E-cadherin engagement stimulates proliferation via Rac1

Abstract

E-cadherin has been linked to the suppression of tumor growth and the inhibition of cell proliferation in culture. We observed that progressively decreasing the seeding density of normal rat kidney-52E (NRK-52E) or MCF-10A epithelial cells from confluence, indeed, released cells from growth arrest. Unexpectedly, a further decrease in seeding density so that cells were isolated from neighboring cells decreased proliferation. Experiments using microengineered substrates showed that E-cadherin engagement stimulated the peak in proliferation at intermediate seeding densities, and that the proliferation arrest at high densities did not involve E-cadherin, but rather resulted from a crowding-dependent decrease in cell spreading against the underlying substrate. Rac1 activity, which was induced by E-cadherin engagement specifically at intermediate seeding densities, was required for the cadherin-stimulated proliferation, and the control of Rac1 activation by E-cadherin was mediated by p120-catenin. Together, these findings demonstrate a stimulatory role for E-cadherin in proliferative regulation, and identify a simple mechanism by which cell-cell contact may trigger or inhibit epithelial cell proliferation in different settings.

Figure 1.

Epithelial cell proliferation is dependent on seeding density and cell spreading. (A) Phase-contrast images of NRK-52E cells seeded at the indicated densities. (B) Graph of the percentage of NRK-52E cells entering S phase after seeding of synchronized cells at different densities, measured by BrdU incorporation. (C) Graph of cell spreading of NRK-52E seeded at different densities. (D–F) The same experiments as in A–C, but performed with MCF-10A cells. Error bars indicate the SEM of three independent experiments. Bars, 50 μm.

Figure 2.

Contact-stimulated proliferation is spreading independent. (A) Phase-contrast images of MCF-10A cells patterned in triangular (left) and bowtie-shaped (middle) microwells and then immunostained for E-cadherin (right). Graph of percentage of single and pairs of NRK-52E (B) and MCF-10A (C) cells entering S phase after being synchronized and seeded onto microwell substrates of indicated sizes. (D) MCF-10A cells embedded in collagen gels and immunostained for E-cadherin (green) and (E) BrdU (red), both of which were counterstained with Hoechst (blue). (F) Graph of percentage of MCF-10A cells within different sized clusters entering S phase after seeding of synchronized cells into collagen gels. Error bars indicate the SEM of three independent experiments. *, P < 0.05, calculated by t test and compared with single cells. Bars: (A) 10 μm; (D and E) 20 μm.

Figure 3.

E-cadherin is required for contact-mediated proliferation. Fluorescence images of E-cadherin in Ad-GFP– (left) and Ad-EΔ–infected (right) NRK-52E (A) and MCF-10A (D) 24 h after seeding. Graph of percentage of NRK-52E (B) and MCF-10A (E) cells synchronized and infected with Ad-GFP and Ad-EΔ entering S phase seeded at different densities. Graph of percentage of pairs and single Ad-EΔ− or Ad-GFP–infected NRK-52E (C) and MCF-10A (F) cells entering S phase. (G–I) The same experiments in MCF-10A cells as in D–F, except the cells were treated with HECD-1 blocking antibody to E-cadherin or mouse IgG (mIgG) control. (J) Western blots of a 120-kD E-cadherin and a 50-kD α-tubulin for control-transfected and E-cadherin siRNA–transfected MCF-10A cells. Graph of percentage of MCF-10A cells transfected with control or E-cadherin siRNA entering S phase when seeded at different densities (K) or in triangular or bowtie-shaped microwells (L). Differential interference contrast images of NRK-52E cells micropatterned onto different sized squares and (M) graph of percentage of synchronized NRK-52E and MCF-10A single cells entering S phase after seeding in different size microwells (N). Error bars indicate the SEM of at least three experiments, and a range of two independent experiments in L. *, P < 0.05, calculated by t test between control (Ad-GFP–infected, mIgG-treated, or control-transfected) and experimental (Ad-EΔ–infected, HECD-1–treated, or siRNA-transfected) cells in the same seeding condition. Bars: (A, D, and G) 20 μm; (M) 10 μm.

Figure 4.

Engagement of E-cadherin receptors stimulates proliferation. (A) Western blot of a 120-kD E-cadherin on hE-Fc–coated or control protein A–coated beads (top) and differential interference contrast image of patterned cell in a 750 μm² microwell with bead attached (bottom). (B) Graph of percentage of cells entering S phase of cells seeded in the indicated microwell sizes, with hE-Fc–coated beads or control protein A–coated beads. Error bars indicate the SEM of at least three experiments. *, P < 0.05, compared with control. P was calculated by t test. Bar, 10 μm.

