Upregulation of MLK4 promotes migratory and invasive potential of breast cancer cells

Abstract

Metastasis to distant organs is a major cause for solid cancer mortality, and the acquisition of migratory and invasive phenotype is a key factor in initiation of malignancy. In this study we investigated the contribution of Mixed-Lineage Kinase 4 (MLK4) to aggressive phenotype of breast cancer cells. Our TCGA cancer genomic data analysis revealed that amplification or mRNA upregulation of MLK4 occurred in 23% of invasive breast carcinoma cases. To find the association between MLK4 expression and the specific subtype of breast cancer, we performed a transcriptomic analysis of multiple datasets, which showed that MLK4 is highly expressed in triple-negative breast cancer compared to other molecular subtypes. Depletion of MLK4 in cell lines with high MLK4 expression impaired proliferation and anchorage-dependent colony formation. Moreover, silencing of MLK4 expression significantly reduced the migratory potential and invasiveness of breast cancer cells as well as the number of spheroids formed in 3D cultures. These in vitro findings translate into slower rate of tumor growth in mice upon MLK4 knock-down. Furthermore, we established that MLK4 activates NF-κB signaling and promotes a mesenchymal phenotype in breast cancer cells. Immunohistochemical staining of samples obtained from breast cancer patients revealed a strong positive correlation between high expression of MLK4 and metastatic potential of tumors, which was predominantly observed in TNBC subgroup. Taken together, our results show that high expression of MLK4 promotes migratory and invasive phenotype of breast cancer and may represent a novel target for anticancer treatment.

Fig. 1

MLK4 is upregulated in invasive breast carcinoma. a, c Amplification and mRNA expression profiling of MLK4 in all invasive breast carcinoma cases (a) and TNBC cases (c) from TCGA dataset. Graphic illustrations taken from cBio Portal. b Expression of MLK4 in breast cancer subtypes for samples from TCGA. P-values between selected distributions were calculated using Mann–Whitney U test. ***P < 0.001. d The difference in MLK4 expression in breast cancer subtypes with respect to copy number alterations of MLK4 gene. Copy number variation is provided in form of a GISTIC score: −1 (deletion), 0 (diploid), 1 (low-level gain), 2 (high-level amplification). The number over given box represents a size of a population constituting a given distribution. e MLK4 expression in three datasets. Samples were assigned to specific subgroups of breast cancer based on the provided immunohistochemical status of ER, PR, and HER2 receptors. The obtained mRNA levels are shown with respect to different subgroups of breast cancer. P-values between selected distributions were calculated using Mann–Whitney U test. **P < 0.01, ***P < 0.001. f Probability of overall survival in breast cancer patients expressing high or low MLK4 levels assessed using KMplotter, with auto-selected best cutoff. Graphic illustrations taken from Kmplot.com

Fig. 2

MLK4 knock-down decreases anchorage-dependent growth of breast cancer cells. a, b MLK4 protein abundance (a) and mRNA expression levels (b) in a panel of breast cancer cell lines. Densitometry analysis was performed from three independent experiments. α-tubulin was used to normalize the results (a). B2M and RPL29 were used as reference genes to determine relative mRNA expression levels (b). Error bars indicate ± SEM from three independent experiments. c A panel of breast cancer cell lines and MCF10A were transfected with siRNA against MLK4 or control siRNA. After 72 h cells were stained with crystal violet and results were quantified by absorbance (upper graph) or cells were lysed and analyzed by western blotting. Error bars indicate ± SEM from three independent experiments (n = 3). Statistical comparison of values was performed using the unpaired two-tailed t-test. ***P < 0.001, **P < 0.01, *P < 0.05. d, e HCC1806_sh2, HCC1806_sh6, parental HCC1806 (d), and HCC1599_sh2, HCC1599_sh6, parental HCC1599 (e) cells were treated with 1 μg/ml doxycycline for 72 h. Whole cell lysates were analyzed by western blotting. f, g HCC1806_sh2, HCC1806_sh6, parental HCC1806 (f), and HCC1599_sh2, HCC1599_sh6, parental HCC1599 (g) cells were treated with 1 μg/ml doxycycline for 6 days. Viability of cells was determined by MTT assay. Error bars indicate ± SEM from three independent experiments (n = 9). Statistical comparison of values was performed using the unpaired two-tailed t-test. **P < 0.01, *P < 0.05. h HCC1806_sh2, HCC1806_sh6 and parental HCC1806 cells were seeded at low density and were grown with or without 1 μg/ml doxycycline. After 2 weeks, cells were stained with crystal violet (pictures on the left), and quantified by absorbance. Error bars indicate ± SEM from five independent experiments (n = 5). Statistical comparison of values was performed using the unpaired two-tailed t-test. ****P < 0.0001

