IK channel activation increases tumor growth and induces differential behavioral responses in two breast epithelial cell lines

Abstract

Many potassium channel families are over-expressed in cancer, but their mechanistic role in disease progression is poorly understood. Potassium channels modulate membrane potential (Vmem) and thereby influence calcium ion dynamics and other voltage-sensitive signaling mechanisms, potentially acting as transcriptional regulators. This study investigated the differential response to over-expression and activation of a cancer-associated potassium channel, the intermediate conductance calcium-activated potassium channel (IK), on aggressive behaviors in mammary epithelial and breast cancer cell lines. IK was over-expressed in the highly metastatic breast cancer cell line MDA-MB-231 and the spontaneously immortalized breast epithelial cell line MCF-10A, and the effect on cancer-associated behaviors was assessed. IK over-expression increased primary tumor growth and metastasis of MDA-MB-231 in orthotopic xenografts, demonstrating for the first time in any cancer type that increased IK is sufficient to promote cancer aggression. The primary tumors had similar vascularization as determined by CD31 staining and similar histological characteristics. Interestingly, despite the increased in vivo growth and metastasis, neither IK over-expression nor activation with agonist had a significant effect on MDA-MB-231 proliferation, invasion, or migration in vitro. In contrast, IK decreased MCF-10A proliferation and invasion through Matrigel but had no effect on migration in a scratch-wound assay. We conclude that IK activity is sufficient to promote cell aggression in vivo. Our data provide novel evidence supporting IK and downstream signaling networks as potential targets for cancer therapies.

Keywords: bioelectricity; cancer; oncochannel; potassium channel.

Figure 1. Functional contribution of IK over-expression to current density and V_mem

(A) IK mRNA expression levels relative to GAPDH in total RNA collected from MDA-MB-231 infected with pMIG-RFP (Cont.) or pMIG-IK (IK) and selected for RFP fluorescence by FACS. Data are presented as mean with standard error of the mean (SEM) of 3 independent replicates (** significant difference p < 0.01, 2 sample t-test). (B) MDA-MB-231 control or MDA-MB-231-IK were fixed on coverslips and immunofluorescence microscopy was performed using antibodies to IK (green) with DAPI (blue) staining of the nuclei. Increased intensity of the green IK staining is evident in the MDA-MB-231-IK sample. (C) Endogenous and 1-EBIO induced current-voltage relationship in control and IK-expressing cells recorded in the cell attached perforated patch configuration from MDA-MB-231 cells. Data are presented as mean ± SEM from a minimum of 7 recordings. The current density was significantly increased in MDA-MB-231-IK 1-EBIO treated cells as compared to control vehicle treated, control 1-EBIO treated, and MDA-MB-231-IK vehicle treated cells (*** p < 0.001, 1-way ANOVA of 0 mV current density) (D) V_mem averaged over 20 seconds from recordings of same cells as C. Data points represent individual cells, bars show mean with SEM (*** indicates significantly different than vehicle treated control p < .001, 1-way ANOVA). (E) V_mem recording of representative MDA-MB-231 control cell (top) and MDA-MB-231-IK cell (bottom) with bath solution exchanged to 1-EBIO solution beginning at 30 seconds and continuing through duration of recording. There is a ~7 second delay between when the solutions are changed and when the new solution reaches the cells.

Figure 2. IK over-expression and activation with 1-EBIO had no effect on MDA-MB-231 proliferation, invasion, or migration

(A) Proliferation of control and IK over-expressing MDA-MB-231 cells was quantified over 4 days for cells treated with vehicle control or 1-EBIO. There was no significant difference in cell number on any day (Day 4: p = 0.33, 1-way ANOVA). (B) Invasion through Matrigel coated transwell after 16 h using FBS as a chemoattractant. Average number of cells per field of view with 4 fields of view per sample and 4 replicates in each of 3 independent experiments. There was no significant difference in the number of invading cells across any condition (p = 0.32, 1-way ANOVA) (C) A confluent layer of MDA-MB-231 control or MDA-MB-231-IK was scratched and healing was assessed after 10 h with vehicle control or 1-EBIO treatment. There was no significant difference in the normalized surface area of wound recovery (p = 0.41, 1-way ANOVA). All data presented as mean with SEM of three independent replicates.

