Regulation of profilin localization in Saccharomyces cerevisiae by phosphoinositide metabolism.

Profilin is an actin- and phosphatidylinositol 4,5-bisphosphate-binding protein that plays a role in the organization of the cytoskeleton and may be involved in growth factor signaling pathways. The subcellular localization of profilin was examined in the yeast Saccharomyces cerevisiae. Immunoblot analysis showed that profilin was localized in both the plasma membrane and cytosolic fractions of the cell. Actin was bound to the profilin localized in the cytosol. The association of profilin with the membrane was peripheral and mediated through interaction with phospholipid. The phospholipid dependence of profilin for membrane binding was examined in vitro using pure profilin and defined unilamellar phospholipid vesicles. The presence of phosphatidylinositol 4,5-bisphosphate in phospholipid vesicles was required for maximum profilin binding. Moreover, the binding of profilin to phospholipid vesicles was dependent on the surface concentration of phosphatidylinositol 4,5-bisphosphate. The subcellular localization of profilin was examined in vivo under growth conditions (i.e. inositol starvation of ino1 cells and glucose starvation of respiratory deficient cells) where plasma membrane levels of phosphatidylinositol 4,5-bisphosphate were depleted. Depletion of plasma membrane phosphatidylinositol 4,5-bisphosphate levels resulted in a translocation of profilin from the plasma membrane to the cytosolic fraction. Profilin translocated back to the membrane fraction from the cytosol under growth conditions where plasma membrane levels of phosphatidylinositol 4,5-bisphosphate were replenished. These results suggested that phosphoinositide metabolism played a role in the localization of profilin.

Profilin is an actin-and phosphatidylinositol 4,5bisphosphate-binding protein that plays a role in the organization of the cytoskeleton and may be involved in growth factor signaling pathways. The subcellular localization of profilin was examined in the yeast Saccharomyces cerevisiae. Immunoblot analysis showed that profilin was localized in both the plasma membrane and cytosolic fractions of the cell. Actin was bound to the profilin localized in the cytosol. The association of profilin with the membrane was peripheral and mediated through interaction with phospholipid. The phospholipid dependence of profilin for membrane binding was examined in vitro using pure profilin and defined unilamellar phospholipid vesicles. The presence of phosphatidylinositol 4,5-bisphosphate in phospholipid vesicles was required for maximum profilin binding. Moreover, the binding of profilin to phospholipid vesicles was dependent on the surface concentration of phosphatidylinositol 4,5-bisphosphate. The subcellular localization of profilin was examined in vivo under growth conditions (i.e. inositol starvation of ino1 cells and glucose starvation of respiratory deficient cells) where plasma membrane levels of phosphatidylinositol 4,5-bisphosphate were depleted. Depletion of plasma membrane phosphatidylinositol 4,5-bisphosphate levels resulted in a translocation of profilin from the plasma membrane to the cytosolic fraction. Profilin translocated back to the membrane fraction from the cytosol under growth conditions where plasma membrane levels of phosphatidylinositol 4,5-bisphosphate were replenished. These results suggested that phosphoinositide metabolism played a role in the localization of profilin.
Profilin is a ubiquitous actin-and PIP 2 1 -binding protein found in eucaryotic cells (1). It is believed that profilin assists in the regulation of the reorganization of the microfilament cytoskeleton (2). Profilin competes with other actin-binding proteins for globular actin (3), catalyzing the exchange of nucleotide phosphates on monomeric actin, thereby activating actin for polymerization (4).
Profilin has also been shown to bind anionic phospholipids (5). However, at physiological salt concentrations, profilin binds exclusively to PIP 2 (6). Profilin cannot bind to actin and PIP 2 simultaneously (7), as the actin-and PIP 2 -binding sites are sufficiently close as to impart mutually exclusive binding (7). The interaction of profilin with PIP 2 may regulate the availability of profilin for interaction with actin (5). Another role proposed for profilin's association with PIP 2 is to prevent the phospholipase C␥1-mediated hydrolysis of PIP 2 (8), and thus the generation of the second messengers inositol trisphosphate and diacylglycerol (9). However, tyrosine phosphorylation of phospholipase C␥1 allows hydrolysis of PIP 2 even in the presence of profilin (10).
