Propellanes as Rigid Scaffolds for the Stereodefined Attachment of σ-Pharmacophoric Structural Elements to Achieve σ Affinity

Following the concept of conformationally restriction of ligands to achieve high receptor affinity, we exploited the propellane system as rigid scaffold allowing the stereodefined attachment of various substituents. Three types of ligands were designed, synthesized and pharmacologically evaluated as σ1 receptor ligands. Propellanes with (1) a 2-methoxy-5-methylphenylcarbamate group at the “left” five-membered ring and various amino groups on the “right” side; (2) benzylamino or analogous amino moieties on the “right” side and various substituents at the left five-membered ring and (3) various urea derivatives at one five-membered ring were investigated. The benzylamino substituted carbamate syn,syn-4a showed the highest σ1 affinity within the group of four stereoisomers emphasizing the importance of the stereochemistry. The cyclohexylmethylamine 18 without further substituents at the propellane scaffold revealed unexpectedly high σ1 affinity (Ki = 34 nM) confirming the relevance of the bioisosteric replacement of the benzylamino moiety by the cyclohexylmethylamino moiety. Reduction of the distance between the basic amino moiety and the “left” hydrophobic region by incorporation of the amino moiety into the propellane scaffold resulted in azapropellanes with particular high σ1 affinity. As shown for the propellanamine 18, removal of the carbamate moiety increased the σ1 affinity of 9a (Ki = 17 nM) considerably. Replacement of the basic amino moiety by H-bond forming urea did not lead to potent σ ligands. According to molecular dynamics simulations, both azapropellanes anti-5 and 9a as well as propellane 18 adopt binding poses at the σ1 receptor, which result in energetic values correlating well with their different σ1 affinities. The affinity of the ligands is enthalpy driven. The additional interactions of the carbamate moiety of anti-5 with the σ1 receptor protein cannot compensate the suboptimal orientations of the rigid propellane and its N-benzyl moiety within the σ1 receptor-binding pocket, which explains the higher σ1 affinity of the unsubstituted azapropellane 9a.


Introduction
In 1976, Martin and coworkers [1] postulated σ receptors as the third type of opioid receptors. The name σ receptor was derived from the benzomorphan SKF-10,047, which caused a unique pharmacological profile in animal studies. Twenty years later, the σ 1 receptor was cloned from different tissues of different speces [2][3][4][5][6]. Subsequently, several models of the structure of the σ 1 receptor were reported, until it was crystallized in 2016, confirming its unique structure [7,8]. The identification of the σ 2 receptor took an even longer time. In 2017, the identity of the σ 2 receptor and the endoplasmic reticulum (ER)resident transmembrane protein 97 (TMEM97) was shown [9,10]. Very recently, the first structure of the human σ 2 receptor was reported. [11] The σ 1 receptor is involved in various neuropsychiatric disorders, such as schizophrenia and depression [12][13][14][15][16]. Several clinically used antidepressants show medium to high σ 1 receptor affinity in addition to their main mechanism of action [17][18][19][20]. σ 1 receptors also play a role in drug/alcohol dependence and neurodegenerative disorders (e.g., Alzheimer's disease) [21][22][23]. The σ 1 receptor antagonist S1RA has been successfully tested in phase II clinical studies for the treatment of neuropathic pain [24,25]. Since the exact signal transduction path of σ 1 receptors is not fully understood so far, analgesic activity in neuropathic pain mouse models is the best method to discriminate σ 1 receptor antagonists from agonists [26,27]. Several human tumors, including prostate, breast and bladder tumors, express a high density of σ 1 receptors. Strong metastasis and poor prognosis were associated with high expression levels of σ 1 receptors. Antagonists at σ 1 receptors were able to reduce tumor cell proliferation [28,29]. Several human tumor cells derived from various tissues (e.g., prostate, breast, colon and lung) overexpress σ 2 receptors. Agonists at the σ 2 receptor are capable of killing tumor cells via apoptotic and non-apoptotic machanisms [30][31][32][33][34].
