Discovery of a small molecule that inhibits Bcl-3-mediated cyclin D1 expression in melanoma cells

Melanoma cancer cell line Mel Juso was cultured in DMEM that was supplemented with 1 g/L Glucose (Corning), 10% FBS (Biosera), and 100 U/ml penicillin/streptomycin (Corning). Melanoma cell lines: Mel Ju, A7, HMCB, WM239A, WM852, and A2058 were cultured in RPMI-1640 supplemented with 10% FBS and 100 U/ml penicillin/streptomycin (Corning). SK-MEL-3 and HT144 cells were cultured in McCoy’s 5A medium supplemented with 10% FBS (Biosera), 100 U/ml penicillin/streptomycin (Corning). Mel Juso, Mel Ju, Mel Ho, and SK-MEL-3 cells were kindly provided by Prof. Anja Bosserhoff (Friedrich-Alexander University Erlangen-Nuremberg, Germany). A7, HMCB, WM239A, WM852, HT144, and A2058 HT144 cells were kindly provided by Prof. Göran Jönsson (Lund University, Sweden). All cell lines were cultured at 37 °C in a humidified atmosphere containing 5% CO₂. Cells were trypsinized with 0.25% trypsin (Corning, Cat No. 25–053-CI). Cell Line authentication has been performed for all the cell lines presented in this study. Immortalized keratinocytes were cultured and transfected as described previously [24]. BCL3ANT was dissolved in DMSO at different concentrations and stored at -20 °C until use.

p-Bcl-3-FLAG was kindly provided by Dr. Roland Schmid, Munich Technical University, Munich, Germany. Cyclin D1 promoter and NFκB mutant cyclin D1 promoter were kindly gifted by Dr. Richard G. Pestell (Pennsylvania Cancer and Regenerative Medicine Research Center, Baruch S. Blumberg Institute, Wynnewood, PA, USA). The NFκB site in the cyclin D1 promoter at 239 to 230 base pairs (bp) was mutated by a polymerase chain reaction from 5’-GGG GAG TTT T-3’ to 5’-GCC CAG TTT T-3’.

Transient transfections were performed in 96-well plates using PolyFect reagent (Qiagen) according to the manufacturer’s instructions. Cells were transfected with a total of 470 ng DNA consisting of: Renilla (20 ng), cyclin D1 promoter (400 ng) or NFκB mutant cyclin D1 promoter (400 ng), together with empty vector (50 ng, Clonetech), or p-Bcl-3-FLAG (50 ng). Cells were transfected for 24 h, after which they were treated with or without drugs for 24 h. Following incubation with or without drugs, cells were washed twice with PBS and analyzed using the Dual-Luciferase Reporter Assay System kit, from now on referred to as: Dual Reporter Assay (Promega) according to the manufacturer’s instructions. Luciferase activity was measured on the Synergy 2 HT Multi-Mode Microplate Reader (Biotek). All experiments were performed three times in triplicates and an average DLR for the control was set to 1. siRNAs (control and Bcl-3) were purchased from Santa Cruz and used according to the manufacturer's instructions. siRNA transfections were made with HiPerFect transfection reagent (QIAGEN).

Mel Juso cells were seeded on 96-well plates at a density of 2 × 10⁴ cells per well (day 0). After cells were attached, transfections were conducted for 24 h using the following plasmids: p-Bcl-3-FLAG, together with reporter plasmid cyclin D1 Luciferase and Renilla-Luc (Day 1). Compounds (Diversity Set, 1368 compounds; National Cancer Institute, Bethesda, MD) were added at a final concentration of 10 μM in DMSO (day 2). Cells were incubated for 24 h, lysed, and analyzed using the Dual Reporter assay (Promega). All experiments were performed three times in triplicates. To assay for transfection efficiency, transfections were performed with a constant amount of Renilla expression plasmid. Luciferase and Renilla activities were assayed 24 h post-transfection according to the Dual Reporter assay manufacturer’s instructions.

