Roles of p-ERM and Rho-ROCK signaling in lymphocyte polarity and uropod formation

Abstract

Front-rear asymmetry in motile cells is crucial for efficient directional movement. The uropod in migrating lymphocytes is a posterior protrusion in which several proteins, including CD44 and ezrin/radixin/moesin (ERM), are concentrated. In EL4.G8 T-lymphoma cells, Thr567 phosphorylation in the COOH-terminal domain of ezrin regulates the selective localization of ezrin in the uropod. Overexpression of the phosphorylation-mimetic T567D ezrin enhances uropod size and cell migration. T567D ezrin also induces construction of the CD44-associated polar cap, which covers the posterior cytoplasm in staurosporine-treated, uropod-disrupted EL4.G8 cells or in naturally unpolarized X63.653 myeloma cells in an actin cytoskeleton-dependent manner. Rho-associated coiled coil-containing protein kinase (ROCK) inhibitor Y-27632 disrupts the uropod but not the polar cap, indicating that Rho-ROCK signaling is required for posterior protrusion but not for ERM phosphorylation. Phosphorylated ezrin associates with Dbl through its NH2-terminal domain and causes Rho activation. Moreover, constitutively active Q63L RhoA is selectively localized in the rear part of the cells. Thus, phosphorylated ERM has a potential function in establishing plasma membrane "posteriority" in the induction of the uropod in T lymphocytes.

Figures

Figure 1.

EL4.G8 cells show polarized morphology with a clear uropod that is regulated by Ser/Thr kinases and the actin cytoskeleton. (A) Phase-contrast view showing the hand mirror–shaped morphology of EL4.G8 cells with clear uropods (left, arrows). Aggregation of EL4.G8 cells by their uropods (left). (B–E) Selective localization of the uropod markers (CD44 and CD43) and p-ERM in the EL4.G8 cell uropod, whereas F-actin and LFA-1 are enriched in the leading edge. Fixed, permeabilized cells were stained with fluorescence-labeled probes for cell surface CD44, CD43, or LFA-1, and for intracellular ezrin, p-ERM, or F-actin. (F) Staurosporine and cytochalasin D disrupt uropod structure. Cells were treated with DMSO (cont.), staurosporine (stauro.), or cytochalasin D (cyt.D) and stained for p-ERM and CD44. (G) Staurosporine abolishes ERM phosphorylation and the interaction between CD44 and ezrin in EL4.G8 cells. The total amount of ezrin and the p-ERM level were determined by Western blotting using lysates from control or staurosporine-treated cells (top and middle). CD44 ezrin interaction was detected by the immunoprecipitation of CD44 and Western blotting for ezrin (bottom). Bars (A–F), 10 μm.

Figure 2.

Intracellular localization of ezrin mutants and the enhancement of uropod size in T567D ezrin transfectants. (A) GFP-tagged ezrin constructs used in this study. In point mutant constructs T567A or T567D, a CT phosphorylation site, Thr567 (T) was replaced by Ala (A) or Asp (D), respectively. (B and C) Intracellular localization of ezrin mutants in EL4.G8 cells. Stable transfectants were examined for the localization of GFP as well as CD44 (B) or F-actin (C). Each set of panels corresponds to the construct shown to the left in A. T567D ezrin was exclusively localized at the uropod (asterisks) and colocalized with the enhanced F-actin signal (C, arrows). Bar, 10 μm. (D) GFP-tagged proteins expressed in stable transfectants were detected by Western blotting using antiezrin or anti-GFP antibody. Polyclonal antiezrin antibody recognizes an epitope in the CT domain and, thus, does not detect the NT fragment. The arrowhead indicates endogenous ezrin. (E) Interaction between CD44 and ezrin mutants was detected by the immunoprecipitation of CD44 and Western blotting for GFP. (F) T567D ezrin increases uropod length. The length of the uropod was measured and expressed as mean ± SD (n > 200). Each bar represents an individual stable clone of a different type of ezrin transfectant.

Figure 3.

T567D ezrin induces a plasma membrane polar cap in uropod-less cells. (A) T567D ezrin can cause formation of the polar cap in staurosporine-treated (arrows) but not in cytochalasin D–treated EL4.G8 cells. WT, T567D, and NT ezrin transfectants were treated with inhibitors and stained for CD44. Bar, 10 μm. (B) Percentage of cells exhibiting the polar cap in A (n > 130). (C) T567D-ΔAB ezrin lacks the ability to induce polar cap formation in staurosporine-treated EL4.G8 cells. Cells were transiently transfected with constructs for T567D or T567D-ΔAB ezrin, and after 24 h the cells were treated with staurosporine and stained for CD44. In contrast with the results for T567D-ΔAB, clear polar caps are observed in T567D ezrin–transfected cells (arrows). Bar, 10 μm. (D) Percentage of cells exhibiting the polar cap in C (n > 130). (E) T567D ezrin can induce organization of the polar cap in X63.653 myeloma cells (arrows), which are naturally unpolarized with regard to the plasma membrane. X63.653 cells were transiently transfected with constructs for WT, T567D, NT, or T567D-ΔAB ezrin and stained for CD44 24 h after the transfection. Bar, 10 μm. (F) Percentage of cells exhibiting the polar cap in C (n > 120).

Figure 4.

