Potent BRD4 inhibitor suppresses cancer cell-macrophage interaction

Small molecule inhibitor of the bromodomain and extraterminal domain (BET) family proteins is a promising option for cancer treatment. However, current BET inhibitors are limited by their potency or oral bioavailability. Here we report the discovery and characterization of NHWD-870, a BET inhibitor that is more potent than three major clinical stage BET inhibitors BMS-986158, OTX-015, and GSK-525762. NHWD-870 causes tumor shrinkage or significantly suppresses tumor growth in nine xenograft or syngeneic models. In addition to its ability to downregulate c-MYC and directly inhibit tumor cell proliferation, NHWD-870 blocks the proliferation of tumor associated macrophages (TAMs) through multiple mechanisms, partly by reducing the expression and secretion of macrophage colony-stimulating factor CSF1 by tumor cells. NHWD-870 inhibits CSF1 expression through suppressing BRD4 and its target HIF1α. Taken together, these results reveal a mechanism by which BRD4 inhibition suppresses tumor growth, and support further development of NHWD-870 to treat solid tumors.

Chemical structures of GSK525762 (I-BET762), CPI-0610, MK8628 (OTX-015) and BMS-986158. Figure 2 Schematic representation of chemical synthesis of NHWD-870 and NHWD-870-HCl. Figure 3 Triple negative breast cancer (TNBC), ovarian cancer and small cell lung cancer (SCLC) cells exhibited high BRD4 expression among several human tumor cell lines. Figure 4 NHWD-870 does not suppress the growth of non-cancerous cells Figure 5 BRD4 loss inhibited c-MYC expression, and suppressed the growth of A375 melanoma cells. Figure 6 Metabolic stability of NHWD-870 in (a) mouse, (b) rat, (c) dog, (d) monkey and (e) human liver microsomes. Figure 7 Pharmacokinetic profile of NHWD-870 using mouse and rat models.        Figure 16 The levels of phosphorylated BRD4 and CSF1 in tumor cells, as well as the number of TAMs were positively correlated in human ovarian tumors. Table 1 Metabolic stability of NHWD-870 in mouse, rat, dog, monkey and human liver microsomes. Table 2 Binding and recovery rates of NHWD-870 in plasma assessed by equilibrium dialysis assays. Table 3 Summary of pharmacokinetic parameters in mouse and rat after intravenous (IV) and oral (PO) administration of NHWD-870. Table 4 Mouse body weight changes (%) from Day 0 after NHWD-870 or BMS-986158 treatment. Table 5 Demographic characteristics of patients with epithelial ovarian cancer and association of CD68 levels with CSF1 and pBRD4 levels and clinical variables of these patients. Table 6 Univariate analysis of 128 patients with epithelial ovarian cancer. Table 7 Multivariate analysis with covariate adjustment of 128 patients with epithelial ovarian cancer. Table 8 List of antibodies Table 9 List of primers for RT-qPCR Supplementary Methods: with anti-CD45, CD11b and CD206 followed by FACS analysis. Shown are representative FACS plots (c) and quantification of the percentage of CD11b + CD206 + cells (d). Data are presented as mean ± SEM from 5 different mice. p values were calculated using two-tailed, unpaired t tests. **, p<0.01; ***, p<0.001. (e-f) TAMs (40,000 cells) treated with or without 25 nM NHWD-870 in medium containing 2% Matrigel were seeded onto the 24-well bottom chamber pre-coated with Matrigel. Representative pictures of TAM spheroid (e: black/white). Scale bar is 50 μm. Quantification of spheroid volumes (f). Data are presented as mean ± SEM from 15 independent experiments. p values were calculated using two-tailed, unpaired t tests. **, p<0.01. (gi) NHWD-870 significantly inhibited tumor cell supported TAM proliferation. A2780 cells (pre-treated with DMSO, 25 or 50 nM NHWD-870 for 48h) were seeded into the top chamber (transwell size: 0.4 μm) and TAMs (1 × 10 5 cells per 6-well) in medium were