Rheb, a Ras-like small GTPase conserved from human to yeast, controls Tor kinase and plays a central role in the regulation of cell growth depending on extracellular conditions. Rhb1 (a fission yeast homolog of Rheb) regulates amino acid uptake as well as response to nitrogen starvation. In this study, we generated two mutants, rhb1-DA4 and rhb1-DA8, and characterized them genetically. The V17A mutation within the G1 box defined for the Ras-like GTPases was responsible for rhb1-DA4 and Q52R I76F within the switch II domain for rhb1-DA8. In fission yeast, two events—the induction of the meiosis-initiating gene mei2+ and cell division without cell growth—are a typical response to nitrogen starvation. Under nitrogen-rich conditions, Rheb stimulates Tor kinase, which, in turn, suppresses the response to nitrogen starvation. While amino acid uptake was prevented by both rhb1-DA4 and rhb1-DA8 in a dominant fashion, the response to nitrogen starvation was prevented only by rhb1-DA4. rhb1-DA8 thereby allowed genetic dissection of the Rheb-dependent signaling cascade. We postulate that the signaling cascade may branch below Rhb1 or Tor2 and regulate the amino acid uptake and response to nitrogen starvation independently.
SIGNALING involving mTOR kinase is evolutionally conserved from yeast to human. In vertebrates, it regulates cell growth depending on the availability of nutrients, energy source, and growth factors (Hay and Sonenberg 2004; Wullschleger et al. 2006; Guertin and Sabatini 2007). The mTOR signaling involves a Ras-like GTPase, Rheb that promotes protein synthesis and thereby cell growth via mTOR (Saucedo et al. 2003; Stocker et al. 2003). When the environment surrounding the cell is not favorable for growth and proliferation, the protein complex of TSC1 and TSC2, which has a GAP GTPase-activating protein (GAP) activity, converts Rheb into a GDP-bound form (Inoki et al. 2003; Tee et al. 2003; Zhang et al. 2003; Li et al. 2004b). A loss of TSC1/2 would thereby result in constitutive upregulation of the Rheb GTPase and of mTOR and, in humans, allow the wide spread of benign tumors termed “hamartomas” in different organs including the brain, eyes, heart, kidney, skin, and lungs (Manning and Cantley 2003; Li et al. 2004a; Pan et al. 2004).
In fission yeast, a prototype of the TOR-signaling system consists of Tsc1/2, Rhb1 GTPase (a homolog of Rheb GTPase), and Tor2 kinase (a homolog of mTOR), as in the mTOR-signaling system in higher eukaryotes (Otsubo and Yamamato 2008). It has been demonstrated by a number of studies that the Tor signaling in fission yeast controls various biological responses to starvation of nitrogen. Upon starvation of nitrogen, fission yeast cells are arrested at G1 with small- and round-cell morphology following two rounds of rapid cell cycling without net growth (Young and Fantes 1987). They also induce expression of the mei2+ gene that is required for initiation of meiosis (Watanabe et al. 1988). Upon a loss of function of tor2+ gene, these events are induced even under a nitrogen-rich condition, indicating that Tor2 is responsible for suppression of the events that are normally induced upon nitrogen starvation (Alvarez and Moreno 2006; Uritani et al. 2006; Hayashi et al. 2007; Weisman et al. 2007; Matsuo et al. 2007). The events adaptive to nitrogen starvation are important for the preservation and survival of the organism, but they are deleterious for actively growing cells. Thus, it is critical for the cell to strictly control the activity of the Tor2 kinase depending on the nutrient condition.