The Rho–ROCK pathway is involved in uropod formation. (A) Y-27632 abolishes uropod but not polar cap formation (arrows). Y-27632–treated parental EL4.G8 cells were stained for p-ERM and CD44 (top panels). GFP-tagged WT and T567D ezrin were also colocalized with CD44 in Y-27632–treated stable transfectants (middle and bottom panels). Bar, 10 μm. (B) Percentage of cells exhibiting the polar cap in A (n > 150). (C) Y-27632 does not alter the phosphorylation status of ERM proteins. The total amount of ezrin and the p-ERM level were determined by Western blotting using lysates from control or Y-27632–treated EL4.G8 cells. (D) WT RhoA shows a diffuse distribution in the cytoplasm, including the leading edge (arrowhead). Constitutively active (Q63L) RhoA is preferentially localized at the rear part of the cells (asterisk), and a dominant-negative form (T19N) abolishes uropod formation but not polar cap formation (arrow). EL4.G8 cells were transiently transfected with WT, Q63L, or T19N RhoA and examined for GFP and CD44 24 h after the transfection. Bar, 10 μm. (E) Percentage of cells exhibiting the uropod in D (n > 75).

Figure 5.

Rho-GEF activity associated with p-ERM and interaction of Dbl and open-form ezrin in EL4.G8 cells. (A) Protein complex containing p-ERM has Rho-GEF activity. p-ERM was immunoprecipitated from EL4.G8 cell lysate, and the associated Rho-GEF activity was measured compared with that when the immunoprecipitation was performed with control antibody. Rho-GTP in the reaction was pulled down and detected by Western blotting (top). The presence of p-ERM was also confirmed by Western blotting (bottom). (B) Ezrin interacts with Dbl. Dbl coimmunoprecipitated with ezrin was detected by Western blotting. (C) Open-form ezrin specifically binds to Dbl by its NT domain. Ezrin mutants were immunoprecipitated with anti-GFP antibody from lysates of stable transfectants and the interaction with Dbl was analyzed. (D) Most of the Rho-GEF activity associated with ezrin is associated with the phosphorylated form. p-ERM was immunodepleted from EL4.G8 cell lysate, and Rho-GEF activity associated with ezrin was measured compared with that when the immunodepletion was performed with control antibody. (D and E) The relative activity compared with the control level was normalized by the amount of immunoprecipitated ezrin and is shown as a percent of control (histograms). (E) Dbl mediates a major part of the association of Rho-GEF activity with ezrin. Dbl was immunodepleted and Rho-GEF activity associated with ezrin was analyzed. (F) Rho-GEF activity is increased in lysates containing T567D compared with WT ezrin and the activity is associated with the NT but not the CT domain. Rho-GEF activity was analyzed in lysates containing ezrin mutants. The activity in lysates containing T567D relative to that in lysates containing WT ezrin is shown (histogram). (G) Overexpression of Dbl-AID significantly inhibits uropod formation in EL4.G8 cells. EL4.G8 cells were transiently transfected with a construct for GFP-tagged Dbl-AID, and the percentage of cells exhibiting the uropod is shown (n > 180).

Figure 6.

A polar cap covers the posterior part of the cell. (A) EL4.G8 cells exhibit a unique cytoplasmic polarity. EL4.G8 cells were stained for CD44 and β-tubulin, F-actin and β-tubulin, β-tubulin and the posterior cytoplasm (C5-ceramide), or GM130 and the posterior cytoplasm (C5-ceramide). Arrowheads and an arrow indicate the positions of the MTOC and the Golgi apparatus, respectively, in each cell. (B) The polar cap is constructed over the posterior part of the cytoplasm (arrows). Y-27632– or staurosporine-treated EL4.G8 cells were stained for CD44 and the posterior cytoplasm. Composite images for β-tubulin/CD44/nucleus are shown (right) together with the MTOC position (arrowheads). (C) The T567D ezrin–associated polar cap covers the posterior cytoplasm in Y-27632– or staurosporine-treated transfectants (arrows). A similar situation is observed in Y-27632 but not in staurosporine treatment in WT ezrin transfectants (arrows). (D) WT and Q63L RhoA show different intracellular localization in the stable transfectants. WT RhoA is distributed diffusely in the cytoplasm, including the leading edge (arrowheads), whereas Q63L RhoA is localized in the posterior cytoplasm beneath the uropod, or beneath the polar cap when the transfectant is treated with Y-27632 (arrows). Similar but less clear localization of WT RhoA in the posterior cytoplasm is also observed in Y-27632-treated WT RhoA transfectants (arrows). (E) Schematic representation of a typical arrangement of cellular structures in an EL4.G8 cell. (F) The T567D ezrin–associated polar cap covers the posterior cytoplasm in X63.653 cells (arrows). Bars (A–D and F), 10 μm.

Figure 7.

Chemotactic migration of EL4.G8 cells toward SDF-1 and augmentation of the activity in T567D ezrin transfectants. Chemotactic activity was determined by constant-period counting using a flow cytometer and is shown as mean ± SD (A–C). (A) EL4.G8 cells exhibit typical chemotaxis toward SDF-1 (10 ng/ml), and this chemotaxis is markedly inhibited by pretreatment of the cells with pertussis toxin (PTx), staurosporine, or cytochalasin D. (B) EL4.G8 cell chemotaxis is dependent on ROCK activity. Cells were pretreated with Y-27632 and the chemotaxis assay was performed in medium containing the drug (in) or in medium only (wash out [w/o]). (C) T567D ezrin enhances chemotaxis. Transfectants preincubated with or without PTx were examined by the chemotaxis assay. Numbers represent the fold increase of chemotaxis toward SDF-1 (specific activity) compared with that of cells incubated in medium alone. A result typical of at least three experiments is shown.

Figure 8.

Model of lymphocyte polarization and uropod formation. During the construction of the motile polarized morphology in lymphocytes in response to chemoattraction or adhesion signals (directional cues), two codependent steps are proposed: “posteriorization” of the plasma membrane mediated by ERM phosphorylation, and “uropod protrusion” mediated by Rho–ROCK activation.