Cell, Vol. 66, 419-421,
August
9, 1991, Copyright
0 1991 by Cell Press
Profilin, a Weak CAP for Actin and RAS
Minireview
Pascal J. Goldschmidt-Clermont’ and Paul A. Janmeyt *Department of Medicine Cardiology Division The Johns Hopkins University School of Medicine Baltimore, Maryland 21205 tExperimental Medicine Division Brigham and Women’s Hospital Harvard Medical School Boston, Massachusetts 02115
Many cells such as platelets and fibroblasts change their shape and move in response to extracellular agonists. This is considered to be a crucial step, because it often occurs when cells become committed to the accomplishment of a function, growth, or even transformation. Much has been learned about the membrane events involved in cell signaling (Cantley et al., 1991). Concurrently, major discoveries have been made concerning the structure of the cytoskeleton that controls cell shape and motility (Stossel, 1989; Pollard, 1990). The connection between membrane activity and cytoskeletal structure has remained a mystery. However, specific interactions between proteins of the actin-based skeleton and key molecules of membrane signaling pathways have been identified in vitro and now in vivo: the actin-binding protein profilin may contribute to a signaling pathway in Saccharomyces cerevisiae that is regulated by RAS, the ubiquitous GTPase that plays a central role in differentiation, growth, and proliferation of diverse cells (Hall, 1990) and by the adenyl cyclase-associated protein known as CAP (Vojtek et al., 1991). In S. cerevisiae, the only cell in which an effector of RAS has been characterized, RAS activates adenylate cyclase. This effect specifically requires the N-terminal domain of CAP (Gerst et al., 1991; Figure 1). The phenotype of yeast cells that are null for CAP or its C-terminal domain resembles that of yeast cells depleted of profilin by gene disruption. They grow slowly, become very large and round, and have budding and nutritional defects. In yeast cells null for
the C-terminal domain of CAP, overexpression of profilin largely reverses the mutant phenotype (Vojtek et al., 1991; Figure 1). This result suggests that profilin participates either directly in the RAS-CAP pathway or in another signaling pathway that intersects the RAS-CAP system (Figure 2). profilin Regulation of Actin Assembly The physiological function of profilin has been controversial since this 12-15 kd protein was isolated from bovine spleen (Carlsson et al., 1976). Profilin was initially described as an inhibitor of actin polymerization that forms complexes with monomeric actin and thereby prevents actin filament formation. Although textbooks generally attribute to profilin the fact that nonmuscle cells maintain large amounts of unpolymerized cytoplasmic actin, recent studies show that the cellular concentration of profilin and its affinity for actin may be insufficient to accomplish this sequestering role in vivo. Careful modeling of its effect on actin polymerization in vitro indicates that profilin may also bind to or “cap” the fast-growing ends of actin filaments (Pollard and Cooper, 1986) and accelerate the exchange of the adenine nucleotide bound to monomeric actin (Goldschmidt-Clermont et al., 1991 b). ATP-actin monomers polymerize faster than ADP-actin (Pollard and Cooper, 1986) and makestifferfilaments(Janmeyet al., 1990). Therefore, under conditions of rapid filament reorganization, where large amounts of ADP-actin monomers are produced, and in the presence of a large excess of ATP over ADP (as in many nonmuscle cells after agonist stimulation), profilin may actually promote actin polymerization. Profilin Connection with Membrane Polyphosphoinositides The regulation of profilin’s interaction with actin is not mediated by Ca2+ or phosphorylation, as is the case for other actin-binding proteins (Hartwig and Kwiatkowski, 1991). Instead, profilin binds tightly and specifically to polyphosphoinositides; such binding dissociates profilin from actin in vitro (Figure 2; Lassing and Lindberg, 1985). Reciprocally, the binding of profilin to clusters of 4-5 molecules of
Figure visiae
Membrane
a
0-u RAS
b CAP Adenylyl cycmse
Cytoplasm
I * NOllllal Growth and Morphology
Abnormal Morphology
Abnormal Response to RAS
d
C
Abnormal Morphology
Rescued Morphology
Abrwmal Respfmse to RAS
1. The RAS-CAP
Pathway
in S. cere-
(a) In budding yeast, RAS activates adenyl cyclase in a complex containing CAP (cyclaseassociated protein), localized at the cytoplasmic surface of the membrane. (b) CAP is a bifunctional signaling protein: its N-terminal domain is necessary for normal responsiveness to activated PAS, and its C-terminal domain participates in specific cell functions, one of which controls cell morphology. (c) In cells null for CAP (that display abnormal morphology), expression of the N-terminal domain of CAP (N) restores cellular responsiveness to RAS. (d) Expression of the C-terminal (COCH) domain of CAP corrects the morphology of ceils, but they are unable to respond to RAS.
