Mutations in the YRB1 Gene Encoding Yeast Ran-Binding-Protein-1 That Impair Nucleocytoplasmic Transport and Suppress Yeast Mating Defects

Corresponding author: Markus Künzler, Ruprecht-Karls-Universität Heidelberg, Biochemie-Zentrum Heidelberg (BZH), Im Neuenheimer Feld 328, 4. OG, D-69120 Heidelberg, Germany., markus.kuenzler{at}urz.uni-heidelberg.de (E-mail)

Communicating editor: M. JOHNSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We identified two temperature-sensitive (ts) mutations in the essential gene, YRB1, which encodes the yeast homolog of Ran-binding-protein-1 (RanBP1), a known coregulator of the Ran GTPase cycle. Both mutations result in single amino acid substitutions of evolutionarily conserved residues (A91D and R127K, respectively) in the Ran-binding domain of Yrb1. The altered proteins have reduced affinity for Ran (Gsp1) in vivo. After shift to restrictive temperature, both mutants display impaired nuclear protein import and one also reduces poly(A)⁺ RNA export, suggesting a primary defect in nucleocytoplasmic trafficking. Consistent with this conclusion, both yrb1^ts mutations display deleterious genetic interactions with mutations in many other genes involved in nucleocytoplasmic transport, including SRP1 (-importin) and several ß-importin family members. These yrb1^ts alleles were isolated by their ability to suppress two different types of mating-defective mutants (respectively, fus1 and ste5^ts), indicating that reduction in nucleocytoplasmic transport enhances mating proficiency. Indeed, in both yrb1^ts mutants, Ste5 (scaffold protein for the pheromone response MAPK cascade) is mislocalized to the cytosol, even in the absence of pheromone. Also, both yrb1^ts mutations suppress the mating defect of a null mutation in MSN5, which encodes the receptor for pheromone-stimulated nuclear export of Ste5. Our results suggest that reimport of Ste5 into the nucleus is important in downregulating mating response.

MATING in the yeast Saccharomyces cerevisiae is the culmination of a complex series of events required for cellular and nuclear fusion of two haploid cells of opposite mating type (SPRAGUE and THORNER 1992 ). Mating pheromones (secreted peptides) bind to G-protein-coupled receptors, stimulating a mitogen-activated protein kinase (MAPK) cascade (BARDWELL et al. 1994 ) that evokes dramatic changes in gene transcription, cell cycle arrest, and pronounced alterations of cell morphology and nuclear reorganization (LEBERER et al. 1997 ; STONE et al. 2000 ). Pheromone-activated and pheromone-induced gene products required for cell fusion are deposited at a localized site on the plasma membrane at the leading edge of a mating projection ("shmoo tip"; MADDEN and SNYDER 1998 ). Proteins required for nuclear fusion are recruited to the nucleus (ROSE 1996 ). How the signal initiated at the plasma membrane is transmitted into the nucleus to activate gene expression is still unclear. Two components of the pathway, Ste5 (PRYCIAK and HUNTRESS 1998 ; MAHANTY et al. 1999 ) and Far1 (BLONDEL et al. 1999 ), shuttle between the nucleus and the cytosol, are predominantly nuclear in naïve cells, but are rapidly ejected from the nucleus in response to pheromone, as observed for the regulated import and export of other nuclear proteins (KAFFMAN and O'SHEA 1999 ).

Proteins and protein-RNA complexes cross the nuclear envelope through nuclear pores comprised of 50 different proteins, termed nucleoporins (RYAN and WENTE 2000 ). Nucleocytoplasmic transport also requires soluble factors. Transport receptors for both import and export (ß-importin and its relatives) bind their cargo and shuttle between the cytosol and the nucleoplasm (GORLICH and KUTAY 1999 ). The small Ras-like GTPase Ran and its associated factors confer directionality to transport (MACARA et al. 2000 ). In S. cerevisiae, GSP1 and GSP2 encode Ran isoforms (BELHUMEUR et al. 1993 ; KADOWAKI et al. 1993 ). Ran exists predominantly in its GTP-bound form in the nucleus; in the cytosol, Ran is mainly GDP-bound. This asymmetry is imposed by the subcellular distribution of Ran regulators: the Ran-specific guanine-nucleotide exchange factor (RanGEF1), the PRP20/SRM1/MTR1 gene product in S. cerevisiae, is confined to the nucleus, whereas the Ran-specific GTPase-activating protein (RanGAP1), the RNA1 gene product in S. cerevisiae, is located in the cytoplasm. GSP1, PRP20, and RNA1 are all essential genes, and recessive mutations in all three block nuclear protein import and poly(A)⁺ RNA export (CORBETT and SILVER 1997 ; OKI et al. 1998 ).

Transport receptors bind specifically to the GTP-bound form of Ran via a conserved domain at their N termini (GORLICH and KUTAY 1999 ). RanGTP-binding to an export receptor enhances its affinity for an export substrate; conversely, binding of RanGTP to an import receptor prevents the binding of an import substrate. Hence, high RanGTP in the nuclear compartment potentiates association of export cargo with export receptors and triggers release of import cargo from import receptors. Ran-binding-protein-1 (RanBP1) is another protein that binds specifically to RanGTP (KUNZLER et al. 2000 ; PLAFKER and MACARA 2000 ). RanBP1 contains a conserved Ran-binding domain (RBD) of 140 residues (BEDDOW et al. 1995 ; VETTER et al. 1999 ), which is necessary and sufficient for high-affinity binding of RanGTP and for nuclear export of RanBP1, at least in yeast (KUNZLER et al. 2000 ). Homologous RBDs are found in other nuclear proteins, like vertebrate RanBP2/NUP358 (YOKOYAMA et al. 1995 ) and RanBP3 (MUELLER et al. 1998 ), and S. cerevisiae nuclear proteins, S. cerevisiae Nup2 (BOOTH et al. 1999 ) and Yrb2 (TAURA et al. 1998 ). Transport receptors block stimulation of Ran-mediated GTP hydrolysis by RanGAP1; in contrast, RanBP1 acts as a coactivator of RanGAP1-stimulated GTP hydrolysis by Ran and, moreover, is required for nucleotide hydrolysis when RanGTP is bound to a transport receptor (GORLICH and KUTAY 1999 ). These biochemical activities, and the fact that RanBP1 is abundant, shuttles between the nucleus and the cytoplasm, but is found almost exclusively in the cytosol at steady state (KUNZLER et al. 2000 ; PLAFKER and MACARA 2000 ), suggest that RanBP1 has a major role in the cytoplasm both in recycling of transport receptors and in release of export cargo (PETERSON et al. 2000 ). Consistent with this view, S. cerevisiae YRB1, encoding yeast RanBP1, is essential for cell viability and is required for both nuclear protein import and poly(A)⁺ RNA export (SCHLENSTEDT et al. 1995 ).

