WO2024206937A2 - Bi-aryl methylated lactam derivatives and related compositions and method of use to inhibit the proliferation of cancerous cells - Google Patents

Abstract

A compound of formula (I): wherein R is formula (II) or formula (III); a composition comprising same; and a method of inhibiting the proliferation of cancerous cells in a patient.

Description

  BI-ARYL METHYLATED LACTAM DERIVATIVES AND RELATED COMPOSITIONS AND METHOD OF USE TO INHIBIT THE PROLIFERATION OF CANCEROUS CELLS CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application claims priority to U.S. provisional patent application no.63/455,855, filed March 30, 2023. STATEMENT OF GOVERNMENT SUPPORT [0002] This invention was made with government support under contract CA207532 awarded by the National Institutes of Health. The government has certain rights in the invention. TECHNICAL FIELD [0003] The present disclosure relates to bi-aryl methylated lactam derivatives, compositions comprising same, and a method of using same to inhibit the proliferation of cancerous cells. BACKGROUND [0004] Chromatin remodelers are multi-subunit epigenetic regulators that modulate the accessibility of DNA. The BRG1-associated factors (BAF) complexes function as ATP- dependent chromatin remodelers and consist of three biochemically distinct complexes: canonical BAF (cBAF), polybromo-associated BAF (PBAF), and GLTSCR1/like-containing BAF (GBAF or ncBAF) (Fig.1A). All three types of BAF complexes share the ATPase and several core subunits, but also contain unique subunits. BAF complexes are the most frequently mutated chromatin regulatory complex in cancer, and different subunits, in cooperation with the catalytic subunits BRM and BRG1, play crucial roles in regulating chromatin accessibility (see Schick et al., Nat Genet 53 (3), 269–278 (2021). Multiple compounds have been developed to target a limited number of BAF complex subunits, including small molecules and proteolysis targeting chimeras (PROTACs); however, most of the subunits have yet to be explored as targets (see, e.g., Shishodia et al., J Med Chem.65(20):13714-13735 (2022)(PBRM1 subunit); Farnaby et al., Nat Chem Biol 15 (7), 672–680 (2019)(BRG1 (SMARCA4) and BRM (SMARCA2) subunits). 1     [0005] In the human proteome, there are forty-two bromodomain-containing proteins that are predicted to recognize acetylated lysines in proteins, most commonly histones. Bromodomains (BDs) are therapeutic targets in multiple diseases, and several BD inhibitors are currently under phase I, II, or III clinical studies (see, e.g., Cochran et al., Nat Rev Drug Discov 18 (8), 609–628 (2019). In BAF complexes, unique bromodomain-containing subunits, such as bromodomain- containing protein 7 (BRD7) in PBAF or its homolog bromodomain-containing protein 9 (BRD9) in GBAF, are likely to mediate sub-complex specific function. BRD7 contains a single BD in its structure that belongs to the family IV of BDs (Fig.1B). BRD7 has been linked to the advancement of several types of tumors such as nasopharyngeal carcinoma, osteosarcoma, and colorectal, breast, ovarian, and prostate cancer (PCa), as well as being involved in the regulation of immune response. Even though BRD7 has been implicated in several disease-related roles, there are currently no selective chemical probes to study its potential as a therapeutic target. BRD9/7 inhibitors (Fig.1C), which are selective for BRD9 (with 73.2% sequence identity to BRD7) over BRD7, have been reported. [0006] In view of the above, it is an object of the present disclosure to provide inhibitors of BRD7. It is another object to provide a method of using such inhibitors to inhibit the proliferation of cancerous cells. These and other objects and advantages, as well as inventive features, will be apparent from the detailed description provided herein. SUMMARY Provided is a compound of formula I: acceptable salt thereof.

[0007] Also provided is the compound of formula I, or a pharmaceutically acceptable salt thereof, wherein R is selected from the group consisting of: 2    

