Relative Dependence of Different Outputs of the Saccharomyces cerevisiae Pheromone Response Pathway on the MAP Kinase Fus3p

Corresponding author: Elaine A. Elion, Harvard Medical School, Department of Biological Chemistry and Molecular Pharmacology, 240 Longwood Ave., Boston, MA 02115., elion{at}bcmp.med.harvard.edu (E-mail)

Communicating editor: E. W. JONES

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fus3p and Kss1p act at the end of a conserved signaling cascade that mediates numerous cellular responses for mating. To determine the role of Fus3p in different outputs, we isolated and characterized a series of partial-function fus3 point mutants for their ability to phosphorylate a substrate (Ste7p), activate Ste12p, undergo G1 arrest, form shmoos, select partners, mate, and recover. All the mutations lie in residues that are conserved among MAP kinases and are predicted to affect either enzyme activity or binding to Ste7p or substrates. The data argue that Fus3p regulates the various outputs assayed through the phosphorylation of multiple substrates. Different levels of Fus3p function are required for individual outputs, with the most function required for shmoo formation, the terminal output. The ability of Fus3p to promote shmoo formation strongly correlates with its ability to promote G1 arrest, suggesting that the two events are coupled. Fus3p promotes recovery through a mechanism that is distinct from its ability to promote G1 arrest and may involve a mechanism that does not require kinase activity. Moreover, catalytically inactive Fus3p inhibits the ability of active Fus3p to activate Ste12p and hastens recovery without blocking G1 arrest or shmoo formation. These results raise the possibility that in the absence of sustained activation of Fus3p, catalytically inactive Fus3p blocks further differentiation by restoring mitotic growth. Finally, suppression analysis argues that Kss1p contributes to the overall pheromone response in a wild-type strain, but that Fus3p is the critical kinase for all of the outputs tested.

HAPLOID Saccharomyces cerevisiae cells (a and ) mate and form diploids through the action of two mitogen-activated protein (MAP) kinases (MAPKs), Fus3p and Kss1p (ELION et al. 1990 , ELION et al. 1991A , ELION et al. 1991B ; GARTNER et al. 1992 ; MA et al. 1995 ). The MAPKs act at the end of a conserved receptor/G protein-coupled protein kinase cascade that is activated by mating pheromone (SPRAGUE and THORNER 1992 ; HERSKOWITZ 1995 ). The ß subunit of the G protein (Ste4p) binds Ste20p, a kinase related to p21-activated kinases (SIMON et al. 1995 ; LEEUW et al. 1998). Ste20p is then thought to activate a MAPKKK Ste11p (LEBERER et al. 1992 ; RAMER and DAVIS 1993 ). Ste11p activates a MAPKK Ste7p (NEIMAN and HERSKOWITZ 1994 ) that activates two MAP kinases, Fus3p and Kss1p (ERREDE et al. 1993 ). Signal transmission requires Ste5p, a scaffolding protein that tethers the MAP kinase cascade enzymes Ste11p, Ste7p, and Fus3p into a high-molecular-weight complex of high specific activity (CHOI et al. 1994 ; KRANZ et al. 1994 ; CHOI et al. 1999 ). In the presence of mating pheromone, Ste5p binds to Ste4p and facilitates activation of Ste11p by Ste20p (LEEUW et al. 1995 ; FENG et al. 1998 ). The ultimate activation of the two MAP kinases is coupled to many responses, including activation of the Ste12p transcription factor, G1 arrest, shmoo formation, cell fusion, and signal attenuation leading to reentry into the mitotic cycle in the absence of mating (ELION et al. 1990 , ELION et al. 1991B ; GARTNER et al. 1992 ; MA et al. 1995 ).

The MAP kinases have overlapping and unique functions for mating, although Fus3p plays a more critical role. While a kss1 null does not have obvious mating defects, a fus3 null is partially defective in mating, G1 arrest, and transcription of FUS1 and FUS2 (ELION et al. 1990 , ELION et al. 1991A , ELION et al. 1991B , ELION et al. 1995 ), both Ste12p-dependent genes (MCCAFFREY et al. 1987 ; ELION et al. 1995 ). Kss1p can perform mating functions because fus3 KSS1 strains mate and fus3 kss1 double-null mutants are completely blocked for activation of Ste12p and sterile (ELION et al. 1991B ; STEVENSON et al. 1992 ), even when the pathway is hyperactivated by overexpression of Gß (ELION et al. 1991B ) or gain-of-function mutations in STE11 (STEVENSON et al. 1992 ). The relative contributions of Fus3p and Kss1p to the different pheromone responses in a wild-type cell is unknown. However, the fact that catalytically inactive Fus3p nearly completely blocks the ability of Kss1p to activate the STE2 gene and to promote mating in the presence of -factor has led to the proposal that Kss1p has no role in the mating pathway in the presence of Fus3p (MADHANI et al. 1997 ).

Analysis of the roles of Fus3p in the pheromone response pathway is complicated by the fact that Fus3p activates Ste12p (ELION et al. 1991B , ELION et al. 1993 ; SONG et al. 1991 ; TEDFORD et al. 1997 ), which activates genes encoding components of the signal transduction cascade as well as effectors of the different pheromone responses (reviewed in SPRAGUE and THORNER 1992 ). Thus, defects associated with fus3 mutations could simply be caused by reductions in the transcription of Ste12p-dependent genes rather than by specific blocks in particular responses. Fus3p associates with and phosphorylates a number of proteins in vitro (ELION et al. 1993 ; TEDFORD et al. 1997 ), suggesting that it mediates the many responses to mating pheromone through direct phosphorylation of effectors. This possibility is consistent with the finding that many of the in vitro substrates that associate with Fus3p are known regulators of the pheromone response pathway [e.g., Ste12p (ELION et al. 1993 ), Far1p (ELION et al. 1993 ; PETER et al. 1993 ; TYERS and FUTCHER 1993 ), Ste5p (KRANZ et al. 1994 ), Ste7p (BARDWELL et al. 1996 ), Ste11p (CHOI et al. 1994 ), and Dig1p/Rst1p and Dig2p/Rst2p (TEDFORD et al. 1997 )].

To analyze the role of Fus3p in the various pheromone responses, we isolated and characterized a series of partial-function fus3 point mutants in a strain lacking KSS1. Our results suggest that Fus3p regulates many outputs through the phosphorylation of multiple substrates, and that Fus3p dosage plays an important role in terminal differentiation. A strong correlation was found between the ability of Fus3p to promote G1 arrest and shmoo formation, suggesting that the two processes may be linked. Suppression analysis argues that Kss1p competes with Fus3p for a subset of substrates in a wild-type strain, but that Fus3p provides essential functions for numerous outputs that cannot be performed by Kss1p. Finally, we show that in a cell containing both Fus3p and Kss1p, catalytically inactive Fus3p preferentially inhibits the ability of active Fus3p to activate Ste12p and promotes recovery from G1 arrest as well as reentry into the mating pathway.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Media and microbiological techniques:
Media preparation (SHERMAN et al. 1986 ), yeast transformations (ITO et al. 1983 ), isolation of yeast genomic DNA (SHERMAN et al. 1986 ), and recovery of plasmid DNA from yeast (HOFFMAN and WINSTON 1987 ) were done as described. Standard bacterial methods were used with Escherichia coli strains HB101, DH5 or KC8 (SAMBROOK et al. 1989 ).

Yeast strain construction:
Yeast strains are listed in Table 1. fus3 mutations (fus3-1, fus3-2, and fus3-201–fus3-213) were put into the FUS3 locus by two-step gene replacement using fus3 derivatives of YIp5 (pYBS16, pYBS19, and pYBS81–pYBS93). Plasmids were linearized at a PmlI site 3' of FUS3 and then integrated next to fus3::LEU2 in EY966 and EY941. ura3 fus3 recombinants were counterselected with 5-fluoroorotic acid (BOEKE et al. 1984 ). fus3 point mutants were identified by loss of LEU2. Southern analysis was done to confirm gene replacement.

Plasmid construction:
fus3-8::ADE2 has a 3-kb BglII ADE2 fragment in the BglII site of a fus3 deletion mutant (pYEE97), creating pYEE1122. The fus3-2 URA3 CEN4 plasmid pYBS40 has the 4-kb HindIII fragment of pYBS16 in the HindIII site of YCp50. The same fragment is cloned into the HindIII site of YIp5 and YEp13 to generate pYBS36 and pYBS29, respectively. Integrating plasmids for fus3-201–fus3-213 (pYBS81–pYBS93) have the BamHI-EcoRI fragment from originally isolated mutants derived from pYEE81 in YIplac211 (GIETZ and SUGINO 1988 ). 2µ plasmids for fus3-201–fus3-213 (pYBS111–pYBS123) have the BamHI-EcoRI fragment in YEplac195.

Cloning of fus3-1 and fus3-2:
The fus3-1 and fus3-2 alleles were cloned from EY473 and EY465 (ELION et al. 1990 ) by eviction. Integrating plasmid pYEE87, with a BamHI site 5' to FUS3 and URA3 (ELION et al. 1990 ), was linearized at BssHII sites 3' of the FUS3 coding sequence and transformed into strains EY473 and EY465. Genomic DNA was isolated from three transformants, digested with BamHI, ligated, and then transformed into E. coli, selecting for ampicillin resistance. Ampicillin-resistant plasmids were screened for additional 5' genomic sequence by digestion with BamHI and BssHII. pYBS16 and pYBS19 were two such plasmids with ~15 kb of genomic sequences 5' to fus3-2 and fus3-1. Each fus3 gene was sequenced on both strands using 15-mer oligonucleotide primers spaced every 150 nucleotides. The fus3-1 and fus3-2 sequences are identical to the published sequence (ELION et al. 1990 ), except for single nucleotide mutations (fus3-1 G838 A838, fus3-2 C431 T431). Wild-type FUS3 from EY441 was also sequenced and was found to have five nucleotide differences from the previously sequenced FUS3 gene: one change (C1264 T1264) results in an amino acid change (H334 Y), three are silent (G A 1194, T C1212, C T1264), and one is a deletion 3' to the coding sequence (1361–1363).