A link between yeast mating and the Ran GTPase cycle was the identification of the srm1-1 mutation, now known to reside in RanGEF1, which suppressed the mating defect of cells lacking pheromone receptors and increased the basal expression of a pheromone-responsive reporter gene (CLARK and SPRAGUE 1989 ). Another connection between mating and nucleocytoplasmic transport was the finding that the ste21 mutation, identified in a screen for enhancers of the mating defect of a temperature-sensitive (ts) mutation in STE4 (encoding the ß-subunit of the pheromone receptor-coupled heterotrimeric G-protein; AKADA et al. 1996 ), resides in MSN5, encoding the nuclear receptor for pheromone-stimulated export of Ste5 (MAHANTY et al. 1999 ) and Far1 (BLONDEL et al. 1999 ). As described here, we isolated a yrb1^ts mutation as a suppressor of a mutant (fus1) defective in the cell fusion step of mating. Fus1 is a pheromone-induced, O-glycosylated, integral membrane protein that acts at a late stage in mating (TRUEHEART and FINK 1989 ). While mapping this yrb1 mutation, we found that it was allelic to a mutation isolated (KATZ et al. 1987 ) as an extragenic suppressor of missense mutations in STE5, which encodes a scaffold protein for the pheromone-activated MAPK cascade (ELION 1998 ). As described here, both yrb1 mutations cause clear defects in nucleocytoplasmic trafficking of various proteins, including Ste5, and are able to suppress the mating defect of a msn5 mutant. These data suggest that preventing efficient reimport of Ste5 after its pheromone-induced release from the nucleus sustains the mating-competent state.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and growth conditions:
Yeast strains used in this study are listed in Table 1. Strains JTY2483 and JTY2484 were obtained by backcrossing strain 381G-42E-P1 three times against either YPH499 or YPH500. msn5::TRP1 strain HMK30 was derived from strain LH90 (BLONDEL et al. 1999 ) by three consecutive backcrosses against the W303-1A derivatives, CRY1 or CRY2. DNA-mediated transformation of yeast cells was performed using a modified version of the lithium acetate method (GIETZ et al. 1992 ). The fus1 mutation deletes 90% of the coding sequence (from the FUS1 promoter to codon 460) and was constructed by a two-step gene disruption method (BOEKE et al. 1987 ). Heterozygous diploid strain JTY2501 was derived from CRY3 by transplacing the YRB1 locus on one homolog of chromosome IV via transformation with an EcoRI-XbaI fragment containing the yrb1::HIS3 construct excised from plasmid pMK112n (Table 2). To construct strain HMK21, JTY2501 was transformed with plasmid pMK103, sporulated, and a MATa His⁺ Ura⁺ 5-fluoro-orotic acid (5-FOA)-sensitive spore was chosen. Strain JTY2486 was obtained by transformation of CRY1 with an EcoRI-SpeI fragment containing the nup2::HIS3 construct excised from plasmid pJON115 (LOEB et al. 1993 ). Strain HMK29 was constructed analogously using a BamHI-HindIII fragment containing a gsp2::LEU2 construct (KADOWAKI et al. 1993 ). Correct transplacements were verified by Southern hybridization analysis.

Unless indicated otherwise, yeast cells were propagated at 30°. Rich medium (YP), synthetic complete medium (SC), and synthetic minimal medium (SM) were prepared as described (KAISER et al. 1994 ). Glucose (Glc) or raffinose (Raf) were added as carbon source at a final concentration of 20 g/liter after autoclaving; induction with galactose (Gal) was performed by adding Gal (final concentration 2%) to Raf-grown cells. Drop-out media (SC lacking the appropriate nutrients) were used to maintain selection for plasmids. Agar plates containing 5-FOA were prepared as described by BOEKE et al. 1987 . Escherichia coli strain DH5 (HANAHAN 1983 ) was used for propagation of plasmid DNAs. Bacteria were cultivated using standard methods (SAMBROOK et al. 1989 ).

Quantitative mating assays:
Quantitative mating assays were performed as previously described (SPRAGUE 1991 ). Briefly, MATa strains to be tested and MAT tester strains were pregrown at 26° to midlogarithmic phase in selective and rich medium, respectively. Cells were washed with water and 10⁶ cells of the MATa strains to be tested were mixed with 10⁷ cells of the MAT tester strain. In the case of the experiment shown in Table 3, the mixture was spread directly onto precooled (14°) SMGlc plates, the plates were incubated for 3 days at 14°, and then for 3 days at room temperature. The resultant diploid colonies were counted and normalized to the titer of input MATa cells (determined by plating the same dilutions on plates selective for the MATa strain to be tested and incubating for 3 days at room temperature). In the case of the experiment shown in Table 5, the mating mixture was collected on a 0.45-µm pore filter and incubated for 6 hr at 30° on YPGlc. After the incubation, cells were resuspended in SMGlc medium and plated in appropriate dilutions onto SMGlc plates with appropriate nutrients to select for diploids. As a control for the number of viable MATa cells used in the mating mixture, 10⁶ cells of the MATa cells were collected on a separate filter, incubated as above, resuspended in YPGlc, and plated on YPGlc plates at appropriate dilutions. Mating efficiency was expressed as percentage of the input MATa haploids that formed diploid colonies.

View this table: In this window In a new window   Table 3. The yrb1-51 (sfo1-1) mutation suppresses the cold-sensitive mating defect of a fus1 mutant

View this table: In this window In a new window   Table 5. Suppression of the partial mating defect of a msn5 null mutant by the yrb1-51 and yrb1-52 mutations

Isolation of yrb1-51:
JTY2024 (MATa fus1) was mutagenized with ethyl methanesulfonate (10⁸ cells/ml; 3% EMS; 1 hr) to 25% survival, and spread on YPGlc plates. After 3 days at 28°, 72,000 colonies were replica plated onto precooled (14°) SMGlc plates containing uracil (20 mg/liter), on which 2 A_{600 nm} units of JY416 (MAT fus1) cells had been spread. These plates were incubated for 4 days at 14°. Candidate clones that gave a positive mating response (35 colonies total) were restreaked from the master plate, retested for suppression, and examined for their ability to grow at various temperatures. A single isolate (JTY2025) displayed a ts phenotype that cosegregated with the ability to suppress the mating defect of the fus1 cells at 14° (data not shown). The mutation conferring these phenotypes was initially named sfo1-1 (suppressor of fus one). In the course of these crosses, it was shown, first, that sfo1 was tightly linked to trp1 (no recombinants in 31 tetrads; distance 1.6 cM) and, second, by complementation tests, that the sfo1-1 mutation was allelic to stp52, another suppressor of mating defects that was mapped to the same region (KATZ et al. 1987 ). Because cloning of the corresponding wild-type DNA by complementation of the ts phenotype of sfo1-1 cells (see below) revealed that it is identical to the YRB1 gene, and because sequence analysis (see below) demonstrated that both the sfo1-1 and stp52 mutations reside in the YRB1 locus, these alleles were renamed yrb1-51 and yrb1-52, respectively.

Recovery and analysis of yrb1^ts alleles:
The base sequence alterations corresponding to the yrb1-51 and yrb1-52 mutations were determined by cloning and sequencing of DNA isolated from the mutants. The polymerase chain reaction (PCR) was used to amplify 636-bp products comprising the entire YRB1 open reading frame (ORF) using genomic DNA from JTY2026 (yrb1-51) and 381G-42E-P1 (yrb1-52) as the template and oligonucleotide primers, 5'-GGG GAT CCG AAT GTC TAG CGA AGA TAA G-3' (OSFO1) and 5'-GGT CTA GAC GCA AGT AAC AAG C-3' (OSFO5), which corresponded, respectively, to positions -2 to +18 and +635 to +616 of the 201-codon YRB1 sequence (where +1 is the first base of the initiator codon of the ORF) and included restriction sites at their 5'-ends to facilitate cloning of the PCR products. Reaction products were isolated, digested with BamHI and XbaI, and inserted into E. coli vector pUC19 for sequencing. Nucleotide sequence of multiple inserts was determined on both strands using the M13/pUC universal and M13/pUC reverse sequencing primers (New England Biolabs, Beverly, MA) and, when necessary, sequence-specific primers. The single-base-pair mutations recovered were tested for their ability to confer a ts phenotype by first substituting the mutant YRB1 ORFs (excised as SalI fragments from the pUC19 derivatives) for the corresponding segment in pMK103 and then introducing the entire yrb1-51 and yrb1-52 genes as EcoRI-XbaI fragments (excised from the pMK103 derivatives) into pRS314, yielding the TRP1-marked plasmids pMK277 and pMK278, respectively. Finally, HMK21 (yrb1 [pMK103]) was transformed with either pMK277, pMK278, or a control plasmid (pMK275) carrying the normal YRB1 gene, plated on 5-FOA plates at 23° to select against the resident URA3-marked YRB1-containing plasmid (pMK103), and the resulting isolates were analyzed for their ability to grow at elevated temperature.

