MEK1 Inhibitor Combined with Irradiation Reduces Migration of Breast Cancer Cells Including miR-221 and ZEB1 EMT Marker Expression

Simple Summary Combined chemotherapy and radiotherapy are an effective treatment for invasive breast cancer. However, some studies suggest that such interventions may increase the risk of metastasis. Cell metastatic behavior is highly dependent on RAS-RAF-MEK pathway and its downstream target activation, including miR-221 overexpression and epithelial-to-mesenchymal transition (EMT). By using MEK1 inhibitor (TAK-733) in combination with radiation therapy for breast cancer cells, significant decrease in migration capacity, including reduction of miR-221 and EMT (ZEB1) marker expression was observed. miR-221 holds great potential as therapeutic biomarker and target for new drug developments, however more insight into efficiency of miR-221 inhibition needs to be followed in the future. Abstract The miR-221 expression is dependent on the oncogenic RAS-RAF-MEK pathway activation and influences epithelial-to-mesenchymal transition (EMT). The Cancer Genome Atlas (TCGA) database analysis showed high gene significance for ZEB1 with EMT module analysis and miR-221 overexpression within the triple-negative breast cancer (TNBC) and HER2+ subgroups when compared to luminal A/B subgroups. EMT marker expression analysis after MEK1 (TAK-733) inhibitor treatment and irradiation was combined with miR-221 and ZEB1 expression analysis. The interaction of miR-221 overexpression with irradiation and its influence on migration, proliferation, colony formation and subsequent EMT target activation were investigated. The results revealed that MEK1 inhibitor treatment combined with irradiation could decrease the migratory potential of breast cancer cells including reduction of miR-221 and corresponding downstream ZEB1 (EMT) marker expression. The clonogenic survival assays revealed that miR-221 overexpressing SKBR3 cells were more radioresistant when compared to the control. Remarkably, the effect of miR-221 overexpression on migration in highly proliferative and highly HER2-positive SKBR3 cells remained constant even upon 8 Gy irradiation. Further, in naturally miR-221-overexpressing MDA-MB-231 cells, the proliferation and migration significantly decrease after miR-221 knockdown. This leads to the assumption that radiation alone is not reducing migration capacity of miR-221-overexpressing cells and that additional factors play an important role in this context. The miR-221/ZEB1 activity is efficiently targeted upon MEK1 inhibitor (TAK-733) treatment and when combined with irradiation treatment, significant reduction in migration of breast cancer cells was shown.


Introduction
Adjuvant radiation therapy for breast cancer (stage I-III) improves disease-free survival and overall survival [1]. Combination of chemotherapy with radiation treatment offers a cost-effective first-line treatment that prolongs 15-year survival for breast cancer patients [1]. Triple-negative breast cancer (TNBC) and human epidermal growth factor receptor-2 (HER2) positive tumours, both have a particularly poor prognosis, with higher rates of recurrence when compared to hormone ER (estrogen receptor) and PR (progesterone receptor) positive cases [2]. Recurrent tumours are often resistant to further radio-chemo-therapies (RCT) and may show increased migratory (metastatic) potential [3]. Consequently, there is no standard therapy for the breast cancer patients with recurrent or metastatic (advanced) breast cancer and it remains a great challenge to select an appropriate treatment [4,5].
In order to find a successful therapeutic approach for advanced breast cancers, the focus should be on the mechanisms involved in cell invasion and metastasis, including regulation of the epithelial-to-mesenchymal transition (EMT). One of the signaling pathways involved in tumor progression and EMT regulation is the RAS-RAF-MEK-MAPK pathway [6,7]. These kinases are overexpressed and hyperactive in various types of cancer. They regulate diverse cellular processes such as proliferation, migration, metastasis, resistance to chemotherapy, and EMT [8]. Furthermore, the concept that non-coding RNAs including miRNAs can promote transformation to more aggressive cancer phenotypes with poor prognosis has already been documented [9].
We have previously shown that the expression of miR-221 was associated with higher risk of metastasis in advanced breast cancers [10] and miR-221 was considered to be a prognostic marker for distinguishing subgroups particularly in advanced (lymph node-positive and HER2-positive) breast cancers [10]. Additionally, miR-221 overexpression may promote the EMT transition by targeting Notch3 [11] and via targeting the 3 UTR of GATA transcriptional repressor (TRPS1), leading to increased cell migration and invasion [12,13] in breast cancer cells. Moreover, it has been reported that miR-221 negatively regulates ER [14]. Increased resistance to tamoxifen, a drug successfully used to treat women with ER-positive breast cancer, was associated with ER and p27 inhibition upon miR-221 overexpression [15,16]. Furthermore, it has been shown that miR-221 confers radioresistance in glioblastoma cells [17]. We now report that miR-221 overexpression is associated with greater cellular migration capacity of breast cancer cells upon radiation treatment. As miR-221 acts downstream of the oncogenic RAS-RAF-MEK pathway, MEK1 inhibitor (TAK-733) was combined with irradiation treatment for subsequent miR-221 functional and phenotypic analysis including cell migration and proliferation activity.

