A Role for GEA1 and GEA2 in the Organization of the Actin Cytoskeleton in Saccharomyces cerevisiae

Corresponding author: Dominick Pallotta, Laval University, Ste-Foy, Québec G1K 7P4, Canada., pallotta{at}rsvs.ulaval.ca (E-mail)

Communicating editor: B. ANDREWS

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Profilin is an actin monomer-binding protein implicated in the polymerization of actin filaments. In the budding yeast Saccharomyces cerevisiae, the pfy1-111 rho2 double mutant has severe growth and actin cytoskeletal defects. The GEA1 and GEA2 genes, which code for paralog guanosine exchange factors for Arf proteins, were identified as multicopy suppressors of the mutant phenotype. These two genes restored the polarized distribution of actin cortical patches and produced visible actin cables in both the pfy1-111 rho2 and pfy1 cells. Thus, overexpression of GEA1 or GEA2 bypassed the requirement for profilin in actin cable formation. In addition, gea1 gea2 double mutants showed defects in budding and in actin cytoskeleton organization, while overexpression of GEA1 or GEA2 led to the formation of supernumerary actin cable-like structures in a Bni1p/Bnr1p-dependent manner. The ADP-ribosylation factor Arf3p may be a target of Gea1p/Gea2p, since overexpression of ARF3 partially suppressed the profilin-deficient phenotype and a deletion of ARF3 exacerbated the phenotype of a pfy1-111 mutant. Gea1p, Gea2p, Arf1p, and Arf2p but not Arf3p are known to function in vesicular transport between the endoplasmic reticulum and the Golgi. In this work, we demonstrate a role for Gea1p, Gea2p, and Arf3p in the organization of the actin cytoskeleton.

THE organization of the actin cytoskeleton in all eukaryotic cells is a complex process involving many structural proteins and intracellular signaling molecules (SCHMIDT and HALL 1998 ; PRUYNE and BRETSCHER 2000A ; DUSTIN 2002 ). In spite of much recent progress, a detailed understanding of the mechanisms controlling actin organization is lacking. The budding yeast, Saccharomyces cerevisiae, has long been used as a model system for studying the organization of the actin cytoskeleton. It undergoes either polarized or isotropic growth during different stages of the life cycle. Isotropic growth during the G₁ phase of the cell cycle allows the mother cell to increase in size, while polarized growth leads to the formation of a bud during the S phase of the cell cycle and a septum before cell division (CHANT 1999 ; PRUYNE and BRETSCHER 2000B ). These changing patterns of growth require the controlled delivery of newly synthesized material to different parts of the cell.

In budding yeast, the actin cytoskeleton is responsible for polarized growth, organelle segregation, and endocytosis (AYSCOUGH 2000 ; SCHAFER 2002 ). The actin cytoskeleton in these cells is formed of cortical patches and cables, both of which contain F-actin and associated proteins. Actin patches are found associated with invaginations of the plasma membrane (MULHOLLAND et al. 1994 ), and they are involved in endocytosis (BRETSCHER et al. 1994 ; SCHOTT et al. 2002 ). On the other hand, actin cables are used for the transport of secretory vesicles from the Golgi to the plasma membrane and in the segregation of organelles (BRETSCHER 2003 ).

The organization of patches and cables changes during the life cycle and these changes are directly correlated with changing patterns of growth (AMBERG 1998 ; KARPOVA et al. 1998A ; YANG and PON 2002 ). During isotropic growth of a cell in the G₁ phase, for example, the patches are distributed throughout the cell and cables are randomly oriented. In cells with small buds, the cortical patches are found nearly exclusively in the bud while the cables are oriented longitudinally from the mother cell toward the patches. Late in the cell cycle, patches accumulate near the bud neck region, and the cables are directed toward them. During mating, an elongated cell, called a shmoo, is formed. In these cells, the cortical patches are located at the tip of the shmoo with the cables oriented toward the patches.

We used a genetic screen to identify proteins involved in the control of the actin cytoskeleton. Profilin is an actin monomer-binding protein implicated in the polymerization of actin filaments. Profilin-deficient yeast cells, pfy1, show a variety of morphological and growth abnormalities, such as the delocalization of cortical patches, the absence of actin cables, and a sensitivity to caffeine and NaCl (HAARER et al. 1990 ; MARCOUX et al. 1998 , MARCOUX et al. 2000 ). Overexpression of MID2, ROM1, ROM2, RHO2, SMY1, SYP1, and WSC1 suppresses the caffeine and NaCl sensitivity of the pfy1 strain. In addition to correcting the growth defects, these suppressors also partially repolarize the actin cortical patch distribution of pfy1 cells without the formation of visible actin cables. These results led to a model in which Mid2p, Rom1p, Rom2p, and Syp1p act through the Rho1p/Rho2p signaling pathway to repolarize cortical patches in pfy1 cells (MARCOUX et al. 1998 , MARCOUX et al. 2000 ).

Our goal in this work was to identify additional proteins that control actin cytoskeleton organization. We were particularly interested in proteins that act either downstream of Rho2p or parallel to the Rho2p pathway. We therefore carried out a genetic screen for multicopy suppressors using a yeast strain with mutations in the PFY1 and the RHO2 genes (pfy1-111 rho2). The GEA1 and GEA2 genes were identified as suppressors of the mutant phenotype. Gea1p and Gea2p have partially overlapping functions and are necessary for proper Golgi structure and function. They play an essential role in vesicular transport between the endoplasmic reticulum and the Golgi (PEYROCHE et al. 1996 , PEYROCHE et al. 2001 ; JACKSON and CASANOVA 2000 ; SPANG et al. 2001 ). In this work, we show a role for these two proteins in the control of the organization of the actin cytoskeleton in budding yeast.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains, media, and transformations:
Yeast strains used in this study are listed in Table 1. Cells were grown in rich YPD medium (1% yeast extract, 2% peptone, and 2% glucose) or in synthetic medium (SC; 0.67% yeast nitrogen base without amino acids, 2% glucose) supplemented with appropriate auxotrophic requirements. Suppressor selection was carried out in SC media containing 1.25 mg/ml caffeine. Cells were transformed by a modified lithium acetate procedure (KAISER et al. 1994 ).

To overexpress the GEA1 and GEA2 genes, the plasmids p88 and pGMS3, supplied by Catherine Jackson, were used (PEYROCHE et al. 2001 ). The gea1-4 allele was amplified by PCR from strain CJY062-10-3, which was also furnished by Catherine Jackson. The oligos 5'-CTGAGCTCATCGCGTTGGATGTAGGTTG and 5'-ACGGTACCGTTGGTCTACTCCTTCTCGAT were used. The resulting fragment was ligated into pRS426, and the recombinant plasmid was transformed into WT strain 22AB1-6A and into strains NHM125, DJP102, and PY3517. Strain PY3517 was generously provided by David Pellman.

