Design, synthesis and characterization of novel gamma‐aminobutyric acid type A receptor ligands

Antinociceptive ligand HZ-166 is a GABAA 2/3 receptor subtype-selective potentiator. It has been shown to exhibit anxiolytic-like effects in rodent and rhesus monkeys, as well as reduced sedative/ataxic liabilities. In order to improve the metabolic stability of HZ-166, the ethyl ester moiety was bioisosterically replaced with 2,4-disubstituted oxazoles and oxazolines. The new analogs of HZ-166 were synthesized, characterized, and evalutated for their biological activity and docked in the human full-length heteromeric α132L GABAA receptor subtype CyroEM structure (6HUO). Importantly no sedation nor ataxia was observed on the rotorod for LKG-I-70 (6) or KPP-III-51 (6c) at 100 and 120 mg/kg, respectively. There was also no loss of righting response for either ligand.


Introduction
Benzodiazepines (BDZs) act on gamma-amino-butyric acid A (GABA A ) receptor ion channels and allosterically potentiate the influx of chloride ions through the channel, imparting a hyperpolarized state to the neuron. The pharmacological action exerted by a BZD is dependent on the discrete subunits of the receptor complex. Convergent evidence from transgenic animals and molecules with selectivity for the proteins required for ligand gating has suggested that α 1  2/3  2 comprised GABA A receptors mediate the tolerance and sedative/ataxic effects of the drugs, whereas the α2 and α3 GABA A receptors mediate the anticonvulsant anxiolytic and antinociceptive effects of ligands distinct from sedative/ataxic effects. 1-3 A compelling clinical opportunity exists in the development of selective α1-sparing subtype GABA A receptor ligands. These new ligands are expected to result in superior treatments for seizures and anxiety without causing amnesia, sedation, ataxia, or the propensity for addiction/dependence.
The five membered heterocyclic oxazolines 5-5d were also employed as bioisosteric replacement of carboxylic acid esters. 17 The -hydroxy amides 2a-2d reacted with thionyl chloride in dichloromethane to provide -halo amides (not shown). The crude -halo amides were then heated with sodium hydroxide in ethanol to afford the cyclized five-membered oxazolines 5-5d in 75-85% yield (Scheme 3). [18][19] The optical rotation of the chiral oxazolines 5a-5d were measured. The chiral oxazolines 5c and 5d were also evaluated by chiral HPLC to detect any racemization. Examination of the data obtain indicated that there was no racemization (Supplementary material).

Molecular modeling
To determine if the ester replacement in 4, 4a, 4c, 5-5d, 6 and 6c could undergo a similar binding pose as HZ-166, molecular docking was performed using AutoDock Vina 1.5.6. 22 A recently published CryoEM structure of the human full-length α1β3γ2L GABA A receptor ion complex with alprazolam (PDB: 6HUO) by Masiulis S. et al 23 was used. The molecular docking scores (binding affinity) of the designed ligands were determined and noted ( Table 1). The binding poses of all ligands described here was found to be similar to HZ-166 (not shown). Illustrated in Figure 1 is the overlay of the binding pose of oxazole 6c and HZ-166. The binding affinities (based on docking scores) of all ligands was found to be within 1 kcal/mol difference from the lead compound HZ-166 (-9 kcal/mol), which suggested all bioisosteres would bind with the receptor complex with a similar affinity to the lead compound ester HZ-166. The halogen substituted benzodiazepines e.g. diazepam and alprazolam undergo a halogen bond interaction with the carbonyl oxygen of the backbone of the 1His102 amino acid in the CryoEM structure ( Figure S1 in Supplementary material). Examination of the docking of 8-bromo substituted bioisosteres 4, 4a, 4c and 5-5d shows a similar halogen bond interaction with the carbonyl oxygen of the backbone of the 1His102 amino acid, but most of the docking software (including AutoDock Vina) 24 does not include halogen bonding in their scoring functions and, therefore, are unable to successfully predict the correct docking score of such complexes. Therefore, it was expected the ligands with bromine at position 8 would bind stronger with the α1β3γ2L GABA A receptor complex, as compared to the 8-ethinyl substituted ligands (6 and 6c) and would exert increased motor side effect. A general method for the molecular modeling is provided in the Supplementary material. -9.1 5d -9.5 5 -9.6 6 -9.9 5a -10 6c -9.6