[0008] Further provided is a composition comprising the compound and a pharmaceutically acceptable carrier, diluent, or excipient. Still further provided is a method of inhibiting the proliferation of cancerous cells in a patient. The method comprises administering to the patient the compound, or composition comprising same, in an amount effective to inhibit cancerous cell proliferation. The patient can have prostate cancer, such as androgen receptor-positive (AR+) prostate cancer. In other embodiments, the patient can have nasopharyngeal carcinoma, osteosarcoma, colorectal cancer, hormone-dependent breast cancer, or hormone-dependent ovarian cancer. BRIEF DESCRIPTION OF THE FIGURES [0009] Fig. 1A illustrates the composition of each of the three BAF subcomplexes: canonical BAF (cBAF), polybromo-associated BAF (PBAF), and GLTSCR1/like-containing BAF (GBAF or ncBAF), including bromodomain-containing protein 7 (BRD7) in PBAF and bromodomain- containing protein 9 (BRD9) in GBAF. [0010] Fig. 1B illustrates a phylogenetic bromodomain tree as described by Filippakopoulos et al., Cell 149 (1), 214–231 (2012). [0011] Fig. 1C shows the chemical structures of: (I) reported BRD7/9 ligands BI-9564, BI-7273, TP472, LP99, and I-BRD9 (with selectivity for BRD9 over BRD7). [0012] Fig. 2A shows a comparison of cartoon illustrations generated using PyMOL of the BD binding pockets of BRD7 (left) and BRD9 (right) crystallized with BI-9564. The comparison suggests BRD7 may allow for a larger hydrophobic moiety compared to BRD9. 3     [0013] Fig.2B is an illustration generated using PyMol identifying binding residues of the BD binding pocket of BRD7. [0014] Figs.3A and 3B provide histograms presenting respective results from thermal shift assays (TSA) for HIS-tagged BRD7 and BRD9 BDs. For each assay, the HIS-tagged BDs were incubated with the compounds at 25, 10 or 1 µM. The Tm of the proteins was calculated based on differential scanning fluorimetry readings at increasing temperatures from four replicates using nonlinear least squares fit on GraphPad Prism 9. The shift in T_m was calculated with respect to vehicle control. [0015] Figs.4A-C illustrate the structure of BI-FAMa (A) and polarization changes of BI-FAMa with increasing concentrations of BRD7 (B) and BRD9 (C), where approximate K_d values were obtained from non-linear One Site – Specific binding fit analyses on GraphPad Prism 9.. [0016] Figs.4D-F illustrate the structure of BI-FAMb.(D) and the polarization changes of BI- FAMb with increasing concentrations of BRD7 (E) and BRD9 (F), where approximate K_d values were obtained from non-linear One Site – Specific binding fit analyses on GraphPad Prism 9. [0017] Fig.5A-5C provides binding mode analysis of respective BI7273, 1-78, and 2-77 compounds docked against BRD7 (PDB: 6V1E) and BRD9 (PDB: 5EU1) performed in the Schrödinger® Maestro suite, and illustrated in PyMOL, in which, in 6A,BI7273, was docked into BRD7 BD (left) and BRD9 BD (right); in 5B,1-78, was docked into BRD7 BD (left) and BRD9 BD (right); and, in 5C, 2-77, was docked into BRD7 BD (left) and RD9 BD (right. The conserved Asn involved in the hydrogen bond and the Tyr involved in the π-π interactions are shown in sticks, the core of BI7273 is shown in a dotted pattern, the core of 1-78 and 2-77 in hash pattern, π-π interactions depicted in dotted lines, and hydrogen bonds depicted in dashed lines, with distance of the interactions given in Å. [0018] Figs.6A and 6B show binding interaction of the 1-78 and 2-77 compounds, where Fig. 7A shows a TREEspot^TM interaction map for each of 1-78 and 2-77 screened in the BROMOscan^TM platform, and where the results for binding interactions for the compounds are reported as % of control (DMSO); and Fig.6B provides a histogram presenting the results from a thermal shift assay using HIS-tagged BRPF1 and the GSK-5959 ligand and 1-78 and 2-77 each at 25, 10 or 1 µM, where the Tm of the proteins in the different conditions was calculated based on differential scanning fluorimetry readings at increasing temperatures from four replicates 4     using nonlinear least squares fit on GraphPad Prism 9, and the shift in Tm was calculated with respect to vehicle. [0019] Figs.7A-7C show cartoon illustrations created in PyMOL of different vantages of the 2- 77 docked against BRD7, where Fig.7A shows the surface of the protein and depicts the amino acids involved in the hydrophobic region right outside the binding pocket as sticks; Fig.7B shows the binding pose of 2-77 docked against BRD7 aligned to the binding pocket of BRD9, showing the surface of the BRD9 and depicting the amino acids involved in the hydrophobic region right outside the binding pocket as sticks, and indicating by an arrow where 2-77 clashes with the binding pocket of BRD9; and Fig.7C shows the binding pose of 2-77 docked against BRD7 aligned to the binding pocket of BRPF1, showing the surface of the BRPF1 and depicting the amino acids involved in the hydrophobic region right outside the binding pocket as sticks, and indicating by a pair of arrows where 2-77 clashes with the binding pocket of BRPF1. [0020] Figs.8A-8C show plotted results of In cellulo assays for the BRD9 inhibitor B1-7273 and the 1-78 and 2-77 compounds, respectively. Each shows NanoBRET assay results treating HEK293T cells treated for two hours with increasing concentrations of each compound, where the BRET ratio in milliBRET units (mBu) was calculated as described in the experimental section, and the IC₅₀ was calculated using nonlinear least squares fit on GraphPad Prism 9. [0021] Figs.9A-9C show the results of in vitro BRD7 knock-down experiments in RWPE-1, LNCaP, and PC-3 cells, where in Fig.9A normalized fold increase in growth is demonstrated after four days of incubation of RWPE-1, LNCaP, and PC-3 cells in which BRD7 has been knocked down with two different shRNA constructs, and cell viability was measured with a CellTiter-Glo® Luminescent Cell Viability Assay at day 0 and day 4. and includes three replicates each from the three separate experiments, and in Figs.9B and 9C, respective LNCaP and PC-3 viability is demonstrated after four days of incubation cells with the BRD9 inhibitor BI-7273, and compounds 1-78 and 2-77, using three replicates each from three separate experiments, where cell viability was measured employing a CellTiter-Glo® Luminescent Cell Viability Assay, error bars in represent s.d. n = 9, and statistical significance determined using multiple t tests with respect to empty vector or vehicle, respectively, and * p = <0.01, ** p = <0.001, *** p = <0.0001, **** p = <0.00001. [0022] Figs.10A-H show the results of in vitro gene expression analysis in RWPE-1, LNCaP, and PC-3 cells. Fig.10A shows a volcano plot of RNA-seq gene expression changes with 72- 5     hour treatment of LNCaP cells with 1 µM of the 2-77 compound compared to treatment with DMSO. The log2 FC indicates the mean expression level for each gene. Each dot represents one gene. The differentially expressed genes are indicated with p_adj<0.05 and FC > 1.5. Fig.10B identifies the top six most significantly enriched pathways represented in the genes that are significantly decreased with 2-77 treatment (left) and significantly increased with 2-77 treatment (right). Fig.10C graphs the results of GSEA analysis of RNA-Seq data from LNCaP cells treated with 2-77 using the MsigDB pathway Hallmark_Androgen_Response. Fig.10D shows a volcano plot of RNA-seq gene expression changes after 72-hour treatment of LNCaP cells with 10 µM enzalutamide (ENZA) or DMSO. The log2 FC indicates the mean expression level for each gene. Each dot represents one gene. The differentially expressed genes are indicated with padj<0.05 and FC > 1.5. Fig.10E illustrates the overlap of differentially expressed genes in LNCaP cells treated with 2-77 or ENZA, and (right) the correlation of all gene expression changes in cells treated with ENZA (x-axis) or 2-77 (y axis). Fig.10F shows a volcano plot of RNA-seq gene expression changes with 72-hour treatment of LNCaP cells with 0.2 µM ACBI, 2 µM BD98, 0.5 µM dBRD9, or 10 µM PB16 compared to DMSO. The log2 FC indicates the mean expression level for each gene. Each dot represents one gene. The differentially expressed genes are indicated in (ACBI1), (BD98), (dBRD9), or (PB16) with p_adj<0.05 and FC > 1.5. Fig. 10G illustrates the overlap of differentially decreased (padj < 0.05, FC > 1.5) genes after ACBI1 2-77 and dBRD9, BD98 or PB16. Fig.10H (left) illustrates the overlap of differentially expressed genes in LNCaP cells treated with 2-77 or PB16, and (right) the correlation of all gene expression changes in cells treated with 2-77 (x-axis) or PB16 (y-axis). [0023] Fig.11 provides a histogram presenting respective results from thermal shift assays (TSA) for binding affinity of Core 3-based 277.1, 277.2, 277.3, 277.4, 277.5, 277.6, 277.7, 277.7, and 277.9 for HIS-tagged BRD7 and BRD9 BDs. For each assay, the HIS-tagged BDs were incubated with the compounds at 25 µM. The Tm of the proteins was calculated based on differential scanning fluorimetry readings at increasing temperatures from four replicates using nonlinear least squares fit on GraphPad Prism 9. The shift in Tm was calculated with respect to vehicle control. [0024] Figs.12A-C present results from competitive fluorescence polarization (FP) assay, where, in Fig.12A, results of FP of BI0FAMb with increasing concentrations of BRD7 are 6     presented; and, in Fig.12B, results of competitive FP of Core 3-based 277.3, 277.4, 277.5, and 277.8, alongside 2-77 and BI-7273 controls, are presented. [0025] Figs.13A and 13B present respective LNCaP and PC-3 viability is demonstrated after four days of incubation cells with 277.3, 277.4, 277.5, 277.8 and BI-7273 and 2-77 controls using three replicates each from three separate experiments, where cell viability was measured employing a CellTiter-Glo® Luminescent Cell Viability Assay, error bars in represent s.d. n = 9, and statistical significance determined using multiple t tests with respect to empty vector or vehicle, respectively, and * p = <0.01, ** p = <0.001, *** p = <0.0001, **** p = <0.00001. DETAILED DESCRIPTION [0026] The present disclosure is predicated, at least in part, on the discovery of an open hydrophobic region adjacent to the acetylated lysine binding pocket of BRD7, which is absent in BRD9. Based on this discovery, a first set of compounds were designed to fit in this uniquely accessible binding region while binding to the acylated Lys binding pocket to achieve selectivity. Two such BRD7 inhibitors, 1-78 and 2-77, show high affinity for the protein in thermal shift assays. Moderate selectivity was observed when screening the compounds via competitive fluorescence polarization (FP) against BI-FAM, a fluorescent probe that binds to both BRD7 and BRD9 in a nanomolar range. Binding mode analyses show that, while fitting in the hydrophobic region in BRD7, 1-78 and 2-77 maintain key interactions with the asparagine and tyrosine residues critical for acetylated lysine binding. The utility and selectivity of the 1-78 and 2-77 compounds were validated in cell-based models of prostate cancer; the compounds were also shown not to affect BRD9 chromatin engagement. [0027] In view of the above, provided is a compound of formula I: 3)

a pharmaceutically acceptable salt

7     [0028] Moreover, based on positive results obtained for 1-78 and 2-77, a set of analog compounds with a shared Core 3 scaffold was designed to fit in the unique BRD7 binding region while also binding to acylated Lys binding pocket and fit in the unique BRD7 binding region to achieve selectivity. Four such BRD7 inhibitors, 277.3, 277.4, 277.5, 277.8, show high affinity for the protein in thermal shift assays. Moderate selectivity was observed when screening the compounds via competitive fluorescence polarization (FP) against BI-FAM. Analysis using Glide XP docking software predicted that, while fitting in the hydrophobic region in BRD7, 277.3, 277.4, 277.5, and 277.8 maintain key interactions with the asparagine and tyrosine residues critical for acetylated lysine binding. The utility and selectivity of 277.3, 277.4, 277.5, and 277.8 were validated in cell-based models of prostate cancer. [0029] In view of the above, provided is a compound of formula I: R (core 3) or a pharmaceutically acceptable

R is selected from the group consisting of:  