Plasmid p4753 containing the FH1 and FH2 domains of Bni1p (Bni1pFH1FH2) was a generous gift from Charles Boone. The plasmid was transformed into WT strain 22AB1-6A and the actin cytoskeleton was observed at 30° by fluorescence staining with 1 µM FITC-conjugated phalloidin.

The coding sequences of ARF1, ARF2, and ARF3 were inserted into the BamHI site of multicopy Yep24 plasmid. The three genes were obtained by PCR amplification using the following oligos: ARF1, 5'-CATCCGCGGCCTAAGACAGT and 5'-TCCATCGACGTTGGCCTCTT; ARF2, 5'-CGGGATCCGAGTGACCTCGCTAGTAAGC and 5'-CGGGATCCAAGGTCCGCATGACTAAACG; and ARF3, 5'-CGGGATCCGGTCTCATAACCCTTTCTTG and 5'-CGGGATCCGGTGTATGCAGATTCAACACC.

The arf3 (EHZ100) strain was created by sporulation of the BY4743 diploid strain obtained from Research Genetics (Birmingham, AL), followed by spore dissection. The deletion of the ARF3 coding sequence in the pfy1-111 strain BHY46 was carried out using the PCR method (BAUDIN et al. 1993 ). The kanamycin gene and the flanking sequences of the ARF3 gene were amplified using the EHZ100 (arf3) strain DNA with the oligonucleotides 5'-CGGGATCCGGTCTCATAACCCTTTCTTG and 5'-CGGGATCCGGTGTATGCAGATTCAACACC. The PCR fragment was transformed into strain BHY46 to obtain the double mutant strain. This strain was verified by PCR amplifications of the two mutated loci.

pfy1-111 rho2 multicopy suppressor screen:
Strain NHM125 (pfy1-111 rho2) was transformed with the YEp24 multicopy library constructed from DNA of the S288C strain (CARLSON and BOTSTEIN 1982 ). About 8000 transformants were selected on SC-URA medium at 30° and subsequently replica plated to SC-URA medium at 37° and SC-URA medium containing 1.25 mg/ml caffeine. A total of 26 transformants grew on caffeine and at 37°. Plasmids were recovered and sequenced.

Phalloidin staining:
Staining of actin filaments was carried out according to a modified protocol for visualizing actin filaments (KARPOVA et al. 1998A ). Cells were grown to exponential phase at 30°. Half of the culture was fixed and stained; the rest was diluted, incubated for 2 hr at 30°, and then shifted to 37° for 3 hr. Cells were fixed at room temperature by the addition of 37% formaldehyde to 3.7% final concentration, directly in the culture media. The cells were washed in PBS and stained with 1 µM FITC-conjugated phalloidin for 90 min on ice or with 0.3 µM AlexaFluor 488-conjugated phalloidin (Molecular Probes, Eugene, OR) for 30 min at room temperature in the dark. Cells were washed repeatedly with PBS and then observed and photographed with a Leitz fluorescence microscope.

Analysis of cell morphology:
Cells were grown to logarithmic phase in YPD or SC medium, fixed with 3.7% formaldehyde, and observed with a Plan Apo x100 objective under a Leitz microscope equipped with Nomarski optics. For the determination of the percentage of cells with a given phenotype, a minimum of 100 cells was observed.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of GEA1 and GEA2 as multicopy suppressors:
The pfy1-111 allele has a mutation that causes a single amino acid substitution in yeast profilin (HAARER et al. 1993 ). Cells carrying this mutation grow comparably to wild-type cells at 30°, but exhibit a slow-growth phenotype, fewer actin cables, and a partial depolarization of actin patches at 37°. The pfy1-111 rho2 double mutant strain, DJP125, has a more severe phenotype. At 37°, the pfy1-111 rho2 cells are large and have no visible actin cables, a completely depolarized actin cytoskeleton, and a marked sensitivity to caffeine (MARCOUX et al. 2000 ). The pfy1-111 rho2 strain was transformed with the YEp24 multicopy library (CARLSON and BOTSTEIN 1982 ). We identified the GEA2 gene as a multicopy suppressor of the caffeine-sensitive phenotype; as expected, PFY1 and RHO2 were also identified in our screen. Since the GEA1 and GEA2 genes have 51% sequence identity and have at least partially overlapping functions, we transformed the pfy1-111 rho2 strain with a multicopy plasmid containing the GEA1 gene. The resulting cells grew well on plates containing 1.25 mg/ml caffeine, indicating that the GEA1 and the GEA2 genes suppressed the caffeine sensitivity of the pfy1-111 rho2 strain (Fig 1A).

View larger version (66K): In this window In a new window Download PPT slide   Figure 1. Suppression of the pfy1 and pfy1-111 rho2 mutant phenotypes by GEA1 and GE2A overexpression. (A) Suppression of the pfy1-111 rho2 caffeine sensitivity by GEA1 and GEA2. Wild-type (WT) and mutant (pfy1-111 rho2) cells and mutant cells overexpressing either GEA1 (pfy1-111 rho2 + GEA1) or GEA2 (pfy1-111 rho2 + GEA2) were tested for caffeine sensitivity. A total of 1 x 105 cells and serial 10-fold dilutions were spotted on SC-URA plates containing 1.25 mg/ml caffeine. The plates were incubated at 30° for 48–72 hr. (B) Overexpression of GEA2 in pfy1 and pfy1-111 rho2 cells repolarizes the cortical actin patches and restores actin cables. Cells from exponential phase cultures, grown at 30°, were stained with FITC- or AlexaFluor488-conjugated phalloidin to visualize actin distribution. WT and mutant (pfy1 and pfy1-111 rho2) cells and mutant cells overexpressing GEA2 (pfy1 + GEA2 and pfy1-111 rho2 + GEA2) are shown. Actin cables are visible in WT cells and mutant cells overexpressing GEA2.

We also tested the effect of the overexpression of the GEA1 and GEA2 genes in pfy1-111 and pfy1 strains. Cells carrying the pfy1 deletion have severe growth defects. They grow with a doubling time of 6 hr in minimal medium at 30° and do not grow at 37° or in the presence of 1.5 mg/ml caffeine (MARCOUX et al. 2000 ). The pfy1-111 and pfy1 strains with multicopy plasmids containing either the GEA1 or the GEA2 gene grew nearly as well as wild-type cells under all conditions, including growth at 37° or in the presence of 1.25 mg/ml caffeine (results not shown). The GEA1 and GEA2 genes can thus suppress the growth deficiencies associated with the pfy1, pfy1-111, and pfy1-111 rho2 mutations.

Gea1p and Gea2p contain the Sec7 domain that is required for guanosine exchange factor (GEF) activity for the small Ras-like GTPases in the Arf protein family (FRANZUSOFF and SCHEKMAN 1989 ; PEYROCHE et al. 1999 ; ROTH 1999 ; JACKSON and CASANOVA 2000 ). Syt1p, another yeast protein containing the Sec7 domain, has GEF activity on Arf1p (JONES et al. 1999 ; SPANG et al. 2001 ). The SYT1 gene was not selected in our genetic screen, and a direct test showed that overexpression of SYT1 was unable to suppress the morphological and growth deficiencies in pfy1 and pfy1-111 rho2 strains. These results show that SYT1 is not a suppressor of the profilin-deficient phenotype (results not shown).

