doi: 10.1016/j.devcel.2020.08.001. Epub 2020 Sep 1.
An Excitable Ras/PI3K/ERK Signaling Network Controls Migration and Oncogenic Transformation in Epithelial Cells
Affiliations
- PMID:32877650
- PMCID: PMC7505206
- DOI: 10.1016/j.devcel.2020.08.001
Item in Clipboard
An Excitable Ras/PI3K/ERK Signaling Network Controls Migration and Oncogenic Transformation in Epithelial Cells
Huiwang Zhan et al. Dev Cell..
Display options
Format
Display options
Format
Abstract
The Ras/PI3K/extracellular signal-regulated kinases (ERK) signaling network plays fundamental roles in cell growth, survival, and migration and is frequently activated in cancer. Here, we show that the activities of the signaling network propagate as coordinated waves, biased by growth factor, which drive actin-based protrusions in human epithelial cells. The network exhibits hallmarks of biochemical excitability: the annihilation of oppositely directed waves, all-or-none responsiveness, and refractoriness. Abrupt perturbations to Ras, PI(4,5)P2, PI(3,4)P2, ERK, and TORC2 alter the threshold, observations that define positive and negative feedback loops within the network. Oncogenic transformation dramatically increases the wave activity, the frequency of ERK pulses, and the sensitivity to EGF stimuli. Wave activity was progressively enhanced across a series of increasingly metastatic breast cancer cell lines. The view that oncogenic transformation is a shift to a lower threshold of excitable Ras/PI3K/ERK network, caused by various combinations of genetic insults, can facilitate the assessment of cancer severity and effectiveness of interventions.
Keywords: ERK; PI(3,4)P2; PI(4,5)P2; PI3K; Ras; excitability; oncogenic transformation; threshold; wave.
Copyright © 2020 Elsevier Inc. All rights reserved.
Conflict of interest statement
Declaration of Interests The authors declare no competing interests.
Figures
(A) Confocal images focusing at the ventral surface of the MDA-MB-231 cell expressing biosensors for F-actin (LifeAct-RFP) and PIP3 (PH-AKT-GFP) (also see Movie S1). Color-coded overlays show progression of waves as a function of time. (B) Relative amount of F-actin and PIP3 across the orange and blue boxes scan in (A). (C) Temporal change of F-actin and PIP3 in the yellow boxes in (A). (D-F) Distribution of the velocity (D), maximum width (E) and duration (F) among waves (mean ± S.D. of N = 20 waves in one cell).(G) Example of a circular wave with one edge propagating outwards to drive the formation of a protrusion upon reaching the cell perimeter (arrow indicates wave and dashed line indicates initial cell perimeter) (also see Movie S1). (H) Time-lapse confocal images of PH-AKT showing transitions between standing (marked with star) and travelling (marked with arrow) waves (also see Movie S1). The unit of time labeled on images is min and scale bar is 20 μm.
(A) Two colliding actin or PIP3 waves in MDA-MB-231 cells annihilated (dashed yellow boxes) (also see Movie S2). (B-D) Responses of ERK-KTR to EGF stimulation of different concentrations in MCF-10A cells. Ratios of cytosolic to nucleus ERK-KTR signal were plotted over time for 19, 21 and 20 cells in one experiment as shown in (B), (C) and (D). EGF stimulations were applied at 78 min. (E-H) Quantification of PIP3 responses to 0.5, 1 and 5 min saturated EGF stimuli in MCF-10A (E and F) and MDA-MB-231 (G and H) cells (mean ± S.E.M. of N = 10 cells of 3 experiments). (I and J) Peak responses to the second of two 1 min stimuli separated by interval T in MCF-10A (I) and MDA-MB-231 (J) cells (mean ± S.E.M. of N = 10 cells of 3 experiments). Example of individual cell response is shown in Figure S1A. Cells were starved in pure DMEM/F-12 (MCF-10A) or DMEM (MDA-MB-231) medium for 24 hours before EGF stimulation. The unit of time labeled on images is min and scale bar is 20 μm.
