperazines. Regarding the σ₁/σ₂ selectivity, it becomes clear that the hydroxyethyl derivatives are more than tenfold selective for the σ₁ over the σ₂ receptor. The PMB-substituted compound 1b provides the best selectivity in this set of compounds with a σ₂ affinity of >1,000 nM. As a result of further chain extension, the σ₂ affinity increases leading to decreased selectivity (1f, K _i (σ₂) = 690 nM, 1g, K _i (σ₂) = 348 nM).

The σ₁ affinity of 1d-g measured with membrane preparations from a human cancer cell line (RPMI 8226) is slightly reduced compared to the affinity measured with the guinea pig brain membrane preparations. Because the same trend was found for the reference compounds haloperidol and (+)-pentazocine it can be assumed that the results of both assays are well comparable.

As the enantiomer of 1d prepared in the same manner from (R)-aspartic acid shows the same σ₁ affinity, it could be assumed that the stereochemistry has only negligible influence on σ₁ receptor affinity and selectivity over the σ₂ subtype (K _i(σ₁) = 1.9 nM, σ₁/σ₂ = 32).

The σ₁ affinity obtained with human receptor preparations was supported by docking of the ligands in the putative binding site of the 3D σ₁ receptor homology model. The calculated free binding energies are in good accordance with their recorded affinities towards the σ₁ receptor. For the most potent human σ₁ receptor ligand 1g (K _i,exp. = 6.8 nM) a ΔG_bind of −10.85 ± 0.36 kcal/mol was calculated which corresponds to an estimated K _i(σ₁)_calcd value of 11.2 nM, consistent with the experimentally determined K _i values (Weber et al. 2014).

The cell growth inhibition potential of piperazinealkanols 1c-g was tested in seven human tumor cell lines. The potent σ₁ receptor ligand 1e inhibited the growth and survival of the bladder cancer cell line 5637, the small cell lung cancer cell line A427, and the multiple myeloma cell line RPMI 8226 in the low micromolar range. Even at high concentrations (20 μM) of 1e, a growth inhibition activity could not be found for the bladder cancer cell line RT4, the large cell lung cancer LCLC-103H, and the pancreas cancer cell line DAN-G. Additionally only low activity was found for the breast cancer cell line MCF-7, indicating a selective mechanism of growth and survival inhibition. Further investigation of the mechanism associated with the inhibitory activity of 1e was performed with RPMI 8226 cells and revealed an increase in the amount of early apoptotic cells after 48 h compared to the untreated control.

Table 1 Inhibition of cancer cell growth by piperazinealkanols 1a-g compared with σ₁ and σ₂ binding affinity

5 Bicyclic Piperazines

In order to investigate the influence of conformational restriction on σ₁ receptor affinity bicyclic compounds of type 2 and 3 with diazabicycloalkan scaffold were designed by intramolecular connection of the 2-hydroxyalkyl side chain of piperazines 1 with C-5 of the piperazine ring. Propano- and butano-bridged homologs of 2 and 3 with diazabicyclo[3.2.2]nonane and diazabicyclo[4.2.2]decane scaffold were synthesized by an expansion of the ethano-bridge by one or two methylene moieties to elucidate the effect of bridge size on σ₁ receptor affinity (Geiger et al. 2007; Weber et al. 2016; Sunnam 2010).

Homologous dioxopiperazines 11 (n = 1–3) with different substitution pattern of the N-atoms were used as starting material for the synthesis of bicyclic compounds 2 and 3 (Scheme 2). The dioxopiperazine 11 (n = 3) was synthesized from racemic-2-aminoadipic acid in the same manner as explained for dioxopiperazines 11 (n = 1,2) in Scheme 1. The mixed methyl silyl ketals 12 were obtained by Dieckmann analogous cyclization of 11. The Dieckmann analogous cyclization gave only low yields for the dioxopiperazines 11 (n = 1) with acetate side chain due to the rigidity of the resulting products 12 (n = 0). The (R)-configuration of the ketalic center of 12 was shown by X-ray crystal structure analysis (Holl et al. 2008). Hydrolysis and reduction of 12 led to the bicyclic alcohols 2 and 3. The enantiomers ent-2 and ent-3 were obtained starting with (R)-configured amino acids.