The structures of σ 1 and σ 2 receptor ligands are quite diverse. Some prototypical σ ligands containing highly flexible structural elements are displayed in Figure 1. Binding of flexible ligands to a biological target is associated with an entropic penalty, since the binding site of the target forces the flexible ligand into a particular conformation leading to loss of conformational freedom of the ligand. We are interested in rather rigid ligands with a defined three-dimensional structure fitting exactly into the binding pocket of the target protein. In this respect, we reported on spiro-and bicyclic ligands with high affinity and selectivity for σ 1 receptors. As an example, the spirocyclic piperidine derivative (S)-fluspidine (1) is depicted in Figure 2 [35][36][37][38]. (S)-Fluspidine (1) interacts with low nanomolar affinity with σ 1 receptors and shows 300-fold selectivity against the σ 2 subtype. The 18-F-labeled analog [ 18 F]1 is currently investigated as the PET tracer for imaging of σ 1 receptors in the brain of patients suffering from major depression [39]. The piperazine derivative 2 rigidified by a propano bridge exhibits even higher σ 1 affinity than 1 and more than 40-fold selectivity over the σ 2 receptor [40]. On the other hand, the rigid granatane derivative 3 reported by Mach and coworkers [41] represents a ligand with a 30-fold preference for the σ 2 receptor. (Figure 2) Inspired by these conformationally restricted spiro-and bicyclic σ ligands 1-3 we introduced the propellane as novel rigid scaffold to achieve high σ 1 and/or σ 2 receptor affinity. With K i values of 77 nM and 82 nM the [4.3.3]propellane syn,syn-4a [42] and the 3-aza [4.4.3]propellane anti-5 [43] represent promising σ 1 receptor ligands. (Figure 2) Herein, we started further exploiting the propellane system as rigid scaffold to attach various functional groups and substituents, designed to address σ 1 and/or σ 2 receptors. (Figure 3) Due to the rigid structure of the propellane scaffold, all substituents adopt an exact orientation. The manuscript contains three parts. In the first part (compounds of type A), the carbamate at the "left" part of the propellane system was kept constant and the substituent at the second cyclopentane ring was modified (compare substituents of granatane 3). The second part deals predominantly with propellanes of type B containing an arylmethylamino moiety (and analogous amino groups) at the "right" side of the propellane system and variations of the "left" side. The third part investigated, whether the amino moiety on the "right" side could be replaced bioisosterically by a urea moiety as shown for compounds of type C. The urea is not basic, but represents a strong H-bond donor and acceptor.

Chemistry
The synthesis of the carbamates 4 started with the mixture of the diastereomeric hydroxyketones anti-6 and syn-6 [42], which was reacted with 2-methoxy-5-methylphenyl isocyanate in the presence of,Bu 2 Sn(OAc) 2 . The carbamates anti-7 and syn-7 were isolated in 36% and 31% yield, respectively. (Scheme 1) The X-ray crystal structure of syn-7 confirmed the successful formation of the carbamate and its syn-configuration at 8-position ( Figure 4). Moreover, the very long conjoining C-C bond, which belongs to all three rings of the propellane, was confirmed by the crystal structure (C1-C6 = 1.562(2) Å). Final reductive amination of the ketones anti-7 and syn-7 with primary amines and NaBH(OAc) 3 [44] provided the amines 8-anti-4 and 8-syn-4. Both type of compounds were obtained as mixture of diastereomers. The ratio of diastereomers was approximately 1:1, respectively.  3 , ClCH 2 CH 2 Cl, rt, 20 h, 50%. DMABn = (4-dimethylamino)benzyl. The residues R are defined in Table 1.  promising σ 1 affinity, which depends on the substitution pattern and the configuration. 43 Therefore, we decided to investigate the σ receptor affinity of the benzylamine 8 and modify the substituents in the 11-position. At first, the ketones 10-12 were reductively aminated with benzylamine and NaBH(OAc) 3 44 to yield the benzylamines 8, 13 and 14. The pure diastereomeric alcohols anti-11 and syn-11 were converted separately into the benzylamines 11-anti-13 and 11-syn-13, respectively, which were not further separated. Whereas the synand anti-configured benzylamines 8 were also not separated, the diastereomeric 1,3-dioxolanes 14 were separated by flash chromatography to obtain diastereomerically pure 8- Furthermore, the unsubstituted ketone 10 was reductively aminated with tryptamine to yield the indolylethylamine 15. The mixture of diastereomeric benzylamines anti-8 and syn-8 was treated with ammonium formate and Pd(OH) 2 removing reductively the benzyl moiety. Upon treatment with aldehydes and NaBH(OAc) 3 [44], the resulting primary amine 16 was further transformed into the dimethylamino substituted benzylamine 17 and the cyclohexylmethylamine 18. (Scheme 2).