Differential miR-221 Expression in Breast Cancer Subgroups and Cell Lines
The miR-221 and HER2 boxplots were generated using GEPIA interactive web server [18] from publicly available RNA sequencing expression databases including the TCGA and the GTEx projects ( Figure 1a). The in silico analysis showed significant difference in miR-221 expression in comparison to normal adjacent tissue samples (Figure 1a, left) in agreement with previously published in house analysis [10]. Within TCGA database 512 samples were analyzed for miR-221 expression including different breast cancer subgroups: HER2+, Luminal A, Luminal B and TNBC (Figure 1a

High miR-221 Expression Levels in Breast Cancer Cells Correlate with Low Attachment, Low Adhesion and High Migration Activity
Spheroid formation of five different breast cancer cell lines ( Figure 1d) showed an association with miR-221 expression levels using an established 3D-microtissue platform [19]. The 100-fold higher miR-221 expression in MCF7 and SKBR3 cells (when compared to T47D), as well as 1000-fold higher miR-221 expression (Figure 1c) detected in MDA-MB-231 cells correlated with reduced capacity to form spheroids. Wound healing assays (Figure 1d) showed that high migration capacity associates with high miR-221 expression levels and the highest migratory activity was seen for MDA-MB-231 cells that express the highest levels of miR-221. In conclusion, the analysis of different breast cancer cell lines showed that high miR-221 expression levels (including TNBC and HER2 + cell lines) strongly correlate with breast cancer cell migration capacity and inhibit self-aggregation properties to form spheroids ( Figure 1).

Breast Cancer 3D-Spheroid Analysis Using MEK1 Inhibitor (TAK-733) in Combination with Irradiation
It was previously shown that miR-221 expression is dependent on the oncogenic RAS-RAF-MEK pathway activity [13]. Therefore, the MEK1 inhibitor (TAK-733) was analyzed for synergistic effects with irradiation using our lately developed 3D microtissue assay platform [19] with T47D cells ( These data were supported using living (Hoechst 33342) and dead (DRAQ7) cell stainings (Figure 1b), showing increased fraction of dead (red) cells upon combined TAK-733 and 4 Gy irradiation treatment. TAK-733 (MEKi) or 4 Gy irradiation alone resulted in detection of more dead cells compared to sham irradiated (DMSO treated) controls and lower numbers of living ("blue") cells were identified. Spheroids with disseminated borders were detected 9 days upon indicated treatments ( Figure 1b). Additionally, the influence of TAK-733 MEK1 inhibitor on miR-221 expression was analyzed from spheroids (three days upon treatment) with 1 µM and 10 µM TAK-733 concentration combined with 4 Gy irradiation. TAK-733 (10 µM) significantly inhibited miR-221 cells compared to sham irradiated (DMSO treated) controls and lower numbers of living ("blue") cells were identified. Spheroids with disseminated borders were detected 9 days upon indicated treatments ( Figure 1b). Additionally, the influence of TAK-733 MEK1 inhibitor on miR-221 expression was analyzed from spheroids (three days upon treatment) with 1 µM and 10 µM TAK-733 concentration combined with 4 Gy irradiation. TAK-733 (10 µM) significantly inhibited miR-221 expression when used in combination with 4 Gy irradiation ( Figure 1c). These data revealed that MEK1 inhibitor treatment combined with irradiation could significantly decrease the growth of 3D breast cancer microtissues including reduction in miR-221 expression.