Construction of plasmids:
Standard techniques were used for the manipulation of recombinant DNA (SAMBROOK et al. 1989 ). Plasmid DNA from E. coli was isolated according to DEL SAL et al. 1988 . Unless specified otherwise, PCR amplifications were performed using Vent DNA polymerase (New England Biolabs). Correct sequence of PCR-generated constructs was verified by nucleotide sequence analysis. Plasmids used in this study are listed in Table 2.

The YRB1 gene was isolated from a yeast genomic library (CARLSON and BOTSTEIN 1982 ) carried in a high-copy-number yeast/E. coli shuttle vector (YEp24) by virtue of its ability to complement the ts phenotype of the yrb1-51 mutant. The smallest original isolate (pSB415) contained the YRB1 gene as well as the neighboring gene NTH1 encoding neutral trehalase. Subsequent subcloning localized the complementing activity to a 1.3-kb chromosomal EcoRI-XbaI fragment containing only YRB1, which was used to construct pMK102, pMK103, and pMK275.

To construct a plasmid (pMK112n) carrying the yrb1::HIS3 deletion construct, an internal BglII fragment was excised from the YRB1-containing insert in pMK101 and replaced by a BamHI fragment containing the HIS3 gene, which was inserted in the same transcriptional orientation as YRB1. Construction of plasmid pNOPPATA-GSP1G21V, which expresses a GTPase-defective mutant form of Gsp1, Gsp1(G21V), fused at its N terminus to a cleavage site (ENLYEQG) for tobacco etch virus (TEV) protease and to two immunoglobulin G (IgG) binding domains of Protein A (ProtA), under control of the NOP1 promoter and the ADH1 terminator, has been described previously (HELLMUTH et al. 1998 ). Plasmids pMK294-51 and pMK294-52, and pMK291-wt, pMK291-51, and pMK291-52, expressing Yrb1-green fluorescent protein (GFP) fusions under control of the authentic YRB1 promoter, were constructed by replacing an internal BglII fragment in the YRB1 coding sequence in pMK284n, which expresses a functional Yrb1-GFP chimera (HELLMUTH et al. 1998 ), with the corresponding fragments from yrb1-51 and yrb1-52, followed by subsequent recloning of the respective YRB1-GFP gene fusions as EcoRI-NotI fragments into pRS424.

Fusions of full-length Yrb1 to the Gal4 transcriptional activation domain (TAD) and full-length Gsp1(G21V) to the E. coli LexA DNA-binding domain (DBD) were generated via PCR, using the two-hybrid vectors pGAD424 and pBTM116, respectively. Fragments comprising the entire YRB1 ORF were synthesized using 5'-CCG AAT TCG GTC CAG GTG GTA GCG AAG ATA AGA AAC CTG TCG-3' (OSFO15) and the M13/pUC reverse sequencing primer (New England Biolabs) as the primers and pUC19 carrying the chromosomal YRB1-containing EcoRI-XbaI fragment (pMK102), or pUC19 carrying the corresponding fragments from the yrb1-51 or yrb1-52 ORFs, as templates. The PCR products were digested with EcoRI and PstI and inserted into the corresponding sites in pGAD424, yielding pMK199-wt, pMK199-51, and pMK199-52, respectively. Similarly, a fragment comprising the entire ORF coding for Gsp1(G21V) was produced using 5'-GCG AGG CCT TGC CCC AGC TGC TAA CGG TGA AG-3' (OGSP7) and RSET (5'-AAC TGC AGC CAA CTC AGC TTC C-3') as the primers, and E. coli expression vector pRSETB (Invitrogen, Carlsbad, CA) carrying a PCR-mutated genomic PvuII-HindIII fragment coding for Gsp1(G21V) as the template. The resulting PCR product was cleaved with StuI and PstI and inserted into the SmaI and PstI sites of pBTM116, yielding plasmid pMK195-GV.

Plasmid YEplac195-AU-L25_NLS-GFP was derived from YEplac195-ADE2-URA3-L25-GFP (HURT et al. 1998 ) by removing most of the RPL25 coding sequence, except for the 5'-end that encodes the first 52 residues of L25 (and contains an intron), using a two-step PCR procedure (GIEBEL and SPRITZ 1990 ; mutagenic primer, 5'-GGG ACA ACT CCA GTG AAA AGT CTT CTC TTT GCT CTC GAG TGG AAC AGC CTT GGA AGC-3'; O. GADAL and E. HURT, unpublished results). Plasmid pNOPGFP2L-STE5 expressing a GFP-Ste5 fusion protein from a multicopy-plasmid under control of the constitutive NOP1-promoter was constructed by inserting a PCR-generated STE5 NcoI-BamHI fragment comprising the entire ORF (using primers OSTE5-1, 5'-GGGGGGCCATGGGTATGGAAACTCCTACAGAC-3', and OSTE5-2, 5'-GGGGGGATCCCTATATATAATCCATATGG-3') into pNOPPATA (HELLMUTH et al. 1998 ) and subsequent recloning of the insert as PstI fragment into pNOPGFP2L. This vector is based on pRS425 and contains a 1.4-kb BamHI-PstI NOP1p-GFP cassette (K. HELLMUTH and E. HURT, unpublished results).

Preparation of rabbit polyclonal anti-Yrb1 antiserum:
To generate a (His)₆-Yrb1 fusion protein containing all but the first 10 residues of Yrb1, the corresponding YRB1 coding sequence was excised as a SalI fragment from pMK102 and ligated into the XhoI site of pRSETA (Invitrogen), yielding pMK104. For expression in E. coli, strain BL21(DE3)/pLysS (STUDIER 1991 ) was transformed with pMK104 and the fusion protein was induced by addition of isopropyl-ß-D-thio-galactopyranoside (IPTG) to a final concentration of 0.4 mM followed by incubation at 37° for 2 hr. Selection for pMK104 had to be maintained by adding 50 mg/liter carbenicillin (Sigma, St. Louis), a more stable derivative of ampicillin, to the medium because the fusion protein was relatively toxic to the cells. The fusion protein was purified from E. coli using Ni²⁺-chelate affinity chromatography (Ni-NTA resin; QIAGEN, Chatsworth, CA), according to the manufacturer's recommendations. The purified protein was used to raise polyclonal antisera in two adult female New Zealand White rabbits (nos. 1390 and 1391), following standard protocols (HARLOW and LANE 1988 ), using 600 µg of protein in 50% Titermax (CytRx, Norcross, GA) for the first immunization and 400 µg of protein in 50% incomplete Freund's adjuvant (Sigma) for each of two subsequent immunizations administered after 3 and 5 weeks, respectively. Bleeds were taken after 4 weeks (2 ml), 6 weeks (50 ml), 7 weeks (2 ml), and 8 weeks (terminal) and stored in 0.02% sodium azide at -70°. For detection of Yrb1 by immunoblotting the resulting antisera (nos. 1390 and 1391) were used as primary antibodies at a dilution of 1:5000.

