INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The ability of Saccharomyces cerevisiae cells to rapidly respond and adapt to changing environmental conditions is essential for viability. A prerequisite for the generation of a proper physiological response is the ability to sense and subsequently transduce information regarding the extra- and intracellular environments. Sensor-initiated signals are used to make dynamic adjustments in patterns of gene expression and protein turnover, processes that enable cells to express the necessary components appropriate for prevailing conditions. For example, in response to nutrient availability, yeasts regulate the expression and activity of proteins involved in nutrient uptake and utilization. Recently, several plasma membrane-localized nutritional sensors that monitor nutrient availability in the extracellular environment have been identified in yeast. These include two glucose sensors, SNF3 and RGT2 (41, 49); a G-protein-coupled receptor that is activated by the presence of fermentable sugars, GPR1 (35, 44, 67); a high-affinity ammonium transporter (MEP2) (45) that may function as an ammonium sensor (43); and an amino acid sensor, SSY1 (19, 30, 33, 34). Little is known regarding whether these primary sensors function alone or in complexes together with other proteins, and the mechanisms by which these sensors transduce nutritionally derived signals remain to be elucidated.

SNF3 and RGT2 encode unique members of the hexose transporter (HXT) family that possess unusually long C-terminal domains. These proteins are poorly expressed compared to the other known functional hexose transporters and pleiotropically affect the expression of multiple HXTs (41, 49). The expression of SNF3 is repressed by the presence of glucose (10), whereas RGT2 is constitutively expressed (49). Recent reports have shown that the cytoplasmically oriented C-terminal extensions of Snf3p and Rgt2p have important roles in signaling. The C terminus of Snf3p expressed as a soluble protein (62), or as a hybrid with the Hxt2p hexose transporter (48), can partly suppress phenotypes associated with an snf3 deletion. The C-terminal extensions of Snf3p and Rgt2p physically interact with two homologous proteins, Std1p and Mth1p (37, 56). Std1p and Mth1p are suggested to act as repressors of Snf3p and Rgt2p target genes in the absence of signaling. Green fluorescent protein-Std1p fusion proteins are localized to the cell periphery and the cell nucleus; thus the possibility exists that Std1p mediates information directly from the cell surface to the nucleus (56).

SSY1 encodes a unique member of the amino acid permease protein family that functions as a sensor of extracellular amino acids (19, 30, 33, 34). Ssy1p has a unique 200-amino-acid N-terminal extension that is required for SSY1 activity (34); however, its precise function remains obscure. Ssy1p is localized to the plasma membrane and is, as are the other members of the amino acid permease family, dependent upon the amino acid-specific packaging chaperone Shr3p (25, 42) to exit the endoplasmic reticulum (34). SSY1 is required for the transcriptional induction of multiple genes encoding amino acid permeases (AGP1, BAP2, BAP3, GNP1, VAP2, and TAT2), the peptide transporter (PTR2), and arginase (CAR1) in response to extracellular amino acids (19, 30, 34). SSY1-mediated signals are also required for full transcriptional repression of the general amino acid permease (GAP1) on ammonium-based media in the presence of amino acids (34). Ssy1p is dependent upon PTR3 to mediate amino acid-derived signals (34). Ptr3p is a peripherally associated plasma membrane protein and contains a domain that shares homology with amino acid permeases and Gcn4p (34).

Several factors required for transcription of SSY1-controlled genes have been identified, including STP1, STP2, and ABF1 (15, 16) and UGA35 (also known as DAL81) and GRR1 (30). STP1 and STP2 were originally identified as genes required for pre-tRNA maturation (66). Abf1p is a general transcription factor involved in global gene activation and silencing (17, 54, 57). UGA35/DAL81 encodes a nonspecific factor required for the full induction of several genes active in nitrogen utilization, including -aminobutyric acid and allophanate-inducible genes (8, 14, 64). Some of the SSY1-responsive genes, including BAP2, also carry promoter binding sites for Gcn4p and Leu3p (18). GCN4 encodes a transcription factor that is translationally regulated by the general amino acid control pathway that monitors intracellular levels of free amino acids. The general control pathway coordinately upregulates biosynthetic genes and transporters in response to amino acid deprivation (28). Leu3p, a transcription factor that is capable of acting as a repressor or an inducer, participates in regulating the transcription of several genes within the branched-chain amino acid biosynthetic pathways (9, 65). GRR1 was previously reported to be involved in glucose signaling and cell cycle control (5, 40). Grr1p is an F-box-containing component of discrete Skp1-Cullin-F-box (SCF) ubiquitin ligase complexes that mark proteins for degradation via the proteasome. Several F-box-containing proteins have been identified in yeast, and they are thought to provide specificity by recruiting distinct protein substrates to SCF complexes for ubiquitination (51). Presumably the inability of grr1 mutants to properly respond to nutritional signals is a consequence of aberrant patterns of regulated protein degradation (32, 50). The large number of factors that affect the transcription of target genes indicates a complex network of regulatory processes that most likely integrate signals derived from different primary nutritional sensors.