Scheme 2

Synthesis of bicyclic diazabicycloalkanols 2 and 3. Reagents and reaction conditions: (a) n = 1: NaHMDS, THF, −78 °C, 40 min, then (H₃C)₃SiCl, −78 °C, 1 h, then rt, 2 h (Weber et al. 2016); n = 2: LiHMDS, THF, −78 °C, 30 min, then (H₃C)₃SiCl, −78 °C, 0.5 h, then rt, 3 h (Geiger et al. 2007); n = 3: LiHMDS, THF, −78 °C, 30 min, then (H₃C)₃SiCl, −78 °C, 2 h, then rt, 0.5 h (Sunnam 2010); (b) n = 0: 1. 0.5 M HCl, THF, rt, 16 h; 2. LiAlH₄, THF, reflux, 16 h (Weber et al. 2016); n = 1: 1. p-TosOH, THF, H₂O, rt, 16 h; 2. LiBH₄, THF, −90 °C, 3.5 h; 3. LiAlH₄, THF, reflux, 16 h (Geiger et al. 2007) n = 2: 1. p-TosOH, THF, H₂O, rt, 16 h; 2. LiBH₄, THF, −90 °C, 2.5 h; 3. LiAlH₄, THF, reflux, 15 h (Sunnam 2010)

In piperazinealkanols 1 the hydroxyalkyl side chain can adopt several conformations and, moreover, the piperazine ring can adopt two conformations, leading to an axial or equatorial orientation of the side chain resulting in different distances between the pharmacophoric elements. In the bicyclic alcohols 2 and 3 the additional bridge over the piperazine ring reduces the flexibility of the ring system and its hydroxyalkyl side chain. As a result of conformational restriction, the pharmacophoric elements are fixed in a defined arrangement minimizing the loss of entropy during binding and thus increasing the overall free binding energy.

The σ₁ and σ₂ receptor affinity of the bicyclic alcohols 2 and 3 was determined with tissue membrane preparations from guinea pig brain (for σ₁) and rat liver (for σ₂). Compounds 2a-d and 3a-d were also tested against human σ₁ receptors from multiple myeloma RPMI 8226 cell line membrane preparations (Geiger et al. 2007; Weber et al. 2016).

Almost all cyclohexylmethyl substituted compounds 2a-d and 3a-d show high σ₁ receptor affinity with K _i values in the low-nanomolar range. The only exceptions are ent-2a and 3c, both with a K _i value of 23 nM. The extension of the ethano-bridge of 2a-c and 3a-c by a methylene moiety does not influence σ₁ receptor affinity since the propane-bridged homologs 2d,e and 3d,e show approximately the same affinity. Only 2e and ent-2e show K _i values in the three-digit nanomolar range (K _i(σ₁) = 125 and 118 nM). However, the introduction of a second methylene moiety leads to a salient decrease in σ₁ receptor affinity, which implies that butano-bridged diazabicycloalkanols 2f and 3f are not tolerated by the σ₁ receptor.

The stereochemistry has only low impact on σ₁ receptor affinity since all four stereoisomers 2a, ent-2a, 3a, and ent-3a show the same σ₁ receptor affinity. However, in case of PMB-substituted derivatives, 2e and ent-2e show lower σ₁ receptor affinity than 3e and ent-3e.

The σ₁ affinity determined with human σ₁ receptor material is in good accordance with the σ₁ affinity obtained with σ₁ receptors from guinea pig brain.

Compared with the flexible hydroxyethyl piperazines 1b, 1d, and 1g, the corresponding ethano-bridged piperazines 2a-c and 3a-c reveal the same σ₁ receptor affinity. However, the conformational restriction of the hydroxypropyl piperazines 1d led to increased σ₁ receptor affinity of 2e and 3e. That is due to the higher flexibility of the hydroxypropyl piperazines 1d,e compared to their shorter hydroxyethyl homologs 1a-d. This is not valid for the butano-bridged piperazines 2f and 3f which display very low σ₁ affinity, indicating that the bridge size is too bulky for the binding pocket. Obviously the size of the butano-bridge outweighed its positive effect of conformational restriction.

The σ₁/σ₂ selectivity varies from low preference for the σ₁ receptor (2c: σ₁/σ₂ = 2) up to high selectivity for the σ₁ receptor (3d: σ₁/σ₂ = 178). The PMB-substituted derivative ent-3e (σ₁/σ₂ = 227) showed the highest σ₁/σ₂ selectivity.

The bicyclic compounds 2a-d and 3a-d were docked into the binding site of the 3D homology model to determine the free binding energies. Figure 5 illustrates the identified interactions between the high affinity compound ent-3a (K _i(σ_1human) = 1.6 nM) and the human σ₁ receptor (Weber et al. 2016).

Fig. 5

Interactions between ent-3a and amino acids of the binding site in the 3D homology model of the human σ₁ receptor

The calculated free binding energies of all docked compounds are in good accordance with the experimentally determined receptor binding data. For ent-3a the calculated ΔG_bind is −10.93 ± 0.34 kcal/mol corresponding to a calculated K _i value of 9.7 nM. This K _i value is in good agreement with the K _i values recorded with σ₁ receptors from guinea pig brain (K _i = 5.7 nM) and from human RPMI 8226 cell (K _i = 1.6 nM) membrane preparations.

The cell growth inhibition potential of compounds 2a-e and 3a-e was evaluated in five different cancer cell lines (Table 2). The naphthylmethyl substituted derivatives 2b and 3b similarly inhibited the growth and survival of all tested cell lines. The benzyl substituted derivatives 3a and ent-2a and the biphenylylmethyl substituted compound 2c show moderate inhibition of cell growth of the bladder cell line 5637. A clear correlation between σ₁ receptor affinity and growth and survival inhibition could not be determined, however, we did discern a trend revealing sensitivity of A-427 cell line against all tested bicyclic compounds. With the exception of 2b and 3b, growth and survival of the other cell lines were not inhibited at compound concentrations up to 10 or 20 μM.

