Discovery and characterization of bromodomain 2–specific inhibitors of BRDT

Significance There is no nonhormonal contraceptive pill for men, although hundreds of genes have been identified to play roles during spermatogenesis and fertilization in the male reproductive tract. To address the absence of contraceptive drugs for men, we established a DNA-encoded chemistry technology (DEC-Tec) platform. Our drug discovery campaign on BRDT, a validated spermatogenic-specific contraceptive target, yielded rapid discovery of potent and specific inhibitors of the second bromodomain of BRDT that have unique binding characteristics to BRDT-BD2 relative to BRDT-BD1. Our study emphasizes the robustness and validation of the DEC-Tec platform where the obtained structure–affinity relationship data would allow us to identify specific protein binders immediately without performing exhaustive medicinal chemistry optimization of compounds with potential as male contraceptives.

2b) Thermal shift assay. The dye SYPRO Orange (ThermoFisher Scientific, USA) was used to perform the protein thermal shift assay. The assay was set up on a 384-well Roche plate where His-bromodomain at a concentration of 2 µM was incubated with the test compound at various concentrations, and SYPRO Orange dye at 5 × in a 10-µL reaction. The melting curve experiment and data analysis was run on a Roche Lightcycler 480 real-time PCR instrument.

2c) NanoBRET target engagement intracellular BET bromodomain assay. The NanoLuc-BRDT-BD1
and NanoLuc-BRDT-BD2 vectors were constructed by subcloning the same BRDT bromodomain sequences applied for protein expression into the NanoLuc-BRD4-BD1 and NanoLuc-BRD4-BD2 vectors (Promega, USA) to replace the corresponding BRD4 bromodomain sequences. The NanoBRET tracer competition assay was performed in transiently transfected HEK293 cells expressing each NanoLucbromodomain vector on a 384-well plate following the manufacturer's protocol (Promega, USA). Tracer titration was performed for each NanoLuc fusion vector to determine the optimized tracer concentration. The phusion protein was allowed to express for 36 h. The cells were then preincubated with tested compounds at different concentrations for 2 h followed by 2 h of incubation with tracer. Freshly prepared NanoBRET Nano-Glo substrate plus extracellular NanoLuc inhibitor were then added to initiate the subsequent bioluminescence resonance energy transfer (BRET) measurements using a CLARIOstar Plus BMG LABTECH plate reader. Data analysis was done by measuring the ratio of acceptor emission to donor emission (BRET ratio) and normalized by subtracting no-tracer-control-background.
2d) Metabolic stability assay in liver microsomes. Compounds 3-10, 15, 17 and 20-22 (2.0 μM) were incubated in the mouse or human liver microsomes (0.5 mg protein/mL) at 37 °C. The samples are collected at specific time-points 0, 5, 10, 20, 40 and 60 min. The reactions are terminated by adding equivalent volume of ice-cold CH 3OH and vortexed. The reaction mixtures are centrifuged at rcf 15,000 for 15 min. Five μL of the supernatant was analyzed by UHPLC-Q Exactive Orbitrap MS (Thermo Fisher Scientific, USA) equipped with 50 mm × 4.6 mm column (XDB C-18, Agilent Technologies, USA). The column temperature was maintained at 40 °C. The flow rate was at 0.3 mL/min with a 30% mobile phase (acetonitrile containing 0.1% formic acid). Q Exactive MS was operated in positive or negative mode with electrospray ionization. Ultra-pure nitrogen was applied as the sheath (45 arbitrary unit), auxiliary (10 arbitrary unit), sweep (1.0 arbitrary unit) and the collision gas. The capillary gas temperature was set at 275 °C and the capillary voltage was set at 3.7 kV. MS data was acquired from 80 to 1,200 Da in profile mode.
2e) BROMOscan bromodomain profiling. BROMOscan bromodomain profiling was provided by Eurofins DiscoverX Corp. (San Diego, CA, USA, http://www.discoverx.com). Determination of the Kd between test compounds and DNA tagged bromodomains was achieved through binding competition against a proprietary reference immobilized ligand.

Crystallography
BRDT-BD2 was co-crystallized with CDD-1102 or CDD-1302 by hanging drop vapor diffusion method. For crystallization, the purified BRDT-BD2 at 10 mg/mL was mixed with 2 and 4 molar excess of CDD-1102 and CDD-1302, respectively. The protein-inhibitor mixture was concentrated using Amicon Ultra-15 centrifugal filters (Millipore Sigma) to 30 mg/mL. Mosquito (TTP labtech) was used to dispense equal volumes of protein and reservoir (250 nL each) against 70 μL reservoir buffer in 96 wells crystallization tray (96-Well clear polystyrene microplate from SPT labtech). For the CDD-1102 complex, crystals were observed after 10 days in drop containing 0.8 M succinic acid (pH 7.0), whereas for the CDD-1302 complex, crystals were observed after 15 days in drop containing 1.1 M sodium malonate 0.1 M, HEPES pH 7.0, 0.5 % v/v Jeffamine ED-2001. The diffraction data were collected at Advanced Light Source (ALS), Beam Line 5.0.2 (UC Berkeley, USA) at wavelength (λ) = 1.0000 Å, by using an ADSC Q210 CCD detector. The data were integrated and scaled by using iMosflm and SCALA, respectively (2,3). The crystal structures of the CDD-1102 and CDD-1302 complexes were determined by molecular replacement (4) using crystal structure of the human C-terminal bromodomain BRD2 (PDB ID: 4UYG) and the determined structure of the CDD-1102 complex as search models. The final models have gone through several rounds of refinement using phenix.refine (5) followed by manual model building using COOT (6). For all structural analysis and figure preparation, the visualization program PyMOL was used (7). The data collection and refinement statistics are summarized in Table S2.

Step 2i: N-(2-Chloro-4-methoxyphenyl)-1-methyl-3-(1-methyl-1H-indol-2-yl)-1H-indazole-5carboxamide (36)
The same procedure for the synthesis of 29 (step 2b) was followed by using 2-chloro-4-methoxyaniline as the amine. The residue was purified by reverse-phase column chromatography on C-18 (CH3CN/water, 5:95 to 100:0) to afford title compound 36 (    The modeled pose was obtained through an extensive small molecule docking study using a homology model of human BRDT-BD2 containing the ZA channel waters of bromodomains (shown as red dots), as these water molecules frequently play a role in binding inside the KAc pocket. To our surprise, the aniline -NH2 group of both CDD-1102 and CDD-1302 successfully replaced a ZA channel water, which allowed the ligands to penetrate into the KAc pocket very deeply. This water molecule usually bridges the conserved Y308 (BRDT-BD2 numbering) and the KAc of the acetylated histone or small molecule BET inhibitors. The computational modeling was performed using the Schrodinger Suite of Programs (8). Table S1. Bromodomain constructs expressed and purified for the DEC-Tec selection and assay validation. All constructs were subcloned into bacterial vectors pET15b or pET28b (addgene) and were purified on Talon cobalt resins. The final buffer was exchanged to 50 mM HEPES pH 7.5, 150 mM NaCl, 1 mM TCEP and 10% glycerol on PD-10 columns (Amersham Biosciences).