seeded into the bottom chamber. CD45 + F4/80 + CD11b + CD206 + TAMs were isolated from tumors of ovarian cancer-bearing donor mice. Shown are schematics of the experiment (g), representative images (h) and quantification of colonies (i). Data are presented as mean ± SEM from 3 independent experiments. p values were calculated using two-tailed, unpaired t tests. **, p<0.01; ***, p<0.001. Source data are provided as a Source Data file. Figure 12. CSF1 was highly expressed in tumor cells, and CSF1R was highly expressed in TAMs. CD11b + monocytes in the orthotopic OC model were harvested from blood and CD11b + CD206 + macrophages (TAMs) in the orthotopic OC model were isolated from tumors. RT-qPCR analyses of CSF1 and CSF1R in monocytes, TAMs (a,b), ID8 mouse ovarian cancer cells (a), and B16 mouse melanoma cells (b). Data are presented as mean ± SEM from 3 independent experiments. p values were calculated using two-tailed, unpaired t tests. n.s, not significant; *, p<0.05; **, p<0.01; ***, p<0.001. Source data are provided as a Source Data file. Figure 13. NHWD-870 inhibited CSF1/CSF1R signaling pathway to reduce TAM proliferation. TAMs (2×10 5 cells) (treated with or without 10 ng/ml CSF1 and 100 nM NHWD-870 for 48 h) were seeded into the 6-well bottom chamber, in the presence or absence of A2780 cells (pre-treated with or without 100 nM NHWD-870 for 48 h) seeded into the top chamber (transwell size: 0.4μm). Shown are schematics of the experiment and representative western blot analysis of the indicated proteins (a), and quantification of the relative levels of p-PI3K, p-AKT1 and p-ERK over their respective total protein (b). Data are presented as mean ± SEM from 3 independent experiments. p values were calculated using two-tailed, unpaired t tests. n.s, no significant; **, p<0.01; ***, p<0.001. Source data are provided as a Source Data file. Figure 14. NHWD-870 and BRD4 deletion partially inhibited proliferation of TAMs induced by CSF1. (a) CD45 + CD11b + F4/80 + TAMs in the orthotopic B16 melanoma model were harvested from tumors. Shown are representative FACS plots of TAMs. (b,c) Representative FACS plots (b) and quantification (c) of Ki67 + macrophages treated with 50 ng/ml CSF1, and DMSO or 25 or 50 nM NHWD-870. Data are presented as mean ± SEM from 3 independent experiments. p values were calculated using two-tailed, unpaired t tests. *, p<0.05; **, p<0.01. (d,e) Representative FACS plots (d) and quantification (e) of Ki67 + control (Ctrl) or BRD4 knockout (KO) macrophages treated with 50 ng/ml CSF1. Data are presented as mean ± SEM from 3 independent experiments. p values were calculated using two-tailed, unpaired t tests. ***, p<0.001. Source data are provided as a Source Data file. (a) Relative HIF1 luciferase reporter activity in control (Ctrl) or OTX-015 treated A375 cells exposed to 20% or 1% O2 for 24 hours. (b) Representative western blot analysis of control and BRD4 knockout (KO) HeLa cells transfected with HIF1α plasmids and exposed to 1% O2 for 24 hours.  1) ND = Not determined (Parameters not determined due to inadequately defined terminal elimination phase); 2) PK parameters were estimated by non-compartmental model using WinNonlin 6.  Body weight loss is more than 10%, but less than 15% Body weight loss is more than 15% Dead Supplementary

Protein extraction and western blot analysis. Freshly dissected unfixed tissue
was homogenized in lysis buffer. The lysates were centrifuged at 13,000g for 10 minutes at 4°C. Supernatants were collected and determined with a Bradford Protein Assay kit (Bio-Rad, Hercules, CA). The cell lysates were subjected to SDS-PAGE followed by immunoblotting (Immobilon P; Millipore, Milford, MA) with specific antibodies followed by detection using an enhanced chemiluminescence kit (Amersham Life Science, Arlington Heights, IL).