Although several of the downstream components required for the transmission of amino acid-induced signals have been identified, little is known regarding the primary component composition of the plasma membrane amino acid-sensing system. As previously indicated, this system not only depends upon Ssy1p, but also requires Ptr3p (34). Even less is known regarding the mechanisms associated with initial signaling events that occur within the sensing system in response to extracellular amino acids. We have focused our efforts on characterizing the proteins comprising the plasma membrane amino acid nutritional sensor. In this paper, we present results that identify Ssy5p as a third component of this signaling system. Mutations in SSY5 have earlier been described to interfere with amino acid uptake (33). We show that ssy5 mutants belong to the same phenotypic epistasis group as both ssy1 and ptr3 mutants. Ssy5p is a peripheral membrane protein that binds to the plasma membrane and is dependent upon SSY1 and PTR3 for wild-type expression. Additionally, Ssy1p and Ptr3p exhibit altered electrophoretic mobilities dependent on the presence of the other sensing components and on the availability of amino acids. Finally, we have documented the dynamic nature of amino acid-induced signaling and have found that each of the components of the Ssy1p1-Ptr3p-Ssy5p (SPS) signaling system is rapidly downregulated in response to amino acid availability.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains and media. The yeast strains used in this study are listed in Table 1. Strains carrying a null mutation of SSY5 were constructed as follows. PLY1 and PLY126 were transformed with a linear 9-kb XbaI fragment from pHK049 (see Table 2; plasmids are described in the following section) containing the ssy51::hisG-URA3-hisG deletion construct. Strain HKY75 was propagated on media containing 5-fluoroorotic acid (5-FOA) (26) to attain strain HKY77 carrying the unmarked ssy52 deletion. Strain HKY82 was obtained by transforming HKY77 with a linear SalI-SpeI fragment of pHK031 containing ssy112::hisG-URA3-hisG. Similarly, strain HKY83 was obtained by transforming HKY77 with a linear EcoRI fragment of pPL341 containing ptr314::hisG-URA3-hisG (34). Strains HKY82 and HKY83 were propagated on 5-FOA, resulting in strains HKY84 and HKY85 with unmarked ssy113 and ptr315 alleles, respectively. The correct integration of each gene replacement was confirmed by Southern analysis. The yeast strain cdc25H was obtained from Stratagene (La Jolla, Calif.).

Plasmids. The plasmids and oligonucleotides used in this study are listed in Table 2. In separate reactions, a 4.9-kb BamHI-BstEII fragment from pPL496 containing SSY5 was ligated to BamHI-digested pRS316 (59) and pRS202 (13), resulting in pHK036 (CEN, ori1), pHK037 (CEN, ori2), and pHK037 (2µm, ori2). A blunt-ended 5-kb BamHI-BglII fragment isolated from pSE1076 (1) containing a hisG URA3 kan^r hisG blaster cassette was inserted into BsrGI-digested pHK036 made blunt by using T4 DNA polymerase. This ssy5 construct (pHK049) removes nucleotides (nt) 28 to +1108 (nucleotide designations are in relation to the first nucleotide of the initiator ATG codon). Plasmid pHK031 was constructed by inserting a SalI-SpeI fragment from pHK030 carrying the ssy112::hisG URA3 kan^r hisG deletion allele (34) into SalI-SpeI-digested pBluescript II KS(+) (Stratagene, La Jolla, Calif.). To remove the BamHI sites within their multicloning sequences, plasmids pPL193 (34) and pHK036 were digested with BamHI, made blunt with T4 DNA polymerase, and religated, creating pHK040 and pHK041, respectively. The SSY5-c-myc epitope-tagged allele in pHK048 was constructed in two steps. First, a BamHI site (reading frame a) was introduced immediately after the ATG initiation codon of SSY5 by using site-directed mutagenesis (36, 63) with single-stranded pHK041 as a template and oligonucleotide POL99-026 as mutagenic primer, resulting in pHK047. In step 2, a BamHI-flanked cloning cassette, encoding the c-myc epitope reiterated three times (c-myc³) was inserted into the unique BamHI site of pHK047, creating pHK048. The construction of plasmids encoding SOS hybrid proteins required multiple steps. Site-directed mutagenesis with single-stranded pHK040 and pHK041 as a template and mutagenic primers POL00-005 and POL00-004 was used to create unique BamHI sites (reading frame b) immediately following the ATG of PTR3 and the ATG of SSY5, resulting in plasmids pHK042 and pHK043, respectively. The vectors contained in the CytoTrap kit were obtained from Stratagene. The BamHI-SalI fragments from plasmids pHK042 and pHK043 were inserted into BamHI-SalI-digested pSOS (Stratagene), creating pHK044 and pHK045. Plasmid pHK050 was constructed by moving the SacI-SalI fragment from pHK035 containing SSY5 into SacI-SalI-digested pAD40 (obtained from M. Wigler, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.). In a single reaction, two XbaI sites were introduced into the SSY1 coding sequence with single-stranded pHK003 as a template (34) and mutagenic primers POL95-039 and POL96-021. The resulting plasmid was digested with XbaI to release the internal fragment of SSY1 and religated, resulting in pHK038. pHK038 encodes only the first 206 N-terminal amino acids of Ssy1p. The 2µm plasmids pHK039 and pHK012 were constructed by inserting the SacII-KpnI fragments from pPL356 (34) or pHK038 into SacII-KpnI-digested pRS202, respectively. Plasmid pHK026 (2µm PTR3) was created by ligating a SalI-SacII fragment containing PTR3 obtained from pPL193 into SalI-SacII-digested pRS202.

Genetic analysis. Strain PLY1 was used to isolate spontaneous mutants resistant to 30 mM histidine (42). The super-high-histidine-resistant (shr) mutants were backcrossed to PLY4 (MAT his429 ura3-52 ade21::URA3), an isogenic derivative of PLY1. Tetrad analysis indicated that the mutant phenotypes segregated 2:2. Strain PLAS10-8C (shr4-10) was obtained as a meiotic segregant from one of these crosses. SHR4 was cloned by complementation of the 30 mM histidine-resistant phenotype; strain PLAS10-8C (shr4-10) was transformed with a plasmid library (pRS202 library, obtained from Philip Hieter [13]), and Ura⁺ transformants unable to grow on selective medium containing 30 mM histidine were identified. Complementing plasmids were isolated and further analyzed. Plasmid pHK035, which contains a 4.9-kb BstEII-BamHI insert with a single open reading frame (ORF) (YJL156c), fully complemented shr4 mutant alleles.