The σ 1 affinity of the 3-azapropellane anti-9b with an OH moiety at 12-position is three-fold higher than the σ 1 affinity of its syn-diastereomer syn-9b. (Table 2) However, the low nanomolar σ 1 affinity of the unsubstituted 3-azapropellane 9a (K i = 17 nM) was unexpected. 42,43,48 Compared to the naked azapropellane 9a, the σ 1 affinity of the naked propellanamine 8 is 35-fold reduced. The 3-azapropellanes 9a and anti-9b and the naked propellanamine 8 show at least 10-fold selectivity for the σ 1 receptor over the σ 2 receptor. The least potent σ 1 ligand syn-9b has only a slight preference for the σ 1 receptor over the σ 2 subtype (Table 2). Modifications of the 11-substituent and the N-substituent led to propellanamines 13-17 with very low σ 1 and σ 2 affinity. However, the cyclohexylmethyl moiety increased the σ 1 and the σ 2 affinity remarkably. With K i values of 24 nM (σ 1 affinity) and 101 nM (σ 2 affinity), 18 represents the most potent σ ligand of this series of compounds. The increase of both σ 1 and σ 2 affinities by introduction of the cyclohexylmethyl moiety instead of the benzyl moiety has already been observed for some other classes of σ ligands [49,50].
In the third part of this project, the amino moiety on the "right" side was replaced by a urea to investigate, whether this H-bond donor and H-bond acceptor group could mimic the basic amino moiety. For this purpose, the propellane-8,11-dione 19 was reductively aminated with benzylamine. The resulting secondary amines syn-20 and anti-20 were separated by flash column chromatography and subsequently reacted with 2-methoxy-5methylphenyl isocyanate to obtain the urea derivatives syn-21 and anti-21, respectively. Final reduction of the ketones syn-21 and anti-21 with NaBH 4 provided the four diastereomeric N-benzylurea derivatives syn,anti-22, syn,syn-22, anti,anti-22 and anti,syn-22 bearing a secondary alcohol in 11-position (Scheme 3). The N-benzylurea derivatives anti-21, syn,anti-22 and anti,syn-22 were crystallized from EtOAc, leading to crystals suitable for X-ray crystal structure analysis. (Figures 6-8).   The crystal structure of anti-21 nicely confirms the anticonfiguration of the urea at the propellane system. The conjoining bond C1-C6 is rather long (1.550(3) Å). The cyclohexane ring adopts a chair conformation and the cyclopentane ring bearing the urea adopts an envelope conformation, which leads to an outward orientation of the large urea ( Figure 6).
The X-ray crystal structures of the diastereomeric N-benzylurea derivatives syn,anti-22 and anti,syn-22 containing an additional OH moiety in 11-position are shown in Figures 7 and 8. Both structures prove the syn,antiand anti,syn-configuration, respectively. The six-membered ring of the propellane scaffold of syn,anti-22 adopts a chair conformation. However, in the diastereomer anti,syn-22 a boat-like conformation was found for the cyclohexane ring. This boat-like conformation leads to an extraordinarily long conjoining bond (C1-C6 = 1.590(2) Å). Both five-membered rings of urea derivative anti,syn-22 adopt unusual envelope conformations and all three rings of the propellane scaffold flap in the same direction ( Figure 8). This pattern was only reported for heterocyclic 8,11-dioxa[4.3.3]propellanes [51].
To obtain urea derivatives without the N-benzyl substituent, the primary amine 16 was reacted with various isocyanates. Since the primary amine 16 was employed as 1.1-mixture of antiand syn-diastereomers, the urea derivatives 23a-d were obtained as 1:1-mixture of antiand syn-diastereomers as well (Scheme 4).   Table 3.
