Activated ERK Signaling Is One of the Major Hub Signals Related to the Acquisition of Radiotherapy-Resistant MDA-MB-231 Breast Cancer Cells

Breast cancer is one of the major causes of deaths due to cancer, especially in women. The crucial barrier for breast cancer treatment is resistance to radiation therapy, one of the important local regional therapies. We previously established and characterized radio-resistant MDA-MB-231 breast cancer cells (RT-R-MDA-MB-231 cells) that harbor a high expression of cancer stem cells (CSCs) and the EMT phenotype. In this study, we performed antibody array analysis to identify the hub signaling mechanism for the radiation resistance of RT-R-MDA-MB-231 cells by comparing parental MDA-MB-231 (p-MDA-MB-231) and RT-R-MDA-MB-231 cells. Antibody array analysis unveiled that the MAPK1 protein was the most upregulated protein in RT-R-MDA-MB-231 cells compared to in p-MDA-MB-231 cells. The pathway enrichment analysis also revealed the presence of MAPK1 in almost all enriched pathways. Thus, we used an MEK/ERK inhibitor, PD98059, to block the MEK/ERK pathway and to identify the role of MAPK1 in the radio-resistance of RT-R-MDA-MB-231 cells. MEK/ERK inhibition induced cell death in both p-MDA-MB-231 and RT-R-MDA-MB-231 cells, but the death mechanism for each cell was different; p-MDA-MB-231 cells underwent apoptosis, showing cell shrinkage and PARP-1 cleavage, while RT-R-MDA-MB-231 cells underwent necroptosis, showing mitochondrial dissipation, nuclear swelling, and an increase in the expressions of CypA and AIF. In addition, MEK/ERK inhibition reversed the radio-resistance of RT-R-MDA-MB-231 cells and suppressed the increased expression of CSC markers (CD44 and OCT3/4) and the EMT phenotype (β-catenin and N-cadherin/E-cadherin). Taken together, this study suggests that activated ERK signaling is one of the major hub signals related to the radio-resistance of MDA-MB-231 breast cancer cells.


Introduction
Breast cancer is one of the major causes of death due to cancer worldwide, especially in women [1]. For breast cancer, many therapies are available such as surgical resection, with respect to this value, we selected around 10 upregulated proteins and 16 downregulated proteins, which are specified in Figure 1A. The highly expressed proteins included mitogen-activated protein kinase 1 (MAPK1), which exhibited about a 2.81-fold increase compared to the p-MDA-MB-231 cells. Next to MAPK1, the highly expressed protein was dual-specificity protein kinase CDC-like kinases (CLK1), which exhibited a 1.32-fold increase.
Among the 16 downregulated proteins, caspase 3 was the most downregulated in the RT-R-MDA-MB-231 cells, which is suggested as one of the mechanisms for RT resistance [20]. Figure 1A shows a graphical representation of the proteins concerning the fold change. These findings suggested that the upregulation of MAPK1, CLK1, and FGF22 and the downregulation of caspase 3 might be involved in the acquisition of radio-resistant MDA-MB-231 cells.   (Table 1).
Among the 16 downregulated proteins, caspase 3 was the most downregulated in the RT-R-MDA-MB-231 cells, which is suggested as one of the mechanisms for RT resistance [20]. Figure 1A shows a graphical representation of the proteins concerning the fold change. These findings suggested that the upregulation of MAPK1, CLK1, and FGF22 and the downregulation of caspase 3 might be involved in the acquisition of radio-resistant MDA-MB-231 cells. Table 1. GO enrichment analysis of differentially expressed proteins. GO enrichment analysis showed that MAPK1 is related to all the suggested signaling pathways involved in the radio-resistance of RT-R-MDA-MB-231 cells.

