|Home | About | Journals | Submit | Contact Us | Français|
Evolutionarily conserved target of rapamycin (TOR) complex 1 (TORC1) responds to nutrients, especially amino acids, to promote cell growth. In the yeast Saccharomyces cerevisiae, various nitrogen sources activate TORC1 with different efficiencies, although the mechanism remains elusive. Leucine, and perhaps other amino acids, was reported to activate TORC1 via the heterodimeric small GTPases Gtr1-Gtr2, the orthologues of the mammalian Rag GTPases. More recently, an alternative Gtr-independent TORC1 activation mechanism that may respond to glutamine was reported, although its molecular mechanism is not clear. In studying the nutrient-responsive TORC1 activation mechanism, the lack of an in vitro assay hinders associating particular nutrient compounds with the TORC1 activation status, whereas no in vitro assay that shows nutrient responsiveness has been reported. In this study, we have developed a new in vitro TORC1 kinase assay that reproduces, for the first time, the nutrient-responsive TORC1 activation. This in vitro TORC1 assay recapitulates the previously predicted Gtr-independent glutamine-responsive TORC1 activation mechanism. Using this system, we found that this mechanism specifically responds to l-glutamine, resides on the vacuolar membranes, and involves a previously uncharacterized Vps34-Vps15 phosphatidylinositol (PI) 3-kinase complex and the PI-3-phosphate [PI(3)P]-binding FYVE domain-containing vacuolar protein Pib2. Thus, this system was proved to be useful for dissecting the glutamine-responsive TORC1 activation mechanism.
The target of rapamycin (TOR) is an evolutionarily conserved protein kinase that regulates cell growth as the catalytic component of rapamycin-sensitive TOR complex 1 (TORC1) (1, 2). TORC1 is activated by nutrients, particularly amino acids, in all tested eukaryotes, and active TORC1 promotes cell growth by activating anabolic processes, including synthesis of protein, lipids, and nucleotides and by inhibiting catabolic processes, such as autophagy. In mammals, TORC1 consists of mammalian/mechanistic TOR (mTOR), raptor, mammalian Lst8 (mLst8), DEP domain-containing mTOR-interacting protein (DEPTOR), and PRAS40, whereas in the yeast Saccharomyces cerevisiae, TORC1 consists of Tor1 or Tor2, the raptor orthologue Kog1, the mLst8 orthologue Lst8, and Tco89. Recent studies have revealed an evolutionarily conserved amino acid-responsive TORC1 activation mechanism, in which heterodimeric small GTPases, RagA or RagB (RagA/B)-RagC/D in mammals and Gtr1-Gtr2 in yeast, play a central role (3,–5). The mammalian Rag heterodimer is anchored to the lysosomal membrane through the Ragulator complex. An amino acid stimulus induces conversion of RagA/B from the GDP-bound state to the GTP-bound state, and the GTP-bound form directly binds to TORC1 (3, 4), thereby recruiting TORC1 to the lysosomal membrane where the TORC1 activator Rheb resides (6). In yeast, the Gtr heterodimer is similarly anchored to the vacuole by the orthologous counterpart of the Ragulator complex composed of Ego1, Ego2, and Ego3, which, together with the Gtr heterodimer, constitute the EGO complex (5, 7, 8). The EGO complex mediates activation of TORC1 in response to leucine and perhaps other amino acids. Although yeast TORC1 also directly binds to and is thereby activated by GTP-bound Gtr1 (5), unlike mammalian TORC1, most of yeast TORC1 constitutively remains on the vacuolar membrane even upon amino acid or nitrogen starvation (5, 9, 10). In addition, the orthologue of Rheb, Rhb1, is not involved in the regulation of TORC1 in yeast (5; our unpublished observations). Therefore, although the EGO complex constitutes a TORC1-activating mechanism similar to the mammalian Ragulator-Rag complex, it remains unknown how vacuolar localization of TORC1 is related to the control of TORC1 activity. In addition, TORC1 in Rag/Gtr-depleted cells remains activated in response to amino acids, particularly glutamine, in mammals and yeast (11,–13). These observations suggest the existence of Rag/Gtr complex-independent sensing machinery, which specifically senses glutamine, although the molecular mechanism remains elusive.