  [0030] The above compounds include isotopic variants and compounds in which one or more hydrogen atoms have been substituted with deuterium. The compounds may contain one or more chiral centers or may otherwise be capable of existing as multiple stereoisomers. In one embodiment, the compounds are not limited to any particular stereochemical requirement, and the compounds, and the compositions, methods, uses, and medicaments that include them, may be optically pure, or may be any of a variety of stereoisomeric mixtures, including racemic and other mixtures of enantiomers, other mixtures of diastereomers, and the like. Such mixtures of stereoisomers may include a single stereochemical configuration at one or more chiral centers, while including mixtures of stereochemical configurations at one or more other chiral centers. [0031] Similarly, the compounds may include geometric centers, such as cis, trans isomers, diastereomers, enantiomers, and E and Z double bonds. In another embodiment, the compounds are not limited to any particular geometric isomer requirement, and the compounds, and the compositions, methods, uses, and medicaments that include them, may be pure, or may be any of a variety of geometric isomer mixtures. Such mixtures of geometric isomers may include a single configuration at one or more double bonds and chiral carbons, while including mixtures of geometry at one or more other double bonds and chiral carbons. [0032] The terms “salts” and “pharmaceutically acceptable salts” refer to derivatives of the compounds wherein the parent compound is modified by making acid or base salts thereof. Examples of pharmaceutically acceptable salts include, but are not limited to, mineral or organic acid salts of basic groups such as amines; and alkali or organic salts of acidic groups such as carboxylic acids. Pharmaceutically acceptable salts include the conventional non-toxic salts or the quaternary ammonium salts of the parent compound formed, for example, from non-toxic inorganic or organic acids. For example, such conventional non-toxic salts include those derived from inorganic acids such as hydrochloric, hydrobromic, sulfuric, sulfamic, phosphoric, and nitric; and the salts prepared from organic acids such as acetic, propionic, succinic, glycolic, stearic, lactic, malic, tartaric, citric, ascorbic, pamoic, maleic, hydroxymaleic, phenylacetic, glutamic, benzoic, salicylic, sulfanilic, 2-acetoxybenzoic, fumaric, toluenesulfonic, methanesulfonic, ethane disulfonic, oxalic, and isethionic, and the like. [0033] Pharmaceutically acceptable salts can be synthesized from the parent compound, which contains a basic or acidic moiety, by conventional chemical methods. In some instances, such salts can be prepared by reacting the free acid or base forms of these compounds with a 9     stoichiometric amount of the appropriate base or acid in water or in an organic solvent, or in a mixture of the two; generally, nonaqueous media like ether, ethyl acetate, ethanol, isopropanol, or acetonitrile are preferred. Lists of suitable salts are found in Remington’s Pharmaceutical Sciences, 17th ed., Mack Publishing Company, Easton, Pa., 1985, the disclosure of which is hereby incorporated by reference for its teachings regarding same. [0034] The term “solvate” means a compound, or a salt thereof, that further includes a stoichiometric or non-stoichiometric amount of solvent bound by non-covalent intermolecular forces. Where the solvent is water, the solvate is a hydrate. [0035] The above compounds, and pharmaceutically acceptable salts and solvates thereof, can be synthesized in accordance with methods known in the art and exemplified herein. [0036] Further provided is a composition comprising the compound and a pharmaceutically acceptable carrier, diluent, or excipient. The compounds can be formulated as pharmaceutical compositions using methods well-known in the art. “Carrier” is used generically herein to refer to pharmaceutically acceptable carriers, diluents, adjuvants, and excipients (see, e.g., Remington, the Science and Practice of Pharmacy, 23rd edition, Philadelphia, PA: Lippincott Williams and Wilkins, which is incorporated herein by reference). [0037] Still further provided is a method of inhibiting the proliferation of cancerous cells in a patient. The method comprises administering to the patient the compound, or composition comprising same, in an amount effective to inhibit cancerous cell proliferation. The patient can have prostate cancer, such as androgen receptor-positive (AR+) prostate cancer. In other embodiments, the patient can have nasopharyngeal carcinoma, osteosarcoma, colorectal cancer, hormone-dependent breast cancer, or hormone-dependent ovarian cancer. [0038] Any suitable route of administration can be used in the above methods. Examples include, but are not limited to, oral, parenteral, intravenous, intratumoral, and peritumoral. An effective amount can be determined by one of ordinary skill in the art using dosage range determining methods known in the art. Typically, a physician (or veterinarian for non-human subjects) will determine the actual dosage, which will be most suitable for an individual patient. The specific dose level and frequency of dosage for an individual may be varied and will depend upon a variety of factors including the activity of the specific compound employed, the metabolic stability and length of action of that compound, the age, body weight, general health, gender, diet, mode and time of administration, rate of excretion, other administered drugs, and the 10     severity of the particular condition. The compounds/compositions described herein can be administered with other biologically active compounds as appropriate. The compounds/compositions described herein can also be administered in combination with other cancer-treatment modalities, such as, but not limited to, chemotherapy, radiotherapy, and surgery. EXAMPLES [0039] The following examples serve to illustrate the present disclosure. The examples are not intended to limit the scope of the claimed invention in any way. Methods Binding site analysis [0040] The binding pockets of the BD of BRD7 bound to BI-9564 (PDB: 5MQ1) and the BD of BRD9 bound to BI-9564 (PDB: 5F1H) were visualized and compared in the molecular visualization software PyMOL version 2.5.2 (Schrödinger®). The same software was used to analyze the binding mode of the ligands in the published crystal structures of BI-9564 bound to BRD7 (PDB: 6V1F) and BRD9 (PDB: 5FH1); of I-BRD9 bound to BRD7 (PDB: 6V17) and BRD9 (PDB: 6V1B); of TP-472 bound to BRD7 (PDB: 6V1F) and BRD9 (PDB: 6V14); of Bromosporine bound to BRD7 (PDB: (6V1H) and BRD9 (5IGM); and of NI-48 bound to BRPF1 (PDB:5T4V). Docking studies [0041] All the molecular docking studies were performed in the Schrödinger® Maestro suite. In brief, the designated ligands were prepared employing LigPrep. For the binding pose analysis, a receptor grid was generated for the crystal structures of BRD7 bound to BI7273 (PDB: 6V1E), BRD9 bound to BI7273 (PDB: 5UE1), and BRPF1 bound to NI-48 (PDB: 5T4V). A docking screening was then performed with Glide with the precision set to SP and adding the Epik state penalties to the docking score. The top poses of each ligand for each BD were visualized and analyzed in PyMOL version 2.5.2 (Schrödinger®). Synthesis General method A: compounds 1-38, 1-70, 1-79, and 2-88 (Core 1) and 1-75 and 1-78 (Core 3) 11     [0042] The aryl bromide was dissolved in anhydrous DMF. Then Pd(dppf)Cl2, Cs2CO3, and the pinacol ester or boronic acid were added to the solution. The reaction was stirred overnight under reflux.  Compound Boronic acid/pinacol ester Aryl bromide 1-38 Br

12     1^-78 _Br   General

method B: fcompounds 2-79 (Core 1), and 2-77 and 2-81 (Core 3) [0042] The aryl bromide was dissolved in 1,4-dioxane anhydrous. Then Pd(dppf)Cl₂, Cs₂CO₃, and the pinacol ester were added to the solution. The reaction was stirred for 48 hours (2-79) or 72 hours (2-77 and 2-81) under reflux.   Compound Pinacol ester Aryl bromide  

General method C: compounds 2-43 and 2-63 (Core 2) and 2-70 and 2-71 (Core 4) 13     [0043] Step 1: The aryl aldehyde was dissolved in THF under argon and the reaction was cooled to 0 °C. Then, the Grignard reagent was added over 10 minutes. The reaction was then quenched. Step 2: After purification, the product from Step 1 was dissolved in DCM and pyridinium dichromate was added.  Compound Grignard reagent Aryl aldehyde 2-43 H O  