Restoration of actin cables and polarized patch distribution by GEA1 and GEA2:
An overexpression of the genes MID2, ROM1, ROM2, SMY1, SYP1, and RHO2 can partially correct many of the abnormal phenotypes associated with profilin-deficient phenotype (MARCOUX et al. 1998 , MARCOUX et al. 2000 ). The pfy1 cells with any of these suppressors are nearly normal in size, although they are still somewhat larger than wild-type cells. There are no visible actin cables in cells with the suppressors, but the actin cortical patches are partially to completely polarized. To investigate the basis of the GEA1/GEA2 suppression in pfy1 cells, we used fluorescence microscopy to assay actin cytoskeleton organization. Identical results were obtained for both genes and only the results for GEA2 are shown.

Overexpression of either GEA1 or GEA2 in pfy1 cells resulted in a wild-type distribution of actin patches, with concentrations in the bud and at the septum in dividing cells. Remarkably, the pfy1 cells overexpressing GEA1 or GEA2 also showed visible actin cables (Fig 1B). This result was unexpected since cables are not observed in pfy1 cells overexpressing MID2, ROM1, ROM2, SMY1, SYP1, or RHO2 (MARCOUX et al. 2000 ). We therefore examined the actin cytoskeleton in the double mutant pfy1-111 rho2 for the presence of actin cables. Cells overexpressing either GEA1 or GEA2 had polarized actin patches and visible actin cables. The GEA1 or GEA2 genes are therefore excellent suppressors of the actin cytoskeletal defects found in profilin-deficient cells and in cells carrying a profilin conditional mutation combined with rho2.

A mutant gea1 allele is not a suppressor of the profilin-deficient phenotype:
Gea1p and Gea2p play a role in protein secretion, Golgi organization, and retrograde vesicular transport between the Golgi and the endoplasmic reticulum (ER). We wished to determine whether a gea1 allele that is defective in these functions is also an ineffective suppressor of actin cytoskeletal defects. The gea1-4 allele has multiple amino acid substitutions, including two in the Sec7 region, which is the likely catalytic domain for GEF activity (PEYROCHE et al. 2001 ). A strain carrying this allele is defective in protein secretion, glycosylation, and Golgi organization (PEYROCHE et al. 2001 ; SPANG et al. 2001 ). The gea1-4 allele was cloned into the multicopy plasmid pRS426 and inserted separately into pfy1 and pfy1-111 rho2 cells. The gea1-4 allele was unable to correct the growth and actin cytoskeleton defects associated with these strains (results not shown).

Actin cytoskeleton defects in gea mutant strains:
Our results showing that GEA1 and GEA2 are multicopy suppressors of the actin defects in profilin-deficient cells suggest that cells carrying mutations in the GEA genes might have problems with cell polarity and actin cytoskeleton organization. Since yeast strains deleted for either GEA1 or GEA2 have a normal phenotype, we assayed the gea1-4 gea2 and gea1-6 gea2 double mutants for actin cytoskeletal defects. The gea2 strain was used for comparison. Since the gea1-6 gea2 strain grows at 30° but is not viable at 37°, we grew all three strains at 30° and then transferred them to 37° for 3 hr before examination. The morphology of the cells at 30° and 37° is shown (Fig 2A and Fig B).

View larger version (90K): In this window In a new window Download PPT slide   Figure 2. Morphology and actin cytoskeleton in gea2, gea1-4 gea2, and gea1-6 gea2 strains. (A and B) Morphology of gea2, gea1-4 gea2, and gea1-6 gea2 strains. Cells from exponential phase culture grown at 30° (A) and 37° (B) were observed by Nomarski optics. The gea1-4 gea2 cells show an aberrant budding pattern at 37° but not at 30°. The gea1-6 gea2 cells are larger and rounder at both temperatures and show a double-bud pattern at 37°. (C) Actin cytoskeleton in gea2, gea1-4 gea2, and gea1-6 gea2 cells. Cells from exponential phase cultures were grown at 30° and then incubated at 37° for 2.5 hr. Actin distribution was visualized with FITC- or AlexaFluor488-conjugated phalloidin.

As expected from previous work, the gea2 strain had a normal morphology at 30° and 37° (PEYROCHE et al. 1996 ; SPANG et al. 2001 ). Likewise, the gea1-4 gea2 strain appeared normal at the permissive temperature. However, at the restrictive temperature 40% of the budding cells had two buds. The gea1-6 gea2 mutant cell had an even more severe phenotype. At the permissive temperature most of the cells were larger and rounder than normal, a phenotype often found in cells with a defective actin cytoskeleton (KARPOVA et al. 1998B ). At the restrictive temperature, all of the cells had the round phenotype and 25% of the budding cells had two or more buds. In some cells the buds were juxtaposed and in others they showed a bipolar distribution.

The actin cytoskeleton was observed in these same cells. At the permissive temperature, the actin cortical patches and cables had a normal distribution in the three mutant cells (results not shown). At 37°, there was no change in the actin cytoskeleton for the gea2 cells. At this temperature, however, the gea1-4 gea2 and gea1-6 gea2 cells had a partially depolarized actin cytoskeleton (Fig 2C). Cortical patches were found in both the mother cells and the small buds. The actin cables were visible in only some cells and were generally shorter when they were present. These results show that cells require a functional copy of either GEA1 or GEA2 to maintain a normal cell polarity and actin cytoskeleton.

Actin cable formation by Gea1p/Gea2p requires functional formins:
The budding yeast contains two formins, Bni1p and Bnr1p, that are essential for the formation of actin cables. Neither BNI1 nor BNR1 is essential for growth, but cells lacking both genes are nonviable (KAMEI et al. 1998 ; VALLEN et al. 2000 ). Cells carrying a temperature-sensitive allele for BNI1 and a deleted BNR1 gene (bni1-FH2#1 bnr1) are viable and have a normal actin cytoskeleton at 24°. At the nonpermissive temperature, these cells have cortical patches but no actin cables (EVANGELISTA et al. 2002 ; SAGOT et al. 2002A ). We tested whether the formation of actin cables by overexpression of GEA1/2 was dependent on the presence of the yeast formins by overexpressing these two genes separately in the bni1-FH2#1 bnr1 cells. After 1 hr at 37°, the double mutant cells had clearly visible but partially depolarized cortical patches and no actin cables. Overexpression of GEA1 or GEA2 had no effect on the actin cytoskeleton; the cortical patches were still partially delocalized and no actin cables were visible (Fig 3A). Thus, the formation of actin cables by overexpression of GEA1 or GEA2 requires the presence of functional formins.

