TPX2 overexpression promotes sensitivity to dasatinib in breast cancer by activating YAP transcriptional signaling

Abstract

Chromosomal instability (CIN) is a hallmark of cancer aggressiveness, providing genetic plasticity and tumor heterogeneity that allows the tumor to evolve and adapt to stress conditions. CIN is considered a cancer therapeutic biomarker because healthy cells do not exhibit CIN. Despite recent efforts to identify therapeutic strategies related to CIN, the results obtained have been very limited. CIN is characterized by a genetic signature where a collection of genes, mostly mitotic regulators, are overexpressed in CIN-positive tumors, providing aggressiveness and poor prognosis. We attempted to identify new therapeutic strategies related to CIN genes by performing a drug screen, using cells that individually express CIN-associated genes in an inducible manner. We find that the overexpression of targeting protein for Xklp2 (TPX2) enhances sensitivity to the proto-oncogene c-Src (SRC) inhibitor dasatinib due to activation of the Yes-associated protein 1 (YAP) pathway. Furthermore, using breast cancer data from The Cancer Genome Atlas (TCGA) and a cohort of cancer-derived patient samples, we find that both TPX2 overexpression and YAP activation are present in a significant percentage of cancer tumor samples and are associated with poor prognosis; therefore, they are putative biomarkers for selection for dasatinib therapy.

Keywords: Hippo‐YAP/TAZ; SFK kinases; TPX2; chromosomal instability; dasatinib; mitosis.

Fig. 1

TPX2 inducible overexpression leads to increased sensitivity to SRC kinase inhibitors. (A) MDA‐MB‐453 cells increase TPX2 expression upon the inducible activation of the TET‐ON system by doxycycline (DOX). Representative blot of three experimental replicates. (B) Drug screening scatter plot of noDOX/+DOX ratios showing the response at each drug concentration tested in MDA‐MB‐453 cells during two cellular doubling rounds (6 days). Ratios for each compound (duplicates) are represented as gray dots, except dasatinib (highlighted in red). (C) 10‐point Concentration‐Response Curves (CRC) and IC50 calculation of dasatinib in MDA‐MB‐453 cells with two doses of DOX and drug incubation during 9 days. Mean cell number normalized vs. DMSO ± SEM (blue = noDOX control, orange = 0.01 μg·mL⁻¹ DOX, red = 0.1 μg·mL⁻¹ DOX). Representative graph of two experimental replicates. (D) 10‐point CRC and IC50 calculation of saracatinib in MDA‐MB‐453 cells with two doses of DOX and drug incubation during 9 days. Mean cell number normalized vs. DMSO ± SEM (blue = noDOX control, orange = 0.01 μg·mL⁻¹ DOX, red = 0.1 μg·mL⁻¹ DOX). Representative graph of two experimental replicates. (E) 10‐point CRC and IC50 calculation of bosutinib in MDA‐MB‐453 cells, with two doses of DOX and drug incubation during 9 days. Mean cell number normalized vs. DMSO ± SEM (blue = noDOX control, red = 0.1 μg·mL⁻¹ DOX). Representative graph of two experimental replicates. (F) Colony formation assay, during 2 weeks, in MDA‐MB‐453 cells upon 0.1, 0.3, and 0.9 μm of Dasatinib treatment. The colony area is normalized vs. the DMSO‐treated cells ± SD. Two‐way ANOVA with Tukey multiple comparisons test: P < 0.0001 (****), P < 0.001 (***), P < 0.05 (*). (blue = noDOX control, orange = 0.01 μg·mL⁻¹ DOX, red = 0.1 μg·mL⁻¹ DOX). Representative graph of three experimental replicates. (G) Colony formation assay, during 2 weeks, in MDA‐MB‐453 cells upon 0.1, 0.3, and 0.9 μm bosutinib treatment. The colony area is normalized vs. the DMSO‐treated cells ± SD. Two‐way ANOVA with Tukey multiple comparisons test: P < 0.05 (*). (blue = noDOX control, red = 0.1 μg·mL⁻¹ DOX). Representative graph of two experimental replicates.