Protein purification [0044] Recombinant N-terminal His-Tev-tagged BRD7 BD (Addgene plasmid No.98245) and N-terminal His-Tev-tagged BRD9 BD (Addgene plasmid No.39012) were expressed in BL21(DE3) E. Coli cells with kanamycin containing-LB broth agar. Transfected cells were incubated in kanamycin containing-LB broth until OD₆₀₀ reached 0.6-1, then IPTG was added to a final concentration of 1 mM. Protein expression was induced for 16 hours at 20 °C. The bacteria were then pelleted by centrifugation at 4000 x g for 10 minutes at 4 °C and resuspended in His binding buffer (NaCl 150 mM, Tris 20 mM, pH = 8, Imidazole 25 mM, protease 14     inhibitors). Cells were lysed by sonication and centrifuged at 15,000 x g for 40 minutes at 4 °C. The supernatant was incubated for two hours with HisPur^TM Ni-NTA Resin equilibrated in binding buffer. The beads were washed three times with binding buffer, and the proteins were then eluted in His elution buffer (NaCl 150 mM, Tris 5 mM, Imidazole 500 mM, pH = 8). SDS- PAGE was used to confirm purity of the protein. Thermal shift assay based on differential scanning fluorimetry [0045] The TSA was performed following a previously reported protocol. (Vivoli et al., JoVE No. 91, e51809(2014), which is incorporated herein by reference). The reaction was run in 20 µL using a StepOnePlus Real-Time PCR System (Applied Biosystems). The reaction was set up in the following buffer: 10 mM HEPES pH = 7.0, 150 mM NaCl, 8X SYPRO Orange S6651 Invitrogen (5000X stock), 0.2 mg/mL HIS-tagged BRD7 or HIS-tagged BRD9 in elution buffer, 5% DMSO containing the inhibitors at designated concentrations. Melting curves were obtained using a temperature gradient of 25-75 ^C over 120 minutes with readings every 0.5 ^C. Melting curves for HIS-tagged BRD7 BD and HIS-tagged BRD9 BD were obtained for four replicates at each ligand concentration, and the T_m values were calculated using nonlinear least squares fit in GraphPad Prism 9. Fluorescent polarization (FP) assay [0046] HIS-tagged BRD7 and BRD9 BDs were purified as described above and dialyzed in the following reaction buffer: Tris 20 mM, NaCl 150 mM, 0.02% Tween 20, pH = 8.0. The assay was completed in 384-well plates (Greiner medium binding Fluotrac or Perkin Elmer Optiplate). The reaction volume was kept at 35 µL per well, and four replicates of each reaction were used. The plates were incubated covered for five minutes before reading them in a Synergy Neo2 HTS multimode microplate reader (Biotek) with a xenon flash lamp as light source. The excitation was set at 485/20 nm, and the emission was set at 528/20 nm. The gain was adjusted to a blank buffer or negative control well, and the fluorescent polarization estimates were determined from parallel and perpendicular intensities given in millipolarization (mP) values. [0047] The relative affinity of the BRD7 BD for the FAM-labeled probe was evaluated by direct binding as follows: starting with a concentration of 80 µM of BRD7 BD, two-fold dilutions were performed to a minimum concentration of 1.22 nM, while keeping the FAM-labeled probe concentration constant at 100 nM. The same procedure was used for the BRD9 BD but starting with a concentration of 20 µM to a minimum of 0.00954 nM. 15     [0048] Based on the relative affinity results, 5 µM of BRD7 BD and 250 nM of BRD9 BD were used for performing the competitive assay, and the FAM-labeled probe was kept constant at 100 nM. Starting with a concentration of 25 µM of ligand, two-fold dilutions were performed to a minimum concentration of 24.4 µM. The results of the assays were graphed in GraphPad Prism 9 and analyzed using a “One site – Specific binding” fit for the direct binding assay and a non- linear “[Inhibitor] vs response – Variable slope (four parameters)” fit for the competitive assay. [0049] K_i estimations based on the experimental IC₅₀ were calculated employing the equation K_i= [I]50/([L]50/Kd + [P]0/Kd + 1); where [I]50 is the concentration of the inhibitor at 50% inhibition, [L]₅₀ is the concentration of free FAM-labeled probe at 50% inhibition, K_d is the dissociation constant of the protein – FAM-labeled probe complex (obtained from the direct binding assay), and [P]0 is the concentration of the free protein at 0% inhibition. BROMOscan bromodomain profiling (bromoKdELECT, BromoMAX panel) [0050] BROMOscan^TM bromodomain profiling was provided by Eurofins DiscoverX Corp. (San Diego, CA, USA). Compounds 1-38, 1-78 and 2-77 were tested at 2000 nM in a BromoMAX panel; BI7273, 1-78, and 2-77 were tested in a concentration gradient starting at 10 µM in a bromoKdELECT assay. The results for binding interactions for the compounds are reported as % of control (DMSO). Cell lines [0051] HEK293T (RRID:CVCL_0063), LNCaP (clone FGC; RRID:CVCL_1379), PC-3 (RRID:CVCL_0035), RWPE-1 (RRID:CVCL_3791), DU145 (RRID:CVCL_0105), 22Rv1 (RRID: CVCL_1045), and C4-2 (RRID:CVCL_4782) cells were purchased from ATCC. PC-3 cells were cultured in F12K medium supplemented with 10% FBS, 100 U/mL penicillin and 100 g/mL streptomycin, 2 mmol/L L-alanyl-L-glutamine (Corning Glutagro) and 2.5µg/mL Plasmocin® (Invivogen). HEK293T cells were cultured in DMEM medium, 1 mM sodium pyruvate, and supplemented as above. LNCaP and 22Rv1 cells were cultured in RPMI-1640 with 1× MEM non-essential amino acids, 1 mM sodium pyruvate, 0.1 M HEPES (Cytiva), and supplemented as above. RWPE-1 cells were cultured in keratinocyte SFM (Gibco 17005–042, Thermo Fisher Scientific) supplemented with 0.05 mg/mL bovine pituitary extract (Gibco 13028-014), 0.005 µg/mL EGF human recombinant (Gibco 10450-013), 100 U/mL penicillin, 100 g/mL streptomycin and 2.5µg/mL Plasmocin® (Invivogen). DU145 and C4–2 cells were 16     cultured following ATCC suggestions supplemented with 100 U/mL penicillin, 100 g/mL streptomycin and 2.5µg/mL Plasmocin® (Invivogen). Lentiviral production [0052] 20,000,000 HEK293T cells were cultured overnight and transfected after 16 hours with pLKO.1 puro empty vector (Addgene plasmid #8453), shBRD7-1 (TRCN0000151186), shBRD7-2 (TRCN0000154102), or shPBRM1 (TRCN0000015994) knockdown constructs in addition to the packaging vectors pMD2.G (Addgene plasmid #12259) and psPAX2 (Addgene plasmid #12260). After 16 hours of incubation the media was changed. After 48 hours of incubation after media change the supernatant was ultracentrifuged at 17,500 x g for two hours at 4 °C. The pellet was then resuspended in 200 µL phosphate-buffered saline buffer and stored at -80 °C. Knockdown experiments [0053] 500,000 – 600,000 cells were seeded in 6-cm plates and incubated for 24 – 72 hours such that the cells were 50 – 80% confluent before adding 5 µL of lentivirus containing pLKO.1 puro empty vector, shBRD7-1, or shBRD7-2. Twenty-four hours after transduction, the cells were selected by treatment with puromycin (2 µg/mL) for 48 hours. Cells were then counted, seeded in a 96-well plate at a density of 5,000 cells per well and incubated for four days. Cell viability was measured with a CellTiter-Glo® kit at day 0 and day 4, and the fold increase in growth was calculated and reported relative to the empty vector control. [0054] To evaluate the efficacy of the constructs, cells were transduced with lentivirus containing pLKO.1 puro empty vector, shBRD7-1, or shBRD7-2, and selected for 48 hours with puromycin (2 µg/mL) after transduction. Cells were collected and resuspended in buffer A (0.05 mM EDTA, 25 mM HEPES, 5 mM MgCl₂, 10% glycerol, 25 mM KCl, 0.1% NP40) and incubated for 15 minutes on ice. The samples were then centrifuged at 600 x g for five minutes at 4 °C. The pellet was resuspended in buffer B (20 mM HEPES, 150 mM NaCl, 7.5 mM MgCl2, 1% Triton X-100) and incubated while rotating for 30 minutes at 4 °C. A cocktail of protease inhibitors was added to the buffers right before each step. The samples were then centrifuged at 10,000 x g for 10 minutes at 4 °C. The protein content of the supernatant of each sample was obtained with a Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific). The samples were prepared for immunoblotting by mixing them in a 1:3 ratio with a mixture of β-mercaptoethanol and 4X Bolt™ LDS Sample Buffer (Invitrogen) in a 1:9 ratio; and then boiling them at 95 °C for 15 17     minutes. Equal amounts of protein from the samples were loaded to a 4 –12% SDS- polyacrylamide gel (Invitrogen) for immunoblot analysis. Immunoblotting [0055] Cell lysate samples were denatured for at least 15 minutes at 95 °C, electrophoresed on 4 –12% SDS-polyacrylamide gels (Invitrogen), and transferred onto a Immobilon®-FL PVDF membranes (Millipore Sigma). The membranes were then blocked for 30 minutes in Immobilon® Signal Enhancer (Millipore Sigma) and stained overnight with primary antibodies. For secondary antibodies staining, the membranes were washed with tris-buffered saline buffer with 0.1% Tween-20 and incubated for one hour with infrared-dye labeled goat anti-mouse or anti-rabbit antibodies (LICOR Biotechnology). Images were obtained using an Odyssey Clx imager (LICOR Biotechnology). Antibodies [0056] BRD7 (B-8), Santa Cruz Biotechnology sc-376180 (1:250); TATA binding protein (TBP), Abcam ab818 (1:2,000); BRD9, Bethyl Laboratories A303-781A (1:1,000); GAPDH (6C5), Santa Cruz Biotechnology sc-32233 (1:500); Histone H3, Active Motif 39064 (1:10,000); β- actin, Santa Cruz Biotechnology sc-47778 (1:2000). Growth inhibition experiments [0057] Cells were seeded in a 96-well plate (no.655098, Greiner Bio-One) at a density of 5,000 cells per well. Treatment with inhibitors at 5, 1 or 0.1 µM started the day after seeding with a final content of 0.1% DMSO in the media. Media with treatment was changed every 48 hours. Cell viability was measured with a CellTiter-Glo® kit (Promega) four days after starting the treatment and reported as a percent of DMSO control. Cell fractionation assay [0058] The cell fractionation assay was performed based on known protocols. (see, e.g., Cer et al., Nucleic Acids Research 37 (suppl 2), W441–W445 (2009) (Cer, 2009); Porter et al., J. Biol. Chem.292 (7), 2601–2610 (2017), each of which is incorporated herein by reference. After 24 hours of treatment with vehicle, iBRD910 µM, 1-7810 µM or 2-7710 µM, 20 million HEK293T cells were harvested for processing. Cells were washed twice with cold phosphate- buffered saline and then resuspended in 500 µL of buffer A (0.05 mM EDTA, 25 mM HEPES, 5 mM MgCl₂, 10% glycerol, 25 mM KCl, 0.1% NP40) and incubated for eight minutes on ice. The samples were then centrifuged at 1300 x g for five minutes at 4 °C; supernatant 1 was collected 18     (cytosolic fraction) and pellet 1 (nuclei) was resuspended in 500 µL of buffer 2 (3 mM EDTA, 0.2 mM EGTA, 150 mM NaCl, 5 mM Tris, pH = 6.9). The nuclear soluble fraction was then collected after centrifugation at 20,000 x g for 15 minutes at 4 °C, and the pellet (chromatin fraction) was resuspended in 500 µL of buffer L (125 mM Tris base, 140 mM SDS, 20% glycerol, 10 % 2-mercaptoethanol, 2 mM MgCl2, pH = 6.9). Two microliters of Benzonase® endonuclease were added to each sample to release the chromatin-bound proteins, and the samples were then incubated with shaking at 500 rpm for 16 hours at 37 °C. The supernatant containing the chromatin-bound proteins was collected for immunoblot analysis. The inhibitors were added to a final concentration of 10 µM to each buffer. DTT to a final concentration of 1 mM and a cocktail of protease inhibitors were added to the buffers right before each step. [0059] For immunoblot analysis, the cytosolic fraction, nuclear soluble fraction, and chromatin- bound protein samples were diluted in a 1:3 ratio with a mixture of β-mercaptoethanol and 4X Bolt™ LDS Sample Buffer (Invitrogen) in a 1:9 ratio; then boiled at 95 °C for 30 minutes to reduce viscosity. Equal volumes of the samples were loaded to the 4 –12% SDS-polyacrylamide gel (Invitrogen). [0060] Whole cell lysates were prepared as well for immunoblot analysis. After 24 hours of treatment with vehicle, iBRD910 µM, 1-7810 µM or 2-7710 µM, HEK293T cells were harvested for processing. Cells were washed once with phosphate-buffered saline and then resuspended in RIPA buffer (50 mM Tris pH = 8, 150 mM NaCl, 1% NP40, 0.1% SDS, 0.5% sodium deoxycholate). A cocktail of protease