The Na+-H+ exchanger-1 induces cytoskeletal changes involving reciprocal RhoA and Rac1 signaling, resulting in motility and invasion in MDA-MB-435 cells

Abstract

Introduction: An increasing body of evidence shows that the tumour microenvironment is essential in driving neoplastic progression. The low serum component of this microenvironment stimulates motility/invasion in human breast cancer cells via activation of the Na+-H+ exchanger (NHE) isoform 1, but the signal transduction systems that underlie this process are still poorly understood. We undertook the present study to elucidate the role and pattern of regulation by the Rho GTPases of this serum deprivation-dependent activation of both NHE1 and subsequent invasive characteristics, such as pseudopodia and invadiopodia protrusion, directed cell motility and penetration of normal tissues.

Methods: The present study was performed in a well characterized human mammary epithelial cell line representing late stage metastatic progression, MDA-MB-435. The activity of RhoA and Rac1 was modified using their dominant negative and constitutively active mutants and the activity of NHE1, cell motility/invasion, F-actin content and cell shape were measured.

Results: We show for the first time that serum deprivation induces NHE1-dependent morphological and cytoskeletal changes in metastatic cells via a reciprocal interaction of RhoA and Rac1, resulting in increased chemotaxis and invasion. Deprivation changed cell shape by reducing the amount of F-actin and inducing the formation of leading edge pseudopodia. Serum deprivation inhibited RhoA activity and stimulated Rac1 activity. Rac1 and RhoA were antagonistic regulators of both basal and stimulated tumour cell NHE1 activity. The regulation of NHE1 activity by RhoA and Rac1 in both conditions was mediated by an alteration in intracellular proton affinity of the exchanger. Interestingly, the role of each of these G-proteins was reversed during serum deprivation; basal NHE1 activity was regulated positively by RhoA and negatively by Rac1, whereas RhoA negatively and Rac1 positively directed the stimulation of NHE1 during serum deprivation. Importantly, the same pattern of RhoA and Rac1 regulation found for NHE1 activity was observed in both basal and serum deprivation dependent increases in motility, invasion and actin cytoskeletal organization.

Conclusion: Our findings suggest that the reported antagonistic roles of RhoA and Rac1 in cell motility/invasion and cytoskeletal organization may be due, in part, to their concerted action on NHE1 activity as a convergence point.

Figure 1

Efficiency of transfection with Rho family constructs. MDA-MB-435 cell monolayers were transfected with variable concentrations of the pBabe vector containing either the dominant negative (a) N17Rac1 or (b) 17NCdc42 construct. Forty-eight hours later the monolayers were fixed with 4% paraformaldehyde, and the expression of the haemagglutinin (HA) tag visualized by immunofluorescent microscopy with an anti-HA antibody. The percentage of cells expressing HA-tagged N17Rac1 increased in a concentration-dependent manner and reached approximately 90% when the cells were transfected with 10 μg plasmid cDNA.

Figure 2

Role of Na⁺–H⁺ exchanger (NHE)1 in serum deprivation-dependent rearrangement of the actin cytoskeleton. (a) To examine the role of NHE1 in serum deprivation-induced reorganization of the actin cytoskeleton underlying motility and invasive ability, MDA-MB-435 cell monolayers were left in fresh whole growth medium (nondeprived [ND]) or were serum deprived (D) in the presence or absence of 2 μmol/l of the specific NHE1 inhibitor 5-(N, N-dimethyl)-amiloride (DMA) for 1 day. The fluorescent photomicrographs show cell groups at a magnification of 40× (bar = 10 μm). (b) typical one-parameter fluorescence distributions of MDA-MB-435 cells labelled with phalloidin–fluorescein isothiocyanate in the above experimental conditions. The lighter coloured curve is from control, ND cells. Cyt B is the fluorescence distribution after treatment with the actin cytoskeleton disruptor cytochalasin B (5 μmol/l for 15 min). (c) Relative intracellular F-actin content measured by flow cytometry in MDA-MB-435 cells in the above experimental conditions. Mean ± standard error. n = 9; **P < 0.001 versus ND cells.

Figure 3

Serum deprivation activates Rac1 and inactivates RhoA without changes in total cellular expression. MDA-MB-435 cells were nondeprived (t0) or deprived of serum for 6 or 24 hours and cell extracts were assayed for (a) Rhotekin or (b) p21-activated kinase (PAK-1)-binding activity, respectively, as described in the Methods section. Representative immunoblots from these experiments are shown. For these experiments, the RhoA bound to the RBD or Rac1 bound to PAK-1 (upper panels) were normalized to the total RhoA or Rac1 content of cell extracts (lower panels). GST, glutathione-S-transferase; RBD, Rho-binding domain of Rhotekin (upper panel) or Rac-binding domain (lower panel); WB, Western blot.