a PIP:!
b
PIP:,
CELL MEMBRANE
CYTOPLASM
C
d
EGFR J\
* Y
PLC ~
Figure
2. Profilin
Connection
with the Polyphosphoinositide
Cycle
and Receptor
Tyrosine
Kinase
Pathway
(a) Binding of profilin (Pro) to small clusters of PIP2 inhibits profilin interaction with actin and thereby the participation of profilin in the reorganization of the cytoskeleton. (b) The affinity of the complex between profilin and PIP, clusters is sufficient to block the access of unphosphorylated phospholipase Cy to its substrates. y, unphosphorylated tyrosine residue. (c) In resting cells, the tyrosine kinase of the epidermal growth factor receptor (EGFR) is not activated. Soluble phospholipase CT is inhibited by the binding of profilin to PIP,. (d) In the presence of ligand, the receptor dimerizes and the activated tyrosine kinase phosphorylates the receptor itself and tyrosine residues on phospholipase CT (Cantley et al., 1991). Phosphorylation of phospholipase Cy allows this enzyme to overcome profilin inhibition and to hydrolyze PIP? to produce soluble inositol trisphosphate (IPJ and diacylglycerol (DAG) that remains in the membrane. It is possible that the hydrolysis of PIP, bound to profilin by phosphorylated phospholipase CT increases the off-rate of profilin from the membrane and thereby enhances profilin interaction with actin. The validity of this model, which is based on studies in vitro, has yet to be demonstrated in viva.
phosphatidylinositol 45bisphosphate (PIP,) in lipid bilayers (Goldschmidt-Clermont et al., 1990; Machesky et al., 1990) protects PIP2 from hydrolysis by soluble phosphatidylinositol-specific phospholipase C (Figure 2). Together, these in vitro studies indicate that profilin may link cell signaling at the membrane level to reorganization of the cytoskeleton; this suggestion is strengthened by in vivo evidence that profilin may play an important role in cell signaling (Vojtek et al., 1991). Other cytoskeletal proteins bind specifically to membrane polyphosphoinositides; these include gelsolin (Janmey and Stossel, 1987) and gCap39 (Vu et al., 1990). Given the limited concentration of polyphosphoinositides in cells, these proteins are likely to compete for binding to lipids. Quantitation of polyphosphoinositides in extracts of A431 cells treated with epidermal growth factor failed to show a direct correlation between polyphosphoinositide levels and the concentration of actin-gelsolin or actinprofilin complexes (Dadaby et al., 1991). These results suggest that the interaction between cytoskeletal proteins and polyphosphoinositides detected in vitro may not regulate their activity in the live cell. However, it is also possible that changes in the turnover rates of pools of polyphosphoinositides bound to specific actin-binding proteins (which
are not detected by measuring the bulk levels of polyphosphoinositides) may represent the mechanism by which the polyphosphoinositides regulate cytoskeletal proteins. Moreover, these interactions occur not in dilute solution but at the interface between membrane and cytoplasm. Therefore, the kinetics and affinity of such reactions may be difficult to predict by measuring their insoluble reactants and soluble products. Although both profilin isoforms present in Acanthamoeba bind actin with similar affinity, only profilin II binds PIP2 with high enough affinity (micromolar Kd) to inhibit unphosphorylated phospholipase Crl (Machesky et al., 1990). In budding yeast deficient for the C-terminal domain of CAP, transfection of profilin II cDNA, but not profilin I, can rescue the mutant cells (Vojtek et al., 1991). This is important, as it indicates that the CAP-rescuing function of profilin is independent of profilin interaction with actin; it is tempting to speculate that the CAP-rescuing function may be related to the interaction of profilin II with the polyphosphoinositide cycle. Profilfn and Receptor Tyrosine Kinase Pathway The traditional model to explain cell locomotion and morphogenesis in response to extracellular signals invokes two separate compartments: the cell membrane and the
Minireview 42t