inhibitors was added to the buffer right before use. The samples were then incubated for 20 minutes on ice and centrifuged at 15,000 x g for 30 minutes at 4 °C. The samples were prepared for immunoblotting by mixing them in a 1:3 ratio with a mixture of β-mercaptoethanol and 4X Bolt™ LDS Sample Buffer (Invitrogen) in a 1:9 ratio; and then boiling them at 95 °C for 15 minutes. Equal amounts of protein from the samples were loaded to the 4 –12% SDS-polyacrylamide gel (Invitrogen). NanoBRET assay [0061] For the NanoBRET screening, the NanoBRET^TM TE Intracellular BRD Assay-02 kit (Promega CS1810C21) was ued. NanoLuc® fusion BRD7-BD-Luc, BRD9-BD-Luc, and BRPF1-BD-Luc were manufactured by Promega. The assay was performed employing HEK293T cells cultivated as described in the Cell Lines section. The assay was set up in white, 19     flat bottom, non-binding surface 96-well plate (Corning 3992), and there were three wells of each concentration or control tested. [0062] On the day before the assay was performed, HEK293T cells were trypsinized and collected to prepare a 15 mL suspension of 200,000 cells/mL per DNA construct in a sterile, conical tube. Lipid:DNA complexes were prepared by first preparing a 10 µg/mL solution of DNA in Opti-MEM^TM without phenol red (Gibco 11058021) containing the following ratios: 9.0 µg/mL of Transfection Carrier DNA, 1.0 µg/mL NanoLuc® fusion DNA, 730 µL of Opti- MEM^TM without phenol red. After mixing thoroughly, 21.8 µL of FuGENE® HD (Promega E2311) were added. The solution was mixed by inversion 5 – 10 times and incubated at room temperature for 20 minutes. The Lipid:DNA mix was then added to the cell suspension and mixed thereby by inverting five times. The cells mixed with lipid:DNA complex were then incubated in a 10-cm plate at least for 24 hours to allow expression to occur. [0063] To evaluate the relative affinity of the NanoLuc® fusion proteins for the NanoBRET Tracer, a direct binding assay was performed. Starting with a concentration of 400 µM, two-fold dilutions of NanoBRET Tracer were performed to a concentration of 1.56 µM in DMSO to make 100X Tracer solutions.1 part of 100X NanoBRET Tracer was mixed with 4 parts of NanoBRET Tracer Dilution Buffer to generate 20X NanoBRET Tracer dilutions. A “no tracer” solution was prepared by mixing 1 part of DMSO to 4 parts of NanoBRET Tracer Dilution Buffer. The cells were then trypsinized, neutralized with regular media, and centrifuged at 250 x g for five minutes. The cells were then resuspended in Opti-MEM without phenol red, and the cell density was adjusted to 200,000 cells/mL. Cell suspension (68 µL) per well was added into a white, non- binding surface, 96-well plate. Four microliters per well of each 20X NanoBRET Tracer dilution or “no tracer” solution were added to three wells containing suspended cells. The plate was then mixed on an orbital shaker for 15 seconds at 500 rpm with 1 mm of diameter of shaking. A 10X “no inhibitor” solution was prepared by mixing one part of DMSO with nine parts of Opti-MEM without phenol red. “No inhibitor” (8 µL) was added to each well. The plate was then mixed on an orbital shaker for 15 seconds at 500 rpm with 1 mm of diameter of shaking and incubated at 37 °C in a 5% CO2 incubator for two hours. The plate was allowed to cool down to room temperature for 15 minutes before proceeding to add the 3X complete NanoBRET^TM Nano-Glo® mix as described below. 20     [0064] Less than 2 hours prior to BRET measurements, the 3X complete NanoBRET^TM Nano- Glo® mix was prepared in Opti-MEM without phenol red. This mixture consisted of a 1:166 dilution of NanoBRET^TM Nano-Glo® Substrate plus a 1:500 dilution of Extracellular NanoLuc Inhibitor in Opti-MEM without phenol red, which were mixed gently by inversion 5 – 10 times in a conical tube. Forty microliters of 3X complete NanoBRET^TM NanoGlo® mix were added to each well, and the plate was incubated for two to three minutes at room temperature. The donor and acceptor emissions were measured at 450 nm and 610 nm, respectively, in a SpectraMax iD5 plate reader employing the LUM-Dual Color Endpoint readout protocol, with an integration time of 1,000 ms and a read height of 1 mm from the plate. The BRET ratio in milliBRET units (mBU) with background correction was calculated employing the following equation: BRET ratio = [(Acceptorsample / Donorsample) – (Acceptorno tracer control / Donorno tracer control)] x 1000, where Acceptorsample and Donorsample are respectively the acceptor and donor emissions of each well, and Acceptor_{no tracer control} and Donor_{no tracer control} are respectively the average of the acceptor and donor emissions of the three wells where the “no tracer” solution was added. [0065] Based on the direct binding assay results, a final concentration of 0.4 µM of tracer was used for performing the competitive assay. A 100X NanoBRET Tracer solution (40 µM) was prepared in DMSO.1 part of 100X NanoBRET Tracer was mixed with 4 parts of NanoBRET Tracer Dilution Buffer to generate the 20X NanoBRET Tracer dilution (8 µM). A “no tracer” solution was prepared by mixing 1 part of DMSO to 4 parts of NanoBRET Tracer Dilution Buffer. Cells were trypsinized, centrifuged, and resuspended in Opti-MEM, and their density was adjusted to 200,000 cells/mL as described above. Cell suspension (68 µL) was added to each well of a white, non-binding surface, 96-well plate. NanoBRET Tracer (20X dilution; 4 µL) was added to wells containing suspended cells; 4 µL of “no tracer” solution were added to three wells containing suspended cells. The plate was then mixed on an orbital shaker for 15 seconds at 500 rpm with 1 mm of diameter of shaking. Starting with a concentration of 10 mM, two-fold dilutions of 1,000X inhibitor were performed to a minimum concentration of 20 µM in DMSO. Solutions (10X) were prepared by mixing one part of 1,000X solution with nine parts of Opti- MEM without phenol red. A “no inhibitor” mixture was prepared by mixing one part of DMSO with nine parts of Opti-MEM without phenol red. Eight microliters of each concentration per well were added to wells containing cells with 1X tracer. For the controls, 8 µL of “no inhibitor” mixture were added to the three wells containing “no tracer” solution and three other wells 21     containing 1X tracer. The plate was then mixed on an orbital shaker for 15 seconds at 500 rpm with 1 mm of diameter of shaking and incubated at 37 °C in a 5% CO2 incubator for two hours. The plate was allowed to cool down to room temperature for 15 minutes before proceeding to add the 3X complete NanoBRET^TM Nano-Glo® mix as described above. The donor and acceptor emissions were measured at 450 nm and 610 nm, respectively, in a SpectraMax iD5 plate reader following the protocol described above. The BRET ratio in mBU was calculated as described above, where Acceptor_{no tracer control} and Donor_{no tracer control} are respectively the average of the acceptor and donor emissions of the three wells where the “no tracer” and “no compound” solutions were added. [0066] The results of the assays were graphed in GraphPad Prism 9 and analyzed using a non- linear “[Agonist] vs. response - - Variable slope (four parameters)” fit for the direct binding assay and a non-linear “[Inhibitor] vs response - - Variable slope (four parameters)” fit for the competitive assay. RNAseq library prep and data analysis [0067] LNCaP cells were treated in triplicate with 1µM 2-77, 10 µM ENZA, 0.2 µM ACBI, 2 µM BD98, 10 µM PB16 or DMSO for 72 hours on 10 cm plates in RPMI growth medium with 10% fetal bovine serum (FBS). After 72 hours of drug treatment, cells were trypsinized, and RNA was extracted using TRIzol™ (Invitrogen™ - 15596026). Next, the TRIzol™-extracted RNA was further purified by column purification using PureLink™ Genomic RNA Mini Kit (Invitrogen™ - 12183018A) with On-column PureLink™ DNase treatment (Invitrogen™ - 12185010). The RNA quality and concentration was determined using ThermoFisher™ Qubit fluorimetry and Agilent™ Bioanalyzer, and all samples had a RIN score of 9.5 or greater. DNA libraries were generated using the Illumina® Stranded mRNA Prep kit (20040534) with sets of IDT® for Illumina® RNA UD Indexes (20040553), and the concentration was determined using Agilent™ Bioanalyzer. Libraries were pooled with equimolar amounts using cluster numbers obtained from MiSeq and sent for 150-bp paired-end sequencing on the NovaSeq 6000 platform (Novogene™, Sacramento, CA). Sequencing data were processed using Partek workflow. Read alignment to human genome build hg38 was performed with STAR 2.7.8a. Partek E/M model was used to assemble gene level expression data from filtered alignments, and differential gene expression analysis was conducted using DESeq2. Differential gene expression from LNCaP cells treated with 0.5 µM dBRD9 for 3 days was previously 22     published. (Alpsoy et al., Cancer Res 81 (4), 820–832021, which is incorporated herein by reference). Correlations were calculated using GraphPad Prism 9. Pathway analysis was performed using enrichr for 2020_MSigDB_Hallmark gene sets. Enrichments of gene sets were performed using GSEA. Quantitative overlap analysis was performed using eulerr. [0068] The cell fractionation assay was performed based on previously reported protocols. After 24 hours of treatment with vehicle, iBRD910 µM, 1-7810 µM or 2-7710 µM, 20 million HEK293T cells were harvested for processing. Cells were washed twice with cold PBS and then resuspended in 500 µL of buffer A (0.05 mM EDTA, 25 mM HEPES, 5 mM MgCl2, 10% glycerol, 25 mM KCl, 0.1% NP40) and incubated for eight minutes on ice. The samples were then centrifuged at 1300 x g for five minutes at 4 °C; supernatant 1 was collected (cytosolic fraction) and pellet 1 (nuclei) was resuspended in 500 µL of buffer 2 (3 mM EDTA, 0.2 mM EGTA, 150 mM NaCl, 5 mM Tris, pH = 6.9). The nuclear soluble fraction was then collected after centrifugation at 20,000 x g for 15 minutes at 4 °C and the pellet (chromatin fraction) was resuspended in 500 µL of buffer L (125 mM Tris base, 140 mM SDS, 20% glycerol, 10 % 2- mercaptoethanol, 2 mM MgCl2, pH = 6.9). Two microliters of Benzonase® endonuclease were added to each sample to release the chromatin bound proteins, and the samples were then incubated with shaking at 500 rpm for 16 hours at 37 °C. For Western blot analysis, the samples were boiled at 95 °C for 30 minutes to reduce viscosity. The inhibitors were added to a final concentration of 10 µM to each buffer. DTT was added to a final concentration of 1 mM, and a cocktail of protease inhibitors was added to the buffers right before each step. Example 1: Rational design and synthesis of BRD7 BD probes. [0069] The BRD9 BD inhibitor BI-9564 also binds to the closely related BD of BRD7, although with lower affinity (K_d values of 19 nM and 117 nM, respectively). BI-9564 forms two hydrogen bonds with BRD9^Asn100, a water-mediated hydrogen bond with BRD9^Tyr57, a π-π interaction with BRD9^Tyr106, a C-H π-interaction with BRD9^Ile53, and a T-stacking interaction with BRD9^Phe44. Although BI-9564 maintains the same contacts with the BRD7 BD, the binding affinity is reduced because of proposed entropic costs associated with the increased flexibility of BRD7 BD in solution. In order to exploit structural differences between the two BDs to develop BRD7- selective inhibitors, the deposited structures of the BRD7 BD (PDB: 5MQ1) and BRD9 BD (PDB: 5F1H) bound to BI-9564 were compared. As illustrated in Fig.2A, a hydrophobic region adjacent to the acetylated lysine binding pocket was identified in BRD7 BD that is not present in 23     BRD9 BD (see Fig.2A, BRD7 BD binding residues). To achieve selectivity for BRD7 over BRD9, a library of ligands was designed that would accommodate the acetylated lysine binding pocket in a similar manner as known BRD7/9 ligands but also extend into this BRD7-specific binding pocket. The thirteen library members contain one of four common cores found in known BRD7/9 ligands and can be synthesized in one or two steps from commercially available building blocks. The compounds are summarized in Table 1 and Scheme 1. Table 1: Structures of designed inhibitors based on four core options. Core Compound R Core 1 1-38

