Forcing the phenyl moiety into the axial position by embedding the 2-phenyl-1,3-dioxane system in a tricyclic benzomorphan scaffold: design, synthesis and biological evaluation

Abstract

The relative configuration and the substitution pattern control the interaction of 2-(2-phenyl1,3-dioxan-4-yl)ethan-1-amines with σ₁ receptors or the PCP binding site of NMDA receptors. In order to investigate the influence of the orientation of the phenyl moiety in 2-position on the receptor interaction, the 2-phenyl-1,3-dioxane system was embedded in a tricyclic benzomorphan scaffold (3) fixing the phenyl moiety in an axial orientation relative to the 1,3-dioxane ring. The key step of the synthesis of tricyclic amines 3 was the addition of lithiated 2-methylbenzamide 7 at pentanone 6 to afford the tertiary alcohol 8. Lactone formation (9), DIBAH reduction (10) and intramolecular transacetalization led to the tricyclic alcohol 11, which was converted into a series of twelve primary, secondary and tertiary amines 3a–m. Although the primary amine 3a is structurally related to the potent PCP antagonist 2a, it did not interact with the PCP binding site of the NMDA receptor. The missing ethyl moiety and/or an unfavorable orientation of the phenyl moiety might be responsible for the lost PCP affinity of 3a. As observed for the flexible 1,3-dioxanes 1b and 2b, introduction of a benzyl moiety at the amino group resulted in high σ₁ receptor affinity of 3b. In accordance with σ₁ pharmacophore models, two small or two large substituents at the amino moiety were less tolerated by the σ₁ receptor, whereas an additional small methyl moiety increased the σ₁ affinity of 3h and 3j. With respect to σ₁ receptor affinity and selectivity over the σ₂ subtype, the methylated cyclohexylmethylamine 3j (K_i(σ₁) = 6.4 nM, 9-fold selectivity) represents the most promising ligand. The highest ligand-lipophilicity efficiency (LLE) was obtained for the secondary cyclohexylmethylamine 3d (LLE = 6.7). However, the highest metabolic stability (phase I metabolism) was determined for the benzylamine 3b (89% intact after incubation for 90 min).

---

1. Introduction

This project deals with target hopping, i.e., a switch from open channel blockers of N-methyl-D-aspartate (NMDA) receptors to σ₁ receptor ligands.

The heterotetrameric NMDA receptor belongs to the class of ligand-gated ion channels. The pore of the NMDA receptor contains binding sites for Mg²⁺ ions and phencyclidine (1-(1-phenylcyclohexyl)piperidine, PCP, Fig. 1).^1–3 Addressing the PCP binding site with ligands leads to inhibition of the ion flux through the open channel.^1–3 NMDA receptor inhibitors interacting with the PCP binding site are termed open-channel blockers, as the ion channel has to be opened by the agonists glutamate and glycine before the ligands can enter the channel pore and reach their binding site. Very potent open-channel blockers, such as phencyclidine, MK-801, and dexoxadrol (Fig. 1)^1–3 are associated with desired analgesic and anesthetic activity, but also with strong undesired psychotomimetic and hallucinogenic effects. However, some low affinity open-channel blockers are clinically used as dissociative anesthetic (ketamine) and antitussive (dextromethorphan) as well as for the treatment of Parkinson's (amantadine) and Alzheimer's disease (memantine).^1–3 The structure of the NMDA receptor has been solved by X-ray crystal structure analysis.^4,5

Fig. 1 Prototypical ligands interacting with the PCP binding site of the NMDA receptor and/or with σ receptors.

The σ₁ receptor is a chaperone predominantly located in the membrane of the endoplasmic reticulum controlling Ca²⁺ flux. It is involved in neurodegenerative and neuropsychiatric diseases such as schizophrenia, depression and addiction as well as Alzheimer's diseases. Moreover, inhibition of the σ₁ receptor can be exploited for the treatment of neuropathic pain and cancer.^6–8 In the forced swimming test, σ₁ receptor knock-out mice showed a depression-like behavior.^9,10 In 2016, the σ₁ receptor was crystallized for the first time exhibiting an unexpected structure with only one transmembrane helix.¹¹ Two years later, structures of the σ₁ receptor in complex with prototypical ligands haloperidol and (+)-pentazocine were reported.¹²

σ receptors and the PCP binding site of NMDA receptors are historically related. After refuting the hypothesis that the σ receptor is an opioid receptor subtype,¹³ it was postulated that the PCP binding site and the σ receptor are identical.¹⁴ This hypothesis arose from the observation that prototypical open channel blockers of the NMDA receptor, e.g., phencyclidine, interacted with high affinity with σ receptors as well. Moreover, prototypical σ ligands such as racemic SKF-10 047 (rac-N-allylnormetazocine, Fig. 1) could also inhibit the NMDA receptor by interacting with the PCP binding site.¹⁴ However, identification of ligands inhibiting selectively the NMDA receptor by interaction with the PCP binding site without addressing the σ receptor (e.g., dexoxadrol) and ligands interacting selectively with σ₁ receptors but not with the PCP binding site (e.g., haloperidol, Fig. 1) led to a clear differentiation of σ receptors and the PCP binding site. Moreover, the distribution patterns of both proteins in the central nervous system are quite different.¹⁵

The confusion regarding binding affinity of benzomorphans originated from their absolute configuration: dextrorotatory (S,S,S)-configured N-allyl-normetazocine ((+)-SKF-10 047) shows high affinity towards the σ₁ receptor subtype, whereas the levorotatory (R,R,R)-configured enantiomer (−)-SKF-10 047 inhibits the NMDA receptor associated cation channel. Small N-substituents (e.g., H or CH₃) of (R,R,R)-(−)-benzomorphans increase the PCP binding affinity and decrease the affinity towards opioid receptors.¹⁶

Derived from the PCP ligands dexoxadrol and etoxadrol 2,4-disubstituted 1,3-dioxanes 1 and 2 have been developed.^17,18 Depending on the substitution pattern and the orientation of the phenyl moiety in 2-position, 1 and 2 showed high affinity and selectivity either for the σ₁ receptor or the PCP binding site of the NMDA receptor. Whereas the benzaldehyde-derived primary amine 1a did neither interact with σ₁ receptors nor the PCP binding site, the analogous benzylamine 1b revealed high σ₁ receptor affinity and selectivity (>170-fold) over the PCP binding site. The propiophenone-derived 1,3-dioxane 2a with a primary amino moiety showed strong interaction with the PCP binding site and high selectivity over the σ₁ receptor (>750-fold). Introduction of a benzyl moiety at the primary amine of 2a led to the secondary amine 2b with high σ₁ receptor affinity and high selectivity over the PCP binding site (>500-fold)¹⁷ (Fig. 2).

Fig. 2 Design of conformationally restricted 1,3-dioxanes 3 with the phenyl ring fixed in axial orientation.

Herein, we report the synthesis and biological evaluation of conformationally restricted analogs of 1 and 2, in which the 2-phenyl-1,3-dioxane system is embedded in a tricyclic benzomorphan scaffold. In the tricyclic acetals 3 the phenyl moiety in “2-position” of the 1,3-dioxane ring is forced to adopt the axial orientation. This axial orientation of the phenyl moiety is not possible for benzaldehyde-derived acetals 1, since the phenyl moiety always adopts the equatorial orientation due to thermodynamic reasons. On the other hand, the “axial orientation” (related to the 1,3-dioxane ring) of the phenyl ring in tricyclic compounds 3 corresponds to the axial orientation in the propiophenone-derived 1,3-dioxanes 2. However, the ethyl moiety at the acetalic center of 2 is missing in the tricyclic acetals 3. The conformationally restricted acetals 3 should provide relationships between structural modifications and affinity towards σ receptors and NMDA receptors. In particular, insights should be gained about the relevance of the orientation of the phenyl moiety and the presence of an additional ethyl substituent in 2-position of the 1,3-dioxane ring. Variation of the substituents at the amnio moiety should further broaden this SAR study.

2. Results and discussion

2.1. Synthesis

The synthesis of the conformationally restricted 1,3-dioxanes 3 started with pentane-1,3,5-triol (4).¹⁹ The primary alcoholic groups of 4 were selectively protected by the tert-butyldimethylsilyl (TBS) protective group to afford the bis-silyl ether 5. Subsequently, the remaining secondary alcohol 5 was oxidized with Dess Martin Periodinane (DMP)²⁰ to give the ketone 6 in 97% yield as key intermediate of this synthesis (Scheme 1).

Scheme 1 Synthesis of conformationally restricted 1,3-dioxanes 3 with axially oriented phenyl moiety in “2-position”. Reagents and reaction conditions: (a) TBS-Cl, imidazole, DMF, rt, 6 h, 71%. (b) Dess Martin Periodinane (DMP), CH₂Cl₂, rt, 4.5 h, 97%. (c) sec-BuLi, THF, −78 °C, 1.5 h; then addition of ketone 6, −50 °C, 2.5 h, 92%. (d) Neat, 186 °C, 16 h, 82%. (e) DIBAH, CH₂Cl₂, −78 °C, 70 min, 93%. (f) HCl, THF, rt, 16 h, 99%. (g) Zn(M₃)₂·2 pyridine, PPh₃, DIAD, THF, rt, 19 h, 64%. (h) H₂ (balloon), Pd/C, EtOAc, rt, 16 h, 75%. (i) RCH=O, NaBH(OAc)₃, CH₂Cl₂, rt. (j) O=HCCH₂CH₂CH₂CH=O, NaBH(OAc)₃, CH₂Cl₂, rt, 18 h, 32%. (k) Formalin, NaBH(OAc)₃, CH₂Cl₂, rt, 18 h, 97% (3h), 87% (3j).

The next step, i.e., the nucleophilic addition of 2-methylbenzamide 7 to the ketone 6 represents the key step of the synthesis. For this purpose, benzamide 7 was deprotonated with sec-BuLi at the CH₃ moiety and the resulting methyllithium species reacted with the ketone 6 to provide the tertiary alcohol 8 in 92% yield. Heating of the hydroxy amide 8 without solvent at 186 °C for 16 h led to the δ-lactone 9, which was reduced with diisobutylaluminum hydride (DIBAH)²¹ to provide the cyclic hemi acetal 10 in 94% yield. Treatment of the hemi acetal 10 with diluted HCl led to cleavage of both silyl ethers, but only one primary alcohol could react with the hemi acetal to form the tricyclic acetal 11. During this reaction step, the symmetry of the molecules was lost leading to chiral compounds (Scheme 1).