Colony formation assay.
Cells were seeded in 6-well plates at 3,000 cells/well and treated with JQ1, NHWD-870, or DMSO control. Media were replaced once every week. After 2-3 weeks, depending on the cell growth rate, cells were fixed in 4% para-formaldehyde for 10 minutes at room temperature and stained with 0.5% crystal violet for 30 minutes. The plates were washed with water and dried before photographing. Colony numbers were calculated with imageJ version 1.52s 1 .

MTT and MTS assay. Cells (5,000 cells per well) were seeded into 96-well plates.
20 μl MTT or MTS reagent was added into each well and incubated for 4 hours at 37℃, 5% CO2. The optical density (OD490nm) was measured using a microplate reader, and the relative absorbance was calculated by subtracting the average background signal.
7. Metabolic stability assay. The metabolic profile of NHWD-870 was determined using mouse, rat, dog, monkey, and human liver microsomes using the protocols described below. Testosterone, Diclofenac, and Propafenone were included as the controls.

4)
After pre-warming, dispense 90 μL/well NADPH regenerating system to 96-well plate as reservoir according to the plate map. Then add 10 μL/well to every plate by Apricot to start reaction.

7)
The sampling plates are shaked for approx 10 min. should have a value of not less than 80% bound, and the recovery rate should be at least 30%.
9. In vivo pharmacokinetics study. Mouse and SD rat models were used to evaluate pharmacokinetic property of NHWD-870. NHWD-870 was administrated by either intravenous injection (IV, 1.75 mg/kg for mouse and 1 mg/kg for rat) or oral gavage (PO, 5 mg/kg) into mice or rats. Concentration of drugs in plasma, lung and tumor tissues were measured for the indicated time.