24     Core 3 1-78 R₃ 2-77 R₅ Scheme 1

Reagents and conditions for the synthesis of the designed BRD7 inhibitors. (a) NaH, DMF (anhydrous), 0 °C, the CH₃I to room temperature, 16 hours; (b) Pd(dppf)Cl₂, Cs₂CO₃, DMF (anhydrous), reflux, 16 hours; (c) Pd(dppf)Cl₂, Cs₂CO₃, 1,4-dioxane (anhydrous), reflux, 48 – 72 25     hours; (d) dissolve aryl aldehyde in THF, 0 °C, under argon, then add Grignard reagent over 10 minutes, then quench; (e) pyridium dichromate, DCM. Example 2: Thermal shift assay (TSA) to evaluate stability of the BRD7 and BRD9 BDs when bound to the ligands. [0070] With reference to Figs.3A and 3B, to screen for ligands of BRD7 BD or BRD9 BD a TSA based on differential scanning fluorimetry was performed to identify compounds that can stabilize the BDs and increase the melting temperature (T_m). Of the compounds with Core 1, 1- 38 stabilized and increased the Tm of both BRD7 BD and BRD9 BD, while 2-88 stabilized only BRD9 BD. Of the compounds with Cores 2 and 4, none of the compounds increased the T_m of either of BD. Of the compounds with Core 3, 1-78 and 2-77 increased the T_m of BRD7 BD but not BRD9 BD. Example 3: Competitive fluorescence polarization (FP) assay shows 1-78 and 2-77 are selective for BRD7 over BRD9. [0071] To further characterize the compounds, a fluorescently labeled probe, BI-FAMa, based on the structure of BI7273, a 4-carbon linker, and a fluorescein label was developed for use in a competitive FP assay (see Fig.4A). With reference to Figs.4B and 4C, BI-FAMa binding to the BRD7 and BRD9 BDs was characterized and a K_d of 508.8 nM for BRD9 was calculated, while it was not possible to calculate a Kd for BRD7 Even though direct binding had been achieved, optimal K_d values to saturate the system at a lower protein concentration for the FP competition assay were sought. With this in mind, BI-FAMb, which contains a 6-carbon linker instead to give more flexibility to the molecule, was developed (see Fig.4D). As a result, a fluorescent probe, which showed a K_d of 705.0 nM for BRD7 and 23.59 nM for BRD9, was obtained (see Figs.4E and 4F). Concentrations of 5 µM and 0.25 µM, respectively, of BRD7 and BRD9 were used to achieve approximately 90% saturation of the FP signal. The compounds were tested, along with BI-7273 as a control, in concentration gradients starting at 25 µM and ten two-fold dilutions. Four of the inhibitors showed binding for both proteins: 1-38 and 2-88 from core option 1 and 1- 78 and 2-77 from core option 3. Using the IC50 values obtained from the competition assay, and the Kd found of each protein for BI-FAMb, Ki values were estimated for the compounds employing the equation reported by Cer, 2009. (Table 2). The calculated K_i of BI7273 was lower for BRD9 than for BRD7. Compounds 1-38 and 2-88 also showed selectivity for BRD9 over 26     BRD7. However, 1-78 and 2-77 showed selectivity for BRD7 over BRD9, which aligned with the results obtained in the TSA. Table 2: Binding affinity of inhibitors to BRD7 and BRD9. Compound IC50 BRD7 (µM)^a Ki BRD7 (µM)^b IC50 BRD9 (µM)^a Ki BRD9 (µM)^b Exam

