3‐[3‐(Phenalkylamino)cyclohexyl]phenols: Synthesis, biological activity, and in silico investigation of a naltrexone‐derived novel class of MOR‐antagonists

The development of novel μ‐opioid receptor (MOR) antagonists is one of the main objectives of drug discovery and development. Based on a simplified version of the morphinan scaffold, 3‐[3‐(phenalkylamino)cyclohexyl]phenol analogs were designed, synthesized, and evaluated for their MOR antagonist activity in vitro and in silico. At the highest concentrations, the compounds decreased by 52% to 75% DAMGO‐induced GTPγS stimulation, suggesting that they acted as antagonists. Moreover, Extra‐Precision Glide and Generalized‐Born Surface Area experiments provided useful information on the nature of the ligand–receptor interactions, indicating a peculiar combination of C‐1 stereochemistry and N‐substitutions as feasibly essential for MOR–ligand complex stability. Interestingly, compound 9 showed the best experimental binding affinity, the highest antagonist activity, and the finest MOR–ligand complex stability. In silico experiments also revealed that the most promising stereoisomer (1R, 3R, 5S) 9 retained 1,3‐cis configuration with phenol ring equatorial oriented. Further studies are needed to better characterize the pharmacodynamics and pharmacokinetic properties of these compounds.


| INTRODUCTION
Despite the serious and potentially fatal adverse effects, μ-opioid receptor (MOR) agonists such as morphine, oxycodone, and fentanyl, have been over/misprescribed in recent years, resulting in a dramatic increase in opioid dependence, illegal opioid use, and opioid-related deaths. Opioid antagonists are a class of drugs that bind competitively to one or more of the opioid receptors, present little or no intrinsic activity, and robustly antagonize the effects of receptor agonists. They are utilized as antidotes for opioid overdose (naloxone) and approved for the treatment of opioid and alcohol dependence (naltrexone). Several structural classes have been identified as MOR antagonists with variable Arch. Pharm. 2023;356:e2200432.
wileyonlinelibrary.com/journal/ardp step by step, from natural opioid morphine structures to simplified ones.
With this in mind, in the present research, exploiting naltrexone as a structural prototype and applying a simple synthetic approach, we disclosed the 3- [3-(phenalkylamino) cyclohexyl]phenol core as the starting background to design new MOR antagonists ( Figure 1). In detail, we elected three crucial structural features to be present in the newly designed moieties: (1) a 3-hydroxyphenyl group, (2) an amino group substituted with a phenylethyl or phenylpropyl chain at a specified distance from the OH group, and (3) a lipophilic core structure combining these elements.
The phenylalkyl moiety has been selected as the N-substituent of choice, as it is present in a variety of morphans, piperidines, and other chemical scaffolds provided by MOR antagonist activity. [3,[7][8][9][10] Additionally, phenylalkyl substituent should increase lipophilicity to the rather small morphinan structure. The methyl group (R) in position 5 to the cyclohexane ring might be important as a conformation stabilizer, also affecting compound potency. In fact, its role as a modulator of MOR functional activity has already been reported for a series of trans-3,4-dimethyl-4-(3-hydroxyphenyl)piperidines. [11] To better elucidate its importance, derivatives without -CH 3 [12] generated in situ from methylmagnesium bromide to cyclohexenone 2, afforded the required 3-(3-methoxyphenyl)-3,5-dimethylcyclohexan-1-one 3.
Acylation of intermediates 4 and 5 with benzyl chloroformate in the presence of triethylamine has been used to ease the separation of products 6 and 7. Cleavage of aryl methoxy group of compounds 6 and 7 by boron tribromide proceeded with concomitant removal of N-Cbz group, providing target amines 8 and 9. Additionally, tertiary amines 10 and 11 have been prepared by reductive alkylation of compounds 8 and 9 respectively, with formaldehyde (Scheme 1). All the compounds were isolated as hydrochloride salts and structurally characterized by means of 1 H-NMR (nuclear magnetic resonance), 13 C-NMR, and HRMS (high-resolution mass spectrometry) spectroscopic methods (see the Supporting Information).