Rheb/Rhb1 GTPase plays a central role as a signal carrier in the Tor signaling. In Drosophila, it has been shown that Rheb promotes cell growth via the Tor-signaling system (Saucedo et al. 2003; Stocker et al. 2003). In fission yeast, a loss of rhb1+ causes a cell cycle arrest at G1 and induction of fnx1+ and mei2+ genes, both of which are normally suppressed by Tor2 (Mach et al. 2000; Matsuo et al. 2007). These genetic studies have indicated that Rheb/Rhb1 positively regulates Tor kinase. Biochemical studies have also demonstrated close interaction between Rheb/Rhb1 GTPase and Tor kinase. It has been shown that when Rheb is ectopically expressed, it binds to mTOR (Long et al. 2005). Subsequently, it has been demonstrated that purified Rheb loaded with GTP, but not with GDP, activates mTOR kinase in vitro (Sancak et al. 2007). Ras-like small GTPases, in general, bind to and stimulate their targets only when they are loaded with GTP. The above two studies, however, have reported that binding between Rheb and mTOR is not influenced by the guanine-nucleotide binding state of Rheb. In contrast, Rhb1 in a GTP-bound form binds to Tor2 kinase in fission yeast (Urano et al. 2005; Uritani et al. 2006). Another study has demonstrated that GTP-bound Rheb directly binds to FKBP38, a member of the FK506-binding protein, and protects mTOR from inhibition by FKBP38 (Bai et al. 2007), suggesting that Rheb activates mTOR rather indirectly, although this model has been disputed by a recent study (Wang et al. 2008). Obviously, further investigation is needed to formulate a general view of a molecular mechanism by which Rheb/Rhb1 GTPase interacts with Tor as well as with other components.
In our previous studies, we showed that deletion of the tsc2+ gene encoding a GAP for Rhb1 GTPase confers a defect in uptake of amino acids (Matsumoto et al. 2002) and prevents induction of the mei2+ gene upon nitrogen starvation (Nakase et al. 2006). Rhb1 GTPase, which is assumed to be in a GTP-bound form in cells lacking Tsc2 (Δtsc2), is likely a cause of these defects. In this study, we generated two dominant active rhb1 mutants and found that they distinctively respond to nitrogen starvation. One mutant (rhb1-DA4) exhibits the phenotypes identical to that of Δtsc2. The other one (rhb1-DA8), however, is defective in uptake of amino acids, but allows induction of the mei2+ gene upon nitrogen starvation. Remarkably, rhb1-DA8 also allows induction of the mei2+ gene in the Δtsc2 background. On the basis of these results, we propose a model for Rhb1 signaling in fission yeast.
MATERIALS AND METHODS
Yeast strains, media, and transformation:
The Schizosaccharomyces pombe strains used in this study are listed in Table 1. The yeast cells were grown in yeast extract with supplements (YES) media and EMM synthetic minimal media with appropriate nutrient supplements as described by Moreno et al. (1991). All yeast transformations were carried out by lithium acetate methods (Okazaki et al. 1990; Gietz et al. 1992).
Mutagenesis of rhb1+:
The rhb1+ gene was amplified by PCR using the forward primer 5′-rhb1 [5′-GGGGGGGGGTCGAC(SalI)ATGGCTCCTATTAAATCTCGTAGA-3′] and the reverse primer 3′-rhb1 [5′-CCCCCCCCGGATCC(BamHI)TTAGGCGATAACACAACCCTTTCC-3′]. To introduce mutations, PCR was performed in the presence of 0.5 mm Mn2+ and 1 mm Mg2+. The resulting fragments were digested with BamHI and SalI and then cloned into pREP81 to construct a pool of rhb1 mutants.