cytosol (reviewed in Cantley et al., 1991). In this model, an agonist switches on key membrane-bound enzymes, resulting in a burst of formation of soluble second messengers, such as Ca2+, inositol trisphosphate, or CAMP. By diffusing into the cytosol, these messengers activate proteins that regulate the actin polymer network and thereby induce the reorganization of the cytoskeleton (Stossel, 1989). At the same time, protein kinases directly affect the functional state of cytoskeletal substrates by phosphorylating specific residues on these proteins. Although this model applies to specific proteins of the cytoskeleton, it may not be the only mechanism involved in the cytoskeletal response to receptor tyrosine kinase activation. Phosphorylation of phospholipase Crl, a major substrate for receptor tyrosine kinases (Cantley et al., 1991) allows phospholipase Cyl to hydrolyze the PIP* pool bound to profilin (Goldschmidt-Clermont et al., 1991 a). In cells activated with epidermal growth factor or plateletderived growth factor, a burst of inositol trisphosphate production is usually observed as part of the early cellular response to agonist stimulation. In a reconstituted system made of artificial membranes containing PIP2 and purified proteins, the activity of tyrosine-phosphorylated phospholipase Cyl is not detectably different from that of the unphosphorylated enzyme. However, when PIP2 is mixed with specific detergents (Nishibe et al., 1990) or when profilin is present in the reaction mixture, the activity of phosphorylated phospholipase Cyl is substantially greater than that of the unphosphorylated enzyme. The mechanism overcoming profilin inhibition is unknown but may involve the aggregation of phospholipase Cyl on the membrane (Todderud et al., 1990). These in vitro experiments suggest that profilin may participate in the pathway activated by receptor tyrosine kinases and may indeed be a part of the mechanism by which receptor tyrosine kinases control the reorganization of the cytoskeleton. This new mechanism of amplification involving the phosphorylation of phospholipase Cyl and the hydrolysis of PIP2 bound to cytoskeletal proteins represents a potential alternative way for receptor tyrosine kinase to control actin-binding proteins (Figure 2). Another interesting regulatory mechanism involves direct phosphorylation by tyrosine kinases of cytoskeletal proteins with a Src homology 2 (SH2) domain (Davis et al., 1991). Profilin and the RAS Pathway The ability of polyphosphoinositide-binding profilin to replace the C-terminal domain of CAP, which is in the yeast RAS pathway, indicates that these profilin isoforms may participate in the RAS pathway. The biochemical link between profilin and CAP remains to be established, and several possibilities exist. First, CAP itself may bind actin. Alternatively, given that CAP contains a rare stretch of six prolines and another proline-rich sequence bridging its N- and C-terminal domains and the highly specific binding of profilin to poly+-proline (Tanaka and Shibata, 1985) it is possible that CAP and profilin bind each other. Currently, however, evidence points to involvement of lipids within the polyphosphoinositide cycle. The biological activity of RAS is controlled by mitogenitally active phospholipids. Two proteins that regulate the
GTPase activity of H-Ras, GTPase-activating protein (GAP) and a new GTPase-inhibiting cytoplasmic protein (GIP), are regulated in opposite ways by specific phospholipids, including polyphosphoinositides and diacylglycerol (Tsai et al., 1990). The net result of the phospholipid regulation of GAP and GIP is the activation of H-Ras. The most potent activator is diacylglycerol, which is produced by phospholipase Crl in cells stimulated by receptor tyrosine kinases (Cantley et al., 1991). Like RAS and profilin, it is possible that CAP may also be regulated by phospholipids, and therefore the effect of profilin on the polyphosphoinositide cycle may have something to do with the ability of the PIPe-binding isoforms of profilin to rescue CAP-deficient yeast cells. Although the discovery of Vojtek et al. does not provide a direct clue about the regulation of the RAS pathway or the actin-based skeleton, it is remarkable because it shows that a protein participating in the organization of the cytoskeleton may also function as a signaling protein. References Canttey, L. C., Auger, K. R., Carpenter, A.. Kapeller, Ft., and Soltoff, S. (1991). Cartsson, Dresdner, 353-366.
