Cell motility: proline-rich proteins promote protrusions.

Many proline-rich proteins participate in delivering actin monomers to specific cellular locations where actin-rich membrane protrusions, such as ruffles, filopodia and microspikes, are formed. These protrusions are necessary for cell motility. Actin monomer is usually delivered to the site of polymerization in the form of profilactin - a complex of G-actin with a polyproline-binding protein, profilin. Here, we describe proline-rich proteins involved in regulating actin polymerization and classify them according to their role in recruiting profilin to the membrane.

Effects of latrunculin reveal requirements for the actin cytoskeleton during secretion from mast cells.

To investigate the role of the actin cytoskeleton in exocytosis, we have tested the effects of latrunculin B, a microfilament-disrupting drug, on secretion from intact and permeabilised rat peritoneal mast cells. The toxin strongly inhibited secretion from intact cells (attached or in suspension) responding to a polybasic agonist, compound 48/80. However, this effect was revealed only after a profound depletion of actin filaments. This was achieved by a long (1 h) exposure of cells to the drug before activation, together with its presence during activation. Maximal inhibition of secretion by such treatment was 85% at 40 microgram/ml latrunculin B. These results indicate that minimal actin structures are essential for the exocytotic response. In contrast, stimulus-induced cell spreading was prevented by latrunculin (5 microgram/ml) applied either before or after activation. The effects of the toxin on intact cells were fully reversible. The responses of permeabilised cells were affected differentially: secretion induced by calcium was more sensitive to latrunculin than that induced by GTP-gamma-S. The calcium response, therefore, is more dependent upon the integrity of the actin cytoskeleton than the response induced by GTP-gamma-S. Again, maximal inhibitory effects (approximately 65 and 25% at 40 microgram/ml) were observed only when cells were exposed to the toxin both before and after permeabilisation. Since the permeabilised cells system focuses on the final steps of exocytosis, the incomplete inhibition suggests that actin plays a modulatory rather than a central role at this stage.

Calmodulin regulates the disassembly of cortical F-actin in mast cells but is not required for secretion.

Secretion is dependent on a rise in cytosolic Ca(2+)concentration and is associated with dramatic changes in actin organization. The actin cortex may act as a barrier between secretory vesicles and plasma membrane. Thus, disassembly of this cortex should precede late steps of exocytosis. Here we investigate regulation of both the actin cytoskeleton and secretion by calmodulin. Ca(2+), together with ATP, induces cortical F-actin disassembly in permeabilized rat peritoneal mast cells. This effect is strongly inhibited by removing endogenous calmodulin (using calmodulin inhibitory peptides), and increased by exogenous calmodulin. Neither treatment, however, affects secretion. Low concentrations ( approximately 1 microM) of a specific inhibitor of myosin light chain kinase, ML-7, prevent F-actin disassembly, but not secretion. In contrast, a myosin inhibitor affecting both conventional and unconventional myosins, BDM, decreases cortical disassembly as well as secretion. Observations of fluorescein-calmodulin, introduced into permeabilized cells, confirmed a strong (Ca(2+)-independent) association of calmodulin with the actin cortex. In addition, fluorescein-calmodulin enters the nuclei in a Ca(2+)-dependent manner. In conclusion, calmodulin promotes myosin II-based contraction of the membrane cytoskeleton, which is a prerequisite for its disassembly. The late steps of exocytosis, however, require neither calmodulin nor cortical F-actin disassembly, but may be modulated by unconventional myosin(s).

Rho controls cortical F-actin disassembly in addition to, but independently of, secretion in mast cells.

Localized disassembly of cortical F-actin has long been considered necessary for facilitation of exocytosis. Exposure of permeabilized mast cells to calcium/ATP induces cortical F-actin disassembly (calmodulin-dependent) and secretion (calmodulin-independent). The delay in the onset of secretion is characteristic for the calcium/ATP response and is abolished by GTP. Here we report that a constitutively active mutant of Rho (V14RhoA) enhanced both secretion and cortical F-actin disassembly. In addition, V14RhoA mimicked GTP by abolishing the delay in secretion. Inhibition of Rho by C3 transferase prevented both secretion ( approximately 80%) and F-actin disassembly (approximately 20%). Thus, both Rho GTPase and calcium/calmodulin contribute to the control of cortical F-actin disassembly. Stabilization of actin filaments by high concentrations of phalloidin or by a calmodulin-inhibitory peptide (based on the calmodulin-binding domain of myosin light chain kinase) did not affect the extent of secretion or the secretion-enhancing effects of V14RhoA. These results further support the existence of divergent, Rho-dependent, pathways regulating actin and exocytosis. Furthermore, compound Y-27632, a specific inhibitor of Rho-associated protein kinase (p160(ROCK)), attenuated the Rho-induced loss of cortical F-actin without affecting secretion. A model is presented in which Rho regulates secretion and cortical F-actin in a manner dependent on and/or synergistic with calcium.