p e : n ng mo e anays s: - an - ose ey nteractons w en nteracting with BRD9. [0072] BRD9/7 inhibitors have a pharmacophore that mimics an acetylated lysine and forms a hydrogen bond with the highly conserved asparagine in the binding pocket (BRD7^Asn211, BRD9^Asn100). Additionally, the aromatic systems in BI-7273, BI-9564, I-BRD9, TP-472, and Bromosporine form π-π interactions with BRD7^Tyr217 or BRD9^Tyr106 (binding mode analyses with published crystal structures are shown in Figs.5A). Using molecular dynamics, 1-78 and 2- 77 were docked in the binding pocket of the structures of the BRD7 and BRD9 BDs crystallized with BI7273 (Fig.5A (top) PDB: 6V1E and (bottom) PDB: 5EU1, respectively).2-77 interacts with BRD7 BD with the same key interactions found between BI7273 and BRD7^Asn211 and BRD7^Tyr217 (Fig.5B (top)). BRD9 BD, however, cannot accommodate the core of 2-77 inside the pocket. Instead, the software docks 2-77 into the BRD9 BD binding pocket via the pyridine moiety of the molecule, where the aromatic ring forms a π-π interaction with BRD9^Tyr106, and a hydrogen bond can be formed between the oxygen of the para-methoxy group and BRD7^Asn211 (Fig.5B (bottom)). In contrast, the software docks 1-78 such that it maintains the π-π interactions between the pharmacophore and BRD7^Tyr217 and BRD9^Tyr106; however, to accommodate the steric bulk of 1-78 in BRD9 BD, the key hydrogen bond between the amide oxygen and BRD9^Asn100 is not maintained (Figs.5C). 27     Example 5: BROMOscan and bromoKdELECT profiling of 1-78 and 2-77. [0073] Compounds 1-78 and 2-77 were screened in the BROMOscan^TM panel (Eurofins DiscoverX Corp.), which gives a selectivity profile over 40 distinct bromodomains. Compound 1-38 was tested as a proof of concept for selectivity for BRD9 over BRD7, and 2-81 as a proof of concept for weak binding to BRD7 (see Figs.6A, TREEspot^TM interaction maps and 6B, results of TSA). To characterize 1-78 and 2-77 in a separate assay, the bromoKdELECT platform (Eurofins DiscoverX Corp.) was used to evaluate the binding affinity of the compounds for BRD7 BD. The results showed that 1-78 and 2-77 have a Kd average of 290 nM and of 340 nM for BRD7, respectively. Interestingly, both compounds show off-target binding for bromodomain and PHD finger-containing protein 1 (BRPF1) in the BROMOscan^TM even though the sequence identity between the BDs of BRFP1 and BRD7 is only 38%. Some known BRPF1 inhibitors share a core similar to that of 1-78 and 2-77, such as NI-48, with the differences relying in the position of the methyl group with respect to the lactam ring and the position at which the R- group binds to the core. The binding mode between NI-48 and the BRPF1 BD (PDB: 5T4V) was found to be similar to that of 1-78 and 2-77. The oxygen of the lactam ring of NI-48 also forms a hydrogen bond with the conserved asparagine BRPF1^Asn708, and the core forms two π-π interactions with BRPF1^Phe714. Additionally, the R-group of NI-48 forms an edge-to-face π-π interaction with BRPF1^Phe714 allowed by the length and flexibility of the chain that connects the R-group to the core. [0074] The BROMOscan^TM is a competitive binding assay and not a direct binding assay; therefore, it is difficult to assess from the BROMOscan alone how well the instant BRD7 inhibitors bind to BRPF1. Therefore, the inhibitors were also tested against the BD of BRPF1 using TSA. GSK-5959, a known selective BRPF1 inhibitor with a IC50 of 80 nM, significantly stabilized BRPF1 at concentrations as low as 1 µM (Figs.6B). Stabilization of BRPF1 was only observed by 1-78, and only at high concentrations (25 µM). Even though the compounds can compete with non-selective ligands for BRPF1 binding in the BROMOscan^TM, the interaction is not strong enough to stabilize the BRPF1 BD in vitro at similar concentrations. [0075] To get a better understanding of why the compounds do not stabilize BRPF1 BD to the same extent as BRD7 BD, their binding mode was evaluated in silico.1-78, 2-77, and GSK-5959 was docketed into the binding pocket of the BRPF1B BD crystallized with NI-48 (PDB: 5T4V). The core of GSK-5959 forms a hydrogen bond with the conserved Asn in BRPF1 (BRPF1^Asn708) 28     and a π-π interaction with BRPF1^Phe714, which is the gatekeeper of the binding pocket instead of the Tyr in BRD7^Tyr217 and BRD9^Tyr106. Even though 1-78 fits in the binding pocket, it lacks the critical hydrogen bond with BRPF1^Asn708, which could explain the reduced BRPF1 stabilization by 1-78 in the TSA. Similarly to what was observed when docking 2-77 into BRD9, BRPF1 cannot accommodate 2-77 in its binding pocket while maintaining critical binding interactions. To further explore why BRD9 and BRPF1 lack the ability to bind to 2-77, the pose of 2-77 bound to BRD7 was aligned with the binding pockets of BRD9 and BRPF1 (Figs.7A-C for BRD7, BRD9, and BRPF1, respectively). The R-group of 2-77 occupies the hydrophobic pocket in BRD7 without any steric hinderance;(Fig.7A, right); however, the R-group clashes with the binding pocket of BRD9 and BRPF1, explaining the selectivity. In all, two compounds, 1-78 and 2-77, which were shown to be selective for BRD7 bromodomains in vitro, were developed. Example 6: BRD7 inhibitors are selective for BRD7 BD in cells. [0076] 1-78 and 2-77 were next tested in cellulo to validate their use of as tools to study the function of BRD7. To that end, a NanoBRET assay was performed to assess intracellular BD engagement by the compounds. To do so the BDs of BRD7 and BRD9 fused with luciferase were expressed in HEK293T cells and first incubated with increasing concentrations of fluorescent tracer designed to generate BRET when bound to the target BDs. Using these binding profiles, 0.4 µM was selected as the tracer concentration for the NanoBRET assay with cells expressing BRD7 BD or BRD9 BD. The assay was performed by treating the cells in a competition assay with increasing concentrations of 1-78 and 2-77, BI7273, and GSK-5959. BI7273 had a lower IC50 for BRD9 (337 nM) than for BRD7 (1.208 µM) (Fig.8A). When treating the cells with 1-78 and 2-77, a lower IC₅₀ was observed for BRD7 (818 nM and 1.476 µM, respectively) than for BRD9 (3.286 µM and 2.469µM, respectively), indicating that the compounds were selective for BRD7 BD over BRD9 BD in cellulo (Figs.8B and 8C). Example 7: BRD7 inhibitors do not displace BRD9 from chromatin. [0077] PBAF contains a total of eight bromodomains across three separate subunits. While PBAF is reliant on single bromodomains for transcriptional function, it is not reliant on single bromodomains for global chromatin association. In contrast, GBAF is completely dependent on the BRD9 BD for global chromatin association, and the BRD9-selective bromodomain inhibitor I-BRD9 can displace BRD9 from chromatin. A cell fractionation assay was performed to determine whether the present BRD7 inhibitors can inhibit and displace BRD9 from chromatin at 29     cellularly active concentrations. HEK293T cells were treated with 10 µM of the compounds or DMSO, and cytosolic, nuclear soluble, and chromatin fractions were collected. Western blotting was used to evaluate the relative amount of BRD9 still bound to chromatin after inhibitor treatment. I-BRD9 treatment significantly reduces BRD9 bound to chromatin while 1-78 and 2-77 do not, indicating that, while the NanoBRET assay detects some off-target binding to BRD9 at 10 µM, it is not sufficient to inhibit BRD9 chromatin binding. Example 8: BRD7 inhibition reduces cell proliferation and AR target gene expression in prostate cancer-cell based models. [0078] Prostate cancer (PCa) is the second leading cause of death and the most frequently diagnosed cancer in men in the US, with 34,500 associated deaths and 268,490 new cases estimated for 2022. Several studies have identified BRD7 as a prognostic marker and facilitator of PCa progression, and the knockout of BRD7 reduces expression of testosterone-response genes in HAP1 cells; however, the therapeutic potential of targeting BRD7 in PCa has not been explored, in part due to a lack of chemical tools. [0079] PCa can be classified as either hormone-naïve or castration-resistant, and either androgen receptor- (AR-) positive or AR-negative. Six different cell lines were employed to model the disease: RWPE-1 (normal prostate epithelial cells), LNCaP (hormone-naïve, AR-positive PCa cells), 22Rv1 (castration-resistant, AR-positive PCa cells, C4-2 (castration-resistant, AR-positive PCa cells), PC-3 (castration-resistant, AR-negative PCa cells), and DU-145 (castration-resistant, AR-negative PCa cells). To determine prostate cancer dependency on BRD7, lentiviral-mediated shRNA was used to knock down BRD7 in all six cell lines. With reference to Fig.9A-9C, reduced expression of BRD7 decreased cell proliferation in AR-positive cells, while having little to no effect on normal epithelial prostate cells or AR-negative PCa cells, an effect similar to, but more pronounced than what was previously observed for PBRM1. Both 1-78 and 2-77 inhibited cell growth of LNCaP cells at all three concentrations (5, 1 and 0.1 µM), while being active in PC-3 only at the highest concentration (5 µM), in agreement with a greater BRD7-dependency in AR-positive PCa. With specific reference to Figs.9A and 9B, to evaluate if the activity of the compounds could be related to off-target BRD9 or BRPF1 inhibition, LNCaP and PC-3 cells were treated with BI7273 and GSK-5959 at 5, 1 and 0.1 µM, and observations showed little to no effect with either compound. 30     [0080] With reference to Figs.10A-10C, to evaluate how BRD7 BD inhibition by 2-77 affects gene expression in PCa, RNA-Seq was performed on LNCaP cells treated with 1 µM 2-77 or DMSO for 72 hours.661 genes were identified that decreased and 859 genes that increased with 2-77 treatment using DESeq (padj < 0.05, fold change (FC) > 1.5) (Fig.10A). Those MSigDB_Hallmark gene sets that were significantly enriched in the respective increased and decreased genes were then identified. Androgen Response, G2/M checkpoint and E2F targets were the most significantly enriched pathways (Fig.10B), which agrees with BRD7 knockdown affecting only AR-positive PCa cell lines with BRD7 knockdown decreasing cell growth and viability. Gene set enrichment analysis (GSEA) was performed for the Hallmark_Androgen_Response gene set with the full RNA-Seq dataset, and significant and strong negative enrichment of androgen response genes was found, indicating that 2-77 decreases AR target gene expression (Fig.10C). [0081] With reference to Figs.10D-10E, to further explore the relationship between 2-77 and androgen response genes, RNA-seq was performed on LNCaP cells treated with 10 µM of the AR antagonist enzalutamide (ENZA) or DMSO for 72 hours.662 genes were identified that decreased with ENZA and 609 genes that increased with ENZA (p_adj < 0.05, FC > 1.5) (Fig. 10D). In agreement with a role for BRD7 in AR target gene expression, a high overlap of genes differentially expressed with 2-77 treatment and genes differentially expressed with ENZA treatment was observed, as well as a high correlation between all gene expression changes induced by treatment with 2-77 or ENZA (Fig.10E). [0082] With reference to Fig.10F, while few studies have specifically addressed the role of PBAF in PCa gene expression, several have identified universal BAF subunits (see references for BRG or SMARCD) or subunits from the other subcomplexes (cBAF and GBAF) to be critical for AR target gene expression. Therefore, RNA-Seq was also performed with a SMARCA2/4 degrader that eliminates all BAF complexes (ACBI1), a cBAF-specific inhibitor (BD98), a BRD9 degrader that eliminates GBAF complexes (dBRD9), and a BD inhibitor specific for the PBAF subunit PBRM1 (PB16). In agreement with their modes of actions, the largest changes in gene expression were found with ACBI1, which eliminates all BAF function, and the next largest with BD98, which targets the most abundant cBAF subcomplex. In contrast, smaller changes in gene expression were found with dBRD9 or PB16, which target the less abundant GBAF and PBAF complexes. In agreement with published findings, a decrease in androgen target gene 31     expression was observed for inhibitors of all three SWI/SNF subcomplexes. Similarly, the genes differentially expressed with ENZA showed significant overlap and correlation with genes differentially expressed with the inhibitors of all three individual BAF subcomplexes. Similar to what was observed with 2-77, gene expression changes from PBRM1 inhibitor PB16 treatment had particularly high correlation with gene expression changes from ENZA treatment. [0083] With reference to Figs.10G and 10H, cBAF complexes also facilitate other transcription factors, such as FOXA1, ERG and MYC, and GBAF complexes also regulate AR-independent BRD4 target genes; therefore, it was hypothesized that each subcomplex would regulate unique target genes in addition to shared AR-target genes. To evaluate this, the overlap of genes regulated by specific BAF subcomplex inhibitors (2-77, BD98, dBRD9, PB16) and genes regulated by ACBI1, which eliminates all three BAF subcomplexes, was determined. Correlation and overlap between the genes regulated by ACBI1 and the genes regulated by all four inhibitors tested was observed, with the strongest overlaps observed for genes that decreased with treatment, consistent with a general role for BAF in gene activation. Therefore, the overlap between genes decreased with ACBI1, genes decreased with 2-77, and genes decreased with subcomplex-specific inhibitors BD98, dBRD9, or PB16 was compared (Fig.10G). Several genes, including AR target genes, were decreased with all drug treatments. Subsets of genes decreased with ACBI1 were decreased with dBRD9 or BD98 but not 2-77, indicating subcomplex-specific inhibition of genes. In contrast, almost all genes that decreased with both ACBI1 and PB16 were also decreased with 2-77, consistent with inhibition of the same subcomplex (PBAF). This trend extended to all the genes differentially regulated by PB16 and genes differentially regulated by 2-77, which showed extremely high overlap and correlation (Fig.10H). The degree of gene regulation was stronger for 2-77, which may be pharmacological, or may be due to an increased dependence of PCa cells on BRD7 compared to PBRM1. This possibility is supported by a larger viability defect with BRD7 knockdown compared to viability defect previously observed with PBRM1 knockdown (see Fig.9A). Example 9: Synthesized Core 3-Analog Candidate BRD7 inhibitors. [0084] A virtual library of ~500 compounds was designed based on a Core 3 scaffold that would accommodate the acetylated lysine binding pocket of BRD7 while also extending into the BRD7-specific binding pocket. Docking was then performed using Glide XP docking software to estimate binding affinities between candidate compounds and respective BRD7 and BRD9 BDs, 32     along with 2-77 and BI-7273 controls. Compounds having XP scores indicating selectivity for BRD7 over BRD9 were then further narrowed using Prime MMGB/SA binding affinity analysis (dG) for BRD7 selectivity over BRD9 and compounds scoring a more negative ΔG value for BRD7 versus BRD9 were then further narrowed still using interaction pattern analysis for selectivity over BRD9, which yielded thirteen candidate BRD7 inhibitor compounds. The thirteen Core 3 analogs, including their structure, their XP score, ΔG value, and BRD7 residues with which they are predicted to interact (as identified in Fig.2B) are summarized in the Table 3. Table 3: Core 3-Analog Candidate BRD7 inhibitors