L., Nystrom, H., Lovgren,
Dadaby, C. Y., Patton, Eiol. 112, 1151-1156.
C., Duckworth, B, Graziani, Cell 64, 281-302.
L.-E., Lindberg, U., Kannan, K. K., CidS., and Jornvall, H. (1976). J. Mol. Biol. 105, E., Cooper,
J. A., and Pike, L. (1991).
J. Cell
Davis, S., Lu, M. L., Lo, S. H., Lin, S., Butler, J. A., Druker, 6. J., Roberts, T. M., An, O., and Chen, L. B. (1991). Science 252,712-715. Gerst, J. E., Ferguson, K.. Vojtek, Mol. Cell. Biol. 17, 1246-1257. Goldschmidt-Clermont, Pollard, T. D. (1990). Goldschmidt-Clermont, S. G., and Pollard,
A., Wigler,
M., and Field, J. (1991).
P. J.. Machesky, L. M., Baldassare, Science 247, 1575-1578.
P. J.. Kim, J. W., Machesky, L. M., Rhee, T. D. (1991a). Science 251, 1231-1233.
Goldschmidt-Clermont, Pollard, T. D. (1991b).
P. J., Machesky, L. M., Doberstein, J. Cell Biol. 113, 1081-1089.
Hall, A. (1990).
249, 635-640.
Nature
Hartwig, 87-97.
J. H., and Kwiatkowski,
Janmey,
P. A., and Stossel,
D. J (1991). T. P. (1987).
Janmey, P. A., Hvidt, S., Oster, G., Lamb, wig, J. H. (1990). Nature 347, 95-99. Lassing,
J. J., and
I., and Lindberg,
U. (1985).
Curr. Opin.
Nature
Cell Biol. 3,
325, 362-364.
J., Stossel,
Nature
Machesky, L. M., Goldschmidt-Clermont. (1990). Cell Reg. 7, 937-950.
S. K., and
T. P., and Hart-
318, 472-474. P. J., and
Pollard,
T. D.
Nishibe, S., Wahl. M. I., Hernandez-Sotomayor, S. M. T.. Tonks, N. K., Rhee, S. G., and Carpenter, G. (1990). Science 250, 1253-1256. Pollard,
T. D. (1990).
Curr. Opin.
Pollard, T. D., and Cooper, 1035. Stossel,
T. P. (1989).
Tanaka,
M., and Shibata.
J. Viol. Chem. H. (1985).
Todderud, G.. Wahl, M. I., Rhee, Science 249, 296-298. Tsai, 965. Vojtek. Wigler.
M.-H.,
Yu, C.-L..
A.. Haarer, M. (1991).
Yu, F., Johnson, 1413-1415.
Cell Biol. 2, 33-40.
J. A. (1986).
and Stacey,
Annu. Rev. Biochem.
55,987-
264, 18261-18264. Eur. J. Biochem.
151, 291-297.
S. G.. and Carpenter, D. W. (1990).
G. (1990).
Science
250, 982
B., Field, J., Gerst, J., Pollard, T. D., Brown, Cell 66, this issue. P., Sudhof,
T., and Yin, H. L. (1990).