Table 2 Pharmacological data for diazabicycloalkanols 2 and 3

The bridge size does not show additional influence on σ₁ receptor affinity. The K _i values of 2d and 3d are in the same range as the K _i values of the ethano-bridged compounds 2a-c and 3a-c. Although IC₅₀ values are not available for compounds 2e and 3e, cell growth of only 33–46 % could be detected for the A-427 cell line, whereas the growth of the other cell lines was not inhibited (Geiger et al. 2007).

Further experiments directed to elucidate the mechanism of cell growth inhibition showed that ent-2a induced apoptosis in A-427 cells in a time-dependent manner (Weber et al. 2016).

6 Spirocyclic σ₁ Receptor Ligands

Spirocyclic compounds 4 represent high affinity σ₁ receptor ligands with a favorable pharmacological profile for use as fluorinated PET tracers. The homologous fluoroalkyl derivatives 4a-d bind σ₁ receptor with K _i values in the low-nanomolar range and show high selectivity over the σ₂ receptor. The (S)-configured fluoroethyl substituted compound (S)-4b is currently investigated as PET tracer for imaging of σ₁ receptors in the CNS of patients suffering from major depression (Fischer et al. 2011; Wang et al. 2013; James et al. 2012). The spirocyclic σ₁ receptor ligands 5 bearing an exocyclic amino moiety allow diverse modifications by the introduction of two N-substituents. Furthermore, the existence of cis/trans isomerism increases the diversity of this compound class (Rack et al. 2011).

6.1 Homologous Fluoroalkyl Derivatives

The homologous fluoroalkyl derivatives 4a-d were developed from the 2-benzofuran 15, a ligand with high σ₁ receptor affinity (K _i = 1.1 nM) and high selectivity over the σ₂ receptor (K _i = 1,280 nM) and over 60 other receptors and ion channels like the hERG K⁺-channel. To eliminate the metabolically unstable acetalic function and to open up the possibility to introduce a fluorine atom into the molecule, a fluoroalkyl residue was installed instead of the acetalic moiety.

The synthesis of fluoroalkyl derivatives 4a-d started with the acetalization of 2-bromobenzaldehyde 13 to yield the dimethyl acetal 14 (Scheme 3). Homologation of 13 with a Wittig reagent provided the α,β-unsaturated acetal 17. Halogen-metal-exchange of the acetals 14 and 17 with n-BuLi followed by addition of 1-benzylpiperidin-4-one and subsequent transacetalization afforded the spirocyclic 2-benzofurans 15 and 18. The 2-benzofuranes 15 and 18 served as key intermediates for the synthesis of alcohols 16. The alcohols were reacted with diethylaminosulfur trifluoride (DAST) to provide the homologous fluoroalkyl derivatives 4 (Maestrup et al. 2009, 2011; Grosse Maestrup 2010; Maier and Wunsch 2002a, b).

Scheme 3

Synthesis of homologous fluoroalkyl derivatives 4. Reagents and reaction conditions (a) HC(OCH₃)₃, p-TosOH, CH₃OH, reflux, 16 h; (b) (1) n-BuLi, 1-benzylpiperidin-4-one, THF, −95 °C, 2 h, rt, 4 h; (2) TosOH, CH₃OH, rt, 7 d; (Maier and Wunsch 2002a) (c) n = 1: (1) trimethylsilyl cyanide, tetracyanoethylene, CH₃CN, reflux, 4 h; (2) H₂SO₄, EtOH, reflux, 7.5 h; (3) LiAlH₄, Et₂O, −15 °C, 30 min; (Maier and Wunsch 2002a) n = 3: (1) allyltrimethylsilane, BF₃ ^.OEt₂, CH₂Cl₂, −25 °C, 20 min, 0 °C, 4 h; (2) 9-BBN, THF, rt, 16 h; (3) H₂O₂, NaOH, −25 °C, 45 min, rt 1 h; (Maestrup et al. 2009) (d) DAST, CH₂Cl₂, −78 °C to rt., 17 h; (Maestrup et al. 2009) (e) [(CH₂O)₂CHCH₂PPh₃Br], K₂CO₃, TDA-1, CH₂Cl₂, reflux, 6 d; (f) (1) n-BuLi, THF, 1-benzylpiperidin-4-one, −78 °C, 1 h, rt, 16 h; (2) HCl, THF, rt, 2 h; (g) n = 2: NaBH₄, CH₃CN, 0 °C, 15 min, rt, 16 h; (Maestrup et al. 2011) n = 4: (1) ethoxycarbonylmethylentriphenylphosphorane, K₂CO₃, THF, reflux, 23 h; (2) H₂, Pd/C, EtOH, 1 bar, rt, 15 min; (3) LiAlH₄, THF, −20 °C, 30 min; (Grosse Maestrup 2010) (h) DAST, CH₂Cl₂, −78 °C to rt, 18–21 h (Maestrup et al. 2011)

The σ₁ and σ₂ receptor affinity of the homologous fluoroalkyl derivatives 4 was determined with receptor material from guinea pig brain (σ₁) and rat liver (σ₂).