Deletion of rhb1+ gene:
A plasmid to delete the rhb1+ gene, pBS-Δrhb1, was constructed by PCR amplification of a 1.1-kbp DNA fragment upstream of the start codon of the rhb1+ gene and a 1.4-kbp DNA fragment downstream of the stop codon. The primer sequences used in the PCR were as follows: for amplification of the upstream sequence, the forward primer was rhb1-NotI-1 [5′-CCCCCCGCGGCCGC(NotI)GTGTAAAGGGAGCCGTTCAAG-3′] and the reverse primer was rhb1-BamHI-2[5′-CCCCCCGGATCC(BamHI)GGCAAATTATTAACTATAGAG-3′]; for the downstream sequence, the forward primer was rhb1-BamHI-3 [5′-CCCCCCGGATCC(BamHI)GCTTCTGCTTGAATTTATC-3′] and the reverse primer was rhb1-SalI-4 [5′-CCCCCCGTCGAC(SalI)GTACGTTCAATTCCTATTC-3′]. The resulting DNA fragments were digested with combinations of appropriate restriction enzymes and then ligated into pBS plasmids to create pBS-rhb1-up and pBS-rhb1-down. The pBS-rhb1-down plasmid was digested with NotI and BamHI and ligated with a NotI-BamHI fragment that contained the rhb1 upstream sequence isolated from pBS-rhb1-up to create pBS-rhb1-up-down. pBS-rhb1-up-down was digested with BamHI and ligated with a DNA fragment that contained the ura4+ gene to create pBS-Δrhb1. Digestion of pBS-Δrhb1 with NotI and SalI generated a 4.3-kbp DNA fragment containing the ura4+ gene flanked by the 1.1-kbp upstream and the 1.4-kbp downstream sequences of the rhb1+ gene, which was used for transformation to delete the rhb1+ gene in a diploid strain.
Cells were cultured in liquid YES or EMM medium at a concentration of 1 × 107 cells/ml, and each culture was diluted by a factor of 10. Five microliters of each suspension was spotted onto appropriate media.
Total cell lysates were prepared as follows. Cells were lysed with glass beads in 1× PBS (140 mm NaCl, 2.7 mm KCl, 1.5 mm KH2PO4, and 8.1 mm Na2HPO4) containing 1 mm MgCl2, 0.5% Triton X-100, and 0.5% deoxycholate. The following protease inhibitors were added to the cell extracts: 1 mm phenylmethylsulfonyl fluoride and 1× protease inhibitor cocktail (Nacalai Tesque). Equal amounts of total proteins were then loaded onto a 15% polyacrylamide gel and transferred to nitrocellulose membranes. Filters were probed with anti-Rhb1 rabbit polyclonal antibodies (Nakase et al. 2006) at a 1:2000 dilution. Blots were also probed with an anti-α-tubulin monoclonal antibody, TAT-1 (a gift from K. Gull, University of Manchester, Manchester, UK), to normalize protein loading. Other procedures used were Northern blotting and microarray analysis, which were performed as described previously (Nakase et al. 2006; Chikashige et al. 2007).
Screen for rhb1 mutants:
We and other groups previously showed that a loss of Tsc2 (GTPase-activating protein for Rhb1 GTPase) causes a reduction in the uptake of amino acids and resistance to canavanine, a toxic analog of arginine (van Slegtenhorst et al. 2004; Nakase et al. 2006). Rheb GTPase, which would predominantly remain as a GTP-bound form in the cells lacking the Tsc2 activity, is likely a cause of the low uptake of amino acids and the resistance to canavanine. We reasoned that a mutation on the rhb1+ gene, which can stably maintain bound GTP or abnormally activate a downstream component, might confer a phenotype similar to that observed in cells lacking Tsc2 (Δtsc2). To obtain such a mutation, a pool of the rhb1 mutants was generated by error-prone PCR. Individual PCR products were cloned into a plasmid, pREP81, which allowed expression of a cloned gene from an inducible promoter, nmt1, and transformed into a wild-type strain. We screened for transformants, which became resistant to canavanine upon expression of the cloned rhb1 mutant. Among ∼40,000 transformants, we found 3 exhibiting strong resistance to canavanine upon expression of the rhb1 gene, which presumably carried a mutation.
Construction of rhb1-DA4 and rhb1-DA8:
The plasmids were recovered from the three transformants, and the nucleotide sequence of the rhb1+ gene (designated rhb1-DA4, -DA5, and -DA8, respectively) was analyzed. Because we found that the rhb1 gene isolated from each transformant carried multiple mutations, we determined the mutation sites responsible for conferring the resistance to canavanine. As shown in Figure 1A, the mutation of rhb1-DA4 responsible for conferring the resistance to canavanine is valine at position 17 replaced with alanine (V17A). The rhb1-DA5 carries the responsible mutation at the same position, but replaced with aspartic acid (V17D). The rhb1-DA8 carries two mutations replacing glutamine at position 52 with arginine (Q52R) and isoleucine at position 76 with phenylalanine (I76F), both of which are necessary for conferring the resistance to canavanine (Figure 1A).
