A Role for the Swe1 Checkpoint Kinase During Filamentous Growth of Saccharomyces cerevisiae

Corresponding author: Curt Wittenberg, Departments of Molecular Biology and Cell Biology, The Scripps Research Institute, 10550 North Torrey Pines Rd., La Jolla, CA 92037., curtw{at}scripps.edu (E-mail)

Communicating editor: S. SANDMEYER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In this study we show that inactivation of Hsl1 or Hsl7, negative regulators of the Swe1 kinase, enhances the invasive behavior of haploid and diploid cells. The enhancement of filamentous growth caused by inactivation of both genes is mediated via the Swe1 protein kinase. Whereas Swe1 contributes noticeably to the effectiveness of haploid invasive growth under all conditions tested, its contribution to pseudohyphal growth is limited to the morphological response under standard assay conditions. However, Swe1 is essential for pseudohyphal differentiation under a number of nonstandard assay conditions including altered temperature and increased nitrogen. Swe1 is also required for pseudohyphal growth in the absence of Tec1 and for the induction of filamentation by butanol, a related phenomenon. Although inactivation of Hsl1 is sufficient to suppress the defect in filamentous growth caused by inactivation of Tec1 or Flo8, it is insufficient to promote filamentous growth in the absence of both factors. Moreover, inactivation of Hsl1 will not bypass the requirement for nitrogen starvation or growth on solid medium for pseudohyphal differentiation. We conclude that the Swe1 kinase modulates filamentous development under a broad spectrum of conditions and that its role is partially redundant with the Tec1 and Flo8 transcription factors.

ORGANISMS are able to respond to changes in environmental conditions via a broad spectrum of cellular responses. The nature and magnitude of those responses differ depending upon the type of cell and the degree to which it must individually cope with those changes. The demands upon free-living unicellular organisms to adapt to environmental change are greater than those of metazoan cells, which may be buffered by homeostatic mechanisms provided by the organism as a whole. Fungi, as free-living organisms, have developed the flexibility to deal with substantial changes in their environment including the quantity, quality, and location of nutrients. The capacity of many species of fungi to switch between a cellular yeast form and filamentous forms in response to environmental changes is one example of that flexibility that is well documented (reviewed in MADHANI and FINK 1998 and BORGES-WALMSLEY and WALMSLEY 2000 ). In some fungal pathogens of plants and animals the ability to switch between the two forms is thought to be an important determinant of pathogenicity (reviewed in MADHANI and FINK 1998 and BORGES-WALMSLEY and WALMSLEY 2000 ). In the budding yeast Saccharomyces cerevisiae the differentiation from the yeast form to the filamentous form is thought to allow otherwise sessile cells to forage for nutrients (GIMENO et al. 1992 ).

Although most laboratory strains of S. cerevisiae are incapable of filamentous differentiation (LIU et al. 1996 ), invasive growth has now been documented in many wild strains and some laboratory strains. The differentiated state is known as haploid invasive growth in haploid cells (ROBERTS and FINK 1994 ) and as pseudohyphal growth in diploid cells (GIMENO et al. 1992 ). Despite the significant differences between both the nature of the invasive forms and the environmental stimuli that lead to their differentiation, haploid and diploid states share significant similarities. The differentiation of both types of invasive cells apparently involves the activation of a number of signaling pathways. Among those, the mitogen-activated protein kinase (MAPK) signaling pathway acting through the Tec1/Ste12 transcription factor and the cyclic-AMP/protein kinase A (cAMP/PKA) pathway acting through the Flo8 transcription factor are the best characterized (reviewed in MADHANI and FINK 1998 and BORGES-WALMSLEY and WALMSLEY 2000 ). There is, at least, partial overlap in the targets of the two transcription factors (LO and DRANGINIS 1998 ; MADHANI et al. 1999 ; PAN and HEITMAN 1999 ; RUPP et al. 1999 ).

A number of features of the cellular responses of haploid and diploid cells to filamentous differentiation signals are similar, although manifested to a different extent by each cell type (GIMENO et al. 1992 ; ROBERTS and FINK 1994 ). One of the most obvious differences between filamentous cells and the equivalent yeast form cells is in cell morphology. First, cells adopt a unipolar budding pattern characterized by the emergence of a new bud predominantly at one end of the cell. In daughter cells the new bud site is distal from the site of cytokinesis whereas in mothers that site is adjacent to the prior bud site. Furthermore, invasive cells adopt an elongated morphology. Finally, abscission of daughter cells from mothers is either delayed or suppressed. Together these changes lead to the formation of the filaments of cells that are characteristic of invasive development. Despite rather extensive characterization of these phenotypes, the extent to which these phenotypic changes affect the capacity of cells to invade solid medium is not entirely clear.

The differentiation of invasive cells is thought to involve significant alteration of cell cycle dynamics (KRON et al. 1994). However, much of the evidence for alteration of the cell cycle derives from studies of populations of cells that have been induced to acquire the invasive phenotype via mutations that activate one of the pathways thought to transduce the differentiation signal. When the transcription factor, PHD1, is ectopically expressed, it induces pseudohyphal differentiation (GIMENO and FINK 1994 ) and an associated shortening of G1 phase and elongation of G2 phase. Furthermore, enhancing the activity of G1 cyclins (CLN1 and CLN2) or decreasing the activity of mitotic cyclins (CLB1 and CLB2) via ectopic expression or mutation leads to similar changes in cell cycle dynamics as well as a number of morphological changes similar to those observed in pseudohyphal growth (KRON et al. 1994 ; LEW et al. 1997 ; AHN et al. 1999 ; EDGINGTON et al. 1999 ; LOEB et al. 1999B ). Finally, single amino acid changes in Cdc28 itself can lead to all of the aforementioned morphogenetic changes (EDGINGTON et al. 1999 ). Taken together these observations suggest that many and perhaps all of the changes in cell morphology and budding pattern associated with invasive growth are a consequence of changes in the activity of the Cdc28 cyclin-dependent kinase (CDK).