Real-time PCR device CFX connect (BioRad) was used to monitor protein unfolding through the increase in fluorescence of fluorophore SYPRO Orange (Invitrogen). Protein samples (GST-Bcl-3-ARD, 5 μg in 25 μl) were diluted in buffer containing 50 mm Tris–HCl pH 8.0, 150 mM NaCl, and incubated on 96-well microplates in the RT-PCR device in the presence of SYPRO Orange and ligand (final concentration of 10 μM). The ligand stock of 10 mM was dissolved in 100% DMSO. Optical foil was used to cover the microplates. Samples were heated at 1 °C per minute, from 20 °C to 80 °C. Fluorescence intensity was measured every 1 °C and plotted as a function of temperature by using the software package within the BioRad instrument. Each point represents the negative derivative or the negative slope at each temperature. The Tm (melting point of the protein sample) corresponds to the lowest point. Two grooves or lowest points can be seen where the lowest point to the left did not change depending on the addition of the ligand, whereas the second lowest point showed ligand-dependent change in Tm during temperature increase. From the analysis of the second lowest point, stabilization of 2–3 degrees was obtained when the ligands were added to the protein (black line) as compared to the protein with ligand (green line, control sample). All measurements were setup in duplicates. The effect of Compound BCL3ANT on the Tm of GST-Bcl-3-ARD was also measured at an 0.1 mg/mL GST-Bcl-3-ARD in 10 mM HEPES, pH 7.4 with 1.9 mM Compound XY. The measurement was done once in duplicate.

Cells were placed on ice and washed once with cold PBS before being harvested in cold 1 × lysis buffer (50 mM Tris–HCl pH 7.4, 150 mM NaCl, 1% Triton-x-100, complemented with 40 μl/ml complete protease inhibitors (Roche Applied Science). Lysates were cleared by centrifugation at 12,000 × g for 10 min at 4 °C and protein content was determined using a Bradford assay according to manufacturer’s instructions (Thermo Fisher Scientific). Equal amounts of protein were electrophoretically separated on 10% SDS/polyacrylamide gels and proteins were transferred onto Immobilion-FL polyvinylidene difluoride (PVDF) membranes (Millipore). Membranes were blocked with 5% non-fat milk in PBS-T for 1 h at room temperature followed by overnight incubation at 4 °C with primary antibodies against α-tubulin (1:4000, Sigma), Bcl-3 (1:500, Santa Cruz), cyclin D1 (1:1000, Abcam). Primary antibodies were detected with horseradish peroxidase-labeled secondary antibodies (1:5000, Dako). Chemiluminescence was detected with a charge-coupled device camera (Fujifilm).

Cells cultured on cover slips were washed twice with PBS and fixed for 4 min using 4% paraformaldehyde in PBS. Following fixation, cells were permeabilized using 0.25% Triton-x-100 in PBS for 10 min. After permeabilization, cells were washed three times in PBS and blocked for one hour in 1% bovine serum albumin (BSA) in PBS. Thereafter, cells were incubated for one hour with primary antibody in PBS followed by washing with PBS and incubation with Alexa Fluor 546 conjugated antibodies (Molecular probes) in PBS. Cover slips were mounted on object slides using Vectashield with diamidino-2-phenylindole (DAPI) (Vector Laboratories). Images were captured using a 40X oil objective on the Zeiss LSM 710 Confocal System (Zeiss).