View larger version (93K): In this window In a new window Download PPT slide   Figure 3. The actin cytoskeleton in a bni1 bnr1 mutant and in WT cells overexpressing GEA1, GEA2, or a BNI1 fragment. (A) GEA1 and GEA2 genes are not suppressors of the bni1-FH2#1 bnr1 mutation. Actin distribution was visualized with FITC- or AlexaFluor488-conjugated phalloidin in bni1-FH2#1 bnr1 cells and bni1-FH2#1 bnr1 cells overexpressing GEA1 or GEA2. (B) Supernumerary actin cable formation in WT cells overexpressing Bni1pFH1FH2, GEA1, or GEA2. Actin distribution was visualized in wild type and wild-type cells overexpressing Bni1pFH1FH2, GEA1, or GEA2 grown at 30°. The arrows indicate aberrant actin cables.

Formation of supernumerary cables by overexpression of GEA1/2:
We showed that formation of cables by overexpression of GEA1/2 is profilin independent but formin dependent. We next examined the effect of GEA1/2 overexpression on cable formation in WT cells. Cells overexpressing either of these genes had normal growth rates and morphologies at 30°. The cortical actin patches appeared normal in number, size, and localization during the cell cycle. Actin cables were clearly visible, but they were different from the cables in WT cells (Fig 3B). In most cells overexpressing GEA1 or GEA2, the cables were longer and more convoluted than those in WT cells. This often gave the appearance of a network of actin cables. In some cells, cable-like structures with a circular appearance were seen. This is in contrast to actin cables in WT cells that are usually linear structures. A network of convoluted actin cable-like structures was also seen with overexpression of GEA1 or GEA2 at 37°. Thus, overexpression of GEA1 or GEA2 in the presence of yeast formins caused the formation of actin cables in profilin-deficient cells and supernumerary cables in WT cells.

The large Bni1 protein contains a GTPase-binding domain, a profilin-binding formin homology 1 domain (FH1), a formin homology 2 domain (FH2) involved in actin polymerization, and a Dia-autoregulatory domain (WASSERMAN 1998 ; RIDLEY 1999 ; ZELLER et al. 1999 ; ALBERTS 2001 ; EVANGELISTA et al. 2003 ). A Bni1p fragment containing the FH1 and FH2 domains (Bni1pFH1FH2) can nucleate unbranched actin filaments in vitro (EVANGELISTA et al. 1997 ; SAGOT et al. 2002B ) and assembles ectopic filamentous actin structures in vivo (PRUYNE et al. 2002 ; SAGOT et al. 2002B ). The Bni1pFH1FH2-induced cables are longer and more convoluted than the cables seen in untreated WT cells. The supernumerary actin filaments formed by the expression of Bni1pFH1FH2 under the control of the inducible GAL1 promoter resembled the actin structures formed by overexpression of GEA1 or GEA2 in WT cells (Fig 3B). These results concerning the formation of supernumerary cables by overexpression of Bni1pFH1FH2, GEA1, and GEA2 in WT cells are compatible with a model in which GEA1 and GEA2 affect actin cable formation by acting through Bni1p.

ARF3 is a multicopy suppressor of the profilin-deficient phenotype:
Arf1p, Arf2p, and Arf3p are small guanosine nucleotide-binding proteins that are members of the ADP-ribosylation factor (ARF) family in yeast (MOSS and VAUGHAN 1998 ; ROTH 1999 ). To determine whether the Gea1 and Gea2 proteins act through an Arf protein to suppress the actin cytoskeleton defects, we overexpressed ARF1, ARF2, and ARF3 separately in pfy1-111 cells. Only the overexpression of ARF3 corrected the caffeine sensitivity of the pfy1-111 cells (Fig 4A). We then tested the effects of overexpression of the Arf family members in pfy1 and pfy1-111 rho2 cells. Overexpression of ARF1 or ARF2 in either of these mutant cells did not correct the growth and actin cytoskeleton defects. The cells were large, did not grow at 37°, and had a depolarized actin cytoskeleton. ARF3, on the other hand, was a partial suppressor (Fig 4B). The pfy1 cells overexpressing ARF3 were smaller than mutant cells alone and had polarized actin patches. However, no visible actin cables were seen. In pfy1-111 rho2 cells overexpressing ARF3, actin patches were partially repolarized, but no actin cables were visible. In addition, 40% of the cells had a single elongated bud or had multiple buds. This polarity effect was restricted to pfy1-111 rho2 cells since overexpression of ARF3 in pfy1 cells did not give this phenotype.

View larger version (85K): In this window In a new window Download PPT slide   Figure 4. Overexpression of ARF1, ARF2, and ARF3 in profilin-deficient cells. (A) Overexpression of ARF3 suppresses the pfy1-111 caffeine sensitivity. Profilin mutant cells (pfy1-111) or profilin mutant cells overexpressing ARF3 (pfy1-111 + ARF3) were tested for caffeine sensitivity. A total of 1 x 105 cells and serial 10-fold dilutions were spotted on SC-URA plates containing 1.25 mg/ml caffeine. The plates were incubated at 30° for 48–72 hr. (B) Overexpression of ARF3 partially repolarizes the cortical patches in profilin-deficient cells. Two profilin mutant cell types (pfy1 and pfy1-111 rho2) with or without the overexpression of ARF1, ARF2, and ARF3 were grown exponentially at 30° and stained for actin distribution with FITC- or AlexaFluor488-conjugated phalloidin. Neither ARF1 nor ARF2 overexpression in pfy1 and pfy1-111 rho2 strains repolarizes the actin cytoskeleton. ARF3 overexpression in pfy1 cells partially repolarizes the cortical patches without restoring actin cables. In pfy1-111 rho2 cells overexpressing ARF3, actin cortical patches localize mainly in the bud. Some of these cells have elongated buds.

arf3 shows a genetic interaction with pfy1-111:
ARF3 is a partial suppressor of the profilin-deficient phenotype, suggesting that this gene is involved in polarity and actin cytoskeleton organization. We therefore examined the phenotype of the arf3 strain (Fig 5). At 30°, this strain had a normal morphology and actin cytoskeleton organization. At 37°, the cells had a normal size but 20% of the budding cells had two or more buds. These results suggest a role for Arf3p in cell polarity. We then constructed the double mutant pfy1-111 arf3 to look for genetic interactions. The double mutant was viable and grew well at all temperatures. At 30°, 30% of the budding cells had two or more buds. This result was not seen with cells carrying either the arf3 or the pfy1-111 mutation. At 37°, again 30% of the cells had multiple buds. In some cases the second bud was formed directly at the tip of the first bud. In other cases, long dumbbell-like structures were formed that have buds at both ends of the cell (Fig 5). These results demonstrate a genetic interaction between the arf3 and pfy1-111 alleles.