After establishment of the tricyclic system 11, the remaining primary alcohol was converted into various amines 3. At first, a Mitsunobu reaction with Zn(N₃)₂·2 pyridine in the presence of PPh₃ and diisopropyl azodicarboxylate (DIAD) transformed the primary alcohol 11 into azide 12, which was reduced with H₂ and Pd/C to afford the primary amine 3a. Finally, reductive alkylation of the primary amine 3a with aldehydes and NaBH(OAc)₃ ²² provided secondary and tertiary amines 3b–m. Stoichiometric amounts of aldehyde led to secondary amines (e.g.; 3d, 3f, 3g), intermediate amounts of benzaldehyde provided a mixture of secondary and tertiary amines 3b and 3c, whereas an excess of aldehyde completely alkylated the primary amine 3a affording tertiary amines 3e, 3k nd 3l. The piperidine ring of 3m was established by reductive alkylation of the primary amine 3a with glutaraldehyde and NaBH(OAc)₃.²² Reductive methylation with formaldehyde and NaBH(OAc)₃ converted the secondary amines 3b and 3d into tertiary methylamines 3h and 3j, respectively.

2.2. Receptor affinity

The affinity of the amines 3a–m towards the PCP binding site of the NMDA receptor and towards σ₁ and σ₂ receptors was determined in radioligand receptor binding studies. In the PCP assay, pig brain cortex was used as receptor material and [³H](+)-MK-801 as radioligand.^17,23 [³H](+)-pentazocine and [³H]di-o-tolylguanidine served as radioligands and guinea pig brain and rat liver as receptor material in the σ₁ and σ₂ receptor assay, respectively.^24,25 The recorded affinities of amines 3 are summarized in Table 1 together with the K_i values of some reference compounds.

Table 1 PCP, σ₁ and σ₂ receptor affinities of tricyclic amines 3 and reference compounds

Entry	Compd	NR2	Ki ± SEM^a^,^b (nM)
			PCP	σ1	σ2
a Ki values represent the mean of at least three independent experiments.b For ligands with low affinity, the Ki value was determined only once.
1	3a	NH2	>1000	>1000	>1000
2	3b	NHBn	>1000	69 ± 9	60 ± 23
3	3c	NBn2	>1000	>1000	>1000
4	3d	NHCH2C6H11	>1000	15 ± 4	26 ± 16
5	3e	N(CH2C6H11)2	>1000	280	375
6	3f	NHCH2C6H4(4-F)	>1000	47 ± 29	30 ± 11
7	3g	NHCH2CH2CH2Ph	>1000	242	62 ± 8
8	3h	N(CH3)Bn	>1000	15 ± 3	137
9	3j	N(CH3)CH2C6H11	>1000	6.4 ± 2.0	55 ± 11
10	3k	N(CH3)2	>1000	>1000	>1000
11	3l	N(C4H9)2	>1000	3410	298
12	3m	N(CH2)5	>1000	766	381
13	Dexoxadrol		25 ± 4	—	—
14	Phencyclidine		59 ± 12	—	—
15	(+)-Pentazocine		—	5.7 ± 2.2	—
16	Haloperidol		—	6.3 ± 1.6	78 ± 2.3
17	Di-o-tolylguanidine		—	89 ± 29	58 ± 18

---

The primary amine 3a did not show high affinity for the PCP binding site of the NMDA receptor. The structurally related lead compound 2a exhibited a strong interaction with the PCP binding site of the NMDA receptor (K_i = 13 nM, see Fig. 1). Both compounds 2a and 3a are primary amines with an axially oriented phenyl moiety in 2-position of the 1,3-dioxane ring. However, 2a has an additional ethyl moiety in 2-position of the 1,3-dioxane ring, which is not present in the tricyclic amine 3a. The missing ethyl moiety or the slightly different orientation of the axially oriented phenyl moiety in 2-postion of the 1,3-dioxnae ring might be responsible for the complete loss of PCP affinity of primary amine 3a.

Introduction of various substituents at the amino moiety did not lead to PCP affinity of 3b–m. This observation is in good accordance with literature describing the highest PCP affinity for primary amines for these types of ligands.¹⁷

In agreement with literature (see Fig. 1), the primary amine 3a did not bind at σ₁ and σ₂ receptors. However, as shown for the lead compounds 1b and 2b, the introduction of a benzyl moiety increased the σ₁ affinity of 3b (K_i = 69 nM) considerably. A cyclohexylmethyl moiety instead of the benzyl moiety increased the σ₁ affinity to 15 nM (3d). A further increase of the σ₁ affinity could be obtained by methylation of the secondary amines 3b and 3d: the analogous methylamines 3h (K_i = 15 nM) and 3j (K_i = 6.4 nM) revealed the highest σ₁ affinity of this class of compounds. Two large (3c, 3e) or two small substituents (3k. 3m) at the amino moiety did not lead to high σ₁ affinity. A larger alkyl chain between the N-atom and the terminal phenyl moiety reduced the σ₁ affinity of the phenylpropyl derivative 3g (K_i = 242 nM).

The secondary amines 3b, 3d, and 3f show very similar affinities towards both σ₁ and σ₂ receptors. However, selectivity for the σ₁ receptor was gained by introduction of an additional CH₃ moiety at the amino group. The methylated benzylamine 3h and the methylated cyclohexylmethylamine 3j displayed 9-fold selectivity for the σ₁ receptor over the σ₂ subtype.

2.3. Physicochemical and pharmacokinetic properties of selected ligands

In order to get an idea about the lipophilicity of the tricyclic compounds, the log D_7.4 value of selected ligands was determined using the micro-shake flask method.^26,27 In brief, the compounds were distributed between n-octanol and MOPS buffer pH 7.4 layers and the amount of the compound in the buffer layer was determined by LC-MS.

The recorded log D_7.4 values are summarized in Table 2. The benzylamine 3b and the cyclohexylmethylamine 3d show the same log D_7.4 value of 1.1. The additional methyl moiety at the amino group of 3j increased the log D_7.4 value by approx. one log unit. The amount of the primary amine 3a in the aqueous layer was higher than its amount in the n-octanol layer leading to a negative log D_7.4 value of −1.9 (Table 2).

Table 2 Physicochemical and pharmacokinetic properties including ligand-lipophilicity efficiency (LLE) of selected σ₁ ligands

Compd	NR2	−log Ki(σ1)	log D7.4 ± SD (n = 3)	LLE^a	Plasma protein binding^b	Metabolic stability ± SD (n = 3)^c
a LLE (ligand-lipophilicity efficiency) = −log Ki − log D7.4.b Interaction with human serum albumin was recorded by HPAC analysis.c Amount (in %) of parent compound after incubation with mouse liver microsomes and NADPH for 90 min.d Mean of two values.e Imipramine was used as reference compound to prove the activity of the microsomes and NADPH in the metabolic experiments.
3a	NH2	<6	−1.9^d		42 ± 0.2	87 ± 4
3b	NHBn	7.2	1.1 ± 0.04	6.1	75 ± 1	89 ± 6
3d	NHCH2C6H11	7.8	1.1 ± 0.05	6.7	77 ± 1	32 ± 4
3j	N(CH3)CH2C6H11	8.2	2.0 ± 0.1	6.2	78 ± 1	45 ± 4
Imipramine^e						14 ± 1

---

The recorded log D_7.4 values were used to calculate the ligand-lipophilicity efficiency (LLE), which modulates the biological activity of a ligand by its lipophilicity (LLE = −log K_i − log D_7.4).^28,29 Due to its increased σ₁ affinity, the LLE value of the cyclohexylmethyl derivative 3d (LLE = 6.7) is higher than the LLE value of the benzyl derivative 3b (LLE = 6.1). However, the increased σ₁ affinity of the methylated derivative 3j is compensated by its increased lipophilicity resulting in a reduced LLE value of 6.2. The LLE value for the primary amine 3a was not calculated, since 3a did not show any interactions with the σ₁ receptor even at a concentration of 1 µM (Table 2).

The plasma protein binding of the amines 3a, 3b, 3d and 3j was determined by high performance affinity chromatography (HPAC) using a stationary phase coated with human serum albumin. The retention time correlates with the affinity towards human serum albumin, which represents the main component of human blood.^30,31 The plasma protein binding of the secondary amines 3b and 3d and the tertiary amine 3j is very similar (75–78%). However, the very polar primary amine 3a exhibited a considerably lower binding to human serum albumin of 42% (Table 2).

In order to investigate the phase I metabolic stability, selected amines were incubated with mouse liver microsomes and NADPH at 37 °C for 90 min. The amount of the residual parent compound was determined by LC-MS.^26,32 The benzylamine 3b revealed high metabolic stability, since 89% of the parent compound were intact after 90 min. The cyclohexylmethylamines 3d and 3j were faster metabolized and only 32% and 45% remained unchanged after 90 min, respectively.

3. Conclusion

The 1,3-dioxane 2a bearing an axially oriented phenyl moiety at 2-position and a primary amino moiety at the ethyl substituent in 4-position showed high affinity towards the PCP binding sit of the NMDA receptor (K_i = 13 nM), but no affinity at σ₁ and σ₂ receptors (Fig. 1). Although structurally related to 2a, the conformationally restricted 1,3-dioxane 3a with a primary amino moiety at the ethyl substituent did not interact with the PCP binding site. We hypothesize that the missing 2-ethyl moiety of 2a and/or an unfavorable orientation of the conformationally restricted phenyl moiety are the reason for the reduced PCP affinity of the primary amine 3a.

Independent of the orientation of the phenyl moiety in 2-position, both 1,3-dioxanes 1b and 2b with a benzylamino moiety displayed high σ₁ affinity (Fig. 1). The same effect was observed for the benzylamine 3b showing also high σ₁ affinity (K_i = 69 nM).

The results obtained for this novel class of σ₁ ligands correlate well with pharmacophore models for σ₁ receptor ligands. According to these models, potent σ₁ ligands should contain a central amino moiety substituted with two lipophilic substituents.^33–35 The introduction of an additional large substituent at the amino group as in 3c and 3e led to considerably reduced σ₁ affinity. On the other hand, two small substituents at the amino moiety (e.g., compounds 3k, 3l, 3m) are not sufficient to bind with high affinity at σ₁ receptors. However, a small methyl moiety in addition to the large benzyl or cyclohexylmethyl moiety at the amino group resulted in increased σ₁ affinity of 3h and 3j.

The highest σ₁ affinity was observed for the cyclohexylmethyl substituted amines 3d (K_i = 15 nM) and 3j (K_i = 6.4 nM). Due to their low lipophilicity (log D_7.4 = 1.1 and 2.0, respectively), both compounds exhibit high LLE values of 6.7 and 6.2, respectively. However, in the presence of mouse liver microsomes and NMDPH (phase I metabolism), both amines 3d and 3j were rapidly metabolized. In contrast, the less potent benzylamine 3b (K_i(σ₁) = 69 nM, LLE = 6.1) revealed high metabolic stability as 89% of the parent compound remained unchanged upon incubation with liver microsomes and NADPH over 90 min.