Science
S., and 250.
August
9, 1991, Copyright
0 1991 by Cell Press
Profilin, a Weak CAP for Actin and RAS
Minireview
Pascal J. Goldschmidt-Clermont’ and Paul A. Janmeyt *Department of Medicine Cardiology Division The Johns Hopkins University School of Medicine Baltimore, Maryland 21205 tExperimental Medicine Division Brigham and Women’s Hospital Harvard Medical School Boston, Massachusetts 02115
Many cells such as platelets and fibroblasts change their shape and move in response to extracellular agonists. This is considered to be a crucial step, because it often occurs when cells become committed to the accomplishment of a function, growth, or even transformation. Much has been learned about the membrane events involved in cell signaling (Cantley et al., 1991). Concurrently, major discoveries have been made concerning the structure of the cytoskeleton that controls cell shape and motility (Stossel, 1989; Pollard, 1990). The connection between membrane activity and cytoskeletal structure has remained a mystery. However, specific interactions between proteins of the actin-based skeleton and key molecules of membrane signaling pathways have been identified in vitro and now in vivo: the actin-binding protein profilin may contribute to a signaling pathway in Saccharomyces cerevisiae that is regulated by RAS, the ubiquitous GTPase that plays a central role in differentiation, growth, and proliferation of diverse cells (Hall, 1990) and by the adenyl cyclase-associated protein known as CAP (Vojtek et al., 1991). In S. cerevisiae, the only cell in which an effector of RAS has been characterized, RAS activates adenylate cyclase. This effect specifically requires the N-terminal domain of CAP (Gerst et al., 1991; Figure 1). The phenotype of yeast cells that are null for CAP or its C-terminal domain resembles that of yeast cells depleted of profilin by gene disruption. They grow slowly, become very large and round, and have budding and nutritional defects. In yeast cells null for
the C-terminal domain of CAP, overexpression of profilin largely reverses the mutant phenotype (Vojtek et al., 1991; Figure 1). This result suggests that profilin participates either directly in the RAS-CAP pathway or in another signaling pathway that intersects the RAS-CAP system (Figure 2). profilin Regulation of Actin Assembly The physiological function of profilin has been controversial since this 12-15 kd protein was isolated from bovine spleen (Carlsson et al., 1976). Profilin was initially described as an inhibitor of actin polymerization that forms complexes with monomeric actin and thereby prevents actin filament formation. Although textbooks generally attribute to profilin the fact that nonmuscle cells maintain large amounts of unpolymerized cytoplasmic actin, recent studies show that the cellular concentration of profilin and its affinity for actin may be insufficient to accomplish this sequestering role in vivo. Careful modeling of its effect on actin polymerization in vitro indicates that profilin may also bind to or “cap” the fast-growing ends of actin filaments (Pollard and Cooper, 1986) and accelerate the exchange of the adenine nucleotide bound to monomeric actin (Goldschmidt-Clermont et al., 1991 b). ATP-actin monomers polymerize faster than ADP-actin (Pollard and Cooper, 1986) and makestifferfilaments(Janmeyet al., 1990). Therefore, under conditions of rapid filament reorganization, where large amounts of ADP-actin monomers are produced, and in the presence of a large excess of ATP over ADP (as in many nonmuscle cells after agonist stimulation), profilin may actually promote actin polymerization. Profilin Connection with Membrane Polyphosphoinositides The regulation of profilin’s interaction with actin is not mediated by Ca2+ or phosphorylation, as is the case for other actin-binding proteins (Hartwig and Kwiatkowski, 1991). Instead, profilin binds tightly and specifically to polyphosphoinositides; such binding dissociates profilin from actin in vitro (Figure 2; Lassing and Lindberg, 1985). Reciprocally, the binding of profilin to clusters of 4-5 molecules of
Figure visiae
Membrane
a
0-u RAS
b CAP Adenylyl cycmse
Cytoplasm
I * NOllllal Growth and Morphology
Abnormal Morphology
Abnormal Response to RAS
d
C
Abnormal Morphology
Rescued Morphology
Abrwmal Respfmse to RAS
1. The RAS-CAP
Pathway
in S. cere-
(a) In budding yeast, RAS activates adenyl cyclase in a complex containing CAP (cyclaseassociated protein), localized at the cytoplasmic surface of the membrane. (b) CAP is a bifunctional signaling protein: its N-terminal domain is necessary for normal responsiveness to activated PAS, and its C-terminal domain participates in specific cell functions, one of which controls cell morphology. (c) In cells null for CAP (that display abnormal morphology), expression of the N-terminal domain of CAP (N) restores cellular responsiveness to RAS. (d) Expression of the C-terminal (COCH) domain of CAP corrects the morphology of ceils, but they are unable to respond to RAS.