Hydroxylamine mutagenesis:
Approximately 10 µg of cesium chloride-purified plasmid DNA (pYEE81: FUS3 URA3 CEN4 ARS1) was incubated for 20 hr at 37° in 0.5 ml of neutral hydroxylamine solution (0.44 N NaOH, 5 mM EDTA, 0.07 g/ml hydroxylamine from Sigma, St. Louis). The DNA was precipitated twice using E. coli tRNA as a carrier, washed with 100% ethanol, and then dissolved in 300 µl H₂O. A control DNA sample was treated identically using 10 mM Tris-HCl, pH 8.0, 1 mM EDTA. E. coli strain KC8 [RX1486 Mod⁺(K12) leuB600 trpC9830 lacx74 StrepA LALUK pyrF::Tn5(KanR) his-B463, a gift from K. Struhl] was transformed with aliquots of mutagenized and nonmutagenized DNA, and the frequency of URA3 mutation was determined to be ~1% of the transformants screened. Three pools of ~50,000 transformants each were made.

Isolation of arrest-defective fus3 mutants that mate:
A reconstruction experiment with MATa bar1 fus3 kss1 cells (EY966) containing either wild-type FUS3 or fus3-2 (pYEE81 or pYBS40) was performed to determine the optimal amount of -factor on the plates. Strain EY966 was then transformed with three pools of hydroxylamine-mutagenized pYEE81. The -factor-resistant mutants were selected directly on SC-uracil plates with 5 µg of -factor. A total of 352 -factor-resistant colonies was picked, retested for -factor resistance, and tested for mating by a qualitative patch mating test with MAT lys9 and MAT lys9 fus1 fus2 tester lawns (ELION et al. 1990 ). A total of 239 retested as resistant; of these, 65 could mate. Twenty-six of the retrieved plasmids conferred -factor resistance and mating.

RNA analysis:
Yeast strains carrying fus3 URA3 CEN4 plasmids were grown at 30° in selective media to an A₆₀₀ of 0.5 and then split into three aliquots for 0-, 15-, and 120-min -factor induction time points. Cells were induced in 50 nM -factor at 30°. Ten milliliters of cells were pelleted and resuspended in 0.5 ml of 25 mM Tris-HCl, pH 8.0, 10 mM EDTA, pH 8.0, 1% SDS, and extracted with an equal volume of 60° water-saturated phenol, incubating at 60° for 30 min with vortexing at 5-min intervals. The aqueous phase was reextracted twice and precipitated with 0.1 volume of 3 M sodium acetate, pH 5.2, and 2 volumes of 100% ethanol at -70° overnight. RNA was pelleted, washed twice with 95% ethanol, and resuspended in 30–100 µl of diethyl pyrocarbonate-treated, doubly distilled water. Five micrograms of total RNA was separated in a 1.2% agarose, 20 mM MOPS, pH 7, 5 mM sodium acetate, 10 mM EDTA, 1 M formaldehyde gel, transferred to nitrocellulose, and hybridized as described (ELION et al. 1990 ). DNA probes were generated by random priming with Klenow and [-³²P]dATP. FUS1 and BIK1 were detected with the 5.5-kb HindIII-PvuII fragment of pJEF518 (TRUEHEART et al. 1987 ). FAR1 was detected with HindIII fragments of pYBS104 (nt 284–2775, CHANG and HERSKOWITZ 1990 ).

Fus1-lacZp assays:
Cells were grown overnight at 30° to an A₆₀₀ of ~0.3, pelleted, and resuspended at an A₆₀₀ of 0.3 in two 10-ml aliquots of prewarmed fresh media. -Factor was added to one of each pair of samples. Cells were shaken at 30° for the indicated length of time, pelleted, and washed once with ice-cold H₂O and frozen at -80°. Pellets were partially thawed on ice in 0.25 ml breaking buffer (0.1 M Tris-HCl, pH 8.0, 20% glycerol v/v, 1 mM dithiothreitol), to which 12.5 µl of 40 mM phenylmethylsulfonyl fluoride in 95% ethanol were added. A predetermined amount of acid-washed glass beads (0.45–0.55 mM diameter, Sigma) was added (to the meniscus), and the samples were vortexed vigorously for four 15-sec pulses, chilling samples intermittently on ice. More breaking buffer was added (0.25 ml), and the samples were vortexed for an additional 15 sec. Samples were then centrifuged for 15 min in a microcentrifuge at 4°, and the supernatant was transferred to a new Eppendorf tube and assayed for total protein using Bio-Rad dye (Bio-Rad, Richmond, CA) and for ß-galactosidase activity as described (CRAVEN et al. 1965 ). Typically, 100 µl of extracts were mixed with Z buffer (CRAVEN et al. 1965 ) for a total volume of 1 ml. Samples were prewarmed in a 30° water bath for 15 min, and 200 µl of 4 mg/ml o-nitrophenyl ß-D-galactopyranoside (Sigma) dissolved in Z buffer was then added. Reactions were stopped by the addition of 500 µl of 1 M Na₂CO₃. Units of ß-galactosidase activity were calculated by the following formula: units = .

Mating, pheromone sensitivity, and morphogenesis assays:
Four-hour patch matings and quantitative matings, halo assays (using 5 µl of 0.1 mM -factor dissolved in 90% methanol or dimethylsulfoxide and stored at -20°), and measurement of the percentage of unbudded cells were performed as described (ELION et al. 1990 ).

DNA sequence analysis:
fus3 mutants were sequenced using synthetic oligonucleotides and the Sequenase Kit version 2.0 (United States Biochemical, Cleveland). The entire coding strand of each mutant was sequenced, and the noncoding strand was sequenced across the region containing the mutation. The 26 plasmids consisted of one isolate of fus3-204(G A 801), 207(G A 904), 209(C T 952), 210(G A 974), 211(G A 1213), 213 (G A 352); two isolates of fus3-203(C T 740), 208(C T 905), 212(G A 1216); three isolates of fus3-201(C T 635), 202(C T 656), 205(G A 846); and five isolates of fus3-206(G A 856). fus3-202, 205, 206, and 212 were each isolated from more than one pool of mutagenized DNA.

Immunoblot analysis of Fus3p and Ste7Mp protein:
Extracts were prepared as described (ELION et al. 1993 ). For Fus3p, 50 µg of whole-cell extract was separated by SDS-PAGE and immunoblotted using Fus3p antiserum (BRILL et al. 1994 ). For Ste7Mp, 1 mg of whole-cell extract was subjected to a 40% ammonium sulfate cut (ZHOU et al. 1993 ); the resulting precipitate was resuspended in 1x loading buffer and run on an SDS-PAGE gel. Ste7Mp was detected with monoclonal antibody 9E10. Fus3p samples were normalized by comparing the relative amounts of Tcm1p ribosomal protein using a monoclonal antibody to Tcm1p (ELION et al. 1995 ). Proteins were visualized with the ECL system (Amersham, Little Chalfont, Buckinghamshire, UK).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of -factor-resistant fus3 mutants that mate:
Fus3p regulates G1 arrest, transcriptional activation, and steps leading to the fusion of two cells into a zygote through independent and/or interdependent mechanisms (ELION et al. 1990 ). To analyze the role of Fus3p in these and other responses, we isolated partial-function fus3 mutants that are defective in one response while functional in others. We first selected for mutants that did not arrest on plates spread with -factor and then screened the -factor-resistant mutants for retention of the ability to mate (MATERIALS AND METHODS). Mating was chosen as a criterion for function because it is dependent upon the activation of the Ste12p transcription factor as well as shmoo formation, partner selection, and cell and nuclear fusion (SPRAGUE and THORNER 1992 ). Previous work has shown that fus3 mutants defective in G1 arrest can be mating competent (ELION et al. 1990 , ELION et al. 1991B ).

The fus3 mutants were isolated and analyzed in a strain lacking the KSS1 gene to ensure that all pheromone responses were FUS3 dependent. The BAR1 gene, which encodes an -factor protease (CIEJEK and THORNER 1979 ), was also deleted to reduce the amount of -factor needed for selection. Figure 1 shows patch mating and halo assays of representative fus3 mutants and scoring criteria. The mutants fall roughly into three classes of -factor resistance: as resistant as a fus3 KSS1 strain (e.g., fus3-203, scored as +), slightly more resistant than a fus3 KSS1 strain (e.g., fus3-206, scored as ++), and as resistant as a fus3 kss1 strain (e.g., fus3-205, scored as +++).

View larger version (70K): In this window In a new window Download PPT slide   Figure 1. -Factor resistance and mating of fus3 mutants. (A) Mating of fus3 mutants. Patches of MATa fus3 mutant strains were mated to a lawn of MAT cells for 4 hr at 30°. Diploids were then selected on a minimal medium plate that was photographed after 2 days at 30°. (B) -Factor resistance of fus3 mutant strains. Halo assays were done using 5 µl of 0.1 mM -factor. Plates were incubated overnight at 30° and then photographed. (C) Dominant negative phenotype of overexpressed fus3 alleles. Halos were done as described in (B). URA3 2µ plasmids carrying the fus3 alleles (except for fus3-2, which is on a LEU2 2µ vector) in a FUS3 KSS1 strain (EY957). All + and - symbols refer to the rating system used during characterization of the mutants and correspond to those in Table 2. For A and B, the fus3 alleles were introduced on URA3 CEN4 plasmids in either fus3 KSS1 (EY941) or fus3 kss1 (EY966) strains. All assays were done at least twice on duplicate samples. Note that wild-type Fus3p behaves as a very weak dominant negative in this strain background (W303); this effect is not seen in S288C (ELION et al. 1990 ).