33    

Example 10: Thermal Shift Assay. [0085] Of the thirteen Core 3 analogs of Table 3, nine were synthesized for thermal shift assays for selectivity for BRD7: 277.1, 277.2, 277.3, 277.4, 277.5, 277.6, 277.7, 277.7, and 277.9. Synthesis was performed in accordance with Scheme 1, herein. For each assay, the HIS-tagged BDs were incubated with the compounds at 25 µM. The Tm of the proteins was calculated based on differential scanning fluorimetry readings at increasing temperatures from four replicates 34     using nonlinear least squares fit on GraphPad Prism 9. The shift in Tm was calculated with respect to vehicle control. As demonstrated in the histogram of Fig.11, of the nine synthesized compounds, 277.3, 277.4, 277.5, and 277.8 showed the greatest shift in the T_m of BRD7 BD versus BRD9 BD. Example 11: Competitive Fluorescence Polarization (FP) Assay. [0086] To further characterize compounds 277.3, 277.4, 277.5, and 277.8, BI-FAMb (see Fig. 4D), based on the structure of BI7273, a 4-carbon linker, and a fluorescein label was used in a competitive FP assay, the results of which are presented in Figs.12A-C. With reference to Fig. 12A, BI-FAMb had a K_d value of 0.319 µM. A concentration of 5 µM of BRD7 was used to achieve approximately 90% saturation of the FP signal. The compounds were tested, along with BI-7273 and 2-77 as controls, in concentration gradients starting at 25 µM and ten two-fold dilutions, and the results are shown in Fig.12B. Using the IC50 values obtained from the competition assay, and the K_d found for BI-FAMb, K_i values were estimated for the compounds employing the equation reported by Cer et al., 2009 (see Table 4). Ki values for each of compounds 277.3, 277.4, 277.5, and 277.8 greater than that of BRD9 inhibitor BI-723, and thus established these compounds as having greater binding affinity for BRD7 than BI-723. Table 4: Binding affinity of 277.3, 277.4, 277.5, and 277.8 Compound IC_50BRD7 (µM) K_iBRD7 (µM)   Example 12: Eff

p y. [0087] Lentiviral-mediated shRNA was used to knock down BRD7 in LnCaP and PC-3 cell lines, as described in Example 8, herein. With reference to Figs.13A and 13B, each of compounds 277.3, 277.4, 277.5, and 277.8 inhibited cell growth of LNCaP cells, primarily at 5 and 1 µM 35     concentrations, while being active in PC-3 only at the 5 µM, similar to performance of 2-77 and in agreement with a greater BRD7-dependency in AR-positive PCa, indicating that compounds 277.3, 277.4, 277.5, and 277.8 will decrease AR target gene expression. [0088] All patents, patent application publications, journal articles, textbooks, and other publications mentioned in the specification are indicative of the level of skill of those in the art to which the disclosure pertains. All such publications are incorporated herein by reference to the same extent as if each individual publication were specifically and individually indicated to be incorporated by reference. In the event of inconsistent usages between this document and those documents so incorporated by reference, the usage in the incorporated reference should be considered supplementary to that of this document; for irreconcilable inconsistencies, the usage in this document controls. [0089] The invention illustratively described herein may be suitably practiced in the absence of any element(s) or limitation(s), which is/are not specifically disclosed herein. Thus, for example, each instance herein of any of the terms “comprising,” “consisting essentially of,” and "consisting of" may be replaced with either of the other two terms. Likewise, the singular forms "a," "an," and "the" include plural references unless the context clearly dictates otherwise. Thus, for example, references to "the method" includes one or more methods and/or steps of the type, which are described herein and/or which will become apparent to those ordinarily skilled in the art upon reading the disclosure. The term "or" is used to refer to a nonexclusive "or" unless otherwise indicated. [0090] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood to one of ordinary skill in the art. The following terms and phrases shall have the meaning indicated. [0091] The term "about," when referring to a number or a numerical value or range (including, for example, whole numbers, fractions, and percentages), means that the number or numerical range referred to is an approximation within experimental variability (or within statistical experimental error) and thus the numerical value or range can vary between 1% and 15% of the stated number or numerical range (e.g., +/- 5 % to 15% of the recited value, such as within 10%, within 5%, or within 1% of a stated value or stated limit of a range) provided that one of ordinary skill in the art would consider equivalent to the recited value (e.g., having the same function or result). The term "substantially" can allow for a degree of variability in a value or range, for 36     example, within 90%, within 95%, 99%, 99.5%, 99.9%, 99.99%, or at least about 99.999% or more of a stated value or of a stated limit of a range. [0092] In addition, it is to be understood that the phraseology or terminology employed herein, and not otherwise defined, is for the purpose of description only and not of limitation. Any use of section headings is intended to aid reading of the document and is not to be interpreted as limiting. Further, information that is relevant to a section heading may occur within or outside of that particular section. 37  

Claims

  WHAT IS CLAIMED IS: 1. A compound of formula I:

a pharmaceutically acceptable salt thereof.

2. A composition comprising the compound of claim 1 and a pharmaceutically acceptable carrier, diluent, or excipient. 3. A method of inhibiting the proliferation of cancerous cells in a patient, which method comprises administering to the patient a composition of claim 2 in an amount effective to inhibit cancerous cell proliferation, whereupon the proliferation of cancerous cells in the patient is inhibited. 4. The method of claim 3, wherein the patient has prostate cancer. 5. The method of claim 4, wherein the prostate cancer is androgen receptor-positive (AR+). 6. The method of claim 3, wherein the patient has nasopharyngeal carcinoma, osteosarcoma, colorectal cancer, hormone-dependent breast cancer, or hormone-dependent ovarian cancer. 7. A compound of formula I: 38     R wherein R is selected from the of:

or a pharmaceutically acceptable salt thereof. 8. A composition comprising the compound of claim 7 and a pharmaceutically acceptable carrier, diluent, or excipient. 9. A method of inhibiting the proliferation of cancerous cells in a patient, which method comprises administering to the patient a composition of claim 8 in an amount effective to inhibit cancerous cell proliferation, whereupon the proliferation of cancerous cells in the patient is inhibited. 10. The method of claim 9, wherein the patient has prostate cancer. 11. The method of claim 10, wherein the prostate cancer is androgen receptor-positive (AR+). 39     12. The method of claim 9, wherein the patient has nasopharyngeal carcinoma, osteosarcoma, colorectal cancer, hormone-dependent breast cancer, or hormone-dependent ovarian cancer. 40