Figure 4

Effect of Rho family mutants on the basal and serum deprivation stimulation of Na⁺–H⁺ exchanger (NHE)1 activity. To determine the role of Rho family G-proteins in the regulation of basal and serum deprivation stimulated Na⁺–H⁺ exchanger (NHE)1 activity, monolayers were transfected with empty pBabe vector or vector containing mutated Rho family genes. After 24 hours, medium was replaced with either serum replete or deplete medium for an additional 24 hours. Intracellular pH recoveries were measured and the initial rate of the NHE1 activity calculated as Na⁺-dependent H⁺ efflux, as described in the Methods section. All measurements were conducted in nominally HCO₃-free, HEPES-buffered solutions in order to measure only the NHE. (a) Effect of dominant negative (dn) Rho family mutants on the serum deprivation-dependent upregulation of NHE1 activity. (b) Effect of constitutively active (ca) RhoA and Rac1 mutants on the serum deprivation-dependent upregulation of NHE1 activity. Data for control cells are shown in the gray solid bars, whereas the stippled bars represent data for Cdc42, stripped bars represent data for Rac1 and hatched bars represent data for RhoA. Data are expressed as means ± standard error for between 9 and 20 observations in each condition. *P < 0.01, **P < 0.001 versus control values; ⁺P < 0.01, ⁺⁺P < 0.001 versus deprived values. Insets show the level of transfection of haemagglutinin (HA)-tagged cDNA in a typical experiment by Western blot of whole cell extracts using anti-HA antibody.

Figure 5

Effect of dominant negative (dn) RhoA and Rac1 mutants on the intracellular (pHi) dependence of net acid extrusion rate. To examine the mechanism of RhoA and Rac1 modification of the serum deprivation-dependent increase in Na⁺–H⁺ exchanger (NHE)1 activity, the intracellular pH (pHi) dependence of NHE1 activity (Δmmol/l H⁺ _i/minute) in MDA-MB-435 cell monolayers was measured. Monolayers were left in fresh whole growth medium (open circles) or serum deprived (closed circles) plus dn-N19RhoA (triangles) or dn-N17Rac1 (squares) for 24 hours. Results from a typical experiment are shown. The plots were computed as described in the Methods section. H⁺ _i, intracellular proton concentration.

Figure 6

Role of RhoA and Rac1 in serum deprivation-dependent rearrangement of the actin cytoskeleton and cell shape. To examine the role of RhoA and Rac1 in serum deprivation-induced reorganization of the actin cytoskeleton underlying motility and invasive ability, MDA-MB-435 cell monolayers were (a) left in fresh whole growth medium (nondeprived) or were serum deprived (deprived) for 1 day after transfection with empty pBabe vector (Cont) or with dominant negative (dn)-N19RhoA, constitutively active (ca)-V14RhoA, dn-Rac1 or V12Rac1 inserted in the same vector (bar = 10 μm). Arrows indicate the presence of lamellipodia whereas pseudopodia are indicated by arrowheads and invadopodia along the cell surface or at the end of an dominant leading edge pseudopodia are indicated by an asterist. (b) The complex actin organization of a long, dominant leading edge pseudopodia extension in a cell transfected with dn-RhoA and deprived of serum at a higher magnification (bar = 5 μm).

Figure 7

Effect of dominant negative (dn) and constitutively active (ca) RhoA and Rac1 mutants on the Na⁺–H⁺ exchanger (NHE)1-dependent stimulationof cell motility, invasive capacity and cell shape by serum deprivation. To examine the effect of RhoA and Rac1 on serum deprivation-induced motility and invasive ability of the cells, MDA-MB-435 cell monolayers were left in fresh whole growth medium (nondeprived [ND]) or were serum deprived (D) after having been transfected with dn or ca mutants of RhoA (stippled bars) or Rac1 (stripped bars), or empty vector (empty bars) for 1 day. (a) For motility, cells were then trypsinized, centrifuged and resuspended in serum-free medium with or without 2 μmol/l 5-(N, N-dimethyl)-amiloride (DMA). Cells (20,000) were added to the upper chamber of a Boyden Chamber, in which the lower chamber containing medium plus 1% serum was separated by a collagen I coated filter with 8 μm pores. After 4 hours of incubation, filters were removed, processed as described in the Methods section, and the cells that had traversed the filter were photomicrographed (20×). The bar is 10 μm. Circles are the 8 μm pores of of the filter. (b) As described in the Methods section, the number of cells that tranversed the collagen I coated filter were counted in photomicrographs taken at 100×. Values are expressed as mean ± standard error. n = 4; **P < 0.001 versus nondeprived cells. For invasion, cells were metabolically loaded with ³H-thymidine for the 24 hours prior to the experiment, MDA-MB-435 cells were trypinsized, centrifuged and resuspended in serum-free medium, and 30,000 cells were added to the confluent MCF-10A monolayer in each well of a 96 well culture plate. After 8 hours incubation, noninvaded cells were removed and (c) the monolayer was photographed at 10× or (d) the number of invaded cells determined as described in the Methods section by counting in a Packard TopCount NXT^® microplate scintillation counter. Values are expressed as mean ± standard error. n = 16; **P < 0.001 versus nondeprived cells.