Two-hybrid assay:
To assess interactions between LexA(DBD)-Gsp1(G21V) and Gal4(TAD)-Yrb1 fusion proteins, strain JTY2500 harboring the E. coli lacZ gene under control of eight LexA-binding sites was cotransformed with the appropriate pBTM116- and pGAD424-based plasmids. Transformants were grown in SCGlc medium lacking leucine and tryptophan to midexponential phase (A_{546 nm} = 1) and assayed for ß-galactosidase acitivity as described previously (KUNZLER and HURT 1998 ).

Preparation of yeast cell extracts:
Yeast cells were washed once with one culture volume of cold phosphate-buffered saline (PBS), aliquoted into 1.5-ml microcentrifuge tubes (20 A_{546 nm} units per tube), and stored as pellets at -70°. Frozen cell pellets were thawed by adding 0.2 ml cold lysis buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 20 mM MgCl₂, 10% glycerol, 2 mM DTT, and 1 mM PMSF) and lysed by vigorous vortexing with 0.2 g of acid-washed glass beads (0.45–0.6 mm diameter) for six 30-sec periods (separated by 1-min periods of cooling on ice). The lysate was clarified by centrifugation for 5 min at 13,000 x g at 4° and the protein concentration was determined by a dye-binding method (BRADFORD 1976 ) using commercially available reagents (Bio-Rad, Hercules, CA) and bovine serum albumin (BSA) as the standard.

Purification of ProtA-TEV-Gsp1(G21V) from yeast:
Transformants of wild-type strain CRY1, coexpressing ProtA-TEV-Gsp1(G21V) from pNOPPATA-GSP1G21V and either Yrb1-GFP, Yrb1(A91D)-GFP, or Yrb1(R127K)-GFP from plasmids pMK284n, pMK294-51, or pMK294-52, respectively, were grown in selective medium at 26° to a A_{546 nm} = 1.5. Purification of Gsp1(G21V) from these cells was performed essentially as described (HELLMUTH et al. 1998 ), with the minor modification that universal buffer (KUNZLER and HURT 1998 ) was used throughout the purification, including cell lysis, washing steps, and elution. Elution by cleavage with TEV-protease (GIBCO-BRL, Gaithersburg, MD) was performed by incubation for 1 hr at room temperature.

Nuclear protein import and RNA export assays:
Temperature-sensitive mutants, and their otherwise isogenic wild-type strains, containing plasmids that express constitutively nuclear transport substrates fused to GFP(S65T), namely SV40_NLS-Gal4(TAD)-GFP (pGADGFP), GFP-Npl3 (pNOPGFPAU-NPL3), and L25_NLS-GFP (YEp195-AU-L25_NLS-GFP), were cultivated to early exponential phase (A_{546 nm} = 0.5) in selective SCGlc medium at 23°, split into two equal portions, and incubated at either 23° or 37° for various periods of time. Strains carrying plasmids expressing SV40_NLS-GFP-ß-galactosidase (pPS817) under control of the GAL1 promoter were pregrown to early exponential phase (A_{546 nm} = 0.5) in selective SCRaf medium at 23° before Gal (2%) was added to the cultures and the cells were incubated at 23° for another hour (to allow mRNA synthesis and export). The induced cultures were split into two equal portions, and one portion was shifted to 37° for 3 hr, while the other portion was maintained at 23° for the same period. Fluorescence microscopy of living yeast cells expressing GFP fusion proteins was done according to HELLMUTH et al. 1998 . Cells were concentrated by brief centrifugation and resuspended in the residual growth medium without any washing steps. To assay mRNA export, cells were cultivated in YPGlc medium as described above for strains harboring constitutively expressed GFP fusion proteins. Poly(A)⁺ RNA was localized by in situ hybridization as described previously (SEGREF et al. 1997 ).

Miscellaneous:
SDS-PAGE and immunoblotting were conducted as described previously (KUNZLER and HURT 1998 ). Multiple sequence alignment was done using the CLUSTALW 1.7 (THOMPSON et al. 1994 ) and BOXSHADE 3.21 [Bioinformatics group of the Swiss Institute for Experimental Cancer Research (ISREC)] programs. Identities to the Yrb1 sequence were calculated on the basis of pairwise alignments using the ALIGN algorithm from the FASTA package (PEARSON and LIPMAN 1988 ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of yrb1^ts mutations as suppressors of mating defects:
The mating deficiency of a fus1 mutant is much more pronounced at 14° than at 30°. At 14°, diploid formation in a MATa fus1 x MAT fus1 cross is <0.5% that of a MATa FUS1 x MAT fus1 cross (Table 3). A screen for extragenic suppressors of this mating defect (see MATERIALS AND METHODS) yielded a single mutation, sfo1-1. This suppressor mutation reproducibly enhanced mating competence of a fus1 mutant 30–50-fold (Table 3), but did not fully restore mating proficiency to the level of a FUS1 cell. The suppressor segregated 2:2 through two backcrosses against a fus1 strain and cosegregated with a recessive ts growth defect (in >15 tetrads analyzed per cross). Genetic mapping of the mutation to the right arm of chromosome IV between CEN4 and the TRP1 gene (data not shown), complementation of the ts growth defect by the wild-type YRB1 gene on a plasmid (data not shown), elimination of the suppression phenotype by plasmid-borne YRB1 (Table 3), and nucleotide sequencing of the mutant DNA (see below) all established that sfo1-1 was a recessive allele of YRB1, which encodes the homolog of mammalian RanBP1 (OUSPENSKI et al. 1995 ; SCHLENSTEDT et al. 1995 ). Hence, sfo1-1 was redesignated yrb1-51.

A recessive ts mutation, stp52 (sterile pseudoreversion), closely linked to TRP1, was isolated as an extragenic suppressor of the mating defect of a MATaste5-3 x MAT ste5-3 cross at restrictive temperature (KATZ et al. 1987 ). The stp52 mutation also suppressed other ste5, ste4, and ste7 missense (ts) alleles (KATZ et al. 1987 ). Although the linkage analysis reported by KATZ et al. 1987 assigned the stp52 mutation to the opposite side of the TRP1 locus from yrb1-51, we found that a yrb1-51/stp52 diploid strain was still ts, and that the ts growth defect of the stp52 mutant could be completely rescued by the cloned YRB1 gene on a plasmid (data not shown). These results demonstrated that the stp52 mutation was another recessive allele of YRB1, as was confirmed by sequencing of the mutant DNA (see below). Hence, stp52 was redesignated yrb1-52.

Phenotypic characterization of the yrb1-51 and yrb1-52 mutations:
To understand how alterations in YRB1 can suppress mating-defective mutants, we examined, first, the physiology of the yrb1 mutants. At 23°, yrb1-51 mutant cells grew nearly as well as wild-type cells, whereas the yrb1-52 mutant cells displayed impaired growth already under these conditions; both yrb1-51 and yrb1-52 cells ceased growth and lost viability within 3–6 hr after shift to 37° (data not shown). Similar results were observed for the corresponding homozygous diploids (data not shown). As judged by immunoblotting of cell lysates (Fig 1A), after shift to 37° for 3 hr, the product of the yrb1-51 allele was hardly detectable, whereas the yrb1-52 product remained relatively stable even 6 hr after temperature shift. Thus, the yrb1-51 mutation appears to destabilize the gene product at higher temperature, whereas the yrb1-52 product is stable under the same conditions. Upon prolonged incubation at 37°, an apparent degradation product of Yrb1 accumulated in yrb1-52 cells, but was also observed in the wild-type control cells. Yrb1 was expressed at similar levels in MATa, MAT, and MATa/MAT cells (data not shown) and its level in MATa cells was not elevated in response to treatment with -factor mating pheromone (data not shown).