View larger version (90K): In this window In a new window Download PPT slide   Figure 5. Morphology of pfy1-111, arf3, and pfy1-111 arf3 strains. Exponentially growing mutant strains at 30° and 37° were visualized by Nomarski optics. The pfy1-111 cells are nearly normal at 30°, but large and round at 37°. The arf3 cells are normal at 30° but show double-bud pattern at 37° in 20% of the cells. The pfy1-111 arf3 double mutant cells show an aberrant morphology at both temperatures, with 30–35% of the cells having two or more buds.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Gea1p and Gea2p play an essential role in secretion (FRANZUSOFF and SCHEKMAN 1989 ; PEYROCHE et al. 1996 ; WOLF et al. 1998 ). These proteins can functionally replace each other, but at least one of the two proteins is needed for viability. Cells with a gea2 deletion and a conditional gea1 mutation have aberrant ER and Golgi structures. They are defective in ER-to-Golgi transport, intra-Golgi transport, and retrotransport between the Golgi and the ER (PEYROCHE et al. 2001 ; SPANG et al. 2001 ). We present evidence in this article that Gea1p and Gea2p are also involved in actin cytoskeleton organization and cell polarity. First, we identified GEA1 and GEA2 as multicopy suppressors of the pfy1, pfy1-111, and pfy1-111 rho2 mutations. Overexpression of GEA1 or GEA2 restored visible actin cables and polarized actin cortical patches in these cells. By contrast, other high-copy suppressors of profilin mutant phenotypes, such as MID2, failed to restore actin cables (MARCOUX et al. 1998 , MARCOUX et al. 2000 ). Second, strains deleted for GEA2 and carrying a mutant allele of GEA1 had defects in budding and in actin cytoskeleton organization. Finally, overexpression of GEA1 or GEA2 in WT cells led to the formation of supernumerary actin cable-like structures.

Two mechanisms for actin polymerization in budding yeast are known. The Arp2/3 complex contains seven proteins, including the actin-related proteins Arp2 and Arp3 (WINTER et al. 1997 , WINTER et al. 1999 ). This complex is associated with the proteins Abp1, Bee1/Las17, Myo3, Myo5, Pan1, and Vrp1 in the cortical patch region of the cell (SCHOTT et al. 2002 ). The actin filaments formed with actin monomers and the Arp2/3 complex are branched structures (VOLKMANN et al. 2001 ; POLLARD and BELTZNER 2002 ). Several lines of evidence indicate that the Arp2/3 complex and its associated proteins are responsible for cortical patch assembly in budding yeast and that the complex is dispensable for cable assembly (WINTER et al. 1997 , WINTER et al. 1999 ; EVANGELISTA et al. 2002 ).

Recently, the yeast formins Bni1p and Bnr1p were identified as a second system for the polymerization of actin. The key finding was the in vitro polymerization of actin by the formins (EVANGELISTA et al. 2002 ; LEW 2002 ; SAGOT et al. 2002B ). This process is stimulated by profilin and is independent of the Arp2/3 complex. The actin filaments formed are linear structures. Yeast cells deficient for the formins have cortical patches but no cables. Overexpression of a truncated Bni1p causes the formation of supernumerary actin cables (EVANGELISTA et al. 2002 ; WINTER et al. 1999 ). Altogether, these results suggest that the Bni1p/Bnr1p-dependent nucleation of actin is responsible for the formation of actin cables in vivo.

In this work, we show the formation of actin cable-like structures by overexpression of GEA1 or GEA2. Actin cable formation by Gea1p/Gea2p is profilin independent, since normal-appearing actin cables were formed in profilin-deficient cells. The actin structures, however, are Bni1p/Bnr1p dependent, since no cable-like structures were formed in cells lacking these proteins.

Gea1p and Gea2p share a region of sequence similarity of 200 amino acids that is also found in Sec7p and Syt1p (JONES et al. 1999 ). This region, called the Sec7 domain, is responsible for the GEF activity on the small GTP-binding proteins of the Arf family, which in S. cerevisiae is composed of three members, Arf1p, Arf2p, and Arf3p (ROTH 1999 ; JACKSON and CASANOVA 2000 ). These proteins and their equivalents in other cells were named for their activity as cofactors for the ADP ribosylation of the -factor of G proteins by the cholera toxin (KAHN and GILMAN 1984 ). Most is known about Arf1p and Arf2p. These proteins are 96% identical and have equivalents in numerous eukaryotic cells (SEWELL and KAHN 1988 ; STEARNS et al. 1990A , STEARNS et al. 1990B ). The S. cerevisiae ARF1 and ARF2 genes have at least a partially redundant function. A deletion of either gene is viable, while the deletion of both genes is lethal. This lethality can be corrected by the expression of mammalian ARF proteins (KAHN et al. 1991 ; LEE et al. 1992 ). The S. cerevisiae Arf1p and Arf2p are 75% identical to the mammalian class I proteins, ARF1–ARF3. S. cerevisiae Arf1p and Arf2p and mammalian ARF1 are necessary for intracellular transport and secretion (ROTH 1999 ; SPRINGER et al. 1999 ; YAHARA et al. 2001 ). They bind to Golgi membranes and are involved in vesicle coat formation. Like other members of the superfamily of small GTPase related to Ras, these proteins bind GTP or GDP. The activated GTP form binds to membranes and recruits coat proteins for the formation of transport vesicles.

Little is known about the S. cerevisiae Arf3p. Only 54% identical to Arf1p and Arf2p, it is not essential for viability and cannot correct the lethality of an arf1 arf2 double mutation. Arf3p is not involved in ER-to-Golgi transport, indicating that it plays a role different from that of the other Arf members in yeast (LEE et al. 1994 ). In this work we present results suggesting a role for the S. cerevisiae Arf3p in the control of cell polarity and the organization of the actin cytoskeleton. First, ARF3 was a partial suppressor of the phenotypes associated with pfy1 and pfy1-111 rho2 cells. In contrast, ARF1 and ARF2 were not suppressors. At 37°, the arf3 cells often had multiple buds. We also showed a genetic interaction between pfy1-111 and arf3. The double mutant cells often had multiple buds at either 30° or 37°. The effect was more dramatic at the higher temperature where elongated cells containing two or more buds were observed frequently. These defects were not corrected by overexpression of either GEA1 or GEA2.

Our results suggest a model in which Gea1p and Gea2p act, at least partially, through Arf3p to correct the phenotypes associated with the profilin-defective cells pfy1 and pfy1-111 rho2. It remains to be determined experimentally that the Gea proteins have GEF activity on Arf3p. It is also necessary to establish whether all the changes associated with overexpression of GEA1 and GEA2 result from an activation of Arf3p. GEA1 and GEA2 are excellent suppressors of the profilin-deficient phenotype. The cells are normal in size and have polarized actin cortical patches and visible actin cables. Overexpression of ARF3 corrects some but not all of the phenotypes associated with these cells. Actin cables, for example, are not produced by overexpression of ARF3.

