Cell motility and the cytoskeleton最新文献

Previous studies have demonstrated that the penetrating weak base NH4Cl and the anesthetics procaine and urethane disrupt the normal attachment of cortical granules to the cortex of the sea urchin egg. Hylander and Summers (1981: Dev. Biol. 86:1-11) hypothesized that this effect may be caused by a pH-induced polymerization of cortical actin. We have tested this hypothesis by measuring the intracellular pH of eggs of the sea urchins S. purpuratus and A. punctulata treated with NH4Cl, procaine, or urethane, and determining the effects of these agents on the organization of cortical actin. Intracellular pH was determined by the ratiometric measurement of the fluorescent dye BCECF, and filamentous actin organization was examined by confocal laser scanning microscopy of BODIPY-phallocidin stained eggs. Treatment of eggs with either NH4Cl or procaine resulted in a rapid and reversible increase in cytoplasmic pH of up to 1 pH unit and a dose-dependent increase in the intensity of fluorescent staining of the cortex, indicating an increase in the content of filamentous actin. While urethane also induced a dramatic polymerization of cortical actin, no effect on cytoplasmic pH could be detected. These results demonstrate that NH4Cl, procaine and urethane all induce an increase in the amount of filamentous actin in the sea urchin egg cortex that may participate in the detachment of cortical granules. However, these compounds do not share a common mechanism of action based on the elevation of cytoplasmic pH.

Previous studies have demonstrated that overexpression of the carboxyl-terminal fragment, CaD39, of human fibroblast caldesmon in Chinese hamster ovary cells protected endogenous tropomyosin from turnover and stabilized actin microfilament bundles [Warren et al., 1994: J. Cell Biol. 125:359-368]. To assess the consequences of having CaD39-stabilized microfilaments in living cell, we characterized the motile behaviors of stable CaD39-expressing lines. We here found that CaD39-expressing cells adhered faster to plastic, glass, fibronectin-coated glass, and collagen-coated glass than control cells. Moreover, the CaD39-expressing cells also exhibited enhanced spreading immediately after attachment. Despite these differences, overexpression of CaD39 had little effect on the velocity of intracellular granule movement, or the velocity and persistence of cellular translocation. However, CaD39-expressing cells were more elongate and encompassed less area than non-expressing cells during migration in a wound-healing assay. In interphase cells, the expressed CaD39 fragments were found associated with tropomyosin-enriched microfilaments. Like endogenous caldesmon, the CaD39 fragment was also modified at mitosis. Although a significant portion of CaD39 underwent only partial modification, the majority of the CaD39 was released from the microfilaments during mitosis. This is consistent with the finding that the CaD39-induced advantage for attachment and spreading was lost during mitosis. In CaD39-expressing cells, an incomplete release of the CaD39 from microfilaments at mitosis was found which may be responsible for the increase in the frequency of multinuclear cells in CaD39-expressing lines.

Reorientation of the cortical microtubule array is an essential component of cellular development in plants. However, mechanistic details of this process are unknown. The cortical microtubule array of freshly isolated protoplasts (obtained from Nicotiana tabacum BY-2 suspension culture) is relatively random, but upon culturing the cell wall regenerates and the microtubules begin to reorganize. Because cortical microtubules are highly dynamic, we postulated that their reorganization is accomplished solely by the depolymerization of disordered microtubules, followed by repolymerization into an ordered array. This hypothesis was tested on freshly isolated protoplasts using drugs that alter the dynamic status of microtubules by either hyperstabilizing the polymer (taxol); or preventing the addition of subunits to the microtubules (amiprophosmethyl; APM). Microtubule arrays that were hyperstabilized with 10 microM taxol not only reordered, but did so more quickly than untreated cells. Moreover, protoplasts treated with taxol and 20 microM APM also showed accelerated reorganization. Control experiments, performed in vivo and in vitro, confirmed that subunit addition was hindered by APM. Thus, microtubules appear capable of reorienting as relatively intact units. Sodium azide (1 mM) and sodium cyanide (1 mM) can prevent reorientation, indicating that cellular energy is required for this event but this energy is not used by the actin-myosin system because the microfilament-disrupting drug cytochalasin D (50 microM) did not affect reorientation. These results indicate that cortical microtubule array reorganization is a complex process that can involve polymer movement.