Northern analysis. Steady-state levels of AGP1 and PTR2 mRNA (see Fig. 2) were determined in strains PLY126 and HKY77 transformed with YCp405 and pRS316. Cells from overnight cultures grown in SD were harvested, washed once, and resuspended in 10× volume of fresh SD at an optical density at 600 nm (OD₆₀₀) of 0.2. Cultures were grown to an OD₆₀₀ of 0.8, one-half of the culture was induced by addition of leucine to a concentration of 0.15 mM, and an identical aliquot of water was added to the remaining half of the culture (uninduced control). After 45 min, 30 ml of cells was harvested, RNA was prepared, and Northern analysis was performed with aliquots (10 µg) of RNA as described previously (34). SSY5 mRNA levels were analyzed in strains HKY77, HKY84, and HKY85 transformed with pHK048 and YCp405 (Fig. 4B). Cells were grown overnight in SD or SD supplemented with 1.3 mM leucine, washed once, and resuspended in a 10× volume of fresh medium at an OD₆₀₀ of 0.2. RNA was isolated when cultures reached a cell density of an OD₆₀₀ of 0.8. Radioactive probes were prepared with the following template DNA fragments: a 1.2-kb EcoRV internal fragment of SSY5 and a 339-bp PCR fragment from AGP1 amplified with oligonucleotide primer pairs POL00-006 and POL00-007. Additionally, the previously described 569-bp PCR fragment of CAR1, 530-bp PCR-fragment from PTR2, and a 1.65-kb BamHI-HindIII fragment containing ACT1 were used (34). The DNA fragments were labeled with [-³²P]dCTP (3,000 Ci/mmol, Amersham, United Kingdom) by using a random-primed DNA labeling kit (MBI Fermentas Molecular Biology) and purified by using Bio-Rad Bio Spin columns. After hybridization, blots were rinsed once with 5× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% sodium dodecyl sulfate (SDS), washed two times with 5× SSC-0.1% SDS, one time with 1× SSC-0.1% SDS, and one time with 0.5× SSC-0.1% SDS if required. Washings were performed at 55°C for 20 min. After washings, blots were visualized and quantified with a Fujix Bio-Image Analyzer BAS1500 (Fuji Photo Film Co., Ltd., Tokyo, Japan).

Protein manipulations. Protein was determined by the method of Markwell et al. (46). Whole-cell protein was examined with lysates prepared from 100 ml of cells by the glass-bead method described by Chang and Slayman (11). The membrane association of Ssy5p in strain HKY77 transformed with plasmids pHK048 and YCp405 was examined as follows. An aliquot of a lysate (100 µg of protein) in low-salt BB buffer (0.3 M sorbitol, 5 mM MgCl₂, 5 mM Tris [pH 7.5]) was diluted 1:1 with either H₂O, 1.6 M urea, or 2 mM EDTA; mixed; and incubated on ice for 30 min. Samples were centrifuged at 100,000 × g for 45 min at 4°C, and protein pellets were resuspended in 2× sample buffer containing low-salt BB buffer. After sonication and denaturation and then incubation for 10 min at 37°C, aliquots (10 µg of protein) were resolved by SDS-polyacrylamide gel electrophoresis (PAGE) and analyzed by immunoblotting.

Amino acid pool size determination. Whole-cell and vacuolar amino acid pool concentrations were determined with cells grown in YPD to an OD₆₀₀ of 1 essentially as described by Ohsumi et al. (47). Appropriate quantities of cultures (3 × 10⁸ cells) were harvested by centrifugation, and cell pellets were washed twice with 1.5 ml of water and resuspended in 1.5 ml of AA buffer (0.6 M sorbitol, 2.5 mM potassium phosphate buffer [pH 6]) containing 10 mM glucose. For the determination of vacuolar amino acid pools, the cells were resuspended in the same buffer containing 0.8 mM CuCl₂ and incubated for 10 min at 30°C. One-milliliter aliquots of cell suspensions were filtered (Whatman GF/F filters), and filters were washed four times with AA buffer. The washed filters were boiled in 3 ml of water for 15 min, and 1-ml aliquots were centrifuged to remove particles of filter. The concentrations of amino acids in 30-µl aliquots were determined.

Time course experiments. Cells from overnight cultures of strains HKY31(pHK018, YCp405), HKY77(pHK048, YCp405), and HKY20(pHK010, YCp405) grown in SD were washed once and resuspended in a 10× volume of fresh SD to an OD₆₀₀ of 0.1 to 0.2. After cultures reached a cell density of OD₆₀₀ of 0.5, the cultures were split into two equal volumes. One-half of the cultures received an addition of L-leucine to a concentration of 1.3 mM (induced); the other half received an aliquot of water (uninduced control). Subsamples (130 ml) were withdrawn immediately prior to the addition of L-leucine (t = 0) and at 10, 30, 60, 120, and 180 min after L-leucine addition. Subsamples were rapidly chilled on ice, total cell protein was prepared from 100 ml of culture, and RNA was isolated from 30 ml of culture.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

ssy5 mutations result in histidine resistance and increased vacuolar pools of arginine and histidine. Yeast strains carrying mutations in SHR4 were isolated in a genetic selection for shr mutants resistant to 30 mM histidine (42). SHR4 was cloned by complementation of the recessive 30 mM histidine-resistant phenotype of strain PLAS10-8C (shr4-10) (Materials and Methods). Subsequent sequence analysis indicated that SHR4 is identical to ORF YJL156c, previously identified as SSY5 (33). In addition to exhibiting resistance to 30 mM histidine, ssy5 mutant strains have increased vacuolar pools of arginine and histidine (Table 3). Compared to the wild-type strain PLY1, the vacuolar levels of histidine and arginine in mutant strain PLAS10-8C (ssy5-410) were increased by two- and threefold, respectively. In contrast, the levels of lysine remained unchanged. The observed increases in vacuolar amino acid pools resulting from the ssy5 mutation are similar to those observed in strains carrying mutations in SSY1 and PTR3 (Table 3) (34).

