Actin Binding to the BAR Domain and Arf GAP Activity of ASAP1 Coordinately Control Actin Stress Fibers and Focal Adhesions

FIGURE 2. Changes in both focal adhesions (FAs) and stress fibers (SFs) are common to four cell lines treated with siRNA targeting ASAP1. The indicated cells were treated with either nontargeting (control) or ASAP1‐targeting siRNA for 3 days. The cells were harvested, ASAP1 levels determined (Panel B) and plated on fibronectin‐coated cover slips for 5–6 h. The cells were then fixed and prepared for immunofluorescence microscopy staining for actin and paxillin. (A) Representative images. Arrows point to examples of SFs and arrowheads point to examples of FAs. (B) Quantifications of bundled actin and FAs. Bundled actin was determined as described in “Section 2.” Briefly, profiles of two orthogonal sets of nine lines in two ∼80 pixel boxes/cell for 10–20 cells were determined. The number of peaks for each set of orthogonal lines was determined and the maximum for a region of interest is reported. Results from one of two experiments with similar results are presented. Cellular FAs were quantified as previously described (Chen and Kroog 2010). The summary of three experiments for C2C12, NIH3T3, and HeLa cells and of two experiments for U2OS are shown. The inset is an immunoblot for ASAP1 in…

FIGURE 3. Effect of knockdown of endogenous ASAP1 on focal adhesions (FAs) and stress fibers (SFs) is rescued by expression of ASAP1 from an ectopic promoter. U2OS cells, with cDNA expression cassettes for ASAP1 under control of a tetracycline promoter integrated into the chromosome using a lentivirus vector, were plated in 35 mm cell culture plates and treated with either nontargeting (control Si) or siRNA targeting the 3′UTR of ASAP1 (ASAP1 siRNA) as indicated. After 24 h, doxycycline (0.5 µg/mL) was added as indicated. Two days later, cells were harvested and replated on fibronectin coated cover slips for 6 h in serum free medium. The cells were then fixed and stained for actin and paxillin. (A) Immunoblot of cell lysates for ASAP1 under the indicated conditions. The top panel is a blot using an antibody to ASAP1, the middle panel uses an antibody to the epitope tag fused to the ectopically expressed ASAP1. Note there is some expression without doxycycline when endogenous ASAP1 is knockdown and expression with doxycycline is greater with reduced endogenous ASAP1. This observation is reproducible and is the subject of another project. β‐Actin was used as a loading control. (B)…

FIGURE 4. Reduced Arf5 expression affects stress fibers (SFs) and focal adhesions (FAs). U2OS cells were treated with either a pool of siRNA or a single siRNA targeting Arf5. After 3 days, cells were harvested. Aliquots were lysed and immunoblotted. Cells were plated on fibronectin coated coverslips for 5–6 h, fixed, stained for actin and paxillin, and imaged. (A) Immunoblot of lysates of U2OS cells treated with the indicated siRNA for the reduction of Arf5 expression. Specificity of knockdown was checked by blotting for Arf6. β‐Actin was used as a loading control. (B) Quantification of the Immunoblot. Blots were quantified using fluorescent secondary antibodies detected using an Odyssey Imager (LiCOR). (C) Effect of Arf5 knockdown on FAs in U2OS. FAs were quantified as described in Figure 2 and in “Section 2.” (D) Effect of Arf5 knockdown on cellular distribution of FAs. The distance of FAs (>1 µm2) from the cell edge was determined as described in “Section 2.” The density of paxillin containing plaques at relative distances from the edge of the cell is plotted, 0 being the edge of the cell. Thirty‐eight cells for the mock condition, 40 cells for the control siRNA condition and 31 cells…

FIGURE 5. Both GTP binding and hydrolysis by Arf5 are required for the maintenance of focal adhesions (FAs) and stress fibers (SFs). (A) Representative images of U2OS cells expressing the indicated mutant of Arf5. Cells were transfected with plasmids for expression of the indicated Arf5 mutant. Control cells were treated with transfection reagent and an empty expression vector. After 24 h, the cells were replated on fibronectin coated coverslips for 6 h, fixed and stained for actin, paxillin, or the ectopic Arf5 (identified by the HA tag fused to Arf5). Arrowheads indicate paxillin containing plaques at the edge of the cell. (B) Distribution of FAs in U2OS cells expressing Arf5 mutants. The density of paxillin containing plaques at relative distance from the edge of the cell is plotted, 0 being the edge of the cell. 36, 36, 36, and 38 cells were analyzed for control, Arf5‐HA, [T31N]Arf5‐HA, and [I46D]Arf5‐HA. (C) Effect of expressing Arf5 and mutants of Arf5 on number and size of FAs in U2OS cells. The number and size of paxillin‐containing plaques in U2OS cells expressing the indicated Arf5 mutants were determined as described in (Chen and Kroog, 2010) and in “Section 2.” (D) Effect of…