A summary for all of the fus3 mutants and their amino acid substitutions appears in Table 2, which also includes the original fus3-2 and fus3-1 alleles (ELION et al. 1990 ) introduced into the same strain background (MATERIALS AND METHODS), as well as a previously engineered, catalytically inactive mutant fus3R42R43. None of the hydroxylamine-induced mutations (fus3-201-fus3-213) overlap with either fus3-1 or fus3-2, which were introduced by ethylmethane sulfonate mutagenesis (ELION et al. 1990 ). Three fus3 alleles exhibit temperature-sensitive mating in the patch mating assay. All the fus3 alleles (except for fus3R42R43) are largely recessive on a centromeric plasmid in a FUS3 KSS1 strain, suggesting that they are loss-of-function alleles. Six fus3 alleles are dominant negatives when overexpressed in a FUS3 KSS1 strain (Table 2; Figure 1C; fus3-2, fus3-203, fus3-204, fus3-205, fus3-206, and fus3R42R43), suggesting that these mutations affect Fus3p interactions with other proteins, such as regulators or substrates. All the mutant proteins are present at levels similar to that of Fus3p (Figure 2A). While the Fus3-212p protein appears less abundant in this particular figure, other immunoblots (such as that in Figure 2C) indicate a level similar to native Fus3p.

View larger version (45K): In this window In a new window Download PPT slide   Figure 2. Immunoblot analysis of Fus3p and Ste7Mp from wild-type and mutant strains. (A) Mutant Fus3p proteins. Whole-cell extracts were prepared from wild-type and fus3 KSS1 strains grown to an A600 of ~0.8. Fifty micrograms of extract was immunoblotted using antisera to full-length Fus3p. The immunoblot was then stripped and reprobed using a monoclonal antibody to ribosomal protein Tcm1p. KSS1 derivatives were used to ensure relatively similar levels of FUS3 transcription, as it is Ste12p dependent. (B) Ste7Mp is hyperphosphorylated in the presence of -factor and cycloheximide. Strain EYL341 harboring CYHpSTE7M CEN TRP1 (pNC318, ZHOU et al. 1993 ) was grown and induced with 2 µM -factor for 15 min, and extracts were prepared as described in MATERIALS AND METHODS. Ste7Mp was detected in the immunoblot with antibody 9E10. The other band is a cross-reacting protein. (C) Ste7Mp in fus3 kss1 strains. fus3 kss1 strains harboring CYHpSTE7M CEN TRP1 (pNC318, ZHOU et al. 1993 ) were grown and induced with 50 nM -factor for 1 hr, and extracts were prepared as described in MATERIALS AND METHODS. Ste7Mp was detected in the immunoblot with antibody 9E10. The band marked by an asterisk is a cross-reacting protein. The right side of this blot was stripped and reprobed with Fus3p antisera.

The fus3 mutations affect conserved residues implicated in interactions with regulators and substrates:
Strikingly, all but one of the 15 fus3 point mutations isolated in the screen affect residues conserved among MAP kinases (Figure 3). When catalytically inactive and active Erk2p crystal structures (ZHANG et al. 1994 ; CANAGARAJAH et al. 1997 ) are used as models, the mutations fall into three classes: those predicted to inhibit enzyme activity, and those predicted to affect interactions with either Ste7p or downstream substrates.

View larger version (97K): In this window In a new window Download PPT slide   Figure 3. fus3 mutations affect amino acid residues conserved among MAP kinases. (A) Alignments of Fus3p (ELION et al. 1990 ), Kss1p (COURCHESNE et al. 1989 ), and Erk2p (BOULTON et al. 1990 ), with fus3 point mutations indicated. (B) Positions of the fus3 mutations on the Erk2p crystal structure (ZHANG et al. 1994 , inactive form).

Seven mutations (fus3-2, fus3-201, fus3-202, fus3-203, fus3-211, fus3-212, and fus3-213) are predicted to affect the enzymatic activity of Fus3p kinase. Four of these mutations [A124V (fus3-201), S131L (fus3-202), D317N (fus3-211), and E318K (fus3-212)] cluster near the hinge between the two domains and may affect either Fus3p catalytic activity or interactions with other proteins. The A124 and S131 mutations are buried and may be more likely to inhibit Fus3p activity, perhaps by affecting domain rotations. In contrast, D317 and E318 are on the surface and may affect interactions with other proteins. The D317 and E318 mutations make Fus3p thermolabile (Table 2). A30T (fus3-213) is at the base of the phosphate anchor (P loop) and may influence ATP binding and overall activity. T56M (fus3-2) is in the C helix, and A159V (fus3-203) is near the Asp-Phe-Gly sequence that interacts with the C helix to maintain the separation between the domains. A159V is adjacent to a key arginine that interacts directly with phosphothreonine 183 in the active form (CANAGARAJAH et al. 1997 ). These mutations are most likely to interfere with overall Fus3p activity, possibly through the ability of Fus3p to interact with Ste7p (BRUNET and POUYSSEGUR 1996 ). Both Fus3-2p and Fus3-203p interfere with G1 arrest when overexpressed in the presence of wild-type Fus3p (Table 2), suggesting that they form nonproductive interactions with regulators or substrates.

Four mutations (fus3-1, fus3-204, fus3-205, and fus3-206) are predicted to affect interactions with Ste7p and activation of Fus3p. Three of these mutations cluster in and near the phosphorylation lip and, therefore, are quite likely to affect interactions with Ste7p and possible dimerization of Fus3p. Mutations E192K (fus3-1), M194I (fus3-205), and A198T (fus3-206) lie near the surface loop of Erk2p (aa 199–205) that interacts with the phosphorylation lip. These mutations could alter the affinity of the lip for Mekp conformational changes associated with Erk2p activation or dimerization (CANAGARAJAH et al. 1997 ). M179I (fus3-204) is located on the phosphorylation lip and, therefore, may also affect interactions with Ste7p. A weakening of the interactions between the lip and the adjacent surface loop may strengthen the interaction between Fus3p and Ste7p (ZHANG et al. 1994 ), and it may provide a possible explanation for the fact that three of these mutations (fus3-204, fus3-205, and fus3-206) generate proteins that are dominant negatives when overexpressed (Table 2). Fus3-205p is more efficiently suppressed by overexpression of the upstream activator Ste5p (KRANZ et al. 1994 ) than by downstream substrates such as Far1p and Ste12p (SATTERBERG 1993 ), which is consistent with a defect in activation.

Four mutations (fus3-207, fus3-208, fus3-209, and fus3-210) are likely to influence the ability of Fus3p to interact with substrates. A214V (fus3-207) and A214T (fus3-208) are in helix F, in the extended substrate binding site identified in cAPK-peptide complexes (ZHANG et al. 1994 ; CANAGARAJAH et al. 1997 ). H230Y (fus3-209) is adjacent to the P +1 specificity pocket, the orientation of which changes dramatically upon phosphorylation of Erk2p to allow substrate interactions (CANAGARAJAH et al. 1997 ). Thus, these mutations may affect substrate interactions rather than catalytic activity. G240D (fus3-210) is located on a region that is an insertion in MAP kinases and may affect interactions with substrates. This substitution causes Fus3p to be thermolabile (Table 2). None of the fus3 mutations predicted to alter the binding of Fus3p to substrates result in dominant negative Fus3p proteins (Table 2), suggesting that these amino acid substitutions are more likely to reduce rather than enhance binding to substrates.

The mutant Fus3p proteins have varying defects in hyperphosphorylation of Ste7p:
As an initial indirect assessment of kinase activity, we monitored the ability of the mutant Fus3p proteins to phosphorylate Ste7p in vivo. Fus3p feedback phosphorylates Ste7p (ERREDE et al. 1993 ), possibly to downregulate the activity of Ste7p (HOU et al. 1993). Because the association between Ste7p and Fus3p is strong (BARDWELL et al. 1996 ), this approach may underestimate a defect in the ability to phosphorylate another substrate, or it may identify a defect that is specific to Ste7p. When activated by -factor, Fus3p multiply phosphorylates the N-terminal regulatory domain of Ste7p, causing it to shift from faster- to slower-migrating species (ERREDE et al. 1993 ; ZHOU et al. 1993 ; Figure 2B). This mobility shift occurs in the presence of cycloheximide, consistent with it being an immediate signaling event (Figure 2B). Either Fus3p or Kss1p is sufficient to hyperphosphorylate Ste7p. A small fraction of Ste7p is still hyperphosphorylated in the fus3 kss1 strain by an undefined mechanism (compare FUS3 KSS1, FUS3 kss1, fus3 KSS1, and fus3 kss1 strains, Figure 2C; ZHOU et al. 1993 ).

Ste7p is hyperphosphorylated to varying degrees in the fus3 kss1 strains, consistent with the isolation of a series of partial-function Fus3p proteins. On the basis of the relative intensities of the hyper- and hypophosphorylated forms of Ste7p, the mutants are classified in Table 2 as wild type (++++), reduced (++), greatly reduced (+), or completely defective (-). Three fus3 kss1 strains (fus3-206, fus3-209, and fus3-213) are wild type for phosphorylation of Ste7p. These mutants may retain wild-type enzymatic activity or have a defect that is not resolvable by using Ste7p as a substrate. Intermediately phosphorylated forms of Ste7p accumulate in most of the mutants, indicating incomplete phosphorylation of Ste7p. Kss1p fully hyperphosphorylates Ste7p in two fus3 mutants tested (fus3-2 KSS1, fus3-205 KSS1; Figure 2C), indicating that the mutant Fus3p proteins present in these strains do not block access of Kss1p to Ste7p.

Dosage response of a FUS3 kss1 strain:
To analyze the mutants phenotypically, we generated an -factor dosage response curve for transcriptional activation, G1 arrest, and morphogenesis in a control FUS3 kss1 strain. Transcription was monitored by both Northern blot analysis of FUS1 expression and a FUS1-lacZ fusion gene (TRUEHEART et al. 1987 ), the expression of which is dependent upon Ste12p and the MAP kinases (MCCAFFREY et al. 1987 ; ELION et al. 1991A ). G1 arrest was monitored by the accumulation of unbudded cells, and morphogenesis was monitored by cell enlargement and shmoo formation.

