|
Figure 1. X-PAK5 belongs to a new PAK subfamily. (A) Multiple sequences alignment (ClustalW) of X-PAK5 NH2-terminal amino acid sequence with that of hsPAK4 (AJ011855), KIAA1264 (BAA86578), hsPAK6 (NP_064553), DmMBT (AJ011578), and CeC45B11 (Z74029); amino acids identity and homology are black and gray boxed, respectively. A putative nuclear localization sequence (NLS), CRIB domain, and potential SH3 binding domains (PXXP) are indicated. (B) Western blot was performed on interphase (inter) or CSF Xenopus egg, XTC cells, and Xenopus tissues extracts using immunopurified Abn122.
|
|
Figure 2. Endogenous X-PAK5 subcellular localization and regulation during cell cycle progression. X-PAK5 subcellular localization was analyzed using immunopurified Abn122 in XTC cells. (A) Subset of X-PAK5 colocalizes with the Golgi apparatus. (B) X-PAK5 costains with actin and tubulin networks. Square regions 1 and 2 were selected on the merged image between X-PAK5 (green), MTs (red), and actin (blue) networks to perform quantitative colocalization. (a) X-PAK5 (green) colocalization in square 1 with microfilaments (blue); white spots represent colocalized voxels. (b) X-PAK5 (green) colocalization in square 1 with MTs (red). Yellow spots represent colocalized voxels. (C) X-PAK5 costains with vimentin and tubulin networks. Square region 3 was selected on the merged image between X-PAK5 (green), MTs (red), and vimentin (blue) networks. (a) X-PAK5 (green) colocalization with vimentin (blue); white spots represent colocalized voxels. (b) X-PAK5 (green) colocalization with MTs (red). Yellow spots represent colocalized voxels. (D) Representative G1 and G2 cells illustrate the more or less filamentous pattern of X-PAK5 during cell cycle progression. Cells were triple stained for X-PAK5, actin, and tubulin in G1 and G2. Arrows show actin-rich structures or single MTs decorated by X-PAK5. During mitosis, cells were triple stained for X-PAK5, DNA, and tubulin.
|
|
Figure 3. X-PAK5 binds MTs and MFs. (A and B) Western blot analysis of MTs copelleting assays. 2 ml equivalent of the pellet fraction (P), the depleted supernatant (S), and the high speed egg extract (T) are loaded in all lanes. Nitrocellulose was probed with b tubulin (E7), Abn122, and cdk2 antibodies. Autoradiography allowed detection of in vitroâtranslated X-PAK5 mutants. (A) Endogenous X-PAK5 associates to taxol stabilized MTs (Taxol) but is not pelleted in NZ-treated extracts (NZ). (B) 35S X-PAK5 wt and 35S X-PAK5 K/R associate to MTs. Binding occurs through the N-Ter regulatory domain (35S X-PAK5 Nter) but not through the C-ter catalytic domain (35S X-PAK5 Cter). (C) Recombinant His6-tagged X-PAK5 (0.1â2 μg) was incubated with MTs (20 μg) polymerized from purified brain tubulin, and MTs were pelleted. Supernatants and pellets were loaded on SDS-PAGE that was Coomassie stained. (D) Actin spin-down binding assays with the in vitroâtranslated X-PAK5 mutants. Most of wt (35S X-PAK5 wt) and kinase dead X-PAK5 (35S X-PAK5 K/R) bind to polymerized MFs. Binding occurs through the N-ter domain, but the C-ter domain also binds microfilaments to some extent. In absence of actin (âactin, bottom) most of the in vitroâtranslated product was recovered in the supernatant fraction. Experiments in Fig. 3 are representative of at least three experiments.
|
|
Figure 9. Activities and subcellular localization of GFPâX-PAK5 mutants. (A) In vitro kinase activity of mock or GFPâX-PAK5â, -K/Râ, -ENâtransfected cells. (Top) Kinase activity was tested against histone H2B substrate. (Bottom) Western blot of the immunoprecipitates using a GFP antibody. (B and C) XTC cells were transfected with GFPâX-PAK5 K/R (B) and GFPâX-PAK5/EN (C) and were costained for actin (blue) and tubulin (red).