One interpretation of this result is that the Gea proteins act on Arf3p and other Rho proteins to correct the actin cytoskeleton defects. It should be noted that in our experiments the Gea1/2 proteins do not act through Rho2p, since genetic suppression by these proteins is seen in the absence of Rho2p. An alternative interpretation of the differences observed with overexpression of GEA1/2 and ARF3 is the possibility of different levels of Arf3p-GTP in the cells. If the Gea proteins have GEF activity on Arf3p, then an overexpression of the Gea proteins should lead to an increase of Arf3p-GTP, the active form of Arf3p. An overexpression of ARF3 should increase intracellular Arf3p, but only some of this protein will be associated with GTP, while the rest will be linked to GDP. Further work is necessary to distinguish between these two alternatives.

Compared to the mammalian Arf proteins, the yeast Arf3p is most closely related to the class III ARF6 protein, with which it shares 60% sequence identity. The mammalian ARF6 protein is located at the cell periphery, and it cycles between the plasma membrane and endosomal vesicles (D'SOUZA-SCHOREY et al. 1995 ; PETERS et al. 1995 ). It has a dual role in cells. It regulates membrane traffic between the plasma membrane and intracellular endosomal vesicles, and it affects the organization of the actin cytoskeleton (RADHAKRISHNA and DONALDSON 1997 ; RADHAKRISHNA et al. 1999 ). Cells expressing wild-type ARF6 form actin-containing surface projections upon treatment with the G protein activator aluminum fluoride (RADHAKRISHNA et al. 1996 ). The effect was not seen upon expression of an inactive form of the protein containing a GTP-binding-defective ARF6 mutation. Expression of a GTPase-defective mutant of ARF6, a constitutively active form of the protein, induces actin polymerization at the cell periphery. No actin rearrangements are seen with the expression of wild-type mammalian ARF1 or its GTPase-defective form, showing that cytoskeletal reorganization is specific to ARF6 (RADHAKRISHNA et al. 1999 ). As with other Rho-type protein, the activity of ARF6 is controlled by its binding to GTP or GDP. EFA6, which contains a Sec7 domain, is a guanine exchange factor for ARF6. Expression of EFA6 induces membrane ruffles that are inhibited by the coexpression of dominant-inhibitory mutant forms of ARF6 or Rac1, another Rho-type protein involved in actin cytoskeleton organization (FRANCO et al. 1999 ).

Collectively these results show that the mammalian ARF6 protein is involved in actin cytoskeletal functions. Our results suggest that the budding yeast protein Arf3p also plays a role in cell polarity and the organization of the actin cytoskeleton, possibly by activation via GEA1/GEA2. A link between Arf signaling and the actin cytoskeleton in yeast has already been suggested by the work on Gcs1p. This Golgi-located protein is a GAP for Arf1p (POON et al. 1996 ). In combination with Age2p, Gcs1p is involved in vesicular transport from the trans-Golgi network to the endosome (POON et al. 2001 ) and with Glo3p it is necessary for retrograde transport from the Golgi to the endoplasmic reticulum (POON et al. 1999 ). In addition, Gcs1p is likely involved in actin cytoskeleton organization since a gsc1 deletion strain has mislocalized actin patches and recombinant Gcs1p binds to actin filaments and stimulates actin polymerization in vitro (BLADER et al. 1999 ).

The mechanism by which Gea1p and Gea2p stimulate actin cable formation in a Bni1p/Bnr1p-dependent manner remains to be determined. An active Sec7 region in Gea1p, which is the probable catalytic domain for GEF activity, is important for actin cytoskeleton activity. The gea1-4 allele that carries mutations in the Sec7 region is unable to stimulate actin cable formation. In addition, Rho family G proteins bind to Bni1p and are likely responsible for their activation. Finally, Arf3p and Rho2p are partial suppressors of the cytoskeletal defects of profilin-deficient yeast cells. These results suggest a functional link between Gea1p, Gea2p, formins, Rho-type proteins, and the actin cytoskeleton.

	ACKNOWLEDGMENTS

We thank Charles Boone, Catherine L. Jackson, and David Pellman for their generous gifts of strains and plasmids, without which this work could not have been completed. This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada. Ewa Zakrzewska and Marjorie Perron were supported by fellowships from the Centre de Recherche sur la Structure, la Fonction et l'Ingénierie des Protéines (CREFSIP). Marjorie Perron was also supported by the Fonds Québécois de Recherche sur la Nature et les Technologies (FQRNT).

Manuscript received April 16, 2003; Accepted for publication July 6, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALBERTS, A. S., 2001  Identification of a carboxyl-terminal diaphanous-related formin homology protein autoregulatory domain. J. Biol. Chem. 276:2824-2830.[Abstract/Free Full Text]

AMBERG, D. C., 1998  Three-dimensional imaging of the yeast actin cytoskeleton through the budding cell cycle. Mol. Biol. Cell 9:3259-3262.[Free Full Text]

AYSCOUGH, K. R., 2000  Endocytosis and the development of cell polarity in yeast require a dynamic F-actin cytoskeleton. Curr. Biol. 10:1587-1590.[Medline]

BAUDIN, A., O. OZIER-KALOGEROPOULOS, A. DENOUEL, F. LACROUTE, and C. CULLIN, 1993  A simple and efficient method for direct gene deletion in Saccharomyces cerevisiae. Nucleic Acids Res. 21:3329-3330.[Free Full Text]

BLADER, I. J., M. J. COPE, T. R. JACKSON, A. A. PROFIT, and A. F. GREENWOOD et al., 1999  GCS1, an Arf guanosine triphosphatase-activating protein in Saccharomyces cerevisiae, is required for normal actin cytoskeletal organization in vivo and stimulates actin polymerization in vitro. Mol. Biol. Cell 10:581-596.[Abstract/Free Full Text]

BRETSCHER, A., 2003  Polarized growth and organelle segregation in yeast: the tracks, motors, and receptors. J. Cell Biol. 160:811-816.[Abstract/Free Full Text]

BRETSCHER, A., B. DREES, E. HARSAY, D. SCHOTT, and T. WANG, 1994  What are the basic functions of microfilaments? Insights from studies in budding yeast. J. Cell Biol. 126:821-825.[Free Full Text]

CARLSON, M. and D. BOTSTEIN, 1982  Two differentially regulated mRNAs with different 5' ends encode secreted with intracellular forms of yeast invertase. Cell 28:145-154.[Medline]

CHANT, J., 1999  Cell polarity in yeast. Annu. Rev. Cell Dev. Biol. 15:365-391.[Medline]

D'SOUZA-SCHOREY, C., G. LI, M. I. COLOMBO, and P. D. STAHL, 1995  A regulatory role for ARF6 in receptor-mediated endocytosis. Science 267:1175-1178.[Abstract/Free Full Text]

DUSTIN, M. L., 2002  Shmoos, rafts, and uropods—the many facets of cell polarity. Cell 110:13-18.[Medline]