hERG channel assay. The potential inhibitory effect of NHWD-870 on human
Ether-à-go-go related gene (hERG) channel was evaluated by manual patchclamp system using the protocols described below. HEK293 cell line stably transfected with hERG gene was used in this study and Dofetilide was used as the benchmark for the quality of the assay. The potency of a test compound inhibiting hERG channel is list as follows: a) Low: IC50 > 10 µM; b) Moderate: 1 µM < IC50 < 10 µM; c) High: IC50 < 1 µM. The FDA criterion for defining a drug as hERG-positive is when IC50 < 1 µM. of 30, 10, 3.33, 1.11 and 0.37 µM. (5) Experimental procedure 1) Remove the coverslip from the cell culture dish and place it on the microscope stage in bath chamber. 2) Locate a desirable cell using the ×10 objective. Locate the tip of the electrode under the microscope using the ×10 objective by focusing above the plane of the cells. Once the tip is in focus, advance the electrode downwards towards the cell using the coarse controls of the manipulator, while simultaneously moving the objective to keep the tip in focus. 3) When directly over the cell, switch to the ×40 objective and use the fine controls of the manipulator to approach the surface of the cell in small steps. 4) Apply gentle suction through the side-port of the electrode holder to form a gigaohm seal. 5) Use the Cfast to remove the capacity current that is in coincidence with the voltage step.
Obtain the whole cell configuration by applying repetitive, brief, strong suction until the membrane patch has ruptured. 6) Set membrane potential to -60 mV at this point to ensure that hERG channels are not open. The spikes of capacity current should then be cancelled using the Cslow on the amplifier. 7) Set holding potential to -90 mV for 1 second; record current at 50 kHz and filter at 10 kHz. Leaking current was tested at -80 mV for 500 ms. 8) The hERG current was elicited by depolarizing at +30 mV for 4.8seconds and then the voltage was taken back to -50 mV for 5.2 seconds to remove the inactivation and observe the deactivating tail current. The maximum amount of tail current size was used to determine hERG current amplitude. 9) Record current for 120 seconds to assess the current stability. Only stable cells with recording parameters above threshold were applied for the drug administrations. 10) Firstly vehicle control was applied to the cells to establish the baseline. Once the hERG current was found to be stabilized for 3 minutes, compound was applied. hERG current in the presence of test compound were recorded for approximately 5 minutes to reach steady state and then 5 sweeps were captured. For dose response testing, 5 doses of compound was applied to the cells accumulatively from low to high concentrations. In order to ensure the good performance of cultured cells and operations, the positive control, Dofetilide, with 5 dosing concentration was also used test the same batch of cells that were used for compounds. (6) Data analysis (6.1) Data acceptance criteria The following criteria were used to determine data acceptability. 1) Initial seal resistance > 1 GΩ; 2) Leak currents < 50% of the control peak tail currents at any time; 3) The peak tail amplitude >250pA; 4) Membrane resistance Rm > 200 MΩ; 5) Access resistance (Ra) < 15 MΩ; 6) Apparent run-down of peak current < 2.5% per min. (6.2) Data analysis Data that met the above criteria for hERG current quality were further analyzed as the following steps. 1) Percent current inhibition was calculated using the following equation.
Note: PatchMaster software was used to extract the peak current from the original data.
2) The dose response curve of test compounds was plotted with %inhibition against the concentration of test compounds using Graphpad Prism 5.0, and fit the data to a sigmoid dose-response curve with a variable slope.
11. Plasmids and CRISPR/Cas9-mediated knockout. pCAG-HIF1α plasmid was purchased from Addgene (#21101) and was described previously 3  14. Immunohistochemical staining. Paraffin-embedded implantation samples isolated from peritoneal metastasis of EOC patients (128 cases) were sectioned at a thickness of 4μm. To stain CD68, CSF1 and p-BRD4, the slides were first deparaffinized in xylene and rehydrated with gradient concentrations of alcohol under standard procedures. After rehydration, the slides were immersed in 0.01 mol/L citrate buffer (pH 6.0) and heated (95°C) for 15 min for antigen retrieval.
Then, the samples were incubated with 3% hydrogen peroxide (H2O2) for 10 minutes followed by 10% normal goat serum blocking for 10 minutes. Subsequently, the sections were incubated with rabbit polyclonal anti-human CD68 antibody (dilution 1:100) (sc-20060, Santa Cruz Biotechnology) and anti-human p-BRD4 (dilution 1:100), or with mouse monoclonal anti-human CSF1 (dilution 1:100) (sc-365779 Santa Cruz Biotechnology) for 1 hour at room temperature. After washing with PBST for 3 times, the sections were incubated with biotin-labeled secondary antibody followed by horseradish peroxidase (HRP)-conjugated streptavidin for 30 minutes individually at room temperature. After applying HRP substrate, 3.3'diaminobenzidine tetrahydrochloride (D3939-1set, Sigma) in 0.01% H2O2, for 10 minutes, the slides were counterstained with Meyer's hematoxylin for 30 to 60 seconds and mounted with mounting medium for visualization under microscope 7,8 . Scoring of CD68, CSF1 and p-BRD4 in EOC samples via IHC staining follows the methods previously published 7,8 . Briefly, CD68, p-BRD4 and CSF1 staining intensity was determined using ImageJ version 1.52s 1 , and normalized with cell area. Cell area was determined by manual delineation of raw IHC staining images.
A minimum of 12 cells were analyzed from two independent experiments. All IHC staining samples from EOC patients were evaluated independently by two experienced pathologists.

ChIP-qPCR analysis.
A2780 cells were treated with 15nM NHWD-870 or DMSO for 3 days followed by crosslinking in DMEM with 1% formaldehyde for 10 minutes. After stopping the cross-linking SDS for DNA extraction, followed by incubation at 65°C overnight for reverse crosslinking. The next day RNase A and proteinase K were added and incubated at 37°C for 2 hours, respectively. Then SanPrep Column PCR Product Purification Kit (Sangon Biotech) was used for DNA purification. ChIP-qPCR analyses were performed using primers F-"CGCGAACGACAAGAAAAAGT", R-"GAGCAGCAGCAGAAACAAAA" for the HIF1A promoter.