A previous report (Urano et al. 2005) described several hyperactive mutations of the fission yeast rhb1 gene, including V17A (identical to rhb1-DA4) and V17G within the G1 box, a domain important for guanine nucleotide binding (Figure 1B and see discussion). Biochemical analysis indicated that the V17G mutation inhibits both GTP and GDP binding. It is thereby plausible that Rhb1 GTPase stimulates its effector regardless of the guanine-nucleotide-binding status if valine at position 17 is mutated to glycine or alanine. The rhb1-DA8 contains mutations at Q52 and I76, both of which are not conserved in other small GTPases and thus its biochemical character is less predictable.
We introduced the rhb1-DA4 and rhb1-DA8 to the native rhb1 locus. In the heterozygous diploid strain rhb1+/Δrhb1, the deletion allele (Δrhb1) was replaced with either the rhb1-DA4 allele or the rhb1-DA8 allele. The resulting heterozygous diploids were sporulated, and each of them was found to successfully produce four viable spores. We determined the genotype of each cell by sequencing the rhb1 locus and confirmed that each diploid produced two wild-type cells and two rhb1 mutants.
Genetic analysis of rhb1-DA4 and rhb1-DA8:
We examined the rhb1-DA4 and rhb1-DA8 mutants for their resistance to canavanine. As shown in Figure 2A, the two mutants were resistant to canavanine, although the rhb1-DA4 mutant was resistant to a lesser extent. We also found that these mutants were defective in uptake of leucine (Figure 2B). The heterozygous diploids rhb1+/rhb1-DA4 and rhb1+/rhb1-DA8 were also resistant to canavanine (Figure 2C), indicating that these two mutations are dominant.
We previously showed that a loss of the cpp1+ gene (cpp1-1) suppresses defects associated with Δtsc2 (Nakase et al. 2006). The cpp1+ gene encodes an enzyme that farnesylates the C terminus of Rhb1 GTPase. We speculate that a failure in farnesylation results in a partial loss of the Rhb1 function, which contributes to the suppression of Δtsc2. To test whether the cpp1-1 mutation could suppress rhb1-DA4 and rhb1-DA8, it was introduced into each of the rhb1 mutants. As shown in Figure 2D, the cpp1-1 mutation fully suppressed the resistance to canavanine in rhb1-DA4. In contrast, the suppression of the resistance by the cpp1-1 mutation was only partial in the rhb1-DA8 mutant. We also examined Rhb1-DA4 and Rhb1-DA8 for their modification by Cpp1. It was shown previously that Rhb1 GTPase in fission yeast (Yang et al. 2000) migrates faster on SDS–PAGE if properly modified. As shown in Figure 2E, Rhb1, Rhb1-DA4, and Rhb1-DA8 were detected as a doublet in cell extracts prepared from the cpp1-1 mutant grown at 26°. Three hours after the shift to the restrictive temperature 36°, the slower-migrating form of Rhb1 increased. In cell extracts prepared from the wild-type background, only faster-migrating forms of Rhb1-DA4 and Rhb1-DA8 were detected. These results indicated that Rhb1-DA8 was not fully modified with a farnesyl group in the cpp1-1 mutant even at 26°. It is thus likely that Rhb1-DA8 GTPase is less dependent on the modification by Cpp1 to cause the resistance to canavanine.
Response to nitrogen starvation:
Fission yeast cells, when grown in nitrogen-free media, undergo stimulated rates of division mostly due to a shortened G2 phase, a period important for cell growth for this organism (Young and Fantes 1987). As a result, cells become small and round. They also allow induction of the meiosis-initiating gene mei2+. These two events are a hallmark of the response to nitrogen starvation in fission yeast. In this study, the two dominant active rhb1 mutants, along with Δtsc2, were examined for their response to nitrogen starvation in detail.