All four fluoroalkyl homologs 4a-d bind with very high affinity to the σ₁ receptor (K _i(σ₁) = 0.59–1.4 nM) with high selectivity over the σ₂ subtype (Table 3). The fluoroethyl derivative 4b, termed fluspidine, shows the most promising ligand-binding profile [K _i (σ₁) = 0.59 nM, K _i (σ₂) = 785 nM)].

Table 3 σ₁ and σ₂ binding affinities of homologous fluoroalkyl derivatives 4

All four compounds 4a-d were also synthesized in their [¹⁸F]-labeled form for the use as PET tracers. For radiosynthesis, the alcohols 14 were transformed into the corresponding tosylates. Nucleophilic substitution of the tosylates with K[¹⁸F]F complexed with the cryptand Kryptofix 2.2.2 led to the [¹⁸F] labeled spirocyclic σ₁ receptor ligands [¹⁸F]4a-d with high radiochemical purity (>98%) and radiochemical yield (40–50%) with reaction times <30 min (Fischer et al. 2011; Maestrup et al. 2009; Maisonial et al. 2011, 2012). Because of the high target affinity and selectivity, [¹⁸F]4b was further evaluated in animal studies with female CD-1 mice. [¹⁸F]4b showed fast and sufficient brain uptake (3.9 and 4.7%ID/g) and high metabolic stability in vivo (>94% parent compound in plasma samples after 30 min, only one metabolite was found). Good concordance between expression of σ₁ receptors and binding site occupancy with [¹⁸F]4b was found by ex vivo brain section imaging (Fischer et al. 2011). Due to the promising properties of the racemic compound [¹⁸F]4b, the enantiomers (R)- and (S)-[¹⁸F]4b were separated by chiral HPLC of the tosylate 13b (Holl et al. 2013). The σ₁ receptor affinity was 0.57 nM for (R)-[¹⁸F]4b and 2.3 nM for (S)-[¹⁸F]4b. Thus, the (R)-enantiomer is the eutomer.

6.2 Spirocyclic Ligands with Exocyclic Amino Moiety

For the synthesis of spirocyclic ligands with exocyclic amino moiety, the dimethyl acetal 19 was reacted in a bromine lithium exchange to give an aryllithium intermediate. After addition to cyclohexane-1,4-dione followed by transacetalization under acidic conditions, reductive amination with benzylamine and NaBH(OAc)₃ led to the diastereomeric benzylamines cis-5e and trans-5e (R = H, Scheme 4). In order to investigate the influence of the second N-substituent, the benzylamines 5e were transformed into different tertiary amines. Each isomer can adopt different conformations with axially or equatorially oriented amino substituents.

Scheme 4

Synthesis of spirocyclic ligands 5 with exocyclic amino moiety. Reagents and reaction conditions (a) (1) (methoxymethyl)triphenylphosphonium chloride, KO^tBu, THF, −10 °C, then rt, 16 h; (2) pTosOH^.H₂O, MeOH, reflux, 72 h; (b) (1) n-BuLi, THF, −78 °C, 20 min; (2) cyclohexane-1,4-dione, −78 °C, 2 h, rt, 1 h; (3) CHCl₃, HCl, rt, 1.5 h; (4) benzylamine, THF, HOAc, NaBH(OAc)₃, rt, 2 h; (5) R-CHO, NaBH(OAc)₃, CH₂Cl₂, rt, 23 h (Rack et al. 2011)

The σ₁ and σ₂ receptor affinity of spirocyclic ligands with exocyclic amino moiety 5 was determined with membrane preparations obtained from guinea pig brain (for σ₁) and rat liver (for σ₂) (Table 4).

Table 4 σ₁ and σ₂ binding affinity of spirocyclic ligands 5 with exocyclic amino moiety

The shift of the basic amino group to a position outside of the spirocyclic ring was envisaged to come closer to the required distances between the pharmacophoric elements (benzene ring and amino moiety) according to the models of Glennon and Laggner. The benzylpiperidin 15 (Fig. 6) shows high σ₁ receptor affinity and high selectivity over the σ₂ subtype and over other receptors and ion channels. It was found that small residues at the N-atom resulted in low σ₁ affinity whereas a benzyl group turned out to be optimal. The important role of the N-benzyl moiety can be explained by the pharmacophore model of Glennon et al. The benzene ring of the annulated pyrane ring interacts with the primary hydrophobic region and the benzyl moiety of the N-atom interacts with the secondary hydrophobic region. However, the distance between the N-atom and the primary hydrophobic region was too small for both conformers with axially and equatorially oriented phenyl ring. Ideally, the distance should be 6–10 Å due to the pharmacophore model of Glennon et al. In case of 15 the distance was found to be 5.7 and 5.1 Å for the equatorial and axial conformer, respectively (Fig. 6). Therefore it was decided to extend the distance between the N-atom and the O-heterocycle-annulated benzene ring by exclusion of the N-atom from the piperidine ring, resulting in spirocyclic compounds 5 with exocyclic amino group. Another advantage of an exocyclic amino moiety is the possibility to install and to modify two different residues at the N-atom.