MAPK1 Is the Most Important Signaling Pathway in Acquiring Radio-Resistant RT-R-MDA-MB-231 Cells
Gene ontology (GO) enrichment analysis of differentially expressed proteins was carried out with the use of KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis. The KEGG pathway analysis showed that the most significant pathway involved in the RT-R-MDA-MB-231 cells was the MAPK1 signaling pathway ( Figure 1B). GO enrichment analysis suggested that MAPK1 is the most important signaling pathway in acquiring radio-resistant RT-R-MDA-MB-231 cells. The differentially expressed proteins were interrogated using the STRING database for the protein-protein interaction network analysis. String analysis of the protein-protein interaction (PPI) network generated an interconnected protein network with a medium confidence level of 0.04, which created a single module. The PPI network analysis of differentially expressed proteins showed a single module with 15 proteins such as MAPK1, CASP3, FGF22, MAPK11, HSP90AA1, and F2R. They are involved in MAPK signaling, NOD-like receptor signaling, PI3K-Akt signaling, and Pathways in cancer. The highly increased MAPK1 is related to all the suggested pathways. In addition, this module revealed that MAPK1 harbored a direct protein-protein interaction with caspase 3, which is crucial in inducing programmed cell death type 1 (apoptosis) (Figure 2). These findings support MAPK1 as being one of the important proteins involved in the acquisition of radio-resistant MDA-MB-231 cells.    These findings suggested that ERK signaling should be important in the cell survival of RT-R-MDA-MB-231 cells, and that the inhibition of ERK signaling might reverse the radio-resistance of MDA-MB-231 cells.

Inhibition of ERK Signaling Reversed the Radio-Resistance of RT-R-MDA-MB-231 Cells
To explore the radio-sensitivity of both p-MDA-MB-231 and RT-R-MDA-MB-231 cells, we performed a colony formation assay. This revealed that RT-R-MDA-MB-231 cells were resistant to radiation (RT) until 4 Gy, whereas p-MDA-MB-231 cells were sensitive to RT treatment ( Figure 4A,B). The colony number of RT-R-MDA-MB-231 cells was higher than that of p-MDA-MB-231 cells, which suggested that the RT-R-MDA-MB-231 cells were highly proliferative compared to p-MDA-MB-231 cells ( Figures 4A,B). To investigate the correlation between activated ERK signaling and radio-resistance in RT-R-MDA-MB-231 cells, we performed an ERK inhibition test with a colony formation assay. As shown in Figures 4C,D, the inhibition of MEK/ERK (at 20 μM of PD98059) reversed the radio-resistance of RT-R-MDA-MB-231 cells. These findings support the importance of activated ERK signaling for the radio-resistance of RT-R-MDA-MB-231 cells.

Inhibition of ERK Signaling Reversed the Radio-Resistance of RT-R-MDA-MB-231 Cells
To explore the radio-sensitivity of both p-MDA-MB-231 and RT-R-MDA-MB-231 cells, we performed a colony formation assay. This revealed that RT-R-MDA-MB-231 cells were resistant to radiation (RT) until 4 Gy, whereas p-MDA-MB-231 cells were sensitive to RT treatment ( Figure 4A,B). The colony number of RT-R-MDA-MB-231 cells was higher than that of p-MDA-MB-231 cells, which suggested that the RT-R-MDA-MB-231 cells were highly proliferative compared to p-MDA-MB-231 cells ( Figure 4A,B). To investigate the correlation between activated ERK signaling and radio-resistance in RT-R-MDA-MB-231 cells, we performed an ERK inhibition test with a colony formation assay. As shown in Figure 4C,D, the inhibition of MEK/ERK (at 20 µM of PD98059) reversed the radioresistance of RT-R-MDA-MB-231 cells. These findings support the importance of activated ERK signaling for the radio-resistance of RT-R-MDA-MB-231 cells. In Figure 3A, we found differences in the morphology between p-MDA-MB-231 cells and RT-R-MDA-MB-231 cells after ERK inhibition. To elucidate the differences in cell morphology between the two types of cells, we performed mitochondria staining, Mayer's hematoxylin staining for the cell structure, and DAPI for the nucleus. MitoTracker ® Red staining is used to show the live time status of mitochondria [21]. The staining revealed that, with the treatment of the MEK/ERK inhibitor, mitochondrial fragmentation was seen in RT-R-MDA-MB-231 cells at the 24 h-inhibition of ERK signaling ( Figure 5A  . The values are represented as mean ± standard deviation (SD) (n = 5). ** p < 0.01; *** p < 0.005.

