Exogenous Nucleation Sites Fail to Induce Detectable Polymerization of Actin in Living Cells

Most nonmuscle cells are known to maintain a relatively high concentration of unpolymerized actin. To determine how the polymerization of actin is regulated, exogenous nucleation sites, prepared by sonicating fluorescein phalloidin-labeled actin filaments, were microinjected into living Swiss 3T3 and NRK cells. The nucleation sites remained as a cluster for over an hour after microinjection, and caused no detectable change in the phase morphology of the cell. As determined by immunofluorescence specific for endogenous actin and by staining cells with rhodamine phalloidin, the microinjection induced neither an extensive polymerization of endogenous actin off the nucleation sites, nor changes in the distribution of actin filaments. In addition, the extent of actin polymerization, as estimated by integrating the fluorescence intensities of bound rhodamine phalloidin, did not appear to be affected. To determine whether the nucleation sites remained active after microinjection, cells were first injected with nucleation sites and, following a 20-min incubation, microinjected with monomeric rhodamine-labeled actin. The rhodamine-labeled actin became extensively associated with the nucleation sites, suggesting that at least some of the nucleation activity was maintained, and that the endogenous actin behaved in a different manner from the exogenous actin subunits. Similarly, when cells containing nucleation sites were extracted and incubated with rhodamine-labeled actin, the rhodamine-labeled actin became associated with the nucleation sites in a cytochalasin-sensitive manner. These observations suggest that capping and inhibition of nucleation cannot account for the regulation of actin polymerization in living cells. However, the sequestration of monomers probably plays a crucial role. I N nonmuscle cells, close to 50% of total actin molecules appear to be present in the unpolymerized form (Bray and Thomas, 1976; Blikstad et al., 1978). The amount of unpolymerized subunits decreases after the stimulation of polymorphonuclear leukocytes (Fechheimer and Zigmond, 1983; Rao and Varani, 1982), Dictyosteliurn (Condeelis et al., 1988), and platelets (Carlsson et al., 1979; Fox and Philips, 1981), coincident with the appearance of new actin filaments. Thus, it is likely that these subunits may serve as building blocks and become assembled into filaments upon stimulation. However, one intriguing question is how resting cells maintain the relatively high concentration of unpolymerized actin. Based on in vitro measurements (Bonder et al., 1983), purified actin has a critical concentration of ,~0.3/~M under physiological ionic conditions. Assuming that the intracellular concentration of actin is 200 #M, one would expect >99% of actin to be in the filamentous form. The most likely way for maintaining a high concentration of unpolymerized actin is through interactions with various actin binding proteins. Since actin polymerization involves nucleation and subsequent addition of subunits to the nuclei or the ends of filaments, inhibition of polymerization can be achieved in two possible ways. First, monomeric actin can be maintained by binding to a protein that inhibits its polymerization activity. Second, polymerization can be inhibited by a combination of proteins that cap the ends of filaments and proteins that inhibit the formation of active nucleation sites. Thus, even though actin subunits may be active, there are no available sites for the assembly to take place. A wide variety of actin-binding proteins have been identiffed in recent years (for reviews, see Stossel et al., 1985; Pollard and Cooper, 1986). For example, gelsolin is a wellcharacterized protein that caps the barbed ends of actin filaments in a Ca-dependent manner (Yin and Stossel, 1979). Profilin can bind actin monomers and inhibit their polymerization (Carlsson et al., 1977); it may also inhibit the selfnucleation of actin (Pollard and Cooper, 1986). Actobindin is also capable of sequestering actin monomers and inhibiting the formation of nucleation sites (Lambooy and Korn, 1986, 1988). However, serious questions remain concerning the possible roles of these proteins in the regulation of actin polymerization in living cells. For example, a study by Lind et al. (1987) indicates that there may not be enough profilin in platelets to account for the maintenance of unpolymerized actin. In the same study, the extent of gelsolin-actin binding was found to increase, rather than decrease, after platelet activation, contrary to what one might expect if gelsolin were © The Rockefeller University Press, 0021-9525/90/02/359/7 $2.00 The Journal of Cell Biology, Volume 1 I0, February t990 359-365 359 on A uust 7, 2017 jcb.rress.org D ow nladed fom involved in the inhibition of polymerization, Furthermore, since the actin binding activity of both gelsolin and profilin can be inhibited by phosphatidylinositol-4,5-bisphosphate (Janmey and Stossel, 1987; Lassing and Lindberg, 1985), a common component of eukaryotic membranes, it is possible that these proteins might be at least partially inactive in resting cells. In this paper, we attempt to delineate the mechanism ofactin regulation by microinjecting living cells with exogenous nucleation sites. If the regulation of polymerization is achieved primarily by the capping of filaments in conjunction with the inhibition of self nucleation, we might observe an increase in actin polymerization after the microinjection. Our results, however, indicate little or no stimulation of actin polymerization, even though the nucleation sites remain capable of binding exogenous, fluorescently labeled actin subunits. The results are thus more consistent with monomer sequestration being the primary mechanism of regulation. Materials and Methods Preparation of Nucleation Sites and Fluorescently Labeled Actins Muscle actin was purified from rabbit back and leg muscles after Spudich and Watt (1971). In some experiments, the actin was further purified by gel filtration chromatography in a Sephadex G-150 column (Sigma Chemical Co., St. Louis, MO), as described by MacLean-Fletcher and Pollard (1980). Nucleation sites were prepared by sonicating fluorescein phalloidin-labeled filaments of muscle actin. Briefly, G-actin was clarified at 25,000 rpm for 20 min in a rotor (42.2 Ti rotor; Beckman Instruments, Inc., Palo Alto, CA), and polymerized in 2 mM Tris-acetate, pH 6.95, 60 mM KCI, 1.4 mM MgCI2, 0.2 mM ATE 0.1 mM DTT. Fluorescein phalloidin (Molecular Probes Inc., Eugene, OR) was dissolved in a microinjection buffer containing 2 mM Tris-acetate, pH 6.95, 100 mM KCI, 2 mM MgCI2, 0.2 mM ATE 0.1 mM DTT. After clarification, the phalloidin solution was mixed with actin at a phalloidin/actin molar ratio of 0.65:1.0. The mixture was then dialyzed for 2 h against the microinjection buffer and sonicated for 1530 s in a bath sonicator immediately before use. The nucleation sites were microinjected at a concentration of 13.3 #M actin. In some experiments, identical results were obtained with actin polymerized and microinjected in the presence of 2 mM MgCI2 and no KCI. N-(l-pyrenyl) iodoacelamide (Molecular Probes Inc.) labeled actin (pyrene actin) was prepared as described by Kouyama and Mihashi (1981) with minor modifications (Cooper et al., 1983). The dye-to-protein molar ratio was determined to be 0.8. Tetramethylrhodamine iodoacetamide (Molecular Probes Inc.) labeled actin (rhodamine actin) was prepared as described previously (Wang, 1984). The dye-to-protein molar ratio was estimated to be 0.7. Rhodamine actin was microinjected in a buffer of 2 mM Tris-acetate, pH 6.95, 0.05 mM MgCI2, 0.2 mM ATE and 0.1 mM DTT at a concentration of 4.0-5.0 mg/ml.