Fig. 2

Dasatinib sensitivity upon TPX2 overexpression in breast cancer cell lines. (A) DepMap portal retrieved data from 36 breast cancer cell lines, showing the correlation between TPX2 expression (transcript per million – TPM) vs the sensitivity to dasatinib (area under the curve – AUC). The lower AUC, the more sensitivity to the drug. Correlation coefficient (R) was obtained by a Pearson correlation test. Statistical significance P < 0.05. (B) Inducible overexpression of TPX2, upon indicated doxycycline (DOX) concentrations, in the ZR‐75‐1, MDA‐MB‐361, and HCC‐1937 breast cancer cell lines. Representative graph of two experimental replicates. (C) Colony formation assay, during 2 weeks, in ZR‐75‐1 cells treated with the indicated concentrations of dasatinib. Representative graph of two experimental replicates. (D) Colony formation assay, during 2 weeks, in MDA‐MB‐361 cells treated with the indicated concentrations of dasatinib. Representative graph of two experimental replicates. (E) Colony formation assay, during 2 weeks, in HCC‐1937 cells treated with the indicated concentrations of dasatinib. Representative graph of two experimental replicates. The colony area is normalized vs. the DMSO‐treated cells ± SD. Two‐way ANOVA with Tukey multiple comparisons test: P < 0.0001 (****), P < 0.001 (***), P < 0.01 (**), P < 0.05 (*). (blue = noDOX control, red = DOX).

Fig. 3

Evaluation of SFK downstream signaling upon TPX2 inducible overexpression. (A) Correlation analysis of TPX2 and SFK kinases genes (SRC, YES, FYN, LYN, SRM, and LCK) expression levels, using breast cancer cell lines data retrieved from the DepMap portal. Correlation coefficient (R) was obtained by a Pearson correlation test. Statistical significance P < 0.05. (B) Downstream SRC signaling analysis by western blot of MDA‐MB‐453 cells expressing TPX2 (DOX) or control cells (ctrl), and treated with 0.1, 0.3, and 0.9 μm of dasatinib. GAPDH expression levels are used as a loading control. Representative graph of three experimental replicates. (C) 10‐point CRC and IC50 calculation of the JNK kinase inhibitor JNK‐IN‐8 in MDA‐MB‐453 cells, expressing TPX2 with two doses of DOX, and drug incubation during 9 days. Mean cell number normalized vs. DMSO ± SEM (blue = noDOX control, orange = 0.01 μg·mL⁻¹ DOX, red = 0.1 μg·mL⁻¹ DOX). Representative graph of two experimental replicates. (D) Colony formation assay, during 2 weeks, in MDA‐MB‐453 cells upon 0.1, 0.3 and 0.9 μm of JNK‐IN‐8 treatment. The colony area is normalized vs. the DMSO‐treated cells ± SD. Two‐way ANOVA with Tukey multiple comparisons test: P < 0.01 (**), P < 0.05 (*). (blue = noDOX control, orange = 0.01 μg·mL⁻¹ DOX, red = 0.1 μg·mL⁻¹ DOX). Representative graph of two experimental replicates.

Fig. 4

TPX2 overexpression leads to increased YAP signaling. (A) Correlation analysis of TPX2 expression levels of and the YAP1 and TAZ transcription factors, using breast cancer cell lines data retrieved from the DepMap portal. Correlation coefficient (R) was obtained by a Pearson correlation test. Statistical significance P < 0.05. (B) Hippo/YAP signaling analysis by western blot of MDA‐MB‐453 cells expressing TPX2 (DOX) or control cells (ctrl), treated with 0.1, 0.3, and 0.9 μm of Dasatinib. GAPDH expression levels are used as a loading control. Representative graph of three experimental replicates. (C) YAP nuclear/cytoplasm ratio analysis in MDA‐MB‐453 cells upon TPX2 overexpression and dasatinib treatment. The red dotted line depicts the cell nucleus. Normalized data to the control untreated cells is represented in the histogram showing control cells (blue bars – ctrl) or TPX2 expressing cells (orange bars – DOX) at 0.1, 0.3 and 0.9 μm of Dasatinib (light colored bars). Each dot represents a single microscopy field. One‐way ANOVA with Tukey multiple comparisons test: P < 0.001 (****). Representative graph of three experimental replicates. Scale Bar = 10 μm. (D) Immunofluorescence of active‐YAP in MDA‐MB‐453 cells upon TPX2 expression with DOX. The nuclear signal is quantified with imagej software, and normalized vs. control data is plotted ± SD. Unpaired T‐Test analysis: P < 0.0001 (****). Representative graph of two experimental replicates. (E) RT‐qPCR gene expression test of Cyr61, CGTF, and TEAD4 as surrogate markers of YAP transcription activity, showing mean mRNA levels (arbitrary units) ± SD. One‐way ANOVA with Tukey's multiple comparisons post hoc test P < 0.05 (*); P < 0.01 (**), P < 0.001 (***), P < 0.0001 (****). Representative graph of two experimental replicates. (F) Colony formation assay, during 2 weeks, in MDA‐MB‐453 cells upon 0.5, 1.0, and 5.0 μm of verteporfin treatment. The colony area is normalized vs. the DMSO‐treated cells ± SD. Two‐way ANOVA with Tukey multiple comparisons test: P < 0.01 (**). (blue = noDOX control, red = 0.1 μg·mL⁻¹ DOX). Representative graph of two experimental replicates.