In conclusion, the methylated cyclohexylmethylamine 3j showed the highest σ₁ affinity (K_i = 6.4 nM) and the highest selectivity over the σ₂ subtype (9-fold). The secondary cyclohexylmethylamine 3d exhibited the highest LLE value (6.7) and the benzylamine 3b was the metabolically most stable σ₁ ligand of this series (89% intact after 90 min).

4. Experimental

4.1. Synthetic procedures

4.1.1. 1,5-Bis(tert-butyldimethylsilyloxy)pentan-3-ol (5).
Pentane-1,3,5-triol¹⁹ (4, 3 g, 25 mmol), tert-butyldimethylsilyl chloride (7.5 g, 50 mmol) and imidazole (10.2 g, 150 mmol) were dissolved in DMF (30 mL) in a Schlenk flask under N₂ atmosphere. The reaction mixture was stirred at rt for 6 h. After addition of water (50 mL) the organic layer was extracted with ethyl acetate (4 × 50 mL). The organic layer was then washed with brine (100 mL), dried (Na₂SO₄), concentrated in vacuo and the residue was purified by fc (∅ = 8 cm, h = 19.5 cm, ethyl acetate : cyclohexane = 0.2 : 9.8, V = 65 mL, R_f = 0.08 (ethyl acetate : cyclohexane = 0.4 : 9.6)). Colorless oil, yield 6.1 g (71%). C₁₇H₄₀O₃Si₂ (348.3). ¹H NMR (CDCl₃): δ [ppm] = 0.07 (s, 12H, 2 × Si(CH₃)₂), 0.89 (s, 18H, 2 × SiC(CH₃)₃), 1.63–1.75 (m, 4H, CH₂CH₂OSi), 3.65 (d, J = 2.3 Hz, 1H, OH), 3.76–3.87 (m, 4H, CH₂OSi), 3.95–4.02 (m, 1H, CH). IR: v [cm⁻¹] = 3524 (b, ν, O–H), 2954 (w, ν, C–H, alkyl). MS (EI): m/z = 349 [M + H]⁺. Elemental analysis: calcd C 58.56, H 11.56; found C 58.54, H 11.80.

4.1.2. 1,5-Bis(tert-butyldimethylsilyloxy)pentan-3-one (6).
Under N₂ atmosphere, a solution of alcohol 5 (2.2 g, 6.4 mmol) in CH₂Cl₂ (90 mL) was added to a solution of Dess Martin Periodinane (3.3 g, 7.7 mmol) in CH₂Cl₂ (75 mL). The reaction mixture was stirred for 4.5 h at rt. Et₂O (150 mL) and 1 M NaOH (63 mL) were added and the mixture was stirred for another 15 min. The Et₂O layer was separated and washed with NaOH (100 mL) and water (100 mL) followed by extraction of the combined aqueous layers with Et₂O (4 × 200 mL). Finally, the organic layer was dried (Na₂SO₄), the solvent was removed in vacuo and the residue was purified by fc (∅ = 6 cm, h = 15.5 cm, ethyl acetate : cyclohexane = 0.3 : 9.7, V = 30 mL, R_f = 0.17). Colorless oil, yield 2.2 g (97%). C₁₇H₃₈O₃Si₂ (346.2). ¹H NMR (CDCl₃): δ [ppm] = 0.05 (s, 12H, 2 × Si(CH₃)₂), 0.87 (s, 18H, 2 × SiC(CH₃)₃), 2.65 (t, J = 6.4 Hz, 4H, CH₂CH₂OSi), 3.88 (t, J = 6.4 Hz, 4H, CH₂OSi). IR: v [cm⁻¹] = 2955, 2857 (w, ν, C–H, alkyl), 1715 (s, ν, C=O). MS (EI): m/z = 347 [M + H]⁺, 289 [M − C(CH₃)₃]⁺. Elemental analysis: calcd C 58.90, H 11.05; found C 59.28, H 11.34.

4.1.3. 2-{4-(tert-Butyldimethylsilyloxy)-2-[2-(tert-butyldimethylsilyloxy)ethyl]-2-hydroxybutan-1-yl}-N,N-diethylbenzamide (8).
Under N₂ atmosphere diethylamide 7 (368 mg, 1.92 mmol) was dissolved in abs. THF (125 mL) and cooled down to −78 °C. Then, 1.3 M sec-butyllithium in cyclohexane/hexane = 92 : 8 (2.0 mL, 2.12 mmol) was added dropwise to the solution and the color changed from colorless to orange. After 1.5 h ketone 6 (1.0 g, 2.9 mmol) dissolved in THF (5 mL) was added using a syringe pump over 10 min. The reaction mixture turned yellow and was allowed to warm to −50 °C during 2.5 h. Water (50 mL) was added at −50 °C and after addition of brine (50 mL), the aqueous layer was extracted with chloroform (4 × 100 mL). The organic layer was dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 6 cm, h = 21 cm, ethyl acetate : cyclohexane = 2 : 8, V = 10 mL, R_f = 0.42). Colorless oil, yield 949 mg (92%). C₂₉H₅₅NO₄Si₂ (537.4). ¹H NMR (CDCl₃): δ [ppm] = 0.05 (s, 12H, 2 × Si(CH₃)₂), 0.88 (s, 18H, 2 × SiC(CH₃)₃), 1.04 (t, J = 7.1 Hz, 3H, CH₂CH₃), 1.26 (t, J = 7.1 Hz, 3H, CH₂CH₃), 1.56–1.80 (m, 4H, CH₂CH₂OSi), 2.55–2.74 (m, 1H, CH₂OSi), 2.86–3.05 (m, 1H, CH₂OSi), 3.09–3.14 (m, 2H, CH₂CH₃), 3.31–3.51 (m, 1H, CH₂OSi), 3.58–3.75 (m, 1H, CH₂OSi), 3.76–3.92 (m, 4H, CH₂CH₃ and CH₂Ph), 4.79 (s, 1H, OH), 7.16 (dd, J = 7.6/1.4 Hz, 1H, CH_arom.), 7.22 (td, J = 7.5/1.3 Hz, 1H, CH_arom.), 7.31 (td, J = 7.6/1.6 Hz, 1H, CH_arom.), 7.53 (dd, J = 7.7/0.9 Hz, 1H, CH_arom.). ¹³C NMR (CDCl₃): δ [ppm] = −5.2 (6C, Si(CH₃)₃), 13.0 (1C, CH₂CH₃), 14.1 (1C, CH₂CH₃), 26.1 (6C, SiC(CH₃)₃), 39.3 (2C, OCH₂CH₂), 41.6 (1C, OCH₂), 43.0 (1C, OCH₂), 43.3 (1C, CH₂CH₃), 60.1 (1C, CH₂Ph), 60.2 (1C, CH₂CH₃), 73.5 (1C, Cq), 125.7 (1C, CH_arom), 126.1 (1C, CH_arom), 128.4 (1C, CH_arom), 132.2 (1C, CH_arom), 135.3 (1C, Cq_arom), 137.5 (1C, Cq_arom), 172.0 (1C, C=O). IR: v [cm⁻¹] = 2951 (s, ν, C–H, alkyl), 1613 (s, ν, C=O), 772 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₉H₅₅NO₄Si₂H [M + H] 538.3742, found 538.3754. HPLC: Compound 8 was not stable during HPLC analysis.

4.1.4. 3,3-Bis-[2-(tert-butyldimethylsilyloxy)ethyl]-3,4-dihydro-2-benzopyran-1-one (9).
Diethylamide 8 (1.0 g, 1.93 mmol) was heated in a silicone bath to 186 °C overnight. The residue was purified by fc (∅ = 4 cm, h = 17.5 cm, ethyl acetate : cyclohexane = 5 : 95, V = 20 mL, R_f = 0.6). Colorless oil, yield 736 mg (82%). C₂₅H₄₄O₄Si₂ (464.3). ¹H NMR (CDCl₃): δ [ppm] = 0.02, 0.03 (2s, 12H, Si(CH₃)₂), 0.87 (s, 18H, SiC(CH₃)₃), 2.00 (t, J = 6.5 Hz, 4H, CH₂CH₂OSi), 3.21 (s, 2H, ArCH₂), 3.79 (td, J = 6.4/1.8 Hz, 4H, CH₂OSi), 7.19 (d, J = 7.5 Hz, 1H, 5-CH_arom), 7.36 (td, J = 7.7/1.1 Hz, 1H, 7-CH_arom), 7.52 (td, J = 7.5/1.4 Hz, 1H, 6-CH_arom), 8.07 (dd, J = 7.7/1.1 Hz, 1H, 8-CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = −5.3 (2C, Si(CH₃)₂), −5.3 (2C, Si(CH₃)₂), 18.4 (2C, C(CH₃)₃), 26.1 (6C, SiC(CH₃)₃), 36.9 (1C, CH₂Ph), 40.3 (2C, CH₂CH₂OSi), 58.9 (2C, CH₂OSi), 84.0 (1C, C-3), 125.2 (1C, CH_arom), 127.5 (1C, CH_arom), 128.3 (1C, CH_arom), 130.0 (1C, CH_arom), 134.0 (1C, Cq_arom), 138.5 (1C, Cq_arom), 165.1 (1C, C=O). IR: v [cm⁻¹] = 2951 (s, ν, C–H, alkyl), 1720 (s, ν, C=O), 752 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₅H₄₄O₄Si₂H [M + H] 465.2864, found 465.2848, calcd for C₂₅H₄₄O₄Si₂ − C(CH₃)₃ + H [M − C(CH₃) + H] 407.2074, found 407.2051. Purity (HPLC): 97.1%, t_R = 28.81 min.

4.1.5. 3,3-Bis[2-(tert-butyldimethylsilyloxy)ethyl]-3,4-dihydro-1H-2-benzopyran-1-ol (10).
Lactone 9 was dissolved in CH₂Cl₂ (33 mL) and cooled down to −78 °C. 1M DIBAH solution in toluene (1.7 mL, 1.7 mmol) was added and the reaction mixture was stirred at −78 °C for 1 h and 10 min. After addition of saturated NaHCO₃ solution (20 mL) at −78 °C and brine (10 mL), the aqueous layer was extracted with chloroform (4 × 25 mL). The organic layer was dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 4 cm, h = 20 cm, ethyl acetate : cyclohexane = 5 : 95, V = 20 mL, R_f = 0.33 (ethyl acetate : cyclohexane = 1.9)). Colorless oil, yield 598 mg (93%). C₂₅H₄₆O₄Si₂ (466.3). ¹H NMR (CDCl₃): δ [ppm] = 0.017, 0.021, 0.027, 0.030 (4s, 12H, Si(CH₃)₂), 0.87, 0.88 (2s, 18H, SiC(CH₃)₃), 1.80 (t, J = 7.3 Hz, 2H, CH₂CH₂OSi), 1.92 (t, J = 6.6 Hz, 1H, CH₂CH₂OSi), 1.93 (t, J = 6.9 Hz, 1H, CH₂CH₂OSi), 2.80 (d, J = 15.8 Hz, 1H, CH₂Ph), 2.99 (d, J = 15.8 Hz, 1H, CH₂Ph), 3.74 (t, J = 7.0 Hz, 2H, CH₂OSi), 3.76 (t, J = 6.8 Hz, 2H, CH₂OSi), 5.52 (s, 1H, CH), 7.06–7.08 (m, 1H, CH_arom), 7.16–7.18 (m, 1H, CH_arom), 7.21–7.26 (m, 2H, CH_arom). IR: v [cm⁻¹] = 3406 (b, ν, O–H), 2951 (s, ν, C–H, alkyl), 748 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₅H₄₆O₄Si₂ − OH [M − OH] 449.2947, found 449.2950. HPLC: Compound 10 was not stable during HPLC analysis.