Inhibition of ERK Signaling-Induced Necroptosis of RT-R-MDA-MB-231 Cells While It Induced the Apoptosis of p-MDA-MB-231 Cells
In Figure 3A, we found differences in the morphology between p-MDA-MB-231 cells and RT-R-MDA-MB-231 cells after ERK inhibition. To elucidate the differences in cell morphology between the two types of cells, we performed mitochondria staining, Mayer's hematoxylin staining for the cell structure, and DAPI for the nucleus. MitoTracker ® Red staining is used to show the live time status of mitochondria [21]. The staining revealed that, with the treatment of the MEK/ERK inhibitor, mitochondrial fragmentation was seen in RT-R-MDA-MB-231 cells at the 24 h-inhibition of ERK signaling ( Figure 5A). With the inhibition of ERK signaling, RT-R-MDA-MB-231 cells showed more fragmentation and swollen mitochondria than p-MDA-MB-231 cells did, suggesting that ERK inhibition con-  . The values are represented as mean ± standard deviation (SD) (n = 5). ** p < 0.01; *** p < 0.005.

ERK Inhibition Induced Caspase Activation and PARP-1 Cleavages in p-MDA-MB-231 Cells, While It Did Increase the Expression of Cyclophilin A (CypA) and AIF in RT-R-MDA-MB-231 Cells
To molecularly confirm the difference in cell death between p-MDA-MB-231 and RT-R-MDA-MB-231 cells, we performed Western blot analysis. Figure 6 demonstrates that, in MDA-MB-231 cells, ERK inhibition induced the cleavage PARP-1 and caspase-3, which is a hallmark for caspase-dependent apoptosis, but that, in the RT-R-MDA-MB-231 cells, ERK inhibition induced AIF (apoptosis-inducing factor), which positively regulates the CypA protein, which is considered a biomarker of necroptosis [22]. These findings support ERK inhibition inducing the apoptosis of p-MDA-MB-231 and the necroptosis of RT-R-MDA-MB-231 cells.

ERK Inhibition Induced Caspase Activation and PARP-1 Cleavages in p-MDA-MB-231 Cells, While It Did Increase the Expression of Cyclophilin A (CypA) and AIF in RT-R-MDA-MB-231 Cells
To molecularly confirm the difference in cell death between p-MDA-MB-231 and RT-R-MDA-MB-231 cells, we performed Western blot analysis. Figure 6 demonstrates that, in MDA-MB-231 cells, ERK inhibition induced the cleavage PARP-1 and caspase-3, which is a hallmark for caspase-dependent apoptosis, but that, in the RT-R-MDA-MB-231 cells, ERK inhibition induced AIF (apoptosis-inducing factor), which positively regulates the CypA protein, which is considered a biomarker of necroptosis [22]. These findings support ERK inhibition inducing the apoptosis of p-MDA-MB-231 and the necroptosis of RT-R-MDA-MB-231 cells. It was reported that CSC markers and EMT phenotypes were highly expressed in RT-

Protein Array Analysis
The

Protein Array Analysis
The total proteins of p-MDA-MB-231 and RT-R-MDA-MB-231 cells were isolated with a radioimmunoprecipitation assay (RIPA) buffer, which contained 0.1% NP-40 and 0.1% sodium dodecyl sulfate in phosphate-buffered saline (PBS) containing a protease inhibitor cocktail (Sigma Aldrich, St. Louis, MO, USA). The expression profiling of proteins was analyzed by a Signaling Explorer Antibody Array (Ebiogen, Seoul, Korea).

Bioinformatics Analysis
The obtained proteins from the antibody array analysis were further submitted to DAVID (The Database for Annotation, Visualization, and Integrated Discovery). DAVID is an online tool that provides a biological understanding between two or more data sets of genes, and it can also be used to determine gene ontology (GO) in terms of biological processes and cellular processes. To determine the pathways involved in the identified genes, KEGG (Kyoto Encyclopedia of Genes and Genomes) (https://www.genome.jp/ kegg/pathway.html, accessed on 20 October 2018) was employed. The selected genes were investigated for potential protein-protein interactions using STRING (Search Tool for the Retrieval of Interacting Genes) database version: 10.5 (https://string-db.org, accessed on 19 February 2021). For the display of protein interactions, selected proteins were uploaded into the STRING database and assessed using Cytoscape Software version Cytoscape_v3.7.1 (https://www.cytoscape.org, accessed on 19 February 2021). To access the interaction of the experimental data and to provide unambiguous comprehensive coverage, the online tool string was used.