Fig. 3

MLK4 depletion leads to cell cycle arrest of TNBC cells with high expression of MLK4. a HCC1806_sh2 and HCC1806_sh6 cells were grown with or without doxycycline (1 μg/ml) for 10 days. At day 6, 8 and 10 cells were counted using Bio-Rad Cell Counter TC-20. Error bars indicate ± SEM from three independent experiments (n = 3). Statistical comparison of values was performed using the unpaired two-tailed t-test. *P < 0.05. b Parental HCC1806 cells were transfected with siRNA against MLK4 or control siRNA. After 48 h EdU was added to cells for the next 24 h, then subjected to measurement of EdU incorporation. Error bars indicate ± SEM from at least three independent experiments (n = 18). Statistical comparison of values was performed using one-way ANOVA. **P < 0.01, *P < 0.05. c 72 h following the transfection of parental HCC1806 with siRNA against MLK4 or control siRNA, cells were harvested and fixed in 70% ethanol. Cells then were treated with RNAse A and stained with propidium iodiode. The distribution of cells in phases of cell cycle by flow cytometry was assessed and the ratio of G1/S was calculated. Error bars indicate ± SEM from four independent experiments (n = 4). Statistical comparison of values was performed using one-way ANOVA. **P < 0.01. d HCC1806_sh2, HCC1806_sh6 and parental HCC1806 cells were treated with 1 μg/ml doxycycline for 72 h. Whole-cell lysates were analyzed by western blotting. The treatment of parental HCC1806 cells with 5,8 nM paclitaxel for 24 h served as a positive control for apoptosis evidenced with PARP and pro-caspase-3 cleavage. e, f Reduction in proliferation after MLK4 knock-down is rescued by overexpression of MLK4-WT. e HCC1806_sh2 and HCC1806_sh6 were treated with 1 μg/ml doxycycline. Next day cells were transiently transfected with MLK4-WT vector or EV (empty vector) control and were left to grow with or without doxycycline. After 72 h cells were stained with crystal violet and results were quantified by absorbance. Error bars indicate ± SEM from three independent experiments (n = 3). Statistical comparison of values was performed using the two-way ANOVA. *P < 0.05. Representative immunoblots showing the level of MLK4-WT in samples lysed after 72 h of transfection are shown below. f HCC1806_sh6 with stable overexpression of MLK4-WT and HCC1806_sh6 control vector cells were seeded at low density and were grown with or without 1 μg/ml doxycycline. After 10 days, cells were stained with crystal violet, and quantified by absorbance. Error bars indicate ± SEM from four independent experiments (n = 4). Statistical comparison of values was performed using the two-way ANOVA. **P < 0.01. Representative immunoblots showing the level of MLK4 in samples lysed after 96 h doxycycline treatment are shown on the right

