Combined XPO1 Inhibition and Parthenolide Treatment Can Be Efficacious in Treating Triple-Negative Breast Cancer

Abstract

Triple-negative breast cancer (TNBC) is an aggressive, heterogeneous subtype of breast cancer with limited treatment options. Our previous work explored repurposing selinexor, an XPO1 inhibitor, as a novel therapeutic option for TNBC. To enhance its efficacy, this study aimed to identify beneficial combination therapies with selinexor and experimentally evaluate their effects in TNBC. Using the computational tool IDACombo, we nominated drugs predicted to improve the efficacy of XPO1 inhibition. The top candidate, parthenolide, was tested in vitro using three transcriptionally distinct TNBC cell lines. Fluorescently labeled cells were co-cultured and treated with selinexor, parthenolide, or their combination. Growth inhibition was assessed across the mixed population and by individual cell line after 96 h, and potential synergy was evaluated using Combenefit. While selinexor and parthenolide monotherapy inhibited the growth of TNBC subtypes, the combination was more effective in suppressing the overall cell population. Synergistic interactions between the two agents were observed in specific TNBC lines but not all, reflecting the combination effect in heterogeneous TNBC patients. Our findings suggest the selinexor-parthenolide combination as a potential therapeutic strategy for TNBC, warranting further investigation. Our study also demonstrates the value of integrative computational-experimental approaches in guiding heterogeneity-informed drug combinations for preclinical evaluation.

Keywords: NFKBIA; XPO1; combination therapy; parthenolide; selective inhibitors of nuclear export (SINEs); triple negative breast cancer (TNBC).

Figure 1

Parthenolide with an XPO1 inhibitor is predicted to be the most efficacious drug combination for breast cancer. (A) Bar plot showing the top 10 IDACombo scores predicted for combinations of leptomycin B (an XPO1 inhibitor) with all other drugs screened in the CTRPv2 dataset (n = 481) using breast cancer cell lines (n = 40). Higher IDACombo score is indicative of higher predicted combination efficacy. (B) 3D plot illustrating measured and predicted average cellular viabilities across 27 breast cancer cell lines. The blue and green spheres represent the viability for leptomycin B and parthenolide monotherapies, respectively, at various concentrations. The pink spheres represent the predicted viability of the combination treatment. The gray plane represents the lowest average viability achievable with either monotherapy alone; therefore, pink spheres that fall below this plane indicate that the combination therapy is predicted to reduce cellular viability more effectively than either monotherapy alone.

Figure 2

Hierarchical clustering of TNBC patient tumor samples and cell lines captures meaningful subpopulations for representing intertumoral heterogeneity. (A) Batch-corrected hierarchical clustering of gene expression data from 326 basal-like patient tumors (METABRIC cohort) and 27 CCLE TNBC cell lines. Representative cell lines selected from each cluster for further functional study are highlighted in black. (B) Box plot showing the calculated growth rate constant (k) for fluorescently labeled TNBC cell lines (MDA-MB-231-BFP, MDA-MB-468-GFP, HCC-1806-RFP) grown in co-culture compared to their respective parental lines in monoculture. Fluorescently labeled cells grown in co-culture proliferated exponentially over a 96 h period in the absence of drug exposure, showing no significant difference in their calculated growth rate compared to their parent lines. Statistical analysis was performed using an unpaired t-test: ns, not significant.

Figure 3

Combination of parthenolide and selinexor exhibit enhanced cellular growth inhibition in a mixed TNBC co-culture model. Three TNBC cell lines were fluorescently labeled using distinct lentiviral vectors: MDA-MB-231 (BFP, blue), MDA-MB-468 (GFP, green), and HCC-1806 (RFP, orange). The cells were mixed in a co-culture model and treated with either media only (Control), parthenolide (6 µM), selinexor (50 nM), or a combination of both agents. Fluorescent images were captured using the Cytation™ Cell Imaging Multi-Mode Reader (Gen5 software, version 3.14). Percent viability was assessed based on normalized fluorescent cell area. (A) Bar graph showing percent viability of the total cell population. (B) Representative fluorescent images of the co-culture under each treatment condition. (C–E) Bar graphs showing individual viability responses of each cell line within the co-culture: (C) MDA-MB-231, (D) MDA-MB-468, and (E) HCC-1806. Data represents mean ± SD from three biological replicates (n = 3). Data and images were collected 96 h post-treatment. Statistical analysis was performed using one-way ANOVA followed by post hoc multiple comparisons. Significance levels: * p < 0.05; ** p < 0.01; *** p < 0.001.

Figure 4

Parthenolide and selinexor combination demonstrates cell line-specific synergistic interactions in TNBC based on Combenefit analysis using Highest Single Agent (HSA) model. Combenefit software (version 2.02) was used to evaluate the combinatorial effects of parthenolide and selinexor on cell viability in three TNBC cell lines independently: MDA-MB-231, MDA-MB-468, and HCC-1806. Cells were treated individually with increasing concentrations of parthenolide, selinexor, and their combination. Cell viability was measured as normalized cell area from three biological replicates (n = 3). Top panel (A–C): 3D surface plots represent drug combination effects in (A) MDA-MB-231, (B) MDA-MB-468, and (C) HCC-1806, mapped according to HSA model. The x- and y-axes represent parthenolide and selinexor concentrations, respectively, and the z-axis represents percent cell viability (% Control). Blue regions indicate synergistic interactions while regions leaning towards red indicate possible antagonistic interactions. Bottom panel (D–F): Corresponding HSA synergy scores are shown as heatmap matrices for (D) MDA-MB-231, (E) MDA-MB-468, and (F) HCC-1806. Numbers in each cell represent the synergy score ± SD from three independent experiments (n = 3). Values are color-coded according to the obtained score, with blue indicating statistically significant synergistic interactions. The vertical color scale bar to the right of each plot or matrix reflects the gradient of synergy or antagonism with synergy towards blue color and antagonism towards red color. Statistical significance was assessed using a one-sample t-test (Significance levels: * p < 0.05).