We are using the yeast Saccharomyces cerevisiae as a model eucaryote to study profilin. The gene (PFY1) encoding for profilin has been cloned from S. cerevisiae (11). S. cerevisiae profilin resembles profilins from higher eucaryotes in size (ϳ15 kDa), ability to bind actin, PIP 2 , and polyproline, and inhibit ATP hydrolysis by monomeric actin (11)(12)(13). Moreover, profilin is required for the proper organization of the actin cytoskeleton into actin cables, which occur at regions of active growth and for proper maintenance of cell polarity (12). It is unclear whether these roles for profilin are related to its ability to bind actin and/or PIP 2 (13). The physiological relevance of profilin's ability to bind polyproline is also unclear (13).
The fact that binding to actin and PIP 2 are mutually exclusive in profilin, and that actin is localized in the cytosol (14) whereas PIP 2 is localized in the plasma membrane (15), raises the suggestion that profilin is localized in these two compartments of the cell. In this report we demonstrated that profilin was indeed localized to the plasma membrane and cytosolic fractions of S. cerevisiae. The majority of profilin was bound to plasma membranes. However, under growth conditions where the plasma membrane levels of PIP 2 were depleted, profilin was found to translocate from the membrane to the cytosol. Profilin translocated back to the membrane fraction from the cytosol under growth conditions where plasma membrane levels of phosphatidylinositol 4,5-bisphosphate were replenished. The results of our studies were consistent with the notion that the metabolism of PIP 2 played a role in the localization of profilin.

Materials
All chemicals were reagent grade. Growth medium supplies were purchased from Difco Laboratories. ATP, phospholipids, phenylmethanesulfonyl fluoride, benzamide, aprotinin, leupeptin, pepstatin, and bovine serum albumin were obtained from Sigma. Protein assay reagent was purchased from Bio-Rad. SDS-polyacrylamide gel electrophoresis and immunochemical reagents were from Life Technologies, Inc. Sephadex G-50 superfine was purchased from Pharmacia Biotech. Octyl-␤-D-glucopyranoside (octyl glucoside) was from Calbiochem. Cen-* This work was supported by United States Public Health Service Grant GM-28140 from the National Institutes of Health, New Jersey State Funds, and the Charles and Johanna Busch Memorial Fund. This is New Jersey Agricultural Experiment Station Publication D-10531-2-95. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ¶ To whom correspondence and reprint requests should be addressed. Tel.: 908-932-9663; Fax: 908-932-6776; E-mail: george@a1.caft1vax. rutgers.edu. 1 The abbreviations used are: PIP 2 , phosphatidylinositol 4, 5-bisphosphate; PIP, phosphatidylinositol 4-phosphate; PC, phosphatidylcholine. tricon-10 concentration filters were purchased from Amicon. Anti-actin antibodies were provided by Susan S. Brown (University of Michigan Medical School).
Purification of Profilin and the Preparation of Anti-profilin Antibodies-Profilin was purified to apparent homogeneity using polyproline affinity chromatography as described by Haarer et al. (12). Antiserum to profilin was raised in New Zealand White rabbits by standard procedures (20). Anti-profilin antibodies were purified from the antiserum as described by Olmsted (21) using nitrocellulose-bound profilin as an affinity adsorbent.