The small GTP-binding proteins, Rac and Rho, regulate cytoskeletal organization and exocytosis in mast cells by parallel pathways.

In mast cells, activation of GTP-binding proteins induces centripetal reorganization of actin filaments. This effect is due to disassembly, relocalization, and polymerization of F-actin and is dependent on two small GTPases, Rac and Rho. Activities of Rac and Rho are also essential for the secretory function of mast cells. In response to GTP-gamma-S and/or calcium, only a proportion of permeabilized mast cells is capable of secretory response. Here, we have compared actin organization of secreting and nonsecreting cell populations. We show that the cytoskeletal and secretory responses are strongly correlated, indicating a common upstream regulator of the two functions. The secreting cell population preferentially displays both relocalization and polymerization of actin. However, when actin relocalization or polymerization is inhibited by phalloidin or cytochalasin, respectively, secretion is unaffected. Moreover, the ability of the constitutively active mutants of Rac and Rho to enhance secretion is also unaffected in the presence of cytochalasin. Therefore, Rac and Rho control these two functions by divergent, parallel signaling pathways. Cortical actin disassembly occurs in both secreting and nonsecreting populations and does not, by itself, induce exocytosis. A model for the control of exocytosis is proposed that includes at least four GTP-binding proteins and suggests the presence of both shared and divergent signaling pathways from Rac and Rho.

Secretion from permeabilised mast cells is enhanced by addition of gelsolin: contrasting effects of endogenous gelsolin.

Permeabilised rat mast cells were exposed to gelsolin and its N-terminal half (S1-3), proteins that sever actin filaments in a calcium-dependent and independent manner, respectively. Gelsolin and S1-3 induced a decrease in cellular F-actin content and an increase in the extent of the secretory response. The calcium sensitivities of both these effects were consistent with the differential calcium requirements of the two proteins. Segment 1 (S1), which binds G-actin and caps filaments but does not sever them, did not show these effects. Thus, secretion of mast cells is promoted as a consequence of the severing activity of exogenous gelsolin or S1-3. Most of the endogenous gelsolin remained within permeabilised, washed mast cells and its distribution in resting state was predominantly cortical. Addition of calcium in the absence of MgATP did not reduce the F-actin content; by contrast, calcium with MgATP induced F-actin loss that was unaffected by the presence of anti-gelsolin. Because this antibody inhibits the severing activity of gelsolin, these results indicate that in permeabilised mast cells the severing activity of the remaining endogenous gelsolin is not involved in cortical actin filaments disassembly. Upon exposure to GTP-gamma-S in the absence of calcium, the content of cortical gelsolin was reduced. This parallels our previous observation of a GTP-gamma-S induced reduction of cortical actin filaments followed by their relocation to the cell's interior (Norman et al. (1994) J. Cell Biol. 126, 1005-1015) and suggests that actin redistribution may be a consequence of dissociation of gelsolin caps brought about by activation of a GTP-binding protein.

The small GTPases Rac and Rho as regulators of secretion in mast cells.

BACKGROUND: Regulated secretion by mast cells is known to be controlled by GTP-binding proteins, but the proteins involved have not been identified. Rac and Rho, two small GTPases related to the oncoprotein Ras, mediate transmission of signals from cell-surface receptors to the actin cytoskeleton. In rat mast cells, both Rac and Rho participate in effecting the centripetal redistribution of filamentous actin that is observed after stimulation of the cells. Rho is responsible for polymerization of actin filaments in the cell interior, whereas Rac is required for the entrapment in the interior of filaments released from the cortex. Such cytoskeletal changes could be important in control of the exocytotic process, so we examined whether Rac and Rho also play a role in regulated secretion by mast cells. RESULTS: We show that the constitutively active mutant proteins, V14RhoA and V12Rac1, enhance regulated secretion from permeabilized mast cells by increasing the proportion of cells that are competent to respond to stimulation. In addition, inhibition of endogenous Rac and Rho activity using inhibitors, N17Rac1 and C3 transferase, respectively, reduces the secretory response of mast cells to stimuli. CONCLUSION: These results provide direct evidence that, in mast cells, both Rac and Rho are components of the signalling pathway that leads to secretion.