EVANGELISTA, M., K. BLUNDELL, M. S. LONGTINE, C. J. CHOW, and N. ADAMES et al., 1997  Bni1p, a yeast formin linking cdc42p and the actin cytoskeleton during polarized morphogenesis. Science 276:118-122.[Abstract/Free Full Text]

EVANGELISTA, M., D. PRUYNE, D. C. AMBERG, C. BOONE, and A. BRETSCHER, 2002  Formins direct Arp2/3-independent actin filament assembly to polarize cell growth in yeast. Nat. Cell Biol. 4:260-269.[Medline]

EVANGELISTA, M., S. ZIGMOND, and C. BOONE, 2003  Formins: signaling effectors for assembly and polarization of actin filaments. J. Cell Sci. 116:2603-2611.[Abstract/Free Full Text]

FRANCO, M., P. J. PETERS, J. BORETTO, E. VAN DONSELAAR, and A. NERI et al., 1999  EFA6, a sec7 domain-containing exchange factor for ARF6, coordinates membrane recycling and actin cytoskeleton organization. EMBO J. 18:1480-1491.[Medline]

FRANZUSOFF, A. and R. SCHEKMAN, 1989  Functional compartments of the yeast Golgi apparatus are defined by the sec7 mutation. EMBO J. 8:2695-2702.[Medline]

HAARER, B. K., S. H. LILLIE, A. E. ADAMS, V. MAGDOLEN, and W. BANDLOW et al., 1990  Purification of profilin from Saccharomyces cerevisiae and analysis of profilin-deficient cells. J. Cell Biol. 110:105-114.[Abstract/Free Full Text]

HAARER, B. K., A. S. PETZOLD, and S. S. BROWN, 1993  Mutational analysis of yeast profilin. Mol. Cell. Biol. 13:7864-7873.[Abstract/Free Full Text]

JACKSON, C. L. and J. E. CASANOVA, 2000  Turning on ARF: the Sec7 family of guanine-nucleotide-exchange factors. Trends Cell Biol. 10:60-67.[Medline]

JONES, S., G. JEDD, R. A. KAHN, A. FRANZUSOFF, and F. BARTOLINI et al., 1999  Genetic interactions in yeast between Ypt GTPases and Arf guanine nucleotide exchangers. Genetics 152:1543-1556.[Abstract/Free Full Text]

KAHN, R. A. and A. G. GILMAN, 1984  Purification of a protein cofactor required for ADP-ribosylation of the stimulatory regulatory component of adenylate cyclase by cholera toxin. J. Biol. Chem. 259:6228-6234.[Abstract/Free Full Text]

KAHN, R. A., F. G. KERN, J. CLARK, E. P. GELMANN, and C. RULKA, 1991  Human ADP-ribosylation factors. A functionally conserved family of GTP-binding proteins. J. Biol. Chem. 266:2606-2614.[Abstract/Free Full Text]

KAISER, C. A., S. MICHAELIS and A. MITCHELL, 1994 Methods in Yeast Genetics: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

KAMEI, T., K. TANAKA, T. HIHARA, M. UMIKAWA, and H. IMAMURA et al., 1998  Interaction of Bnr1p with a novel Src homology 3 domain-containing Hof1p. Implication in cytokinesis in Saccharomyces cerevisiae. J. Biol. Chem. 273:28341-28345.[Abstract/Free Full Text]

KARPOVA, T. S., J. G. MCNALLY, S. L. MOLTZ, and J. A. COOPER, 1998a  Assembly and function of the actin cytoskeleton of yeast: relationships between cables and patches. J. Cell Biol. 142:1501-1517.[Abstract/Free Full Text]

KARPOVA, T. S., S. L. MOLTZ, L. E. RILES, U. GULDENER, and J. H. HEGEMANN et al., 1998b  Depolarization of the actin cytoskeleton is a specific phenotype in Saccharomyces cerevisiae. J. Cell Sci. 111:2689-2696.[Abstract]

LEE, F. J., J. MOSS, and M. VAUGHAN, 1992  Human and Giardia ADP-ribosylation factors (ARFs) complement ARF function in Saccharomyces cerevisiae. J. Biol. Chem. 267:24441-24445.[Abstract/Free Full Text]

LEE, F. J., L. A. STEVENS, Y. L. KAO, J. MOSS, and M. VAUGHAN, 1994  Characterization of a glucose-repressible ADP-ribosylation factor 3 (ARF3) from Saccharomyces cerevisiae. J. Biol. Chem. 269:20931-20937.[Abstract/Free Full Text]

LEW, D. J., 2002  Formin' actin filament bundles. Nat. Cell Biol. 4:E29-E30.[Medline]

MANNING, B. D., R. PADMANABHA, and M. SNYDER, 1997  The Rho-GEF Rom2p localizes to sites of polarized cell growth and participates in cytoskeletal function in Saccharomyces cerevisiae. Mol. Biol. Cell 8:1829-1844.[Abstract/Free Full Text]

MARCOUX, N., Y. BOURBONNAIS, P. M. CHAREST, and D. PALLOTTA, 1998  Overexpression of MID2 suppresses the profilin-deficient phenotype of yeast cells. Mol. Microbiol. 29:515-526.[Medline]

MARCOUX, N., S. CLOUTIER, E. ZAKRZEWSKA, P. M. CHAREST, and Y. BOURBONNAIS et al., 2000  Suppression of the profilin-deficient phenotype by the RHO2 signaling pathway in Saccharomyces cerevisiae. Genetics 156:579-592.[Abstract/Free Full Text]

MOSS, J. and M. VAUGHAN, 1998  Molecules in the ARF orbit. J. Biol. Chem. 273:21431-21434.[Free Full Text]

MULHOLLAND, J., D. PREUSS, A. MOON, A. WONG, and D. DRUBIN et al., 1994  Ultrastructure of the yeast actin cytoskeleton and its association with the plasma membrane. J. Cell. Biol. 125:381-391.[Abstract/Free Full Text]

PETERS, P. J., V. W. HSU, C. E. OOI, D. FINAZZI, and S. B. TEAL et al., 1995  Overexpression of wild-type and mutant ARF1 and ARF6: distinct perturbations of nonoverlapping membrane compartments. J. Cell Biol. 128:1003-1017.[Abstract/Free Full Text]

PEYROCHE, A., S. PARIS, and C. L. JACKSON, 1996  Nucleotide exchange on ARF mediated by yeast Gea1 protein. Nature 384:479-481.[Medline]

PEYROCHE, A., B. ANTONNY, S. ROBINEAU, J. ACKER, and J. CHERFILS et al., 1999  Brefeldin A acts to stabilize an abortive ARF-GDP-Sec7 domain protein complex: involvement of specific residues of the Sec7 domain. Mol. Cell 3:275-285.[Medline]

PEYROCHE, A., R. COURBEYRETTE, A. RAMBOURG, and C. L. JACKSON, 2001  The ARF exchange factors Gea1p and Gea2p regulate Golgi structure and function in yeast. J. Cell Sci. 114:2241-2253.