Preparation of Subcellular Fractions-All steps were carried out at 5°C. Cell extracts were prepared by disruption with glass beads using a Bead-Beater (Biospec Products) in buffer A (50 mM Tris-HCl (pH 7.5), 1 mM Na 2 EDTA, 0.3 M sucrose, 10 mM 2-mercaptoethanol, 0.5 mM phenylmethanesulfonyl fluoride, 1 mM benzamide, 5 g/ml aprotinin, 5 g/ml leupeptin, and 6 g/ml pepstatin) as described previously (22). Unbroken cells and glass beads were removed by centrifugation at 1,500 ϫ g for 5 min. Mitochondrial (23) and microsomal (22) fractions were collected from cell extracts by differential centrifugation. The postmicrosomal supernatant was used as the cytosolic fraction. The cytosolic fraction was concentrated with a Centricon-10 concentration filter. Plasma membranes were isolated by the method of Serrano (24) as modified by Monk et al. (25). Measurement of plasma membrane (vanadate-sensitive) ATPase activity (26) was used to assess the purity of plasma membranes. Vanadate-sensitive ATPase activity accounted for 90% of the total ATPase activity in plasma membranes. Membrane fractions were suspended in buffer A to a volume equivalent to that of the cytosolic fraction.
Preparation of Phospholipid Vesicles-Large (average diameter ϭ 90 nm) unilamellar phospholipid vesicles were prepared in 50 mM Tris-HCl buffer (pH 8.0) by removing octyl glucoside from octyl glucoside/phospholipid (molar ratio of 15:1) mixed micelles (30) by Sephadex G-50 superfine column chromatography (31,32). Pure phospholipids and phospholipids extracted (33) from yeast plasma membranes were used for the preparation of phospholipid vesicles. The phospholipid composition of the vesicles is indicated in figure legends.
Analysis of Profilin Binding to Plasma Membranes and Phospholipid Vesicles-Pure profilin (5 g) was incubated for 30 min at room temperature with the indicated concentrations of plasma membranes or phospholipid vesicles in 50 mM Tris-HCl buffer (pH 8.0) in a total volume of 0.5 ml. Following incubation, the profilin bound to plasma membranes or phospholipid vesicles was separated from unbound profilin by Sephadex G-50 superfine column chromatography (32) or by centrifugation at 100,000 ϫ g for 1 h. Both methods for the separation of bound and unbound profilin gave the same results. The amount of bound and unbound profilin was quantitatively determined by scanning densitometry of immunoblots using pure profilin to generate a standard curve. The curve was linear over the concentration range of 50 -5000 ng of profilin. The lower limit of sensitivity for the binding assay was 6 g of profilin/mg of phospholipid phosphate. Profilin binding assays were conducted in triplicate with an average standard deviation of Ϯ3%. The amount of profilin bound to membranes and phospholipid vesicles was normalized to phospholipid phosphate.
Analysis of Protein and Phospholipid Phosphate-Protein was determined by the method of Bradford (34) using bovine serum albumin as the standard. The phosphate concentration of membranes and phospholipid vesicles was determined by the method of Bartlett (35).

Subcellular
Localization of Profilin-Affinity-purified antiprofilin antibodies were prepared and used to examine the subcellular localization of profilin. The anti-profilin antibodies could be used to immunoprecipitate profilin from cell extracts (data not shown) and to analyze profilin by immunoblotting (Fig. 1A). Immunoblot signals were optimized by analyzing a number of antigen and antibody concentrations and were in the linear range of detectability. Profilin was localized to the total membrane and cytosolic fractions of the cell (Fig. 1A). About 80% of the total amount of profilin was associated with membranes, whereas the remainder was associated with the cytosolic fraction (Fig. 1B). Subcellular membrane fractions were isolated and examined for the association of profilin. Essentially all of the membrane-associated profilin was found with the plasma membrane fraction (Fig. 1C).
We next questioned whether the association of profilin to membranes was integral or peripheral. Peripheral membrane proteins are associated with membranes through noncovalent interactions and can be released by treatment with high ionic strength. Integral membrane proteins are more tightly associated and cannot be released by changes in ionic strength. Membranes were treated with NaCl, and the amount of profilin remaining bound to the membrane was analyzed. Treatment with NaCl resulted in a dose-dependent dissociation of profilin from the membranes (Fig. 2). Maximum dissociation was obtained with 0.1 M NaCl. Thus, profilin was a peripheral membrane protein.