ssy51::hisG-URA3 kan

ura3-52 his429

ura3-52 lys2201

ssy51

ura3-52 his429 ssy51

MAT ura3-52 his429 shr4-10

ssy5 null mutant strains exhibit levels of resistance to toxic amino acids and azetidine carboxylate similar to those of ssy1 and ptr3 mutants. The growth characteristics of isogenic wild-type, ssy113, ptr315, ssy52, ssy113 ptr315, ssy52 ssy113, and ssy52 ptr315 strains were examined. All strains grew well on SPD and SD (Fig. 1A and D, respectively). Only the growth of the wild-type strain was inhibited on medium containing toxic levels of histidine (Fig. 1B), methionine (Fig. 1C), or L-azetidine-2-carboxylate (Fig. 1E). L-Azetidine-2-carboxylate is a proline analogue (38) that pleiotropically inhibits multiple amino acid permeases, including the general amino acid permease (GAP1) (29). In contrast to the wild-type strain, the ssy52, ssy113, and ptr315 mutant strains grew equally well on each of the selective media and formed colonies of similar size. Thus, the ssy5 null mutation manifests identical levels of resistance to either ssy1 or ptr3 null mutations. Additionally, strains carrying the possible double mutant combinations ssy113 ptr315, ssy52 ssy113, and ssy52 ptr315 did not exhibit any additive affects.

View larger version (109K): [in this window] [in a new window]   FIG. 1.   Growth characteristics of strains carrying single and double mutant combinations of ssy5, ssy1, and ptr3 null alleles. Strains PLY126 (wild type [WT]), HKY20 (ssy113), HKY31 (ptr315), HKY77 (ssy52), HKY33 (ssy113 ptr315), HKY84 (ssy52 ssy113), and HKY85 (ssy52 ptr315) were streaked onto the following media: SPD (plus uracil and lysine) (A), SPD (plus uracil and lysine) containing 30 mM L-histidine (B), SPD (plus uracil and lysine) containing 30 mM L-methionine (C), SD (plus uracil and lysine) (D), and SD (plus uracil and lysine) containing 100 µg of L-azetidine-2-carboxylate ml1 (E). Plates were incubated for 3 days at room temperature and photographed.

SSY5 is required for amino acid-induced expression of permeases. Wild-type (PLY126) and ssy52 (HKY77) strains carrying plasmids pRS316 and YCp405, which complement the ura3-52 and lys2201 auxotrophies, respectively, were grown on SD medium without nutritional supplements to an OD₆₀₀ of 0.8. Total RNA was isolated 45 min after the addition of 0.15 mM leucine or an equal volume of water, and the transcript levels of the broad-range amino acid permease (AGP1) (30) and the peptide transporter (PTR2) (52) were determined by Northern analysis (Fig. 2). The levels of expression were quantitated by phosphorimaging, and ACT1 transcript levels were used to standardize quantitations. AGP1 transcripts were detected in wild-type cells (Fig. 2A, lanes 1 and 2), but not in ssy5 null mutant cells (Fig. 2A, lanes 3 and 4). The lack of detectable AGP1 transcripts in ssy5 mutant cells indicates that SSY5 is required to maintain the basal AGP1 expression observed in wild-type cells grown in the absence of exogenously added amino acids (Fig. 2, compare lanes 1 and 3). When leucine was added to wild-type cells, AGP1 mRNA levels increased by fivefold (Fig. 2A, compare lanes 1 and 2). Similarly, leucine-induced wild-type cells have approximately 20-fold more PTR2 transcripts than uninduced cells (Fig. 2B, compare lanes 1 and 2). In contrast, when leucine was added to ssy52 cells, the levels of AGP1 transcripts did not increase (Fig. 2A, lane 4), and PTR2 (Fig. 2B, lane 4) did not accumulate to wild-type levels. These results indicate that ssy5 mutations mimic the observed transcriptional defects exhibited by ssy1 and ptr3 mutations (4, 19, 30, 34).

View larger version (35K): [in this window] [in a new window]   FIG. 2.   Expression of AGP1 and PTR2 in wild-type (WT) and ssy5 null mutant strains. Strains PLY126 (wild type) and HKY77 (ssy52) were grown in SD to an OD600 of 0.8, and total RNA was isolated 45 min after the addition of water (lanes 1 and 3) or 0.15 mM leucine (lanes 2 and 4). Expression levels of AGP1 (A) and PTR2 (B) were determined by Northern analysis and quantitated by phosphorimaging. The levels of actin (ACT1) transcript were used to standardize quantitations (lower panels). The relative expression levels were normalized to the expression levels in the uninduced wild-type strain.

Ssy5p is a peripheral membrane protein that associates with the plasma membrane. SSY5 encodes a 76-kDa protein comprised of 687 amino acids that does not share significant sequence homology with other known proteins. A functional epitope-tagged version of Ssy5p was created by introducing the c-myc epitope at the extreme N terminus (see Materials and Methods). The SSY5-c-myc allele was judged to be functional based on its ability to complement the 30 mM histidine-resistant phenotype of ssy5 mutants. The level of Ssy5p-c-myc present in whole-cell extracts was examined by SDS-PAGE and immunoblotting. Ssy5p-c-myc migrated as a 67-kDa protein, significantly faster than its predicted molecular weight, and the signal strength of the immunoreactive Ssy5p-c-myc band was rather weak. The low levels of Ssy5p within extracts are consistent with the low codon index bias (0.017) of the SSY5 ORF.