4.1.6. 2-(1,5-Epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)ethan-1-ol (11).
Lactol 10 (1.3 g, 2.8 mmol) was dissolved in abs. THF (104 mL) and acidified with 1M HCl (pH 1, 11 mL). The reaction mixture was stirred overnight at rt. After removing the THF in vacuo, brine (25 mL) was added and the aqueous layer was extracted with chloroform (4 × 50 mL). The residue was dried (Na₂SO₄) and purified by fc (∅ = 8 cm, h = 18 cm, ethyl acetate : cyclohexane = 1 : 1–>6 : 4, V = 65 mL, R_f = 0.28). Colorless oil, yield 613 mg (99%). C₁₃H₁₆O₃ (220.1). ¹H NMR (CDCl₃): δ [ppm] = 1.49 (d broad, J = 13.3 Hz, 1H, 4-H_eq), 1.92 (td, J = 5.7/2.8 Hz, 2H, CH₂CH₂OH), 2.30–2.39 (m, 1H, 4-H_ax), 2.76 (d, J = 17.5 Hz, 1H, CH₂Ph), 3.23 (d, J = 17.5 Hz, 1H, CH₂Ph), 3.66–3.70 (m, 2H, 3-H_ax and 3-H_eq), 3.91 (ddd, J = 6.4/5.3/1.4 Hz, 2H, CH₂OH), 5.94 (s, 1H, 1-H), 7.14 (d, J = 7.6 Hz, 1H, CH_arom), 7.17 (dd, J = 7.5/1.5 Hz, 1H, CH_arom), 7.24 (t, J = 11.1 Hz, 1H, CH_arom), 7.30 (td, J = 7.4/1.6 Hz, 1H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 36.6 (1C, C-4), 36.8 (1C, C-6), 45.0 (1C, CH₂CH₂OH), 57.0 (1C, C-3), 58.9 (1C, CH₂OH), 72.8 (1C, C-5), 94.2 (1C, C-1), 126.6 (1C, CH_arom), 126.6 (1C, CH_arom), 127.3 (1C, CH_arom), 128.7 (1C, CH_arom), 131.9 (1C, Cq_arom), 135.0 (1C, Cq_arom). IR: v [cm⁻¹] = 3406 (b, ν, O–H), 2943 (s, ν, C–H, alkyl), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₁₃H₁₆O₃H [M + H] 221.1172, found 221.1201. Purity (HPLC): 99.7%, t_R = 13.65 min.

4.1.7. 5-(2-Azidoethyl)-1,5-epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocine (12).
Alcohol 11 (649 mg, 2.94 mmol) was dissolved in dry THF (49 mL). Then Zn(N₃)₂·2 pyridine (907 mg, 2.94 mmol), PPh₃ (1.5 g, 5.9 mmol) and THF (31 mL) were added. DIAD (1.2 mL, 5.9 mmol) was added dropwise and the reaction mixture was stirred at rt for 19 h. After addition of 1 M NaOH (50 mL) and brine (25 mL), the aqueous layer was extracted with ethyl acetate (4 × 50 mL). The organic layer was dried (Na₂SO₄) and after evaporation of the solvent the residue was purified by fc (∅ = 8 cm, h = 17.5 cm, ethyl acetate : cyclohexane = 4 : 6, V = 65 mL, R_f = 0.74 (cyclohexane : ethyl acetate = 6 : 4)). Colorless oil, yield 702 mg (97%). C₁₃H₁₅N₃O₂ (245.1). ¹H NMR (CDCl₃): δ [ppm] = 1.53 (d, J = 13.4 Hz, 1H, 4-H_eq), 1.94 (td, J = 7.1/0.9 Hz, 2H, CH₂CH₂N₃), 2.15–2.24 (m, 1H, 4-H_ax), 2.78 (d, J = 17.4 Hz, 1H, CH₂Ph), 3.12 (d, J = 17.4 Hz, 1H, CH₂Ph), 3.50 (t, J = 7.4 Hz, 2H, CH₂N₃), 3.60–3.70 (m, 2H, 3-H_ax and 3-H_eq), 5.93 (s, 1H, 1-H), 7.13 (d, J = 7.4 Hz, 1H, CH_arom), 7.17 (dd, J = 7.5/1.3 Hz, 1H, CH_arom), 7.24 (t, J = 7.4 Hz, 1H, CH_arom), 7.29 (td, J = 7.4/1.6 Hz, 1H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 36.46 (1C, C-4), 36.51 (1C, C-6), 42.3 (1C, CH₂CH₂N₃), 46.1 (1C, CH₂N₃), 56.9 (1C, C-3), 70.5 (1C, C-5), 94.4 (1C, C-1), 126.7 (2C, CH_arom), 127.3 (1C, CH_arom), 128.7 (1C, CH_arom), 132.0 (1C, Cq_arom), 134.7 (1C, Cq_arom). IR: v [cm⁻¹] = 2951 (s, ν, C–H, alkyl), 2091 (s, ν, N=N=N), 1103 (s, δ, C–O), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₁₃H₁₅N₃O₂ − N₂ + H [M − N₂ + H] 218.1181, found 218.1161. Purity (HPLC): 97.3%, t_R = 19.11 min.

4.1.8. 2-(1,5-Epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)ethan-1-amine (3a).
Azide 12 (702 mg, 2.86 mmol) was dissolved in ethyl acetate (27 mL) and Pd/C (112 mg, 16%) was added. The mixture was stirred under a H₂ atmosphere (1 atm) overnight. It was filtered through Celite®, the filtrate was concentrated in vacuo and the residue was purified by fc (∅ = 4.5 cm, h = 18 cm, methanol : CH₂Cl₂ : NH₃ = 50 : 50 : 1, V = 20 mL, R_f = 0.28 (ethyl acetate: methanol = 8 : 2)). Colorless oil, yield 472 mg (75%). C₁₃H₁₇NO₂ (219.1). ¹H NMR (CDCl₃): δ [ppm] = 1.50 (d, J = 13.3 Hz, 1H, 4-H_eq), 1.81 (t, J = 7.8 Hz, 2H, CH₂CH₂NH₂), 2.17–2.25 (m, 1H, 4-H_ax), 2.74 (d, J = 17.4 Hz, 1H, CH₂Ph), 2.90 (t, J = 7.5 Hz, 2H, CH₂NH₂), 3.12 (d, J = 17.4 Hz, 1H, CH₂Ph), 3.59–3.69 (m, 2H, 3-H_ax and 3-H_eq), 5.92 (s, 1H, 1-H), 7.12 (d, J = 7.4 Hz, 1H, CH_arom), 7.17 (dd, J = 7.4/1.3 Hz, 1H, CH_arom), 7.22 (t, J = 7.3 Hz, 1H, CH_arom), 7.28 (td, J = 7.4/1.7 Hz, 1H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 36.6 (1C, C-4), 36.7 (1C, C-6), 36.8 (1C, CH₂CH₂NH₂), 47.8 (1C, CH₂NH₂), 57.7 (1C, C-3), 71.3 (1C, C-5), 94.4 (1C, C-1), 126.5 (1C, CH_arom), 126.6 (1C, CH_arom), 127.3 (1C, CH_arom), 128.6 (1C, CH_arom), 132.3 (1C, Cq_arom), 135.2 (1C, Cq_arom). IR: v [cm⁻¹] = 3364 (b, ν, N–H), 2940 (s, ν, C–H, alkyl), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₁₃H₁₇NO₂ + H [M + H] 220.1332, found 220.1312. Purity (HPLC): 99.9%, t_R = 10.18 min.

4.1.9. N-Benzyl-2-(1,5-epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)ethan-1-amine (3b) and N,N-dibenzyl-2-(1,5-epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)ethan-1-amine (3c).
NEt₃ (40 mg, 0.39 mmol) was added to a solution of 3a-HCl (50 mg, 0.20 mmol) in CH₂Cl₂ (6 mL) to obtain the free primary amine. Then, benzaldehyde (30 µL, 0.29 mmol) and NaBH(OAc)₃ (83 mg, 0.39 mmol) were added. The reaction mixture was stirred overnight at rt. After addition of a saturated NaHCO₃ solution (10 mL) and brine (10 mL), the aqueous layer was extracted with chloroform (4 × 20 mL). Finally, the combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 1.5 cm, h = 13.0 cm, methanol : CH₂Cl₂ : NH₃ = 4 : 96.9 : 0.1, V = 5 mL). The fraction with the R_f = 0.55 of the first fc purification was purified by another fc (∅ = 1.5 cm, h = 15 cm, cyclohexane : ethyl acetate : N,N-dimethylethylamine = 94 : 6 : 0.1, V = 5 mL).

Compound 3b (R_f = 0.33 (methanol : ethyl acetate = 2 : 8)): colorless oil, yield 9 mg (16%). C₂₀H₂₃NO₂ (309.2). ¹H NMR (CDCl₃): δ [ppm] = 1.49 (d, J = 13.3 Hz, 1H, 4-H_eq), 1.80 (s broad, 1H, NH), 1.88 (t, J = 7.9 Hz, 2H, CH₂CH₂NH), 2.16–2.24 (m, 1H, 4-H_ax), 2.74 (d, J = 17.5 Hz, 1H, 6-H), 2.84 (t, J = 7.4 Hz, 2H, CH₂NH), 3.12 (d, J = 17.5 Hz, 1H, 6-H), 3.60–3.66 (m, 2H, 3-H_ax and 3-H_eq), 3.82 (s, 2H, CH₂Ph), 5.93 (s, 1H, 1-H), 7.11 (d, J = 7.4 Hz, 1H, CH_arom), 7.17 (d, J = 7.4 Hz, 1H, CH_arom), 7.21–7.33 (m, 7H, CH_arom).