Actin filament organization in activated mast cells is regulated by heterotrimeric and small GTP-binding proteins.

Rat peritoneal mast cells, both intact and permeabilized, have been used widely as model secretory cells. GTP-binding proteins and calcium play a major role in controlling their secretory response. Here we have examined changes in the organization of actin filaments in intact mast cells after activation by compound 48/80, and in permeabilized cells after direct activation of GTP-binding proteins by GTP-gamma-S. In both cases, a centripetal redistribution of cellular F-actin was observed: the content of F-actin was reduced in the cortical region and increased in the cell interior. The overall F-actin content was increased. Using permeabilized cells, we show that AIF4-, an activator of heterotrimeric G proteins, induces the disassembly of F-actin at the cortex, while the appearance of actin filaments in the interior of the cell is dependent on two small GTPases, rho and rac. Rho was found to be responsible for de novo actin polymerization, presumably from a membrane-bound monomeric pool, while rac was required for an entrapment of the released cortical filaments. Thus, a heterotrimeric G-protein and the small GTPases, rho and rac, participate in affecting the changes in the actin cytoskeleton observed after activation of mast cells.

Calcium and GTP-gamma-S as single effectors of secretion from permeabilized rat mast cells: requirements for ATP.

Treatment with metabolic inhibitors and addition of exogenous MgATP exerted different effects on secretion from streptolysin-O-permeabilized mast cells, responding to calcium and GTP-gamma-S as single effectors (i.e., independently of each other). These effects were also strongly dependent on the experimental conditions. Thus cells, triggered by Ca2+ at the time of permeabilization, did not require MgATP, but after metabolic inhibition rapidly became absolutely dependent on its provision, requiring high (> mM) concentrations. AMP-PNP was not effective. After longer treatment with metabolic inhibitors, the absolute dependence on MgATP was also exhibited by cells responding to dual effectors (i.e., Ca2+ and GTP-gamma-S applied together). In contrast, calcium independent secretion due to GTP-gamma-S was more resistant to metabolic inhibition, exhibiting no absolute requirements for MgATP. Once the responsiveness to GTP-gamma-S had been lost, it could not be restored by addition of MgATP. MgATP, in fact, inhibited the response of permeabilized cells to GTP-gamma-S. This effect could be mimicked by AMP-PNP. When permeabilized cells were washed before triggering, MgATP (0.1-1 mM concentration range) was no longer inhibitory but stimulatory. These differences between Ca(2+)- and GTP-gamma-S-induced responses indicate that ATP utilization is essential to the calcium, but not to the guanine nucleotide, pathway to secretion. The rate of the response to calcium/MgATP was much slower in the absence than in the presence of GTP-gamma-S. The onset of secretion occurred after an initial delay. This lag phase was abolished by addition of GTP-gamma-S, suggesting that a GTP-binding protein may control a reaction which constitutes a rate-limiting step in the secretory process.

Calcium-induced secretion from permeabilized rat mast cells: requirements for guanine nucleotides.

Cells used in this work were permeabilized by streptolysin-O and then washed to remove freely soluble components. The secretory responsiveness of these cells to various combinations of calcium, MgATP and guanine nucleotide was characterized and in most respects was found to be similar to that of the metabolically inhibited (unwashed) cells. The content of adenosine and guanine nucleotides, which remain within the permeabilized cells after washing, was estimated as 0.83 and 0.12 mM (extrapolated to intact cells), which constitutes 18 and 25%, respectively, of the total nucleotide content of mast cells. High (> mM) concentrations of MgATP, required for the calcium-induced secretion, were reduced to microM levels by suboptimal concentrations of GTP, which also markedly increased both the rate and extent of the response. Similarly, microM concentrations of MgATP reduced the requirements of the calcium-dependent secretion for GTP. The synergy of the GTP and ATP effects suggests that, together, the two nucleotides can maintain a pool of free GTP, presumably as a result of transphosphorylation from ATP to GDP. Thus, MgATP may work by transphosphorylating the endogenous GDP. However, neither GTP nor GTP-gamma-S were effective as substitutes for MgATP in the calcium-induced secretion, particularly that from metabolically inhibited cells. This indicates that MgATP does not act simply by providing GTP but is needed to maintain a phosphorylated state of the system. The synergistic effects of ATP and GTP were observed only in the presence of calcium. To test whether calcium/MgATP-induced secretion requires an activated G protein, the effects of G-protein inactivators were studied. GDP, deoxy GDP and GDP-beta-S exerted differing degrees of inhibition on secretory responses induced by various combinations of effectors. The response to calcium/MgATP was less sensitive to these inhibitors than that to GTP-gamma-S (with or without calcium). However, all three 'inhibitors' were also capable of stimulating calcium/MgATP-dependent secretion, indicating a transphosphorylation, producing GTP, dGTP and GTP-beta-S. Thus, in the absence of any specific inhibitors for either G proteins or the transphosphorylation reaction, the degree of dependence of the calcium-induced secretion on a G protein remains unclear.

