The Rho-family GEF Asef2 activates Rac to modulate adhesion and actin dynamics and thereby regulate cell migration

Abstract

Asef2 is a recently identified Rho-family guanine nucleotide exchange factor (GEF) that has been implicated in the modulation of actin, but its function in cell migration and adhesion dynamics is not well understood. In this study, we show that Asef2 is an important regulator of cell migration and adhesion assembly and disassembly (turnover). Asef2 localizes with actin at the leading edge of cells. Knockdown of endogenous Asef2 impairs migration and significantly slows the turnover of adhesions. Asef2 enhances both Rac1 and Cdc42 activity in HT1080 cells, but only Rac1 is crucial for the Asef2-promoted increase in migration and adhesion turnover. Phosphoinositide 3-kinase (PI3K) and the serine/threonine kinase Akt are also essential for the Asef2-mediated effects on migration and adhesion turnover. Consistent with this, Asef2 increases the amount of active Akt at the leading edge of cells. Asef2 signaling leads to an overall decrease in Rho activity, which is crucial for stimulating migration and adhesion dynamics. Thus, our results reveal an important new role for Asef2 in promoting cell migration and rapid adhesion turnover by coordinately regulating the activities of Rho-family GTPases.

Fig. 1.

Asef2 increases the amount of active Rac1 and Cdc42, but decreases Rho activity in HT1080 cells. (A) Schematic of the domain structure of Asef2. APC-binding region (ABR), Src homology-3 (SH3), Dbl homology (DH) and Pleckstrin homology (PH) domains are shown. (B) Immunoblot for Asef2 from HT1080 cells stably expressing GFP-Asef2. The ∼75 kDa band represents endogenous Asef2 whereas the ∼100 kDa band corresponds to exogenously expressed GFP-Asef2. (C) Quantification of the amount of GFP-Asef2 in stably expressing HT1080 cells relative to endogenous levels of the protein. Error bar represents s.e.m. from four separate experiments. (D) The GTP-bound (active) forms of Rac1, Cdc42 and RhoA were pulled down from lysates of GFP (control) or GFP-Asef2 stable cells. The total amount of each of these GTPases in cells is included as a loading control. Quantification of the amount of active GTPases from blots from four separate experiments is shown (lower panels). Error bars represent s.e.m. (^*P<0.04; ^**P<0.003).

Fig. 2.

Asef2 localizes with actin at the leading edge and promotes cell migration. (A) Cells stably expressing GFP (top panels) or GFP-Asef2 (bottom panels) were fixed and stained for actin using TRITC-phalloidin (red). Overlays are shown (right panels). Scale bar: 10 μm. (B) Quantification of the normalized leading edge fluorescent intensity for GFP (control) and GFP-Asef2 stable cells is shown. The normalized fluorescent intensity was significantly greater at the leading edge of GFP-Asef2 stable cells compared with GFP cells, confirming an enrichment of GFP-Asef2 at this region. Unlike Asef2, the normalized leading edge fluorescent intensity in GFP stable cells was negative because the fluorescent intensity in the cytosol was higher than at the leading edge, indicating that GFP is diffusely distributed in cells. Error bars represent s.e.m. for 20 cells from three separate experiments (^*P<0.0001). (C) Enlargements of the boxed regions in A. Scale bar: 5 μm. (D) Wild-type HT1080 cells were fixed and co-immunostained for endogenous Asef2 (green) and actin (red). The overlay is shown (far right panel). Scale bar: 5 μm. (E) GFP (control) and GFP-Asef2 stable cells were plated on fibronectin and images were collected every 5 minutes using time-lapse microscopy. Rose plots showing individual migration tracks for these cells are shown. Quantification of the migration velocity of GFP (control) and GFP-Asef2 stable cells is shown (far right). Error bars represent s.e.m. for 28-32 cells from at least three independent experiments (^*P<0.0001). (F) Plot of the migratory distance traveled for individual GFP (control) and GFP-Asef2 stable cells are shown.

Fig. 3.

Knockdown of endogenous Asef2 significantly impairs migration. (A) Wild-type HT1080 cells were transfected with empty pSUPER vector, scrambled siRNA (scr siRNA) or ASEF2 siRNAs. In some experiments, wild-type HT1080 cells were co-transfected with ASEF2 siRNA #2 and a truncated form of Asef2 (Asef2Δ204) (right panels). Cell lysates were immunoblotted for Asef2 or α-tubulin (loading control). (B) Quantification of endogenous amounts of Asef2 from cells transfected with the indicated constructs is shown. Error bars represent s.e.m. from four independent experiments (^*P<0.0001). (C) Wild-type HT1080 cells were transfected with empty pSUPER vector, scrambled siRNA (scr siRNA) or ASEF2 siRNAs and used in live-cell migration assays 3 days later. To show that the migration phenotype observed with cells expressing ASEF2 siRNA was due to endogenous loss of the protein, Asef2Δ204, which is a truncated, active form of Asef2, was co-expressed with ASEF2 siRNA #2. Rose plots with individual migration tracks for cells expressing the indicated constructs are shown. (D) Quantification of the migration velocity of cells transfected with constructs from C. Error bars represent s.e.m. for 30-35 cells from four separate experiments (^*P<0.009). (E) HT1080 cells were transfected with empty pSUPER vector, scrambled siRNA (scr siRNA) or ASEF2 siRNA #1 and cell lysates were assayed for active Rac1, Cdc42 and RhoA. Quantification of the amount of active GTPases from blots from four separate experiments is shown (lower panels). Error bars represent s.e.m. (^*P<0.003; ^**P<0.03). For B, D and E, asterisks denote a statistically significant difference compared with pSUPER-transfected cells.