The FUS3 kss1 strain has a different dosage response curve for each output across the -factor concentration range that brackets the conditions used in our experiments (Figure 4). Northern blot analysis shows that expression of the FUS1 gene is maximal at 5 nM -factor after either 15 or 60 min of induction, while FUS1-lacZ expression is only ~65% maximal by 10 nM and ~90% maximal at 25 nM. This difference may be caused by posttranscriptional events that must occur to generate the Fus1-ß-galactosidase fusion protein. G1 arrest does not occur at all at 10 nM -factor, but is ~85% maximal at 17.5 nM -factor. The pattern of FUS1-lacZ expression across this range of -factor concentrations indicates that a proper dosage response is occurring despite the absence of G1 arrest. Cell morphogenesis and shmoo formation only reach ~80 and ~65% maximal levels, respectively, at 50 nM -factor. Thus, transcription requires the least amount of signaling, G1 arrest requires more, and shmoo formation requires the most signaling. These results are consistent with those of MOORE 1983 , who found that increasing -factor was needed for agglutination, G1 arrest, and shmoo formation in a BAR1 strain that is presumably FUS3 and KSS1.

View larger version (24K): In this window In a new window Download PPT slide   Figure 4. Dosage response of a FUS3 bar1 kss1 strain. Strain EY1119 was grown to an A600 of ~0.3 and then induced with the indicated amount of -factor for 2 hr before fixing with formaldehyde and monitoring for unbudded cells and shmoo formation (MATERIALS AND METHODS; legend to Table 5). Strain EY2524 (isogenic strain with FUS1-lacZ inserted into the LYS2 locus; lys2::FUS1-lacZ) was grown identically and induced for 3 hr with -factor, and then cell extracts were prepared and assayed for ß-galactosidase activity (MATERIALS AND METHODS).

Increasing Fus3p function is required for transcriptional activation, G1 arrest, and shmoo formation:
To determine whether increasing Fus3p function might be a determinant for the strength of the different outputs, we measured the relative abilities of the fus3 kss1 strains to transcribe the FUS1 gene (Figure 5; Table 3), arrest in G1 phase (Table 4), enlarge, and form shmoos (Table 5) in the presence of -factor. A summary of these findings is in Table 2. Transcriptional induction of FUS1, a rapid response, was measured by Northern blot analysis after incubating the cells for 15 and 120 min in a saturating concentration of -factor (50 nM). G1 arrest was monitored after a 3-hr incubation in 25 nM -factor (the minimal concentration that confers maximal arrest and a length of time that enables all the cells within a cell cycle to arrest), and shmoo formation was measured after a 3-hr incubation in 50 nM -factor (the slowest response requiring the most -factor).

View larger version (79K): In this window In a new window Download PPT slide   Figure 5. Northern analysis of FUS1 and FAR1 mRNA in fus3 mutants. All wild-type and mutant FUS3 alleles were maintained on URA3 CEN4 plasmids in a fus3 kss1 double mutant (EY966). Strains were grown and RNA extracted as described in MATERIALS AND METHODS. Five micrograms of total RNA was used for the Northern analysis. The RNA blot was simultaneously probed with radiolabeled fragments of FAR1 and FUS1/BIK1.

Overall, the fus3 kss1 strains have a wide range of output capabilities. They exhibit a ninefold range of FUS1 transcription (11–102% wild type), with three of the mutants (fus3-204, fus3-209, and fus3-213) having essentially wild-type levels of transcription (87, 95, and 102%, respectively). They also exhibit a broad spectrum of G1 arrest defects, ranging from little or no arrest defect (fus3-203, fus3-204, fus3-206, fus3-208, fus3-209, and fus3-213) to approximately as defective as a fus3 kss1 double-null mutant (fus3-210, fus3-211, and fus3R42R43). The fus3 kss1 strains are, on average, more severely defective in shmoo formation than the other outputs. Although most strains are able to undergo some morphogenesis (tallied as cell enlargement without an obvious projection), few are able to form complete projections (tallied as shmoo + multishmoo). As expected from the design of our screen, none of the point mutants are as defective as the catalytically inactive fus3R42R43 mutant, which expresses 0.3% wild-type levels of FUS1-lacZ and does not undergo G1 arrest or shmoo formation. None of the mutants display a complete uncoupling of G1 arrest and transcriptional activation (i.e., no G1 arrest with maximal transcription), although our screen allowed for the isolation of such a class of mutants.

We looked for correlations among the various mutants for relative amounts of FUS1 transcription, G1 arrest, morphogenesis, and shmoo formation (Figure 6, A–C). Simplest best-fitting curves were computer generated using subsets of mutants and the FUS3 kss1 and fus3 kss1 strains (shown as circles in all the graphs). Mutants that fall near the curve but do not define it are indicated as open triangles. Mutants that appear to be outliers from the curves are indicated as solid triangles. In many instances, the level of FUS1 transcription roughly correlates with the level of G1 arrest, morphogenesis, and shmoo formation, which is consistent with the fact that activation of Ste12p is required for the expression of a variety of genes that are required for all the outputs. Mutants with lower levels of transcription are more defective in G1 arrest, morphogenesis, and shmoo formation, while mutants with high levels of transcription are most efficient for all the outputs. These correlations are more clearly revealed by the mutants predicted to have reduced Fus3 kinase activity rather than by the mutants predicted to have defects in substrate interactions. Mutants with a greater capacity to undergo G1 arrest are also better able to enlarge and form shmoos (Figure 6D and Figure E). Collectively, this analysis argues that increasing Fus3 function is required for transcriptional activation, G1 arrest, morphogenesis, and shmoo formation, respectively.

View larger version (44K): In this window In a new window Download PPT slide   Figure 6. Correlations among G1 arrest, FUS1 transcription, and morphogenesis in fus3 kss1 mutants. Values from Table 3 Table 4 Table 5 are renormalized relative to the FUS3 kss1 strain and graphed as indicated. Each curve is optimized to go between points for the fus3 kss1 and FUS3 KSS1 strains. See text for details. FUS1 mRNA levels are taken from the 2-hr -factor induction. Simplest best-fitting curves for the data were derived by comparing linear, quadratic, polynomial (third order), and exponential functions.

The ability to undergo G1 arrest, undergo morphogenesis, and form shmoos does not strictly correlate with the ability to activate Ste12p:
Next, we used the same comparative analysis to ask how likely it is that Fus3p phosphorylates only one substrate (such as Ste12p) for all of the responses. If this were the case, then one might imagine that a strict correlation should be found among the various outputs, and that the relative strengths of these outputs might also align with the dosage response curve shown in Figure 4. By contrast, a lack of correlation might be expected if it is necessary to phosphorylate more than one substrate, and if the mutants have different capacities to phosphorylate these substrates. This would be revealed by the existence of outliers from the curves derived from comparisons of the relative output capabilities of the different mutants (Figure 6), as well as an inability to align the values derived from the dosage response curve for the FUS3 kss1 strain with those of the mutants.

We do not find a strict correlation between the level of Ste12p activation and the other outputs, suggesting that Fus3p phosphorylates multiple effectors in addition to Ste12p to mediate G1 arrest, morphogenesis, and shmoo formation. This is shown in two different ways: by comparing the level of FUS1 expression at which the FUS3 kss1 strain mediates the other outputs to that of the mutants, and by the existence of outliers in the correlative graphs in Figure 6A–C. First, it is not possible to align the values of the dosage response curve with respect to FUS1 expression. Many fus3 point mutants undergo some G1 arrest, morphogenesis, and shmoo formation despite greatly reduced levels of FUS1 expression (Table 2), although the FUS3 kss1 strain only mediates these outputs at -factor levels that induce maximal FUS1 expression (Figure 4). In addition, mutants with the same level of FUS1 expression can have varying capacities to promote G1 arrest (Figure 6A, compare the G1 arrest capacities of mutants with ~20% wild-type levels of FUS1 mRNA). Second, two classes of outliers are found when the mutants are compared among themselves for all three outputs (Figure 6, A–C). Class I outliers constitute mutants with a greater than expected ability to promote either G1 arrest, morphogenesis, or shmoo formation (solid triangles to the right of the curves in Figure 6, A–C), while class II outliers constitute mutants with greater-than-expected defects in either G1 arrest, morphogenesis, or shmoo formation for their relative levels of FUS1 expression (solid triangles above the curves in Figure 6, A–C). The absence of correlation is not explained by aberrant expression of the FUS1 gene because the mutants express essentially wild-type levels of another Ste12p-dependent gene, FAR1 (CHANG and HERSKOWITZ 1990 ; OEHLEN et al. 1996 ), in response to -factor (Figure 5; Table 3). Taken together, these findings argue for a direct role for Fus3p in mediating all three outputs, and they suggest that other aspects of Fus3p function distinct from transcriptional activation of Ste12p also mediate G1 arrest, morphogenesis, and shmoo formation.

Remarkably, the fus3 mutations that constitute each class of outliers are predicted to have similar functional consequences on Fus3p. Class I outliers are all predicted to have defects in enzymatic activity or in their ability to be activated by Ste7p (fus3-203, fus3-204, fus3-206, and fus3-213), while class II outliers are all predicted to be defective in interactions with substrates (fus3-207, fus3-208, fus3-209, and fus3-210). Class I outliers may have alterations in the absolute level of Fus3p kinase activity or the duration of the period of Fus3p activation. The enhanced capability of the fus3-204 mutant to undergo G1 arrest, enlarge, and form projections raises the possibility that this is a gain-of-function mutation. Class II outliers are most likely to have reduced capability to bind to substrates involved in G1 arrest, morphogenesis, and shmoo formation while maintaining greater affinity for binding to substrate(s) involved in transcriptional activation (such as Ste12p). Curiously, an increasing number of class II outliers are found for G1 arrest (fus3-207 and fus3-210), morphogenesis (fus3-207, fus3-208, and fus3-210), and shmoo formation (fus3-207, fus3-208, fus3-209, and fus3-210), respectively, perhaps reflecting increasing constraints for binding to substrates in these outputs.

Optimal shmoo formation strongly correlates with G1 arrest:
The fact that the mutants were originally selected on the basis of -factor resistance and are most defective in morphogenesis and shmoo formation (Table 2) raises the possibility that G1 arrest and morphogenesis responses are connected. This possibility is strongly supported by correlative analysis (Figure 6, D–E). The most striking correlation is found for shmoo formation (Figure 6E). Here, optimal shmoo formation is achieved only in the strains that undergo wild-type levels of G1 arrest. Moreover, a nearly superimposable curve is found using the values from the dose response curve of the FUS3 kss1 strain (squares), substantiating the findings with the fus3 point mutants. These relationships strongly suggest that shmoo formation is linked to prior G1 arrest.