A glioma produces some of the most heterogeneously growing, angiogenic, and invasive primary brain tumor cells known. To dissect cellular individuality, and therefore tumor heterogeneity, multiple morphological and molecular processes in single living human glioma cells were measured using multimode light microscopy. Feature extraction of time-lapse image series of spreading, locomoting, and interacting cells either in the presence or absence of physiological modulators was performed by defining five parameters that described cell shape, movement, and cell-cell contacts. Concurrent visualization of all five parameters with a scatterplot matrix revealed temporal as well as time-independent relationships between the parameters that were sufficient to define the individuality of normal and transformed glial cells. Because the actin-cytoskeleton plays a role in regulating the cellular processes described above, the dynamics of a fluorescent analog of non-muscle actin within motile glioma cells were measured in addition to the morphological parameters. The actin-cytoskeleton within the thin sweeping lamellipodia of a glioma exhibited a paucity of large stress fibers, a rich collection of microvillar structures containing actin, and dynamics that were distinct from those of normal motile cells. This approach can therefore potentially be used to dissect the molecular origins of transformation using a small number of representative tumor cells.

We previously showed that axoplasmic organelles from the squid giant axon move toward the barbed ends of actin filaments and that KI-washed organelles separated from soluble proteins by sucrose density fractionation retain a 235-kDa putative myosin. Here, we examine the myosin-like activities of KI-washed organelles after sucrose density fractionation to address the question whether the myosin on these organelles is functional. By electron microscopy KI-washed organelles bound to actin filaments in the absence of ATP but not in its presence. Analysis of organelle-dependent ATPase activity over time and with varying amounts of organelles revealed a basal activity of 350 (range: 315-384) nmoles Pi/mg/min and an actin-activated activity of 774 (range: 560-988) nmoles/mg/min, a higher specific activity than for the other fractions. By video microscopy washed organelles moved in only one direction on actin filaments with a net velocity of 1.11 +/- .03 microns/s and an instantaneous velocity of 1.63 +/- 0.29 microns/s. By immunogold electronmicroscopy, 7% of KI-washed organelles were decorated with an anti-myosin antibody as compared to 0.5% with non-immune serum. Thus, some axoplasmic organelles have a tightly associated myosin-like activity.

Locomoting blebbing cells have been used as a model to obtain novel insight into the mechanisms of cell locomotion. We tested the hypothesis that locomotion can be due to progressive one-sided protrusion of cellular volume into pseudopods. The hypothesis is supported by the finding that the rate and direction of locomotion of individual Walker carcinosarcoma cells can be predicted by sequential measurement of protrusive activity. Protrusive activity at the front is closely associated with forward movement of the rear part of the cell. During bleb formation the cell membrane of Walker carcinosarcoma cells is pushed forward faster (1.2-4.1 microns/sec) than known rates of actin elongation.

Centrin, a approximately or equal to 20 kDa calcium-binding protein also known as caltractin, is a component of centrosome-associated algal flagellar roots capable of calcium-mediated contraction, and is also found in the centrosomes of vertebrate cells. Our analysis of a centrin gene from a protist, the amoeboflagellate Naegleria gruberi, reveals conserved features that distinguish centrins from calmodulin. Antibodies to bacterially expressed Naegleria centrin, which also recognize yeast Cdc31p, were employed to localize centrin immunoreactivity in selected organisms possessing specialized microtubule-organizing centers (MTOCs) or accessory structures. There is a striking morphological diversity of such structures. In the simplest associations, as found in Naegleria flagellates and vertebrates tracheal epithelium, centrin is intimately associated with the cylinder of the basal bodies. In cells with unfocused mitotic spindles, Naegleria amoebae and onion root tips, no localization of centrin was detected. In Dictyostelium discoideum and Saccharomyces cerevisiae, which lack centrioles, centrin immunoreactivity was observed as punctate cytoplasmic bodies but not associated with spindle pole MTOCs. In Paramecium multimicronucleatum, centrin immunoreactivity is localized to the infraciliary lattice, previously shown to exhibit calcium-mediated contraction. In Vorticella microstoma, known for the calcium-induced rapid contraction of its stalk, centrin immunoreactivity is localized to the contractile spasmoneme and myonemes. Similar antigens from Paramecium and Vorticella are detected by anti-centrin and anti-spasmin. The pattern of localization of centrin immunoreactivity supports the conjecture that a contractile system involving centrin, initially associated with centriolar structures, was recruited during evolution to build specialized organelles in different organisms and cell types.