Fig. 5

Dasatinib strengthens the mitotic arrest mediated by TPX2 overexpression. (A) Cell cycle profiling by DAPI DNA staining and flow cytometry analysis of MDA‐MB‐453 cells. Cells incubated with DOX (0.01 and 0.1 μg·mL⁻¹) for TPX2 expression, were also treated with 0.1, 0.3 and 0.9 μm of Dasatinib. The G₁ and G₂/M peaks are indicated at the bottom of the cell cycle profiles, and the quantification of the G₂/M percentage of cells (±SD) is shown in the bottom histogram. Two‐way ANOVA with Tukey multiple comparisons test comparing dasatinib impact: P < 0.0001 (****) P < 0.01 (**), P < 0.05 (*); or TPX2 expression impact versus no DOX treated cells: P < 0.05 (#). (blue = no DOX control, orange = 0.1 μg·mL⁻¹ DOX, red = 1.0 μg·mL⁻¹ DOX). Representative graph of three experimental replicates. (B) Mitotic index quantification by phospho‐Ser10 Histone H3 (pH3) immunofluorescence upon DOX incubation and dasatinib treatment. The upper panel shows a representative image of MDA‐MB.453 cells stained with DAPI for DNA (light blue) and pH3 depicting mitotic cells (dark purple). The bottom histogram shows the quantification (±SD) at different concentrations of dasatinib. Two‐way ANOVA with Tukey multiple comparisons test, comparing TPX2 expression impact: P < 0.0001 (****) P < 0.001 (***), P < 0.05 (*); or dasatinib impact with in the TPX2 expressing cohort; P < 0.0001 (####) P < 0.01 (##) P < 0.05 (#). (blue = no DOX control, red = 1.0 μg·mL⁻¹ DOX). Representative graph of three experimental replicates. Scale Bar = 25 μm. (C) Biochemical analysis by western blot of mitotic markers activation, and YAP‐Ser397 phosphorylation, of MDA‐MB‐453 cells expressing TPX2 (DOX) or control cells (ctrl), treated with 0.1, 0.3, and 0.9 μm of Dasatinib. GAPDH expression levels are used as a loading control. Representative graph of three experimental replicates.

Fig. 6

Correlation analysis of TPX2 expression and YAP activation of a breast cancer tumoral microarray. (A) Immunohistochemistry of TPX2 and pS127‐YAP in 99 grade‐3 breast cancer paraffin‐embedded samples. The upper row shows an example of breast carcinoma with a positive signal for both biomarkers. The middle row is an example of breast carcinoma with no TPX2 expression and with S127‐pYAP positive signal (indicative of YAP inactivation). The bottom panel indicates an example of breast carcinoma with positive expression of TPX2 and negative S127‐pYAP expression (indicative of YAP activation). Pictures were obtained at 40× magnification. Scale Bar = 100 μm. (B) Quantification and statistical analysis table of 99 samples stained as in panel (A) showing the relationship between pS127‐YAP signal and other immunohistochemical features, such as expression of YAP, TPX2 and HER2, metastasis capacity, and tumoral subtypes. (C) Gene Set Enrichment Analysis (GSEA) of breast invasive ductal carcinoma samples from the METABRIC project, accordingly to TPX2 expression levels, showing enhanced YAP/TAZ signaling [53] and SRC signaling (SRC_UP.V1_DN) [111]. Normalized Enrichment Score (NES) and FDR qValue are indicated in each plot. (D) Correlation analysis of TPX2 and the YAP/TAZ signature expression (z‐score) in breast cancer samples from the METABRIC project, evaluating the influence on Relapse Free survival (RF0 = no relapse; RF1 = positive relapse). Statistical analysis by adjusted Fisher test: P < 0.0001 (****); P < 0.01 (**).