¹³C NMR (CDCl₃): δ [ppm] = 36.6 (1C, C-4), 36.7 (1C, C-6), 43.8 (1C, CH₂CH₂NH), 43.8 (1C, CH₂NH), 54.3 (1C, CH₂Ph), 57.1 (1C, C-3), 71.4 (1C, C-5), 94.4 (1C, C-1), 126.5 (1C, CH_arom), 126.6 (1C, CH_arom), 127.1 (2C, CH_arom), 127.3 (1C, CH_arom), 128.3 (2C, CH_arom), 128.6 (2C, CH_arom), 132.2 (1C, Cq_arom), 135.2 (1C, Cq_arom), 140.2 (1C, Cq_arom). IR: v [cm⁻¹] = 3329 (w, ν, N–H), 3028 (s, ν, C–H, aryl), 2932 (s, ν, C–H, alkyl), 1605 (s, ν, C=C, aryl), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₀H₂₃NO₂H [M + H] 310.1802, found 310.1802. Purity (HPLC): 99.5%, t_R = 15.90 min.

Compound 3c (R_f = 0.55 (methanol : ethyl acetate = 2 : 8)): colorless oil, yield 35 mg (45%). C₂₇H₂₉NO₂ (399.2). ¹H NMR (CDCl₃): δ [ppm] = 1.29 (d, J = 13.6 Hz, 1H, 4-H_eq), 1.79 (ddd, J = 9.1/6.5/2.6 Hz, 2H, CH₂CH₂N), 1.94 (td, J = 12.4/7.2 Hz, 1H, 4-H_ax), 2.47 (d, J = 17.2 Hz, 1H, 6-H), 2.49–2.53 (m, 2H, CH₂N), 2.85 (d, J = 17.5 Hz, 1H, 6-H), 3.40–3.54 (m, 6H, CH₂Ph and 3-H_ax and 3-H_eq), 5.77 (s, 1H, 1-H), 6.92 (d, J = 7.4 Hz, 1H, CH_arom), 7.03 (d, J = 7.4 Hz, 1H, CH_arom), 7.07–7.16 (m, 4H, CH_arom), 7.20 (t, J = 7.4 Hz, 4H, CH_arom), 7.27 (d, J = 7.1 Hz, 4H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 36.3 (1C, C-6), 36.6 (1C, C-4), 41.1 (1C, CH₂CH₂N), 47.1 (1C, CH₂N), 57.0 (2C, CH₂Ph), 58.4 (1C, C-3), 71.2 (1C, C-5), 94.3 (1C, C-1), 126.4 (1C, CH_arom), 126.6 (1C, CH_arom), 127.0 (1C, CH_arom), 127.3 (1C, CH_arom), 128.3 (4C, CH_arom), 128.5 (2C, CH_arom), 129.0 (4C, CH_arom), 132.3 (1C, Cq_arom), 135.2 (1C, Cq_arom), 139.7 (2C, Cq_arom). IR: v [cm⁻¹] = 3059 (s, ν, C–H, aryl), 2936 (s, ν, C–H, alkyl), 1601 (s, ν, C=C, aryl), 733 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₇H₂₉NO₂H [M + H] 400.2271, found 400.2223; calcd for C₂₇H₂₉NO₂ − CH₂Ph [M − CH₂Ph] 308.1651, found 308.1649. Purity (HPLC): 98.6%, t_R = 19.61 min.

4.1.10. N-(Cyclohexylmethyl)-2-(1,5-epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)ethan-1-amine (3d).
Cyclohexanecarbaldehyde (45 µL, 0.37 mmol) and NaBH(OAc)₃ (158 mg, 0.74 mmol) were added to a solution of the primary amine 3a (82 mg, 0.37 mmol) in CH₂Cl₂ (11.5 mL). The reaction mixture was stirred for 18 h at rt. After addition of a saturated NaHCO₃ solution (10 mL) and brine (10 mL), the aqueous layer was extracted with chloroform (4 × 20 mL). Finally, the combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 3 cm, h = 1.5 cm, methanol : CH₂Cl₂ : NH₃ = 20 : 80 : 0.1, V = 10 mL, R_f = 0.13 (methanol : ethyl acetate = 20 : 80)). Colorless oil, yield 41 mg (35%). C₂₀H₂₉NO₂ (315.2). ¹H NMR (CDCl₃): δ [ppm] = 0.85–0.94 (m, 2H, CH(CH₂)₅), 1.12–1.28 (m, 3H, CH(CH₂)₅), 1.49 (d, J = 13.2 Hz, 1H, 4-H_eq), 1.64–1.75 (m, 5H, CH(CH₂)₅), 1.86 (t, J = 7.8 Hz, 2H, CH₂CH₂N), 2.17–2.24 (m, 2H, 4-H_ax, CH(CH₂)₅), 2.47 (d, J = 6.6 Hz, 2H, CH₂CH(CH₂)₅), 2.74 (d, J = 17.5 Hz, 1H, 6-H), 2.79 (t, J = 7.4 Hz, 2H, CH₂N), 3.12 (d, J = 17.5 Hz, 1H, 6-H), 3.62–3.69 (m, 2H, 3-H_ax and 3-H_eq), 5.91 (s, 1H, 1-H), 7.11 (d, J = 7.4 Hz, 1H, CH_arom), 7.16 (d, J = 7.4 Hz, 1H, CH_arom), 7.21 (t, J = 7.2 Hz, 1H, CH_arom), 7.27 (td, J = 7.4/1.6 Hz, 1H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 26.1 (2C, CH(CH₂)₅), 26.7 (2C, CH(CH₂)₅), 31.5 (1C, CH(CH₂)₅), 36.6 (1C, C-4), 36.7 (1C, C-6), 37.8 (1C, CH(CH₂)₅), 43.7 (1C, CH₂CH₂NH), 44.4 (1C, CH₂NH), 56.9 (1C, CH₂CH(CH₂)₅), 57.0 (1C, C-3), 71.3 (1C, C-5), 94.4 (1C, C-1), 126.5 (1C, CH_arom), 126.6 (1C, CH_arom), 127.3 (1C, CH_arom), 128.6 (1C, CH_arom), 132.2 (1C, Cq_arom), 135.2 (1C, Cq_arom). IR: v [cm⁻¹] = 2920 (s, ν, C–H, alkyl), 1103 (s, δ, C–O), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₀H₂₉NO₂H [M + H] 316.2271, found 316.2234; calcd for C₂₀H₂₉NO₂ − Cy [M − Cy] 232.1338, found 232.1306. Purity (HPLC): 95.6%, t_R = 17.15 min.

4.1.11. N,N-Bis(cyclohexylmethyl)-2-(1,5-epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)ethan-1-amine (3e).
NEt₃ (40 mg, 0.39 mmol) was added to a solution of 3a-HCl (50 mg, 0.20 mmol) in CH₂Cl₂ (6 mL) to obtain the free primary amine. Then, cyclohexanecarbaldehyde (36 µL, 0.29 mmol) and NaBH(OAc)₃ (83 mg, 0.39 mmol) were added. The reaction mixture was stirred for 18 h at rt. After addition of a saturated NaHCO₃ solution (10 mL) and brine (10 mL), the aqueous layer was extracted with chloroform (4 × 15 mL). Finally, the combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 1.5 cm, h = 13 cm, ethyl acetate : cyclohexane : N,N-dimethylethylamine = 6 : 94 : 0.1, V = 5 mL, R_f = 0.34 (methanol : ethyl acetate = 20 : 80)). Colorless oil, yield 41 mg (51%). C₂₇H₄₁NO₂ (411.3). ¹H NMR (CDCl₃): δ [ppm] = 0.77–0.85 (m, 5H, CH(CH₂)₅), 1.13–2.25 (m, 6H, CH(CH₂)₅), 1.38 (s broad, 2H, CH(CH₂)₅), 1.51 (d, J = 13.3 Hz, 1H, 4-H_eq), 1.67–1.80 (m, 11H, CH(CH₂)₅), 2.10–2.21 (m, 5H, CH₂CH₂N, CH₂CH(CH₂)₅, 4-H_ax), 2.49 (d, J = 7.8 Hz, 2H, CH₂N), 2.73 (d, J = 17.5 Hz, 1H, 6-H), 3.16 (d, J = 17.5 Hz, 1H, 6-H), 3.62–3.65 (m, 2H, 3-H_ax and 3-H_eq), 5.93 (s, 1H, 1-H), 7.12 (d, J = 7.4 Hz, 1H, CH_arom), 7.17 (d, J = 7.4 Hz, 1H, CH_arom), 7.21 (t, J = 6.9 Hz, 1H, CH_arom), 7.27 (td, J = 7.4/1.6 Hz, 1H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 26.4 (4C, CH(CH₂)₅), 27.1 (2C, CH(CH₂)₅), 32.1 (4C, CH(CH₂)₅), 36.4 (2C, CH(CH₂)₅), 36.6 (1C, C-6), 36.8 (1C, C-4), 41.4 (1C, CH₂CH₂N), 48.9 (1C, CH₂N), 57.1 (1C, C-3), 62.7 (2C, CH₂CH(CH₂)₅), 71.2 (1C, C-5), 94.4 (1C, C-1), 126.4 (1C, CH_arom), 126.6 (1C, CH_arom), 127.3 (1C, CH_arom), 128.5 (1C, CH_arom), 132.4 (1C, Cq_arom), 135.4 (1C, Cq_arom). IR: v [cm⁻¹] = 2920 (s, ν, C–H, alkyl), 1103 (s, δ, C–O), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₇H₄₁NO₂H [M + H] 412.3210, found 412.3201. Purity (HPLC): 96.1%, t_R = 22.02 min.

4.1.12. 2-(1,5-Epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)-N-(4-fluorobenzyl)ethan-1-amine (3f).
4-Fluorobenzaldehyde (30 µL, 0.28 mmol) and NaBH(OAc)₃ (119 mg, 0.56 mmol) were added to a solution of the primary amine 3a (62 mg, 0.28 mmol) in CH₂Cl₂ (8.5 mL). The reaction mixture was stirred for 18 h at rt. After addition of a saturated NaHCO₃ solution (10 mL) and brine (10 mL), the aqueous layer was extracted with chloroform (4 × 20 mL). Finally, the combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 2 cm, h = 20 cm, methanol : CH₂Cl₂ : NH₃ = 5 : 95 : 0.1, V = 10 mL, R_f = 0.46 (methanol : ethyl acetate = 20 : 80)). Colorless oil, yield 50 mg (55%). C₂₀H₂₂FNO₂ (327.4). ¹H NMR (CDCl₃): δ [ppm] = 1.46 (d, J = 13.2 Hz, 1H, 4-H_eq), 1.84 (t, J = 8.0 Hz, 2H, CH₂CH₂N), 1.84 (s broad, 1H, NH), 2.13–2.22 (m, 1H, 4-H_ax), 2.71 (d, J = 17.5 Hz, 1H, 6-H), 2.80 (t, J = 7.4 Hz, 2H, CH₂N), 3.09 (d, J = 17.5 Hz, 1H, 6-H), 3.57–3.66 (m, 2H, 3-H_ax and 3-H_eq), 3.75 (s, 2H, CH₂Ph), 5.90 (s, 1H, 1-H), 6.96 (t, J = 8.8 Hz, 2H, CH_arom), 7.09 (d, J = 7.5 Hz, 1H, CH_arom), 7.14 (dd, J = 7.5/1.2 Hz, 1H, CH_arom), 7.20 (t, J = 7.0 Hz, 1H, CH_arom), 7.24–7.28 (m, 3H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 36.7 (1C, C-4), 36.7 (1C, C-6), 43.8 (2C, CH₂CH₂N), 53.6 (1C, CH₂Ph), 57.1 (1C, C-3), 71.4 (1C, C-5), 94.4 (1C, C-1), 115.3 (1C, CH_arom), 115.5 (1C, CH_arom), 126.6 (1C, CH_arom), 126.7 (1C, CH_arom), 127.4 (1C, CH_arom), 128.7 (1C, CH_arom), 129.8 (1C, CH_arom), 129.9 (1C, CH_arom), 132.3 (1C, Cq_arom), 135.2 (1C, Cq_arom), 160.9 (1C, Cq_arom), 163.2 (1C, Cq_arom). IR: v [cm⁻¹] = 3333 (b, ν, N–H), 2932 (s, ν, C–H, alkyl), 1601 (s, ν, C=C, aryl), 1099 (s, δ, C–O), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₀H₂₂FNO₂H [M + H] 328.1707, found 328.1731. Purity (HPLC): 99.3%, t_R = 16.25 min.