Cell Culture
RT-MDA-MB-231 cells were established as previously described [8]. Briefly, MDA-MB-231 cells were fractionated with X-ray irradiation until a final dose of 50 Gy was reached. p-MDA-MB-231 and RT-MDA-MB-231 cells were cultured in RPMI-1450 medium supplemented with 10% heat-inactivated FBS and 1% penicillin/streptomycin. The cells were maintained at 37 • C in a 5% CO 2 incubator. The cells were grown with 80% confluence and were treated with a MEK/ERK inhibitor (PD98059) dissolved in DMSO or DMSO alone.

Cell Viability Assay
We used a calorimetric assay, MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide), to analyze the cell viability. The cells were seeded in 24-well plates with a confluence of 1 × 10 5 cells/well, they were treated with the MEK/ERK inhibitor, and they were maintained for 24 and 48 h at 37 • C in a 5% CO 2 incubator. After incubation, 50 µL of MTT (0.5 mg in 1× PBS) was added to each well and incubated for about 2 h at 37 • C in a 5% CO 2 incubator. The media were removed and the formazan crystals that formed in the live cells were dissolved with the 500 µL of DMSO. The solubilized formazan crystals were transferred to 96-well plates and the absorbance was read by an enzyme-linked immunosorbent assay (ELISA) reader at 540 nm. The cell viability was quantified in percentage, while vehicle-treated control cells were set at 100%.

Colony Formation Assay
P-MDA-MB-231 or RT-R MDA-MB-231 cells (1 × 10 3 cells/well) were seeded in sixwell plates, treated with the indicated doses of the MEK/ERK inhibitor, and maintained at 37 • C in a 5% CO 2 incubator. The cells were irradiated with a given concentration, and the media were discarded after 24 h and replaced with fresh complete media every 2-3 days. After 14 days, the medium was discarded and the cells were washed with 1× PBS thrice. The colonies were fixed with absolute methanol for 10 min, stained with Giemsa staining solution, and then maintained at room temperature. The number of colonies was counted using ImageJ software.

Mitotracker Red Analysis
For mitochondrial morphology analysis, Mitotracker Red dye was used. The cells were seeded with a confluence of 1 × 10 5 cell/well in 12-well plates, they were treated with the MEK/ERK inhibitor for 24 h, and they were washed with 1× PBS and then stained with 0.5 µL of Mitrotracker red in 500 µL of 1× PBS. The cells were incubated for 30 min in a 5% CO 2 incubator at 37 • C. After incubation, the cells were viewed under a fluorescent microscope for the analysis of the live mitochondrial status after the treatment of a MEK/ERK inhibitor.

Hematoxylin Staining
The cells were seeded in 12-well plates with a confluence of 1 × 10 5 cells/well and were grown for 24 h with the MEK/ERK inhibitor at 37 • C in a CO 2 incubator. After incubation, the cells were washed with 1× PBS and then fixed with 4% para-formaldehyde overnight. The fixed cells were washed thrice with 1× PBS for about 5 min per wash, they were stained with 200 µL of Mayer's hematoxylin staining solution, and they were incubated for 20 min in the dark at room temperature. Then, the cells were washed thoroughly with 1× PBS, followed by 1 mL of 90% glycerol, and they were observed under a phase-contrast microscope.

DAPI (4 ,6-diamidino-2-phenylindole) Staining
For the nuclear morphological changes, DAPI staining was performed. The cells were seeded in 12-well plates at a density of 1 × 10 5 cells/well with the treatment of the MEK/ERK inhibitor for about 24 h, they were washed with 1× PBS, and they were fixed overnight with 4% formaldehyde. After fixation, the cells were washed with 1× PBS thrice for about 5 min per wash. DAPI solution (0.5 µL) was added to the 500 µL of 1× PBS, which was incubated for 30 min at 37 • C with a 5% CO 2 incubator. After incubation, the cells were washed with 1× PBS and were fixed with 90% glycerol in 1× PBS. The cells were viewed under a fluorescent microscope (Leica Microsystems GmbH, Wetzlar, Germany).