The morphogenetic changes associated with invasive growth are most consistent with hyperaccumulation of G1 cyclins and/or delayed activation of mitotic forms of the CDK. Although there is little reason a priori to exclude other mechanisms, inhibitory phosphorylation has received the most attention and is, therefore, the mechanism for which there is the best experimental understanding. Manipulation of Swe1, the budding yeast homolog of the fission yeast Wee1 protein kinase that catalyzes inhibitory tyrosine phosphorylation of Clb- associated forms of the Cdc28 CDK (BOOHER et al. 1993 ), is normally activated in response to morphogenetic aberrations (LEW and REED 1995 ; SIA et al. 1996 ). Its activity leads to a delayed mitotic progression resulting in elongation of G2 phase and to the acquisition of hyperpolarized cell morphology. Indeed, ectopic activation of Swe1 either by overexpression or by inactivation of its negative regulators, Hsl1 and Hsl7, has been shown to promote morphological changes reminiscent of invasive cells (MA et al. 1996 ; BARRAL et al. 1999 ; EDGINGTON et al. 1999 ; MCMILLAN et al. 1999A ; SHULEWITZ et al. 1999 ). Furthermore, inactivation of Hsl1 has been shown to promote invasive growth in a noninvasive yeast strain (EDGINGTON et al. 1999 ). Finally, Hsl7 has been proposed to repress pseudohyphal growth via inhibition of the Ste20 MAPK kinase (FUJITA et al. 1999 ). On the other hand, inactivation of Swe1 has been shown to have different effects on the invasive properties of haploid and diploid strains. Inactivation of SWE1 inhibits haploid invasive growth (EDGINGTON et al. 1999 ), whereas it has no effect on pseudohyphal differentiation in diploid cells (KRON et al. 1994 ; AHN et al. 1999 ).

We have reexamined the role of Swe1 and its regulators in filamentous differentiation using the well-characterized invasive strain of S. cerevisiae, 1278b (GIMENO et al. 1992 ; GRENSON et al. 1966 ). We have extended previous observations (EDGINGTON et al. 1999 ) by demonstrating that inactivation of either Hsl1 or Hsl7 enhances the invasive behavior of both haploid and diploid cells and that the enhancement is mediated via the Swe1 protein kinase. Whereas Swe1 contributes noticeably to the effectiveness of haploid invasive growth under all conditions tested, its contribution to pseudohyphal growth is limited to morphological differentiation under standard assay conditions consistent with the recent report of Ahn and colleagues (AHN et al. 1999 ). However, we find that Swe1 is essential for all aspects of pseudohyphal differentiation under a number of nonstandard assay conditions as well as in the absence of Tec1. Moreover, Swe1 is required for the induction of filamentation of both haploid and diploid cells promoted by butanol, a recently described phenomenon related to pseudohyphal development (LORENZ et al. 2000 ). Finally, inactivation of Hsl1 is insufficient to promote filamentous growth in the absence of both Tec1 and Flo8 or to bypass the requirement for nitrogen starvation or growth on solid medium for pseudohyphal differentiation. We conclude that Swe1 modulates the effectiveness of invasive growth under a broad spectrum of conditions.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast media:
Growth, manipulation, and construction of yeast strains was performed by standard procedures unless otherwise indicated. Synthetic low ammonia dextrose plate (SLAD) contained 50 µm ammonium sulfate, 6.8 g/liter yeast nitrogen base without amino acids or ammonium sulfate, 2% dextrose, and 2% washed agar (GIMENO et al. 1992 ). Synthetic low ammonia raffinose-galactose plates were prepared with 1% raffinose and 1% galactose instead of 2% dextrose as carbon source. Synthetic leucine dextrose plate and synthetic proline dextrose plate were prepared with 0.75 mg/ml leucine and 0.1 mg/ml proline, respectively, as the only nitrogen source. Agar was washed five times as 2% wt/vol suspension with deionized water for 30 min each wash. After the final wash, it was sterilized by autoclaving at 4% wt/vol in deionized water and diluting to 2% final concentration with 2x liquid media filter sterilized. Yeast extract, peptone, yeast nitrogen base are from Difco (Detroit), and agar is from Angus. Other reagents were obtained from Sigma (St. Louis).

Yeast strains and plasmids: All yeast strains were derived in the 1278b background (GRENSON et al. 1966 ; GIMENO et al. 1992 ). Yeast strains were as described in Table 1. Standard genetic methods were used for genetic manipulation of S. cerevisiae. Marker segregation or PCR assay determined genotypes. Strain RLVsc69 was constructed by using a PCR-mediated disruption method with the use of the G418 resistance cassette of plasmid pFA6-kanMX2 (WACH et al. 1994 ). Plasmids are described in Table 2.

Filamentous assay conditions:
The 1278b yeast strain is the prototypical strain for analysis of both haploid invasive and pseudohyphal growth (GIMENO et al. 1992 ). We refer to both of these responses using the general descriptions, filamentous or invasive growth.

For haploid invasive growth, the strains were patched on YEPD and pregrown at 30° for 2 days. Cells from these plates were plated in small patches on YEPD, YEP-GAL, or SC plates using a round glass rod to avoid disturbing the surface of the agar and incubated for 2 days at 30°. Haploid invasive growth was evaluated based upon three criteria. First, we analyzed the capacity of cells to resist removal from the surface of the plate under a stream of water (water washed). Next, the ability of cells to invade the agar was evaluated following "scrubbing" the surface of the plate with a gloved finger while washing under a stream of water (mechanically washed). Third, the cells remaining in the plate after mechanical washing were subjected to microscopic examination (invasive growth). Although these are qualitative assays, together they provide a number of criteria to judge the capacity of cells to differentiate. These include the ability of the cells to penetrate the agar, changes in cell morphology, and conversion of the budding pattern of cells from axial to unipolar budding.

For pseudohyphal development, strains were pregrown for 2 days on SD medium at 30° and then transferred to SLAD or modified SLAD plates, as indicated. To avoid disturbing the agar surface and the colony density-dependent inhibition of pseudohyphae formation (WRIGHT et al. 1993 ), single unbudded cells were carefully placed at 1 cm from each other by using the needle of a dissecting microscope. A total of 20–100 colonies, each derived from a single unbudded cell, were analyzed for each strain. The pictures shown are of representative colonies. Pseudohyphal growth was evaluated by multiple criteria. First, colony and cell morphology was monitored after one day of growth on SLAD plates (an assay of the extent of early morphological differentiation). The length/width ratio of cells forming filaments was evaluated after 1 day of growth on SLAD at 30° by measuring both dimensions of cells in imaged filaments. Second, cell shape and budding pattern were evaluated as described by MOSCH and FINK 1997 , using cells that were washed off plates and analyzed by microscopic examination. Third, the cell and colony morphology was evaluated after 5 days of growth on SLAD before (total growth) and after mechanically washing the noninvasive cells from the plate surface (invasive growth). Butanol-induced filamentous differentiation was evaluated as previously described (LORENZ et al. 2000 ).