Fig. 4

Depletion of MLK4 attenuates migratory and invasive properties of cells. a HCC1806_sh2 and HCC1806_sh6 cells were treated with 1 μg/ml doxycycline for at least 72 h, then cells were serum-starved and seeded on transwell inserts. The cells were allowed to migrate for no longer than 24 h. Cells at the bottom of inserts were stained with crystal violet. Five pictures of every condition were taken. Analysis was performed using ImageJ. Error bars indicate ± SEM from three independent experiments (n = 15). Statistical analysis was done using unpaired two-tailed t-test. ***P < 0.001, *P < 0.05. Below, representative immunoblots showing the level of MLK4 knock-down. b, c HCC1806_sh2 and HCC1806_sh6 cells were treated with 1 μg/ml doxycycline for at least 72 h and then were subjected to wound-healing assay. Four pictures were taken at each condition. Representative pictures and immunoblots showing the level of MLK4 knock-down are shown (b). Quantification was performed using ImageJ (c). Error bars indicate ± SEM from four independent experiments (n = 12). Statistical comparison of values was performed using the unpaired two-tailed t-test. *P < 0.05. d, e HCC1806_sh6 cells were treated with 1 μg/ml doxycycline for 72 h. Next, single cell migration was monitored for 20 h. Analysis was performed using ImageJ (d). Error bars indicate ± SEM (n = 28). Statistical comparison of values was performed using the unpaired two-tailed t-test. ***P < 0.001. Representative immunoblots showing the level of MLK4 knock-down are shown on the right. Representative tracks of five randomly chosen cells (e). The intersection of the x- and y-axes was taken as the starting point of each cell path to which other points of tracks were normalized. f MDA-MB-436 cells were transfected with siRNA against MLK4 or control siRNA. After 48 h, cells were serum-starved and seeded on transwell inserts. The cells were allowed to migrate for no longer than 24 h. Cells at the bottom of inserts were stained with crystal violet. Five pictures of every condition were taken. Analysis was performed using ImageJ. Error bars indicate ± SEM from three independent experiments (n = 15). Statistical analysis was done using one-way ANOVA. **P < 0.01, *P < 0.05. Below, representative immunoblots showing the level of MLK4 knock-down. g MCF10A cells were transfected with EV (empty vector) control, MLK4-WT, MLK4-KA (kinase active) and MLK4-KD (kinase dead). After 48 h, cells were serum-starved and seeded on transwell inserts. The cells were allowed to migrate for no longer than 24 h. Cells at the bottom of inserts were stained with crystal violet. Five pictures of every condition were taken. Analysis was performed using ImageJ. Error bars indicate ± SEM from three independent experiments (n = 15). Statistical analysis was done using one-way ANOVA. **P < 0.01, ***P < 0.001. Below, representative immunoblots showing the level of MLK4 overexpression. h HCC1806_sh2 and HCC1806_sh6 cells were treated with 1 μg/ml doxycycline for at least 72 h, then cells were serum-starved and seeded on transwell inserts coated with Matrigel. The cells were allowed to invade for no longer than 24 h. Cells were stained with crystal violet. Five pictures of every condition were taken. Analysis was performed using ImageJ. Error bars indicate ± SEM from three independent experiments (n = 15). Statistical analysis was done using unpaired two-tailed t-test. ****P < 0.0001. Below, representative immunoblots showing the level of MLK4 knock-down

Fig. 5

MLK4 knock-down impairs spheroid formation in 3D model in vitro and tumor growth in vivo. a HCC1806_sh2 and HCC1806_sh6 cells were seeded on plates coated with Matrigel. Cells were grown for 10 days with or without 1 μg/ml doxycycline. Pictures were taken and the number and size of spheroids were quantified in ImageJ. Error bars indicate ± SEM from three independent experiments. Statistical comparison of values was performed using the unpaired two-tailed t-test. ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. b HCC1806_sh2 and HCC1806_sh6 cells were seeded on low cell attachment surface plates. Cells were grown with or without 1 μg/ml doxycycline. Pictures were taken on day 10 and size of spheroids was quantified in ImageJ. Error bars indicate ± SEM from three independent experiments. Statistical comparison of values was performed using the unpaired two-tailed t-test. *P < 0.05. c HCC1806_sh6 cells were injected into mammary fat pads of RAG2^−/− mice. Doxycycline induction of MLK4 knock-down started one day after the injection. Tumors were measured twice a week. Error bars indicate ± SEM (n = 12 for control group and n = 13 for doxycycline-treated group). Statistical comparison of values was performed using the unpaired two-tailed t-test. **P < 0.01, *P < 0.05