a PIP:!
b
PIP:,
CELL MEMBRANE
CYTOPLASM
C
d
EGFR J\
* Y
PLC ~
Figure
2. Profilin
Connection
with the Polyphosphoinositide
Cycle
and Receptor
Tyrosine
Kinase
Pathway
(a) Binding of profilin (Pro) to small clusters of PIP2 inhibits profilin interaction with actin and thereby the participation of profilin in the reorganization of the cytoskeleton. (b) The affinity of the complex between profilin and PIP, clusters is sufficient to block the access of unphosphorylated phospholipase Cy to its substrates. y, unphosphorylated tyrosine residue. (c) In resting cells, the tyrosine kinase of the epidermal growth factor receptor (EGFR) is not activated. Soluble phospholipase CT is inhibited by the binding of profilin to PIP,. (d) In the presence of ligand, the receptor dimerizes and the activated tyrosine kinase phosphorylates the receptor itself and tyrosine residues on phospholipase CT (Cantley et al., 1991). Phosphorylation of phospholipase Cy allows this enzyme to overcome profilin inhibition and to hydrolyze PIP? to produce soluble inositol trisphosphate (IPJ and diacylglycerol (DAG) that remains in the membrane. It is possible that the hydrolysis of PIP, bound to profilin by phosphorylated phospholipase CT increases the off-rate of profilin from the membrane and thereby enhances profilin interaction with actin. The validity of this model, which is based on studies in vitro, has yet to be demonstrated in viva.
phosphatidylinositol 45bisphosphate (PIP,) in lipid bilayers (Goldschmidt-Clermont et al., 1990; Machesky et al., 1990) protects PIP2 from hydrolysis by soluble phosphatidylinositol-specific phospholipase C (Figure 2). Together, these in vitro studies indicate that profilin may link cell signaling at the membrane level to reorganization of the cytoskeleton; this suggestion is strengthened by in vivo evidence that profilin may play an important role in cell signaling (Vojtek et al., 1991). Other cytoskeletal proteins bind specifically to membrane polyphosphoinositides; these include gelsolin (Janmey and Stossel, 1987) and gCap39 (Vu et al., 1990). Given the limited concentration of polyphosphoinositides in cells, these proteins are likely to compete for binding to lipids. Quantitation of polyphosphoinositides in extracts of A431 cells treated with epidermal growth factor failed to show a direct correlation between polyphosphoinositide levels and the concentration of actin-gelsolin or actinprofilin complexes (Dadaby et al., 1991). These results suggest that the interaction between cytoskeletal proteins and polyphosphoinositides detected in vitro may not regulate their activity in the live cell. However, it is also possible that changes in the turnover rates of pools of polyphosphoinositides bound to specific actin-binding proteins (which
are not detected by measuring the bulk levels of polyphosphoinositides) may represent the mechanism by which the polyphosphoinositides regulate cytoskeletal proteins. Moreover, these interactions occur not in dilute solution but at the interface between membrane and cytoplasm. Therefore, the kinetics and affinity of such reactions may be difficult to predict by measuring their insoluble reactants and soluble products. Although both profilin isoforms present in Acanthamoeba bind actin with similar affinity, only profilin II binds PIP2 with high enough affinity (micromolar Kd) to inhibit unphosphorylated phospholipase Crl (Machesky et al., 1990). In budding yeast deficient for the C-terminal domain of CAP, transfection of profilin II cDNA, but not profilin I, can rescue the mutant cells (Vojtek et al., 1991). This is important, as it indicates that the CAP-rescuing function of profilin is independent of profilin interaction with actin; it is tempting to speculate that the CAP-rescuing function may be related to the interaction of profilin II with the polyphosphoinositide cycle. Profilfn and Receptor Tyrosine Kinase Pathway The traditional model to explain cell locomotion and morphogenesis in response to extracellular signals invokes two separate compartments: the cell membrane and the
Minireview 42t