Microscopy and imaging:
Microcolonies and colonies growing on plates were imaged from below through the agar and plastic petri dish with a Nikon Labophot microscope. Pixera VCS image-acquisition software and Pixera charge-coupled device camera were used to capture images at 1280 x 1024 resolution. Images were copied in Adobe Photoshop converted to gray scale, enhanced in contrast, and filtered to remove noise. Cells from suspension cultures were imaged with a Nikon Eclipse E800 microscope by using IPLAB Spectrum software and a Photmometrix Quantix camera. Images were flattened and copied into Photoshop. The size of the final images was reduced and then the cropped images were assembled into figures using Canvas 6 (Deneba).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Inactivation of Hsl1 and Hsl7 enhances filamentous differentiation via activation of the Swe1 protein kinase:
The similarity between the morphological phenotype of filamentous cells and that resulting from a decrease in the activity of mitotic forms of CDK has been well documented (KRON et al. 1994 ; LEW et al. 1997 ; AHN et al. 1999 ; EDGINGTON et al. 1999 ; LOEB et al. 1999B ). Furthermore, mutations that reduce the activity of Cdc28 or enhance the activity of Swe1, a CDK inhibitory kinase, have been shown to promote haploid invasive growth in a laboratory yeast strain that is only modestly invasive (EDGINGTON et al. 1999 ). This prompted us to assess the importance of Swe1 activity for pseudohyphal and haploid invasive growth using the well-characterized invasive yeast strain, 1278b (GIMENO et al. 1992 ).

Diploid cells carrying disrupted alleles of either HSL1 or HSL7, which encode negative regulators of Swe1 (MA et al. 1996 ; BARRAL et al. 1999 ; EDGINGTON et al. 1999 ; MCMILLAN et al. 1999A ; SHULEWITZ et al. 1999 ), or carrying disrupted alleles of both genes, were constructed and their behavior in pseudohyphal growth assays was examined (Fig 1A). Early in the process of pseudohyphal development on SLAD plates the hsl1/hsl1, hsl7/hsl7, and hsl1/hsl1 hsl7/hsl7 diploid cells were substantially more filamentous and exhibited more robust pseudohyphal differentiation than wild-type cells grown on the same medium (Fig 1A and Fig 2C; Table 3). This is an adaptive response as demonstrated by the analysis of cell shape and budding pattern of the same cells in liquid YEPD (Table 3). We conclude that the hsl1 and hsl7 mutations enhance not only the cellular phenotype of the filamentous cells but also the penetrance of that phenotype in both haploid and diploid cells. Moreover, the two mutations do not have additive effects consistent with the epistasis analysis of these mutations in vegetative cells, indicating that HSL1 and HSL7 lie in the same pathway (MCMILLAN et al. 1999A ; SHULEWITZ et al. 1999 ). We obtained similar results with haploid mutant strains; however, the enhancement of filamentous differentiation was less dramatic than that caused by the same mutations in diploid strains (data not shown).

View larger version (64K): In this window In a new window Download PPT slide   Figure 1. Inactivation of HSL1 and HSL7 enhances filamentous growth via a SWE1-dependent pathway. (A) Enhancement of pseudohyphal growth by inactivation of HSL1 and HSL7. Cell and colony morphology of the indicated strains is shown after 1 and 5 days of growth on a SLAD plate at 30° before and after mechanical washing of the plate to remove noninvasive cells. Wild-type (WT, RLVsc291), hsl1/hsl (RLVsc63), hsl7/hsl7 (RLVsc182), and hsl1/hsl1 hsl7/hsl7 (RLVsc284) strains are shown. (B) Inactivation of SWE1 suppresses the enhancement of pseudohyphal growth caused by hsl1 and hsl7. Pseudohyphal growth of wild-type (RLVsc291), hsl1/hsl1 swe1/swe1 (RLVsc297), hsl7/hsl7 swe1/swe1 (RLVsc285) and hsl1/hsl1 hsl7/hsl7 swe1/swe1 (RLVsc296) strains.

View larger version (32K): In this window In a new window Download PPT slide   Figure 2. The role of SWE1 during filamentous development. (A) Swe1 modulates the extent of haploid invasive growth. Invasive growth assay of wild-type (L5487) and swe1 (RLVsc47) strains. Yeast strains were pregrown on YEPD plates and patched onto fresh YEPD plates. Patches were photographed before and after water or mechanical washing to evaluate the extent of invasion and morphological changes following 2 days of incubation at 30°. (B) SWE1 is dispensable for the invasiveness of diploid strains undergoing pseudohyphal differentiation. Pseudohyphal growth of wild-type (RLVsc291) and swe1/swe1 (RLVsc64) strains. (C) SWE1 is required for cell elongation during pseudohyphal differentiation. The ratio of length to width of cells forming filaments was evaluated as described in MATERIALS AND METHODS.

Hsl1 and Hsl7 are thought to act as negative regulators of Swe1 by virtue of their ability to recruit Swe1 to the bud neck and regulate its phosphorylation and subsequent degradation (MCMILLAN et al. 1999A ; SHULEWITZ et al. 1999 ). The finding that inactivation of negative regulators of Swe1 enhances filamentous growth suggested to us that either these regulators or Swe1 might be a target of the filamentous growth signaling pathway. We asked whether the enhancement of haploid invasive and pseudohyphal growth observed in the hsl1 and hsl7 mutants was mediated via SWE1. To do so, homozygous diploid hsl1 swe1, hsl7 swe1, and hsl1 hsl7 swe1 strains were compared to the wild-type strain in pseudohyphal growth assay (Fig 1B). The enhancement of filamentous differentiation resulting from inactivation of HSL1, HSL7, or both was suppressed by inactivation of Swe1. The hsl1 swe1, hsl7 swe1, and hsl1 hsl7 swe1 mutants all exhibited a substantial reduction in cell elongation, extent of unipolar budding, and filamentation relative to the same strains having wild-type SWE1 (Fig 1 and data not shown). However, despite the fact that inactivation of SWE1 suppressed the enhancement of filamentation caused by the hsl1 and the hsl7 mutations, the capacity of homozygous diploid hsl1 swe1, hsl7 swe1, and hsl1 hsl7 swe1 strains to undergo pseudohyphal differentiation appeared to be unaffected relative to the wild-type strain (Fig 1B). We have shown that swe1 suppresses the enhancement of haploid invasive growth observed in hsl1, hsl7, and hsl1 hsl7 mutants (data not shown). Thus, the enhancement of both forms of invasive growth by inactivation of HSL1 and/or HSL7 appears to be a consequence of hyperactivation of SWE1.

