Cell culture

The human TNBC cell lines (MDA-MB-231, Hs578T) and the human NK cell line NK-92MI were obtained from the American Type Culture Collection. ELK3-knockdown (KD) MDA-MB-231 and Hs578T cell lines were previously described [12, 23]. Individual clones with stable suppression of ELK3, achieved through shRNA targeting to the 3’-UTR region (5’-GCCACAATTAAGGAC TCAT-3’), were selected and used in this study. Control cell lines for ELK3KD TNBC cells were established by transducing the cells with a non-silencing shRNA in a matching vector. ELK3 rescue in ELK3KD cells were achieved by transiently transfecting an ELK3-expressing plasmid into the ELK3KD cells. MDA-MB-231, Hs578T, and ELK3KD cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco/Life Technologies, GrandIsland, New York, USA). The medium used to culture Hs578T cells was supplemented with 0.01 mg/mL insulin (Sigma-Aldrich, St. Louis, USA). The NK-92MI cell line was cultured in α-minimum essential medium (α-MEM; Gibco/Life Technologies) supplemented with 2 mM L-glutamine (Gibco/Life Technologies), 0.1 mM β-mercaptoethanol (Gibco/Life Technologies), 0.02 mM folic acid (Sigma-Aldrich), and 0.2 mM inositol (Sigma-Aldrich). All media were supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin (Gibco/Life Technologies).

Plasmid DNA and siRNA

MDA-MB-231 and Hs578T cells were genetically engineered to achieve stable knockdown of ELK3 using retroviral vectors expressing shRNA targeting ELK3, following described in previous studies [23]. The LifeAct-mEGFP-7 expressing plasmid was obtained from Addgene (Massachusetts, USA), and the DNA encoding LifeAct-mEGFP-7 was used to yield a pCDH-LifeAct-mEGFP plasmid. Detailed information regarding the DNA plasmid constructs and siRNA utilized in this study can be found in Supplemental Table 1.

RNA-sequencing data analysis

To confirm which of the genes encoding each WAVE component shows a negative correlation with ELK3 expression, Gene Expression Omnibus (GEO) (RNA-seq; GSE197575) data deposited at the National Center for Biotechnology Information was used [12].

Cell migration assay

An 8.0 μm Transwell insert system (Corning, Arizona, USA) was utilized to analyze migration of cancer cells. Briefly, 1 × 104 cells in serum-free DMEM were seeded onto the insert, which was then placed in a 24-well plate containing complete medium. Following a 48 h incubation at 37 °C, the cells that migrated to the lower surface of the insert filter were fixed with 4% paraformaldehyde and stained with crystal violet (Sigma-Aldrich) at room temperature (RT) for 30 min. The number of cells that migrated to the lower chamber were examined under an optical microscope.

Cell adhesion assay

Briefly, 2.5 × 105 cells were seeded onto a 96-well plate pre-coated with collagen type 1 (Sigma-Aldrich). After incubating the plate at 37 °C for 30 min, non-adherent cells were washed away and the remaining cells were stained with crystal violet (Sigma-Aldrich) at RT for 10 min. Adherent cells were examined under an optical microscope.

RNA extraction and quantitative RT-PCR

Total RNA was extracted from cancer cells using Trizol (Invitrogen, California, USA) and 1 μg of total RNA was used for cDNA synthesis using the LeGene 1st strand cDNA synthesis system (LeGene Biosciences, California, USA). Quantitative PCR (qPCR) was carried out using TOPreal TM qPCR 2XPreMIX (Enzynomics, Daejeon, Republic of Korea) and the CFX Connect Real-Time PCR Detection System (Bio-Rad Laboratories, California, USA). Expression of mRNA was normalized to that of GAPDH using the comparative cycle method. The primers used in this study are listed in Supplemental Table 2.

Luciferase assay

A 1.5 kbp fragment of the CYFIP2 promoter region (−1450 bp ~ + 50 bp) was synthesized by Cosmo Genentech (Daejeon, Republic of Korea) and cloned into the pGL3 basic plasmid. ELK3KD MDA-MB-231 cells were transfected with the plasmid using Lipofectamine 2000 (Invitrogen). After 48 h, the transfected cells were harvested and lysed using cell lysis buffer (Cell Signaling Technology, Massachusetts, USA). The cell lysate was analyzed for luciferase activity using the Dual-Luciferase Reporter Assay System (Promega, Wisconsin, USA). The activity of firefly luciferase was normalized to the corresponding values for Renilla luciferase.

Western blot analysis

Cancer cells were lysed in radioimmunoprecipitation assay (RIPA) buffer (Cell Signaling Technology) supplemented with protease and phosphatase inhibitor cocktail buffer (GenDEPOT, Texas, USA). The concentration of the extracted proteins was determined in a BCA assay. After proteins were denatured by heating at 95 °C for 10 m, 60 µg of protein was separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene fluoride membrane (Bio-Rad Laboratories). The membrane was blocked with 4% bovine serum albumin (BSA) and incubated overnight at 4 °C with a primary antibody, followed by a secondary antibody at RT for 1 h. Immunoreactivity was detected using Enhanced Chemiluminescence solution (Thermo Fisher Scientific, Massachusetts, USA) and visualized using ChemiDoc™ XRS + system or ImageQuant LAS 4000 system (GE HealthCare, Barrington, IL, USA). The antibodies used in this study are listed in Supplemental Table 3.