Since profilin is an actin-binding protein (1), we questioned whether the cytosolic-associated profilin was bound to actin in the cytosol. Profilin was purified from the cytosolic fraction by affinity chromatography with polyproline-Sepharose (12). The polyproline-Sepharose chromatography procedure was modified such that the column was not washed with buffer containing 3 M urea prior to the elution of profilin with 6 M urea (12). In this way, if profilin was bound to actin in the cytosol, it should remain bound to actin until dissociated from the polyproline-Sepharose column with 6 M urea. The actin-and polyproline-binding sites on profilin are not the same (13). The purified profilin preparation was subjected to immunoblot analysis using affinity-purified anti-actin antibodies. Indeed, actin was present in the profilin preparation purified from the cytosolic fraction (Fig. 3).
Binding of Pure Profilin to Plasma Membranes and Phospholipid Vesicles Prepared from Plasma Membrane Phospholipids-The nature of profilin's association with the plasma membrane was examined by binding studies using pure profilin and plasma membranes. Plasma membranes were washed with 0.5 M NaCl to remove profilin and other peripheral membrane proteins. Pure profilin was then incubated with the salt-washed plasma membranes, and the amount of profilin which bound to the membranes was analyzed. Profilin bound to plasma membranes in a concentration-dependent manner (Fig.  4). Maximum profilin binding was obtained with 80 g/ml phospholipid phosphate. The concentration of plasma membranes that gave half-maximum binding was 30 g/ml of phospholipid phosphate. These results indicated that the association of profilin with plasma membranes was not through interaction with other peripheral membrane proteins.
Phospholipids were extracted from plasma membranes and used for the preparation of phospholipid vesicles. These vesi-cles were used to examine profilin binding using phospholipid phosphate concentrations similar to that used in the experiment shown in Fig. 4. Profilin bound to these phospholipid vesicles in a concentration-dependent manner similar to that found with the salt-washed plasma membranes (Fig. 5). Profilin binding to these vesicles saturated at 80 g/ml phospholipid phosphate and half-maximum binding was at 38 g/ml phospholipid phosphate. The results of this experiment indicated that the association of profilin with plasma membranes was not through interaction with integral membrane proteins. Thus, the association of profilin with the plasma membrane was through interaction with membrane phospholipids.
Effect of Phospholipid Vesicle Composition on Profilin Binding-We examined the effect of phospholipid composition on the ability of profilin to bind vesicles. Phosphatidylcholine vesicles were prepared with phospholipids known to be present in the plasma membranes of S. cerevisiae (36,37). Phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol, phosphatidylserine, and phosphatidate are plasma membrane phospholipids that are also localized in most cellular organelles (37). PIP and PIP 2 are minor phospholipids localized in the plasma membrane (15). The only phospholipid vesicles that showed the ability to bind a significant amount of profilin were those containing PIP 2 (Fig. 6). The addition of PIP to PC:PIP 2 vesicles did not have a major effect on profilin binding (Fig. 6). Binding to PC:PIP 2 vesicles was dependent on the concentration of profilin (Fig. 7). The K d value for profilin was 0.13 M.
Profilin Binding to PC:PIP 2 Vesicles Follows Surface Dilution Kinetics-Lipid-dependent enzymes, which bind their substrates at a membrane surface, follow surface dilution kinetics  (38). Surface dilution kinetics takes into account both the surface and bulk concentrations of lipid when defining kinetic parameters of lipid-dependent enzymes (38,39). Profilin was analogous to a lipid-dependent enzyme in that it bound to PIP 2 at a membrane surface. The surface concentration of PIP 2 , as opposed to the bulk concentration, should be the relevant concentration when considering the binding of profilin to a membrane in vivo. We examined the dependence of the surface concentration of PIP 2 (expressed in mol %) in vesicles on profilin binding. These experiments were performed such that profilin binding was independent of the bulk concentration of phospholipid vesicles (i.e. at a bulk vesicle concentration of 100 g/ml phospholipid phosphate, Figs. 4 and 5). Profilin binding to PC:PIP 2 vesicles was indeed dependent on the surface concentration of PIP 2 (Fig. 8). The binding dependence of profilin to PIP 2 appeared cooperative with a Hill number of 2. However, these data do not rule out a bimolecular reaction. The concentration of PIP 2 that gave half-maximum binding was 8.5 mol %.