View larger version (61K): In this window In a new window Download PPT slide   Figure 1. Effects of yrb1-51 and yrb1-52 mutations on in vivo stability of Yrb1 and cell morphology. (A) Stability of normal and mutant Yrb1 at restrictive temperature. Haploid strain YPH499 (YRB1) and its congenic derivatives, JTY2026 (yrb1-51) and JTY2483 (yrb1-52), were grown at 23° in SCGlc medium to midexponential phase, shifted to 37°, and samples were withdrawn at the indicated times. Total protein was extracted from each sample and analyzed by SDS-PAGE and immunoblotting using a rabbit polyclonal anti-Yrb1 antiserum (no. 1390). A band that cross-reacts nonspecifically with the anti-Yrb1 antiserum served as a loading control. The asterisk indicates a major degradation product of Yrb1. (B) Morphology of homozygous diploid yrb1ts cells. Strains YPH501 (YRB1/YRB1), JTY2482 (yrb1-51/yrb1-51), and JTY2485 (yrb1-52/yrb1-52) were cultivated on YPGlc plates at 16° or 30°, as indicated, and viewed by Nomarski optics.

Examination of the cell morphology revealed that haploid yrb1-51 cells arrested mostly as enlarged cells with a large bud or as large unbudded cells upon shift to 37° (data not shown; BAUMER et al. 2000 ), which is reminiscent of cell cycle progression mutants. Similar results were previously observed for stp52/yrb1-52 cells (KATZ et al. 1987 ; OUSPENSKI 1998 ). Another striking phenotype of both yrb1^ts alleles was the appearance of chains of elongated nonseparated cells, most evident in homozygous diploids grown on plates at a semipermissive temperature (30°; Fig 1B). Such cell elongation is diagnostic of mutations that delay G2-M progression (LEW 2000 ). We observed a similar morphological defect in srp1-31^ts/srp1-31^ts diploids (data not shown). Srp1/-importin is the adaptor necessary for recognition and nuclear import of proteins that contain a classical nuclear localization signal (NLS) by the Kap95/ß-importin receptor (ENENKEL et al. 1995 ). At 37°, srp1-31 cells are impaired in import of NLS-containing reporter proteins and arrest uniformly as large-budded cells indicative of a defect in mitosis (LOEB et al. 1995 ). Correspondingly, degradation of Clb2, whose destruction is required for exit from mitosis, is impaired in srp1-31 cells (LOEB et al. 1995 ); likewise, degradation of Clb2 and of two anaphase inhibitors, Pds1 and Sic1, is also impaired in yrb1-51 cells (BAUMER et al. 2000 ). Despite these similarities, the absence of a uniform cell cycle arrest phenotype distinguishes the yrb1-51 and yrb1-52 mutations from the srp1-31 mutation and from canonical cell division cycle (cdc) mutations, and might indicate a role of Yrb1 at multiple stages of the cell cycle. Consistent with such a notion, the yrb1-51 mutation interferes with both the G1/S transition and the passage through mitosis (BAUMER et al. 2000 ), and displayed synthetic growth defects when combined with two different cdc28 alleles (cdc28-4 and cdc28-1N) that are diagnostic for different cell cycle stages (G1 and G2/M, respectively; Table 4).

Yrb1-51 and Yrb1-52 are altered in conserved residues of the Ran-binding domain and defective for Ran-binding in vivo:
To determine the nature of the alterations in the mutant proteins, PCR was used to recover the YRB1 coding sequences from the mutant strains (see MATERIALS AND METHODS). The DNA sequence of each mutant ORF contained a single point mutation, both of which alter a highly conserved residue in the RBD (Fig 2A). The yrb1-51 mutation is a C-to-A transversion on the coding strand at position 272 (where +1 is the first base of the initiator ATG), which substitutes Asp for Ala at codon 91 (A91D). The yrb1-52 allele is a G-to-A transition on the coding strand at position 380, which substitutes Lys for Arg at codon 127 (R127K). On the basis of homology modeling of Yrb1 on the crystal structure of the first RBD (RanBD1) in mammalian Nup358 (RanBP2) complexed with Ran bound to a nonhydrolyzable GTP analog (VETTER et al. 1999 ), A91D replaces a nonpolar residue in the hydrophobic core of Yrb1 with a bulkier, charged residue (Fig 2B). This change should destabilize the global fold of Yrb1, consistent with the rapid degradation of this mutant protein observed at restrictive temperature (Fig 1A). Two other existent yrb1 alleles, yrb1-1 and yrb1-2 (SCHLENSTEDT et al. 1995 ), alter residues (F187 and L93, respectively) that project into the same hydrophobic pocket as A91. In contrast, R127K makes a seemingly modest change in a surface-exposed residue that forms a bridge to residues in the long C-terminal "arm" of Ran that embraces the RBD (Fig 2B), an alteration unlikely to disrupt the overall structure, consistent with the observed stability of the mutant protein at restrictive temperature (Fig 1A). We confirmed that each mutation was both necessary and sufficient to confer the ts phenotype of the corresponding allele by inserting each mutant DNA into a plasmid and introducing it into a yrb1 background (see MATERIALS AND METHODS).

View larger version (153K): In this window In a new window Download PPT slide   Figure 2. Positions of the altered residues in Yrb1 resulting from the yrb1-51 and yrb1-52 mutations. (A) Alignment of various Ran-binding domains (RBDs). Sequences shown are grouped into three subfamilies (RanBP1, RanBP2, and RanBP3), on the basis of certain shared sequence characteristics, and include the following (with GenBank accession numbers): S. cerevisiae Yrb1 (L38489), Schizosaccharomyces pombe Sbp1 (D86381), mouse RanBP1 (X56045), human RanBP1 (X83617), Xenopus laevis RanBP1 (U09128), Arabidopsis thaliana RanBP1 (U62742), mouse RanBP2 nucleoporin (X87337), human NUP358 (D38076), Bos taurus RanBP2 nucleoporin (L41691), Caenorhabditis elegans Ranup96 (Z34801), S. cerevisiae Nup2 (X69964), S. pombe Hba1 (U38783), and human RanBP3 (Y08697). Sequence of S. cerevisiae Yrb2/Nup36 is from the Swiss Protein Database (accession no. P40517). The mouse and bovine RanBP2 are incomplete because they are derived from partial cDNA clones. An insert of 24 residues (possibly an intron) was omitted from the actual C. elegans Ranup96 sequence to optimize its alignment to the other RBDs. Identities shared by 11 (or more) of the RBDs shown are indicated by white-on-black letters; chemically similar residues are shown as black-on-grey letters. The positions mutated in yrb1-51, A91A, and in yrb1-52, R127K, are indicated at the top. (B) Positions of the residues (A91 and R127) altered in the yrb1-51 and yrb1-52 mutants, respectively, have been modeled on the first RBD (RanBD1) in human NUP358 complexed with Ran bound to a nonhydrolyzable GTP analog (VETTER et al. 1999 ). Blue, Yrb1; purple, Ran; and red, GTP analog.