View larger version (71K): [in this window] [in a new window]   FIG. 3.   SSY5 encodes a peripherally associated PM protein. (A) The membrane association of Ssy5p was examined by using whole-cell lysates prepared from strain HKY77 expressing SSY5-c-myc. Aliquots of total protein lysate (Tot) were diluted 1:1 with H2O, 1.6 M urea, or 2 mM EDTA; mixed; and incubated on ice for 30 min. Membrane pellet (P) and soluble (S) fractions, obtained after centrifugation at 100,000 × g for 45 min at 4°C, were resolved by SDS-PAGE and analyzed by immunoblotting. As a control, the membrane association of the PM ATPase (Pma1p) was monitored. (B) The ability of Ssy5p to associate with the PM was assessed by using the SOS membrane recruitment system. Strains cdc25H (No vector) and cdc25H transformed with plasmids pSOS, pSOS-PTR3 (pHK044), pSOS-SSY5 (pHK045), pAD-SSY5 (pHK050), pSOS-Col1, and pSOS-MAFB were grown on YPD. Culture plates were incubated at room temperature (RT [permissive]) and 37°C (nonpermissive) as indicated, and after 4 days, the plates were photographed.

SSY1 and PTR3 are required for the proper expression of Ssy5p. We compared the levels and electrophoretic properties of Ssy5p-c-myc in whole-cell extracts isolated from wild-type (HKY77), ssy113 (HKY84), and ptr315 (HKY85) null mutant strains (Fig. 4A). As previously indicated, Ssy5p-c-myc migrates as a 67-kDa protein (Ssy5p*) in extracts isolated from wild-type cells (Fig. 4A, lane 1). We were unable to detect Ssy5p-c-myc in extracts prepared from ssy1 null mutant cells (Fig. 4A, lane 2). In extracts derived from ptr3 null mutants, Ssy5p-c-myc migrated as a 76-kDa protein, a mobility that corresponds to the predicted molecular mass of Ssy5p (Fig. 4A, lane 3). Although the data presented in Fig. 4 were obtained by using extracts isolated from strains grown in SD medium without amino acids, similar observations regarding the behavior of Ssy5p-c-myc in ssy1 and ptr3 null mutants were made when strains were grown in SC medium. Faint immunoreactive protein bands, corresponding to twice the molecular weight of the expressed Ssy5p, were observed in extracts prepared from wild-type and ptr3 cells (most easily seen in Fig. 4A, lane 3). Consistent with Ssy5p being a membrane protein, the intensity of the slower-migrating bands varied, dependent upon the denaturing conditions used, and increased when samples were subjected to higher denaturing temperatures.

View larger version (38K): [in this window] [in a new window]   FIG. 4.   Functional expression of Ssy5p requires SSY1 and PTR3. (A) Strains HKY77 (wild type [WT]; lane 1), HKY84 (ssy113; lane 2), and HKY85 (ptr315; lane3) expressing SSY5-c-myc were grown in SD to an OD600 of 0.8, and the levels of Ssy5p-c-myc were analyzed in whole-cell lysates by SDS-PAGE and immunoblotting. Prior to electrophoresis, samples were denatured for 10 min at 37°C. (B) Total RNA was prepared from strains HKY77 (wild-type; lanes 1 and 2), HKY84 (ssy113; lane 3), and HKY85 (ptr315; lane 4) expressing SSY5-c-myc grown to an OD600 of 0.8 in SD (lanes 1, 3, and 4) or SD supplemented with 1.3 mM leucine (lane 2). The levels of SSY5 mRNA were analyzed by Northern blotting, and the levels of actin (ACT1) transcripts were used to control loading.

ssy1

ssy1

Ssy1p and Ptr3p exhibit altered electrophoretic properties dependent upon amino acid availability and sensor component interactions. Our finding that the functional expression of Ssy5p requires both SSY1 and PTR3 (Fig. 4) prompted us to examine the levels and electrophoretic properties of Ssy1p and Ptr3p. Whole-cell extracts containing functional epitope-tagged Ssy1p (Ssy1p-HA1) (34) were isolated from wild-type strains grown in media without (SD) and with (SC) added amino acids. In addition, extracts were prepared from various mutant strains lacking sensor function. In wild-type (HKY20) cells grown in SD (Fig. 5A, lane 1), Ssy1p migrates as a single major band (Ssy1p*). In cells grown in SC, a second lower band (Ssy1p) becomes evident, and the intensity of the upper Ssy1p* band is clearly diminished (Fig. 5A, lane 4). The pattern of Ssy1p staining observed in ptr3 and ssy5 null mutant strains is similar to that observed in wild-type strains grown in SC (Fig. 5A, compare lanes 5 and 6 with lane 4). This pattern is also seen in mutant strains grown in the absence of amino acids (Fig. 5A, lanes 2 and 3).

