Co-ordinate regulation of the cytoskeleton in 3T3 cells overexpressing thymosin-beta4

In several cell types, short-term increases in the concentration of the G-actin-sequestering peptide thymosin-beta4 (Tbeta4) cause the disassembly of F-actin bundles. To determine the extent of cell adaptability to these reductions in F-actin, we overexpressed Tbeta4 in NIH 3T3 cells. In cell lines with Tbeta4 levels twice those of vector controls, G-actin increased approximately twofold as expected. However, F-actin did not decrease as in short-term experiments but rather also increased approximately twofold so that the G-F ratio remained constant. Surprisingly, the cytoskeletal proteins myosin IIA, alpha-actinin, and tropomyosin also increased nearly twofold. These increases were specific; DNA, total protein, lactic dehydrogenase, profilin, and actin depolymerizing factor levels were unchanged in the overexpressing cells. The Tbeta4 lines spread more fully and adhered to the dish more strongly than vector controls; this altered phenotype correlated with a twofold increase in talin and alpha5-integrin and a nearly threefold increase in vinculin. Focal adhesions, detected by indirect immunofluorescence with antivinculin, were increased in size over the controls. Northern blotting showed that mRNAs for both beta-actin and vinculin were increased twofold in the overexpressing lines. We conclude that 1) NIH 3T3 cells adapt to increased levels of G-actin sequestered by increased Tbeta4 by increasing their total actin so that the F-actin/G-actin ratio remains constant; 2) these cells coordinately increase several cytoskeletal and adhesion plaque proteins; and 3) at least for actin and vinculin, this regulation is at the transcriptional level. We therefore propose that the proteins of this multimember interacting complex making up the actin-based cytoskeleton, are coordinately regulated by factors that control the expression of several proteins. The mechanism may bear similarities to the control of synthesis of another multimember interacting complex, the myofibril of developing muscle cells.

Cap Z, a calcium insensitive capping protein in resting and activated platelets

Capping of the barbed-ends of actin filaments is an important mechanism for control of the cytoskeleton. In platelets, a valuable model system, it has been thought that gelsolin was the major capping protein. We now report that platelets contain approximately 2 microM Cap Z, a calcium insensitive heterodimeric capping protein; two major and additional minor isoforms of both alpha and beta subunits are present. In lysates from resting platelets 75-80% of the Cap Z sediments with the high speed pellet, but if the platelets are activated with thrombin for 10 s, about 15% of the Cap Z leaves the pellet fraction and is found in the high speed supernatant where it is not bound to actin. This translocation of Cap Z to the supernatant is also observed when resting platelets are lysed into buffer containing 50-100 microM GTP gamma S and 10 mM EGTA. Our results suggest that release of Cap Z from some actin filaments could generate free filament barbed-ends. An increase in free barbed-ends has been reported in platelet lysates prepared shortly after thrombin activation.

Increasing intracellular concentrations of thymosin beta 4 in PtK2 cells: effects on stress fibers, cytokinesis, and cell spreading

Thymosin beta 4 (T beta 4) binds to G-actin in vitro and inhibits actin polymerization. We studied the effects of increasing T beta 4 concentration within living PtK2 cells, comparing its effects on the disassembly of stress fibers and membrane-associated actin with its ability to inhibit cytokinesis and cell spreading after mitosis. We chose PtK2 cells for the study because these cells have many striking actin bundles in both stress fibers and cleavage furrows. They also have prominent concentrations of membrane-associated actin and remain flattened during mitosis. We have found that PtK2 cells contain an endogenous homologue of T beta 4 at a concentration (approximately 28 microM) sufficient to complex a third or more of the cell's unpolymerized actin. Intracellular T beta 4 concentrations were increased by three different methods: 1) microinjection of an RSV vector containing a cDNA for T beta 4; 2) transfection with the same vector; and 3) microinjection of purified T beta 4 protein. The plasmid coding for T beta 4 was microinjected into PtK2 cells together with fluorescently labeled alpha-actinin as a reporter molecule. Immediately after microinjection fluorescently labeled alpha-actinin was detected in a periodic pattern along the stress fibers just as in control cells injected solely with the reporter. However, after 13 h, cells microinjected with reporter and plasmid showed marked disassembly of the fiber bundles. PtK2 cells transfected with this RSV vector for 2-3 days showed disassembly of stress fibers as detected by rhodamine-phalloidin staining; in these cells the membrane actin was also greatly diminished or absent and the border of the cells was markedly retracted. Microinjection of pure T beta 4 protein into interphase PtK2 cells induced disassembly of the stress fibers within 10 min, while membrane actin appeared only somewhat reduced. If the PtK2 cells were mitotic, similar microinjection of pure thymosin beta 4 protein at times from early prophase to metaphase resulted in an unusual pattern of delayed cytokinesis. Furrowing occurred but at a much slower rate than in controls and the amount of actin in the cleavage furrow was greatly reduced. The cells constricted to apparent completion, but after about 30 min the furrow regressed, forming a binucleate cell, much as after treatment with cytochalasin B or D. Postcytokinesis spreading of these T beta 4-injected cells was often inhibited. These experiments suggest that an insufficient number of actin filaments prolongs the contractile phase of cytokinesis and abolishes the final sealing process.