4.1.13. 2-(1,5-Epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)-N-(3-phenylpropyl)ethan-1-amine (3g).
3-Phenylpropionaldehyde (37 µL, 0.28 mmol) and NaBH(OAc)₃ (119 mg, 0.56 mmol) were added to a solution of the primary amine 3a (62 mg, 0.28 mmol) in CH₂Cl₂ (8.5 mL). The reaction mixture was stirred for 18 h at rt. After addition of a saturated NaHCO₃ solution (10 mL) and brine (10 mL), the aqueous layer was extracted with chloroform (4 × 20 mL). Finally, the combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 2 cm, h = 20 cm, methanol : CH₂Cl₂ : NH₃ = 5 : 95 : 0.1–>10 : 90 : 0.1, V = 10 mL, R_f = 0.19 (methanol : ethyl acetate = 20 : 80)). Colorless oil, yield 44 mg (46%). C₂₂H₂₇NO₂ (337.5). ¹H NMR (CDCl₃): δ [ppm] = 1.50 (d, J = 13.4 Hz, 1H, 4-H_eq), 2.02–2.10 (m, 4H, CH₂CH₂N and CH₂CH₂Ph), 2.20–2.28 (m, 1H, 4-H_ax), 2.74 (t, J = 8.0 Hz, 2H, CH₂CH₂CH₂Ph), 2.78 (d, J = 17.7 Hz, 1H, 6-H), 2.88 (t, J = 7.2 Hz, 2H, NCH₂CH₂CH₂Ph), 3.03 (t, J = 7.5 Hz, 2H, CH₂N), 3.17 (d, J = 17.5 Hz, 1H, 6-H), 3.64–3.70 (m, 2H, 3-H_ax and 3-H_eq), 4.64 (s broad, 1H, NH), 5.97 (s, 1H, 1-H), 7.16–7.37 (m, 9H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 29.9 (1C, CH₂CH₂Ph), 33.3 (1C, CH₂Ph), 36.4 (1C, C-4), 36.5 (1C, C-6), 41.5 (1C, CH₂CH₂N), 43.8 (1C, CH₂N), 48.8 (1C, CH₂N), 56.9 (1C, C-3), 71.0 (1C, C-5), 94.3 (1C, C-1), 126.2 (1C, CH_arom), 126.6 (1C, CH_arom), 126.6 (1C, CH_arom), 127.3 (1C, CH_arom), 128.5 (2C, CH_arom), 128.6 (2C, CH_arom), 128.7 (1C, CH_arom), 131.9 (1C, Cq_arom), 134.8 (1C, Cq_arom), 141.2 (1C, Cq_arom). IR: v [cm⁻¹] = 3329 (s, ν, N–H), 2932 (s, ν, C–H, alkyl), 1605 (s, ν, C=C, aryl), 1099 (s, δ, C–O), 737 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₂H₂₇NO₂H [M + H] 338.2115, found 338.2127; calcd for C₂₂H₂₇NO₂ − CH₂CH₂CH₂Ph [M − CH₂CH₂CH₂Ph] 218.1181, found 218.1178. Purity (HPLC): 98.2%, t_R = 17.69 min.

4.1.14. N-Benzyl-2-(1,5-epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)-N-methylethan-1-amine (3h).
Formalin solution (37%, 114 µL, 4.10 mmol) and NaBH(OAc)₃ (130 mg, 0.62 mmol) were added to a solution of the benzylamine 3b (63 mg, 0.21 mmol) in CH₂Cl₂ (7 mL). The reaction mixture was stirred for 18 h at rt. After addition of a saturated NaHCO₃ solution (10 mL) and brine (10 mL), the aqueous layer was extracted with chloroform (4 × 20 mL). Finally, the combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 1.5 cm, h = 15 cm, methanol : CH₂Cl₂ : NH₃ = 5 : 95 : 0.1, V = 10 mL, R_f = 0.62 (methanol : ethyl acetate = 3 : 7)). Colorless oil, yield 64 mg (97%). C₂₁H₂₅NO₂ (323.2). ¹H NMR (CDCl₃): δ [ppm] = 1.52 (d, J = 13.3 Hz, 1H, 4-H_eq), 1.89 (dd, J = 8.9/6.6 Hz, 2H, CH₂CH₂N), 2.13–2.19 (m, 1H, 4-H_ax), 2.23 (s, 3H, CH₃), 2.58 (td, J = 7.0/3.0 Hz, 2H, CH₂N), 2.73 (d, J = 17.5 Hz, 1H, 6-H), 3.10 (d, J = 17.5 Hz, 1H, 6-H), 3.52 (s, 2H, CH₂Ph), 3.57–3.65 (m, 2H, 3-H_ax and 3-H_eq), 5.93 (s, 1H, 1-H), 7.10 (d, J = 7.4 Hz, 1H, CH_arom), 7.17 (d, J = 7.4 Hz, 1H, CH_arom), 7.20–7.36 (m, 7H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 36.6 (1C, C-4), 36.7 (1C, C-6), 41.6 (1C, CH₂CH₂N), 42.4 (1C, NCH₃), 51.3 (1C, CH₂N), 57.1 (1C, C-3), 62.5 (1C, CH₂Ph), 71.2 (1C, C-5), 94.4 (1C, C-1), 126.5 (1C, CH_arom), 126.6 (1C, CH_arom), 127.1 (1C, CH_arom), 127.3 (1C, CH_arom), 128.4 (2C, CH_arom), 128.5 (1C, CH_arom), 129.2 (2C, CH_arom), 132.3 (1C, Cq_arom), 135.3 (1C, Cq_arom), 139.1 (1C, Cq_arom). IR: v [cm⁻¹] = 3024 (s, ν, C–H, aryl), 2951 (s, ν, C–H, alkyl), 1604 (s, ν, C=C, aryl), 1134 (s, δ, C–O), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₁H₂₅NO₂H [M + H] 324.1958, found 324.1938; calcd for C₂₁H₂₅NO₂ − Bn [M − Bn] 232.1338, found 323.1306; calcd for C₂₁H₂₅NO₂ − CH₂CH₂N(CH₃)(Bn) [M − CH₂CH₂N(CH₃)(Bn)] 175.0759, found 175.0753. Purity (HPLC): 99.5%, t_R = 16.50 min.

4.1.15. N-(Cyclohexylmethyl)-2-(1,5-epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)-N-methylethan-1-amine (3j).
Formalin solution (37%, 52 µL, 1.88 mmol) and NaBH(OAc)₃ (60 mg, 0.28 mmol) were added to a solution of the cyclohexylmethylamine 3d (30 mg, 0.09 mmol) in CH₂Cl₂ (3 mL). The reaction mixture was stirred for 18 h at rt. After addition of a saturated NaHCO₃ solution (10 mL) and brine (10 mL), the aqueous layer was extracted with chloroform (4 × 20 mL). Finally, the combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 1.5 cm, h = 17 cm, methanol : CH₂Cl₂ : NH₃ = 5 : 95 : 0.1, V = 10 mL, R_f = 0.25 (methanol : ethyl acetate = 20 : 80)). Colorless oil, yield 27 mg (87%). C₂₁H₃₁NO₂ (329.2). ¹H NMR (CDCl₃): δ [ppm] = 0.86–0.90 (m, 2H, CH(CH₂)₅), 1.10–1.25 (m, 3H, CH(CH₂)₅), 1.45–1.50 (m, 1H, CH(CH₂)₅), 1.54 (d, J = 8.0 Hz, 1H, 4-H_eq), 1.64–1.72 (m, 3H, CH(CH₂)₅), 1.75–1.79 (m, 2H, CH(CH₂)₅), 1.82–1.86 (m, 2H, CH₂CH₂N), 2.14–2.22 (m, 3H, CH₂CH(CH₂)₅ and 4-H_ax), 2.25 (s, 3H, CH₃), 2.52–2.55 (m, 2H, CH₂N), 2.76 (d, J = 17.5 Hz, 1H, 6-H), 3.14 (d, J = 17.5 Hz, 1H, 6-H), 3.59–3.68 (m, 2H, 3-H_ax and 3-H_eq), 5.92 (s, 1H, 1-H), 7.12 (d, J = 7.4 Hz, 1H, CH_arom), 7.16 (d, J = 7.4 Hz, 1H, CH_arom), 7.22 (t, J = 7.2 Hz, 1H, CH_arom), 7.27 (td, J = 7.3/1.6 Hz, 1H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 26.2 (2C, CH(CH₂)₅), 26.9 (2C, CH(CH₂)₅), 32.0 (1C, CH(CH₂)₅), 35.8 (1C, CH(CH₂)₅), 36.6 (1C, C-6), 36.7 (1C, C-4), 41.0 (1C, CH₂CH₂N), 43.1 (1C, CH₂N), 52.2 (1C, NCH₃), 57.1 (1C, C-3), 65.1 (1C, CH₂CH(CH₂)₅), 71.2 (1C, C-5), 94.4 (1C, C-1), 126.5 (1C, CH_arom), 126.6 (1C, CH_arom), 127.3 (1C, CH_arom), 128.5 (1C, CH_arom), 132.3 (1C, Cq_arom), 135.3 (1C, Cq_arom).

IR: v [cm⁻¹] = 2920 (s, ν, C–H, alkyl), 1103 (s, δ, C–O), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₂₁H₃₁NO₂H [M + H] 330.2540, found 330.2504; calcd for C₂₁H₃₁NO₂ − CH₂Cy [M − CH₂Cy] 232.1338, found 232.1391. Purity (HPLC): 99.2%, t_R = 17.98 min.