We first examined cell morphology and rates of cell division after nitrogen starvation. As shown in Figure 3, 4 hr after the shift to the nitrogen-free media, the wild-type cells became round and small. Its growth rate was stimulated upon nitrogen starvation (Figure 4 ). In contrast, Δtsc2 and the rhb1-DA4 mutant responded minimally to nitrogen starvation. Their morphology did not change dramatically (Figure 3), and their rates of division were lower in the absence of a nitrogen source (Figure 4). We also examined the rhb1-DA8 mutant and unexpectedly found that it behaved just like the wild-type strain in the nitrogen-free media. As shown in Figures 3 and 4, it exhibited shorter and round morphology after stimulated cell division upon the shift to the nitrogen-free media.
We next examined expression of three genes, which are abnormally expressed in Δtsc2 upon the shift to the nitrogen-free media. The mei2+ gene that is induced in the wild-type strain is not induced in Δtsc2. On the other hand, the inv1+ gene and Tf2 (a transposable element) that are not induced in the wild-type strain are induced in Δtsc2 (Nakase et al. 2006).
As shown in Figure 5A, in the rhb1-DA4 mutant, the mei2+ gene was not induced and the inv1+ gene and Tf2 were induced. In addition, we found that introduction of cpp1-1, a suppressor of Δtsc2, also suppressed the phenotype of rhb1-DA4 (Figure 5B). These results indicated that the rhb1-DA4 mutation causes phenotypes indistinguishable from that of Δtsc2. In contrast, expression of the three genes was normal in rhb1-DA8. Upon nitrogen starvation, the mei2+ gene was induced and the two genes, inv1+ and Tf2, were not (Figure 5A). We also analyzed the gene expression profile in a genomewide scale by microarrays. As shown in Table 2, while a number of genes were abnormally expressed in Δtsc2 and the rhb1-DA4 mutant, they were normally expressed in the rhb1-DA8 mutant.
Suppression of Δtsc2 by rhb1-DA8:
The results described above uncovered a very unique feature of the rhb1-DA8 mutation. Although it strongly confers the resistance to canavanine in a dominant manner, the rhb1-DA8 mutation does not affect the responses to nitrogen starvation that are induction of the mei2+ gene and stimulated cell division with no net growth. Because these two events are regulated by the Tor2 kinase (Alvarez and Moreno 2006; Uritani et al. 2006; Hayashi et al. 2007; Matsuo et al. 2007; Weisman et al. 2007), we speculated that (1) the rhb1-signaling cascade branches into two pathways, the Tor2-dependent pathway and an unknown one, and (2) the Rhb1-DA8 GTPase might preferentially stimulate the unknown target that is responsible for conferring the resistance to canavanine. To further understand the nature of the rhb1-DA8 mutation, we constructed a double mutant, Δtsc2 rhb1-DA8, and examined it for the response to nitrogen starvation.
As shown in Figure 4 and Figure 6A, upon the shift to the nitrogen-free media, the double mutant became shorter and round following a stimulated division. It also allowed induction of the mei2+ gene (Figure 6B). These results suggested that the Rhb1 GTPase, which would remain in a GTP-bound form in Δtsc2, cannot stimulate the Tor2 kinase in the rhb1-DA8 background.