Cultured cells were washed in cold PBS and total RNA was extracted using the RNeasy extraction kit (5 Prime) according to manufacturer’s instruction. The purity of RNA was analyzed and quantified using a Nanodrop spectrophotometer (Saveen Werner) and cDNA synthesis was performed using qPCR cDNA synthesis kit (Stratagene) following manufacturer’s instructions. RT-qPCR runs were performed in the Mx3005P real-time thermocycler (Stratagene) with the following program: 2 min 50 °C, 10 min at 95 °C, followed by 40 three-step cycles consisting of 95 °C for 20 s, 60 °C for 30 s, and 72 °C for 1 min. The following primers were used:

Cell proliferation was measured using the WST-1 assay (Roche). In brief, equal number of cells were seeded on 96-well flat-bottom plates in 100 μl of complete cell culture medium for 24 h prior to drug treatments. After 24–96 h of drug treatment, 10 μl of WST-1 reagent was added to each well and plates were incubated for 3 h at 37 °C. Absorbance was measured at 450 nm using the Synergy 2 HT Multi-Mode Microplate Reader (Biotek).

Mel Ju and SK-MEL-3 cells were seeded in 6-well plates at a density of 2 × 10⁵ cells per well and incubated overnight. Medium was changed to fresh medium supplemented with or without compounds. After 48 h the medium was collected and centrifuged at 400 × g for 5 min. Meanwhile, adherent cells were washed with PBS, trypsinized, suspended in complete medium and centrifuged at 400 × g for 5 min. After aspirating medium, the cell pellets from adherent and suspended cells were washed in PBS and centrifuged at 400 × g for 5 min. Cell pellets were carefully resuspended in 150 μl PBS before 1.35 ml 70% EtOH was added to fix the cells at 4 °C overnight. The following day, cells were centrifuged, resuspended in PBS, and centrifuged at 400 × g for 5 min. Cell pellets were resuspended in 50 μl PBS supplemented with 1 μg/ml DAPI (Sigma Aldrich) and 0.1% Triton-x-100. Cells were analysed using the DNA fragmentation protocol on the NucleoCounter 3000 (Chemometec).

The half-maximal inhibitory concentration (IC₅₀) was determined by seeding equal number of cells on 96-well plates. Following attachment, medium was replaced with fresh medium with or without BCL3ANT at ascending concentrations and cells were incubated for 5 days. After 5 days the WST-1 assay was performed and cells were incubated for 3 h at 37 °C with WST-1 reagent before absorbance was measured at 450 nm using the Synergy 2 HT Multi-Mode Microplate Reader (Biotek). Data is presented as IC50 (50% reduction in cell viability).

Mel Ju and SK-MEL-3 cells were seeded in 6-well plates at a density of 5 × 10⁴ cells per well and incubated for 24 h. At indicated time points cells were washed with PBS, trypsinized for 5 min and resuspended in complete medium before being centrifuged for 5 min at 400 × g. Cell pellets were resuspended in PBS, centrifuged, and counted using the “Viability and Cell Count” protocol on the NucleoCounter NC-3000 (Chemometec).

Cells were seeded on 6-well plates at a density of 2 × 10⁵ cells per well and incubated overnight. Medium was changed to fresh medium supplemented with or without A27. After 48 h medium was collected and centrifuged at 400 × g for 5 min. Adherent cells were washed with PBS, trypsinized, and suspended in complete medium before being centrifuged for 5 min at 400 × g. After aspirating medium from the cell pellet, cells were washed in PBS and centrifuged for 5 min at 400 × g. Cell pellets were carefully resuspended in PBS before 70% ethanol was added and cells were fixed overnight at 4 °C. Fixed cells were centrifuged for 5 min at 400 × g and resuspended in PBS supplemented with 1 μg/ml DAPI (Sigma Aldrich) and 0.1% Triton-x-100. Cells were analysed using the “fixed cell cycle-DAPI” protocol on the NucleoCounter NC-3000 (Chemometec).