Investigating the nutrient-sensing mechanism upstream of TORC1 through in vivo experiments is troublesome because which amino acid or metabolite directly activates TORC1 is difficult to determine. Even if a single amino acid is supplemented, it will not directly be reflected at the intracellular level of the amino acid, as any amino acid can be metabolized, sequestered, and consumed in the cell. Furthermore, the intracellular amino acid level is directly affected by the efficiency of amino acid uptake. In addition, TORC1 itself regulates amino acid uptake by controlling the expression and plasma membrane localization of amino acid transporters (14,–16). These problems must be solved if amino acid-responsive activation of TORC1 is to be reproduced by in vitro experiments, as the effects of metabolism and uptake of amino acids can be excluded. Moreover, an in vitro experiment will enable examination of the effects on TORC1 activation of chemicals that cannot permeate the cell or conditions that are lethal to the cell. However, although in vitro TORC1 kinase assays have been conducted by many groups, the amino acid-responsive activation of TORC1 has not been successfully reproduced regardless of the experimental system or organism. The fact that immunopurified TORC1 cannot be activated by adding amino acids suggests that TORC1 itself does not have the ability to respond to amino acids.
In this study, we observed amino acid-responsive activation of TORC1 for the first time, using permeabilized yeast cells or purified vacuoles as TORC1 sources. Using this assay system, we revealed that the predicted glutamine-specific sensing machinery, which is independent of the EGO complex, resides on vacuolar membranes. We also show that the phosphoinositol (PI) 3-kinase complex Vps34-Vps15 and the FYVE domain-containing protein Pib2 are involved in the sensing machinery.
TORC1 itself is unlikely to have the ability to respond to amino acids. As TORC1 localizes in cellular membrane compartments, particularly the vacuole, we expected that membrane-resident factors would be crucial for amino acid responsiveness (17, 18). Thus, we tried to use permeabilized yeast cells as a TORC1 source for an in vitro kinase assay. Here, we called the permeabilized cells semi-intact cells in line with previous studies (19, 20). The semi-intact cells were prepared by partially digesting the cell wall, perforating the plasma membrane, and washing out the soluble cytoplasmic components. We employed Thr37/46 phosphorylation of 4EBP1, which is one of the best-characterized mammalian TORC1 substrates, as the endpoint of TORC1 activity. 4EBP1 has been used for in vitro yeast TORC1 kinase assays as a substrate (21,–23). After adding ATP, the 4EBP1 phosphorylation level increased with time (Fig. 1A). Importantly, the 4EBP1 phosphorylation level increased about twofold after the addition of a physiological level of l-glutamine (14 mM) (Fig. 1B). Both the basal and glutamine-activated kinase activity observed were attributable to TORC1, as 4EBP1 phosphorylation was suppressed by rapamycin (TORC1 inhibitor) or torin-1 (TORC1 and TORC2 inhibitor) (Fig. 1B). TORC2 did not contribute to 4EBP1 phosphorylation, as the semi-intact cells prepared from a TORC2-deficient strain (avo3Δ) showed l-glutamine-responsive 4EBP1 phosphorylation similar to that of wild-type cells (Fig. 1C). These results indicate that TORC1 is responsible for the phosphorylation of 4EBP1 and that adding l-glutamine induced the activation of TORC1 in vitro. 4EBP1 phosphorylation level increased in proportion to glutamine concentration under our experimental conditions (Fig. 1D). Importantly, TORC1 activity did not respond to d-glutamine (Fig. 1E), which strongly suggests that the activation of TORC1 by l-glutamine in vitro was not caused by nonspecific effects of adding glutamine, such as a change in pH or ionic strength, but was caused by a specific interaction of l-glutamine with unidentified physiological sensing machinery.