Swe1 contributes to filamentous growth under standard assay conditions:
The finding that inactivation of SWE1 blocked the enhancement of pseudohyphal growth caused by inactivation of HSL1 or HSL7 without diminishing that capacity relative to wild-type cells suggested that SWE1 is not required for pseudohyphal differentiation. Although a significant effect of swe1 on the haploid invasive behavior of wild-type cells in the D273 background was reported by EDGINGTON et al. 1999 , those conclusions were based on microscopic examination since a relatively small number of those cells remain in the agar. Thus, the effect of inactivation of SWE1 on filamentous behavior of the more invasive 1278b background warranted further examination. We evaluated the behavior of haploid and diploid swe1 mutants in the haploid invasive and pseudohyphal growth assays. swe1 mutants show a modest but reproducible reduction in the effectiveness of haploid invasive growth based upon visual examination (Fig 2A). Consistent with previous results (EDGINGTON et al. 1999 ), the cells are more rounded and a smaller proportion of the cells remain in the agar following gentle or mechanical washing. In contrast, diploid wild-type and swe1/swe1 strains appear equally invasive after 5 and 10 days of growth (Fig 2B), consistent with the findings of AHN et al. 1999 . However, there is a noticeable and reproducible difference in the extent of cell elongation and unipolar budding in the swe1/swe1 cells as compared to the wild-type control (Table 3; Fig 2C). The length to width ratio of swe1/swe1 cells forming filaments on SLAD plate is lower than that of wild-type cells. Moreover, swe1/swe1 cells growing on SLAD exhibit less cell elongation and a lower frequency of unipolar budding in response to the pseudohyphal growth signals (nitrogen starvation and plate contact; Table 3). We conclude that swe1 is required for a full morphological response but not for substrate invasion during pseudohyphal growth and contributes to all aspects of haploid invasive growth under standard assay conditions.

Relationship between Swe1 and the MAPK and cAMP/PKA pathways:
Inactivation of SWE1 suppresses the strong enhancement of filamentous differentiation caused by hsl1 and hsl7 mutations and reduces the extent of morphological differentiation of otherwise wild-type pseudohyphal cells but does not appear to play a fundamental role in their invasive properties. There are two possible explanations for these results: either activation of Swe1 enhances the extent of filamentous differentiation via a pathway that is distinct from that used by wild-type cells or Swe1 activation modulates a pathway normally used to promote filamentous differentiation. To distinguish between these possibilities we examined the relationship between Swe1 and the signal transduction pathways involved in filamentous growth.

The cAMP and the MAPK pathways are known to play roles in transduction of the signals leading to differentiation of both pseudohyphal and haploid invasive cells (reviewed by MADHANI and FINK 1998 and BORGES-WALMSLEY and WALMSLEY 2000 ). Inactivation of HSL1 and/or HSL7, presumed negative regulators of SWE1, stimulates both types of filamentous growth, presumably through repression of mitotic forms of the CDK (the targets of Swe1). We, therefore, asked whether inactivation of HSL1 or HSL7 was sufficient for filamentous differentiation or whether its activity simply modulates the effects of those pathways. The genes encoding the Flo8 or Tec1 transcription factors, downstream targets of the cAMP and MAPK pathways (GAVRIAS et al. 1996 ; MOSCH and FINK 1997 ; LORENZ and HEITMAN 1998 ; RUPP et al. 1999 ), respectively, were individually inactivated in either wild-type or hsl1-deficient strains and then the resulting strains were evaluated for their capacity to undergo filamentous differentiation (Fig 3). As previously reported, inactivation of either FLO8 or TEC1 resulted in a loss or substantial reduction in the capacity of both haploid and diploid cells to grow invasively (GAVRIAS et al. 1996 ; LIU et al. 1996 ; MOSCH and FINK 1997 ; LORENZ and HEITMAN 1998 ). Furthermore, either the tec1 or the flo8 mutation results in a substantial reduction in the invasiveness of haploid hsl1 strains (Fig 3A). However, despite the comparable effect of those mutations on invasiveness, there was a noticeable difference in the effect of the hsl1 mutation on haploid flo8 and tec1 strains (Fig 3A). Inactivation of FLO8 had little effect on the morphology of haploid hsl1 mutants growing on solid rich media whereas inactivation of TEC1 strongly suppressed the morphological response of the hsl1 cells.

View larger version (67K): In this window In a new window Download PPT slide   Figure 3. Relationship between Swe1 activation and the filamentous signaling pathways. (A) Inactivation of HSL1 can partially rescue either the tec1 or flo8 defect in haploid invasive growth. Invasive growth of wild-type (WT, L55- 28), tec1 (RLVsc69), hsl1 tec1 (RLVsc227), flo8 (RLV-sc70), hsl1 flo8 (RLVsc166), and hsl1 tec1 flo8 (RLVsc-382) strains. (B) Inactivation of HSL1 can fully complement the pseudohyphal growth defect of either tec1/tec1 or flo8/flo8. Pseudohyphal growth of wild-type (WT, RLVsc291), tec1/tec1 (RLVsc294), hsl1/ hsl1 tec1/tec1 (RLVsc286), flo8/flo8 (RLVsc66), hsl1/hsl1 flo8/flo8 (RLVsc295), and hsl1/hsl1 tec1/tec1 flo8/flo8 (RLVsc392) strains. (C) Pseudohyphal development promoted by inactivation of HSL1 still requires the normal nitrogen starvation signal. Pseudohyphal growth assay of wild type (WT, RLVsc291) and hsl1/hsl1 (RLVsc63) on SLAD plate supplemented with a high concentration of ammonium sulfate (5 mM).