Mating capacity of fus3 mutants does not strictly correlate with ability to activate Ste12, promote G1 arrest, or form shmoos:
Because Ste12p function is critical for mating (SPRAGUE and THORNER 1992 ), it is surprising that many of the mutants originally selected as efficient maters in the qualitative mating assay (Figure 1) induce low levels of FUS1 expression (Table 3). Quantitative mating assays were done on representative fus3 kss1 strains that exhibit varying levels of transcription, G1 arrest, and shmoo formation (Table 2). A critical dosage of Fus3p function that pertains to Ste12p-dependent transcription is essential for mating. The fus3R42R43 kss1 strain with the lowest levels of FUS1 expression is nearly as defective in mating as the fus3 kss1 double-null mutant, while the fus3-205 kss1 strain with somewhat higher levels of expression (18% wild type) is 36-fold defective in mating, and the fus3-209 kss1 strain with high levels of FUS1 expression (95% wild type) mates efficiently.

As with the other outputs, however, transcriptional activation of FUS1 is not always a predictor of mating capacity. For example, the fus3 KSS1 strain has a 250,000-fold defect in mating, even though Kss1p induces FUS1 to 46% wild-type levels. By contrast, fus3-2 kss1, fus3-206 kss1, fus3-211 kss1, and fus3-212 kss1 strains have much more modest defects in mating despite equivalent or lower levels of FUS1 transcription than in a fus3 KSS1 strain. Furthermore, mating ability does not correlate with either G1 arrest or shmoo formation (Table 2). Thus, Fus3p appears to have mating functions that are independent of G1 arrest, activation of Ste12p, and shmoo formation and for which Kss1p cannot be substituted. This view is consistent with a specific requirement for Fus3p in cell fusion (ELION et al. 1990 ).

Fus3p promotes partner selection by a distinct mechanism:
During the course of this analysis, we discovered that bar1 and fus3 null mutations cause synthetic sterility (Table 2 and legend). The synthetic sterility with bar1 suggests that Fus3p is required for chemotropic partner selection, a process that involves orienting the projection toward the pheromone gradient produced by a candidate partner cell (DORER et al. 1995 ; VALTZ et al. 1995 ). We measured the ability of fus3 mutants to choose a partner in a "confusion assay" (VALTZ et al. 1995 ). As expected, the mating of wild-type cells is inhibited 100-fold by exogenously added -factor, while that of the bar1 null is only somewhat inhibited as a result of a partner selection defect (Table 6). The fus3 KSS1 null mutant is nearly insensitive to the added -factor (only 1.9-fold inhibition), indicating that Fus3p is required for partner selection. The fus3 bar1 double mutant is more defective in partner selection than either single mutant, suggesting that Fus3p regulates partner selection by a different mechanism than does Bar1p. [The fus3 bar1 strain mates much better in this liquid mating assay than it does in the solid mating assay (Table 2). Further work is needed to determine why this is so.] Because the fus3 null behaves like a far1 null in the confusion assay (VALTZ et al. 1995 ), Fus3p may regulate partner selection, at least in part, through phosphorylation of Far1p. Our findings are consistent with those of DORER et al. 1997 , who have observed synthetic sterility between fus3 and sst2 mutants.

We measured partner selection in a subset of fus3 kss1 bar1 strains with varying levels of transcription, G1 arrest, and shmoo formation to determine whether their ability to mate better than a fus3 KSS1 bar1 strain correlates with better partner selection (Table 7). Three of these strains (fus3-2, fus3-209, and fus3-213) are as defective as the fus3 KSS1 strain, suggesting that their mating capacity does not result from better partner selection. The fus3-204 kss1 bar1 strain, however, chooses partners more efficiently than the fus3 KSS1 bar1 strain, nearly as well as a FUS3 kss1 bar1 strain. The enhanced partner selection correlates with enhanced shmoo and mating capabilities (Table 2 and Table 6). Thus, Fus3p may regulate partner selection by a mechanism that is distinct from transcriptional activation and G1 arrest but possibly related to shmoo formation.

KSS1 suppresses the fus3 point mutants to varying degrees and contributes to shmoo formation in the presence of Fus3p:
Comparative analysis of fus3 kss1 and fus3 KSS1 strains shows that Kss1p is unable to substitute for Fus3p for essential functions in partner selection and mating (Table 2 and Table 6). To assess the relative contribution of Kss1p in other outputs, we examined whether the KSS1 gene suppresses G1 arrest, and transcriptional and morphological defects of the fus3 point mutants. Kss1p restores more efficient transcription of the FUS1 gene, G1 arrest, and shmoo formation to most of the mutants (Table 3 and Table 7, compare fus3 kss1 and fus3 KSS1 strains). In most instances, FUS1 transcription, G1 arrest, and shmoo formation are not restored to the level of a FUS3 KSS1 strain, although all of the fus3 alleles (except fus3R42R43) have partial function. This result is consistent with the fact that Kss1p does not fully substitute for Fus3p for activation of Ste12p (ELION et al. 1991A ; MADHANI et al. 1997 ).

Two additional observations support the view that Kss1p helps to promote multiple outputs in a wild-type strain. First, Kss1p fully suppresses the defects in G1 arrest and shmoo formation of fus3-211 and fus3-212 mutants while only partially suppressing their transcription defects (Table 7). The more efficient suppression of the outputs requiring more signaling is noteworthy because it suggests that Kss1p provides G1 arrest and shmoo formation functions to mutants that are more selectively defective in activation of Ste12p. Second, assessment of kss1 and KSS1 strains shows that Kss1p is required for efficient shmoo formation (Table 5, compare FUS3 KSS1-73%, FUS3 kss1-34%, average of triplicate experiments). Taken together, these data suggest that, while Kss1p is unable to provide essential functions for partner selection and mating in a wild-type strain, it does partially contribute to the activation of Ste12p, G1 arrest, and shmoo formation.

Many of the fus3 mutants promote recovery:
Surprisingly, seven fus3 kss1 strains undergo efficient G1 arrest (fus3-203, fus3-204, fus3-206, fus3-208, fus3-209, and fus3-213) despite the fact that they are -factor resistant in a halo assay (Table 2). Because the amount of growth in a halo assay is the sum of G1 arrest and recovery processes, the -factor resistance of these strains could result from an enhanced ability to promote recovery. The fus3-2 mutant may also have an increased ability to recover, on the basis of the observation that KSS1 restores better G1 arrest to the fus3-2 strain (87% wild-type arrest capacity, Table 4), but it does not restore -factor sensitivity in a halo assay (data not shown). Interestingly, the arrest-competent, -factor-resistant fus3 alleles fall into two classes with respect to Ste7p phosphorylation (Table 2): those with wild-type activity (fus3-206, fus3-209, and fus3-213) and those with reduced activity (fus3-2, fus3-203, fus3-204, and fus3-208). While the use of Ste7p as a substrate may underestimate a catalytic defect, this raises the possibility that Fus3 promotes recovery through two mechanisms, one that requires catalytic activity and one that does not.

To test whether Fus3p promotes recovery through a mechanism distinct from G1 arrest, we analyzed the kinetics of arrest and recovery for a mutant from each class (fus3-206 and fus3-208) in addition to fus3-2. Logarithmically growing cells were treated with 25 nM -factor, and the percentage of unbudded cells was monitored during a period of 9 hr. As shown in Figure 7, all three fus3 kss1 strains recover from G1 arrest more rapidly than the FUS3 kss1 strain. The FUS3 kss1 strain is maximally arrested after 2 hr in -factor and begins to recover 4–5 hr later, as shown by the drop in the percentage of unbudded cells. In contrast, the fus3 mutants begin to recover within 1–2 hr of their maximal arrest time points, leading to more budded cells at the 9-hr time point. Thus, at least three fus3 alleles encode proteins that promote faster recovery. These data argue that Fus3p promotes recovery by a mechanism that is distinct from its ability to promote G1 arrest.

View larger version (29K): In this window In a new window Download PPT slide   Figure 7. A subset of fus3 point mutants promotes recovery. The percentage of unbudded cells for FUS3 kss1, fus3-2 kss1, fus3-206 kss1, and fus3-208 kss1 strains was monitored over a 9-hr -factor time course. Strains were grown to an A600 of 0.3 and then incubated with 25 nM -factor for the indicated times. Samples were fixed with formaldehyde and counted for the percentage of each cell type.

Catalytically inactive Fus3p inhibits the function of both Fus3p and Kss1p:
It has been argued that catalytically inactive forms of Fus3p interfere solely with the function of Kss1p, with no effect on Fus3p (MADHANI et al. 1997 ). However, interference with the function of Fus3p could explain why Fus3p proteins with reduced abilities to phosphorylate Ste7p are able to promote more rapid recovery (i.e., fus3-2, fus3-203, fus3-204, and fus3-208). Such a dominant negative effect would be consistent with the observation that catalytically inactive Fus3p (Fus3R42p) inhibits tyrosine phosphorylation of Fus3p by Ste7p (GARTNER et al. 1992 ).

We first tested whether catalytically inactive Fus3p inhibits the function of Fus3p by comparing the expression of FUS1-lacZ in FUS3 KSS1, FUS3 kss1, and fus3 KSS1 strains harboring the catalytically inactive fus3R42R43 gene on a CEN plasmid. As shown in Table 8, fus3R42R43 inhibits -factor-dependent activation of FUS1-lacZ in both FUS3 strains, in addition to inhibiting expression in the fus3 KSS1 strain. Furthermore, fus3R42R43 causes nearly the same level of inhibition in both FUS3 KSS1 and FUS3 kss1 strains, indicating that the major target of inhibition in a wild-type strain is FUS3 rather than KSS1. fus3Y182F and fus3T180A also inhibit FUS1 expression in FUS3 KSS1 and FUS3 kss1 strains (monitored by a FUS1-HIS3 reporter gene, data not shown), indicating that the unactivated form of Fus3p is also inhibitory. Thus, catalytically impaired Fus3p preferentially inhibits the function of Fus3p compared to Kss1p, providing a possible explanation for the enhanced recovery of fus3 mutants with reduced catalytic activity.