Chromatin immunoprecipitation (ChIP) assay

Briefly, 1 × 106 ELK3KD MDA-MB-231 cells were seeded in a 100 mm dish. On the next day, the cells were transfected with a Flag-ELK3-expressing plasmid and incubated at 37 °C for 24 h. The cells were then fixed with 1% paraformaldehyde for 15 min to cross-link proteins and genomic DNA. To stop the cross-linking reaction, glycine was added at a final concentration of 125 mM. Cell lysates were prepared using cell lysis buffer (Cell Signaling Technology) supplemented with a protease/phosphatase inhibitor cocktail (Thermo Fisher Scientific). After fragmentation of genomic DNA by sonication, the cell lysates were centrifuged at 15,493 × g for 15 min at 4 °C and the supernatant containing fragmented genomic DNA was collected for immunoprecipitation. The protein concentration and total volume of lysates from all samples were equalized. Immunoprecipitation was performed at 4 °C overnight using protein A/G magnetic beads (Thermo Fisher Scientific) and either an anti-Flag antibody or rabbit IgG. The genomic DNA–protein-antibody complexes were washed sequentially (twice for 10 min each) with 1 × RIPA buffer, 1 × RIPA buffer with 300 mM NaCl (twice for 10 min each), LiCl buffer (twice for 10 min each), and TE buffer (once for 10 min). To separate the DNA–protein complexes, a solution containing proteinase K (Sigma-Aldrich) and 10% SDS was added to TE buffer, and the mixture was incubated at 65 °C overnight. The immunoprecipitated DNA was purified using the phenol/chloroform solution and then used for ChIP-qPCR analysis. The amount of immunoprecipitated chromatin was calculated as a percentage of the input.

CFSE/7-AAD assay

NK cells were stained for 20 min with 1 µM Cell Trace CFSE (Invitrogen) and then co-cultured for 4 h with cancer cells as target cells. Then, the cells were stained with 7-aminoactinomycin D (7-AAD, Thermo Fisher Scientific) to identify dead cells using a CytoFLEX flow cytometer (Beckman Coulter, Indiana, USA).

Immunocytochemistry

To visualize actin dynamics, 1 × 105 cells were seeded on coverslips, placed in a 12-well plate, and incubated overnight at 37 °C. The cells were then fixed for 30 min with 4% paraformaldehyde (PFA), followed by permeabilization for 10 min with 0.1% Triton X-100. Afterwards, the cells were pre-incubated for 30 min at RT with 1% BSA blocking solution. Cells were incubated overnight at 4 °C with primary antibodies (diluted 1:500), followed by a secondary antibody (Alexa Fluor 594 or Alexa Fluor 488 phalloidin, both 1:1000; Thermo Fisher Scientific). The stained cells were observed under a confocal microscope, and the length and number of filopodia were quantified using Image J software (ImageJ, Bethesda, MD, USA).

Time-lapse observation of the actin accumulation of cancer cells in the presence of NK cells

NK-92MI cells, which were stained with a cell trace (Invitrogen) and LifeAct-mEGFP-expressing cancer were seeded into a 96-well confocal plate. The actin movement in cancer cells was assessed under a time-lapse microscope. A modified Olympus FV3000 microscope, fitted with a 40 × (UPlanXApo, NA = 0.98) and a 60 × (UIPlanXApo, NA-1.42) objective lens, and an ANDOR Zyla 4.2 sCOMS camera was used for time-lapse observations. To image living cells, the microscope stage was equipped with an Olympus FV3000 incubation system (Live Cell Instruments, Seoul, Republic of Korea), which maintained the cell cultures at 37 °C, 5% CO2. Images acquired during these experiments were analyzed by CellSense software (Olympus, Ishikawa-machi, Hachioji, Tokyo, Japan) and Image J.

In vivo animal experiments

MDA-MB-231 lung metastasis model were established in 6-week-old female NSG mice (NOD-prkdcscidem1Il-2rgem1) obtained from JA BIO (Gyeonggi-do, Republic of Korea). Briefly, 1 × 106 MDA-MB-231 cells, engineered to express GFP and luciferase, control, ELK3KD and CYFIP2-silenced ELK3KD cells were injected intravenously. Then, to evaluate in vivo responses to NK cells, 3 × 106 NK-92MI cells were injected after 1 h. The GFP positive cells in the lung was quantified 3 days later using flow cytometry. GFP-labeled tumor cells were visualized by observing frozen lung tissue (sectioned at 8 µm) under a Zeiss LSM510 microscope.

All mice were maintained in a semi-specific pathogen-free animal facility at CHA University (Seongnam, Republic of Korea). Animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC 230135) of CHA University and were carried out in accordance with approved protocols.

Kaplan–Meier analysis

The probability of survival was calculated using the Kaplan–Meier (KM) Plotter Online Tool (http://www.kmplot.com) to evaluate the relationship between the differential expression of gene and survival in patients with breast cancer [24]. Patients were selected 392 based on ‘negative’ estrogen receptor status, ‘negative’ progesterone receptor status, a ‘negative’ human epidermal growth factor receptor 2. The ratio of genes was calculated as numerator: 215785_s_at (CYFIP2)/ denominator: 206127_at (ELK3). The probability of survival was calculated using the KM method, and log-rank tests were used to calculate the p-values.

Statistical analysis

Statistical analysis was conducted using GraphPad Prism software version 7 (GraphPad Software, San Diego, CA, USA). Data are presented as the mean ± standard deviation (SD), or as the standard error of the mean (SEM). Statistical significance was defined as follows: *P < 0.05, ** P < 0.01, *** P < 0.001, or **** P < 0.0001. The term “NS” denotes non-significant results.