View larger version (45K): [in this window] [in a new window]   FIG. 5.   SPS sensor component interactions. (A) The pattern of Ssy1p migration during SDS-PAGE is dependent upon Ptr3p and Ssy5p and is altered in response to amino acids. Whole-cell lysates were prepared from strains HKY20 (wild-type [WT]; lanes 1 and 4), HKY33 (ptr315; lanes 2 and 5), and HKY84 (ssy52; lanes 3 and 6) transformed with pHK010 (SSY1-HA1) grown in SD (lanes 1 to 3) and SC (lanes 4 to 6) to an OD600 of 0.8. The levels of Ssy1p-HA1 in extracts were analyzed by SDS-PAGE and immunoblotting. (B) The electrophoretic mobility of Ptr3p is altered in response to amino acids. Whole-cell lysates were prepared from wild-type strain HKY31 transformed with pHK018 (PTR3-HA1) grown in SD (lane 1) and SC (lane 2) to an OD600 of 0.8. The levels of Ptr3p-HA1 in extracts were analyzed by SDS-PAGE and immunoblotting. (C) The pattern of Ptr3p migration during SDS-PAGE is dependent upon Ssy1p and Ssy5p. Whole-cell lysates were prepared from strains HKY31 (wild type; lanes 1 and 4), HKY33 (ssy1; lanes 2 and 5), and HKY85 (ssy5; lanes 3 and 6) transformed with pHK018 (PTR3-HA1) grown in SD (lanes 1 to 3) and SC (lanes 4 to 6) to an OD600 of 0.8. The levels of Ptr3p-HA1 in extracts were analyzed by SDS-PAGE and immunoblotting. Note that lane 2 in panel B and lane 4 in panel C are derived from the same sample. A longer exposure time was used for a more detailed analysis of the protein bands in lanes 4 to 6.

Overexpression of Ssy1p or the N terminus of Ssy1p disrupts amino acid sensor function. We examined the effects of individually overproducing SSY1, PTR3, and SSY5. The wild-type strain PLY1 was separately transformed with 2µm-based plasmids containing these genes as inserts. Ura⁺ transformants were selected on SC  Ura, cells derived from single colonies were resuspended in water, and dilution series were analyzed on minimal SPD medium supplemented with 0.16 mM histidine (SPD) or toxic levels of histidine (SPD plus 5 mM His). Transformants carrying plasmids PTR3 (pHK026) or SSY5 (pHK037) grew well on SPD containing 0.16 mM histidine, indicating that the overexpression of Ptr3p and Ssy5p did not have any deleterious effects on growth; however, these strains were unable to grow on SPD supplemented with toxic levels of histidine.

View larger version (43K): [in this window] [in a new window]   FIG. 6.   Overexpression of Ssy1p or the N terminus of Ssy1p exerts a dominant-negative effect on amino acid sensor function. Dilution series of strain PLY1 transformed with pRS202 (vector; lane 1), pHK012 (2µm-SSY1; lane 2), pHK039 (2µm-SSY1NT; lane 3), and strain HKY37 (ssy113) transformed with pRS202 (vector; lane 4) were spotted onto SPD and SPD supplemented with 5 mM histidine as indicated. Culture plates were incubated for 4 days at room temperature and photographed.

Time course of amino acid sensor-dependent transcriptional induction. Previous work has demonstrated that leucine is a potent inducer of PTR2 (4) and CAR1 (21) transcription. We have shown that the leucine-induced transcription of these genes is dependent on all three sensor components, SSY1, PTR3, and SSY5 (Fig. 2) (34). The time course of leucine-induced transcription of PTR2 and CAR1 was examined. The level of PTR2 transcripts increased almost 40-fold within 30 min after cells received an aliquot of 1.3 mM leucine (Fig. 7A, lanes 2 to 6). After 30 min, the level of PTR2 transcripts gradually decreased, and after 180 min, the level of PTR2 transcripts was down to half of the maximum level. The time course of leucine-induced CAR1 transcription (Fig. 7B, lanes 2 to 6) exhibited a faster response; however, in general, the pattern of induction was similar. Ten minutes after the addition of leucine, the level of CAR1 transcripts increased twofold, and after 60 min, CAR1 transcripts were restored to basal levels. No increase in PTR2 or CAR1 transcription was observed in parallel uninduced cultures (Fig. 7, lanes 7 to 11). Thus, in the presence of inducing amino acids, cells transiently upregulate the expression of PTR2 and CAR1. Similar results regarding the transcription of the branched-chained amino acid transporter genes BAP2 and BAP3 have been reported (15).

View larger version (31K): [in this window] [in a new window]   FIG. 7.   Time course of L-leucine-induced transcription of PTR2 and CAR1. A liquid culture of strain HKY77 carrying plasmid pHK048 (SSY5-c-myc) was grown in SD to an OD600 of 0.5, and the culture was split into two parts (t = 0). One-half of the cell culture received an aliquot of L-leucine (+leu; lanes 2 to 6) to a final concentration of 1.3 mM, and the other received an equal volume of water (leu; lanes 7 to 11). Both cultures were incubated at 30°C for an additional 180 min, and at the times indicated, subsamples were withdrawn and total RNA was isolated. The levels of PTR2 (A) and CAR1 (B) expression were analyzed by Northern blotting (upper panels), and after background correction, signal strengths (arbitrary units) relative to the levels of actin mRNA (ACT1) were quantitated (lower panels).

Time course of amino acid-induced sensor component modifications and sensor downregulation. We examined whether the observed amino acid-induced changes in the electrophoretic mobility of Ssy1p and Ptr3p (Fig. 5) and Ssy5p occur in a similar time frame as sensor-dependent transcriptional induction (Fig. 7). Strains expressing SSY1-HA1, PTR3-HA1, and SSY5-c-myc were grown in liquid cultures of SD to an OD₆₀₀ of 0.5. Each culture was divided into two culture flasks: one received an aliquot of leucine to a final concentration of 1.3 mM, and the other received an equal volume of water. At various times, subsamples were removed, whole-cell lysates were prepared, and the levels and electrophoretic characteristics of Ssy1p-HA1, Ptr3p-HA1, and Ssy5p-c-myc were analyzed by immunoblotting (Fig. 8).