Two independent approaches demonstrated that the yrb1-51 and yrb1-52 mutations interfere with Yrb1-Ran (Gsp1) interaction in vivo. The GTPase-deficient form of Gsp1, Gsp1(G21V), binds more strongly to Yrb1 than normal Gsp1 (SCHLENSTEDT et al. 1995 ); hence, we used Gsp1(G21V) in our analyses. First, we applied the two-hybrid method using full-length wild-type Yrb1, Yrb1(A91D), or Yrb1(R127K) fused to the Gal4 transcriptional activation domain [Gal4(TAD)] and full-length Gsp1(G21V) fused to the LexA DNA-binding domain [LexA(DBD)] in a reporter strain carrying a chromosomally inserted copy of E. coli lacZ under control of eight LexA-operator sites. In this system, Yrb1(A91D) showed a reproducible reduction (>3-fold) and Yrb1(R127K) showed a dramatic reduction (>50-fold) in interaction with Gsp1(G21V) compared to wild-type Yrb1 (Fig 3A). Immunoblotting with anti-Yrb1 antiserum (see MATERIALS AND METHODS) and anti-Gsp1 antibodies (gift of P. Belhumeur) showed that all constructs were expressed at equivalent levels (data not shown).

View larger version (102K): In this window In a new window Download PPT slide   Figure 3. Yrb1(A91D) (yrb1-51) and Yrb1(R127K) (yrb1-52) bind Gsp1(G21V) with reduced affinity in vivo. (A) Interaction between wild-type Yrb1, Yrb1(A91D), or Yrb1(R127K) and Gsp1(G21V) was determined using the two-hybrid method as described in MATERIALS AND METHODS. In brief, Yrb1 proteins were fused to the Gal4(TAD), and Gsp1(G21V) was fused to the LexA(DBD). In the recipient strain (JTY2500), the E. coli lacZ gene is under control of eight LexA-operator elements. ß-Galactosidase activity is expressed in arbitrary units. Each value represents the average of single measurements made on three independent transformants; error bars indicate standard deviation of the mean. (B) Binding of Yrb1 to Gsp(G12V) was assessed by copurification. Cultures of strain CRY1 (YRB1 GSP1) carrying CEN plasmids expressing a ProtA-TEV-Gsp1(G21V) fusion from the NOP1 promoter and either wild-type Yrb1-GFP, Yrb1(A91D)-GFP, or Yrb1(R127K)-GFP expressed from the authentic YRB1 promoter were grown in selective SCGlc medium at 26°. Extracts were prepared and the ProtA-TEV-Gsp1(G21V) was purified on IgG-Sepharose (Pharmacia, Uppsala, Sweden) and eluted by digestion with recombinant TEV protease (GIBCO-BRL, Gaithersburg, MD). Equal fractions of the load and the eluate of each column were resolved by SDS-PAGE and analyzed, as indicated, by Coomassie blue staining and immunoblotting using a rabbit polyclonal anti-Yrb1 antiserum (no. 1391). Endogenous Yrb1 served as a control to confirm equivalent loading and functionality of the immobilized Gsp1(G21V).

These results were confirmed by a biochemical procedure (Fig 3B). Yrb1-GFP, Yrb1(A91D)-GFP, or Yrb1(R127K)-GFP, produced from the authentic YRB1 promoter on CEN plasmids, were expressed in wild-type cells (strain CRY1) also producing a ProtA-(TEV site)-Gsp1(G21V) from the constitutive NOP1 promoter on a CEN plasmid. After growth at 26°, protein complexes bound to bead-immobilized ProtA-(TEV site)-Gsp1(G21V)were recovered from cell extracts as described in HELLMUTH et al. 1998 , eluted by cleavage with TEV-protease (see MATERIALS AND METHODS), and analyzed by SDS-PAGE, Coomassie blue staining, and immunoblotting using a polyclonal anti-Yrb1 antisera (no. 1391). Examination of the input material (Fig 3B, load), and flow-through fractions (data not shown), demonstrated that expression of normal and mutant Yrb1-GFP fusions was comparable, as were their stabilities during purification. No detectable Yrb1(R127K)-GFP copurified with Gsp1(G21V), whereas significant amounts of both wild-type Yrb1-GFP and endogenous Yrb1 were retained by the same beads (Fig 3B). A detectable amount of Yrb1(A91D)-GFP copurified with ProtA-Gsp1(G21V), but its level was markedly less than the amount of wild-type Yrb1-GFP retained under the same conditions (Fig 3B).

yrb1-51 and yrb1-52 mutants are defective in nuclear protein and RNA transport:
To determine if yrb1-51 and yrb1-52 cause defects in nuclear protein import and RNA export, as observed before for the yrb1-1 and yrb1-2 alleles (SCHLENSTEDT et al. 1995 ), we first examined the distribution of poly(A)⁺-RNA by in situ hybridization. Mutant or wild-type control cells were grown to midexponential phase at permissive temperature (23°), and the cultures were then split into two equal portions, one of which was maintained at 23° and the other shifted to 37°. Samples were withdrawn at various times for analysis. Results were more readily visualized in diploid cells because of their larger size; however, similar findings were made with haploid cells (data not shown). In homozygous yrb1-51/yrb1-51 diploids shifted to 37° for 2 hr, there was a rapid and clear-cut nuclear accumulation of poly(A)⁺-RNA in every cell (Fig 4A); this effect was readily apparent in 50% of the cells even 1 hr after the shift (data not shown). Poly(A)⁺ RNA accumulated in the nucleus in a distinctly punctate pattern, a feature seen in mutants that have a strong RNA export defect, such as mex67-5 (SEGREF et al. 1997 ). However, onset of the RNA export defect in the yrb1-51/yrb1-51 cells was slower than that in mex67-5 mutants and nuclear RNA accumulation was not as complete (some cytosolic poly(A)⁺ RNA signal remains even 2 hr after shift to 37°). In striking contrast, yrb1-52/yrb1-52 diploids did not show any accumulation of poly(A)⁺ RNA (even 5 hr after shift to 37°), just like the wild-type control (Fig 4A). Although yrb1-51 cells manifested a clear defect in RNA export, neither yrb1-51 mutants nor yrb1-52 mutants had any detectable effect on the nuclear export of proteins containing a leucine-rich NES, such as SV40_NLS-PKI_NES-GFP (STADE et al. 1997 ) or yAP1-GFP (YAN et al. 1998 ; data not shown).

View larger version (352K): In this window In a new window Download PPT slide   Figure 4. Nuclear transport defects in yrb1ts mutants. (A) To examine poly(A)+RNA export, cultures of homozygous diploid strains YPH501 (YRB1/YRB1), JTY2482 (yrb1-51/yrb1-51), and JTY2485 (yrb1-52/yrb1-52) were grown in YPGlc at 23° to early exponential phase, split into two equal portions, and incubated for another 2 hr either at 23° or 37°. After fixation with formaldehyde, cells were stained with the DNA dye 4',6-diamidino-2-phenylindole, analyzed by in situ hybridization using a CY3-labeled oligo(dT) probe to visualize the subcellular distribution of poly(A)+RNA, and viewed by fluorescence microscopy using appropriate band-pass filters. (B and C) To examine nuclear protein import, the same strains as in A were transformed with multicopy plasmids expressing either SV40NLS-Gal4TAD-GFP (B) or L25NLS-GFP (C), respectively, cultivated and shifted as in A, and viewed directly by fluorescence microscopy and Nomarski optics.

To monitor the effect of the yrb1-51 and yrb1-52 mutations on nuclear protein import, four different GFP fusions of nuclear proteins were examined. To assess the -importin/Srp1 and ß-importin/Kap95/Rsl1-dependent pathway, two chimeras containing the SV40 NLS were used: a galactose-inducible SV40_NLS-GFP-ß-galactosidase, which is so large it cannot diffuse out of the nucleus after it has been delivered there (LEE et al. 1996 ), and a constitutively expressed SV40_NLS-Gal4_TAD-GFP, which, due to its small size, can diffuse out of the nucleus unless ongoing import occurs continuously (KUNZLER and HURT 1998 ). The third reporter was a fusion of GFP to Npl3, an mRNA-binding protein, whose nuclear entry depends on the importin, Kap111/Mtr10 (SENGER et al. 1998 ). GFP-Npl3 accumulates rapidly in the cytoplasm if import is impaired because Npl3 continuously shuttles between the nucleus and the cytosol. The fourth transport substrate was constitutively expressed and composed of GFP fused to the NLS of ribosomal protein L25 (O. GADAL and E. HURT, personal communication). Transport of L25_NLS-GFP into the nucleus utilizes two different import receptors, Kap121/Pse1 and Kap123/Yrb4 (SCHLENSTEDT et al. 1997 ). Like SV40_NLS-Gal4_TAD-GFP, L25_NLS-GFP is small enough to diffuse out of the nucleus unless its import occurs continuously.