Inactivation of HSL1 or HSL7 was much more effective in suppressing the defect in pseudohyphal differentiation resulting from inactivation of either TEC1 or FLO8 than it was in suppressing the defect in haploid invasive growth caused by the same mutations (Fig 3B). Whereas both tec1/tec1 strains and flo8/flo8 strains exhibited a substantial defect in the capacity to grow invasively (GAVRIAS et al. 1996 ; LIU et al. 1996 ; MOSCH and FINK 1997 ; LORENZ and HEITMAN 1998 ), the same cells exhibited near-wild-type levels of pseudohyphal differentiation when HSL1 was also inactivated. This was most dramatically illustrated by comparison of the flo8/flo8 mutants to the hsl1/hsl1 flo8/flo8 mutants. Whereas the former shows neither morphological differentiation nor invasive growth, the latter is both highly elongated and highly invasive (Fig 3B). In contrast to the flo8/flo8 mutants, the tec1/tec1 strain retained some capacity to invade the agar and form filaments (MOSCH and FINK 1997 ; LORENZ and HEITMAN 1998 ). However, invasive cells were apparent only after the plate was washed (there were no invasive cells on the periphery of the colony). Furthermore, the extent of morphological differentiation and the frequency of invasive cells were significantly reduced relative to that observed with wild-type cells. In contrast, inactivation of HSL1 in a tec1 mutant suppressed both the morphological defect and the inability to invade (Fig 3B). Finally, the suppression of the filamentous growth defect of both haploid and diploid tec1 and flo8 strains caused by inactivation of HSL1 is dependent upon SWE1 (data not shown).

We have shown that the effect of inactivation of HSL1 upon both haploid invasive growth and pseudohyphal development is dependent upon SWE1 (Fig 1). Furthermore, it is sufficient to bypass the requirement for either TEC1, the primary output of the MAPK pathway, or FLO8, the primary output of the cAMP pathway, in mediating pseudohyphal differentiation (Fig 3B). Indeed, both haploid and homozygous diploid hsl1 tec1 swe1 and hsl1 flo8 swe1 strains were unable to invade or to undergo significant morphological differentiation (data not shown). Together, these results suggested that inactivation of HSL1 might be sufficient to promote pseudohyphal differentiation even in the absence of the filamentous signaling pathways. However, those same mutations in a tec1 flo8 strain are insufficient to promote filamentous differentiation in either haploid or diploid strains (Fig 3A and Fig B). In addition, we found that homozygous diploid hsl1 mutants were unable to undergo pseudohyphal differentiation, even after 20 days of growth if the SLAD medium was supplemented with a high level of nitrogen (Fig 3C). Finally, although hsl1/hsl1 cells are modestly elongated in liquid medium (Table 3), they still exhibit a dramatic morphological response to low nitrogen and solid growth medium (Fig 1A and Fig 2C; Table 3). Thus, although inactivation of HSL1 is able to fully suppress the defect in the invasiveness of diploid cells caused by inactivation of either TEC1 or FLO8, hsl1 cells still depend upon the activity of at least one of those pathways (MAPK or cAMP/PKA) as well as on the normal stimuli for filamentous differentiation (contact with solid substrate and nitrogen starvation).

SWE1 is required for invasiveness of haploid and diploid cells in the absence of TEC1:
The findings described above suggested that activation of SWE1 might also be important for the residual invasive behavior exhibited by tec1 mutants. To determine whether that is the case, we analyzed invasive behavior of both haploid and diploid tec1 mutants having or lacking functional SWE1. In both cases, the modest invasiveness of tec1 mutants was found to depend upon SWE1 (Fig 4). This was most apparent when comparing the frequency of invasive tec1/tec1 cells to that of tec1/tec1 swe1/swe1 cells on washed SLAD plates but could also be observed by comparing washed plates of tec1 and tec1 swe1 haploid strains. Furthermore, the extent of morphological differentiation appeared to be more compromised in tec1 swe1 mutants than in cells having a tec1 mutation alone. Thus, despite our failure to detect a contribution of SWE1 to the invasive behavior of otherwise wild-type strains, we have observed a substantial effect of the swe1 mutation upon the invasiveness of cells that are partially compromised in that capacity.

View larger version (81K): In this window In a new window Download PPT slide   Figure 4. Both SWE1 and TEC1 contribute to the same aspects of filamentous development. Top section: SWE1 contributes to haploid invasive growth in the absence of TEC1. Invasive growth assay of wild-type (WT, L5528), tec1 (RLVsc69), and swe1 tec1 (RLVsc176) strains. Bottom section: Swe1 inactivation leads to a decrease in the residual invasive capacity of tec1/tec1 strain under pseudohyphal growth conditions. Pseudohyphal growth of wild-type (WT, RLVsc291), tec1/tec1 (RLVsc294), and swe1/swe1 tec1/tec1 (RLVsc323) strains.

Swe1 is required for filamentous growth under suboptimal induction conditions:
The finding that SWE1 contributes to both haploid invasive and pseudohyphal growth in cells lacking TEC1 suggested that the contribution of SWE1 might be most significant when the signals for filamentous differentiation are compromised. Although pseudohyphal growth is most often assayed on minimal medium containing 50 µM ammonium sulfate (SLAD) at 30°, we also detect filamentous differentiation and invasiveness in the presence of 10-fold higher ammonium sulfate as well as on SLAD plates at lower or higher temperatures. To evaluate whether SWE1 is more important for filamentous growth under these "suboptimal" conditions, the effect of inactivation of SWE1 on the invasive behavior of cells was analyzed.

Inactivation of SWE1 has little if any effect upon the extent of pseudohyphal differentiation in otherwise wild-type cells under standard conditions. In contrast to its dispensability in cells growing on SLAD medium at 30° (Fig 2B), inactivation of SWE1 compromised pseudohyphal growth on the same medium at either 20° (data not shown) or 35° (Fig 5A). This is most obvious when the extent of morphological differentiation of the cells is examined. The swe1 mutants exhibit no elongation or filament formation after 1 day of growth. There is also a noticeable effect on the invasiveness and morphological differentiation of the cells observed after 5 days of growth as evidenced by the frequency of cells retained on the plate after washing and microscopic examination. Similarly, pseudohyphal growth is compromised in swe1 mutants growing at 30° on medium containing 500 µM (rather than 50 µM) ammonium sulfate (Fig 5A). However, the morphogenesis defect of swe1 mutants grown at higher nitrogen levels is less apparent. Furthermore, we have examined the effect of alternative carbon sources (raffinose/galactose) and nitrogen sources (leucine and proline) on pseudohyphal growth of wild-type and swe1 mutant cells. In all cases, inactivation of SWE1 had a readily detectable effect on the ability of cells to undergo pseudohyphal differentiation (data not shown). Finally, under all conditions tested inactivation of hsl1 led to the enhancement of pseudohyphal differentiation (Fig 5A and data not shown).

