Repurposing mebendazole against triple-negative breast cancer CNS metastasis

Abstract

Purpose: Triple-negative breast cancer (TNBC) often metastasizes to the central nervous system (CNS) and has the highest propensity among breast cancer subtypes to develop leptomeningeal disease (LMD). LMD is a spread of cancer into leptomeningeal space that speeds up the disease progression and severely aggravates the prognosis. LMD has limited treatment options. We sought to test whether the common anti-helminthic drug mebendazole (MBZ) may be effective against murine TNBC LMD.

Methods: A small-molecule screen involving TNBC cell lines identified benzimidazoles as potential therapeutic agents for further study. In vitro migration assays were used to evaluate cell migration capacity and the effect of MBZ. For in vivo testing, CNS metastasis was introduced into BALB/c athymic nude mice through internal carotid artery injections of brain-tropic MDA-MB-231-BR or MCF7-BR cells. Tumor growth and spread was monitored by bioluminescence imaging and immunohistochemistry. MBZ was given orally at 50 and 100 mg/kg doses. MBZ bioavailability was assayed by mass spectrometry.

Results: Bioinformatic analysis and migration assays revealed higher migratory capacity of TNBC compared to other breast cancer subtypes. MBZ effectively slowed down migration of TNBC cell line MDA-MB-231 and its brain tropic derivative MDA-MB-231-BR. In animal studies, MBZ reduced leptomeningeal spread, and extended survival in brain metastasis model produced by MDA-MB-231-BR cells. MBZ did not have an effect in the non-migratory MCF7-BR model.

Conclusions: We demonstrated that MBZ is a safe and effective oral agent in an animal model of TNBC CNS metastasis. Our findings are concordant with previous efforts involving MBZ and CNS pathology and support the drug's potential utility to slow down leptomeningeal spread.

Keywords: Breast cancer; Drug repurposing; Leptomeningeal disease; Mebendazole.

Fig. 1

Benzimidazoles as potential treatment for migratory cancers, such as TNBC. a Small molecule screen used to identify pharmacologic compounds active against TNBC cell line SUM149PT, but not against a non-tumorigenic breast cell line MCF10a [22]. b Benzimidazoles are more effective against metastatic TNBC cell line MDA-MD-231 compared to a metastatic luminal A breast cancer cell line MCF7. (a,b) Diagonal line is placed for agents equally effective against indicated cell lines. Benzimidazoles are labeled in red. c Breast cancer cells with high migratory capacity (such as TNBC and HER2 subtype) are more sensitive to tubulin binders. Result of a query of the Genomics of Drug Sensitivity in Cancer (GDSC) database, which included 53 breast cancer cell lines and microtubule inhibitors docetaxel, paclitaxel, vinblastine, and vinorelbine. d GO enrichment terms associated with migration are overrepresented in nematodes and underrepresented in flatworms. The data are taken from the study [31] comparing gene expression in nematodes vs flatworms. e Disruption of ligand-receptor interactions (important for cell migration and metastasis) represent a common consequence of benzimidazole treatment in mammalian cells (a query of the Drug-Path database [34], The pathways strongly associated with cell migration (red color) are affected by the majority of tested benzimidazoles (albendazole (A), fenbendazole (F), mebendazole (M), nocodazole (N), parmendazole (N), and thiabendazole (T)) and have a low false discovery rate (FDR). f Mechanism: Lower concentrations of tubulin binders are needed to inhibit migration. Therefore, migratory cancers should be more responsive to these drugs

Fig. 2

Mebendazole (MBZ) as a potential treatment against leptomeningeal cancer spread. (a, b) Sensitivity to MBZ of TNBC cell line MDA-MB-231 and of hormone receptor positive cell line MCF7 (a), and of their brain-tropic derivatives MDA-MB-231-BR and MCF7-BR (b). Brain-tropic MDA-MB-231-BR cell line is slightly more resistant to MBZ than the parental cell line MDA-MB-231: MDA-MB-231-BR (IC50 = 0.16 µM), MDA-MB-231 (IC50 = 0.14 µM), MCF7-BR (IC50 = 0.19 µM) and MCF7 (IC50 = 0.19 µM). c Infra-red spectra (FT-IR) of MBZ polymorphs revealing the presence of MBZ polymorph C (MBZ-C) and polymorph B (MBZ-B) in the MBZ from Sigma, CAS # 31,431–39-7. d MBZ given at an oral dose of 100 mg/kg reaches therapeutic concentrations in the cerebrospinal fluid (CSF) of NuNu mice (median [MBZ] = 105.9 ng/ml ~ 0.36 µM). Plasma MBZ concentrations represent total quantity of MBZ, and CSF concentrations represent free, unbound MBZ. A horizontal line at 59 ng/mL corresponds to the IC50 = 0.20 µM of MBZ. MDA231 = MDA-MB-231. MDA231-BR = MDA-MB-231-BR. Significance: ***, p < 0.001, NS = not significant