View larger version (42K): [in this window] [in a new window]   FIG. 8.   Time course analysis of the physical alterations to Ssy1p, Ptr3p, and Ssy5p after induction of sensor function by L-leucine. Strains HKY20, HKY33, and HKY77 expressing SSY1-HA1, PTR3-HA1, and SSY5-c-myc, respectively, were grown in liquid cultures of SD to an OD600 of 0.5 (t = 0; lane 1). At t = 0, each culture received an aliquot of L-leucine to a final concentration of 1.3 mM (lanes 2 to 6). At the times indicated, subsamples were removed, whole-cell lysates were prepared, and proteins were analyzed by immunoblotting (upper panel). Lane 7 shows protein levels present in an uninduced control culture similarly incubated for 180 min. The corresponding chemiluminescent signals were quantitated (lower panel).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We have found that SSY5 encodes a third component of the yeast plasma membrane sensor of extracellular amino acids. Ssy5p functions together with the two previously characterized sensor components, Ssy1p1 and Ptr3p, to regulate diverse metabolic processes important for proper amino acid uptake and compartmentalization. The conclusion that Ssy5p is a component of the plasma membrane amino acid sensor is based on several observations. First, mutations in SSY5 belong to the same epistasis group as ssy1 and ptr3 mutations. Ssy5 mutants exhibit similar increases in vacuolar pools of histidine and arginine (Table 3). ssy5 null mutants display identical levels of resistance to toxic amino acids and azetidine carboxylate, and the ssy5 ssy1, ssy5 ptr3, and ssy1 ptr3 double mutant strains exhibit levels of resistance identical to those of each of the single mutant strains. Additionally, SSY5 is required for amino acid-induced transcription of two genes, AGP1 and PTR2, known to be controlled by SSY1 and PTR3 (Fig. 2). The resistance to toxic amino acids and amino acid analogues is likely to be a consequence of the altered uptake and increased capacity to compartmentalize amino acids (Table 3) (34). Second, functional epitope-tagged Ssy5p-c-myc fractionates as a peripherally bound membrane protein, and h-SOS-Ssy5p fusion proteins are recruited to the cytosolic face of the plasma membrane in an Ssy5p-dependent manner (Fig. 3). The observation that Ssy5p, which is predicted to be a soluble protein, is able to associate with the plasma membrane, directly implicates this protein as a constituent of this sensing system. Third, the proper expression of Ssy5p, but not the transcription of SSY5, requires both SSY1 and PTR3 (Fig. 4). Finally, the electrophoretic properties of both Ssy1p and Ptr3p are altered in ssy5 null mutant strains (Fig. 5A and C).

The demonstrated ability of Ssy1p, Ptr3p, and Ssy5p to localize to the plasma membrane is consistent with the possibility that these components associate within a sensor complex, although there is as yet no direct evidence for a physical association between them. Of the three identified sensor components, Ssy1p is the only integral membrane-spanning component; thus, Ssy1p is likely to be the component that transmits signals across the PM. The in vivo membrane topology of the general amino acid permease (Gap1p) has recently been determined; both the N- and C-terminal domains are oriented towards the cytoplasm (24). Based upon the sequence and structural homology that exists between Ssy1p and the other members of the amino acid permease gene family members, it is likely that the N terminus of Ssy1p is also cytoplasmically oriented. The large N-terminal extension of Ssy1p may serve to organize the assembly of the other peripheral membrane components. Ssy5p is a strong candidate for directly interacting with Ssy1p. In the absence of Ssy1p, Ssy5p is unstable, as evidenced by our inability to detect Ssy5p in protein lysates derived from ssy1 null mutants (Fig. 4A). The fact that Ssy5p is degraded in the absence of Ssy1p prevented us from directly assessing whether its localization to the PM is dependent upon Ssy1p. We have previously found that the Ptr3p localizes to the plasma membrane independently of Ssy1p (34).

We further investigated the functional relationships between the SPS sensor components and made several observations that indicate that these three components do indeed intimately interact with one another. We have found that each of the components displays physical properties that are dependent upon the availability of amino acids and on the presence of the other two components. In wild-type cells grown in SD, Ssy1p^* is the predominating form of Ssy1p, whereas in ptr3 or ssy5 null mutants it is not (Fig. 5A). Thus, in mutant cells lacking either PTR3 and SSY5, Ssy1p exhibits characteristics that mimic those of Ssy1p isolated from wild-type cells grown in the presence of amino acids. Additionally, we have consistently observed that cells lacking either Ptr3p and Ssy5p express less Ssy1p (Fig. 5A). Ptr3p migrates as several bands that exhibit significant differences in mobility on SDS gels (Fig. 5B). The presence of the slower-migrating Ptr3p^* species in amino acid-containing medium is dependent on SSY1, but not SSY5 (Fig. 5C); however, in SD-grown cells lacking SSY5, Ptr3p exhibits characteristics identical to those of SC-grown wild-type cells (Fig. 5C). In wild-type cells, Ssy5p is apparently proteolytically modified in a PTR3-dependent manner (Fig. 4A). Because the c-myc epitope used to visualize Ssy5p is located within the N terminus, the Ptr3p-dependent processing event is likely to occur within the C-terminal portion of Ssy5p. The inability to detect Ssy5p in whole-cell extracts prepared from strains carrying ssy1 mutations (Fig. 4A) is presumably the consequence of a proteolytic cleavage event that minimally removes its N terminus.