MEK1 Inhibitor (TAK-733) Combined with Irradiation Inhibits ZEB1 (EMT Marker) Activity
In order to find the genes which were highly associated with EMT hallmarks and potentially influence the mode of cell migration, patient data sets were analyzed for ZEB1, Vimentin and uPAR expression (Figure 4a).

Modulation of miR-221 Expression Alters Migration Capacity
According to different levels of miR-221 expression in breast cancer cell lines (Figure 1c), SKBR3 cells were used to overexpress miR-221 and MDA-MB-231 cells (1000-fold naturally overexpressing miR-221) were used for knock-down experiments by lentiviral approach (Figures 5 and 6). Initially, the migration capacity of SKBR3 cells was analyzed 24 h following irradiation ( Figure S3). Compared to sham irradiated controls (0 Gy, Figure S3a) higher radiation doses (4 Gy and 8 Gy) slightly induced the migratory phenotype of SKBR3 cells 24 h upon treatment. This was in line with apoptosis induction and increased caspase 3/7 enzyme activity detection after irradiation treatment, whereas cell viability was not dramatically decreased 24 h upon irradiation ( Figure S3). To confirm that migration activity is linked with higher miR-221 expression, miR-221-overexpressing SKBR3 cells  The ZEB1 expression showed the highest gene significance with EMT module analysis (r = 0.067; p < 0.0001) from 123 TNBC patients within TCGA cohort. ZEB1 transcription factor expression (a central EMT regulator) was shown to be significantly induced after irradiation alone (Figure 4b,c and Figure S2) which supports radiation resistance for MDA-MB-231 cells, providing stable migration activity, seen in the controls (DMSO) upon 0 Gy and 4 Gy irradiation (Figure 3c,d). Finally, the results revealed that MEK1 inhibitor (when combined with radiation) significantly reduces the migratory potential of MDA-MB-231 breast cancer cells including reduction of miR-221 expression ( Figure 3a) and corresponding downstream ZEB1 and uPAR (EMT) marker expression (Figure 4b-d).