Fig. 3

MBZ reduces the migration of TNBC MDA-MB-231 and MDA-MB-231-BR cells. a, b TNBCs have higher migratory capacity compared to other breast cancer subtypes. Migration scores from Nair et al. [7] were compared among breast cancer subtypes in TCGA patient data (a) and cell lines (b). c, d MDA-MB-231 and MDA-MB-231-BR have higher migration capabilities than MCF7 and MCF7-BR cells. Migratory capability of MDA-MB-231 cells increases upon acquiring brain-tropic status and is effectively inhibited by MBZ. Neither MCF-7 nor MCF-7-BR migrated significantly during 20 h. (c) Representative inverse fluorescence images of DAPI-stained membranes from Boyden chamber during 20 h. DAPI-stained cell nuclei are shown as dark gray spots in the background of white membrane pores. Scale bar: 200 μm. (d) Quantitation of migration in MCF7, MDA-MB-231, and brain-tropic MCF7-BR and MDA-MB-231-BR cells from (c). The numbers of migrated cells are normalized to survival from (e). e Survival during 20 h treatment with MBZ. Plating efficiencies of untreated cell lines were not significantly different. MDA231 = MDA-MB-231, MDA231-BR = MDA-MB-231-BR. FOV = field of view. Significance analysis: ANOVA, **, p < 0.01, ***, p < 0.001, ****, p < 0.0001

Fig. 4

The internal carotid injection (ICA) model of brain metastasis describes leptomeningeal disease (LMD). a Schematics of an internal carotid artery injection of tumor cells. b H&E-stained section from mouse brain affected by LMD. Arrows point to cancer cells in leptomeningeal space. c Patient brain T1W + C MRI sequence shows the anatomical location of LMD (red arrow). (d-h) Immunofluorescence images depict dissemination of neoplastic cells into leptomeningeal space. BP = brain parenchyma. Arrows point to cancer cells in leptomeningeal space. d, e Vimentin, a marker of epithelial-to-mesenchymal transition, is highly expressed in MDA-MB-231 cells. Antibody against human vimentin (hVim) identifies MDA-MB-231-BR breast cancer cells. (f) Antibody against pan-cytokeratin (PanCK) identifies MCF7-BR breast cancer cells. e, f Antibody against laminin (Lam) shows the location of pia. g Bioluminescence images reveal intracranial disease and spinal dissemination (red arrows). (e) Spinal metastases identified by bioluminescence were verified by subsequent H&E staining. Right panel is a magnified version of a region indicated in the left panel. Black arrows point to the same spinal metastasis in the 4 × image and a magnified (× 20) image

Fig. 5

MBZ effectively reduces leptomeningeal dissemination in the TNBC mouse model. a Experimental timeline. Tumor size at the start of treatment is between 200–500 cells and corresponds to the BLI signal of (2 – 5) × 10⁵ for MDA-MB-231-BR cell line and (5–10) × 10⁵ for the MCF7-BR cell line. b Representative bioluminescence images of MDA-MB-231-BR mice from control and treatment groups (at 50 and 100 mg/kg). See Fig.S5 for original images. c, d MBZ slows metastatic growth as detected by bioluminescence imaging (c) and improves survival (d) in the MDA-MB-231-BR mouse model. Number of animals per group: control (n = 11), 50 mg/kg MBZ (n = 7), 100 mg/kg (n = 9). e, f MBZ shows no effect on metastatic growth (e) and survival (f) in the MCF7-BR model. Number of animals per group: control (n = 6), 50 mg/kg MBZ (n = 5). Experiments in (c, e) were analyzed using repeated measures method. Post hoc pairwise comparisons were performed using a Tukey adjustment for multiple comparisons. Black arrows in (c, e) point to a day of treatment start. g MBZ effectively reduces leptomeningeal dissemination (single cell spread and small metastases) in MDA-MB-231-BR model, with no significant effect on big metastases (see Fig. S3 for images). Small metastases are defined as clusters of cells with ≤ 50 cells/cluster without co-option of blood vessels. Data collected from n ≥ 3 mice. Single-cell and small metastatic populations are (20–50)-fold lower in MCF7-BR compared to MDA-MB-231-BR CNS metastasis. Data are mean ± SD. Significance analysis: t-test;*, p < 0.05, ***, p < 0.001, NS = not significant