Control strains or yrb1-51 and yrb1-52 mutants carrying the reporter plasmids described above were cultivated in selective medium, shifted to restrictive temperature, and examined by fluorescence microscopy. For the inducible reporter, transformants were grown to midexponential phase in Raf-containing medium at permissive temperature and induced for 1 hr by addition of 2% Gal (to allow for mRNA synthesis and export) before shift to 37°. For all four reporter proteins, there was a significant cytoplasmic accumulation in yrb1-51 and yrb1-52 mutants after shift to 37°, compared to wild-type cells, indicating a general defect in nuclear protein import. Results for SV40_NLS-Gal4_TAD-GFP (Fig 4B) and L25_NLS-GFP (Fig 4C) reporters in homozygous diploid strains are shown; but, similar results were obtained in haploids and for the other two reporters (data not shown). In yrb1-52 cells, the defect was noticeable even at permissive temperature. Our results showing a defect in nuclear protein import in yrb1-52 cells are at odds with a previously published report on the same mutant (OUSPENSKI 1998 ).

To exclude the possibility that cytoplasmic localization of the reporter proteins in yrb-51 and yrb1-52 cells was due to "leakiness" of the mutant nuclei, accumulation of constitutively expressed SV40_NLS-GFP-ß-galactosidase (encoded by pPS815; LEE et al. 1996 ) was examined in the same strains before and after temperature shift. In contrast to the short-term assay with the inducible version of the same reporter protein (see above), no increased cytoplasmic GFP signal was observed (data not shown), demonstrating that the nuclei in the yrb1-51 and yrb1-52 mutants were not more "leaky" than wild-type nuclei.

Genetic interactions of yrb1 mutations with nucleocytoplasmic transport factors:
As an independent means to confirm that yrb1-51 and yrb1-52 compromise nucleocytoplasmic trafficking even at permissive temperature, genetic interactions of these alleles with mutations in genes encoding a variety of other factors involved in nucleocytoplasmic transport were examined. Strains carrying mutations of interest were crossed with strains carrying the yrb1-51 or yrb1-52 mutation, and the resulting diploids were sporulated. Double mutant segregants from tetratype asci were compared to each single mutant segregant and to the wild-type segregant for their ability to grow at various temperatures (see Fig 5). Mutations tested included alterations in genes encoding components of the Ran GTPase cycle, nuclear import and export receptors, and nucleoporins (see Table 4). The plc1::HIS3 mutation (FLICK and THORNER 1993 ) was tested since there is evidence for a role of PLC1-encoded phosphatidylinositol-specific phospholipase C in mRNA export (YORK et al. 1999 ; J. FLICK, personal communication). The yrb1-51 mutation displayed deleterious genetic interactions with many of these other classes of mutants that affect nucleocytoplasmic transport (Table 4). Combination of the yrb1-52 mutation with at least 12 of these mutations revealed essentially the same growth defects. For xpo1-1 yrb1-52 and yrb2::HIS3 yrb1-52, the growth defect was even more severe than for the corresponding double mutant with yrb1-51 (Table 4). Equally pronounced growth defects were observed when yrb1-51 was combined with mutations in genes encoding certain nuclear transport receptors (Table 4). In all these cases, the double mutant was not viable at any temperature (synthetic lethality); for the other genetic interactions observed, double mutants were viable at the permissive temperature (23°) but revealed a restrictive temperature that was considerably (>3°; ++) or slightly (3°; +) lower than the one of any of the two single mutants (synthetic growth defect). Thus, the functions of Yrb1(A91D) and Yrb1(R127K) are at least partially defective even under permissive conditions.

View larger version (73K): In this window In a new window Download PPT slide   Figure 5. Genetic interactions between yrb1ts alleles and components of the nucleocytoplasmic transport machinery. yrb1-51 (or yrb1-52) mutant strains were crossed with strains carrying a mutation in another gene of interest. The resulting diploids were subjected to sporulation, and growth of individual spores from tetratype asci was examined at various temperatures. Left, genetic interaction of yrb1-51 with nup2::HIS3 is manifested by the extremely poor growth ("+" in Table 4) of the yrb1-51 nup2::HIS3 double mutant at 28°, a temperature clearly permissive for the congenic yrb1-51 and nup2::HIS3 single mutants (yrb1-51 alone has a restrictive temperature of 31° under these conditions and nup2::HIS3 alone has no obvious growth defect even at higher temperatures). Right, a yrb1-51 mutant carrying a URA3-marked multicopy plasmid expressing wild-type YRB1 (pMK103) was crossed with a strain carrying a ts mutation, srp1-31, in -importin. The resulting diploid was sporulated and individual spores from a tetratype ascus were streaked on medium containing 5-FOA to counterselect against the plasmid. Genetic interaction of yrb1-51 with srp1-31 is manifested by the inviability or "synthetic lethality" ("sl" in Table 4) of the yrb1-51 srp1-31 double mutant at any temperature, whereas the congenic yrb1-51 and srp1-31 single mutants are able to grow at permissive temperatures (here shown at 26°). Other double mutant combinations tested and their phenotype are listed in Table 4.

Nucleocytoplasmic trafficking of Ste5 is altered in yrb1-51 and yrb1-52 cells:
Our results indicate that impaired nucleocytoplasmic transport is the primary cause of the phenotypes of yrb1-51 and yrb1-52 mutants, including, presumably, the ability of these mutations to suppress mating defects. Far1 and Ste5 are currently the only components of the mating pheromone response pathway known to shuttle between nucleus and cytoplasm (BLONDEL et al. 1999 ; MAHANTY et al. 1999 ) and yrb1-52 was isolated as an extragenic suppressor of a ste5 missense mutation (KATZ et al. 1987 ). Hence, we constructed a functional GFP-Ste5 chimera and examined its subcellular localization in yrb1-51 and yrb1-52 mutants (and in control cells) in the absence of pheromone to avoid the complications of signal-induced changes. In wild-type cells at steady state, GFP-Ste5 accumulated in the nucleus, even though cytoplasmic staining was also evident (Fig 6), as observed before (MAHANTY et al. 1999 ; PRYCIAK and HUNTRESS 1998 ). In contrast, no nuclear accumulation was observed in either of the two yrb1^ts mutants, even at permissive temperature (Fig 6). Since Ste5 shuttles continuously between nucleus and cytoplasm (MAHANTY et al. 1999 ), the observed mislocalization in cells with defective Yrb1 could be due, in principle, either to inhibition of Ste5 import into the nucleus or enhancement of Ste5 export from the nucleus. Since we have demonstrated that yrb1-51 and yrb1-52 clearly impede nuclear import of various reporter proteins, the former possibility seems more likely.

View larger version (97K): In this window In a new window Download PPT slide   Figure 6. Localization of GFP-Ste5 in yrb1ts mutants. Haploid strains YPH499 (YRB1), JTY2026 (yrb1-51), and JTY2483 (yrb1-52) expressing a GFP-Ste5 fusion protein from a multicopy plasmid were cultivated, shifted, and viewed as described in the legend to Fig 4B and Fig C.

