ELK3-CYFIP2 axis-mediated actin remodeling modulates metastasis and natural killer cell responses in triple-negative breast cancer

Abstract

Triple-negative breast cancer (TNBC) is an aggressive, highly metastatic disease with a poor prognosis. E26 transformation-specific transcription factor (ELK3) is highly expressed in TNBCs, and functions as a regulator of epithelial-mesenchymal transition and immune responses. Because metastatic migration and immune evasion by TNBC cells are critical factors for successful metastasis, unravelling the underlying mechanisms and developing effective immunotherapeutic strategies is urgent. Here, TNBC cell lines MDA-MB-231 and Hs578T were examined to determine the relationship between ELK3 expression and filopodia protrusion on the cell membrane, as well as actin accumulation at contact sites with natural killer (NK) cells. RNA-sequencing analysis and molecular experiments were conducted to identify and validate downstream target genes of ELK3 associated with migration and attachment of TNBC cells. The immune response of TNBC to NK cells was evaluated through imaging and flow cytometry analyses. Clinical significance was assessed through Kaplan-Meier analysis of survival outcomes of TNBC patients. Gene expression profiling and molecular analysis revealed that oncogenic ELK3 directly suppresses expression of cytoplasmic FMR1 interacting protein2 (CYFIP2), a repressor of actin accumulation. Further molecular and pharmacological analyses confirmed that the ELK3-CYFIP2 axis serves a dual role in TNBC cell lines by (1) controlling filopodia-mediated migration and adhesion by regulating actin accumulation, and (2) regulating sensitivity to NK cells by modulating actin accumulation at contact sites. Kaplan-Meier analysis suggested that ELK3-CYFIP2 axis is associated with survival of TNBC patients, and that ELK3 suppresses transcription of CYFIP2. Thus, the ELK3-CYFIP2 axis plays a pivotal role in regulating actin, emphasizing its significance in controlling both cancer cell migration and NK cell responses in TNBC.

Keywords: Actin accumulation; CYFIP2; ELK3; Filopodia; Natural killer cell; Triple-negative breast cancer.

Fig. 1

ELK3 expression levels are associated with filopodia protrusion in TNBC cells. A Protrusion of filopodia from MDA-MB-231 and Hs578T cells was compared with that from ELK3KD cells transfected with an empty matching plasmid (control vector; CV) or an ELK3-expressing vector (ELK3). Cells were stained with DAPI and phalloidin. Actin accumulation associated with filopodia formation was visualized using fluorescence microscopy; representative protrusions are indicated by red arrows. Scale bar, 20 µm. B The number of filopodia per cell were quantified and is presented as individual dots. (MDA-MB-231 cells, n = 29, 30, and 30, respectively, Hs578T cells, n = 16, 67, and 35, respectively.) C The length of filopodia are presented in a graph. (MDA-MB-231 cells, n = 29, 30, and 30, respectively, Hs578T cells, n = 14, 17, and 17, respectively.) Data are presented as the standard error of the mean (SEM). Control (Cont) = sh control of MDA-MB-231 or Hs578T cells; ELK3KD = ELK3KD of MDA-MB-231 or Hs578T cells. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 2

ELK3 transcriptionally represses expression of CYFIP2 in TNBCs. A Heat map displaying the expression levels of WAVE complex subunits (CYFIP2, WASF3, BRK1, CYFIP1, NCKAP1, ABI2) identified by RNA-sequencing analysis of MDA-MB-231 control, ELK3KD cells and ELK3KD + ELK3 rescued cell. B–C Quantitative RT-PCR and immunoblot analysis of CYFIP2 in control and ELK3KD TNBC cells that were transfected with a control vector (CV) or an ELK3-expressing vector (ELK3). Control (Cont) = sh control of MDA-MB-231 or Hs578T cells; ELK3KD = ELK3KD of MDA-MB-231 or Hs578T cells. D Luciferase reporter assay showing activity of the CYFIP2 promoter. ELK3KD MDA-MB-231 cells were transfected with the indicated plasmid combinations for 48 h, followed by a luciferase assay. E ChIP-qPCR analysis of ELK3 binding to the CYFIP2 promoter. A Flag-ELK3-expressing plasmid or Flag-control plasmid was transfected into ELK3KD cells for 48 h, and Flag-immunoprecipitates were subjected to qPCR using primers specific for the CYFIP2 promoter region (−1450 to + 50 bp). All data were derived from at least three independent biological experiments. Data are presented as the mean ± standard deviation (SD). NS indicates no statistical significance. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 3

ELK3-CYFIP2 axis regulates metastatic nature of TNBCs by modulating filopodia protrusion. A Immunoblot analysis confirms the activity of siRNA targeting CYFIP2 (siCYFIP2) in ELK3KD MDA-MB-231 and Hs578T cells. ELK3KD TNBC cells transfected with a non-specific siRNA (siNS) or siCYFIP2. Relative band intensity of ELK3 and CYFIP2. Data are presented as the mean ± SD. B Filopodia formation was observed after staining with DAPI and phalloidin. Actin accumulation of filopodia formation was visualized using fluorescence microscopy; representative protrusions are indicated by red arrows. Scale bar, 20 µm. C The number of filopodia per cell were quantified, and is presented as individual dots. (MDA-MB-231 cells, n = 30, 30, and 26, respectively, Hs578T cells, n = 14, 30, and 30, respectively.) D The length of filopodia are presented in a graph. (MDA-MB-231 cells, n = 30, 30, and 26, respectively, Hs578T cells, n = 16, 30, and 30, respectively.) Data are presented as the SEM. E–F Representative images showing migration and adhesion of the indicated cells. Scale bar, 200 µm. All data were derived from at least three independent biological experiments. Data are presented as the mean ± SD. Control (Cont) = sh control of MDA-MB-231 or Hs578T cells; ELK3KD = ELK3KD of MDA-MB-231 or Hs578T cells. NS indicates no statistical significance. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 4