Finally, the dioxolane substituted benzylamine 14 (mixture of synand anti-diastereomers) was debenzylated by a transfer hydrogenolysis using NH 4 HCO 2 in the presence of Pd(OH) 2 . The mixture of diastereomeric primary amines 24 was isolated in 75% yield and converted into urea upon treatment with 3,4-difluorophenyl isocyanate. After hydrolysis of the dioxolane, the diastereomeric difluorophenylurea derivatives syn-25 and anti-25 were isolated in 42% and 35% yield, respectively. Reduction of the ketones syn-25 and anti-25 with NaBH 4 provided diastereomeric alcohols 26. Whereas the mixture of syn,anti-26 and syn,syn-26 was obtained as 1:1-mixture of diastereomers, anti,anti-26 and anti,syn-26 were separated by flash chromatography (Scheme 5). The syn-configuration of syn-25 was confirmed unequivocally by X-ray crystal structure analysis. In addition to the syn-configuration the chair conformation of the cyclohexane ring and the long conjoining bond C1-C6 = 1.546(3) Å was demonstrated ( Figure 9). The σ 1 and σ 2 affinities of the urea derivatives were determined in receptor binding studies. Unfortunately, the synthesized urea were not able to compete with the radioligands proving that the urea could not replace bioisosterically the basic amino moiety (Table 3).

Computational Studies
Within the class of propellanamines and azapropellanes some promising σ 1 receptor ligands were identified. Therefore, molecular dynamics (MD) simulations were performed to shed light on their mechanism of binding. The starting structure for the σ 1 receptor was obtained from the RCSB Protein Data Bank (PDB ID 5HK1) [7]. Following a consolidated computational protocol [49,52], the binding modes of compounds 9a ( Figure 10A,C) and anti-5 ( Figure 10B,D) were initially recognized. A MM/PBSA (molecular mechanics/Poisson-Boltzmann surface area) approach [53] provided the binding free energy (∆G) of the complexes of both compounds with the σ 1 receptor. The obtained energetic values are in good agreement with their different σ 1 affinity. The following ∆G values were obtained: −10.02 kcal/mol for 9a (K i (σ 1 ) = 17 nM) and −8.87 kcal/mol for anti-5 (K i (σ 1 ) = 82 nM).
As expected, both σ 1 ligands share the same thermodynamics pattern; their binding is enthalpy driven characterized by favorable van der Waals and electrostatic interactions. Specifically, our analysis resulted in an enthalpy contribution (∆H) of −18.47 kcal/mol and −17.58 kcal/mol for 9a and anti-5, respectively. Instead, the entropic components (−T∆S) penalize the binding with the σ 1 receptor with the corresponding values of 8.45 kcal/mol and 8.71 kcal/mol for 9a and anti-5, respectively. The compounds are shown as atom-colored sticks-and-balls (C, grey, N, blue and O, red) while the side chains of σ 1 residues mainly interacting with the ligands are depicted as colored sticks and labeled. Hydrogen atoms, water molecules, ions, and counterions are omitted for clarity. 2D schematic representation of the general stabilizing interactions for σ 1 /9a (C) and σ 1 /anti-5 (D) complexes. (E) Per-residue binding free energy decomposition (∆H res ) of the main involved amino acids in σ 1 /9a (light sea green) and σ 1 /anti-5 (firebrick) complexes. (F) MD distance between the carboxyl oxygen atom (O2) of E172 and the NH group of the ligand detected for σ 1 /9a (light sea green) and σ 1 /anti-5 (firebrick) complexes. (G) MD distance between the OH group of Y103 and the carboxyl oxygen atom (O1) of E172 detected for σ 1 /9a (light sea green) and σ 1 /anti-5 (firebrick) complexes.