Modulation of miR-221 Expression Alters Migration Capacity
According to different levels of miR-221 expression in breast cancer cell lines (Figure 1c), SKBR3 cells were used to overexpress miR-221 and MDA-MB-231 cells (1000-fold naturally overexpressing miR-221) were used for knock-down experiments by lentiviral approach (Figures 5 and 6). Initially, the migration capacity of SKBR3 cells was analyzed 24 h following irradiation ( Figure S3). Compared to sham irradiated controls (0 Gy, Figure S3a) higher radiation doses (4 Gy and 8 Gy) slightly induced the migratory phenotype of SKBR3 cells 24 h upon treatment. This was in line with apoptosis induction and increased caspase 3/7 enzyme activity detection after irradiation treatment, whereas cell viability was not dramatically decreased 24 h upon irradiation ( Figure S3). To confirm that migration activity is linked with higher miR-221 expression, miR-221-overexpressing SKBR3 cells were analyzed in parallel with control cells (containing empty virus (EV) control, Figure 5). miR-221-overexpressing SKBR3 cells showed (10 fold and 100 fold) higher miR-221 expression levels when compared to non-irradiated (0 Gy) empty vector (EV) control cells, 24 h (** p < 0.0100) to 72 h after treatment with 2 Gy, 4 Gy and 8 Gy (*** p < 0.0001, ANOVA between SKBR3 (EV) and +miR-221 overall, Figure 5a,b). Additionally, miR-221 expression levels were not changing dramatically 24 h and 72 h upon 4 Gy (and 8 Gy) irradiation (Figure 5a,b), showing that irradiation alone is not significantly changing expression of miR-221. Slight miR-221 reduction was detected upon 8 Gy irradiation (Figure 5a), but the levels of miR-221 were still significantly higher in miR-221-overexpressing SKBR3 cells when compared to corresponding control (EV) cells 24 h and 72 h after irradiation treatment (Figure 5a,b). Higher miR-221 expression induces higher proliferation capacity of SKBR3 miR-221-overexpressing cells (Figure 5c) 72 h after seeding for the relative proliferation index analysis.    Colony formation assays confirmed an increased radiation resistance of SKBR3 cells overexpressing miR-221 ten days upon irradiation (*** p < 0.0030, ANOVA between SKBR3 (EV) and +miR-221 overall, Figure 5d). Additionally, 24 h post-irradiation, no effects on proliferation were detected either in EV control cells nor in miR-221-overexpressing SKBR3 cells (Figure 5e), whereas significant reduction in SKBR3 cell proliferation was detected 72 h upon 8 Gy irradiation independent of miR-221 overexpression (Figure 5f). Correspondingly, significant upregulation in migration capacity was detected for miR-221-overexpressing SKBR3 cells, 24 h after irradiation treatment (** p < 0.0106, Figure 5g), that retained to be notably higher even 72 h upon 2 Gy, 4 Gy and 8 Gy irradiation (*** p < 0.0003, ANOVA between SKBR3 (EV) and +miR-221 overall, Figure 5h). These data confirm high migration activity of miR-221-overexpression in SKBR3 cells upon irradiation treatment.
Additionally, MDA-MB-231 cells (naturally expressing 1000-fold high levels of miR-221) were used for anti-miR-221 (knock-down) experiments including radiation treatment ( Figure 6). Reduction in miR-221 expression (Figure 6a), when compared to empty vector (EV), confirmed reduced proliferation capacity for anti-miR-221 MDA-MB-231 cells (Figure 6b), 72 h after seeding for relative proliferation index analysis. As the knockdown effect was lowering miR-221 levels (Figure 6a), significant radiosensitization effects of anti-miR-221 treatment were not detected in colony formation assays (Figure 6c), as probably much higher reduction of miR-221 expression level is required. Significant reduction in MDA-MB-231 cell proliferation was detected 24 h to 48 h upon 8 Gy irradiation independent of miR-221 knock-down (Figure 6d,e), including considerably reduced migration of MDA-MB-231 cells by anti-miR-221, when compared to corresponding (EV) controls upon different radiation doses (2 Gy, 4 Gy and 8 Gy, *** p < 0.0001, ANOVA 24 and 48 h between MDA-MB-231 (EV) and +miR-221 overall, Figure 6f,g). The data observed in both breast cancer cell lines suggested that higher miR-221 expression levels significantly influence cellular migration capacity and EMT turnover. Moreover, miR-221 expression levels were very stable upon radiation treatment.

Discussion
Taken together, the results presented here showed that novel therapy design including MEK1 inhibitor (TAK-733) can be seriously considered in the future for the heavily pre-treated advanced breast cancers (e.g., TNBC and HER2+), while efficiencies may depend on different drug combinations and respective time and dose points. TAK-733 (as 1 µM) showed significant effects on miR-221/ZEB1 reduction and breast cancer cell proliferation and migration capacity (Figures 3 and 4). In the future the clear and precise low concentration selection of substances without toxic side effects in combination with radiation therapy would be of high importance due to the benefits for the patients [4,20].
Compounds such as the inhibitors of mitogen-activated protein kinase kinase (MAPKK or MEK1) demonstrate promising effects on cancer cells [21,22]. MEK1 as part of the RAF/MEK/ERK signaling pathway is documented as important transmitter of extracellular signals from receptors on plasma membrane to the cell nucleus (Figure 7). An effective inhibitor for this signaling pathway is the selective, ATP-non-competitive MEK1 inhibitor represented here as TAK-733. The small molecule inhibitor binds to the allosteric site of MEK1 and shows antitumor activity in many mouse xenograft models, including a model for breast cancer [23]. Despite the fact, that the TAK-733 homologue Trametinib is currently FDA-approved for treatment of melanomas and is further in frequent use regarding phase 1 and 2 clinical trials (www.cancer.gov), the mechanisms behind the synergistic effects with radiation treatment still remain to be fully elucidated.
MEK1 inhibitor (TAK-733) in combination with radiation may be an effective and advantageous treatment, especially for differentially miR-221-overexpressing cells (e.g., TNBC-very high and HER2 + high miR-221 expressing cells) that are recognized as highly migrative and resistant to radiation treatment alone. Overexpression of miR-221 can be effectively inhibited using antisense oligonucleotides. However, several challenges regarding the stability and delivery strategies of the anti-miRs still need to be answered before they can be useful as therapeutics. In parallel, novel drugs known to prevent cardiovascular disease by lowering the levels of miR-221 are highly appreciated for future cancer treatments. With our work presented here, we have demonstrated that MEK1 inhibitor (TAK-733) significantly reduces miR-221 expression. By lowering miR-221 expression, migrative and proliferative features of breast cancer cells were inhibited and prevention of cardiotoxicity as a major side effect of breast cancer radiation therapy [35] is conceivable.