POLLARD, T. D. and C. C. BELTZNER, 2002  Structure and function of the Arp2/3 complex. Curr. Opin. Struct. Biol. 12:768-774.[Medline]

POON, P. P., X. WANG, M. ROTMAN, I. HUBER, and E. CUKIERMAN et al., 1996  Saccharomyces cerevisiae Gcs1 is an ADP-ribosylation factor GTPase-activating protein. Proc. Natl. Acad. Sci. USA 93:10074-10077.[Abstract/Free Full Text]

POON, P. P., D. CASSEL, A. SPANG, M. ROTMAN, and E. PICK et al., 1999  Retrograde transport from the yeast Golgi is mediated by two ARF GAP proteins with overlapping function. EMBO J. 18:555-564.[Medline]

POON, P. P., S. F. NOTHWEHR, R. A. SINGER, and G. C. JOHNSTON, 2001  The Gcs1 and Age2 ArfGAP proteins provide overlapping essential function for transport from the yeast trans-Golgi network. J. Cell Biol. 155:1239-1250.[Abstract/Free Full Text]

PRUYNE, D. and A. BRETSCHER, 2000a  Polarization of cell growth in yeast. J. Cell Sci. 113:571-585.[Abstract]

PRUYNE, D. and A. BRETSCHER, 2000b  Polarization of cell growth in yeast. I. Establishment and maintenance of polarity states. J. Cell Sci. 113:365-375.[Abstract]

PRUYNE, D., M. EVANGELISTA, C. YANG, E. BI, and S. ZIGMOND et al., 2002  Role of formins in actin assembly: nucleation and barbed-end association. Science 297:612-615.[Abstract/Free Full Text]

RADHAKRISHNA, H. and J. G. DONALDSON, 1997  ADP-ribosylation factor 6 regulates a novel plasma membrane recycling pathway. J. Cell Biol. 139:49-61.[Abstract/Free Full Text]

RADHAKRISHNA, H., R. D. KLAUSNER, and J. G. DONALDSON, 1996  Aluminum fluoride stimulates surface protrusions in cells overexpressing the ARF6 GTPase. J. Cell Biol. 134:935-947.[Abstract/Free Full Text]

RADHAKRISHNA, H., O. AL-AWAR, Z. KHACHIKIAN, and J. G. DONALDSON, 1999  ARF6 requirement for Rac ruffling suggests a role for membrane trafficking in cortical actin rearrangements. J. Cell Sci. 112:855-866.[Abstract]

RIDLEY, A. J., 1999  Stress fibres take shape. Nat. Cell Biol. 1:E64-E66.[Medline]

ROTH, M. G., 1999  Snapshots of ARF1: implications for mechanisms of activation and inactivation. Cell 97:149-152.[Medline]

SAGOT, I., S. K. KLEE, and D. PELLMAN, 2002a  Yeast formins regulate cell polarity by controlling the assembly of actin cables. Nat. Cell Biol. 4:42-50.[Medline]

SAGOT, I., A. A. RODAL, J. MOSELEY, B. L. GOODE, and D. PELLMAN, 2002b  An actin nucleation mechanism mediated by Bni1 and profilin. Nat. Cell Biol. 4:626-631.[Medline]

SCHAFER, D. A., 2002  Coupling actin dynamics and membrane dynamics during endocytosis. Curr. Opin. Cell Biol. 14:76-81.[Medline]

SCHMIDT, A. and M. N. HALL, 1998  Signaling to the actin cytoskeleton. Annu. Rev. Cell Dev. Biol. 14:305-338.[Medline]

SCHOTT, D., T. HUFFAKER, and A. BRETSCHER, 2002  Microfilaments and microtubules: the news from yeast. Curr. Opin. Microbiol. 5:564-574.[Medline]

SEWELL, J. L. and R. A. KAHN, 1988  Sequences of the bovine and yeast ADP-ribosylation factor and comparison to other GTP-binding proteins. Proc. Natl. Acad. Sci. USA 85:4620-4624.[Abstract/Free Full Text]

SPANG, A., J. M. HERRMANN, S. HAMAMOTO, and R. SCHEKMAN, 2001  The ADP ribosylation factor-nucleotide exchange factors Gea1p and Gea2p have overlapping, but not redundant functions in retrograde transport from the Golgi to the endoplasmic reticulum. Mol. Biol. Cell 12:1035-1045.[Abstract/Free Full Text]

SPRINGER, S., A. SPANG, and R. SCHEKMAN, 1999  A primer on vesicle budding. Cell 97:145-148.[Medline]

STEARNS, T., R. A. KAHN, D. BOTSTEIN, and M. A. HOYT, 1990a  ADP ribosylation factor is an essential protein in Saccharomyces cerevisiae and is encoded by two genes. Mol. Cell. Biol. 10:6690-6699.[Abstract/Free Full Text]

STEARNS, T., M. C. WILLINGHAM, D. BOTSTEIN, and R. A. KAHN, 1990b  ADP-ribosylation factor is functionally and physically associated with the Golgi complex. Proc. Natl. Acad. Sci. USA 87:1238-1242.[Abstract/Free Full Text]

VALLEN, E. A., J. CAVISTON, and E. BI, 2000  Roles of Hof1p, Bni1p, Bnr1p, and myo1p in cytokinesis in Saccharomyces cerevisiae. Mol. Biol. Cell 11:593-611.[Abstract/Free Full Text]

VOLKMANN, N., K. J. AMANN, S. STOILOVA-MCPHIE, C. EGILE, and D. C. WINTER et al., 2001  Structure of Arp2/3 complex in its activated state and in actin filament branch junctions. Science 293:2456-2459.[Abstract/Free Full Text]

WASSERMAN, S., 1998  FH proteins as cytoskeletal organizers. Trends Cell Biol. 8:111-115.[Medline]

WINTER, D., A. V. PODTELEJNIKOV, M. MANN, and R. LI, 1997  The complex containing actin-related proteins Arp2 and Arp3 is required for the motility and integrity of yeast actin patches. Curr. Biol. 7:519-529.[Medline]

WINTER, D. C., E. Y. CHOE, and R. LI, 1999  Genetic dissection of the budding yeast Arp2/3 complex: a comparison of the in vivo and structural roles of individual subunits. Proc. Natl. Acad. Sci. USA 96:7288-7293.[Abstract/Free Full Text]

WOLF, J., M. NICKS, S. DEITZ, E. VAN TUINEN, and A. FRANZUSOFF, 1998  An N-end rule destabilization mutant reveals pre-Golgi requirements for Sec7p in yeast membrane traffic. Biochem. Biophys. Res. Commun. 243:191-198.[Medline]

YAHARA, N., T. UEDA, K. SATO, and A. NAKANO, 2001  Multiple roles of Arf1 GTPase in the yeast exocytic and endocytic pathways. Mol. Biol. Cell 12:221-238.[Abstract/Free Full Text]

YANG, H. C. and L. A. PON, 2002