Western Blot Analysis
P-MDA-MB-231 and RT-MDA-MB-231 cells were seeded in 100 mm plates with a cell density of 2 × 10 6 cells/plate. The cells were treated with the MEK/ERK inhibitor or DMSO as a vehicle control and were maintained for 48 h at 37 • C with a 5% CO 2 incubator. After 48 h, the cells were harvested and transferred to 15 mL falcon tubes, and they were centrifuged for 5 min at 2000 rpm. The supernatant was discarded, and the tubes were centrifuged again to remove the residual supernatant. After complete removal of the supernatant, 500 µL of the 2X sample buffer containing 100 mM of Tris-Cl (pH 6.8), 4% (w/v) sodium dodecyl sulphate (SDS), 0.2% (w/v) bromophenol blue, and 200 mM of dithiothreitol was added. The protein lysates were collected in 1.5 mL Eppendorf tubes and kept at 100 • C for 10 min. The protein concentration was determined by the Bradford assay. In addition, 30 µg of the proteins was resolved in 8-12% SDS-PAGE and was transferred to a polyvinylidene difluoride membrane. After transfer, the membranes were blocked with 3% skimmed milk in Tris-buffered saline containing 1% Tween 20 (TBST) buffer for 30 min at room temperature, and they were incubated at 4 • C overnight with antibodies against actin (A5441, 1:  1:2000, Abcam). After overnight incubation in primary antibodies, the membranes were washed with TBST buffer thrice for about 10 min per wash. Then, the membranes were incubated in horseradish peroxidase (HRP)conjugated secondary antibody for 2 h at room temperature with 1:2000 dilution. The membranes were later washed with TBST buffer three times (10 min/wash) and were developed with ECL (electrochemiluminescence) solutions (Bio-Rad Laboratory, Hercules, CA, USA).

Statistical Analysis
All experiments were performed at least in triplicate, and all analyses were performed with the use of GraphPad Prism 7 software (GraphPad Software, San Diego, CA, USA). One-way ANOVA followed by the Newman-Keuls post hoc test was performed to compare various treatment groups. The data were presented as mean ± standard deviation (SD). A p-value <0.05 was considered statistically significant.

Discussion
Radiation therapy is one of the common and essential parts of breast cancer treatment. Around half of the cancer patients go through radiation therapy at some point in their treatment [23]. Ionizing radiation (IR) induces DNA damage through oxidative stress. The free OH radicals are capable of promoting single-stranded and double-stranded DNA breaks (SSB and DSB, respectively), which, in turn, triggers cell death [24]. Thus, cells develop IR resistance by counteracting the four 'R's,' which is DNA damage by DNA repair, redistribution, repopulation, and reoxygenation through the activation of various pathways [25]. Several studies have described the role of irradiation in breast cancer pathways and the involvement of several proteins in the development of resistance against radiation, which we have listed in Table 2. Only a small number of studies have investigated mechanisms of acquired radio-resistance through the generation of radio-resistant cell lines, which tend to focus on a single specific pathway. Thus, it is paramount to focus on the signaling mechanism through the generation of the radio-resistant breast cancer cell line.
This study was designed to find the hub signaling involved in the RT resistance of RT-R-MDA-MB-231 cells and to investigate the importance of the hub signaling in the reversal of radio-resistance and the regulation of the CSC and EMT phenotype that is highly associated with radio-resistance. We found that ERK signaling was highly activated in RT-R-MDA-MB-231 cells compared to in p-MDA-MB-231 cells and that ERK signaling was essential for the survival of both p-MDA-MB-231 and RT-R-MDA-MB-231 cells. In addition, the RT resistance of RT-R-MDA-MB-231 cells was reversed by the inhibition of ERK signaling (Figure 4). Furthermore, we demonstrated that the activated ERK signaling was associated with cancer stemness and EMT phenotype (Figure 7). Considering all these findings, we can conclude that activated ERK signaling is one of the major hub signaling related to the acquisition of radio-resistant MDA-MB-231 cells. Table 2. List of studies involving radiation resistance and sensitivity in MDA-MB-231 cells or TNBC, and their possible mechanisms behind it. The reports were mainly categorized into three parts as follows: (i) specific signaling pathways involved in radio-resistant breast cancer cells, (ii) signaling pathways involved in radio-sensitivity, (iii) altered expression of gene/proteins involved in radio-resistant breast cancer cells. (iv) Studies that did not fall into any of the three above categories. Abbreviations: IFIT2-interferon-induced protein with tetratricopeptide repeats 2, TRIB3-tribbles homolog 3, ESM-1-endothelial cell-specific molecule-1, DLX2-vertebrate distal-less homeobox 2.