Effect of Inositol Starvation on the Subcellular Localization of Profilin-We sought evidence that the association of profilin with the plasma membrane was dependent on the concentration of PIP 2 in vivo. The availability of the ino1 mutant facilitated experiments to address this question. Strain MC13 is auxotrophic for inositol due to a mutation in the INO1 gene encoding for inositol-1-phosphate synthase (40,41). When ino1 cells are deprived of inositol, they undergo changes in their metabolism which ultimately result in a loss of cell viability, commonly referred to as "inositol-less death" (42,43). Among the early biochemical consequences of inositol starvation of ino1 cells is a 6-fold decrease in the plasma membrane levels of PIP and PIP 2 (36). PIP and PIP 2 account for 0.6 and 0.2 mol %, respectively, of the total plasma membrane phospholipids in S. cerevisiae (36). PIP and PIP 2 are derived from inositol via phosphatidylinositol synthase and the phosphoinositide kinases (44). Plasma membrane-associated phosphatidylinositol 4-kinase has been shown to be inhibited during inositol starvation of ino1 cells (36). Inositol-supplemented ino1 cells were grown to the exponential phase of growth. Cells were washed and then incubated for 4 h in fresh growth medium lacking inositol. As reported previously (43), the inositol-starved cells grew equally as well as the control inositol-supplemented cells during the 4-h incubation period. Following the incubation, the association of profilin with the membrane and cytosolic fractions of the cells was analyzed. This analysis showed that the percentage of profilin bound to membranes decreased from 83% to 66% upon inositol starvation (Fig. 9). At the same time, the amount of profilin associated with the cytosol increased from 17% to 34% (Fig. 9).
Effect of Glucose Starvation on the Subcellular Localization of Profilin-To further support the notion that PIP 2 metabo- lism played a role in the subcellular localization of profilin, we utilized respiratory deficient cells. Talwalkar and Lester (45) have shown that glucose starvation of respiratory deficient cells results in a decrease in the levels of PIP and PIP 2 of 3.6and 9.5-fold, respectively. The addition of glucose to glucosestarved cells has the opposite effect (45). Changes in the levels of these phosphoinositides have been ascribed to fluctuations in the cellular levels of ATP and ADP (45). These nucleotides regulate plasma membrane-associated phosphatidylinositol 4-kinase activity needed for the synthesis of PIP and PIP 2 (46). Glucose-supplemented cells were grown to the exponential phase of growth. Cells were washed and incubated in fresh growth medium without glucose. Glucose starvation resulted in a time-dependent decrease in the amount of profilin bound to the membrane fraction of the cells (Fig. 10). After 20 min of glucose starvation, the percentage of profilin bound to membranes decreased from 81% to 17% (Fig. 10). When glucose was added back to glucose-starved cells, the percentage of profilin bound to membranes increased from 17% to 62% (Fig. 10). Although not shown in Fig. 10, the amount of profilin associated with the cytosolic fraction changed proportionally with the amount of profilin associated with the membranes. DISCUSSION In this work we examined the regulation of profilin localization by PIP 2 metabolism in S. cerevisiae. The initial aim of this study was to examine the subcellular localization of profilin. About 80% of the cellular profilin was localized to the plasma membrane, whereas the remainder was localized to the cytosol. That profilin could be dissociated from membranes with high ionic strength buffer indicated that profilin was a peripheral membrane protein. Pure profilin bound to salt-washed plasma membranes and to unilamellar phospholipid vesicles prepared from phospholipids extracted from plasma membranes. These results raised the suggestion that profilin's association with plasma membranes was not mediated through interaction with other peripheral membrane proteins or integral membrane proteins. Instead, profilin's peripheral associated with plasma membranes was through interaction with membrane phospholipids. This conclusion would be expected based on profilin's ability to bind PIP 2 vesicles (13). PIP 2 is a minor phospholipid found in S. cerevisiae that is localized to plasma membranes (15). Thus, the localization of profilin to plasma membranes would also be expected. Actin was associated with profilin purified from the cytosol. This was consistent with profilin's role in the organization of the actin cytoskeleton (12).