Interaction between Tor and Rhb1-DA:
To further explore the functional relationship between Tor2 and Rhb1, we attempted to examine the effect of the rhb1-DA4 and rhb1-DA8 mutations on a temperature-sensitive allele of tor2, tor2-ts6, or tor2-ts10 (Matsuo et al. 2007). As shown in Figure 7A, the tor2-ts6 mutant was resistant to canavanine, but to a lesser extent compared with the rhb1-DA4 or Δtsc2 single mutants. We introduced rhb1-DA4, a mutation previously characterized as hyperactive (Urano et al. 2005), into tor2-ts6 and found that the double mutant exhibited a canavanine resistance similar to that of the rhb1-DA4 single mutant (Figure 7A). We also constructed the double mutant tor2-ts6 Δtsc2 and found that it exhibited a resistance to canavanine similar to that of the Δtsc2 mutant. These results thereby indicated that the resistance to the drug conferred by rhb1-DA4 or Δtsc2 was not influenced by tor2-ts6. The resistance to canavanine conferred by tor2-ts6 was likely due to a defect in uptake because the tor2-ts6 leu1-32 mutant, whose growth was dependent on leucine supplemented in synthetic media, grew poorly on EMM media containing leucine at a concentration of 40 μg/ml (Y. Nakase, unpublished result).
The crosses between the tor2-ts6 mutant and the rhb1-DA8 mutant produced tetrads in which only two or three spores were viable (Figure 7B). We determined the genotype of each viable spore by sequence analysis to identify the rhb1-DA8 mutation and by testing the temperature sensitivity to identify the tor2-ts6 mutation. We could thereby deduce the genotype of the nonviable spores. The results indicated that all the nonviable spores were double mutants (tor2-ts6 rhb1-DA8). Other tetrads produced four viable spores, two of which were tor2-ts6 and the other two of which were rhb1-DA8 (not shown). We thus concluded that the rhb1-DA8 mutation caused a synthetic lethality with the tor2-ts6 mutation. The crosses between the tor2-ts10 mutant and the rhb1-DA8 mutant produced double mutants (tor2-ts10 rhb1-DA8), which could occasionally form tiny colonies after incubation for >10 days. We, however, failed to isolate a double-mutant strain due to spontaneous reversion (not shown). These results indicated that the rhb1-DA8 allele produced a synthetic lethal or sick phenotype when combined with the tor2-ts alleles. Rhb1-DA8 could not normally stimulate Tor2.
Because we could not construct a double mutant (rhb1-DA8 tor2-ts6), the rhb1-DA8 mutation was introduced ectopically into the tor2-ts6 mutant by transformation. As shown in Figure 7C, upon introduction of the rhb1-DA8 gene into the tor2-ts6 mutant, the mutant became more resistant to canavanine. Similarly, the rhb1-DA4 gene conferred a stronger resistance to the drug when it was introduced ectopically into the tor2-ts6 mutant.
We finally examined the functional relationship between Tor1 and Rhb1. As shown in Figure 7D, the resistance to canavanine conferred by rhb1-DA4, rhb1-DA8, or Δtsc2 was not fully suppressed by Δtor1, indicating that Tor1 does not directly interact with Rhb1 to cause the resistance to the drug.
In this study, we generated two dominant active rhb1 mutants, rhb1-DA4 and rhb1-DA8, and characterized them genetically in fission yeast. Although the two mutants were isolated through the same genetic screen, they responded to nitrogen starvation quite differently.
Two dominant active mutants, rhb1-DA4 and rhb1-DA8:
In the rhb1-DA4 mutant, valine at position 17 is replaced with alanine. Ras and other small GTPases are built with several conserved functional domains, namely, G1, G2, G3, G4, and G5 boxes (Bourne et al. 1991). The V17A mutation is located in the G1 box (Figure 1B) that contacts phosphates of GTP or GDP. This mutation (V17A) and a similar one (V17G) in fission yeast Rhb1 GTPase were also identified previously in a screen similar to that reported in this study. It was demonstrated that the mutation of V17G abolishes the guanine nucleotide binding (Urano et al. 2005). Activating mutations in budding yeast Ras2 at an analogous position such as V21D and V21G were also shown to abolish the nucleotide binding (Dalley and Cannon 1996). Our analysis in this study indicated that the rhb1-DA4 mutation mimics Δtsc2, in which Rhb1 GTPase would remain as a GTP-bound form due to loss of the Tsc2-dependent GAP activity. Taken together, we speculate that rhb1-DA4 can stimulate its downstream elements regardless of the guanine-nucleotide-binding status.
