Discovery of Fluorescence Polarization Probes and Inhibitors for Testis-Specific Bromodomain and Extra-Terminal (BRDT) Proteins

Abstract

The testis-specific bromodomain (BRDT) protein is a member of the Bromodomain and Extra-Terminal (BET) protein family. BRDT interacts with acetylated lysine (Kac) residues of histones and is important for the process of spermatogenesis. The expression of BRDT in humans is testis-restricted and BRDT-1 knockout mice are infertile. Therefore, the inhibition of BRDT is a promising strategy for male contraception. However, no selective BRDT inhibitors have been discovered yet. High sequence identity between BET proteins has impeded the discovery of selective BRDT inhibitors.In our research, novel probes were prepared for use in a fluorescence polarization assay. Probes were designed based on the potent pan-BET inhibitor SG3-179. A series of SG3- 179 analogs linked to the BODIPY fluorophore was synthesized and photo-physical properties and affinities were measured. Among these probes, 2.18 exhibited the best polarization signal and affinity against BET family proteins (Kd= 4–16 nM). The IC50 values that were obtained for published BET inhibitors correlated well with the value obtained using the AlphaScreen assay. Based on the knowledge that BRDT has a unique Arg54 in the Kac site, compounds were designed based on the SG3-179 scaffold with the goal to engage Arg54 in compound binding. Two binding poses of SG3-179 in the Kac binding site inspired us to generate two series of compounds. Negatively charged side chains were introduced for potential interaction with Arg54. Series I compounds contained a single substituted side chain attached to ring A of the SG3-179 scaffold. They were weak BET inhibitors (8.6 ~ 43 μM) in Alpha Screen and FP assays. Series II compounds with two side chains showed double digital nanomolar activity for BRDT-1 inhibition and up to 13-fold selectivity for BRDT- 1 over BRD4-1. As SG3-179 is also a JAK2 and FLT3 inhibitor the new analogs were also tested for inhibition of these two kinases. Compound 3.61 showed more than 400-fold selectivity for BRDT-1 over JAK2 or FLT3. The cocrystal structures of compound 3.61 revealed differences in the binding modes with BRD4-1 and BRDT-1. Series III compounds, which lack the piperidine moiety present in SG3-179, had decreased BRDT/BRD4 activity (0.33 ~ 2.70 μM). In the future, further modification of the most BRDT-1 selective compound 3.61 will be performed to improve BRDT selectivity. Bivalent SG3-179-derived analogs were designed and synthesized as an alternative strategy. We hypothesized that BRDT selectivity can be achieved through BRDT specific intramolecular or intermolecular binding. Through iterative optimization of chemotypes, attachment, spacer length and spacer position (N38 and C3 site), symmetric and non- symmetric bivalent inhibitors with PEG-linkers were prepared. Compared to the parent compound 2.11, two series of bivalent compounds showed up to 70-fold improvements in potency for BRDT-1 in the AlphaScreen assay. The best compound 4.17 exhibited 238- fold higher selectivity for BRDT-1 over BRD4-1. The selectivity ratio initially increased and then decreased with the elongation of the PEG linker in these two series of compounds. The ideal linker length varied in different molecule series. To achieve a better understanding of the binding mode, four cocrystal structures were obtained. We observed that bivalent molecules induced the dimerization of BRDT-1 and BRD4-1. However, for tandem BRD constructs, the inhibitory selectivity of these compounds decreased significantly. In cellular assays, these large molecules were able to penetrate the cell membrane and exhibited double digital nanomolar IC50 values for inhibition of multiple myeloma (MM1.S) cell proliferation.