Thymosin beta 4 (T beta 4) in activated platelets

When resting human blood platelets are stimulated with thrombin, 50 to 60% of the G-actin polymerizes to F-actin within 60 seconds. The increase in F-actin is correlated with a corresponding decrease in the complex of G-actin with T beta 4. Within 5 seconds after stimulation, nucleation sites for pyrene actin polymerization increase 1.5 times in Triton-lysed supernatants. Cytochalasin D, known to inhibit the increase in F-actin after thrombin, also inhibits the fall in T beta 4-actin complex and the increase in nucleation sites. Phosphorylation of T beta 4 could not be detected in either control or activated cells. Increased T beta 4 corresponding to that lost from the T beta 4-actin complex is present in lysates from activated platelets and retains the ability to complex with actin. The data, taken together with previous estimates for the dissociation constant of the T beta 4-actin complex, show that actin polymerization following platelet activation could be controlled primarily by the increased availability of free barbed ends of actin filaments which have a higher affinity for G-actin than does T beta 4 and suggest that the increased free T beta 4 may serve to limit the degree of polymerization.

Formation and contraction of a microfilamentous shell in saponin-permeabilized platelets

To study the mechanism of granule centralization in platelets, we permeabilized with saponin in either EGTA (5 mM) or calcium (1 or 10 microM). Under all conditions, platelets retained 40-50% of their total actin and greater than 70% of their actin-binding protein (ABP) but lost greater than 80% of talin and myosin to the supernatant. Thin sections of platelets permeabilized in EGTA showed a microfilament network under the residual plasma membrane and throughout the cytoplasm. Platelets permeabilized in calcium contained a microfilament shell partly separated from the residual membrane. The shell stained brightly for F-actin. A less dense microfilament shell was also seen in sections of ADP-stimulated intact platelets subsequently permeabilized in EGTA. In the presence of 1 mM ATP gamma S and calcium, myosin was retained (70%) and was localized by indirect immunofluorescence in bright central spots that also stained intensely for F-actin. Electron micrographs showed centralized granules surrounded by a closely packed mass of microfilaments much like the structures seen in thrombin-stimulated intact platelets subsequently permeabilized in EGTA. Permeabilization in calcium, ATP, and okadaic acid, produced the same configuration of centralized granules and packed microfilaments; myosin was retained and the myosin regulatory light chain became phosphorylated. Microtubule coil disassembly before permeabilization did not inhibit granule centralization. These results suggest a possible mechanism for granule centralization in these models. The cytoskeletal network first separates from some of its connections to the plasma membrane by a calcium-dependent mechanism not involving ABP proteolysis. Phosphorylated myosin interacts with the microfilaments to contract the shell moving the granules to the platelet's center.

Vinculin in relation to stress fibers in spread platelets

To investigate the function of vinculin in blood platelets, we studied its localization in relation to other cytoskeletal proteins as well as its state of phosphorylation in platelets allowed to spread on fibrinogen-coated surfaces. By 5 minutes after loading the platelets onto the surfaces the 47 and 20 kDa polypeptides became phosphorylated, indicating activation. By 30 minutes, platelets formed small, typical bundles of fibers which stained brilliantly with rhodamine phalloidin. Myosin and tropomyosin, detected with specific antibodies, were localized in periodic arrays along these bundles. By indirect immunofluorescence, a discrete patch of vinculin was observed at each end of every actin-containing bundle. Vinculin phosphorylation was not detected in immunoprecipitates protected against phosphatases. Interference reflection images showed that regions of close binding to the substratum (adhesion plaques) closely matched the vinculin staining sites. Talin appeared diffusely localized. It could be shown to be present in the plaques when platelets were stabilized with ZnCl2 by the method of Geiger and then sonicated to remove some of the surface membrane. Localizations of vinculin and myosin were unaltered by this treatment. Talin phosphorylation or proteolysis could not account for vinculin translocation. We conclude that platelets, in response to an appropriate physiological surface, form typical actin bundles with vinculin at the termination of each bundle, in close relation to adhesion plaques. The signal for this translocation does not appear to depend on phosphorylation of vinculin or on phosphorylation or proteolysis of talin. Our findings support the conclusion that in platelets, as in nucleated cells, vinculin serves as at least part of the connection between bundled actin fibers and the extracellular matrix. Such a connection seems required for platelets' known ability to exert tension on surfaces.

Isolation of a 5-kilodalton actin-sequestering peptide from human blood platelets

Resting human platelets contain approximately 0.3 mM unpolymerized actin. When freshly drawn and washed platelets are treated with saponin, 85-90% of the unpolymerized actin diffuses out. Analysis by polyacrylamide gel electrophoresis under nondenaturing conditions shows that the bulk of this unpolymerized actin migrates with a higher mobility than does pure G-actin, profilactin, or actin-gelsolin complex. When muscle G-actin is added to fresh or boiled saponin extract, the added muscle actin is shifted to the high-mobility form. The saponin extract contains an acidic peptide having a molecular mass in the range of 5 kDa, which has been purified to homogeneity by reverse-phase HPLC. This peptide also shifts muscle actin to the high-mobility form. Addition of either boiled saponin extract or the purified peptide to muscle G-actin also strongly and stoichiometrically inhibits salt-induced polymerization, as assayed by falling-ball viscometry and by sedimentation. We conclude that this peptide binds to the bulk of the unpolymerized actin in platelets and prevents it from polymerizing.