Next, we tested whether other amino acids could activate TORC1 in vitro. Among all amino acids soluble at concentrations compatible with the assay, only l-glutamine and l-cysteine activated TORC1, whereas only l-aspartic acid inhibited TORC1 (Fig. 1F). In contrast to the previous in vivo observations (5, 11), l-leucine did not activate TORC1 in vitro. d-Cysteine did not activate TORC1 in vitro (Fig. 1G), indicating that l-cysteine also specifically activates TORC1 like l-glutamine. On the other hand, d-aspartic acid inhibited TORC1 in vitro (Fig. 1G), suggesting that aspartic acid is likely to inhibit TORC1 by nonspecific effects. Taken together, these results suggest that the amino acid-sensing machinery in semi-intact cells is highly specific to l-glutamine and l-cysteine.
TORC1 localizes to the vacuolar membrane in yeast (5, 17), and purified vacuoles undergo fragmentation in a TORC1 activity-dependent manner (24); thus, it is possible that purified vacuoles contain all of the necessary components for the glutamine-responsive TORC1 activation observed under the conditions described above. To test this possibility, we purified vacuoles by density gradient centrifugation and performed an in vitro kinase assay using purified vacuoles as the TORC1 source. Indeed, purified vacuoles exhibited l-glutamine-responsive activation of TORC1 similar to semi-intact cells (Fig. 1H). This result suggests that all components necessary for glutamine responsiveness reside on the vacuolar membranes.
The amino acid preference for TORC1 activation in vitro is different from that in vivo. In our in vitro experiments, only glutamine and cysteine activated TORC1 (Fig. 1F), whereas various amino acids activated TORC1 in vivo, which was explained by their ability to activate the EGO complex (11). Hence, we examined the relationship between in vitro TORC1 activation and the EGO complex. TORC1 activity of semi-intact cells prepared from a gtr1Δ mutant exhibited glutamine responsiveness similar to those prepared from wild-type cells (Fig. 2A). Furthermore, vacuoles purified from gtr1Δ or ego3Δ cells also exhibited glutamine responsiveness (Fig. 2B and andC).C). These results suggest that our newly developed assay captured the glutamine-sensing machinery independent of the EGO complex. Our results also suggested that TORC1 localized to the vacuolar membranes in an EGO complex-independent manner, although the EGO complex was reported to anchor TORC1 to the vacuolar membrane (5, 9). Actually, we confirmed that vacuolar localization of Kog1-GFP (Kog1 labeled with green fluorescent protein [GFP]) as well as Tor1-GFP was retained in gtr1Δ, ego1Δ, or ego3Δ cells (Fig. 2B to toEE and data not shown).
These results suggest that the EGO complex-independent machinery resides on the vacuolar membranes; therefore, we tested whether the membrane potential is required to activate the machinery. Treating semi-intact cells with the proton uncoupling reagents carbonyl cyanide m-chlorophenylhydrazone (CCCP) and nigericin, which inhibit the activity of amino acid transporters that reside in vacuolar membranes, and nystatin, which forms pores in the vacuolar membrane by binding to ergosterol (25), did not impede the glutamine responsiveness of TORC1 (Fig. 3). These results indicate that the membrane potential and accumulation of glutamine in the vacuolar lumen are not necessary for glutamine responsiveness and suggest that glutamine sensing occurs on the outer surface of the vacuolar membrane and not in the lumen of the vacuole.
Next, we tried to identify the molecules involved in glutamine responsiveness. First, we tested known TORC1 regulators. The SEA complex, which functions as GAP for Gtr1, was not necessary for glutamine-responsive activation of TORC1 in vitro, as Iml1, Npr2, Npr3, Seh1, and Sea2, which are components of the SEA complex, were dispensable for glutamine responsiveness (data not shown). This observation is consistent with the finding that glutamine responsiveness is independent of the EGO complex. In addition, Tco89, which is a nonessential TORC1 component, was also dispensable for glutamine responsiveness (data not shown). These results suggest that unidentified molecules activate TORC1 in response to glutamine. Hence, we tried to identify the molecules by screening for genes whose disruption impairs glutamine-responsive TORC1 activation among genes that genetically interact with TORC1 components or whose products localize to vacuoles. The screening was performed by preparing semi-intact cells from the corresponding knockout collection strains and testing their glutamine responsiveness, one by one.