(A) Confocal images of the MDA-MB-231 cell showing the response of PIP3 and actin waves to global EGF stimulation (2 ng/ml, added at 0 min) (also see Movie S3). Cells were starved in pure DMEM medium for 2 hours before EGF stimulation. (B) Quantification of PIP3 change at the center (blue box) or periphery (orange box) of the cell ventral surface in (A). (C) Internal and external views of stacked frames of ROI (dashed yellow box) in (B). (D) Response of PIP3 wave to local EGF stimulation. Yellow star indicates the position of micropipette (filled with 10 μg/ml EGF, applied at 0 min). Also see Movie S3. Cells were starved in pure DMEM medium for 2 hours before EGF stimulation. (E) Internal and external views of stacked frames of ROI (dashed yellow box) in (D). (F) Wave maximum width versus distance to the micropipette in (D). The dashed black line is the power trend line. (G) Computational simulation of waves response to global stimulus based on an activator-inhibitor scheme. Green is activator (X) while red is inhibitor (Y). Global stimulus was added at time 0 (a.u.) (also see Movie S4). (H) Computational simulation of waves response to local stimulus. Green is activator while red is inhibitor. Yellow star indicates the position of stimulus (also see Movie S4). The unit of time labeled on images is min and scale bar is 20 μm.
(A) Time-lapse confocal images of MCF-10A cell expressing Lyn-FRB and FKBP-Inp54p (channel shown). 1μM rapamycin was added at 0 min (also see Movie S5). (B) Kymograph of lamellipodia length around the perimeter of the cell in (A). Quantification of cell area change is shown in Figure S2F. (C) Time-lapse confocal images of the bottom surface of MCF-10A cell expressing Lyn-FRB (shown in Figure S3A), FKBP-Inp54p, and RBD (channel shown) after 1μM rapamycin indicate active Ras enriched at the oscillatory protrusions. Also see Movie S5. (D) Quantification of the boxes scan in (C) and Figure S3A. The bold are the smoothed lines. (E) Confocal images of the middle plane of MCF-10A cell expressing Lyn-FRB, FKBP-Inp54p, and RBD (channel shown) before and after 1μM rapamycin. (F) Quantification of the boxes scan in (E). (G) Confocal images of MCF-10A cells expressing Lyn-FRB, FKBP-Inp54p, and ERK-KTR (channel shown) before and after 1μM rapamycin (also see Movie S5). (H and I) Ratios of cytosolic to nucleus ERK-KTR signal in population average (H, mean ± S.D. of N = 31 cells in one experiment) and individual (I) MCF-10A cells. (J and K) Time-lapse confocal images showing increased PIP3 waves (J, also see Movie S6) or Ras waves (K, also see Movie S6) formation after lowering PI(4,5)P2 in MDA-MB-231 cells. 1μM rapamycin was added at 0 min. T-stack analysis of (J) is presented in Figure S3D. (L) Quantification of wave number change in (J and K). Individual waves were followed from origin to end in videos. The total wave number for each cell was quantified during 1 h imaging windows before and after 1μM rapamycin (mean ± S.D. of N = 48 cells, 5 experiments). Paired t test, P = 0.0004. The unit of time labeled on images is min and scale bar is 20 μm.
(A-D) Responses of ERK-KTR to 10 pg/ml EGF stimulation in MCF-10A cells without (A and B) and with (C and D) PIP2 reduction by rapamycin-induced recruitment of Inp54p. Ratios of cytosolic to nucleus ERK-KTR signal over time for 12 and 19 individual cells are plotted in (B) and (D). (E) Fraction of cells showing at least 50% increase in the ratio of cytosolic to nucleus intensity of ERK-KTR in response to 10 pg/ml EGF global stimulation without or with PIP2 reduction by rapamycin-induced recruitment of Inp54p (mean ± S.D. of N = 10 experiments). Unpaired t test with Welch’s correction, P< 0.0001. (F) Fraction of cells responding to global stimulation of various concentrations of EGF in cells without or with PIP2 reduction by rapamycin-induced recruitment of Inp54p (mean ± S.D. of N = 10 experiments). Cells were starved in pure DMEM/F-12 medium for 24 hours before stimulation. Scale bar labeled on images is 20 μm.