Fig. 6

Distance calculation of spirocyclic σ₁ receptor ligands with endocyclic N (15) and exocyclic amino moiety (trans-5a and cis-5a)

As a result of the shift of the basic group, the distances between the N-atom and the O-heterocycle-annulated benzene ring of cis-5a and trans-5a are in good concordance with the distances postulated by Glennon et al. However, the decrease in σ₁ affinity (K _i(σ₁) = 24 and 43 nM) (Rack et al. 2011) provides an example of how receptor binding affinity does not strictly correspond with pharmacophore models. Other considerations like entropic factors should be noted. The introduction of the benzylamino moiety leads to an increased flexibility of the N-substituent.

The σ₁ receptor affinity of 5 also depends on the second N-substituent. Only small groups are tolerated. For small methyl and ethyl groups, the K _i values are 24 and 107 nM, respectively. Bulky residues like pentyl- or benzyl substituents lead to a salient decrease in σ₁ binding affinity (K _i(σ₁) > 1,000 nM). Generally, cis-configured diastereomers show higher σ₁ binding affinity than their trans-configured diastereomers.

7 1,3-Dioxanes

Racemic 1,3-dioxane 6c represents a very potent σ₁ receptor antagonist (Utech et al. 2011). With these compounds σ₁ binding affinity depends on the relative configuration of the substituents at the 2- and 4-position, size of the oxygen containing heterocycle, and length of the aminoalkyl side chain. Since the racemic compound 6c, consisting of a six-membered O-heterocycle combined with an aminoethyl side chain, was found to be a promising candidate as σ₁ receptor ligand, the enantiomers were synthesized and their pharmacology evaluated.

For the synthesis of enantiomerically pure 1,3-dioxanes 6, the enantiomeric azidodiols (S)-22 and (R)-22 were prepared from diester 20 in high enantiomeric excess (Scheme 5). After silylation of 20 and subsequent reduction, the resulting diol 21 was converted into the azidodiols (S)-22 and (R)-22 following two different pathways using lipases as chiral catalysts. Stereoselective acetalization of (S)-22 and (R)-22 with benzaldehyde or acetophenone led to enantiomerically pure azido-1,3-dioxanes, which were subsequently reduced with H₂ and Pd/C to obtain the primary amines 6a and 6b. Further functionalization of the amino moiety was performed by reductive monobenzylation with benzaldehyde and NaB(OAc)₃ to yield the benzylamines 6c and 6d.

Scheme 5

Synthesis of enantiomerically pure 1,3-dioxanes 6. Reagents and reaction conditions. (a) (1) Me₂PhSiCl, imidazole, CH₂Cl₂; (2) LiBH₄, Et₂O; (b) (1) IPA, lipase Candida Antarctica B, MTBE; (2) lipase Burkholderia cepacia, NaHCO₃; (3) Zn(N₃)₂ ^.(pyridine)₂, DIAD, PPh₃, toluene; (4) K₂CO₃, CH₃OH; (5) HCl; (c) (1) IPA, lipase Burkholderia cepacia, MTBE; (2) Zn(N₃)₂ ^.(pyridine)₂, DIAD, PPh₃, toluene; (3) K₂CO₃, CH₃OH; (4) HCl; (Kohler and Wunsch 2006, 2012) (d) (1) Ph-C(=O)R¹, pTosOH, toluene, Dean Stark apparatus, 4 h; (2) H₂, Pd/C, rt, 5 h; R² = PhCH₂: (3) benzaldehyde, NaBH(OAc)₃, CH₂Cl₂, rt, 16 h (Kohler et al. 2012). IPA isopropenyl acetate, MTBE methyl tert-butyl ether, DIAD diisopropyl azodicarboxylate

The σ₁ and σ₂ receptor affinity of 1,3-dioxanes 6 was determined with tissue membrane preparations from guinea pig brain (for σ₁) and rat liver (for σ₂). The 1,3-dioxanes 6 were also tested against the PCP binding site of the NMDA receptor using pig brain cortex membrane preparations (Table 5).