Intracellular mechanisms regulating exocytotic secretion in mast cells.

The release of inflammatory mediators from mast cells occurs by a regulated exocytotic process. We have been able to study the intracellular events in this pathway by permeabilizing the plasma membrane of rat peritoneal mast cells and stimulating exocytosis by providing both Ca2+ and a guanine nucleotide. By this approach we have obtained evidence for the participation of at least two guanine nucleotide binding proteins in the control of exocytosis. We have also shown that ATP is unnecessary for the final events, but that it does have a number of modulatory functions, for instance in the control of the effective affinity of the proteins that bind Ca2+ and GTP. There is also evidence for a protein dephosphorylation in the later stages of the control pathway.

Changes in the state of actin during the exocytotic reaction of permeabilized rat mast cells.

The major part of mast cell actin is Triton-soluble and behaves as a monomer in the DNase I inhibition assay. Thus, actin exists predominantly in monomeric or short filament form, through filamentous actin is clearly apparent in the cortical region after rhodamine-phalloidin (RP) staining. The minimum actin content is estimated to be approximately 2.5 micrograms/10(6) cells (cytosolic concentration approximately 110 microM. After permeabilization of mast cells by the bacterial cytolysin streptolysin-O, approximately 60% of the Triton-soluble actin leaks out within 10 min. However, the staining of the cortical region by RP remains undiminished, and the cells are still capable of exocytosis when stimulated by GTP-gamma-S together with Ca2+. In the presence of cytochalasin E the requirement for Ca2+ is decreased, indicating that disassembly of the cytoskeleton may be a prerequisite for exocytosis. This disassembly is likely to be controlled by Ca2(+)-dependent actin regulatory proteins; their presence is indicated by a Ca2(+)-dependent inhibition of polymerization of extraneous pyrene-G-actin by a Triton extract of mast cells. The effect of cytochalasin E on secretion is similar to that of phorbol myristate acetate, an activator of protein kinase C; both agents enhance the apparent affinity for Ca2+ and cause variable extents of Ca2(+)-independent secretion. Exposing the permeabilized cells to increasing concentrations of Ca2+ caused a progressive decrease in F-actin levels as measured by flow cytometry of RP-stained cells. In this respect, both cytochalasin E and phorbol ester mimicked the effects of calcium. GTP-gamma-S was not required for the Ca2(+)-dependent cortical disassembly. Thus, since conditions have not yet been identified where secretion can occur in its absence, cortical disassembly may be essential (though it is not sufficient) for exocytosis to occur.

The role and mechanism of the GTP-binding protein GE in the control of regulated exocytosis.

Recent advances in the understanding of the terminal stages of the pathway of regulated secretion (exocytosis) have depended in large part on the use of permeabilized secretory cells in which the cytosol is directly accessible to manipulation from the outside. As well as Ca2+, a role for GTP, and hence a GTP-binding protein (GE), is plainly apparent in the exocytotic mechanism of some cells. In metabolically depleted and permeabilized mast cells, the combination of Ca2+ and a guanine nucleotide are a sufficient stimulus for the release of the total cell content of histamine and hexosaminidase. ATP is not required and so a phosphorylation reaction does not comprise a necessary step in the terminal pathway. Phosphorylating nucleotides, however, can modulate the exocytotic reaction in a manner that suggests that the reverse reaction, namely protein dephosphorylation, might be an enabling step. Support for this idea is derived from experiments in which the cells are conditioned to become responsive to Ca2+ alone; in these circumstances the dependence of secretion on the presence of a guanine nucleotide can be reimposed by okadaic acid, an inhibitor of protein phosphatases.