Rotarod study
The ethinyl analog 6c was evaluated for potential motor side effects. The motor/sensory study was carried out in CFW mice on a rotating rod (rotarod). The experiment was conducted by placing mice on the rotarod for a maximum of 3 minutes after oral administration of 6c at doses of 40, 80 and 120 mg/kg. The mice were also observed for loss of righting reflex, an indication of undesired CNS effects. The mice exhibited no sedation nor ataxia, or loss of righting reflex; in contrast diazepam (5mg/kg) significantly impaired rotarod performance ( Figure 2). The ligand (6c) exhibited sensorimotor steadiness at all three time-points, which indicated no sedative/ataxic effects ( Figure 2). The rotarod experiment was performed according to the previously published protocol. 25 In addition, LKG-I-70 (6) was assayed on the rotarod and no sign of sedation nor ataxia was observed even up to 100 mg/kg. There was also no loss of righting response. Based on this early data both 1,3-oxazoles have potential clinical use. There is much work to be done to determine if this holds up in other cases.

Figure 2.
The rotarod study in CFW mice.

Conclusion
The 2-substituted novel oxazole and oxazoline analogs of α2/α3-subtype selective imidazodiazepine HZ-166 were synthesized and characterized. The new chiral oxazolines were analyzed for any racemization by measuring the optical rotation, as well as chiral HPLC; no racemization was detected. These new analogs (especially 6 and 6c) should be more metabolically stable than HZ-166 with potentially less adverse effects and provide better clinical candidates. Examination of the molecular docking study suggests these molecules can bind with the α1β3γ2L GABA A receptor in a similar pose as HZ-166 but not as tightly. Ligands 6 and 6c" were also evaluated for motor side effects (rotarod) and neither exhibited sedative nor ataxic effects. Therefore, 2substituted oxazole and oxazoline bioisosteres of ethyl ester HZ-166 appear to be novel targets for the potential treatment of epilepsy, anxiety and neuropathic pain, by acting via selective positive allosteric amplification of GABA A signaling via α2/α3-containing GABA A receptors. From examination of the data further modification of 8-bromo oxazolines 5-5d to 8-ethinyl oxazolines is suggested and ongoing. We are also in the process of analyzing these new analogs in various animal models (anxiolytic, anticonvulsant and antinociception).

Experimental Section
General: For all reactions oven-dried round-bottom flasks or screw-cap test tubes were used, unless otherwise specified. All chemicals were purchased from commercial suppliers and purified by standard methods, if required. For all organic reactions anhydrous solvents were employed unless specified.
The aq layer was extracted with ethyl acetate (25 mL x 2). The combined organic layer was washed with brine (20 mL) and dried (Na 2 SO 4 ). The solvents were removed under reduced pressure and the residue was purified

2-(8-Bromo-6-(pyridin-2-yl)-4H-benzo[f]imidazo[1,5-a][1,4]diazepin-3-yl)-4,5-dihydrooxazole (5).
The carboxamide 2 (500 mg, 1.2 mmol,) was dissolved in CH 2 Cl 2 (50 mL) and the solution was cooled to 0 °C. Thionyl chloride (837 mg, 7.2 mmol) was added to the solution dropwise and then the reaction mixture was allowed to stir at rt for 16 h. After completion of the process the reaction mixture was cooled to 0 °C and a sat. aq solution of NaHCO 3 (100 mL) was added dropwise. The solution was allowed to stir for 30 min at 0 °C. The organic layer was separated and the aq layer was extracted with CH 2 Cl 2 (3 × 10 mL). The combined organic layer was dried (Na 2 SO 4 ). The solvent was evaporated, and a white solid was obtained directly for the next step without further purification. The white solid was dissolved in in ethanol (25 mL) and NaOH (300 mg, 7.5 mmol) was added. The solution obtained was refluxed for 2 hours. The reaction mixture was cool to rt and ethanol was removed under reduced pressure to afford a solid residue. The solid residue was dissolved in CH 2 Cl 2 (100 mL) and washed with a saturated aq solution of NaHCO 3 (100 mL). The organic layer was separated and an aq layer was extracted with CH 2 Cl 2 (3 × 100 mL). The combined organic layer was dried (Na 2 SO 4 ). The organic solvent was removed and the solid residue obtained was purified by crystallization (5% MeOH in EtOAc) to afford compound 5 as white colored needle shaped crystals (407 mg, 85%). mp 230-231 °C (crystallized from 5% MeOH in EtOAc). 1  The oxazoline 5a was prepared from 2a by following the same procedure described for 5 from 2. The chloroalkane intermediate was purified by crystallization (15% DCM and hexane). The chloroalkane obtained was a light yellow colored solid. The ethyl oxazoline 5a, so obtained, was purified by crystallization using 5% MeOH in EtOAc to afford compound 5a as white colored needle shaped crystals (230 mg, 82%