| Biology
The racemic 3-[3-(phenalkylamino)cyclohexyl]phenols 8,9,10,11,14, and 15 were then tested for their ability to displace specific Subsequently, to determine their potential ability as MOR antagonists, functional characterization of the synthesized compounds was performed using the GTPγS binding assay, a wellvalidated functional assay for G protein-coupled receptors, [13] using membranes from rat brain cortex.  [14,15] Each compound was docked as a chiral positively charged amine since at physiological pH the selected drugs are present as protonated forms. [16] Moreover, to get more reliable results in terms of binding affinities predictions, an MM-GBSA postprocessing method was applied. [17,18] In fact, although molecular docking methods provide excellent binding poses of the ligands within the receptor pocket, they, alone, often fail in predicting suitable binding affinities. [19][20][21] Ideally, the MM-GBSA scores should be compared to experimental binding affinities (K i ) and measurements taken from human opioid receptors. Docking, ligand-protein binding free energy, and experimental K i [22] data are reported in Table 3.
The superimposition of the lowest energy poses of the three antagonists in the MOR binding pocket is depicted in Figure 2. This helped us to identify the common crucial ligand-receptor areas of interaction.
Using the naltrexone molecule as a representative structure, we evidenced the relevant ligand-receptor interactions ( Figure 3 T A B L E 3 K i , Glide-XP, and MM-GBSA scores of BF0, naltrexone, and naloxone  Figure 4 and Table 4.  Table 4 shows that compound (1R, 3R, 5S) 9, with the lowest MM-GBSA score, exhibited the best MOR-ligand complex stability. Conversely, replacement of NH with N-CH 3 , appeared detrimental for 9 as the emerging derivative (1S, 3R, 5R) 11 showed the worst free binding energy score ( Table 4). As already observed for 10, the steric effect of the N-methyl group might be suggestive of the stereochemistry inversion at C-1 and C-5 and the 1,3-cis orientation.
Interestingly, compounds 9 and 11 best poses, although similarly placed within MOR, differed mainly in the orientation of either the phenyl alkyl and the phenol substituents that are now directed toward TMH 2 and TMH 6, respectively ( Figure 8).
Moreover, Figure 9 shows that compound 11, apart from the ionic and hydrogen bond with Asp 147, displayed only a water bridge between OH phenolic group and His 297.
Besides, to better explore the receptor conformational features involved in the deactivation mechanism, Glide-XP and MM-GBSA experiments were also carried out with the most promising (1R, 3R, 5S) 9 on the active form of MOR bound to the crystal morphinan agonist BU72.
Even here, we observed that the strong ionic contact with Asp 147 was crucial to locating the ligand inside the receptor as well as the hydrophobic interactions with receptor sub-pocket residues.
F I G U R E 7 Binding site molecular models of MOR/10 complex obtained with Glide-XP. Glide-XP, extra-precision Glide; MOR, μ-opioid receptor.
showed a novel π-π interaction between the phenol moiety and Tyr 148, not displaying the water bridge between OH group and Lys 233 anymore (Figures 10 and 11).
Thus, preliminary investigations carried out on our first 3-[3-       (11) To the solution of amines 8 or 9 (1 equiv) as free bases in MeOH (1 ml per 0.1 mmol) was added 37% aqueous formaldehyde (2 equiv) and stirred for 30 min, then NaBH 3 CN (3 equiv) was added. The reaction mixture was stirred at room temperature for 16 h, then it was concentrated under reduced pressure. To the residue was added aqueous NaHCO 3 , the mixture was extracted with EtOAc and the organic extract was dried over Na 2 SO 4 . After solvent evaporation, the crude was dissolved in 4 N HCl and purified by flash chromatography on the reversed-phase column to give compounds 10 or 11 as hydrochloride salts.

| Molecular modeling
The crystal structure of the complex «receptor-BF0 ligand» was used as the scaffold of choice to build the μ-opioid receptor virtual model and derived from 2.80 Å (4DKL) RCSB-PDB. Then, naltrexone has been used to validate our docking protocol. The crystal structures of (Mus musculus) MOR in the active form bound to the BU72 agonist (PDB ID, 5C1M) were obtained from the RCSB database at 2.07Å (5C1M).

CONFLICTS OF INTEREST
The authors declare no conflicts of interest.