The observation that overexpression of the first 206 amino acids comprising the N-terminal extension of Ssy1p disrupts sensor function in a dominant-negative manner (Fig. 6) is also consistent with the existence of a multicomponent sensor complex. This finding clearly shows that the N-terminal domain of Ssy1p has an important role in sensor function, a conclusion supported by our previous observation that small in-frame mutations within the N-terminal domain abolish signaling (34). These results suggest that functional SPS sensor complexes assemble with a precise stoichiometry. Consequently, the overproduction of the N-terminus Ssy1p would interfere with SPS sensor function by forming nonproductive complexes with proteins normally interacting with Ssy1p. These nonproductive interactions would effectively decrease the availability of the limiting components to form functional sensor complexes. Similarly, dominant-negative phenotypes associated with the mutations in the cytoplasmically oriented C terminus of the G-protein-coupled alpha-factor receptor (STE2) have been observed to exert their effects by sequestering G-proteins (20, 39). The overexpression of full-length Ssy1p also exhibited dominant-negative effects; this unexpected observation may be the result of a fraction of Ssy1p being mislocalized. Accordingly, limiting sensor components would be sequestered at inappropriate intracellular membranes.

When leucine is added to wild-type cells grown in medium without supplementary amino acids, the transcription of PTR2 and CAR1 is transiently induced (Fig. 7). The response is quite rapid; within 10 min, there is a 15-fold induction of PTR2 and a 2-fold induction of CAR1. After reaching maximum levels (PTR2, 30 min; CAR1, 10 min), their transcript levels slowly adjust back to basal levels. Similar patterns of induction have been reported for the branched-chain amino acid permeases (BAP2 and BAP3) (15). Concurrent with its effect on transcription, leucine stimulates the components of the SPS sensor to become physically modified and causes the levels of each of the SPS sensor components to diminish (Fig. 8). The rapid physical alterations and reduced levels of sensor components are consistent with their being downregulated in response to amino acid availability. It is important to note that in cells grown in medium supplemented with amino acids, the downregulated sensor components are necessary to maintain the steady-state transcript levels of AGP1 and PTR2 (Fig. 2) and the glutamine permease (GNP1) (34). Additionally, the downregulated SPS sensor is required to fully repress the functional expression of Gap1p (34).

Regarding the leucine-induced mobility changes and downregulation of SPS sensor components (Fig. 8), we have defined three sensor states, each associated with defined patterns of component expression (see Fig. 9 for a schematic presentation). In the absence of extracellular amino acids, the state I, or preactivation, conformation, Ssy1p migrates predominantly in its slower-migrating form (Ssy1p^*), Ptr3p is primarily present in its faster-migrating form, and Ssy5p is readily detected in its Ptr3p-dependent low-molecular-mass processed form (Ssy5p^*). Within minutes after the addition of leucine, rapid changes are observed, resulting in the transient state IIa sensor complex (Fig. 9). The characteristics of each component within the state IIa sensor correspond to those observed after 10 min of leucine addition (Fig. 8). The state IIa sensor is defined by the low levels of Ssy5p; the electrophoretic characteristics of the other two components appear unchanged. With increasing time (30- and 60-min time points) (Fig. 8), shifts in the migration of Ssy1p and Ptr3p become increasingly obvious (Fig. 9, state IIb conformation). Within the transitional state IIb complex, the levels of Ssy1p^* decrease and the relative proportion of the slower-migrating form of Ptr3p (Ptr3p^*) increases. The state III configuration, evident 120 min after leucine addition, is indistinguishable from the downregulated sensor conformation that exists in cells grown in SC media. It is possible that the downregulated sensor conformation is actually comprised of a mixed sensor population; the patterns of bands associated with the state I conformation are still evident in the state III sensor.

View larger version (33K): [in this window] [in a new window]   FIG. 9.   Schematic diagram summarizing the dynamic characteristics of the SPS sensor component interactions. The state I sensor is the complex present in cells grown in the absence of amino acids. The SPS components are present in high levels, represented by the heavy outlines. In analogy to the G-protein-coupled -factor receptor complex in MATa cells (39), the state I conformation may represent a preactivation complex. States IIa and IIb are transient complexes that rapidly form when cells grown in the absence of amino acids are induced by amino acids. The components in the transient state IIa and IIb sensor undergoing dynamic changes in expression levels are represented by the dashed outlines. The state III conformation is the downregulated complex, and the diminished levels of the components are represented by light outlines. For additional details, see text.

In our analysis to date, we have defined three components of a plasma membrane sensor of extracellular amino acids. Further biochemical and genetic analysis is necessary to ascertain if these components represent the entire primary sensing complex or if other components exist. The isolation of dominant constitutively activated alleles in any of the genes encoding SPS sensor components would enable epistasis relationships to be better defined. Because the SPS sensor components are localized to the plasma membrane, it is possible that the SPS sensor components, particularly Ssy1p, are subject to regulation by posttranslational mechanisms known to regulate amino acid permeases and other metabolite transporters. In response to ammonium, Gap1p is dephosphorylated and polyubiquitinated (27, 60). Similar mechanisms, as well as substrate-induced feedback inhibition, operate to regulate the uracil permease (23, 58). Finally, other proteins associated with the plasma membrane regulate amino acid uptake. For example, the target of rapamycin pathway, in a manner resembling the SPS sensor, inversely regulates the activity of general and specific amino acid permeases (6, 7, 55). ScRheb (YCR027c) is a member of a new class of farnesylated small G-proteins of the Ras superfamily, that negatively regulates the uptake of arginine by Can1p (61). This regulation is thought to occur at the plasma membrane directly with or through other proteins interacting with the Can1p permease. The fact that yeast plasma membrane nutrient sensors have only recently been discovered reveals how little is understood regarding the molecular signals that enable yeast to adapt to constantly changing environments. Many more novel discoveries can be expected.