View larger version (43K): In this window In a new window Download PPT slide   Figure 5. SWE1 is essential for filamentous differentiation in suboptimal conditions. (A) SWE1 inactivation leads to inhibition of pseudohyphal development at elevated temperatures or at moderate nitrogen levels. Pseudohyphal growth of wild-type (WT, RLVsc291), hsl1/hsl1 (RLVsc63), and swe1/swe1 (RLVsc64) strains on SLAD plates at 35° and on SLAD plates supplemented with a moderate level of ammonium sulfate (0.5 mM). (B) Inactivation of Swe1 abolishes haploid invasive growth on SC plates. Invasive growth of wild-type (WT, L5487), hsl1 (RLVsc46), and swe1 (RLVsc47) strains on the indicated solid growth medium.

We obtained similar results with haploid mutant strains growing on solid rich medium containing either glucose (YEPD) or galactose (YEPGal, data not shown) as a carbon source as well as on synthetic complete medium with glucose as carbon source (SC; Fig 5B). Whereas on YEPD medium swe1 mutants were only slightly compromised in their capacity to invade, no invasive growth was observed under the nonstandard conditions. In all cases, hyperactivation of SWE1 caused a modest but noticeable increase in invasiveness. We conclude that Swe1 is essential for filamentous differentiation under a variety of nonstandard induction conditions.

Swe1 is essential for induction of filamentous growth by butanol:
It has been recently shown that fusel alcohols, notably butanol, induce filamentous growth in haploid cells on both solid and liquid media (DICKINSON 1996 ; LORENZ et al. 2000 ). This phenomenon is dependent upon the MAPK/Tec1 pathway, but not on Flo8, and results in alterations in cell morphology similar to pseudohyphal differentiation. We, therefore, addressed whether SWE1 is required for alcohol-induced filamentous differentiation. As reported previously (LORENZ et al. 2000 ), wild-type haploid cells grown in either rich liquid medium or solid SLAD medium containing 1% butanol exhibited elongated cell morphology and unipolar budding. In liquid this leads to the formation of chains of elongated cells clustered in florets (LORENZ et al. 2000 and Fig 6A). These features were even more pronounced in the hsl1 strain (data not shown). Interestingly, the swe1 strain was substantially compromised in its capacity to undergo butanol-induced differentiation. This was manifested by an inability to differentiate into elongated cells despite acquisition of the unipolar budding pattern in at least some of the cells (Fig 6A). Although haploid cells do not normally undergo invasive differentiation on SLAD plates, they do so in the presence of 1% butanol (LORENZ et al. 2000 and data not shown). Inactivation of Swe1 suppresses most of the phenotypic response observed in those cells (data not shown).

View larger version (52K): In this window In a new window Download PPT slide   Figure 6. SWE1 is essential for the butanol-mediated induction of filamentous growth. (A) Inactivation of Swe1 leads to a decrease in the filamentous response of haploid cells growing in liquid media supplemented with butanol. Wild-type (WT, L5487) and swe1 (RLVsc47) strains are shown during growth in YEPD with or without 1% butanol. (B) Inactivation of Swe1 diminishes the invasiveness and cell elongation of diploid cells induced by growth on butanol-containing SLAD plates. Pseudohyphal development of wild-type (WT, RLVsc291) and swe1/swe1 (RLVsc64) strains growing on SLAD plate supplemented with 1% butanol.

In contrast to its effect on haploid cells, diploid cells growing on SLAD plates containing 1% butanol exhibited a partial inhibition of pseudohyphal development (Fig 6B). As in other nonstandard conditions for pseudohyphal differentiation that we have examined, inactivation of SWE1 resulted in a strong reduction in invasive growth in the presence of butanol (Fig 1 and Fig 6B). That reduction, as in butanol-induced filamentation in haploid cells, was accompanied by failure to develop elongated cell morphology and the retention of unipolar budding. Although ethanol was also shown to stimulate pseudohyphal development of diploid strains (LORENZ et al. 2000 ), we found no reduction in the effectiveness of ethanol-induced pseudohyphal differentiation caused by inactivation of SWE1 (data not shown). Thus, butanol induces filamentation under a wide range of conditions in both haploid and diploid cells and most of those responses are suppressed by inactivation of Swe1.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have performed an in-depth investigation of the role of the Swe1 protein kinase, a negative regulator of Clb-associated CDK activity (BOOHER et al. 1993 ; LEW and REED 1995 ; SIA et al. 1996 ) in filamentous differentiation in the budding yeast. Two approaches were taken to the manipulation of Swe1 activity. Swe1 was either activated by elimination of its negative regulators Hsl1 and Hsl7 (MA et al. 1996 ; BARRAL et al. 1999 ; EDGINGTON et al. 1999 ; MCMILLAN et al. 1999A ; SHULEWITZ et al. 1999 ) or inactivated directly by disruption of the SWE1 gene (BOOHER et al. 1993 ). Inactivation of either Hsl1 or Hsl7 has been shown to promote polarized growth of yeast cells (MA et al. 1996 ; EDGINGTON et al. 1999 ; MCMILLAN et al. 1999A ; SHULEWITZ et al. 1999 ) and to lead to a unipolar budding pattern and invasiveness (EDGINGTON et al. 1999 ). In the absence of those proteins Swe1 fails to localize to the bud neck and is no longer properly regulated with regard to bud emergence (MCMILLAN et al. 1999A ; SHULEWITZ et al. 1999 ). As a consequence, Clb-associated CDK activity fails to be activated in a timely manner, leading to polarized growth promoted by persistence of G1 cyclin-associated CDK activity (MCMILLAN et al. 1999A ; SHULEWITZ et al. 1999 ). Our study extends previous observations (KRON et al. 1994 ; AHN et al. 1999 ; EDGINGTON et al. 1999 ; LOEB et al. 1999A ) by establishing that inactivation of either Hsl1 or Hsl7 enhances all aspects of filamentous growth in both haploid and homozygous diploid strains. When compared to wild-type cells, the mutant cells are more invasive, more elongated, and more filamentous. In that context, it is noteworthy that two of the mutations identified by Myers and colleagues (BLACKETER et al. 1995 ) that promote both haploid invasive and pseudohyphal growth of the D273 background are allelic with HSL1 and HSL7 (elm2 and elm5, respectively; A. MYERS, personal communication). However, we show that, despite their enhanced filamentous behavior, the hsl1 and hsl7 mutant strains still depend upon the normal filamentous growth signals (contact with a solid substrate, nitrogen starvation, and a functional MAPK or cAMP/PKA pathway). Moreover, we show that the enhancement of both haploid invasive growth and pseudohyphal growth that occurs in response to inactivation of either Hsl1 or Hsl7 is dependent upon a functional Swe1. Thus, activation of Swe1 is sufficient to explain the enhancement of invasive growth resulting from Hsl1 and Hsl7 inactivation.