The rhb1-DA8 carries two mutations replacing glutamine at position 52 with arginine (Q52R) and isoleucine at position 76 with phenylalanine (I76F). The amino acids at these positions are conserved in human Rheb (Figure 1B). Of the two mutations in rhb1-DA8, I76F is located within switch II, a domain considered to be important for interaction with GAP and effectors (Vetter and Wittinghofer 2001; Yu et al. 2005). A recent study of human Rheb has demonstrated that a double mutation of I76A D77A almost completely abolishes the ability to stimulate its effector, mTOR, although the mutation does not affect binding to mTOR (Long et al. 2007). These results would suggest that the I76F mutation in rhb1-DA8 might affect the ability of Rhb1 GTPase to stimulate the Tor2 kinase. While the rhb1-DA8 mutant exhibits strong resistance to canavanine, it allows response to nitrogen starvation. The mutant divides without net growth and allows induction of the mei2+ gene upon the shift to the nitrogen-free media. Because these events that are adaptive to nitrogen starvation are normally suppressed by the Tor2 kinase, the phenotype of the rhb1-DA8 mutant would also suggest a weak activity of Rhb1-DA8 GTPase to stimulate Tor2. More directly, the rhb1-DA8 mutation causes a synthetic lethality with the tor2-ts6 mutation (Figure 7B). On the basis of the results from our genetic analysis and the preceding biochemical and structural studies of human Rheb, it is likely that the Rhb1-DA8 GTPase is partially defective in activating the Tor2 kinase.
Roles of the Tor2 kinase in uptake:
It was previously shown that overexpression of tor2+ caused a resistance to canavanine in a wild-type strain, but not in a strain deleted for the tor1+ gene (Weisman et al. 2007). Activation of Tor2 kinase may thereby inhibit uptake of amino acids in a tor1+-dependent manner. In this study, we found that a temperature-sensitive allele of tor2+, tor2-ts6, was defective in uptake of leucine and conferred a resistance to canavanine. The two results are seemingly inconsistent.
In the fission yeast genome, ∼20 amino acid permeases are encoded, of which 11 permeases are activated at the transcriptional level upon nitrogen starvation. On the other hand, 5 are unchanged or even inactivated (Sanger Center Gene DB http://www.genedb.org/genedb/pombe/index.jsp and Nakase et al. 2006). It is likely that some permeases are required when the cell is starved for nitrogen. Other permeases, whose expression is not influenced by the availability of nitrogen, are responsible for uptake under actively growing conditions. The Tor2 kinase, as an essential component for growth, may be responsible for activating a set of permeases under actively growing conditions. A partial loss of function of tor2+ would thereby result in reduction of a total activity of uptake and consequently confer a resistance to canavanine. It is also possible that the Tor2 kinase may be responsible for regulation of a set of permeases, which are expressed at a basal level under actively growing conditions, but derepressed upon nitrogen starvation. Overexpression of tor2+ would result in suppression of the activity of these permeases and thereby confer a resistance to canavanine. Indeed, expression two permeases, Isp5 and 7G5.06, have been shown to be repressed in Δtsc2, a genetic background in which Tor2 kinase would be activated (Weisman et al. 2007).
The mechanism to control the uptake of amino acids is still to be elucidated on the molecular basis. In addition to regulation at the transcriptional level, the total activity of the uptake is likely regulated by stability and/or localization of individual amino acid permeases according to the nutrient condition. It was demonstrated recently that a loss of tsc2+ (Δtsc2) caused mislocalization of Cat1, a cationic amino acid permease, and that a mutation in the pub1+ gene coding a ubiquitin ligase suppressed the mislocalization in Δtsc2 (Aspuria and Tamanoi 2008). Pub1 would be a key component in determining the activity of the uptake at the post-translational level. Its functional relationship with Rhb1 and Tor2 is to be elucidated in the future.