4.1.16. 2-(1,5-Epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)-N,N-dimethylethan-1-amine (3k).
NEt₃ (100 µL, 145 mg, 0.72 mmol) was added to a suspension of 3a-HCl (62 mg, 0.24 mmol) in CH₂Cl₂ (7 mL) to obtain the free primary amine 3a. Then, formalin solution (37%, 133 µL, 4.83 mmol) and NaBH(OAc)₃ (154 mg, 0.72 mmol) were added. The reaction mixture was stirred overnight at rt. Saturated NaHCO₃ solution (10 mL) and brine (10 mL) were added and the aqueous layer was extracted with chloroform (4 × 20 mL). The combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 2 cm, h = 15.5 cm, methanol : CH₂Cl₂ : NH₃ = 10 : 90 : 1, V = 10 mL, R_f = 0.13 (methanol : ethyl acetate = 2 : 8)). Yellow oil, yield 30 mg (50%). C₁₅H₂₁NO₂ (247.3). ¹H NMR (CDCl₃): δ [ppm] = 1.53 (d, J = 8.0 Hz, 1H, 4-H_eq), 1.81–1.85 (m, 2H, CH₂CH₂N), 2.14–2.22 (m, 1H, 4-H_ax), 2.27 (m, 6H, N(CH₃)₂), 2.46 (t, J = 7.5 Hz, 2H, CH₂N), 2.76 (d, J = 17.4 Hz, 1H, 6-H), 3.11 (d, J = 17.5 Hz, 1H, 6-H), 3.58–3.65 (m, 2H, 3-H_ax and 3-H_eq), 5.92 (s, 1H, 1-H), 7.11 (d, J = 7.4 Hz, 1H, CH_arom), 7.16 (dd, J = 7.5/1.4 Hz, 1H, CH_arom), 7.21 (t, J = 7.1 Hz, 1H, CH_arom), 7.27 (td, J = 7.4/1.6 Hz, 1H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 36.6 (1C, C-6), 36.6 (1C, C-4), 41.8 (1C, CH₂CH₂N), 45.6 (2C, N(CH₃)₂), 53.5 (1C, CH₂N), 57.0 (1C, C-3), 71.0 (1C, C-5), 94.4 (1C, C-1), 126.5 (1C, CH_arom), 126.6 (1C, CH_arom), 127.3 (1C, CH_arom), 128.6 (1C, CH_arom), 132.3 (1C, Cq_arom), 135.1 (1C, Cq_arom). IR: v [cm⁻¹] = 2947 (s, ν, C–H, alkyl), 1103 (s, δ, C–O), 764 (s, δ, 1,2-disubst. aryl). MS (APCI): calcd for C₁₅H₂₁NO₂ + H [M + H] 248.1645, found 248.1666. Purity (HPLC): 99.0%, t_R = 12.09 min.

4.1.17. N,N-Dibutyl-2-(1,5-epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)ethan-1-amine (3l).
Butyraldehyde (33 µL, 0.36 mmol) and NaBH(OAc)₃ (154 mg, 0.73 mmol) were added to a solution of the primary amine 3a (80 mg, 0.36 mmol) in CH₂Cl₂ (11 mL). The reaction mixture was stirred for 18 h at rt. After addition of a saturated NaHCO₃ solution (10 mL) and brine (10 mL), the aqueous layer was extracted with CHCl₃ (4 × 10 mL). Finally, the combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 2 cm, h = 17 cm, methanol : CH₂Cl₂ : NH₃ = 10 : 90 : 0.1, V = 10 mL, R_f = 0.43 (methanol : ethyl acetate = 20 : 80)). Colorless oil, yield 8 mg (7%). C₂₁H₃₃NO₂ (331.5). ¹H NMR (CDCl₃): δ [ppm] = 0.94 (t, J = 7.3 Hz, 6H, NCH₂CH₂CH₂CH₃), 1.29–1.39 (m, 4H, NCH₂CH₂CH₂CH₃), 1.53–1.60 (m, 5H, NCH₂CH₂CH₂CH₃, 4-H_eq), 1.94–1.97 (m, 2H, CH₂CH₂N), 2.13–2.22 (m, 1H, 4-H_ax), 2.59–2.74 (m, 4H, NCH₂CH₂CH₂CH₃), 2.79 (d, J = 17.7 Hz, 1H, 6-H), 2.81–2.94 (m, 2H, CH₂N), 3.13 (d, J = 17.7 Hz, 1H, 6-H), 3.63–3.66 (m, 2H, 3-H_ax and 3-H_eq), 5.91 (s, 1H, 1-H), 7.13 (d, J = 7.5 Hz, 1H, CH_arom), 7.17 (dd, J = 7.4/1.1 Hz, 1H, CH_arom), 7.23 (t, J = 7.2 Hz, 1H, CH_arom), 7.29 (td, J = 7.5/1.6 Hz, 1H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 14.0 (2C, NCH₂CH₂CH₂CH₃), 20.7 (2C, NCH₂CH₂CH₂CH₃), 27.6 (2C, NCH₂CH₂CH₂CH₃), 36.4 (1C, C-6), 36.7 (1C, C-4), 40.0 (1C, CH₂CH₂N), 47.8 (1C, CH₂N), 53.3 (2C, NCH₂CH₂CH₂CH₃), 57.0 (1C, C-3), 70.9 (1C, C-5), 94.4 (1C, C-1), 126.60 (1C, CH_arom), 126.63 (1C, CH_arom), 127.4 (1C, CH_arom), 128.7 (1C, CH_arom), 132.1 (1C, Cq_arom), 134.9 (1C, Cq_arom). IR: v [cm⁻¹] = 2955 (s, C–H), 1609 (s, C–C_arom), 1103 (s, C–O), 764 (s, 1,2-disubst. benzene). MS (APCI): calcd for C₂₁H₃₃NO₂H [M + H] 332.2584, found 332.2598; calcd for C₂₁H₃₃NO₂ − butyl [M − butyl] 274.1807, found 274.1806. Purity (HPLC): 96.3%, t_R = 18.18 min.

4.1.18. 1-[2-(1,5-Epoxy-3,4,5,6-tetrahydro-1H-2-benzoxocin-5-yl)ethyl]piperidine (3m).
NEt₃ (90 µL, 0.65 mmol) was added to a solution of 3a-HCl (83 mg, 0.33 mmol) in CH₂Cl₂ (10 mL) to obtain the free primary amine 3a. Then, glutaraldehyde (31 µL, 0.33 mmol) and NaBH(OAc)₃ (138 mg, 0.65 mmol) were added. The reaction mixture was stirred for 18 h at rt. After addition of a saturated NaHCO₃ solution (10 mL) and brine (10 mL), the aqueous layer was extracted with CHCl₃ (4 × 10 mL). Finally, the combined organic layers were dried (Na₂SO₄) and the solvent was removed in vacuo. The residue was purified by fc (∅ = 2 cm, h = 18 cm, methanol : CH₂Cl₂ : NH₃ = 10 : 90 : 0.1, V = 10 mL, R_f = 0.17 (methanol : ethyl acetate = 20 : 80)). Colorless oil, yield 30 mg (32%). C₁₈H₂₅NO₂ (287.2). ¹H NMR (CDCl₃): δ [ppm] = 1.52–1.56 (m, 3H, 4-H_eq, 4-CH₂ pip), 1.77–1.85 (m, 4H, 3-CH₂, 5-CH₂ pip), 2.05–2.21 (m, 3H, 4-H_ax, CH₂CH₂N), 2.62–2.86 (m, 6H, CH₂N, 2-CH₂, 6-CH₂ pip), 2.79 (d, J = 17.8 Hz, 1H, 6-H), 3.13 (d, J = 17.8 Hz, 1H, 6-H), 3.62–3.65 (m, 2H, 3-H_ax and 3-H_eq), 5.89 (s, 1H, 1-H), 7.11–7.16 (m, 2H, CH_arom), 7.21 (t, J = 7.1 Hz, 1H, CH_arom), 7.28 (td, J = 7.4/1.6 Hz, 1H, CH_arom). ¹³C NMR (CDCl₃): δ [ppm] = 23.4 (1C, C-4 pip), 24.4 (2C, C-3, C-5 pip), 36.3 (1C, C-6), 36.7 (1C, C-4), 38.7 (1C, CH₂CH₂N), 52.8 (1C, CH₂N), 54.3 (2C, C-2, C-6 pip), 56.9 (1C, C-3), 70.7 (1C, C-5), 94.4 (1C, C-1), 126.6 (2C, CH_arom), 127.4 (1C, CH_arom), 128.7 (1C, CH_arom), 132.0 (1C, Cq_arom), 134.8 (1C, Cq_arom). IR: v [cm⁻¹] = 2932 (s, C–H), 1099 (s, C–O), 764 (s, 1,2-disubst. benzene). MS (APCI): calcd for C₁₈H₂₅NO₂H [M + H] 288.1958, found 288.1981. Purity (HPLC): 97.4%, t_R = 13.89 min.

4.2. Receptor binding studies

4.2.1. σ₁ receptor assay^24,25. The assay was performed with the radioligand [³H]-(+)-pentazocine (22.0 Ci mmol⁻¹; Perkin Elmer). Guinea pig brains were commercially available (Envigo, Horst, Netherlands). The thawed membrane preparation of guinea pig brain (about 100 μg of the protein) was incubated with various concentrations of test compounds, 2 nM [³H]-(+)-pentazocine, and TRIS buffer (50 mM, pH 7.4) at 37 °C. The non-specific binding was determined with 10 μM unlabeled (+)-pentazocine.

4.2.2. Further assays to record receptor affinity. The assays to determine the affinity towards σ₂ receptors,^24,25 and towards the PCP binding site of NMDA receptors,^17,23 were conducted as reported in literature. Details of the receptor binding studies are given in the SI.

4.3. Determination of log D_7.4 values, plasma protein binding and metabolic stability in vitro

log D_7.4 values were recorded using the micro shake flask method reported in ref. 26 and 27. Plasma protein binding was recorded by HPAC.^30,31 Metabolic stability was determined by incubation with mouse liver microsomes.^26,32 Details are given in the SI.

Conflicts of interest

The authors declare no conflict of interest.

Abbreviations

DIAD	Diisopropyl azodicarboxylate
DIBAH	Diisobutylaluminum hydride
DMP	Dess-Martin-Periodinane
HPAC	High performance affinity chromatography
LLE	Ligand-lipophilicity efficiency
LC-MS	Liquid chromatography combined with mass spectrometry
MOPS	3-Morpholinopropanesulfonic acid
NADPH	Nicotinamide adenine dinucleotide phosphate
NMDA	N-Methyl-D-aspartate
NMR	Nuclear magnetic resonance
PCP	1-(1-Phenylcyclohexyl)piperidine (phencyclidine)
SEM	Standard error of the mean
TBS	tert-Butyldimethylsilyl
THF	Tetrahydrofuran

Data availability

All data will be made available on request to the corresponding author of this manuscript.