To obtain further evidence that suppression of mating defects by yrb1-51 and yrb1-52 arises from impairment of nuclear import of Ste5, we tested whether these yrb1^ts mutations could suppress the mating defect of a msn5 mutant. Msn5 is thought to be the nuclear receptor required for pheromone-stimulated export of Ste5 from the nucleus (MAHANTY et al. 1999 ). Each yrb1^ts mutant was crossed against a msn5 strain and the wild-type single mutant, and double mutant spores from the resulting tetratype asci were tested for their relative mating proficiency using a quantitative mating assay performed at semipermissive temperature (30°) for the yrb1^ts alleles. Both the yrb1-51 and yrb1-52 mutations restored the mating efficiency of the msn5 mutant to essentially the wild-type level (Table 5). This suppression is consistent with the idea that a higher cytoplasmic pool of Ste5 (due to its inefficient nuclear import in the yrb1^ts mutants) compensates for its inefficient pheromone-stimulated export from the nucleus (due to the msn5 mutation).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Ran GTPase is one of the most highly conserved proteins in nucleated cells (MACARA et al. 2000 ; SAZER and DASSO 2000 ). Ran action has been implicated in nucleocytoplasmic transport, microtubule assembly, nuclear envelope formation, maintenance of chromatin structure and nuclear (and nucleolar) organization, chromosome segregation, DNA replication, RNA metabolism, and cell cycle progression. It is still a matter of some debate whether the pleiotropic phenotypes of alterations in the Ran GTPase cycle are due solely to the established role of Ran in nucleocytoplasmic transport (GORLICH and KUTAY 1999 ) or whether Ran has additional roles in the cell. Recent studies using in vitro systems have implicated Ran function directly in microtubule organization (KAHANA and CLEVELAND 1999 ) and nuclear envelope formation (HETZER et al. 2000 ; ZHANG and CLARKE 2000 ). Our evidence indicates that the Ran GTPase cycle is linked to the signaling events required for mating as a consequence of the role of Ran in nucleocytoplasmic transport.

A decade ago, before the function of the essential Ran regulator, RanGEF1, was fully appreciated, a ts mutation (srm1-1) in its yeast homolog (SRM1/PRP20/MTR1) was isolated as a suppressor of the mating defect of haploid cells lacking pheromone receptors (CLARK and SPRAGUE 1989 ). Unlike other ts mutations in yeast RanGEF1 identified subsequently (AMBERG et al. 1993 ), srm1-1 does not cause dramatic nuclear accumulation of poly(A)⁺ RNA at restrictive temperature (KADOWAKI et al. 1993 ; and its effect on nuclear protein import was not examined), which left open the possibility that the Ran GTPase cycle had some role in mating distinct from its function in nucleocytoplasmic transport. We have shown here that two independently isolated suppressors of early and late defects in the mating pathway are alterations in another essential Ran regulator, RanBP1, that clearly impair nuclear protein import, even at permissive temperature. We found further that the Ste5 scaffold protein, which shuttles between nucleus and cytoplasm and is required for pheromone response, is mislocalized to the cytosol in these mutants. Consistent with this finding, both yrb1^ts alleles suppressed the partial mating defect of a null mutation in MSN5, which encodes the nuclear receptor required for pheromone-stimulated export of Ste5.

In retrospect, it may seem obvious that perturbation of nucleocytoplasmic transport could affect yeast mating since response to pheromone requires: (a) that some signal carrier enter the nucleus to induce the transcription of genes, (b) that RNPs containing newly synthesized mRNAs for mating-specific components exit the nucleus, and (c) that the translation products of some of those transcripts be recruited back into the nucleus in support of late mating events, like karyogamy (ROSE 1996 ). Accordingly, it is now known that many components of the pheromone response machinery, both preexisting and pheromone induced, are localized to the nucleus. On the other hand, it has been shown that two components, Far1 (BLONDEL et al. 1999 ) and Ste5 (PRYCIAK and HUNTRESS 1998 ; MAHANTY et al. 1999 ), rapidly relocalize from the nucleus to the site of cell fusion (shmoo tip) upon pheromone treatment. Our results suggest the following mechanism to explain the suppression of mating defects by mutations in nucleocytoplasmic transport factors.

Because both yrb1^ts alleles described here are defective in nuclear protein import, but only one seems impaired in poly(A)⁺ RNA export, yet both act as suppressors of mating defects, it is presumably the import defect that leads to suppression. Also, it should be recalled that yrb1-51 was isolated on the basis of rescue of the mating debility of a fus1 mutant, which has an intact pheromone response pathway and is only partially mating defective. Likewise, yrb1-52 was isolated on the basis of its ability to rescue ste5 missense mutations (and can do so for ste4 and ste7 missense mutations), but is unable to rescue null alleles in these same genes (KATZ et al. 1987 ). These considerations indicate that residual signaling in the pheromone response pathway must be present for suppression to occur. How could impairment of nuclear protein import improve the mating efficiency of such partially mating-defective mutants? One reasonable scenario is that nuclear import of certain signaling components results in downregulation of the mating response. If so, mutations that reduce the rate of nuclear entry of such a factor(s) should enhance the efficiency of mutants that are only partially defective in mating. Both Far1 and Ste5 are reasonable candidates for such factors. For Far1, its pheromone-induced relocalization from the nucleus to the shmoo tip seems to be required for cytoskeletal reorganization, which is necessary for polarized cell growth toward the mating partner (BLONDEL et al. 1999 ; SHIMADA et al. 2000 ). The role of recruitment of Ste5 to the shmoo tip may be similar, given the interaction of Ste5 with Bem1, a protein involved in polarized cell growth (LEBERER et al. 1997 ). Since relocalization of Far1 (BLONDEL et al. 1999 ; SHIMADA et al. 2000 ) and Ste5 (PRYCIAK and HUNTRESS 1998 ; MAHANTY et al. 1999 ) is transient, a nuclear import defect that kept these proteins out of the nucleus longer might increase mating efficiency in cells in which the cell fusion process would otherwise be nonoptimal.

In agreement with the above model, both the yrb1-51 and the yrb1-52 mutations increased the cytoplasmic concentration of Ste5 and were able to suppress the mating defect of a msn5 null mutation. The MSN5 gene encodes a nuclear export receptor of the ß-importin family (KAFFMAN et al. 1998 ; DEVIT and JOHNSTON 1999 ) and was also identified as STE21 because a null mutation at this locus confers a partial mating defect (AKADA et al. 1996 ; ALEPUZ et al. 1999 ; BLONDEL et al. 1999 ). Impaired nuclear protein import in the yrb1^ts mutants keeps Ste5 (and perhaps other proteins required for mating) in the cytosol longer, and thus there is less need for their continuous Msn5-dependent reexport from the nucleus. Also consistent with the above model, MATa yrb1-51 and MATa yrb1-52 mutants are able to respond normally to -factor under conditions (by inducing the pheromone-responsive reporter FUS1-lacZ) and by undergoing G1 arrest (as judged by the standard halo bioassay; data not shown); hence, these mutants are not defective in early signaling events. If, however, proteins required for late events in mating are maintained longer in the cytosol of yrb1 mutants due to defective nuclear protein import, the efficiency of mating would be enhanced, as observed. Finally, such a model would also explain the spectrum of mating defects that are suppressed by yrb1-51, yrb1-52, and srm1-1. All of the mutations suppressed are in gene products that are targeted, directly or indirectly, to the shmoo tip after pheromone treatment and most