Signaling by Rhb1 GTPase:
While amino acid uptake was prevented by both rhb1-DA4 and rhb1-DA8 in a dominant fashion, the response to nitrogen starvation was prevented only by rhb1-DA4. rhb1-DA8 thereby allows genetic dissection of the Rheb-dependent signaling cascade.
Although both overexpression of tor2+ and the rhb1-DA8 mutant causes resistance to canavanine, we speculate that they cause the resistance to the drug by independent mechanisms for the following reasons: (1) the synthetic lethality of the rhb1-DA8 tor2-ts6 double mutant suggests that the rhb1-DA8 mutant activates the Tor2 kinase only partially, and (2) the rhb1-DA8 mutant, unlike overexpression of the tor2+ gene, confers the resistance to canavanine independently from the function of Tor1. In addition, resistance to canavanine exhibited by a particular mutant could indicate that the mutant is tolerant to non-native proteins that have misincorporated canavanine. In this respect, we should not uniformly consider mutants resistant to the drug to be defective in uptake of amino acids.
The signaling cascade would branch below the Rhb1 GTPase or the Tor2 kinase. If it branches just below the Rhb1, the Rhb1 GTPase may regulate Tor2 and an element yet to be identified. Tor2, when stimulated by the Rhb1 GTPase, is responsible for suppressing the events adaptive to nitrogen starvation. The other one would be responsible for controlling amino acid uptake. While Rhb1-DA4 would constitutively stimulate both of these elements, Rhb1-DA8 would largely stimulate the one controlling the uptake. We demonstrated in this study that the Rhb1-DA8 GTPase could not stimulate Tor2 even in the Δtsc2 background, a condition under which the Rhb1-GTPase would exist as a GTP-bound form. Thus, the ability of the Rhb1-DA8 GTPase to stimulate its downstream elements may not be influenced by the guanine-nucleotide-binding status. It is rather likely that the Rhb1-DA8 GTPase may have an altered specificity toward its downstream elements and preferentially stimulate the one controlling the uptake of amino acids.
It is equally possible that the signal cascade branches below Tor2. The Tor2 kinase might exist as at least two different forms: one to support cell growth and suppress the response to nitrogen starvation and the other one to regulate the activity of a particular set of amino acid permeases at the transcriptional, translational, or post-translational level. Rhb1-DA8 may specifically stimulate the later form of Tor2 and reduce a total activity of the amino acid uptake. A partial loss of Tor2 suppresses a defect in the uptake of adenine caused by Δtsc2 (Matsuo et al. 2007). Another form of Tor2 might be involved in the uptake of nucleotides.
Tuberous sclerosis complex pathology:
Mutations in the human genes TSC1 and TSC2 are predisposing to a disease, tuberous sclerosis complex (TSC) (European Chromosome 16 Tuberous Sclerosis Consortium 1993; van Slegtenhorst et al. 1997). Because we showed that the rhb1-DA4 mutation mimics deletion of the tsc2+ gene in fission yeast, it is plausible that an analogous mutation in human Rheb would cause a symptom similar to that found in patients with TSC. Human Rheb could be a third TSC gene. Because our model study in fission yeast demonstrated that the phenotypes significantly varied depending on the allele, the symptoms of the Rheb-induced TSC could also be variable.
We thank M. Yamamoto (University of Tokyo, Tokyo, Japan) for strains. This work was supported by a grant from U. S. Army Medical Research and Material Command “Tuberous Sclerosis Complex Research Program” (to T. Matsumoto), by a grant from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to Y.H.), and by a grant from the Japan Society for the Promotion of Science (JSPS) (to Y.C.). T. Murai was a recipient of a scholarship from Teijin (Osaka, Japan) and a predoctoral fellowship from JSPS. Y.N. was supported by an exploratory research grant from JSPS.
↵1 Present address: Yamada Bee Farm, 194 Ichiba, Kagamino-cho Tomada-gun, Okayama 708-0393, Japan.
Communicating editor: O. Cohen-Fix
- Received May 22, 2009.
- Accepted July 16, 2009.
- Copyright © 2009 by the Genetics Society of America