Swe1 is likely to exert its effect on filamentous differentiation via its capacity to phosphorylate and inactivate the Cdc28 CDK (SIA et al. 1996 ), its only known target. However, as others have observed using the same mutation (SIA et al. 1996 ; EDGINGTON et al. 1999 ), we have found that eliminating the Swe1 phosphorylation site on Cdc28 has a less dramatic effect on the capacity of cells to differentiate in butanol-containing medium than does inactivation of Swe1 (data not shown). Although this appears to support a role for Swe1 independent of CDK phosphorylation, it has been reported that Swe1 can inhibit the activity of their target CDKs without phosphorylation (MCMILLAN et al. 1999B ). Consequently, this issue remains to be resolved.

It was recently reported that inactivation of Hsl7 enhances pseudohyphal growth whereas overexpression leads to its repression (FUJITA et al. 1999 ). That study found that the effect of Hsl7 overexpression was suppressed by overexpression of Ste20, the Pak family kinase of the MAPK signaling pathway that targets Tec1 (FUJITA et al. 1999 ). Furthermore, the enhancement of polarized growth resulting from inactivation of Hsl7 was suppressed by inactivation of Ste20 in both haploid and diploid cells. Although our data does not specifically address the effect of Hsl7 overexpression, we demonstrate that the effect of Hsl7 inactivation can be explained entirely by the activation of Swe1. Since inactivation of Hsl7 in cells lacking Swe1 results in no enhancement of invasive growth, we conclude that the inhibitory effect of Hsl7 on Ste20 does not contribute significantly to invasive growth under standard assay conditions. In fact, there was no condition in which hsl7 swe1 mutants were more invasive than wild-type cells. Nevertheless, the two findings might not be contradictory. It is known that Swe1 activation, via inhibition of mitotic forms of the CDK, prolongs the period during which G1 cyclin-associated CDK remains active and, thereby, the extent of polarized growth (LEW and REED 1993 ). It is possible that the contribution of hyperactive Swe1 to morphogenesis is a consequence of the capacity of G1 cyclin-associated CDK to modulate Ste20 activity (OEHLEN and CROSS 1998 ).

We demonstrate that deregulation of Swe1 strongly enhances filamentous growth. Although cells with activated Swe1 still depend upon the normal signals for filamentous growth, Swe1 activation is sufficient to promote invasive behavior even in the absence of either Tec1 or Flo8, downstream targets of the MAPK and cAMP/PKA pathways, respectively. Together with previous studies, these findings support a model in which modulation of CDK activity plays a central role in filamentation (KRON et al. 1994 ; AHN et al. 1999 ; EDGINGTON et al. 1999 ; LOEB et al. 1999A ; MADHANI et al. 1999 ). Indeed, Swe1, most likely acting through its capacity to inhibit Clb/CDK activity, can override a defect in either the MAPK or cAMP/PKA pathways, suggesting that either or both of these pathways may act, at least in part, by inhibiting Clb/CDK activity. Although it is possible that in cells lacking Tec1 or Flo8 the signal for filamentous growth is transduced via Swe1, we consider it more likely that Swe1 activation suppresses the deficiency in those pathways by providing an alternative or supplementary mechanism for Clb/CDK inhibition. Indeed, Swe1 activation is unable to promote filamentous differentiation when both the Tec1 and Flo8 transcription factors have been inactivated, suggesting that the morphogenesis checkpoint pathway provides a modulatory function as opposed to being a primary pathway for transduction of the signals for filamentous differentiation.

Despite the capacity of hyperactivated Swe1 to effectively suppress the defect in invasiveness caused by inactivation of either Flo8 or Tec1, there are striking differences between the effect of flo8 and tec1 mutants on filamentous differentiation. First, in diploid cells, TEC1, but not FLO8, is required for cell elongation (MOSCH and FINK 1997 ; PAN and HEITMAN 1999 ). Conversely, cell-cell adhesion is defective in the noninvasive flo8 mutant population (PAN and HEITMAN 1999 ) but appears to be unaffected in tec1 mutants (MOSCH and FINK 1997 ; PAN and HEITMAN 1999 ). We find that although activation of Swe1 suppresses the defect in cell elongation in tec1 mutants, that suppression is inefficient. This suggests that Swe1 cannot activate but, instead, modulates the Tec1-dependent filamentous signal that promotes cell elongation. Furthermore, Swe1 activation induces cell-cell adhesion in an otherwise defective flo8 strain, perhaps indirectly through modulation of a target of Tec1 such as Flo11 (GAVRIAS et al. 1996 ; MOSCH and FINK 1997 ; LORENZ and HEITMAN 1998 ; RUPP et al. 1999 ). Thus, cell-cell adhesion appears to be a consequence of activation of a downstream element that can act as a target for the MAPK, cAMP/PKA, and the morphogenesis checkpoint pathway.

Our results demonstrate that Swe1 contributes to pseudohyphal and haploid invasive growth under standard assay conditions. Although inactivation of Swe1 has a noticeable effect on the efficiency of invasive growth of haploid cells and on morphological differentiation of diploid cells, the invasiveness of diploid swe1 mutants is largely indistinguishable from that of wild-type cells (AHN et al. 1999 ). We have furthered prior observations by demonstrating a requirement for either Swe1 or Tec1 in the standard assay. Indeed, inactivation of Swe1 eliminates the residual invasive growth observed under optimal conditions in both haploid and homozygous diploid tec1 mutants. This suggests that these two proteins act coordinately to induce filamentous differentiation. Moreover, consistent with previous observations (MADHANI et al. 1999 ), we find that the well-established invasive growth assays may not be sufficient to measure all of the properties important for filamentous differentiation. As in that earlier analysis, we have shown that alteration of the standard assay conditions reveals important factors for differentiation that were not revealed under the standard conditions. We have established a requirement for SWE1 for both forms of filamentous growth under nonstandard conditions (suboptimal temperature and quality or quantity of nutrients). Furthermore, we found that Swe1 is essential for butanol-mediated induction of filamentous growth, a related phenomenon (DICKINSON 1996 ; LORENZ et al. 2000 ). We have confirmed that the Flo8 transcription factor is largely dispensable for butanol-induced filamentation in both haploid and diploid cells. This suggests that, even under those conditions, FLO8 is not required for development of polarization or a unipolar budding pattern (LORENZ et al. 2000 and data not shown). In contrast, haploid and homozygous diploid cells carrying swe1 (Fig 6) or tec1 (LORENZ et al. 2000 and data not shown) mutations exhibit unipolar budding and cell-cell adhesion but are not elongated when growing on either liquid and solid butanol media. Our findings lead to the conclusion that inactivation of Swe1 has a dramatic effect in terms of both kinetics and extent of filamentous differentiation of haploid and diploid cells under many conditions.