Through the per-residue binding free energy deconvolution (PRBFED) of the enthalpic terms (∆H res ), the main amino acid residues of the σ 1 receptor involved in ligand binding were identified. Basically, by elucidation of the specific ligand/protein interactions, the PRBFED analysis ( Figure 10E) allowed to better understand the preference of the σ 1 receptor for the smaller azapropellane 9a. Specifically, azapropellane 9a is provided with the basic chemical features of a prototypical σ 1 receptor ligand: the rings of the propellane system can perform stabilizing hydrophobic and van der Waals interactions with the σ 1 residues W89, F107, Y120 and W164 (∆H res = −3.27 kcal/mol, Figure 10E) while the N-benzyl group is perfectly encased in an hydrophobic cavity underlying σ 1 residues I124,V152, and H154 (∆H res = −1.64 kcal/mol). However, the most peculiar interaction is definitively performed by its charged N-atom, involved in a permanent ionic interaction with the carboxylic group of E172 (∆H res = −4.93 kcal/mol) with a detected average dynamics length (ADL) of 1.75 ± 0.12 Å ( Figure 10A,F). Moreover, the optimal orientation of this interaction is guaranteed by a stable, internal hydrogen bond between the other O-atom of E172 and the OH moiety of Y103 (ADL = 1.73 ± 0.12 Å, Figure 10A,G). In the same series, the propellanamine derivative 18 exhibited a good σ 1 affinity (K i (σ 1 ) = 34 nM). Our MD study confirmed a very similar binding mode and interaction spectra as observed for the azapropellane 9a ( Figure S1, Supplementary Materials). Indeed, the charged secondary amine maintained the ionic interaction with E172 (∆H res = −4.44 kcal/mol, Figure S1) and the cyclohexyl moiety of 18 is able to positively interact in the hydrophobic cavity with σ 1 residues I124, V152 and H154 (∆H res = −1.72 kcal/mol). Accordingly, the σ 1 /18 complex is provided with a favorable ∆G value of −9.78 kcal/mol. Compared to 9a, the enthalpic contribution of anti-5 was almost 1 kcal/mol lower, although its 2-methoxy-5-methylphenyl carbamate moiety could provide additional interactions in the σ 1 receptor cavity ( Figure 10B,D). However, these additional interactions of anti-5 with residues L95, L182 and Y206 (∆H res = −1.68 kcal/mol, Figure 10E) were not specific and could not compensate the decrease of the other interactions due to a not optimal orientation in the binding site. Indeed, anti-5 cannot establish an optimal H-bond with the carboxylate side chain of E172 (∆H res = −3.65 kcal/mol) as demonstrated by the less stable ADL detected in our MD simulation (ADL = 2.12 ± 0.35 Å, Figure 10B,F). Furthermore, the assumed binding pose of anti-5 does not allow an optimal position of its N-benzyl moiety in the hydrophobic cavity underlying I124, V152 and H154 (∆H res = −0.53 kcal/mol, Figure 10E). Moreover, optimal stabilizing interactions of the propellane system of anti-5 with W89, F107, Y120 and W164 (∆H res = −2.58 kcal/mol) were not reached.

Conclusions
In this manuscript, the rigid propellane system was used to attach σ 1 and σ 2 pharmacophoric substituents in a defined three-dimensional orientation. It was shown, that propellanes with both substituents at the five-membered rings adopting syn-configuration (e.g., syn,syn-4a) exhibited high σ 1 affinity underlining the importance of the stereochemistry. Urea instead of the amino moiety led to the loss of σ 1 and σ 2 affinity. High σ 1 affinity was achieved by incorporation of the basic amino moiety into the propellane scaffold. The azapropellanes anti-5 and 9a demonstrated high σ 1 affinity, which is enthalpy driven. Although the carbamate moiety of anti-5 contributed to the binding free energy of anti-5 within the σ 1 receptor binding pocket, it forces the complete propellane scaffold and its benzyl moiety into a less favorable orientation in the binding pocket. As a result, the σ 1 affinity of anti-5 is lower than the σ 1 affinity of 9a.

General Procedure A for the Synthesis of Carbamates and Ureas
Under N 2 , propellanol of propellanamine (1 equi.), the respective isocyanate (1.2 eq.) and the catalyst Bu 2 Sn(OAc) 2 (0.2 eq.) were dissolved in THF (5 mL per 100 mg of propellanamine) and the mixture was stirred at rt for 24-48 h. Water (5 mL) was added and the mixture was stirred vigorously for 20 min. The mixture was extracted with EtOAc (3×), the combined EtOAc layers were washed with brine (1×), dried (Na 2 SO 4 ), filtered, the filtrate was concentrated in vacuo and the residue was purified by fc.
Exceptions and special features: For compound syn-25 three independent molecules were found in the asymmetric unit. All these three molecules present different groups disordered over two positions. Several restraints (SADI, SAME, ISOR and SIMU) were used in order to improve refinement stability.