(A) Quantification of area changes of MCF-10A cells expressing Lyn-FRB and FKBP-Inp54 (images shown in Figure S5C and S5D) before and after 1μM rapamycin (mean ± S.E.M. of N = 15 cells of 5 experiments). Cells were either pre-treated with DMSO or 10 μM mTOR inhibitor PP242 for 2h before imaging. (B) Quantification of area changes of MCF-10A cells expressing Lyn-FRB and FKBP-Inp54p (images shown in Figure S5G–I) transfected with scrambled, mTOR, or Rictor shRNAs before and after 1μM rapamycin (mean ± S.E.M. of N = 20 cells of 5 experiments). (C) Quantification of area changes of MCF-10A cells before and after treated with DMSO or 10 μM Ulixertinib (ERK inhibitor) (mean ± S.E.M. of N = 20 cells for Ulix and 22 for DMSO of 3 experiments). (D) Quantification of area changes of MCF-10A cells before and after rapamycin-induced PIP2 reduction (mean ± S.E.M. of N = 18 cells for Ulix and 17 for DMSO of 3 experiments). Cells were either pre-treated with DMSO or 10 μM Ulixertinib for 2h before imaging. (E) Confocal images of MCF-10A cells expressing PH-TAPP1-GFP and PH-AKT-RFP. (F) Merged images of both fluorescence channels in (E). Green is PH-TAPP1 and red is PH-AKT. (G) Quantification of the boxes scan in (F). Similar results for MDA-MB-231 cells are shown in Figure S6A–C. (H) PI(3,4)P2 and PI(3,4,5)P3 responses to global EGF (2 ng/ml) stimulation in MCF-10A cell. Thin vertical lines mark 50% values. Cells were starved in pure DMEM/F-12 medium for 24 hours before EGF stimulation. (I) Time-lapse confocal images of MCF-10A cell expressing Lyn-FRB and FKBP-INPP4B (channel shown). 1μM rapamycin was added at 0 min. (J) Quantification of the cell area changes of individual MCF-10A cells (colored lines) or population average (black line, mean ± S.E.M. of N = 15 cells of 3 experiments) before and after rapamycin-induced PI(3,4)P2 reduction. The phosphatase inactive control is shown in Figure S6D. (K) Time-lapse confocal images of the ventral surface of MCF-10A cell expressing Lyn-FRB (channel shown), FKBP-INPP4B, and RBD (channel shown). 1μM rapamycin was added at 0 min. Quantification is shown in Figure S6E and S6F. (L) Diagram of the working model. Green arrows indicate activating interactions while red bars indicate inhibitory interactions. The unit of time labeled on images is min and scale bar is 20 μm.
(A) Time-lapse confocal images of MCF-10A cell expressing Lyn-FRB and FKBP-Kras_G12V (channel shown). 1μM rapamycin was added at 0 min. (B) Time-lapse confocal images of MCF-10A cell expressing Lyn-FRB and FKBP-CDC25 (channel shown). 1μM rapamycin was added at 0 min. (C) Time-lapse confocal images of MDA-MB-231 cell expressing Lyn-FRB, FKBP-CDC25, and PH-AKT (channel shown). 1μM rapamycin was added at 0 min. Also see Movie S7. (D) Time-lapse confocal images of control or Kras_G12V transformed MCF-10A cells expressing LifeAct. Also see Figure S7C–F, and Movie S7 for additional examples. (E) Quantification of fraction of cells with wave in control or Kras_G12V transformed MCF-10A cells during a 2-hour imaging window (mean ± S.E.M. of N = 4 independent experiments, 306 control cells and 347 Kras transformed cells). Unpaired t test with Welch’s correction, P = 0.0071. (F and G) Heat map of ratios of cytosolic to nucleus ERK-KTR signal of control (F) or Kras_G12V transformed (G) MCF-10A cells. N = 25 cells of 2 experiments for each group. (H) Time-lapse confocal images of M1~M4 MCF-10A cells expressing LifeAct. Also see Movie S7 for more examples. Yellow arrows indicate example waves. (I) Quantification of fraction of cells with wave in M1~M4 MCF-10A cells during a 2-hour imaging window (mean ± S.D. of N = 4 independent experiments, 501 M1 cells, 608 M2 cells, 234 M3 cells, and 302 M4 cells). M2 VS M1: Unpaired t test with Welch’s correction, P = 0.0013; M3 VS M1: Unpaired t test, P< 0.0001; M4 VS M1: Unpaired t test, P< 0.0001; M3 VS M2: Unpaired t test, P< 0.0001; M4 VS M3: Unpaired t test, P = 0.0639. The unit of time labeled on images is min and scale bar is 20 μm.
References
- Aoki K, Kumagai Y, Sakurai A, Komatsu N, Fujita Y, Shionyu C, and Matsuda M (2013). Stochastic ERK activation induced by noise and cell-to-cell propagation regulates cell density-dependent proliferation. Mol. Cell 52, 529–540. - PubMed
Publication types
MeSH terms
Substances
Grants and funding
LinkOut - more resources
Full Text Sources
Research Materials
Miscellaneous