It has been suggested that CDKs are the targets of pseudohyphal growth signals (AHN et al. 1999 ). The modulatory function of Swe1 on filamentous growth may not be evident under optimal conditions because its contribution is masked by activation of another pathway, such as that involving TEC1. However, the Swe1 pathway may become more important when repression of Clb-associated CDK activity via those other pathways becomes limiting. Our experiments support a model in which at least three distinct pathways (MAPK, cAMP/PKA, and the morphogenesis checkpoint pathway) play a role in filamentous differentiation of both haploid and diploid strains with each pathway influencing the extent of the phenotypic responses. Moreover, the extent of signaling via each pathway may be strongly influenced by differences in physiological conditions. Indeed, under a variety of conditions, inactivating each of these pathways has a striking consequence in terms of specific phenotypic responses. For example, inactivation of Swe1 results in defects in cell elongation and budding pattern under optimal conditions and affects a broad range of responses under the nonstandard induction conditions that were analyzed. Thus, the effect of Swe1 inactivation on haploid invasive growth may be more readily apparent simply because the strength of one or both of the other signals is lower under haploid invasive growth conditions than it is in diploid cells undergoing pseudohyphal differentiation. Indeed, we, as well as others (EDGINGTON et al. 1999 and data not shown), have noticed that inactivation of Swe1 has a readily detectable effect on the weak haploid invasive behavior of noninvasive laboratory strains. On the basis of our results, we suggest that when filamentous signaling via the MAPK/Tec1 pathway or the cAMP/Flo8 pathway is diminished, Swe1-dependent inhibition of Clb/CDK function becomes essential for filamentous differentiation.

The model proposed above raises important questions. Is the Swe1 protein kinase regulated in response to signals that promote invasive growth and, if so, via what mechanism? This question has yet to be adequately addressed. Both direct and indirect mechanisms by which invasive growth signals might enhance Swe1 activity can be envisioned. Since SWE1 is a target of the G1-specific transcription system (LEW and REED 1995 ), alteration of cell cycle kinetics can substantially alter the accumulation of the SWE1 transcript. Specifically, mechanisms that delay the accumulation of B-type cyclin-associated CDK activity (including Swe1, itself) can enhance the accumulation of G1-specific transcripts. Swe1 is also regulated at the level of turnover via the Cdc34-SCF^Met30 ubiquitin ligase (KAISER et al. 1999 ). It is not clear whether its capacity to act as a substrate is controlled via regulation of the Hsl1 protein kinase or via other mechanisms. Alternatively, the activity of the Swe1 kinase may be regulated via nutritional mechanisms.

The ability to switch from a yeast form to a filamentous form is also typical of many fungal pathogens (reviewed in MADHANI and FINK 1998 and BORGES-WALMSLEY and WALMSLEY 2000 ). The multiple signaling pathways that control dimorphism are conserved among evolutionarily distant fungi and are required for virulence and infection of fungal pathogens. In Ustilago maydis, a smut fungus that infects and induces tumors in corn, both the MAPK (BANUETT and HERSKOWITZ 1994 ) and the cAMP/PKA (GOLD et al. 1994 , GOLD et al. 1997 ) pathways have been shown to be essential for filamentous development and pathogenesis. In the same way, the two pathways are essential for the appressorium formation, a specialized infectious structure, and subsequent infection of rice by the fungus Magnaportha grisea (MITCHELL and DEAN 1995 ; XU and HAMER 1996 ). Finally, the cAMP and MAPK pathways have been shown to be essential for dimorphism and pathogenicity of Criptococcus neoformans (ALSPAUGH et al. 1997 ) and Candida albicans (LIU et al. 1994 ; SINGH et al. 1994 , SINGH et al. 1997 ; KOHLER and FINK 1996 ; LEBERER et al. 1996 , LEBERER et al. 1997 ) respectively, two human opportunistic fungal pathogens that are seen with greatly increased prevalence in immunocompromised hosts and frequently lead to elevated morbidity and mortality (IMAM et al. 1990 ; CASSONE 1996 ). Here we propose that SWE1 kinase plays an important modulatory role in filamentous differentiation of S. cerevisiae. Precedence for a role of a Swe1-like kinase in cell differentiation has recently been reported in the fungus Aspergillus nidulans where the unphosphorylatable CDK NIMX^cdc2AF has been shown to block conidia formation (YE et al. 1999 ). Consistent with the high conservation of the other signaling pathways involved in dimorphism we propose also that SWE1 could play a relevant role in that process and, thereby, in the virulence of fungal pathogens. Indeed, cyclin-dependent protein kinases have been demonstrated to play an important modulatory role in the hyphal development and the virulence of C. albicans (LOEB et al. 1999B ).

	FOOTNOTES

¹ Permanent address: Department of Bacteriology and Medical Mycology, Istituto Superiore di Sanità, 00161 Rome, Italy.

	ACKNOWLEDGMENTS

We thank Anne Dranginis, Tracie Bierwagen, Alan Myers, Danny Lew, Monique Smeets, Steven Reed, Jonathan Loeb, and Haoping Liu for strains and plasmids. We also thank Alan Myers and members of the TSRI Cell Cycle Group (the S. Reed, P. Russell, C. McGowan, and C. Wittenberg laboratories) for their comments and support during the course of this research. R.L.V. is supported by an AIDS fellowship from the National AIDS Program, Istituto Superiore di Sanità, Italy. Support for this project was from US Public Health Service grant GM-43487 to C.W.

Manuscript received November 29, 2000; Accepted for publication February 21, 2001.

	LITERATURE CITED