Table 5 σ₁ and NMDA receptor binding affinities of 1,3-dioxanes 6a-d

We found that both enantiomers of the primary amines 6a and 6b do not bind σ₁ (K _i(σ₁) > 10,000 nM) in this assay. According to the pharmacophore model of Glennon et al. (Glennon 2005; Glennon et al. 1994) affinity for σ₁ should increase by introducing an N-benzyl group as a second hydrophobic residue flanking the basic amino moiety as it is shown for the secondary amines 6c and 6d. In the case of benzylamines the orientation of the phenyl ring at the 1,3-dioxane ring has minimal influence on σ₁ affinity as 6a and 6b show comparable K _i values of 6.0 and 17 nM for the (4R)-configured enantiomers and 50 and 11 nM for the (4S)-configured enantiomers. Compound (S,R)-6c with equatorially oriented 2-phenyl moiety shows high σ₁ affinity (K _i(σ₁) = 6.0 nM).

Regarding σ₁/σ₂ selectivity, we found that primary amines 6a and 6b do not bind σ₁ and the benzyl amines 6c and 6d bind σ₂ with low affinity (K _i(σ₂) > 200 nM).

Depending on the absolute configuration the primary amines 6b with axially oriented phenyl moiety reveal high affinity to the PCP binding site of the NMDA receptor with K _i values of 46 nM ((R,R)-6b) but only 6,120 nM for (S,S)-6b, respectively. The equatorial orientation of the phenyl ring (6a) as well as the introduction of a benzyl group at the amino moiety (6c and 6d) led to complete loss of NMDA affinity (K _i(NMDA) > 10,000 nM).

The benzyl substituted 1,3-dioxane (S,R)-6c represents the most potent candidate among the secondary amines with high σ₁ affinity (K _i(σ₁) = 6.0 nM) and high selectivity over the σ₂ subtype and the NMDA receptor.

In further studies performed with racemic 6c (K _i(σ₁) = 19 nM), promising results were obtained in a capsaicin-induced pain assay with mice. In these studies, even a very low dose of 0.25 mg/kg, rac-6c has high anti-allodynic activity (Utech et al. 2011).

8 Arylalkenylamines

Arylpropenylamines of type 7 show high σ₁ binding affinity and high σ₁/σ₂ selectivity. The influence of the novel σ₁ ligands on nerve growth factor (NGF)-induced neurite outgrowth was evaluated in the in vitro PC12 cell neurite sprouting assay.

Michael addition of cylic amines to unsaturated ketone 23 led to the β-aminoketones 24. Subsequent nucleophilic aryllithium addition followed by dehydration with HCl provided the arylalkenylamines 7, which were crystallized as HCl salts (E)-7a-f (Scheme 6). The racemic arylalkylamine 7g was obtained by catalytic hydrogenation of 7a.

Scheme 6

Synthesis of arylalkenylamines 7a-f. Reagents and reaction conditions (a) HNR₂, PEG 400, rt; (b) (1) Ar-Br, t-BuLi, Et₂O, −78 °C to rt; (2) 37% HCl, rt; (3) 1 M NaOH; (4) 37% HCl, rt; (5) crystallization from acetone (Rossi et al. 2011). Naph naphthyl

The σ₁ and σ₂ receptor binding affinity of compounds 7a-g ^.HCl was tested using guinea pig brain (for σ₁) and rat liver (for σ₂) membrane preparations. Additionally, the selectivity towards the PCP binding site of the NMDA receptor and against μ- and κ-opioid receptors was determined (Table 6).

Table 6 σ₁ and σ₂ affinity of arylalkenylamines 7a-g

Piperidinyl substituted compounds 7a and 7b reveal high σ₁ receptor affinity independent of the aromatic residue (K _i(σ₁) = 0.86 and 0.97 nM). Interestingly, naphthalen-2-yl- or biphenyl-4-yl residues appear to be important for high σ₁ binding affinity of morpholinyl substituted compounds since only low σ₁ affinity is found when a phenyl substituent is present as aromatic moiety (7f, K _i(σ₁) >1,000 nM). The tested compounds show high selectivity over the σ₂ receptor subtype, opioid receptors, and NMDA receptors (the PCP binding site). High σ₂ binding affinity was found only for 4-benzylpiperidinyl substituted derivatives 7c and 7d, with K _i values of 18 and 16 nM, respectively. 7a represents the most potent and selective σ₁ receptor ligand of this set of compounds (K _i(σ₁) = 0.86 nM, σ₁/σ₂ = 129). Therefore the corresponding arylalkylamine 7g was included in this study. Receptor binding studies revealed similar σ₁ binding affinity (K _i(σ₁) = 0.70 nM) and selectivity (σ₁/σ₂ = 147) (Rossi et al. 2011).

clogP and clogD values were calculated for 7a-g. Their drug-like properties were confirmed according to Lipinski’s “rule of five.” With the exception of 7c (logD > 5) all compounds fulfill the “rule of five,” i.e., clogD <5, molecular weight < 500, H-bond acceptors <10, and H-bond donors <5.

In order to determine whether the top compounds 7a and 7g function as agonists or antagonists their influence on nerve growth factor (NGF)-induced neurite outgrowth in PC12 cells was evaluated. 7g potentiated the neurite outgrowth at lower concentrations, consistent with agonist activity. This effect was blocked by co-treatment with the σ₁ receptor antagonist BD-1063, demonstrating the participation of σ₁ receptors. In contrast, (E)-7a did not significantly increase neurite sprouting (Rossi et al. 2011).