cytosol (reviewed in Cantley et al., 1991). In this model, an agonist switches on key membrane-bound enzymes, resulting in a burst of formation of soluble second messengers, such as Ca2+, inositol trisphosphate, or CAMP. By diffusing into the cytosol, these messengers activate proteins that regulate the actin polymer network and thereby induce the reorganization of the cytoskeleton (Stossel, 1989). At the same time, protein kinases directly affect the functional state of cytoskeletal substrates by phosphorylating specific residues on these proteins. Although this model applies to specific proteins of the cytoskeleton, it may not be the only mechanism involved in the cytoskeletal response to receptor tyrosine kinase activation. Phosphorylation of phospholipase Crl, a major substrate for receptor tyrosine kinases (Cantley et al., 1991) allows phospholipase Cyl to hydrolyze the PIP* pool bound to profilin (Goldschmidt-Clermont et al., 1991 a). In cells activated with epidermal growth factor or plateletderived growth factor, a burst of inositol trisphosphate production is usually observed as part of the early cellular response to agonist stimulation. In a reconstituted system made of artificial membranes containing PIP2 and purified proteins, the activity of tyrosine-phosphorylated phospholipase Cyl is not detectably different from that of the unphosphorylated enzyme. However, when PIP2 is mixed with specific detergents (Nishibe et al., 1990) or when profilin is present in the reaction mixture, the activity of phosphorylated phospholipase Cyl is substantially greater than that of the unphosphorylated enzyme. The mechanism overcoming profilin inhibition is unknown but may involve the aggregation of phospholipase Cyl on the membrane (Todderud et al., 1990). These in vitro experiments suggest that profilin may participate in the pathway activated by receptor tyrosine kinases and may indeed be a part of the mechanism by which receptor tyrosine kinases control the reorganization of the cytoskeleton. This new mechanism of amplification involving the phosphorylation of phospholipase Cyl and the hydrolysis of PIP2 bound to cytoskeletal proteins represents a potential alternative way for receptor tyrosine kinase to control actin-binding proteins (Figure 2). Another interesting regulatory mechanism involves direct phosphorylation by tyrosine kinases of cytoskeletal proteins with a Src homology 2 (SH2) domain (Davis et al., 1991). Profilin and the RAS Pathway The ability of polyphosphoinositide-binding profilin to replace the C-terminal domain of CAP, which is in the yeast RAS pathway, indicates that these profilin isoforms may participate in the RAS pathway. The biochemical link between profilin and CAP remains to be established, and several possibilities exist. First, CAP itself may bind actin. Alternatively, given that CAP contains a rare stretch of six prolines and another proline-rich sequence bridging its N- and C-terminal domains and the highly specific binding of profilin to poly+-proline (Tanaka and Shibata, 1985) it is possible that CAP and profilin bind each other. Currently, however, evidence points to involvement of lipids within the polyphosphoinositide cycle. The biological activity of RAS is controlled by mitogenitally active phospholipids. Two proteins that regulate the
GTPase activity of H-Ras, GTPase-activating protein (GAP) and a new GTPase-inhibiting cytoplasmic protein (GIP), are regulated in opposite ways by specific phospholipids, including polyphosphoinositides and diacylglycerol (Tsai et al., 1990). The net result of the phospholipid regulation of GAP and GIP is the activation of H-Ras. The most potent activator is diacylglycerol, which is produced by phospholipase Crl in cells stimulated by receptor tyrosine kinases (Cantley et al., 1991). Like RAS and profilin, it is possible that CAP may also be regulated by phospholipids, and therefore the effect of profilin on the polyphosphoinositide cycle may have something to do with the ability of the PIPe-binding isoforms of profilin to rescue CAP-deficient yeast cells. Although the discovery of Vojtek et al. does not provide a direct clue about the regulation of the RAS pathway or the actin-based skeleton, it is remarkable because it shows that a protein participating in the organization of the cytoskeleton may also function as a signaling protein. References Canttey, L. C., Auger, K. R., Carpenter, A.. Kapeller, Ft., and Soltoff, S. (1991). Cartsson, Dresdner, 353-366.