Omics Data Analysis Using Web Databases with Clinical Follow-Up
HER2 and miR-221 expression markers were generated by GEPIA web server during 2019 and 2020 [18] analyzing the RNA sequencing expression data from 1085 breast tumors (BRCA) and 291 normal samples from the TCGA and the GTEx projects (Figure 1a). Additionally, RNA-Seq, stemloop miRNA-seq and clinic datasets from patients with BRCA (breast cancer) were downloaded from the UCSC Xena browser (https://xenabrowser.net/). Box plots of miR-221 expression levels ( Figure  1a) were created using ggplot2 (https://xenabrowser.net/) by analyzing 515 samples in total and including HER2+ (52 samples), Luminal A (223 samples), Luminal B (121 samples) and TNBC (119 samples). The data for EMT and the target gene (ZEB1, Vimentin and uPAR) analysis (Figure 4a) were retrieved from 123 TNBC patients within TCGA cohort in 2019 and 2020. The correlation was calculated with Pearson's coefficient. The EMT score was obtained with the single-sample GSEA method analysis performed with "GSVA" package in R software (version: 3.63) [36]. The RAS-RAF-MEK-MAPK pathway is altered in~40% of all human cancers, mainly due to mutations in BRAF (~10%) and its upstream activator RAS (~30%) [24,25]. Despite initial high response rates with targeted therapies against RAF and MEK, these therapies are limited due to the emergence of drug resistance and, over the last years, a number of acquired resistance mechanisms were discovered (reviewed in [26,27]). In recent years, a number of studies suggested that drugs targeting epigenetic alterations could be applied in synergy with other anticancer therapies or, importantly, in reversing acquired therapy resistance (reviewed in [25,[28][29][30]). Therefore, combination therapies not only result in higher overall response rates and prolonged time to disease progression when compared with single agents, but could also offer an option for women with anthracycline-and taxane-pre-treated metastatic breast cancer [4,31]. Cardiotoxicity is a well-known side effect of anthracyclines (e.g., Doxorubicine) and the mechanisms leading to this phenomenon include high expression of miR-221 [32,33].

TAK-733 reduces breast cancer cell growth and miR-221 expression after irradiation
It is known that through an ERK1/2 phosphorylation MEK1 is able to activate many transcription factors and processes such as cell growth, migration and epithelial-to-mesenchymal transition (EMT) [34] as indicated in Figure 7. The miR-221 expression is dependent on the oncogenic RAS-RAF-MEK pathway activation ( Figure 7) and potentially influences different cellular processes including EMT [12,13]. In parallel by regulating the level of miR-221 expression after combined MEK1 and 4 Gy irradiation, additional downstream targets may be affected (Figure 7), as e.g., TRPS1 previously shown to regulate ZEB1/2 transcription factor activity [12]. Data presented here highly suggest that MEK1 inhibitor (TAK-733) in combination with radiation may be an effective and advantageous treatment, especially for differentially miR-221-overexpressing cells (e.g., TNBC-very high and HER2 + high miR-221 expressing cells) that are recognized as highly migrative and resistant to radiation treatment alone.
Overexpression of miR-221 can be effectively inhibited using antisense oligonucleotides. However, several challenges regarding the stability and delivery strategies of the anti-miRs still need to be answered before they can be useful as therapeutics. In parallel, novel drugs known to prevent cardiovascular disease by lowering the levels of miR-221 are highly appreciated for future cancer treatments. With our work presented here, we have demonstrated that MEK1 inhibitor (TAK-733) significantly reduces miR-221 expression. By lowering miR-221 expression, migrative and proliferative features of breast cancer cells were inhibited and prevention of cardiotoxicity as a major side effect of breast cancer radiation therapy [35] is conceivable.