X-ray Crystal Structure Analysis of syn-7
A colorless plate-like specimen of C 21 H 27 NO 4 , approximate dimensions 0.100 mm × 0.200 mm × 0.350 mm, was used for the X-ray crystallographic analysis. The X-ray intensity data were measured. The integration of the data using a monoclinic unit cell yielded a total of 3302 reflections to a maximum θ angle of 67.19 • (0.84 Å resolution), of which 3302 were independent (average redundancy 1.000, completeness = 97.6%, R sig = 2.02%) and 3062 (92.73%) were greater than 2σ(F 2 ). The final cell constants of a = 6.9404(2) Å, b = 12.2651(4) Å, c = 22.3938(10) Å, β = 98.330(2) • and volume = 1886.15(12) Å 3 were based upon the refinement of the XYZ-centroids of reflections above 20 σ(I). Data were corrected for absorption effects using the multi-scan method (SADABS). The calculated minimum and maximum transmission coefficients (based on crystal size) are 0.7920 and 0.9330. The structure was solved and refined using the Bruker SHELXTL-2014/7 version Software Package, using the space group P2 1 /c, with Z = 4 for the formula unit, C 21 H 27 NO 4 . The final anisotropic full-matrix least-squares refinement on F 2 with 241 variables converged at R1 = 3.98%, for the observed data and wR2 = 9.96% for all data. The goodness-of-fit was 1.054. The largest peak in the final difference electron density synthesis was 0.198 e − /Å 3 and the largest hole was −0.192 e − /Å 3 with an RMS deviation of 0.034 e − /Å 3 . On the basis of the final model, the calculated density was 1.259 g/cm 3 and F(000), 768 e − . The hydrogen at N1 atom was refined freely. CCDC number: 2073466.

X-ray Crystal Structure Analysis of anti-21
A colorless prism-like specimen of C 28 H 34 N 2 O 3 , approximate dimensions 0.060 mm × 0.240 mm × 0.260 mm, was used for the X-ray crystallographic analysis. The Xray intensity data were measured. The integration of the data using a monoclinic unit cell yielded a total of 4159 reflections to a maximum θ angle of 67.08 • (0.84 Å resolution), of which 4159 were independent (average redundancy 1.000, completeness = 96.3%, R sig = 2.90%) and 3509 (84.37%) were greater than 2σ(F 2 ). The final cell constants of a = 8.5091(4) Å, b = 15.9061(5) Å, c = 17.8829(6) Å, β = 94.139(3) • and volume = 2414.08(16) Å 3 were based upon the refinement of the XYZ-centroids of reflections above 20 σ(I). Data were corrected for absorption effects using the multiscan method (SADABS). The calculated minimum and maximum transmission coefficients (based on crystal size) are 0.8530 and 0.9630. The structure was solved and refined using the Bruker SHELXTL Software Package, using the space group P2 1 /n, with Z = 4 for the formula unit, C 28 H 34 N 2 O 3 . The final anisotropic full-matrix least-squares refinement on F 2 with 304 variables converged at R1 = 7.05%, for the observed data and wR2 = 20.59% for all data. The goodness-of-fit was 1.041. The largest peak in the final difference electron density synthesis was 0.482 e − /Å 3 and the largest hole was −0.353 e − /Å 3 with an RMS deviation of 0.056 e − /Å 3 . On the basis of the final model, the calculated density was 1.229 g/cm 3 and F(000), 960 e − . The hydrogen at N2 atom was refined freely. CCDC number: 2073467. 5.5.4. X-ray Crystal Structure Analysis of syn,anti-22 A colorless plate-like specimen of C 28 H 36 N 2 O 3 , approximate dimensions 0.150 mm × 0.170 mm × 0.270 mm, was used for the X-ray crystallographic analysis. The X-ray intensity data were measured. The integration of the data using a monoclinic unit cell yielded a total of 4152 reflections to a maximum θ angle of 67.18 • (0.84 Å resolution), of which 4152 were independent (average redundancy 1.000, completeness = 96.1%, R sig = 2.57%) and 3721 (89.62%) were greater than 2σ(F 2 ). The final cell constants of a = 7.5886(3) Å, b = 16.0072(4) Å, c = 20.2378(5) Å, β = 100.377(2) • and volume = 2418.12(13) Å 3 were based upon the refinement of the XYZ-centroids of reflections above 20 σ(I). Data were corrected for absorption effects using the multi-scan method (SADABS). The calculated minimum and maximum transmission coefficients (based on crystal size) were 0.8490 and 0.9120. The structure was solved and refined using the Bruker SHELXTL Software Package, using the space group P2 1 /c, with Z = 4 for the formula unit, C 28 H 36 N 2 O 3 . The final anisotropic full-matrix least-squares refinement on F 2 with 307 variables converged at R1 = 4.36%, for the observed data and wR2 = 11.81% for all data. The goodness-of-fit was 1.031. The largest peak in the final difference electron density synthesis was 0.175 e − /Å 3 and the largest hole was −0.164 e − /Å 3 with an RMS deviation of 0.032 e − /Å 3 . On the basis of the final model, the calculated density was 1.232 g/cm 3 and F(000), 968 e − . The hydrogens at N2 and O1 atoms were refined freely. CCDC number: 2073468.