L., Nystrom, H., Lovgren,
Dadaby, C. Y., Patton, Eiol. 112, 1151-1156.
C., Duckworth, B, Graziani, Cell 64, 281-302.
L.-E., Lindberg, U., Kannan, K. K., CidS., and Jornvall, H. (1976). J. Mol. Biol. 105, E., Cooper,
J. A., and Pike, L. (1991).
J. Cell
Davis, S., Lu, M. L., Lo, S. H., Lin, S., Butler, J. A., Druker, 6. J., Roberts, T. M., An, O., and Chen, L. B. (1991). Science 252,712-715. Gerst, J. E., Ferguson, K.. Vojtek, Mol. Cell. Biol. 17, 1246-1257. Goldschmidt-Clermont, Pollard, T. D. (1990). Goldschmidt-Clermont, S. G., and Pollard,
A., Wigler,
M., and Field, J. (1991).
P. J.. Machesky, L. M., Baldassare, Science 247, 1575-1578.
P. J.. Kim, J. W., Machesky, L. M., Rhee, T. D. (1991a). Science 251, 1231-1233.
Goldschmidt-Clermont, Pollard, T. D. (1991b).
P. J., Machesky, L. M., Doberstein, J. Cell Biol. 113, 1081-1089.
Hall, A. (1990).
249, 635-640.
Nature
Hartwig, 87-97.
J. H., and Kwiatkowski,
Janmey,
P. A., and Stossel,
D. J (1991). T. P. (1987).
Janmey, P. A., Hvidt, S., Oster, G., Lamb, wig, J. H. (1990). Nature 347, 95-99. Lassing,
J. J., and
I., and Lindberg,
U. (1985).
Curr. Opin.
Nature
Cell Biol. 3,
325, 362-364.
J., Stossel,
Nature
Machesky, L. M., Goldschmidt-Clermont. (1990). Cell Reg. 7, 937-950.
S. K., and
T. P., and Hart-
318, 472-474. P. J., and
Pollard,
T. D.
Nishibe, S., Wahl. M. I., Hernandez-Sotomayor, S. M. T.. Tonks, N. K., Rhee, S. G., and Carpenter, G. (1990). Science 250, 1253-1256. Pollard,
T. D. (1990).
Curr. Opin.
Pollard, T. D., and Cooper, 1035. Stossel,
T. P. (1989).
Tanaka,
M., and Shibata.
J. Viol. Chem. H. (1985).
Todderud, G.. Wahl, M. I., Rhee, Science 249, 296-298. Tsai, 965. Vojtek. Wigler.
M.-H.,
Yu, C.-L..
A.. Haarer, M. (1991).
Yu, F., Johnson, 1413-1415.
Cell Biol. 2, 33-40.
J. A. (1986).
and Stacey,
Annu. Rev. Biochem.
55,987-
264, 18261-18264. Eur. J. Biochem.
151, 291-297.
S. G.. and Carpenter, D. W. (1990).
G. (1990).
Science
250, 982
B., Field, J., Gerst, J., Pollard, T. D., Brown, Cell 66, this issue. P., Sudhof,
T., and Yin, H. L. (1990).
Science
S., and 250.