Omics Data Analysis Using Web Databases with Clinical Follow-Up
HER2 and miR-221 expression markers were generated by GEPIA web server during 2019 and 2020 [18] analyzing the RNA sequencing expression data from 1085 breast tumors (BRCA) and 291 normal samples from the TCGA and the GTEx projects (Figure 1a). Additionally, RNA-Seq, stem-loop miRNA-seq and clinic datasets from patients with BRCA (breast cancer) were downloaded from the UCSC Xena browser (https://xenabrowser.net/). Box plots of miR-221 expression levels ( Figure 1a)  . The virus particles were harvested 48 h after transfection of 293T cells cultivated in water humified 37 • C incubator with 5% CO 2 , cleared and concentrated as previously described [37][38][39]. According to virus titer determination virus productions ranged between 10 8 to 10 9 TU/mL and viral infection of breast cancer cells was performed using previously described protocols [37][38][39][40][41]. Briefly, 2 × 10 5 cells per well were infected with 4 × 10 5 TU/mL (2 MOI) for 16-24 h in water humified 37 • C incubator with 5% CO 2 . After infection the cells were washed of virus (2xPBS) and cultivated further in corresponding medium. Three days after infection, GFP expression was monitored and cells were seeded in petri dishes for subsequent analysis. Additionally, GFP expressing cells were seeded in 96-well GravityTRAP ULA plates (InSphero AG, Schlieren, Switzerland and Perkin Elmer, Waltham, MA, USA) for 3D microtissue assay analysis or in 12-well Ibidi chambers (Ibidi, Munich, Germany) for wound healing (migration) assay analysis and treated as described below.

Generation of Tumour 3D-Microtissues and Treatment with Compounds and Irradiation
3D microtissues were generated using 500 cells (per well) showing adequate growth kinetics and low interwell variations (bellow 10%). Cell density in media was estimated using a hemocytometer (Z1 Coulter, Beckman, Brea, CA, USA). 3D-microtissues were formed by seeding T47D, MDA-MB-361, MCF7, SKBR3 and MDA-MB-231 cells into the GravityTRAP plate and maturing them for 3 days followed by treatment with 1µM or 10µM compound (TAK-733) where necessary. Compound concentrations were made that 1% of the total volume per well consists of compound-DMSO dilutions. After one day of substance treatment (defined as day 1 of treatment), tissues were sham irradiated (0 Gy) or irradiated with a single acute dose of 4 Gy with a X-Strahl RS225 (X-Strahl LTD, Camberlay, UK) with delivered dose rate of 0.82 Gy/min and including a 3mm aluminum filter. The exposed and sham irradiated 3D-microtissues were subsequently incubated at 37 • C with 5% CO 2 for indicated time points. The experiment was repeated for each dose in quadruplicates and in three independent experiments (n = 3).
Growth of 3D-microtissues was followed in assay plates for 16 days or labelled with 1µM DRAQ7 (red) or 0.1 µM Hoechst 3322 (blue) for 20min at room temperature and then further processed using Operetta ® High Content Imaging System (Perkin Elmer, Waltham, MA, USA) device. Images from a single plate were acquired (when necessary) with the GFP (green), Hoechst 3322 (blue), DRAQ7 (red) and Brightfield channels using the 10 × NA objective in wide field mode. Direct quantification of 3D-microtissue fluorescent area using a high imaging platform accelerates captures the full range of microtissue phenotypes during analysis using Harmony ® 3.1 High Content Imaging and Analysis Software (Perkin Elmer, Waltham, MA, USA). The Find Image Region Building Block was applied to the GF (green), red or Hoechst channel to detect the microtissues in the well. As a next step, the Calculate Morphology Building Block was added to calculate the tissue area (µm 2 ) as the final readout for green channel. For statistical analysis the Student's t-test was used.