FIGURE 6. Effect of ASAP1 on actin cytoskeleton and focal adhesions (FAs) requires GAP activity and the BAR domain. Lentivirus was used to integrate cDNA with the open reading frame for the indicated mutants of ASAP1 ([R485K]ASAP1‐HA or [ΔBAR]ASAP1‐HA) under control of a tet‐on promoter. Each cell line was treated with either nontargeting (control) or siRNA targeting the 3′UTR of ASAP1 (ASAP1 SiRNA) for a day, and then treated with either doxycycline (0.5 µg/mL) or vehicle for an addition 2 days before plating on fibronectin coated covers for 6 h, fixing, staining for actin and paxillin, and imaging. (A) Representative images. Examples of stress fibers (SFs) are indicated by arrows, FAs by arrowheads. (B) Effect of reduced ASAP1 and expression of [R485K]ASAP1‐HA on SFs and FAs. (i) Bundled actin, quantified by determining the number of peaks from line profiles, was taken as a measure of SFs. The summary of two experiments, analyzing 10–20 cells/condition/experiment, is shown. (ii) FAs. Summary data from two experiments for [R485K]ASAP1 and three experiments for [ΔBAR]ASAP1 is presented. (iii) Protein levels determined by immunoblotting. Total ASAP1 (includes endogenous and the mutants) was…

FIGURE 7. BAR and GAP domains must be in the same polypeptide to support formation of focal adhesions (FAs) and stress fibers (SFs). Lentivirus was used to integrate cDNA with open reading frames for both [ΔBAR]ASAP1‐HA and [R485K]ASAP1‐myc under control of a tet‐on promoter into the chromosome of U2OS cells. Cells were treated with either nontargeting siRNA (control) or siRNA targeting the 3′UTR of ASAP1 (ASAP1 siRNA) for 1 day before addition to vehicle or 0.5 µg/mL doxycycline for 2 additional days before plating on fibronectin‐coated coverslips for 6 h, fixing and staining for actin and paxillin. (A) Representative images. Examples of SFs are indicated with arrows, examples of FAs are indicated with arrowheads. (B) Bundled actin. Quantification from two experiments, with 10–20 cells/condition/experiment of the number of peaks from line profiles was determined as described in “Section 2.” (C) FAs in cells expressing mutant ASAP1. Summarized data from two experiments is shown. (D) Protein levels in U2OS cells. Immunoblot was used to assess the expression of the mutant ASAP1. Proteins were detected with an antibody against a C‐terminal epitope of ASAP1, which would detect both endogenous…

FIGURE 8. Examples of possible parallel and integrated mechanisms for ASAP1 control of stress fiber (SF) dynamics. (A) Parallel function of the BAR and GAP domains. If reduction of Arf•GTP levels increases actin polymerization, as described for Git (Nishiya et al. 2005), then Arf GAP activity would promote the formation of actin filaments. In parallel, the BAR domain promotes actin bundling necessary for the formation of stress fiber precursors that are subsequently assembled into stress fibers. (B) Integrated activity of the BAR and GAP domains. Based on the findings that GTP binding and hydrolysis are necessary for SFs, that both actin binding and GAP activity are necessary and that the two activities be in the same polypeptide, we propose ASAP1 regulates SFs by a mechanism that integrates the two activities. One possible mechanism is illustrated. In Step 1, ASAP1 bundles actin. In Step 2, Arf•GTP binds to ASAP1. The complex of actin/ASAP1/NM2 then fuses end‐to‐end with another stress fiber precursor containing α‐actinin. Hydrolysis of GTP on Arf results in dissociation of ASAP1 from the complex. Note that this is one possibility of many but might be a reasonable starting hypothesis for…