Whereas actin-binding proteins (ABPs) regulate network formation during the cell cycle, it is not known whether ABPs also function to sequester or target isoactins to specific subcellular compartments. Recently, we have shown that ezrin indirectly associates with beta, but not alpha actin filaments in a calcium- and cytochalasin-sensitive manner [Shuster and Herman, 1995: J. Cell Biol. 128:837-848]. To identify the beta actin-specific binding protein that fosters ezrin-beta actin interactions, we developed an isoactin affinity fractionation and F-isoactin overlay/Western blotting technique. Results reveal that a 73 kd polypeptide that co-precipitates with ezrin and beta actin [Shuster and Herman, 1995: J. Cell Biol. 128:837-848] can also binds directly to filaments of beta, but not alpha actin by isoactin overlay. In an effort to establish whether p73 plays a role in regulating beta actin dynamics in cells, we produced monoclonal antibodies by immunizing BALB/c mice with p73-containing lamellar lysates or high salt elutions from beta actin affinity columns. Two monoclonal antibodies were cloned that react with p73 present in fractions released from beta actin Sepharose-4B or purified to homogeneity by DEAE chromatography. Anti-p73 Western blots reveal that there is a 16-fold difference in p73 binding to beta actin vs. alpha actin affinity columns when experiments are performed in physiological salts. To characterize p73-beta actin binding in vitro and establish whether p73 binds along the lengths or at the barbed end of the beta actin filament, we asked whether cytochalasin D (CD) could displace p73 pre-bound to beta actin-Sepharose 4B. Anti-p73 Western blotting reveals that nanomolar concentrations of CD are capable of selectively eluting p73 and ezrin from beta actin Sepharose 4B, indicating that p73 binds beta actin via the barbed end. Simultaneous double antibody localization studies using anti-beta actin IgG and anti-p73 IgM reveal that p73 and beta actin are co-localized in the forward aspects of motile cytoplasmic domains, in close proximity to the plasma membrane. Because of its isoform-specific interactions with the barbed end of beta actin filaments, we have named this molecule beta cap73. These results indicate that isoform-specific actin-binding proteins can be identified from cortical cytoplasm, and suggest that beta cap73 may not only act to spatially regulate the intracellular distribution of isoactins, but may also facilitate forward protrusion formation through the regulated release of free filament ends during cell motility.

The regulation of microtubule dynamics in vitro by microtubule-associated proteins (MAPs) was examined, using purified porcine MAP1B and MAP2. MAP1B has a significantly smaller effect on the observed critical concentration for microtubule assembly than MAP2. Assembly is faster in the presence of either MAP, and the resulting microtubules are shorter, indicating that nucleation is substantially promoted by the MAPs. Both MAPs stabilise the microtubule lattice as observed from podophyllotoxin-induced disassembly, but the effect of MAP1B is weaker than the effect of MAP2. At steady-state of assembly MAP1B still allows microtubule dynamic instability to occur as inferred from microtubule length changes. The comparison of the effects of MAP1B and MAP2 indicates that the reduction of the observed critical concentration is attributable to the reduction of the depolymerisation rate and correlates with the extent of suppression of dynamic instability. Numerical simulations illustrate that microtubule dynamics are strongly influenced by relatively small changes in the strength of a limited subset of subunit interactions in the lattice. The observed characteristic differences between the MAPs may be important for the regulation of distinct populations of microtubules which coexist in the same cell, where differences in stability and dynamics may be essential for their different spatial roles as, for example, in developing neurons.