The ELK3-CYFIP2 axis regulates immune sensitivity of TNBCs to NK cells by regulating actin accumulation. A Time-lapse images showing the process of filopodia protrusion and actin accumulation in MDA-MB-231 control or ELK3KD cells at the region of contact with NK-92MI cells (E:T ratio = 1:1). The cancer cells were engineered to express LifeAct-mEGFP, which enables visualization of actin. Scale bar, 5 µm. B Quantification of MDA-MB-231 control or ELK3KD cells showing actin accumulation at the contact site with NK-92MI cells. Cancer cells showing an actin response are denoted as AR (in yellow), and cancer cells with no actin response are denoted as Non-AR (in gray). C The time taken to lyse of MDA-MB-231 control or ELK3KD cells after contact with NK-92MI. Data are presented as individual dots. (Cont: n = 51, and ELK3KD: n = 57.) D Actin accumulation at the contact site between MDA-MB-231 control cells, ELK3KD cells, ELK3KD cells transfected with siCYFIP2, and NK-92MI cells was observed under a fluorescence microscope. E Actin intensity at the contact site with NK-92MI cells was quantified. (Cont: n = 9, ELK3KD + siNS: n = 17 and ELK3KD + siCYFIP2: n = 16.) Data are presented as the standard error of the mean (SEM). F Immune response of MDA-MB-231 and Hs578T control cells, and ELK3KD cells, transfected with non-specific siRNA (siNS) or siCYFIP2 to NK-92MI. Cytotoxic activity of NK-92MI against cancer cells was measured in a CFSE/7-AAD assay (E:T ratio = 10:1). Control (Cont) = sh control of MDA-MB-231 or Hs578T cells; ELK3KD = ELK3KD of MDA-MB-231 or Hs578T cells. The experiments were performed in triplicate. Data are presented as the mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 5

Chemical inhibition of actin remodeling sensitizes immune response of MDA-MB-231 cells, but not ELK3KD cells, to NK cells. A Schematic showing the hypothesis underlying the immune response of CK-666 to inhibit actin polymerization. B Immune response of control and ELK3KD TNBC (MDA-MB-231 and Hs578T) in the presence or absence of CK-666. The cytotoxic activity of NK-92MI against cancer cells was measured in a CFSE/7-AAD assay (E:T ratio = 10:1). The experiments were performed in triplicate. Data are presented as the mean ± SD. C Time-lapse images showing filopodia protrusion and actin accumulation in MDA-MB-231 control or ELK3KD cells at the region of contact with NK-92MI cells (E:T ratio = 1:1) in the presence or absence of CK-666. Cancer cells were engineered to express LifeAct-mEGFP, which enables visualization of actin. Red dotted lines denote cancer cells undergoing lysis after contact with NK-92MI cells. Scale bar, 5 µm. D Quantification of MDA-MB-231 control or ELK3KD cells showing actin accumulation at the contact site with NK-92MI cells in the presence or absence of CK-666. Cancer cells showing an AR are denoted in yellow, and those showing Non-AR in gray. E Time taken for lysis of MDA-MB-231 control or ELK3KD cells after contact with NK-92MI in the presence or absence of CK-666. (MDA-MB-231 cells, -: n = 30, + : n = 29 and ELK3KD, -: n = 25, + : n = 56 respectively.) F Actin accumulation at the site at which MDA-MB-231 control and ELK3KD cells contact NK-92MI cells was assessed in the presence or absence of CK-666 and observed under a fluorescence microscope. G Actin intensity at the contact site between MDA-MB-231 control and ELK3KD cells and NK-92MI cells in the presence or absence of CK-666. (MDA-MB-231 cells, -: n = 15, + : n = 15 and ELK3KD, -: n = 12, + : n = 15 respectively.) Control (Cont) = sh control of MDA-MB-231 or Hs578T cells; ELK3KD = ELK3KD of MDA-MB-231 or Hs578T cells. Data are presented as the SEM. *P < 0.05; **P < 0.01; ***P < 0.001

Fig. 6

ELK3-CYFIP2 axis-mediated regulation of metastasis and NK responses in a mouse model bearing MDA-MB-231 tumors, and its clinical significance in TNBC patients. A Schematic of the in vivo experiment: GFP-expressing control, ELK3KD and CYFIP2-silenced ELK3KD MDA-MB-231 cells were injected intravenously into NSG mice. NK cells were injected intravenously 1 h later (each group n = 4). B Fluorescence images of GFP, indicating extravasated tumor cells in the lungs of mice at 3 days. Nuclei was stained with DAPI. Scale bar, 100 μm. C GFP positive cells in the lungs from each group of mice were quantified by flow cytometry. Control (Cont) = sh control of MDA-MB-231 ELK3KD = ELK3KD of MDA-MB-231 cells. Data represents the mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001. D Kaplan–Meier survival analysis of TNBC patients (n = 392), who were stratified into ‘high’ and ‘low’ groups based on the CYFIP2/ELK3 ratio. E Correlation between ELK3 and CYFIP2 expression levels in 392 patients with TNBC