RNA Isolation for miRNA Expression Analysis
Total RNA was isolated from each of the breast cancer cell lines or after viral transductions (including SKBR3 and MDA-MB-231 cells) and upon TAK-733 and irradiation treatment when indicated. The cells were pelleted by centrifugation at 1500 rpm for 5 min and washed with 1 mL Dulbecco's phosphate-buffered saline (PBS) without MgCl 2 and CaCl 2 (Invitrogen, Thermo Fisher, Waltham, MA, USA). Small RNAs (<200 nucleotides) were isolated from the cells using the miRNA Tissue isolation kit and Maxwell16 device (Promega, Madison, WI, USA) following the protocol for total RNA isolation. The quantity and quality of the total RNA including miRNA fraction was measured with the Nanodrop spectrophotometer (PeqLab Biotechnology, Erlangen, Germany).

TaqMan-miRNA Assays and Data Analyses
Specific single TaqMan-miRNA assays (Applied Biosystems, Thermo Fisher, Waltham, MA, USA) were used for miR-221 expression analysis (Cat.Nr. 4427975; Assay ID 000524) in total RNA isolations from cell culture pellets according to previously described protocols. Quantitative RT-PCR was performed using the StepOnePlus Detection System (Thermo Fischer, Waltham, MA, USA). The relative expression values of specific miR-221 were calculated by using the 2 -∆∆CT method [38] and normalized to the control miRNA (RNU44, # 4427975; Assay ID 001095) [10] and relative to the T47D or SKBR3 (EV) or MDA-MB-231 (EV) as control cells (used as calibrator) alone or after subsequent lentiviral transductions or compound and radiation treatment. All reactions were performed in duplicates and three independent experiments (n = 3) were conducted.

Western Blot Analysis
The isolation of proteins, immunoblotting and quantifications were performed as previously described [10,42]. The respective target proteins regulated by the miRNAs were detected with

Statistics
For statistical analyses of the in vitro proliferation, caspase assays, 3D-microtissue and migration assays, as well as for Western blots, Student's t-test was used. Data show the mean of independent biological experiments with the standard deviation (±SD). The two-sided paired, unpaired or the one-sample t-test were used for statistical analysis and a p-value ≤ 0.05 was deemed statistically significant, while a p-value < 0.01 was considered highly significant. Additionally, migration assays and western blots were confirmed with two-way ANOVA analysis.
Since the ANOVA only provides information about significant differences in general, for example a general effect of irradiation or compound treatment, post hoc testing was performed to compare individual treatments. Bonferroni, Sidak or Dunnett multiple comparisons were done with GraphPad Prism5 to identify the individual significant differences. To analyze significant effects on specific marker proteins by the inhibitors alone irradiated compound-treated cells were compared to their irradiated control (4 Gy, DMSO) and the sham-irradiated compound-treated samples were compared to their sham-irradiated controls (DMSO). For the analysis of interaction effects between irradiation and inhibitor treatment 0 Gy treated samples were compared to the 4 Gy treated samples. A p-value < 0.05 was deemed statistically significant and p-values < 0.01 were considered to be highly significant.

Conclusions
The results presented here confirm that MEK1 inhibitor treatment combined with irradiation could significantly decrease the migratory potential of breast cancer cells including reduction in miR-221 and EMT (ZEB1) marker expression changes. By using MEK1 inhibitor in combination with radiation therapy for advanced breast cancer, the miR-221 holds great potential as therapeutic marker of metastatic breast cancer and as promising target for new drug developments. However, more insight into efficiency of miR-221 inhibition needs to be followed in the future.