Identification of a Novel Bcl-2 Inhibitor by Ligand-Based Screening and Investigation of Its Anti-cancer Effect on Human Breast Cancer Cells

Abstract

Bcl-2 family protein is an important factor in regulating apoptosis and is associated with cancer. The anti-apoptotic proteins of Bcl-2 family, such as Bcl-2, are overexpression in numerous tumors, and contribute to cancer formation, development, and therapy resistance. Therefore, Bcl-2 is a promising target for drug development, and several Bcl-2 inhibitors are currently undergoing clinical trials. In this study, we carried out a QSAR-based virtual screening approach to develop potential Bcl-2 inhibitors from the SPECS database. Surface plasmon resonance (SPR) binding assay was performed to examine the interaction between Bcl-2 protein and the screened inhibitors. After that, we measured the anti-tumor activities of the 8 candidate compounds, and found that compound M1 has significant cytotoxic effect on breast cancer cells. We further proved that compound M1 downregulated Bcl-2 expression and activated apoptosis by inducing mitochondrial dysfunction. In conclusion, we identified a novel Bcl-2 inhibitor by QSAR screening, which exerted significant cytotoxic activity in breast cancer cells through inducing mitochondria-mediated apoptosis.

Keywords: Bcl-2; QSAR; breast cancer cell; small molecule inhibitors; virtual screening.

Figure 1

Flowchart of the virtual screening strategy.

Figure 2

The performance of different models. (A) Receiver operating characteristic (ROC) plot of RF classification models based on CATS, MACCS, and MOE2D descriptors for the test set. (B) Correlations between experimental and predicted activities of the training set (blue) and test set (red) of the RF regression model.

Figure 3

The plot of −Q² vs. the number of descriptors”.

Figure 4

(A) SOM projection of positive set and screened compounds. Black shading indicates compound density. Compounds were represented by CATS descriptors. The number in the grid represents the number of compounds. The position of red number in grid represents the source of the hit compounds we finally selected. In this research, the SOM technique can be used to identify preferred hit compounds. (B) The scaffold structure of selected molecules. Scaffold 1, 3, and 5 are located in cluster (1/4), scaffold 2 is located in cluster (1/1), scaffold 4, 6 are located in cluster (2/1), scaffold 7 is located in cluster (2/5), scaffold 8 is located in cluster (1/3), scaffold 9 is located in cluster (3/5), scaffold 10 is located in cluster (3/2), scaffold 11 is located in cluster (2/4).

Figure 5

The interaction between the compounds and Bcl-2 was determined by SPR assay. (A) The binding curves for the eight hit compounds are represented by a series of colored lines. (B) the binding curves for the different concentration gradients of the compounds M1 are represented by four colored lines.

Figure 6

Chemical structures of the compound M1-M8.

Figure 7

Compound M1 inhibits cell proliferation in breast cancer cells. (A) MDA-MB-231, T47D, and BT549 cells were treated with a series of doses of compound M1 for 72 h. (B) MDA-MB-231, T47D, or BT549 cells were treated with 5, 2, or 3 μM compound M1 for difference time periods. After treatment, cell viability was estimated by CCK-8 assay. Compound M1 vs. control, ^*p < 0.05, ^**p < 0.01, student t-test, n = 3, means ± SD. (C) MDA-MB-231 cells were treated with 4 or 8 μM compound M1 for 48 h. The EdU-marked replicating cells were observed by fluorescent microscope. Compound M1 vs. control, ^*p < 0.05, ^**p < 0.01, student t-test, n = 3, means ± SE. (D) MDA-MB-231 cells were treated with 4 or 8 μM compound M1 for 48 h, and the cell colonies were fixed and stained for camerawork. Compound M1 vs. control, ^*p < 0.05, ^**p < 0.01, student t-test, n = 3, means ± SE.

Figure 8

Compound M1 downregulations Bcl-2 expression and induces mitochondrial dysfunction in breast cancer cells. MDA-MB-231, T47D, and BT549 cells were treated with different doses of compound M1 for 48 h (A), or were treated with 8 μM M1 for difference time periods (B), and the expressions of Bcl-2 were estimated by western blotting. β-actin was shown as a loading control. Compound M1 vs. control, ^*p < 0.05, ^**p < 0.01, student t-test, n = 3, means ± SE. (C) MDA-MB-231 cells were treated with 2 or 4 μM compound M1 for 48 h. The MMP was observed by JC-1 staining by fluorescent microscopic. (D) MDA-MB-231 cells were treated with 2 or 4 μM compound M1 for 48 h. The levels of ROS were measured by staining with DCF-DA and the fluorescent intensity observed by fluorescent microscopic. The levels of ROS were measured by TECAN SPARK 10M. Compound M1 vs. control, ^*p < 0.05, ^**p < 0.01, student t-test, n = 3, means ± SD. (E) T47D cells were treated with 4 or 8 μM compound M1 for 48 h. The levels of Cytochrome c in mitochondrial or cytosol fraction were measured by western blotting. COX-IV and β-actin were shown as loading controls for the mitochondria and cytosol, respectively. Compound M1 vs. control, ^*p < 0.05, ^**p < 0.01, ^##p < 0.01 student t-test, n = 3, means ± SE.

Figure 9

Compound M1 induces apoptosis in breast cancer cells. (A) T47D cells were treated with different concentrations of compound M1 for 48 h, and apoptosis was measured by Annexin V-FITC and PI staining. Compound M1 vs. control, ^*p < 0.05, ^**p < 0.01, student t-test, n = 3, means ± SE. (B) T47D and BT549 cells were treated with different doses of compound M1 for 48 h, and the levels of cleaved caspase-3 and cleaved PARP were measured by western blotting. β-actin was shown as a loading control. Compound M1 vs. control, ^*p < 0.05, ^**p < 0.01, ^#p < 0.05 student t-test, n = 3, means ± SE.