NUDE mice were purchased from Janvier Laboratory (France). The protocol for the in vivo metastasis assay on mice has been approved by the Center of ethical committee on animal experiments in Sweden "Centrala försöksdjurnämnden". Ethical Dnr: M129-15. All animals were maintained under specific pathogen-free conditions at the Medicon Village animal facility at Lund University. All animal experiments were performed according to the national and international guidelines of the European Union. All mice were housed under pathogen-free conditions in the animal facility and received autoclaved water and food. Ten control and ten BCL3ANT treated 6-week-old mice were used in this study. The mice were subcutaneously transplanted with 1.10⁶ cells/ml of SK Mel 3 melanoma cell lines. Six days after transplantation, the mice were treated intraperitoneally (IP) with 10 mg/kg BCL3ANT or vehicle control. The vehicle solution used to dissolve BCL3ANT and as a control was composed by 2% DMSO in PBS. The mice were treated every 3 days with BCL3ANT or vehicle and sacrificed after 16 days post transplantation. Tumor volume was measured using CALIPER. Treatment was repeated every third day until day 16. The mice were euthanized by exposure to CO₂ without removing animals from their home cage.

Expression levels of Bcl-3 and survival data were gathered for melanoma patients with metastases, excluding those with brain metastases, from three pivotal studies [31‐33]. Z scores for Bcl-3 expression were calculated for each dataset. Bcl-3 expression was categorized as low for negative z-scores and high for positive z scores. The data were visualized using GraphPad Prism 9. Data represented are mean ± SEM of at least three separate experiments. Statistical comparisons were carried out by student´s T-test and significance is indicated by p > 0.05, *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001.

Initially, we investigated the survival curve of melanoma patients with metastasis stratified by the Bcl-3 expression level and found that the reduced survival rate of melanoma patients with metastases correlated with high levels of Bcl-3 expression (Fig. 1A). Analysis of Bcl-3 expression levels in nine human melanoma cell lines showed high levels of Bcl-3 expression in all except Mel Juso cells (Fig. 1B). We decided to use high-throughput screening of small molecules for the identification of Bcl-3-mediated cyclin D1 expression. To avoid interfering with the endogenous high expression of Bcl-3, we selected Mel Juso cells to control the level of Bcl-3 overexpression in these cells. A cell-based luciferase reporter assay was performed as a high-throughput screening model by utilizing a reporter construct containing only the part of the cyclin D1 promoter (-64/ + 10) that can be occupied by Bcl-3. As a control, a mutant form of the Bcl-3 binding site at the cyclin D1 promoter (-64/ + 10-mut) was selected for this screening assay. The procedure of the assay is presented in Fig. 1D. As expected, activation of cyclin D1 could be observed with -64/ + 10 cyclin D1 but not with the -64/ + 10-mut cyclin D1 κB mutant plasmid (Fig. 1C). Screening 1368 small molecules from a chemical library of the National Cancer Institute resulted in the identification of 79 small molecules that could decrease cyclin D1 promoter activity by 50% (Fig. 1E). After repeating the screening process with the 79 identified small molecules (Fig. 1F), we found 15 small molecules that could effectively reduce cyclin D1 promoter activity (Fig. 1G).

To reduce the number of small molecules, a fluorescence-based thermal shift assay was performed to analyze the linkage between the Bcl-3 protein and ligand binding. For this experiment, the ankyrin repeats of the Bcl-3 protein, which is the part of the protein that can interact with p50 and p52, were produced and purified. The fluorescence-based thermal shift assay identified 2 ligands among 15 to achieve a temperature shift up to 3 °C compared to the control (Fig. 2A). These two ligands were 2-(((4-(3,4-dichlorophenyl)-1,3-thiazol-2-yl) amino) methylene) malonamide hydrochloride and methoxyellipticine. In silico analysis using the GLIDE program revealed that 2-(((4-(3,4-dichlorophenyl)-1,3-thiazol-2-yl) amino) methylene) showed a better docking score than methoxyellipticine. The 2-(((4-(3,4-dichlorophenyl)-1,3-thiazol-2-yl) amino) methylene) compound is synonymous with NSC-659509. The molecular formulation of NSC-659509 is C₁₃H₁₁C_l3N₄O₂S (Fig. 2B) with a molecular weight of 394 Da. We also assessed the stability of the Bcl-3 protein in the presence of NSC-659509 with a thermal stability assay. At 1.9 mM, NSC-659509 showed a ΔT_m of 4.1 °C, a sample with Bcl-3 without ligand was used as the reference control. To confirm these results, the unfolding of the Bcl-3 protein indicated by the increase in the fluorescence of the fluorophore SYPRO Orange was investigated. As is shown in Fig. 2C, stabilization of 3 degrees was obtained when the ligands were added to the protein (black line); protein without ligand was used as the control (green line, Fig. 2C).