9 Morpholinoethoxypyrazoles

Substituted 1-arylpyrazoles with a basic amino function represent a promising class of σ₁ receptor antagonists. For high σ₁ binding affinity, the distance between the basic amino moiety and the pyrazole ring is of major importance. In previous studies an ethylenoxy spacer and a morpholino residue as the N-component resulted in high σ₁ affinity and excellent selectivity over the σ₂ subtype.

For the synthesis of morpholinoethoxypyrazoles 8a-f, the 3-hydroxypyrazole 28 was prepared in a two-step reaction sequence starting from arylhydrazines 25. At first the terminal amino group of hydrazines 25 was protected by acetylation (Scheme 7). Reaction of 26 with β-ketoesters 27 led to the 3-hydroxypyrazoles 28 with high regioselectivity. Subsequent reaction with 4-(2-chloroethyl)morpholine provided the morpholinoethoxypyrazoles 8a-f (Diaz et al. 2012).

Scheme 7

Synthesis of morpholinoethoxypyrazoles 8a-f. Reagents and reaction conditions (a) Ac₂O, toluene, rt; (b) PCl₃, 50 °C, 2 h; NaH, DMF, 60 °C, 4 h; (c) 4-(2-chloroethyl)morpholine, K₂CO₃, NaI, DMF, 95 °C, 18 h (Diaz et al. 2012)

The σ₁ and σ₂ binding affinity of morpholinoethoxypyrazoles 8a-f was determined with human σ₁ receptors (from transfected HEK-293 cell membrane preparations) and membrane preparations from guinea pig brain (for σ₂).

Generally, the morpholinoethoxypyrazoles 8a-f show only very low affinity for the σ₂ receptor. Affinity for σ₁ depends on the substitution pattern of the pyrazole ring. Substitution at position 1 with aromatic residues (8b-f) produces high σ₁ binding affinity. Non-aromatic residues (tert-butyl, 8a) produce a salient decrease in σ₁ binding affinity (Table 7). Only small residues (e.g., CH₃, H) are tolerated at 4- and 5-position of the pyrazole ring. In the naphthyl series even a methyl group in position 4 seems to be detrimental for high σ₁ affinity (8d, K _i(σ₁) = 139 nM). The introduction of larger moieties in position 4 (e.g., 8f, C(=O)Me) produces a salient decrease in σ₁ binding affinity (K _i(σ₁) = 741 nM). The 5-alkoxy regioisomer of the most promising ligand 8c (K _i(σ₁) = 17 nM) shows a complete loss of σ₁ affinity (K _i > 1,000 nM) (Diaz et al. 2012).

Table 7 σ₁ and σ₂ binding affinity of morpholinoethoxypyrazoles 8a-f

The high σ₁ binding affinity of naphthylpyrazole 8c (also known as S1RA and E-52862) cannot be explained completely by the common pharmacophore models. The 2-[1-(2-naphthyl)pyrazol-3-yloxy]ethyl moiety fits well into the primary hydrophobic region of the Glennon model, tolerating sterically demanding residues. However, the morpholine ring does not fulfill the requirements to address the second hydrophobic region.

Ligands 8 with excellent σ₁ receptor binding affinity and selectivity were further evaluated for their activity at the hERG channel and for efficacy in mouse models of neuropathic pain. Naphthylpyrazole 8c proved to be the most promising candidate with regard to metabolic stability, interaction with the hERG channel (IC₅₀ > 10 μM), and analgesic activity in different pain models (Diaz et al. 2012). It was found that 8c shows dose-dependent analgesic effects in both the capsaicin-induced hypersensitivity and the formalin-induced pain model. In the partial sciatic nerve ligation mouse model, 8c shows dose-dependent inhibition of thermal hypersensitivity and mechanical allodynia, comparable to the effects of pregabalin, the gold-standard for the treatment of neuropathic pain.

The selectivity of 8c towards 170 other targets including various receptors and ion channels was shown. With the exception of moderate affinity for the human serotonin 5-HT_2B receptor (K _i(5-HT_2B) = 328 nM) other targets were not engaged by 8c. The antagonist profile of 8c was verified using phenytoin as an allosteric modulator of σ₁.

The chemical properties of 8c meet Lipinski’s “rule of five.” The pharmacokinetic properties were evaluated in mice. Due to the acceptable solubility and high metabolic stability, a good oral bioavailability can be assumed.

In light of all of the aforementioned properties, 8c entered clinical trials. Passing the single and multiple dose phase I clinical study provided proof-of-concept that 8c is safe and well tolerated by healthy humans. Thus, the development of 8c will be continued into phase II clinical trials (Abadias et al. 2013).

