Asymmetric synthesis and σ receptor affinity of enantiomerically pure 1,4-disubstituted tetrahydro-1H-3-benzazepines

Synthesis of Chiral Tetrahydro-3-benzazepine Motifs by Iridium-Catalyzed Asymmetric Hydrogenation of Cyclic Ene-carbamates

A highly efficient N,P-ligated iridium complex is presented for the simple preparation of chiral tetrahydro-3-benzazepine motifs by catalytic asymmetric hydrogenation. Substrates bearing both 1-aryl and 1-alkyl substituents were smoothly converted to the corresponding hydrogenated product with excellent enantioselectivity (91-99% ee) and in isolated yield (92-99%). The synthetic value of this transformation was demonstrated by a gram-scale hydrogenation and application in the syntheses of trepipam and fenoldopam.

Asymmetric synthesis of 3-substituted tetrahydro-2-benzazepines

The enantiomerically and diastereomerically pure tricyclic oxazolidine cis-10 was prepared in a five step synthesis starting with 1-bromo-2-iodobenzene. Me3SiCN and allylSiMe3 reacted with cis-10 in the presence of TiCl4 to form the nitrile (3S)-11 and the allyl derivative (3S)-12 with high diastereoselectivity. The hydrogenolytic removal of the chiral auxiliary failed, since the endocyclic benzyl-N-bond was cleaved simultaneously. Therefore the N-(hydroxyethyl)amide of (3S)-12 was transformed into the enamide 27, which was hydrolyzed to afford the secondary amide 28. The enamide strategy to remove the chiral auxiliary from (3S)-11 led to complete racemization due to fast deprotonation in α-position of the cyano moiety. Two pairs of enantiomers 30a-b/ent-30a-b with prototypical σ substituents at the N-atom were prepared. The low σ1 affinity of the tetrahydro-2-benzazepines (ent-30b, Ki = 407 nM) is attributed to the short distance between the two lipophilic aromatic moieties.

5D-QSAR for spirocyclic sigma1 receptor ligands by Quasar receptor surface modeling

Based on a contiguous and structurally as well as biologically diverse set of 87 sigma(1) ligands, a 5D-QSAR study was conducted in which a quasi-atomistic receptor surface modeling approach (program package Quasar) was applied. The superposition of the ligands was performed with the tool Pharmacophore Elucidation (MOE-package), which takes all conformations of the ligands into account. This procedure led to four pharmacophoric structural elements with aromatic, hydrophobic, cationic and H-bond acceptor properties. Using the aligned structures a 3D-model of the ligand binding site of the sigma(1) receptor was obtained, whose general features are in good agreement with previous assumptions on the receptor structure, but revealed some novel insights since it represents the receptor surface in more detail. Thus, e.g., our model indicates the presence of an H-bond acceptor moiety in the binding site as counterpart to the ligands' cationic ammonium center, rather than a negatively charged carboxylate group. The presented QSAR model is statistically valid and represents the biological data of all tested compounds, including a test set of 21 ligands not used in the modeling process, with very good to excellent accuracy [q(2) (training set, n=66; leave 1/3 out) = 0.84, p(2) (test set, n=21)=0.64]. Moreover, the binding affinities of 13 further spirocyclic sigma(1) ligands were predicted with reasonable accuracy (mean deviation in pK(i) approximately 0.8). Thus, in addition to novel insights into the requirements for binding of spirocyclic piperidines to the sigma(1) receptor, the presented model can be used successfully in the rational design of new sigma(1) ligands.