Induction of HEXIM1 activities by HMBA derivative 4a1: Functional consequences and mechanism

Abstract

We have been studying the role of Hexamethylene bisacetamide (HMBA) Induced Protein 1 (HEXIM1) as a tumor suppressor whose expression is decreased in tamoxifen resistant and metastatic breast cancer. HMBA was considered the most potent and specific inducer for HMBA inducible protein 1 (HEXIM1) prior to our studies. Moreover, the ability of HMBA to induce differentiation is advantageous for its therapeutic use when compared to cytotoxic agents. However, HMBA induced HEXIM1 expression required at mM concentrations and induced dose limiting toxicity, thrombocytopenia. Thus we structurally optimized HMBA and identified a more potent inducer of HEXIM1 expression, 4a1. The studies reported herein tested the ability of 4a1 to induce HEXIM1 activities using a combination of biochemical, cell phenotypic, and in vivo assays. 4a1 induced breast cell differentiation, including the stem cell fraction in triple negative breast cancer cells. Clinically relevant HEXIM1 activities that are also induced by 4a1 include enhancement of the inhibitory effects of tamoxifen and inhibition of breast tumor metastasis. We also provide mechanistic basis for the phenotypic effects of 4a1. Our results support the potential of an unsymmetrical HMBA derivative, such as 4a1, as lead compound for further drug development.

Keywords: Antiestrogens; Breast cells; Differentiation; HEXIM1; HMBA; Metastasis.

Figure 1. Regulation of the expression of HEXIM1 by HMBA and 4a1

(A) Structures of HMBA and 4a1. (B) MCF7 were treated with vehicle (DMSO), HMBA, or 4a1 using indicated concentrations for 8 h. (C) MDA-MB-231, MDA-MB-468, or BT-474 cells were treated with vehicle (DMSO), 5 mM HMBA, or 50 uM 4a1 for 8 h. Cells were processed for Western blot analyses of HEXIM1 expression. The figures are representative of 3 experiments. * p < 0.01 relative to DMSO treated cells.

Figure 2. HEXIM1 is required for HMBA- and 4a1- induced p21/p27 expression and differentiation

MCF7 cells transfected with control siRNA or HEXIM1 siRNA were treated with vehicle (DMSO), HMBA (5 mM), or 4a1 (50 uM) for 8, 18 or 72 h. (A) Cells were processed for Western blot analyses of HEXIM1 or p21 expression or (B) stained with Nile Red to determine lipid vacoule formation (differentiation marker). MDA-MB-231 cells transfected with control or FLAG-tagged HEXIM1 expression vector or MDA-MB-468 cells transfected with control siRNA or HEXIM1 siRNA were treated with vehicle (DMSO), HMBA (5 mM), or 4a1 (50 uM) for 8, 18 or 72 h. (D) Cells were processed for Western blot analyses of HEXIM1 or p27 expression or (B) stained with Nile Red. All photographs of stained cells were taken at 40x magnification. The figures are representative of 3 experiments.

Figure 3. HEXIM1 restored sensitivity to the inhibitory effects of tamoxifen in a tamoxifen resistant cell line

(A) Western blot analyses of HEXIM1 expression in MCF7 and MCF7/TOTR cells. MCF7/TOTR clone 1 cells were also treated with vehicle (DMSO), 5 mM HMBA, or 50 uM 4a1. (B) MCF7 or MCF7/TOTR (clone 1) or (C) MDA-MB-231, MDA-MB-468, MCF10A, and HBL-100 cells were treated as indicated for 7 days. MTT assays were then performed to assess proliferation. The right panel shows western blot analyses of HEXIM1 expression in MCF10A and HBL-100 cells. The figures are representative of 3 experiments. * p < 0.01 relative to DMSO treated cells.

Figure 4. Induction of differentiation of stem cells in TNBC cells by 4a1

(A) MDA-MB-231 and MDA-MB-231-Nanog-GFP1 cells were treated with vehicle (DMSO) or 4a1 for 8 or 18 h and processed for Western blot analyses of indicated proteins. (B) MDA-MB-231-Nanog-GFP1 cells were treated with vehicle (DMS0) or 4a1 for 72 h and then stained with Nile Red. All photographs were taken at 40x magnification. The figures are representative of 3 experiments. * p < 0.01 relative to DMSO treated cells.

Figure 5. Regulation of the expression of HEXIM1 direct targets by HMBA and 4a1

(A) and (B) MCF7 cells were treated with vehicle (DMSO), 5 mM HMBA, or 50 uM 4a1 for 8 h, (C) MCF7 cells transfected with control siRNA or HEXIM1 siRNA were treated with vehicle (DMSO) or 50 uM 4a1 for 8 h. Cells were processed for RT-PCR analyses or Western blot analyses of HEXIM1 targets relative to GAPDH. The figures are representative of 3 experiments. * p < 0.01 relative to DMSO treated cells.

Figure 6. Injection of PLGA-4a1 resulted in increased HEXIM1 expression and decreased metastasis

After the appearance of palpable mammary tumors in PyMT mice, PLGA or PLGA-4a1 (50 uM, 50 ul volume) were injected into the tumors every other week. Mammary glands, blood, and lungs were then collected. (A) HEXIM1 expression in mammary tumors of PLGA ± 4a1 treated mice were determined by Western blot analyses. (B) Body weights were monitored weekly as indicated. (C) Platelet levels were determined using the HEMAVET 950FS Multi-species Hematology System. (D) Left panel shows lungs from PLGA or PLGA-4a1 treated mice. Right panel shows quantification of tumor area in H&E stained lung tissue sections. (E) Western blot analyses of PyMT expression in the lungs of PLGA or PLGA-4a1 treated mice. In (A) and (E), expression of HEXIM1 and PyMT, respectively, are expressed relative to expression of GAPDH, a loading control. Panels represent 5 mice per group (PLGA ± 4a1). * p < 0.01 relative to PLGA treated cells.

Figure 7. HMBA and 4a1 induced CDK9 recruitment to the HEXIM1 promoter

MCF7 and MDA-MB-231 cells were treated with vehicle (DMSO), 5 mM HMBA, or 50 uM 4a1 for 90 min. ChIP assays were performed with CDK9 antibodies. PCR primers were for the coding region of HEXIM1. The figures are representative of 3 experiments. * p < 0.01 relative to DMSO treated cells.