Figure 5.

Rac1 is involved in cell–cell contact–stimulated proliferation. (A) Fluorescence images of E-cadherin in MCF-10A cells seeded for the indicated number of hours. (B) Western blot of Rac-GTP and total Rac1 levels (21 kD) in synchronized MCF-10A cells seeded at 2 × 10⁴ cells/cm² for the indicated time (top), and a graph of averaged relative Rac1 activity across three separate experiments (bottom). (C) Western blot of Rac1-GTP and total Rac1 levels (21 kD) of synchronized MCF-10A cells seeded at the indicated densities 8 h after seeding (top), and graph of averaged relative Rac1 activity across three separate experiments (bottom). (D) Fluorescence images of phalloidin-stained cells seeded at indicated densities. Arrow indicates the presence of lamellipodia. (E) Graph of percentage of synchronized MCF-10A cells infected with Ad-RacN17 entering S phase seeded at different densities. (F) Graph of percentage of pairs and single Ad-RacN17– or Ad-GFP–infected MCF-10A cells entering S phase. (G) Fluorescence image of E-cadherin in Ad-RacN17–infected cells 24 h after seeding. (H) Graph of percentage of MCF-10A cells synchronized and seeded at different densities in the presence of Y27632 entering S phase. Error bars indicate the SEM of at least three experiments. *, P < 0.05, between intermediate density compared with low or high density seeded cells (C), or between control (Ad-GFP or no treatment) and experimental (Ad-RacN17– or Y27632-treated) cells (E, F, and H). P was calculated by t test. Bars, 20 μm.

Figure 6.

Rac1 activity is downstream of E-cadherin. (A) Western blot of Rac-GTP and total Rac1 levels (21 kD) of synchronized MCF-10A cells infected with Ad-EΔ or Ad-GFP control and seeded for 8 h (top), and graph of averaged relative Rac1 levels across three separate experiments (bottom). (B) Fluorescence images of phalloidin-stained Ad-GFP– (left) and Ad-EΔ–infected (right) synchronized MCF-10A cells seeded at 2 × 10⁴ cells/cm² for 8 h. Arrow indicates the presence of lamellipodia. (C) Graph of cell area of synchronized MCF-10A cell infected with Ad-GFP or Ad-EΔ and seeded at varying densities for 24 h. Error bars indicate the SEM of at least three experiments in A, or the range between two experiments in C. *, P < 0.05, between Ad-GFP– or Ad-EΔ–infected cells at the indicated density. P was calculated by t test. Bar, 20 μm.

Figure 7.

p120-catenin is involved in E-cadherin stimulation of Rac1 and proliferation. (A) Western blot of p120 levels (120 kD) and GAPDH (38 kD) in control- or p120-siRNA–infected (top) and fluorescence images of p120 in control- (bottom left) and siRNA-infected (bottom right) MCF-10A cells. (B) Graph of percentage of synchronized control or p120-siRNA MCF-10A cells entering S phase when seeded at different densities. (C) Western blot of Rac1-GTP and total Rac1 levels (21 kD) in control- or siRNA-infected MCF-10A cells synchronized and then seeded at 10⁴ cells/cm² (top), and a graph of averaged relative Rac1 activity across three separate experiments (bottom). (D) Graph of percentage of synchronized control- or p120-siRNA–infected MCF-10A cells, which were also infected with Ad-GFP or Ad-RacN17 entering S phase when seeded at different densities. (E) Western blot of p120 levels (120 kD), Rac1-GTP, and total Rac1 levels (21 kD) in cells infected with control, p120-siRNA, and p120-siRNA with murine p120 or p120Δ, synchronized, and seeded at 2 × 10⁴ cells/cm² (top), and graph of averaged relative Rac1 activity across three experiments (bottom). (F) Graph of percentage of cells infected with control, p120-siRNA, and p120-siRNA with murine p120 or p120Δ, synchronized, and seeded at different densities entering S phase. Bar, 20 μm. Error bars indicate the SEM of at least three experiments. *, P < 0.05, compared with control. P was calculated by t test.

Figure 8.

Schematic of signaling pathway leading cell–cell contact to inhibition or stimulation of proliferation. Cell–cell contact stimulates proliferation through activation of Rac1 that depends on E-cadherin engagement and p120. Inhibition of proliferation by cell–cell contact results from a crowding-dependent, cadherin-independent decrease in cell–ECM spreading.