10 3,4-Dihydro-2(1H)-quinolones

3,4-Dihydro-2(1H)-quinolones 9 were developed following the idea of combining a piperidine or morpholine basic element (as realized in the σ₁ receptor antagonist 8c) with the quinolone scaffold, which was identified as the interacting element at the σ₁ receptor (Oshiro et al. 2000). The aim was to obtain compounds with high affinity for the σ₁ receptor and potent anti-nociceptive properties as shown for S1RA.

For the synthesis of quinolones 9, 7-hydroxyquinolone 29 was alkylated with dibromoalkans with various length of the alkyl group (Scheme 8). Subsequent reaction with morpholine or piperidine led to the amines 30, which were converted into 1-alkylated quinolones 9 by reaction with benzyl bromides or iodomethane in presence of NaH (Lan et al. 2014).

Scheme 8

Synthesis of quinolones 9. Reagents and reaction conditions (a) (1) Br(CH₂)_nBr, K₂CO₃, acetone, reflux; (2) HNR₂, K₂CO₃, KI, CH₃CN, reflux; (b) NaH, DMF or THF, 0–50 °C (Lan et al. 2014)

The σ₁ and σ₂ affinity of quinolones 9 was determined in competition experiments using guinea pig brain membrane preparations (Table 8).

Table 8 σ₁ and σ₂ binding affinity of 3,4-dihydro-2(1H)-quinolinones 9

The distance between the quinolone scaffold and the morpholine ring has a strong impact on σ₁ binding affinity. Whereas 9a with an ethylene spacer has negligible affinity for σ₁ (K _i(σ₁) > 2,000 nM), the corresponding homolog 9b bearing three CH₂ moieties in the side chain binds σ₁ with high affinity (K _i(σ₁) = 14.8 nM). Elongation by introduction of additional methylene moieties decreased σ₁ affinity in the order n = 3 > n = 4 > n = 5 > n = 6. Compound 9i, with a hexamethylene linker, had the lowest σ₁ binding affinity of this set of compounds with a K _i value of 682 nM. Replacement of the morpholine ring of the most potent ligand 9b of this series by a piperidine ring led to an almost tenfold increase in σ₁ binding affinity (K _i(σ₁) = 1.84 nM). For the promising piperidine derivatives, the effect of the quinolinone N-substituent on σ₁ receptor binding was evaluated by synthesizing different substituted analogs 9c, 9d, and 9f. It has been found that small residues led to decreased σ₁ affinity compared with the N-benzyl substituted derivative 9e (K _i(σ₁) = 1.84 nM). In the case of substitution with a proton or a methyl group, the K _i values were only 89 nM (9c) or 34 nM (9d). The substitution of the phenyl ring with an electron-withdrawing fluorine atom led to a slight increase in σ₁ binding affinity (9f, K _i(σ₁) = 1.22 nM).

Regarding σ₁/σ₂ selectivity, it was found that the most potent σ₁ ligand 9f has also the highest selectivity with a σ₂ K _i > 1,000 nM. For the piperidine derivatives 9c-f, the σ₂ affinity increased with decreased size of substituents. The secondary lactam 9c shows the highest σ₂ affinity and the lowest σ₁/σ₂ selectivity of this set of compounds (K _i(σ₂ = 288 nM, σ₁/σ₂ = 3). The chain length between the quinolone scaffold and the morpholine residue also influences affinity for the σ₂ receptor. Compounds 9a (n = 2), 9h (n = 5), and 9i (n = 6) do not show σ₂ affinity. However, compounds 9b and 9g with a trimethylene or tetramethylene linker displayed moderate σ₂ affinity, with a K _i of approximately 500 nM.

The most potent ligand 9f was further evaluated for its anti-nociceptive activity in the formalin-induced pain assay. It was found that 9f dose-dependently reduced both phases of the pain response with ED₅₀ values of 49.4 ± 4.1 and 50.5 ± 2.5 mg/kg for the acute phase I and the longer-lasting tonic phase II, respectively. The σ₁ antagonist activity of 9e was shown using phenytoin as an allosteric modulator of σ₁.

11 Conclusion

During the past several years, the fields of σ₁ receptor chemistry and pharmacology have made remarkable progress. Various pharmacophore models of σ₁ ligands, a 3D homology model of the σ₁ receptor, its structure in solution (NMR), and its structure in the solid state (X-ray crystallography) have been reported, allowing a closer look at the binding properties of σ₁ receptors to their ligands. Evidence of σ₁ as a promising target for the development of new therapeutic approaches has been demonstrated. The σ₁ antagonist S1RA (8c) is currently in clinical trials for the treatment of neuropathic pain. Bicyclic piperazines 2 and 3 inhibit the growth of small cell lung cancer cells (A-427 cells) in a dose-dependent manner, demonstrating their potential as new tumor therapeutics. The (S)-configured spirocyclic σ₁ antagonist 4b (fluspidine) with a fluoroethyl side chain has been developed as tracer for positron emission tomography (PET) and is currently in clinical trials for imaging and analysis of the brain of patients suffering from major depression.