Fig. 4.

Asef2 induces the formation of small, leading edge adhesions that turn over very rapidly. (A) GFP (control) and GFP-Asef2 stable cells were immunostained for endogenous paxillin or vinculin and visualized with TIRF microscopy. Scale bar: 10 μm. (B) Cells were transfected with mCherry-paxillin and imaged in red fluorescence. Time-lapse images show adhesions at the leading edge assemble and disassemble on a much more rapid time scale in GFP-Asef2 cells compared with control cells (arrowheads). Scale bar: 5 μm. (C) Quantification of the apparent t_1/2 for adhesion assembly and the t_1/2 for adhesion disassembly is shown (^*P<0.003). Error bars represent s.e.m. from 15-23 individual adhesions in 4-6 cells from at least three independent experiments.

Fig. 5.

Rac, but not Cdc42, is necessary for the Asef2-mediated effect on migration. (A) Wild-type HT1080 cells were transfected with empty pSUPER vector, scrambled siRNA (scr siRNA) or RAC siRNAs. Three days later, cells were lysed and immunoblotted for Rac or actin (loading control). Quantification of the amount of endogenous Rac from blots of cells transfected with the indicated constructs is shown (lower panels). Error bars represent s.e.m. from four independent experiments (^*P<0.0001). (B) HT1080 cells were transfected with empty pSUPER vector, scrambled siRNA (scr siRNA) or CDC42 siRNAs. After 3 days, cell lysates were immunoblotted for Cdc42 or α-tubulin (loading control). Quantification of endogenous levels of Cdc42 from blots of cells transfected with the indicated constructs is shown (lower panels). Error bars represent s.e.m. from four independent experiments (^*P<0.0001). For A and B, asterisks denote a statistically significant difference compared with pSUPER-transfected cells. (C,D) GFP and GFP-Asef2 stable cells were transfected with empty pSUPER vector, scrambled siRNA (scr siRNA), RAC or CDC42 siRNAs and used in migration assays. Rose plots with individual migration tracks are shown for cells expressing RAC siRNAs (C) or CDC42 siRNAs (D). (E,F) Quantification of the migration velocity of cells transfected with the indicated constructs is shown. Error bars represent s.e.m. for 30-35 cells from four separate experiments (E, ^*P<0.0001; F, ^*P<0.0002). For E and F, asterisks denote statistically significant differences compared with GFP (control) cells.

Fig. 6.

PI3K is necessary for Asef2-mediated migration. (A) GFP (control) and GFP-Asef2 stable cells were incubated with vehicle (DMSO), wortmannin or LY294002 (50 μM) and then used in live-cell migration assays. Rose plots with individual migration tracks are shown. (B) Quantification of the migration velocity of GFP (control) and GFP-Asef2 stable cells treated with DMSO, wortmannin or LY294002 is shown. Error bars represent s.e.m. for 30-35 cells from four separate experiments (^*P<0.0001). (C) GFP (control) and GFP-Asef2 stable cells were treated with vehicle (DMSO) or 10 nM wortmannin for 2 hours and then cell lysates were assayed for active Rac. (D) Quantification of blots from five separate experiments is shown. Error bar represents s.e.m. (^*P<0.0001). For B and D, asterisks denote statistically significant differences compared with control cells.

Fig. 7.

Akt has a role in Asef2-promoted migration. (A) GFP (control) and GFP-Asef2 stable cells were fixed and co-immunostained for active Akt using a phospho-specific antibody against Thr308, and for actin with Alexa Fluor 647-phalloidin (false-colored purple). (B) Quantification of the normalized fluorescent intensity of active Akt at the leading edge of GFP (control) and GFP-Asef2 stable cells is shown. Error bars represent s.e.m. for 20 cells from three separate experiments (^*P<0.0001). (C) GFP (control) and GFP-Asef2 stable cells were transfected with DN-Akt, KD-Akt or empty vector, and used in live-cell migration assays. Rose plots with individual migration tracks are shown. (D) Quantification of the migration velocity of GFP (control) and GFP-Asef2 stable cells are shown. Error bars represent s.e.m. for 30-35 cells from three separate experiments (^*P<0.0001). Asterisk denotes a statistically significant difference compared with GFP (control) stable cells.

Fig. 8.

Knockdown of endogenous Akt significantly inhibits Asef2-promoted migration. (A) Wild-type HT1080 cells were transfected with empty pSUPER vector, scrambled siRNA (scr siRNA) or AKT siRNAs. Cell lysates were immunoblotted for Akt or α-tubulin (loading control). (B) Quantification of the amount of endogenous Akt in cells transfected with the indicated constructs. Error bars represent s.e.m. from four independent experiments (^*P<0.0003). Asterisks denote a statistically significant difference compared with pSUPER-transfected cells. (C) GFP (control) and GFP-Asef2 stable cells were transfected with AKT siRNAs and used in live-cell migration assays. Rose plots with individual migration tracks are shown. (D) Quantification of the migration velocity of GFP (control) and GFP-Asef2 stable cells transfected with the indicated constructs is shown. Error bars represent s.e.m. for 30-35 cells from three separate experiments (^*P<0.0001). Asterisks denote statistically significant differences compared with GFP (control) stable cells.