Supplementary information (SI): general synthetic methods, the method to determine the purity of the compounds, details of the receptor binding studies including the assays for σ₁ and σ₂ receptor affinity as well as the assay to determine the affinity towards the PCP binding site of the NMDA receptor. Experimental details to determine log D_7.4 values, plasma protein binding and metabolic stability are given. Moreover, ¹H and ¹³C NMR spectra as well as HPLC traces for all test compounds. See DOI: https://doi.org/10.1039/d6ob00129g.

Acknowledgements

This work was supported by the Deutsche Forschungsgemeinschaft (DFG), which is gratefully acknowledged.

References

1. R. Dingledine, K. Borges, D. Bowie and S. F. Traynelis, The Glutamate Receptor Ion Channels, Pharmacol. Rev., 1999, 51(1), 7–61,  DOI:10.1016/S0031-6997(24)01394-2.
2. S. F. Traynelis, L. P. Wollmuth, C. J. McBain, F. S. Menniti, K. M. Vance, K. K. Ogden, K. B. Hansen, H. Yuan, S. J. Myers and R. Dingledine, Glutamate Receptor Ion Channels: Structure, Regulation, and Function, Pharmacol. Rev., 2010, 62(3), 405–496,  DOI:10.1124/pr.109.002451.
3. H. Stark, U. Reichert and S. Graßmann, Structure, function and potential therapeutic significance of NMDA receptors. Part 2: Therapeutic concepts and new receptor ligands, Pharm. Unserer Zeit, 2000, 29, 228–236.
4. E. Karakas and H. Furukawa, Crystal structure of heterotetrameric NMDA receptor ion channel, Science, 2014, 344, 992–997.
5. E. Karakas, N. Simorowski and H. Furukawa, Subunit arrangement and phenylethanolamine binding in GluN1/GluN2B NMDA receptors, Nature, 2011, 475, 249–253,  DOI:10.1038/nature10180.
6. C. G. Rousseaux and S. Greene, Sigma receptors [sRs] : biology in normal and diseased states, J. Recept. Signal Transduction, 2016, 36, 327–388.
7. T. Maurice and T. P. Su, The pharmacology of sigma-1 receptors, Pharmacol. Ther., 2009, 124(2), 195–206.
8. S. Collina, R. Gaggeri, A. Marra, A. Bassi, S. Negrinotti, F. Negri and D. Rossi, Sigma receptor modulators: a patent review, Expert Opin. Ther. Pat., 2013, 23, 597–613.
9. F. Langa, X. Codony, V. Tovar, A. Lavado, E. Giménez, P. Cozar, M. Cantero, A. Dordal, E. Hernández, R. Pérez, X. Monroy, D. Zamanillo, X. Guitart and L. Montoliu, Generation and phenotypic analysis of sigma receptor type I (σ1) knockout mice, Eur. J. Neurosci., 2003, 18, 2188–2196.
10. V. Sabino, P. Cottone, S. L. Parylak, L. Steardo and E. P. Zorrilla, Sigma-1 receptor knockout mice display a depressive-like phenotype, Behav. Brain Res., 2009, 198, 472–476.
11. H. R. Schmidt, S. D. Zheng, E. Gurpinar, A. Koehl, A. Manglik and A. C. Kruse, Crystal structure of the human sigma 1 receptor, Nature, 2016, 532(7600), 527–530.
12. H. R. Schmidt, R. M. Betz, R. O. Dror and A. C. Kruse, Structural basis for sigma(1) receptor ligand recognition, Nat. Struct. Mol. Biol., 2018, 25(10), 981–987.
13. W. R. Martin, C. G. Eades, J. A. Thompson, R. E. Huppler and P. E. Gilbert, The effects of morphine- and nalorphine- like drugs in the nondependent and morphine-dependent chronic spinal dog, J. Pharmacol. Exp. Ther., 1976, 197, 517–532.
14. S. B. Hellewell and W. D. Bowen, A sigma-like binding site in rat pheochromocytoma (PC12) cells: decreased affinity for (+)-benzomorphans and lower molecular weight suggest a different sigma receptor form from that of guinea pig brain, Brain Res., 1990, 527, 244–253.
15. S. B. Hellewell, A. Bruce, G. Feinstein, J. Orringer, W. Williams and W. D. Bowen, Rat liver and kidney contain high densities of sigma 1 and sigma 2 receptors: characterization by ligand binding and photoaffinity labeling, Eur. J. Pharmacol., 1994, 268, 9–18.
16. F. I. Carroll, P. Abraham, K. Parham, X. Bai, X. Zhang, G. A. Brine, S. W. Mascarella, B. R. Martin, E. L. May, C. Sauss, L. Di Paolo, P. Wallace, J. M. Walker and W. D. Bowen, Enantiomeric N-substituted N-normetazocines: a comparative study of affinities at sigma, PCP, and mu opioid receptors, J. Med. Chem., 1992, 35, 2812–2818.
17. J. Köhler, K. Bergander, J. Fabian, D. Schepmann and B. Wünsch, Enantiomerically pure 1,3-dioxanes as highly selective NMDA and σ₁ receptor ligands, J. Med. Chem., 2012, 55, 8953–8957.
18. T. Utech, J. Köhler, H. Buschmann, J. Holenz, J. M. Vela and B. Wünsch, Synthesis and pharmacological evaluation of a potent and selective σ₁ receptor antagonist with high antiallodynic activity, Arch. Pharm. Chem. Life Sci., 2011, 344, 415–421.
19. E. Quick, D. Schepmann and B. Wünsch, Synthesis and structure-affinity relationships of σ receptor ligands with 1,3-dioxane structure, Pharmaceuticals, 2025, 18, 1300,  DOI:10.3390/ph18091300.
20. R. J. Boeckman and K. M. George, 1,1,1-Triacetoxy-1,1-dihydro-1,2-benziodoxol-3(1 H)-one, in Encyclopedia of Reagents for Organic Synthesis, John Wiley & Sons, Ltd, Chichester, UK, 2009, p. 15.  DOI:10.1002/047084289x.rt157m.pub2.
21. E. Winterfeldt, Synthesis, 1975, 617–630.
22. A. F. Abdel-Magid and S. J. A. Mehrman, A Review on the Use of Sodium Tiacetoxyborohydride in the Reductive Amination of Ketones and Aldehydes, Org. Process Res. Dev., 2006, 10, 971–1031.
23. S. Gawaskar, D. Schepmann, A. Bonifazi, D. Robaa, W. Sippl and B. Wünsch, Benzo[7]annulene-based GluN2B selective NMDA receptor antagonists: Surprising effect of a nitro group in 2-position, Bioorg. Med. Chem. Lett., 2015, 25(24), 5748–5751.
24. N. Kopp, G. Civenni, D. Marson, E. Laurini, S. Pricl, C. V. Catapano, H.-U. Humpf, C. Almansa, F. R. Nieto, D. Schepmann and B. Wünsch, Chemoenzymatic synthesis of 2,6-disubsttuted tetrahydropyrans with high σ₁ receptor affinity, antitumor and analgesic activity, Eur. J. Med. Chem., 2021, 219, 113443.
25. K. Miyata, D. Schepmann and B. Wünsch, Synthesis and σ receptor affinity of regioisomeric spirocyclic furopyridines, Eur. J. Med. Chem., 2014, 83, 709–716.
26. F. Börgel, F. Galla, K. Lehmkuhl, D. Schepmann, S. M. Ametamey and B. Wünsch, Pharmacokinetic properties of enantiomerically pure GluN2B selective NMDA receptor antagonists with 3-benzazepine scaffold, J. Pharm. Biomed. Anal., 2019, 172, 214–222.
27. F. Galla, C. Bourgeois, K. Lehmkuhl, D. Schepmann, M. Soeberdt, T. Lotts, C. Abels, S. Ständer and B. Wünsch, Effects of polar κ receptor agonists designed for the periphery on ATP-induced Ca²⁺ release from keratinocytes, MedChemComm, 2016, 7, 317–326,  10.1039/C5MD00414D.
28. P. D. Leeson and B. Springthorpe, The influence of drug-like concepts on decision-making in medicinal chemistry, Nat. Rev. Drug Discovery, 2007, 6(11), 881–890,  DOI:10.1038/nrd2445.
29. A. Tarcsay, K. Nyíri and G. M. Keseru, Impact of lipophilic efficiency on compound quality, J. Med. Chem., 2012, 55, 1252–1260.
30. V. Butsch, F. Börgel, F. Galla, K. Schwegmann, S. Hermann, M. Schäfers, B. Riemann, B. Wünsch and S. Wagner, Design, (Radio)Synthesis, and in Vitro and in Vivo evaluation of highly selective and potent matrix metalloproteinase 12 (MMP-12) inhibitors as radiotracers for positron emission tomography, J. Med. Chem., 2018, 61, 4115–4134,  DOI:10.1021/acs.jmedchem.8b00200.
31. C. P. Konken, K. Heßling, I. Thale, S. Schelhaas, J. Dabel, S. Maskri, E. Bulk, T. Budde, O. Koch, A. Schwab, M. Schäfers and B. Wünsch, Imaging of the calcium activated potassium channel 3.1 (K_Ca3.1) in vivo using a senicapoc-derived PET tracer, Arch. Pharm., 2022, e2200388,  DOI:10.1002/ardp.202200388.
32. C. Wiese, E. Große Maestrup, F. Galla, D. Schepmann, A. Hiller, S. Fischer, F.-A. Ludwig, W. Deuther-Conrad, C. K. Donat, P. Brust, L. Büter, U. Karst and B. Wünsch, . Comparison of in silico, electrochemical, in vitro and in vivo metabolism of a homologous series of (radio)fluorinated σ₁ receptor ligands designed for positron emission tomography, ChemMedChem, 2016, 11, 2445–2458,  DOI:10.1002/cmdc.201600366.
33. S. Brune, D. Schepmann, K. H. Klempnauer, D. Marson, V. Dal Col, E. Laurini, M. Fermeglia, B. Wünsch and S. Pricl, The sigma enigma: in vitro/in silico site-directed mutagenesis studies unveil σ₁ receptor ligand binding, Biochemistry, 2014, 53(18), 2993–3003.
34. E. Laurini, V. Dal Col, M. G. Mamolo, D. Zampieri, P. Posocco, M. Fermeglia, L. Vio and S. Pricl, Homology model and docking-based virtual screening for ligands of the σ₁ receptor, ACS Med. Chem. Lett., 2011, 2(11), 834–839.
35. E. Rack, R. Fröhlich, D. Schepmann and B. Wünsch, Design, synthesis and pharmacological evaluation of spirocyclic σ1 receptor ligands with excocyclic amino moiety (increased distance 1), Bioorg. Med. Chem., 2011, 19, 3141–3151.

---