Soluble proteins as modulators of the exocytotic reaction of permeabilised rat mast cells.

This study addresses the question of the role of cytoplasmic proteins in exocytosis from permeabilised rat mast cells. We have used two different methods of cell permeabilisation (ATP4- and streptolysin O) to regulate the size of the plasma membrane lesions, and thus to dictate the rate and extent of efflux of the cytosolic proteins, and compared the secretory response of the two preparations. We report evidence for the existence of two factors present in the cytosol, which affect the exocytotic mechanism in opposing manners. One of these is required for the maintenance of cell responsiveness; it is retained for more than 120 min by ATP4- -permeabilised cells but lost within 60 min from cells permeabilised by streptolysin O. The other factor, which leaks immediately from cells treated from streptolysin O, but only gradually from cells treated with ATP4-, has the effect of suppressing the affinity for both Ca2+ and guanine nucleotide in the exocytotic reaction.

Enhanced number of actin binding sites on plasma membranes of polyoma virus-transformed fibroblasts.

Plasma membranes, isolated from normal (C13) and polyoma virus-transformed (J1) cultured BHK cells were incubated with G-actin under polymerizing conditions, followed by a low-speed centrifugation. The amount of actin attached to the pelleted BHK-J1 plasma membranes was at least twice that on BHK-C13 membranes, indicating a greater number of actin attachment sites on the former. This result was confirmed by the observation that the plasma membranes from the transformed cells were also more active in nucleating polymerization of pyrene-labelled actin. Most of the actin attachment sites could be solubilized by Triton or low-salt extraction treatment.

Identification of two species of actin depolymerizing factor in cultures of BHK cells.

High-speed supernatant obtained from the lysate of cultured BHK cells has been chromatographed on Sepharose-4B, DEAE-cellulose and hydroxyapatite columns, and a fraction has been identified with characteristics similar to an actin depolymerizing factor (ADF), a small protein previously isolated from embryonic chick brain. Using a rabbit antibody against the chick brain protein, two immunoreactive forms were identified: a 19 kDa form co-migrating in SDS-polyacrylamide gels with embryonic chick brain ADF, and a 20 kDa form. The two species could be separated on a hydroxyapatite or green A dye matrix columns and only the 20 kDa protein was active when assayed for effects on pyrene-G-actin assembly. It enhanced the rate of F-actin assembly, but only after an initial lag phase, and decreased the final proportion of actin in filamentous form. These effects were calcium-independent. Actin depolymerizing factor constituted at least 0.5% of the total protein in the cytoplasmic fraction. A Triton extract of plasma membrane-enriched fraction from BHK cells was fractionated on a Sepharose-4B column and again, a fraction was found which had an ADF-like activity and also contained the two immuno-cross-reactive forms, 19 kDa and 20 kDa. These results suggest a novel regulation of the microfilament system in eukaryotic cells via the control of the ADF activity.

Activation of myosin ATPase by actin isolated from cultured BHK cells and the effect of gelsolin.

Activation of skeletal muscle myosin and myosin subfragment-1 (S1) by actin purified from the cytoplasm of cultured BHK cells was studied using the fluorescence of pyrene-labelled BHK F-actin and its quenching by S1 and by an enzyme-linked ATPase assay. At non-saturating concentrations, both muscle and BHK actin activated skeletal muscle myosin to the same degree: at 30 degrees C and an ionic strength of 108 mM, 1 microM actin approximately doubled the ATPase of myosin or of S1. The association between BHK actin and S1 was also followed in a fluorescence stop flow: the rate of ATP binding monitored by the loss of light scattering upon dissociation of actin was again the same for BHK and muscle actin. The similarity of activation of myosin ATPase by BHK and muscle actin at low actin concentrations (i.e. the similarity of Vmax/Km) suggests that both Vmax and Km are similar for the two types of actin. The effect of varying filament length on actin activation of myosin ATPase was examined using pig plasma or BHK gelsolin to regulate the length. For both types of actin, maximum enhancement of the actomyosin ATPase activity was observed at an actin/gelsolin ratio of about 30:1, whereas inhibition was observed at lower ratios. Both activation and inhibition of actomyosin ATPase were apparent in the absence or presence of calcium; differences were observed only in the extent and the time course of the effect.