Fig. 6

MLK4 activates NF-κB pathway and promotes mesenchymal phenotype of breast cancer cells. a, b HCC1806_sh2 and HCC1806_sh6 cells were treated with 1 μg/ml doxycycline. Whole cell lysates were analyzed by western blotting (a). Quantification analysis of immunoblots was performed using ImageJ (b). Error bars indicate ± SEM from four independent experiments (n = 4). Statistical comparison of values was performed using the unpaired two-tailed t-test. **P < 0.01, *P < 0.05. c HCC1806_sh6 cells were treated with 1 μg/ml doxycycline for 72 h, and then were stimulated with 20 ng/ml TNF-α for 1 h, lysed and subjected to TransAM DNA binding assay. Error bars indicate ± SEM from at least three independent experiments (n = 3). Statistical comparison of values was performed using the unpaired two-tailed t-test. *P < 0.05. d, e MDA-MB-436 cells were transfected with siRNA against MLK4 or control siRNA. After 72 h whole-cell lysates were analyzed by western blotting (d). Quantification analysis of immunoblots was performed using ImageJ (e). Error bars indicate ± SEM from three independent experiments (n = 3). Statistical comparison of values was performed using one-way ANOVA. **P < 0.01, *P < 0.05. f, g MCF10A cells were transfected with EV (empty vector) control, MLK4-WT, MLK4-KA (kinase active) and MLK4-KD (kinase dead). After 48 h whole-cell lysates were analyzed by western blotting (f). Quantification analysis of immunoblots was performed using ImageJ (g). Error bars indicate ± SEM from three independent experiments (n = 3). Statistical comparison of values was performed using one-way ANOVA. **P < 0.01, *P < 0.05. h, i MCF10A cells were transfected with EV (empty vector) control or MLK4-WT. After 8 h medium was changed and cells were treated with 10 µM BAY-11-7082 or DMSO for 48 h. Whole cell lysates were analyzed by western blotting (h). Quantification analysis of immunoblots was performed using ImageJ (i). Error bars indicate ± SEM from three independent experiments (n = 3). Statistical comparison of values was performed using two-way ANOVA. **P < 0.01, *P < 0.05. j MCF10A cells were transfected with EV (empty vector) control or MLK4-WT. After 48 h cells were seeded on transwell inserts. When cells attached to the bottom of inserts, the medium was changed and cells were treated with 10 µM BAY-11-7082 or DMSO overnight. Cells that migrated through the inserts were stained with crystal violet. Five pictures of every condition were taken. Analysis was performed from three independent experiments using ImageJ (n = 15). Statistical analysis was done using two-way ANOVA. ****P < 0.0001

Fig. 7

High MLK4 expression is associated with lymph nodes metastasis in TNBC patients. a Immunohistochemical staining for MLK4 was performed in paraffin-embedded blocks of tumor tissues of 129 breast cancer samples (TNBC – 46, HER2-positive – 26, luminal A – 26, luminal B – 31). The staining was evaluated using the H-score system. Information on occurrence of lymph node metastasis was collected from pathologic reports. The intensity of immunohistochemical staining expressed by H-score was calculated using Mann–Whitney U test. *P < 0.05. b NST (no specific type) of triple-negative carcinoma showing lymphovascular invasion (arrows) in hematoxylin eosin staining and strong MLK4 staining in the intravascular component (2, 3) comparing to the main bulk of the tumor (1). Scale bar 100 µm (×10 magnification) and 20 µm (×40 magnification)