|
|
Figure 4. X-PAK5 disorganizes the actin and MTs networks. XTC cells transfected with GFPâX-PAK5 were costained for actin (blue), IFs (red), and tubulin (red). (A) In 51.8% of transfected cells (n = 162), X-PAK5 colocalizes with the disorganized MTs. (B) When X-PAK5 colocalizes with MTs, IFs are collapsed (in 85% of transfected cells, n = 120), whereas only 32% (n = 260) of untransfected cells had a similar phenotype. (C) In 9.8% of transfected cells (n = 162), X-PAK5 colocalizes with actin stress fibers and lamellipodia.
|
|
Figure 5. X-PAK5 rearranges the MT network. XTC cells were transfected with GFPâX-PAK5 and stained with specific antibodies against TyrMTs and GluMTs (A). (B) The R1/R2 ratios in GFP- (left) or GFPâX-PAK5â (middle) transfected cells are plotted against GFP intensity. R1/R2 was below 1 for most GFPâX-PAK5âtransfected cells.
|
|
Figure 6. X-PAK5âbound MTs are stable against dilution-induced depolymerization. GFPâX-PAK5âtransfected XTC cells were permeabilized and immediately fixed (t = 0, top) or incubated in PEM buffer for 30 min (t = 30, bottom) and fixed. GFPâX-PAK5 was costained with TyrMTs.
|
|
Figure 7. X-PAK5 associates to newly nucleating MTs and induces their morphological change. (A) GFPâX-PAK5âtransfected XTC cells were treated with 10 μM NZ for 1 h, washed in fresh medium for 15 min (1 h NZ, 15'Rec), and stained for TyrMTs (top) or MTs (bottom). Merged images are on the right. (B) GFPâX-PAK5âtransfected XTC cells were treated with 10 μM NZ for 2.2 h and washed for 2 h more. Live cell behavior was recorded by time-lapse microscopy (see videos 1 and 2 http://www.jcb.org/cgi/content/full/jcb.200104123/DC1 for the zoomed region). GFPâX-PAK5âbound MTs are seen at 0 and 30 min and 1 and 2 h after NZ treatment and after 0, 8, and 15 min and 2 h recovery. The area surrounding the centrosome is magnified in recovery time points.
|
|
Figure 8. Dynamic of MTs in A6 and H1H10 cells. Elapsed time is indicated at the top left (A) and bottom right (B and C) (min:sec). (A) Dynamic of GFP-labeled MTs in live H1H10 cells. The boxed area is zoomed on the top right; the cross pinpoints the initial position of one MT. Also see video 3. (B) Dynamics of GFPâX-PAK5âbound MTs in live A6 cells. The boxed area is zoomed on the top right. The two crosses pinpoint at two thick GFPâ X-PAK5âbound MTs that essentially stay still and paused. Also see video 4. (C) Dynamic of RFPâX-PAK5âbound GFP MT bundles (the dot pinpoints the initial position of the free end of a bundle) versus unbound GFP MTs (the cross pinpoints the initial position of one MT) were overlaid. Also see video 5; all videos available at http://www.jcb.org/cgi/content/full/jcb.200104123/DC1.
|
|
Figure 10. X-PAK5 kinase is not activated but relocalizes upon GTPase expression. XTC cells were cotransfected with GFPâX-PAK5 and pMT-mycCdc42V12 or -RacV12. (A, top) Kinase activities were assayed over H2B. (A, bottom) Western blot of the immunoprecipitates using a GFP antibodies indicates that equal amount of kinase were immunoprecipitated. (B) Representative myc-tagged active GTPase and GFPâX-PAK5âcoexpressing cells were stained for GFPâX-PAK5 (green), GTPases using myc polyclonal antibody (Red), and tubulin antibodies or actin (blue, respectively). Merged micrographs are on the right. (C) Quantification of cells that present GFPâX-PAK5 binding to the MTs in cells expressing either GFPâX-PAK5 alone or coexpressed with active GTPases. In cells in which GFPâX-PAK5 still binds few MTs (like in B with RacV12), the GTPase is also bound to these few MTs. (D) Western blot analysis of MTs copelleting assays as in Fig. 3. Egg extracts were incubated for 30 min with purified GST or CDC42V12-GST prior to MT pelleting. Cdc42V12GST but not GST prevents most of X-PAK5 binding to the MT fraction (P).
|