No. Author Key Findings
Year References   2020 [40] (iv) Studies that did not fall into any of the three above categories. Before concluding, we should discuss some questions. The first question would be whether activated ERK signaling is the main mechanism for the radio-resistance of MDA-MB-231 cells. In Figures 4 and 7, the MEK/ERK inhibition test revealed that MEK/ERK inhibition induced cell death and suppressed the expression of CSC markers and the EMT phenotypes of both p-MDA-MB-231 cells and RT-R-MDA-MB-231 cells. This finding also suggested that ERK signaling is essential for the survival of MDA-MB-231 cells and that it may not be related to RT resistance per se. We also agree with the point. In addition, it was reported that other signaling pathways such as PI3K/Akt, and STAT or other anti-apoptotic proteins are important in the radio-resistance or radio-sensitivity of MDA-MB 231 cells (Table 2). However, there is some evidence to support that the ERK signaling is related to the RT-resistance of RT-R-MDA-MB-231 cells. Recent studies have depicted that the ERK1/2 activation prevails over the cell cycle arrest in the G2/M phase where cancer cells are susceptible to IR, thus inducing radio-resistance [5]. Another is that MDA-MB-231 cells are a triple-negative breast cancer cell line, which is known to have high CSC properties [49]. The third is that in our previous report, RT-R-MDA-MB-231 cells also showed an increased STAT 3 activity, which was reported to be related to cancer stemness and EMT, but that the inhibition of STAT 3 activity by the JNK inhibitor or Janus-activated kinase 2 (JAK2) inhibitor could not suppress the increased expression of CSC markers [50]. To solve this question, we carried out this study with antibody microarray analysis, which revealed that highly increased MAPK1 was enriched in all of the upregulated pathways of RT-R-MDA-MB-231 cells (MAPK signaling, NOD-like receptor signaling, PI3K-Akt signaling, and Pathways in cancer). In addition, the KEGG pathway enrichment analysis showed that all enriched pathways include MAPK1, and the PPI network analysis of differentially expressed proteins showed that MAPK1 could be related to the resistance of cell death. The third is that the inhibition of ERK signaling reversed RT resistance ( Figure 4). The inhibition of ERK signaling was reported to increase the anti-cancer efficacy of RT [51,52]. This finding could support the reversal of RT resistance caused by the inhibition of ERK signaling. With all of these findings, we can say that activated ERK signaling is one of the main mechanisms for the radio-resistance of MDA-MB-231 cells.
The second point to discuss would be the relationship between ERK signaling and EMT, as well as the CSC phenotype of RT-R-MDA-MB-231 cells, because it is mentioned that several other signaling pathways such as JAK/STAT, Hedgehog, Wnt, Notch, PI3K/PTEN, and nuclear factor-κB (NF-κB) signaling pathways, compared to ERK signaling, are closely related to CSC properties [53][54][55], and the blocking of these pathways involved might be an effective way to target CSCs [55]. Even though it is not common, PRMT6-dependent CRAF/ERK signaling was reported to regulate CSC plasticity [56]. In addition, it was reported that the CSC properties-related signal is frequently complexed, and there is cross-talk between and among the mentioned various pathways [57]. In addition, the influence of ERK activity in regulating the CSC phenotype is also reported in gemcitabineresistant pancreatic cells [58], cisplatin-resistant non-small cell lung cancer cells [59], and docetaxel and carboplatin-resistant ovarian cancer cells [60]. To determine the real cause of the activated ERK signaling of RT-R-MDA-MB-231 cells, we performed whole genome sequencing. In the study, we could not find any mutations on the linear line for the activation of ERK signaling (data not shown), such as EGFR, SOS, B-raf, Ras, or MEK. Regarding this question, we could not give the audience clear evidence. Therefore, further studies are required.