Catalytically inactive Fus3p hastens recovery but does not block G1 arrest or shmoo formation:
We next tested whether catalytically inactive Fus3p promotes recovery in the presence of wild-type Fus3p and Kss1p by determining whether the fus3R42R43 centromere plasmid causes a wild-type strain to become -factor resistant in a halo assay and recover faster in an -factor time course experiment. FUS3 KSS1 and FUS3 kss1 strains harboring the fus3R42R43 gene are more resistant to -factor, as shown by turbid halos (Figure 8A). Slightly greater resistance is found for the FUS3 kss1 strain, which is consistent with the greater inhibition of FUS1 expression (Table 8). Figure 8B shows that, while both FUS3 fus3R42R43 and FUS3 strains arrest with identical kinetics and efficiency in the presence of -factor, the FUS3 fus3R42R43 strain initiates recovery approximately two cell cycles (3 hr) before the FUS3 strain. Furthermore, fus3R42R43 does not inhibit shmoo formation (Figure 8C) despite the lower levels of FUS1 expression. Indeed, a greater number of multishmoo cells accumulate in the presence of fus3R42R43 (Figure 8D), as if fus3R42R43 allows cells to more rapidly reset a program of shmoo formation. Thus, catalytically inactive Fus3p promotes exit from G1 arrest and reentry into the mitotic cycle. In addition, it may promote the formation of a second projection under conditions that preclude exit from G1 arrest.

View larger version (44K): In this window In a new window Download PPT slide   Figure 8. fus3R42R43 promotes recovery. (A) Halo assays of FUS3 KSS1 and FUS3 kss1 strains (EY2522 and EY2524) harboring either fus3R42R43 (pRT6) or vector (ZM43). Photographs were taken after 2 days of incubation at 30°. (B) Percentage of unbudded cells, (C) percentage of shmoos, and (D) percentage of multishmoos (a multishmoo is defined as a single cell body with two or more projections) during an -factor time course. The FUS3 KSS1 strain (EY2522) harboring either pRT6 or ZM43 was grown as described in Figure 7.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fus3p regulates numerous outputs through distinct functions:
Analysis of the role of Fus3p in mating is complicated by the fact that Fus3p activates the expression of Ste12p-dependent genes involved in many aspects of mating. Thus, any response to mating pheromone attributed to Fus3p could be a secondary consequence of Ste12p activation. Relative output dependencies are not revealed by analysis of a FUS3 kss1 strain, which has maximal FUS1 expression at all levels of G1 arrest and shmoo formation, and maximal G1 arrest at all but the lowest levels of morphogenesis and shmoo formation. Through a comparative analysis of a panel of fus3 point mutants, however, we find functional evidence in favor of a model in which Fus3p regulates G1 arrest, shmoo formation, partner selection, mating, and recovery through the phosphorylation of distinct effectors in addition to Ste12p. Our findings are consistent with initial in vitro analysis suggesting the existence of numerous Fus3p substrates (ELION et al. 1993 ). In addition, they suggest that, while activation of Ste12p is essential for each of these outputs, a high level of Ste12p-dependent expression may not be essential.

What determines Fus3p signaling capacity?
Our results suggest that Fus3p dosage may be an important determinant of signaling. A simple view of signal transduction during mating might be one in which different amounts of active Fus3p kinase are needed to execute different responses, given that Fus3p levels increase severalfold in the presence of -factor (ELION et al. 1990 , ELION et al. 1993 ), and that increasing amounts of -factor are required to elicit transcriptional activation of Ste12p, agglutination, G1 arrest, and shmoo formation, a late response (Figure 4; MOORE 1983 ). This view is consistent with the observation that a CUP1-FUS3 promoter fusion gene fully complements a fus3 null for G1 arrest only when it is expressed at levels equivalent to the FUS3 gene in the presence of -factor (F. FARLEY and E. ELION, unpublished results).

One might imagine that certain fus3 mutations cause differential effects on the various outputs as a consequence of a decrease in the overall level of Fus3p kinase activity and preferential phosphorylation of certain substrates. This possibility fits with our correlative data and the predicted structural defects of some of the fus3 mutants. The fact that on average the mutants are less defective in transcription and most defective in shmoo formation suggests that increasing Fus3p function is required for transcriptional activation, G1 arrest, and shmoo formation. This is consistent with the dosage response curves for the FUS3 kss1 strain (Figure 4). It is possible that persistent activation of Fus3p is required to generate the morphogenesis response.

Our analysis also suggests that factors other than the absolute level of kinase activity affect Fus3p function. Several fus3 mutants that are predicted to be defective in interactions with activators and/or substrates exhibit varying defects in the individual outputs (Table 2; Figure 8). Fus3p localizes in the cytoplasm and nucleus (CHOI et al. 1999 ) and associates with a number of proteins (ELION et al. 1993 ; CHOI et al. 1994 ; KRANZ et al. 1994 ; BARDWELL et al. 1996 ; LYONS et al. 1996 ; TEDFORD et al. 1997 ), suggesting that it participates in a complex array of physical interactions at a number of cellular locales. Thus, Fus3p signaling capacity is likely to derive from a combination of factors that affect its activation and association with regulators and substrates. Such factors may include the relative concentrations of active and inactive Fus3p, regulators and substrates, and the ability of Fus3p to localize to key targets.

Fus3p regulates recovery through a mechanism that is distinct from its ability to promote G1 arrest:
Our analysis provides the first functional evidence that Fus3p promotes recovery through a mechanism that is distinct from its ability to promote G1 arrest. This function may involve the phosphorylation of effectors that promote recovery. Several of the fus3 mutants that appear to have enhanced recovery exhibit normal hyperphosphorylation of Ste7p (fus3-206, fus3-209, and fus3-213), indicating that the -factor resistance in these strains is not caused by a defect in hyperphosphorylation of Ste7p.

Catalytically inactive Fus3p is also likely to promote recovery through the inhibition of active Fus3p and Kss1p. First, catalytically defective mutants can promote -factor resistance without blocking G1 arrest (e.g., fus3-2, fus3-203, fus3-204, fus3-206, fus3-208, and fus3-209). Second, many catalytically defective mutants were found to inhibit the function of active Fus3p (fus3R42R43) and active Kss1p (fus3R42R43, fus3-2, fus3-205, fus3-206, and fus3-210), as assessed by activation of the FUS1 gene. Third, time course analysis demonstrates that catalytically inactive fus3R42R43 hastens recovery in a FUS3 KSS1 strain without blocking G1 arrest or shmoo formation.

How might catalytically inactive Fus3p promote recovery? The most obvious explanation is that inactive forms of Fus3p bind to and sequester upstream activators or downstream substrates from active Fus3p or Kss1p. Fus3p, rather than Kss1p, is likely to be the major target of inhibition, because catalytically inactive Fus3p inhibits activation of FUS1 in a FUS3 kss1 strain nearly equally as well as in a FUS3 KSS1 strain. Consistent with this possibility, catalytically inactive Fus3p forms stable complexes with both activators and substrates (ELION et al. 1993 ; KRANZ et al. 1994 ). Precedence for a dominant negative form of regulation by Fus3p already exists. Catalytically inactive Fus3p has been shown to inhibit the invasive response (MADHANI et al. 1997 ). This inhibition is likely to occur in a wild-type cell because the invasive response occurs in the absence of pheromone when Fus3p is largely inactive. Time course studies of activation of Fus3p by -factor indicate the existence of a large pool of catalytically inactive Fus3p after 60 min (CHOI et al. 1999 ). Thus, the inhibition we observe with Fus3R42R43p is likely to reflect a natural function for Fus3p.

Does catalytically inactive Fus3p play a role in the timing of the pheromone response?
Several results raise the possibility that the regulation of G1 arrest by Fus3p may play a role in the timing of a cell's initiation of, and exit from, a program of terminal differentiation. First, the ability of Fus3p to promote shmoo formation was found to be closely linked to its ability to promote G1 arrest. Second, faster recovery may correlate with an ability to form multiple shmoos (see multishmoo tallies for FUS3 KSS1, fus3R42R43, fus3-204 kss1, and fus3-213 kss1, Table 5; Figure 8). Our previous work suggests that proteins required for shmoo formation and G1 arrest may colocalize with the MAP kinase cascade and be jointly regulated by Fus3p as well as functionally interdependent (LYONS et al. 1996 ). Thus, termination of G1 arrest through inactive forms of Fus3p might provide a simple route to end terminal differentiation and couple it to reentry into the mitotic cycle. Perhaps when signaling is still sufficient for terminal differentiation, the same attenuating mechanisms permit a second shmoo projection to form in an effort to find a mating partner. Our results raise the interesting possibility that the relative level of inactive Fus3p during the course of exposure to pheromone may function as a clock to time the exit from the response.

Kss1p is likely to phosphorylate a subset of the same substrates as Fus3p:
The suppression analysis suggests that, in a wild-type strain, Fus3p performs the bulk of phosphorylation of effector proteins in the mating signaling cascade, with Kss1p playing a much less critical role for many outputs. This view is consistent with earlier analyses (ELION et al. 1991A , ELION et al. 1991B ; GARTNER et al. 1992 ; MA et al. 1995 ). More recently, it has been proposed that Kss1p has no function whatsoever in the mating pathway in the presence of Fus3p, and that it only substitutes when Fus3p is deleted (MADHANI et al. 1997 ). Our findings argue that Kss1p does function in the mating pathway in the presence of Fus3p, but in a much more limited way than Fus3p. First, a kss1 null strain is partially defective in activation of the FAR1 gene and shmoo formation. Second, Kss1p is able to suppress to varying degrees the transcription and G1 arrest defects of fus3 mutants that produce stable Fus3p proteins. That Kss1p does function in the pheromone response pathway in a wild-type cell is consistent with the sterile phenotype of a fus3 kss1 double-null mutant (ELION et al. 1991A , ELION et al. 1991B ), the observation that Kss1p is activated by pheromone in the presence of Fus3p (GARTNER et al. 1992 ; MA et al. 1995 ), and additional studies suggesting that Kss1p provides unique functions for the regulation of G1 arrest and recovery (CHERKASOVA et al. 1999 ). Thus, Fus3p may phosphorylate a number of substrates more readily than Kss1p and/or have substrates that are not phosphorylated by Kss1p. Differential substrate phosphorylation could result from different catalytic efficiencies for substrates (KALLUNKI et al. 1994 ) as well as unequal access to substrates. Kss1p is predominantly nuclear (MA et al. 1995 ), while Fus3p is more evenly distributed in the nucleus and cytoplasm (CHOI et al. 1999 ), consistent with its much greater role in mating.