We confirmed that NSC-659509, hereafter referred to as BCL3ANT, could interfere with Bcl-3-mediated cyclin D1 promoter activity, but this effect was not observed when the cyclin D1 κB mutant was expressed (Fig. 3A). Additionally, BCL3ANT reduced cyclin D1 luciferase activity in Bcl-3-overexpressing melanoma cells in a concentration-dependent manner (Fig. 3B). We treated Mel Ju cells with increasing concentrations of BCL3ANT and found that the IC₅₀ value was 40.4 μM (Fig. 3C). Furthermore, we showed that Mel Ju cells treated with BCL3ANT had reduced cyclin D1 mRNA and protein levels compared to control-treated cells (Fig. 3D). No differences could be observed in Bcl-3 mRNA levels after treatment of Mel Ju cells with BCL3ANT compared to DMSO-treated cells (Fig. 3E). SiRNA-induced knockdown of Bcl-3 in immortalized keratinocytes [24] and treatment with BCL3ANT did not affect cell growth and cell proliferation of the cells (Fig. 3F right panel) compared to siRNA-control cells (Fig. 3F left panel). In conclusion, these results suggest that BCL3ANT-mediated cyclin D1 repression occurs at the promoter level.

After establishing that BCL3ANT interferes with the expression of cyclin D1, we wanted to examine the effect of Bcl-3 inhibition in melanoma cells. Since Bcl-3 is predominantly a nuclear protein but can be found within the cytoplasm of cells, we initially analyzed whether inhibition of Bcl-3 affects its cellular localization. Treatment of Mel Ju cells with BCL3ANT did not alter the subcellular localization of Bcl-3 compared to control treated cells (Fig. 4A). Additionally, no differences were found in the amount of DNA fragmentation or cell viability after treatment of Mel Ju cells with BCL3ANT within 24 h of treatment (Fig. 4B and C), indicating that cell survival is not affected by BCL3ANT treatment. Next, cell proliferation was monitored in the presence of BCL3ANT. We treated Mel Ju and SK-MEL-3 cells with BCL3ANT for 72 h, and observed a significant reduction in the cell proliferation rate compared to that of control cells (Fig. 4D). Accordingly, we observed an accumulation of cells in the G1 and S phase of the cell cycle after treatment of cells with BCL3ANT in comparison to control cells (Fig. 4E). To examine whether Bcl-3 inhibition affected cell migration, a wound healing assay was performed. Melanoma cells treated with BCL3ANT migrated more slowly than control-treated cells (Fig. 4F). Finally, we transplanted melanoma cells into immunocompromised nude mice and treated these animals with control or BCL3ANT for 16 days. A significant delay in tumor growth was observed in animals treated with BCL3ANT compared to control-treated animals (Fig. 4G). According a database [dtp.cancer.gov/databases], the animals were treated with 0.04 mg/kg, 0.08 mg/kg, or 0.16 mg/kg of BCL3ANT, and all animals beside one from 0.16 mg/kg were survived at each of the dose levels (see https://​dtp.​cancer.​gov/​databases_​tools/​data.​search.​htm) (Fig. 4H). Taken together, these data suggest that treatment of melanoma cells with BCL3ANT reduces cell proliferation by blocking the G1-to-S cell cycle transition via inhibition of cyclin D1 expression and limits the growth of the cells in vivo.