X-ray Crystal Structure Analysis of anti,syn-22
A colorless needle-like specimen of C 28 H 36 N 2 O 3 , approximate dimensions 0.020 mm × 0.070 mm × 0.270 mm, was used for the X-ray crystallographic analysis. The X-ray intensity data were measured. The integration of the data using a monoclinic unit cell yielded a total of 4164 reflections to a maximum θ angle of 66.92 • (0.84 Å resolution), of which 4164 were independent (average redundancy 1.000, completeness = 95.5%, R sig = 3.04%) and 3332 (80.02%) were greater than 2σ(F 2 ). The final cell constants of a = 7.9714(2) Å, b = 27.7657(10) Å, c = 11.1029(5) Å, β = 95.285(3) • and volume = 2446.97(15) Å 3 were based upon the refinement of the XYZ-centroids of reflections above 20 σ(I). Data were corrected for absorption effects using the multiscan method (SADABS). The calculated minimum and maximum transmission coefficients (based on crystal size) are 0.8500 and 0.9880. The structure was solved and refined using the Bruker SHELXTL-2014/7 version Software Package, using the space group P2 1 /c, with Z = 4 for the formula unit, C 28 H 36 N 2 O 3 . The final anisotropic full-matrix least-squares refinement on F 2 with 308 variables converged at R1 = 5.00%, for the observed data and wR2 = 13.22% for all data. The goodness-of-fit was 1.031. The largest peak in the final difference electron density synthesis was 0.131 e − /Å 3 and the largest hole was −0.239 e − /Å 3 with an RMS deviation of 0.044 e − /Å 3 . On the basis of the final model, the calculated density was 1.218 g/cm 3 and F(000), 968 e − . The hydrogens at N2 and O1 atoms were refined freely. CCDC number: 2073469.

X-ray Crystal Structure Analysis of syn-25
A colorless prism-like specimen of C 19 H 22 F 2 N 2 O 2 , approximate dimensions 0.100 mm × 0.140 mm × 0.180 mm, was used for the X-ray crystallographic analysis. The X-ray intensity data were measured. The integration of the data using a triclinic unit cell yielded a total of 9040 reflections to a maximum θ angle of 67.31 • (0.84 Å resolution), of which 9040 were independent (average redundancy 1.000, completeness = 96.6%, R sig = 2.87%) and 7415 (82.02%) were greater than 2σ(F 2 ). The final cell constants of a = 12.9305(5) Å, b = 13.0504(5) Å, c = 17.0932(4) Å, α = 112.037(2) • , β = 97.638(2) • , γ = 97.0010(10) • and volume = 2603.27(16) Å 3 were based upon the refinement of the XYZ-centroids of reflections above 20 σ(I). Data were corrected for absorption effects using the multiscan method (SADABS). The calculated minimum and maximum transmission coefficients (based on crystal size) were 0.8630 and 0.9200. The structure was solved and refined using the Bruker SHELXTL Software Package, using the space group P-1, with Z = 6 for the formula unit, C 19 H 22 F 2 N 2 O 2 . The final anisotropic full-matrix least-squares refinement on F 2 with 928 variables converged at R1 = 6.16%, for the observed data and wR2 = 18.77% for all data. The goodness-of-fit was 1.038. The largest peak in the final difference electron density synthesis was 1.261 e − /Å 3 and the largest hole was −0.361 e − /Å 3 with an RMS deviation of 0.044 e − /Å 3 . On the basis of the final model, the calculated density was 1.333 g/cm 3 and F(000), 1104 e − . The hydrogens at N1A, N2A, N1B, N2B, N1C and N2C atoms were refined freely. CCDC number: 2073470.

Receptor Binding Studies
The affinity towards σ 1 and σ 2 receptors was recorded according to the procedures given in the Supplementary Materials and ref [45][46][47].
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/ijms22115685/s1, The Supplementary Materials contain purity data of all prepared compounds, experimental procedures of receptor binding studies, details of the X-ray crystal structure analyses, molecular dynamics simulations of compound 18 displayed in Figure S1, 1 H and 13 C NMR spectra including some 2D NMR spectra of prepared compounds and selected HPLC traces.