	ACKNOWLEDGMENTS

We thank R. Tung for constructing the fus3R42R43 mutant, F. Walker for technical assistance during the early stages of this work, B. Errede for pNC318 (Ste7Mp), and J. Wilsbacher for help with structural analysis of the fus3 mutants. We also thank an anonymous reviewer for very useful criticisms of the manuscript. This research was supported by a Young Investigator Grant from the Harcourt Charitable Foundation, National Institutes of Health grant RO1GM46962, a grant from the Council for Tobacco Research, an American Cancer Society Junior faculty award to E.A.E., and a predoctoral National Science Foundation Fellowship to B.S.

Manuscript received August 27, 1998; Accepted for publication January 14, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BARDWELL, L., J. G. COOK, E. C. CHANG, B. R. CAIRNS, and J. THORNER, 1996  Signaling in the yeast pheromone response pathway: specific and high-affinity interaction of the mitogen-activated protein (MAP) kinases Kss1 and Fus3 with the upstream MAP kinase kinase Ste7. Mol. Cell. Biol. 16:3637-3650[Abstract].

BOEKE, J. D., J. TRUEHEART, G. NATSOULIS, and G. R. FINK, 1984  A positive selection for mutants lacking orotidine 5'-phosphate decarboxylase activity in yeast: 5-fluoroorotic acid resistance. Mol. Gen. Genet. 1818:288-291.

BOULTON, T. G., G. D. YANCOPOULOS, J. S. GREGORG, C. SLAUGHTER, and C. MOOMAN et al., 1990  An insulin-stimulated protein kinase similar to yeast kinases involved in cell cycle control. Science 249:64-67[Abstract/Free Full Text].

BRILL, J. A., E. A. ELION, and G. R. FINK, 1994  A role for autophosphorylation revealed by activated alleles of FUS3, the yeast MAP kinase homolog. Mol. Biol. Cell 5:297-312[Abstract].

BRUNET, A. and J. POUYSSEGUR, 1996  Identification of MAP kinase domains by redirecting stress signals into growth factor responses. Science 272:1652-1655[Abstract].

CANAGARAJAH, B., A. KHOKHLATCHEV, M. H. COBB, and E. J. GOLDSMITH, 1997  The structure of phosphorylated MAP kinase Erk2p at 2.5 Å resolution. Cell 90:859-869[Medline].

CHANG, F. and I. HERSKOWITZ, 1990  Identification of a gene necessary for cell cycle arrest by a negative growth factor of yeast: FAR1 is an inhibitor of the G1 cyclin CLN2. Cell 63:999-1011[Medline].

CHERKASOVA, V., D. M. LYONS, and E. A. ELION, 1999  Fus3p and Kss1p control G1 arrest in Saccharomyces cerevisiae through a balance of distinct arrest and proliferative functions that operate in parallel with Far1p. Genetics 151:989-1004[Abstract/Free Full Text].

CHOI, K.-Y., B. SATTERBERG, D. M. LYONS, and E. A. ELION, 1994  Ste5 tethers multiple protein kinases in the MAP kinase cascade required for mating in S. cerevisiae.. Cell 78:499-512[Medline].

CHOI, K.-Y., J. E. KRANZ, S. K. MAHANTY, and E. A. ELION, 1999  Characterization of Fus3 localization: active Fus3 localizes in complexes of varying size and specific activity. Mol. Biol. Cell in press.

CIEJEK, E. and J. THORNER, 1979  Recovery of S. cerevisiae a cells from G1 arrest by -factor requires endopeptidase action. Cell 18:623-635[Medline].

COOK, J. G., L. BARDWELL, S. J. KRON, and J. THORNER, 1996  Two novel targets of the MAP kinase Kss1 are negative regulators of invasive growth in the yeast Saccharomyces cerevisiae.. Genes Dev. 10:2831-2848[Abstract/Free Full Text].

COURCHESNE, W. E., R. KUNISAWA, and J. THORNER, 1989  A putative protein kinase overcomes pheromone-induced arrest of cell cycling in S. cerevisiae.. Cell 58:1107-1119[Medline].

CRAVEN, G. R., E. J. STEERS, J. L. BETHUNE, and C. B. ANFINSEN, 1965  Purification, composition, and molecular weight of the ß-galactosidase of Escherichia coli K12. J. Biol. Chem. 240:2468-2477[Free Full Text].

DORER, R., P. M. PRYCIAK, and L. H. HARTWELL, 1995  Saccharomyces cerevisiae cells execute a default pathway to select a mate in the absence of pheromone gradients. J. Cell Biol. 131:845-861[Abstract/Free Full Text].

DORER, R., C. BOONE, T. KIMBROUGH, J. KIM, and L. H. HARTWELL, 1997  Genetic analysis of default mating behavior in Saccharomyces cerevisiae.. Genetics 146:39-55[Abstract].

ELION, E. A., 1995  Ste5: a meeting place for MAP kinases and their associates. Trends Cell Biol. 5:322-327[Medline].

ELION, E. A. and J. R. WARNER, 1986  An RNA polymerase I enhancer in Saccharomyces cerevisiae.. Mol. Cell. Biol. 6:2089-2097[Abstract/Free Full Text].

ELION, E. A., P. GRISAFI, and G. R. FINK, 1990  FUS3 encodes a cdc2/CDC28-related kinase required for the transition from mitosis into conjugation. Cell 60:649-664[Medline].

ELION, E. A., J. A. BRILL, and G. R. FINK, 1991a  Functional redundancy in the yeast cell cycle: FUS3 and KSS1 have both overlapping and unique functions. Cold Spring Harbor Symp. Quant. Biol. 56:41-49[Abstract/Free Full Text].

ELION, E. A., J. A. BRILL, and G. R. FINK, 1991b  FUS3 represses CLN1 and CLN2 and in concert with KSS1 promotes signal transduction. Proc. Natl. Acad. Sci. USA 88:9392-9396[Abstract/Free Full Text].

ELION, E. A., B. SATTERBERG, and J. E. KRANZ, 1993  Fus3 phosphorylates multiple components of the mating signal transduction cascade: evidence for Ste12 and Far1. Mol. Biol. Cell 5:495-510.

ELION, E. A., J. TRUEHEART, and G. R. FINK, 1995  Fus2 localizes near the site of cell fusion and is required for both cell fusion and nuclear alignment during zygote formation. J. Cell Biol. 130:1283-1296[Abstract/Free Full Text].

ERREDE, B., A. GARTNER, Z. ZHOU, K. NASMYTH, and G. AMMERER, 1993  MAP kinase-related FUS3 from S. cerevisiae is activated by STE7 in vitro.. Nature 362:261-264[Medline].

FENG, Y., L. Y. SONG, E. KINCAID, S. K. MAHANTY, and E. A. ELION, 1998  Functional binding between Gß and the LIM domain of Ste5 is required to activate the MEKK Ste11. Curr. Biol. 8:267-278[Medline].

GARTNER, A., K. NASMYTH, and G. AMMERER, 1992  Signal transduction in Saccharomyces cerevisiae requires tyrosine and threonine phosphorylation of FUS3 and KSS1. Genes Dev. 6:1280-1292[Abstract/Free Full Text].

GIETZ, R. D. and A. SUGINO, 1988  New yeast-Escherichia coli shuttle vectors constructed with in vitro mutagenized yeast genes lacking six-base pair restriction sites. Gene 74:527-534[Medline].

HERSKOWITZ, I., 1995  MAP kinase pathways in yeast: for mating and more. Cell 80:187-197[Medline].

HOFFMAN, C. S. and F. WINSTON, 1987  A ten minute DNA preparation from yeast efficiently releases autonomous plasmids for transformation of E. coli.. Gene 57:267-272[Medline].

ITO, H., Y. FUKADA, K. MURATA, and A. KIMURA, 1983  Transformation of intact yeast cells with alkali cations. J. Bacteriol. 153:163-168[Abstract/Free Full Text].

KALLUNKI, T., B. SU, I. TSIGELNY, H. K. SLUSS, and B. DERIJARD et al., 1994  JNK2 contains a specificity-determining region responsible for efficient c-Jun binding and phosphorylation. Genes Dev. 8:2996-3007[Abstract/Free Full Text].

KRANZ, J. E., B. SATTERBERG, and E. A. ELION, 1994  The MAP kinase Fus3 associates with and phosphorylates the upstream signaling component Ste5. Genes Dev. 8:313-327[Abstract/Free Full Text].

LEBERER, E., D. DIGNARD, D. HARCUS, D. Y. THOMAS, and M. WHITEWAY, 1992  The protein kinase homolog Ste20 is required to link the yeast pheromone response G-protein ß subunits to downstream signaling components. EMBO J. 11:4815-4824[Medline].

LEEUW, T., A. FOUREST-LIEUVIN, C. WU, J. CHENEVERT, and K. CLARK et al., 1995  Pheromone response in yeast: association of Bem1 with proteins of the MAP kinase cascade and actin. Science 270:1210-1213[Abstract/Free Full Text].

LYONS, D. M., S. K. MAHANTY, K.-Y. CHOI, M. MANANDHAR, and E. A. ELION, 1996  The SH3-domain protein Bem1 coordinates MAP kinase cascade activation with cell cycle control. Mol. Cell. Biol. 16:4095-4106[Abstract].