Figure 3. MDA-MB-231 soft agarose colony formation is decreased by both IK expression and IK activation

(A) Bright field images of crystal violet stained MDA-MB-231 control and MDA-MB-231-IK cells grown in soft agarose and treated with vehicle control or 1-EBIO for 28 days. (B) Quantification of colonies from A. Data presented as mean with standard deviation of three independent replicates, *** indicates significantly different than vehicle treated control, +++ significant difference between indicated samples (p < 0.001 1-way ANOVA).

Figure 4. 1-EBIO treatment decreased MCF-10A proliferation and invasion but had no effect on migration

(A) Control and IK over-expressing MCF-10A proliferation was quantified over 4 days for cells treated with vehicle control or 1-EBIO (** p < 0.01 indicates significantly different than vehicle control, 1-way ANOVA). (B) Invasion through Matrigel-coated transwell after 24 h using horse serum and EGF as a chemoattractant. Average number of cells per field of view with 4 fields of view per sample and 4 replicates in each of 3 independent experiments (*** p < 0.001 indicates significantly different than vehicle treated control, +++ p < 0.001 indicates significant difference between indicated samples, 1-way ANOVA). (C) A confluent layer of MCF-10A control or MCF-10A-IK was scratched and treated with vehicle control or 1-EBIO. Healing was assessed after 12 h by quantifying the surface area of wound recovery (no significant difference across all conditions; p = 0.28, 1-way ANOVA). All data presented as mean with SEM of three independent replicates.

Figure 5. 1-EBIO treatment increased G2 phase accumulation in MCF-10A but had no effect on MDA-MB-231

Control and IK over-expressing MCF-10A (A) and MDA-MB-231 (C) cells were treated with vehicle control or 1-EBIO for 24 h, stained with annexin-FITC and violet dead stain, and analyzed by flow cytometry to quantify the percentage of positively stained cells (no significant difference in annexin-V or dead stained cells across any condition, 1-way ANOVA). (B, D) Plot of cell count versus propidium iodide (PI) fluorescence to analyze cell cycle. Arrow marks fluorescence intensity corresponding to G2 phase. Plot is representative of three independent replicates (MCF-10A **p < 0.01, MDA-MB-231 p = 0.29, 1-way ANOVA). All data presented as mean with SEM of three independent replicates.

Figure 6. IK over-expression increases in vivo tumor growth of MDA-MB-231

(A) MDA-MB-231 control (n = 11) and MDA-MB-231-IK (n = 9) cells were injected into the inguinal mammary fat pad of 6 wk old female mice and tumor size was plotted over time. Data presented as mean with SEM (* p < .05, ** p < .01, ***p < .001 2 sample t-test). (B) Fixed primary tumor sections stained with antibody to human nuclear antigen (green) with DAPI (blue) staining of the nuclei (top) or with anti-IK antibody (green; bottom). (C) Quantification of Ki67 positive (left) and active Caspase-3 positive (right) cells in primary tumor sections with representative images of Ki67 and active caspase-3 (green) staining with DAPI (blue) staining of the nuclei shown below. There was no significant difference in the percentage of either Ki67 or active caspase-3 positive cells (p = 0.27, 2-sample t-test and p = 0.15, 2-sample t-test).

Figure 7. No evidence of increased aggression in primary tumor

(A) Primary tumor sections stained with H&E. (B) Quantification of vessel density in primary tumors (top) with representative CD31 stained sections (bottom, green - CD31, blue - nuclei stained with DAPI; p =0.14, 2-sample t-test).

Figure 8. IK over-expression increases metastasis to the lung

(A) Quantification of HNA positive cells in lung tissue sections collected 4 weeks after mice were injected in the mammary fat pad with MDA-MB-231 cells. Data are mean with SEM of 3 sections per mouse control n =8, IK n = 9 p < 0.01, 2-sample t-test). (B) Representative image of HNA staining (green) with DAPI (blue) counterstain in full lung section with arrows indicating positive cells.