The third point to discuss is why the phenotype of cell death induced by ERK inhibition differed between p-MDA-MB-231 cells and RT-R-MDA-MB-231 cells, while ERK inhibition induced cell death and suppressed the increased expression of CSC markers and the EMT phenotype of both p-MDA-MB-231 cells and RT-R-MDA-MB-231 cells. We speculate that the reason could be that RT-R-MDA-MB-231 cells exhibit a decreased activity of caspase. It has been reported that the cancer cells harboring caspase defects frequently undergo necroptosis or necrosis instead of apoptosis when the death signal appears [61]. Initially, we thought that the defects in caspase 3 activity were the main cause that was associated with increased ERK signaling, which was revealed by string analysis of the PPI network. By whole genome sequencing, we also found that RT-R-MDA-MB-231 cells harbor a nonsynonymous single nucleotide mutation in CASP9 (Arg173His; rs2308950), which is known to be involved in the pathogenesis of various cancers (data not shown) [62,63]. We speculate that this mutation is also responsible for RT-R-MDA-MB-231 cells undergoing necroptosis during the inhibition of ERK signaling.
The fourth point to discuss would be the role of the other upregulated and downregulated proteins in RT-R-MDA-MB-231 cells. Although we could not discuss here all of the 26 proteins, recent studies have suggested that the inhibition of CLK1 also decreases cell proliferation [64]. CLK1 and FGF22 are oncogenes in cancer and their inhibition leads to the inhibition of breast cancer growth in cell culture and xenograft models [65,66]. This supports the possible contribution of upregulated CLK1 to the rapid growth of RT-R-MDA-MB-231 cells. The most downregulated protein, caspase 3, may also contribute to the radio-resistance of RT-R-MDA-MB-231 cells, by avoiding IR-induced apoptosis [67]. All of these findings suggest that the changes in the expression of proteins may be involved in the biological phenotype of RT-R-MDA-MB-231 cells. Regarding these points, further studies are warranted.
The fifth point to discuss would be the mechanisms driving the upregulation of ERK signaling in RT-R-MDA-MB-231 cells. As we know that radiotherapy works by damaging the DNA of cancer cells, our first thought was that the upregulation of ERK signaling would be related to some of the mutations in the Ras-Raf-MEK-ERK pathway. Therefore, we performed whole genome sequencing, but there was no additional mutation of ERFR, PI3K/Akt, Ras, Raf, MEK, or ERK molecules of MDA-MB-231 cells (data not shown). In this paper, we inhibited ERK signaling with PD98059, a non-adenosine triphosphate competitive MAPK (MEK) inhibitor [68]. Therefore, we can speculate that the upregulation of ERK signaling would be caused by some hidden mutations in the upstream of the Ras-Raf-MEK-ERK pathway.
The weakness of this study is that we performed the experiment with only one cell line. It is in question whether the main mechanism of the radio-resistance of RT-R-MDA-MB-231 cells can be applied to all radiation-resistant breast cancer cell lines or can be generalized to triple-negative breast cancer cells. In addition, even regarding the radio-resistance of RT-R-MDA-MB 231 cells, other signaling pathways are also suggested as a key signaling pathway involved in the resistance. Similar to the signaling involved in CSC, the signaling involved in the radio-resistance of RT-R-MDA-MB 231 cells could also be complexed. However, aberrantly upregulated ERK signaling contributes to cancer cell proliferation, survival, and metastasis [14], and many other reports have suggested that ERK signaling is an important signaling pathway in radio-resistance [5, 15,52]. Therefore, further research is also warranted regarding ERK signaling on the radio-resistance of breast cancer, especially on TNBC.

Conclusions
In summary, we found that ERK signaling was highly activated in RT-R-MDA-MB-231 cells compared to p-MDA-MB-231 cells. The activated ERK signaling was associated with an increased cancer stemness and EMT phenotype. In addition, the RT resistance of RT-R-MDA-MB-231 cells was reversed by the inhibition of ERK signaling. Furthermore, the inhibition of ERK suppressed the CSC marker proteins. With all of these findings, we conclude that activated ERK signaling is one of the major hub signals related to the acquisition of radio-resistant MDA-MB-231 cells. This study suggests a distinct and advantageous therapeutic value of the targeting of the ERK signaling pathway in MDA-MB-231 cells. Further research is also warranted regarding ERK signaling on the radio-resistance of breast cancer, especially on TNBC. interferon-induced protein with tetratricopeptide repeats 2 TRIB3 Tribbles homolog 3 ESM-1 Endothelial cell-specific molecule-1 DLX2 Vertebrate distal-less homeobox 2