MA, D., J. G. COOK, and J. THORNER, 1995  Phosphorylation and localization of Kss1, a MAP kinase of the Saccharomyces cerevisiae pheromone response pathway. Mol. Biol. Cell 6:889-909[Abstract].

MADHANI, H. D., C. A. STYLES, and G. R. FINK, 1997  MAP kinases with distinct inhibitory functions impart signaling specificity during yeast differentiation. Cell 91:685-694[Medline].

MCCAFFREY, G., F. J. CLAY, K. KELSAY, and G. F. SPRAGUE, JR., 1987  Identification and regulation of a gene required for cell fusion during mating of the yeast Saccharomyces cerevisiae.. Mol. Cell. Biol. 7:2680-2690[Abstract/Free Full Text].

MOORE, S. A., 1983  Comparison of dose-response curves for -factor-induced cell division arrest, agglutination, and projection formation of yeast cells. J. Biol. Chem. 258:13849-13853[Abstract/Free Full Text].

NEIMAN, A. M. and I. HERSKOWITZ, 1994  Reconstitution of a yeast protein kinase cascade in vitro: activation of the yeast MEK homologue STE7 by STE11. Proc. Natl. Acad. Sci. USA 91:3398-3402[Abstract/Free Full Text].

OEHLEN, L. J. W. M., J. D. MCKINNEY, and F. R. CROSS, 1996  Ste12p and Mcm1 regulate cell cycle-dependent transcription of FAR1. Mol. Cell. Biol. 16:2830-2837[Abstract].

PETER, M., A. GARTNER, J. HORECKA, G. AMMERER, and I. HERSKOWITZ, 1993  FAR1 links the signal transduction pathway to the cell cycle machinery in yeast. Cell 73:747-760[Medline].

RAMER, S. A. and R. W. DAVIS, 1993  A dominant truncation allele identifies a gene, STE20, that encodes a putative protein kinase necessary for mating in Saccharomyces cerevisiae.. Proc. Natl. Acad. Sci. USA 90:452-456[Abstract/Free Full Text].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual, Ed. 2, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SATTERBERG, B., 1993 G1 cell cycle arrest in S. cerevisiae, Ph.D. Thesis. Harvard University, Cambridge, MA.

SHERMAN, F., G. R. FINK and J. B. HICKS, 1986 Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SIMON, M., C. DE VIRGILIO, B. SOUZA, J. R. PRINGLE, and A. ABO et al., 1995  Role for the Rho-family GTPase Cdc42 in yeast mating pheromone signal pathway. Nature 376:702-705[Medline].

SONG, O., J. W. DOLAN, Y. O. YUAN, and S. FIELDS, 1991  Pheromone-dependent phosphorylation of the yeast STE12 protein correlates with transcriptional activation. Genes Dev. 5:741-750[Abstract/Free Full Text].

SPRAGUE, G. F., JR., and J. W. THORNER, 1992 Pheromone response and signal transduction during the mating process of Saccharomyces cerevisiae. pp. 657–744 in The Molecular and Cellular Biology of the Yeast Saccharomyces, edited by E. JONES, J. PRINGLE and J. BROACH. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

STEVENSON, B. J., N. RHODES, B. ERREDE, and G. F. SPRAGUE, JR., 1992  Constitutive mutants of the protein kinase STE11 activate the yeast pheromone response pathway in the absence of the G protein. Genes Dev. 6:1293-1304[Abstract/Free Full Text].

TEDFORD, K., S. KIM, D. SA, K. STEVENS, and M. TYERS, 1997  Regulation of the mating pheromone and invasive growth responses in yeast by two MAP kinase substrates. Curr. Biol. 7:228-238[Medline].

TRUEHEART, J., J. D. BOEKE, and G. R. FINK, 1987  Two genes required for cell fusion during conjugation: evidence for a pheromone-induced surface protein. Mol. Cell. Biol. 7:2316-2328[Abstract/Free Full Text].

TYERS, M. and B. FUTCHER, 1993  Far1 and Fus3 link the mating pheromone signal transduction pathway to three G1-phase Cdc28 kinase complexes. Mol. Cell. Biol. 13:5659-5660[Abstract/Free Full Text].

VALTZ, N., M. PETER, and I. HERSKOWITZ, 1995  FAR1 is required for oriented polarization of yeast cells in response to mating pheromones. J. Cell Biol. 131:863-873[Abstract/Free Full Text].

ZHANG, F., A. STRAND, D. ROBBINS, M. H. COBB, and E. J. GOLDSMITH, 1994  Atomic structure of the MAP kinase Erk2 at 2.3 Å resolution. Nature 367:704-710[Medline].

ZHOU, Z., A. GARTNER, R. CADE, G. AMMERER, and B. ERREDE, 1993  Pheromone-induced signal transduction in Saccharomyces cerevisiae requires the sequential function of three protein kinases. Mol. Cell. Biol. 13:2069-2080[Abstract/Free Full Text].

This article has been cited by other articles:

				A. I. Goranov, M. Cook, M. Ricicova, G. Ben-Ari, C. Gonzalez, C. Hansen, M. Tyers, and A. Amon The rate of cell growth is governed by cell cycle stage Genes & Dev., June 15, 2009; 23(12): 1408 - 1422. [Abstract] [Full Text] [PDF]

				R. J. Taylor, D. Falconnet, A. Niemisto, S. A. Ramsey, S. Prinz, I. Shmulevich, T. Galitski, and C. L. Hansen Dynamic analysis of MAPK signaling using a high-throughput microfluidic single-cell imaging platform PNAS, March 10, 2009; 106(10): 3758 - 3763. [Abstract] [Full Text] [PDF]

				K. D. Craven, H. Velez, Y. Cho, C. B. Lawrence, and T. K. Mitchell Anastomosis Is Required for Virulence of the Fungal Necrotroph Alternaria brassicicola Eukaryot. Cell, April 1, 2008; 7(4): 675 - 683. [Abstract] [Full Text] [PDF]

				L. Yu, M. Qi, M. A. Sheff, and E. A. Elion Counteractive Control of Polarized Morphogenesis during Mating by Mitogen-activated Protein Kinase Fus3 and G1 Cyclin-dependent Kinase Mol. Biol. Cell, April 1, 2008; 19(4): 1739 - 1752. [Abstract] [Full Text] [PDF]

				R. K. Esch, Y. Wang, and B. Errede Pheromone-Induced Degradation of Ste12 Contributes to Signal Attenuation and the Specificity of Developmental Fate Eukaryot. Cell, December 1, 2006; 5(12): 2147 - 2160. [Abstract] [Full Text] [PDF]

				A. Fleissner, S. Sarkar, D. J. Jacobson, M. G. Roca, N. D. Read, and N. L. Glass The so Locus Is Required for Vegetative Cell Fusion and Postfertilization Events in Neurospora crassa Eukaryot. Cell, May 1, 2005; 4(5): 920 - 930. [Abstract] [Full Text] [PDF]

				Y. Wang, W. Chen, D. M. Simpson, and E. A. Elion Cdc24 Regulates Nuclear Shuttling and Recruitment of the Ste5 Scaffold to a Heterotrimeric G Protein in Saccharomyces cerevisiae J. Biol. Chem., April 1, 2005; 280(13): 13084 - 13096. [Abstract] [Full Text] [PDF]

				K. S. Bruno, F. Tenjo, L. Li, J. E. Hamer, and J.-R. Xu Cellular Localization and Role of Kinase Activity of PMK1 in Magnaporthe grisea Eukaryot. Cell, December 1, 2004; 3(6): 1525 - 1532. [Abstract] [Full Text] [PDF]

				L. Staleva, A. Hall, and S. J. Orlow Oxidative Stress Activates FUS1 and RLM1 Transcription in the Yeast Saccharomyces cerevisiae in an Oxidant-dependent Manner Mol. Biol. Cell, December 1, 2004; 15(12): 5574 - 5582. [Abstract] [Full Text] [PDF]

				A. Flotho, D. M. Simpson, M. Qi, and E. A. Elion Localized Feedback Phosphorylation of Ste5p Scaffold by Associated MAPK Cascade J. Biol. Chem., November 5, 2004; 279(45): 47391 - 47401. [Abstract] [Full Text] [PDF]

				S. Maleri, Q. Ge, E. A. Hackett, Y. Wang, H. G. Dohlman, and B. Errede Persistent Activation by Constitutive Ste7 Promotes Kss1-Mediated Invasive Growth but Fails To Support Fus3-Dependent Mating in Yeast Mol. Cell. Biol., October 15, 2004; 24(20): 9221 - 9238. [Abstract] [Full Text] [PDF]

				A. B. Kusari, D. M. Molina, W. Sabbagh Jr., C. S. Lau, and L. Bardwell A conserved protein interaction network involving the yeast MAP kinases Fus3 and Kss1 J. Cell Biol., January 19, 2004; 164(2): 267 - 277. [Abstract] [Full Text] [PDF]

				V. M. Blanc and J. Adams Evolution in Saccharomyces cerevisiae: Identification of Mutations Increasing Fitness in Laboratory Populations Genetics, November 1, 2003; 165(3): 975 - 983. [Abstract] [Full Text] [PDF]

				Y. Wang and E. A. Elion Nuclear Export and Plasma Membrane Recruitment of the Ste5 Scaffold Are Coordinated with Oligomerization and Association with Signal Transduction Components Mol. Biol. Cell, June 1, 2003; 14(6): 2543 - 2558. [Abstract] [Full Text] [PDF]

				Y. Wang and H. G. Dohlman Pheromone-dependent Ubiquitination of the Mitogen-activated Protein Kinase Kinase Ste7 J. Biol. Chem., May 3, 2002; 277(18): 15766 - 15772. [Abstract] [Full Text] [PDF]

				E. A. Elion The Ste5p scaffold J. Cell Sci., March 13, 2002; 114(22): 3967 - 3978. [Abstract] [Full Text] [PDF]

				S. A. Burchett, A. Scott, B. Errede, and H. G. Dohlman Identification of Novel Pheromone-response Regulators through Systematic Overexpression of 120 Protein Kinases in Yeast J. Biol. Chem., July 6, 2001; 276(28): 26472 - 26478. [Abstract] [Full Text] [PDF]