The phospholipid requirement for profilin binding to membranes was examined with pure profilin and well defined unilamellar phospholipid vesicles. The vesicles were prepared with phospholipids found in S. cerevisiae plasma membranes. Under the conditions of our experiments, binding of profilin to these vesicles was dependent on the presence of PIP 2 . The dissociation constant for profilin binding to PC:PIP 2 vesicles (0.13 M) was in the range reported for profilin from higher eucaryotes (5,8). In addition, profilin binding to PC:PIP 2 vesicles was dependent on the surface concentration of PIP 2 . Thus, profilin followed surface dilution kinetics (39), a characteristic common to lipid-dependent enzymes that bind lipids at a membrane surface (38).
The physiological relevance of the dependence of profilin binding on the surface concentration of PIP 2 was addressed by two independent approaches using S. cerevisiae strains where the plasma membrane levels of PIP 2 are known to fluctuate under certain growth conditions. Inositol starvation of ino1 cells (36,43) and glucose starvation of respiratory deficient cells (45) were used to deplete plasma membrane levels of PIP 2 . After a 4-h starvation period of ino1 cells for inositol, a small but reproducible amount of profilin was found to translocate from membranes to the cytosol. Inositol starvation also causes fluctuations in the plasma membrane levels of phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol, phosphatidylserine, and phosphatidate, and PIP (36). However, it was unlikely that the translocation of profilin was due to changes in these phospholipids. None of these phospholipids had a significant effect on binding of pure profilin to vesicles. In a more dramatic fashion, glucose starvation of respiratory deficient cells over a 20-min period caused a significant amount of profilin to translocate from membranes to the cytosol. Moreover, the addition of glucose back to glucose-starved cells caused profilin to translocate back to membranes from the cytosol. These translocations occurred too rapidly to be attributed to a genetic mechanism. Taken together, these results raised the suggestion that the plasma membrane concentration of PIP 2 played a role in the regulation of profilin localization. Our studies did not rule out the possible regulation of profilin localization by other plasma membrane-associated components or other cellular factors (e.g. cytoskeleton), which could have changed in response to inositol starvation of ino1 cells or glucose starvation of respiratory deficient cells.
In summary, this work addressed the fundamental question of where profilin was localized in the cell. We demonstrated that profilin was localized to the plasma membrane and cytosolic fractions and that profilin binding to the membrane was dependent on PIP 2 . It is known that PIP 2 levels change under FIG. 9. Effect of inositol starvation on the localization of profilin in ino1 cells. Inositol-supplemented ino1 cells were grown to the exponential phase of growth. Cells were washed and incubated for 4 h in fresh growth medium without inositol. Control cells were incubated in growth medium with inositol. Following the 4-h incubation, the subcellular localization of profilin from inositol-supplemented and inositol-starved cells was determined as described in the legend to Fig. 1. The data shown are representative of two independent experiments.
FIG . 10. Effect of glucose starvation on the localization of profilin in respiratory deficient cells. Glucose-supplemented cells were grown to the exponential phase of growth. Cells were washed and incubated for 30 min in fresh growth medium without glucose. Following the 30-min incubation, the culture was divided and glucose was added back (indicated by the arrow) to one of the cultures. The incubation of the cultures was continued for an additional 30 min. At the indicated time intervals, samples of the cultures were taken and analyzed for the association of profilin with the membrane and cytosolic fractions of the cell as described in the legend to Fig. 1. The data shown are representative of two independent experiments. various growth conditions. Our studies showed that conditions which affect PIP 2 metabolism played a role in profilin's localization. Future studies will be directed toward gaining insight into the role PIP 2 metabolism plays in profilin function.