Among the tested candidate genes, semi-intact cells prepared from the pib2Δ and vps34Δ strains exhibited impaired glutamine responsiveness (Fig. 4A). The impairment was observed over a wide range of ATP concentration (Fig. 4B). In addition, purified vacuoles from pib2Δ and vps34Δ cells also lost glutamine responsiveness, although vps34Δ vacuoles showed somewhat higher background activities for unknown reasons (Fig. 4C). These results suggest that Pib2 and Vps34 are involved in glutamine-responsive TORC1 activation. One possibility for the role of Pib2 and Vps34 in the activation of TORC1 is that they are required for vacuolar localization of TORC1. However, Kog1-GFP and Tor1-GFP were observed in vacuoles of pib2Δ and vps34Δ cells by using a microscope (Fig. 4D and andE).E). In addition, the amount of TORC1 in vacuoles purified from pib2Δ cells was comparable to that from wild-type cells (Fig. 4F). These results indicate that Pib2 and Vps34 do not control the vacuolar localization of TORC1. The decreased TORC1 level in vacuoles purified from vps34Δ cells was probably due to the pleiotropic effects of vps34Δ, as the vacuolar marker protein Vph1 was similarly decreased (Fig. 4F).
Vps34 is the only phosphatidylinositol (PI) 3-kinase in yeast that synthesizes PI-3-phosphate [PI(3)P]. Vps34 is recruited to the membrane through interaction with the membrane-tethered protein kinase Vps15 (26). Vps34 has roles in autophagy and vacuolar protein sorting after forming two different multimeric complexes called complex Ι and II, respectively (27). Complex Ι-deficient cells (atg14Δ), complex II-deficient cells (vps38Δ), and cells deficient in both complexes (vps30Δ) exhibited intact in vitro glutamine responsiveness (Fig. 5A). In contrast, vps15Δ cells lost glutamine responsiveness (Fig. 5B). The marginal TORC1 activation observed in vps34Δ cells in Fig. 5B may be due to residual Pib2 on the vacuole, although the significance is currently unclear. These results indicate that the Vps34-Vps15 heterodimer is required but that Vps34 complexes Ι and II are not required for glutamine-responsive TORC1 activation. The existence of an alternative Vps34-Vps15 complex distinct from complex I or II has been proposed, although its molecular composition and physiological function have been elusive (27). Our results show that one function of the alternative Vps34-Vps15 complex is to activate TORC1 in an EGO-independent glutamine-responsive manner. The role of Vps34-Vps15 in TORC1 activation is probably to recruit Pib2 to the vacuole, as Pib2 has a PI(3)P-interacting FYVE domain, localizes on the vacuolar membrane through the domain, and delocalizes to the cytoplasm in vps34Δ cells (28).
Next, we tested whether Pib2 physically interacts with TORC1 by coimmunoprecipitation experiments using purified vacuoles. Purified vacuoles were incubated with l- or d-glutamine and then treated with the chemical cross-linker 3,3′-dithiobis(sulfosuccinimidylpropionate) (DTSSP) prior to immunoprecipitation to capture transient or unstable interaction on the vacuolar membranes. Tor1, as well as Kog1, was coprecipitated with Pib2 (Fig. 6). Remarkably, l-glutamine, but not d-glutamine, enhanced Pib2-TORC1 interaction (Fig. 6). This specificity for l-glutamine is in good agreement with the results of the in vitro kinase assay that l-glutamine, but not d-glutamine, activates TORC1 (Fig. 1E).
Finally, we tested the role of Pib2 and Vps34 during amino acid-responsive TORC1 activation in vivo. The phosphorylation status of Sch9-T737, which is a well-defined TORC1 substrate (29), was examined to monitor activation of TORC1 in vivo. After 40 min of nitrogen starvation, the cells were stimulated with glutamine or arginine for the indicated times. Nitrogen starvation of wild-type cells deactivated TORC1, whereas adding glutamine or arginine immediately reactivated TORC1, but TORC1 activity decreased subsequently and then increased again, as reported previously (11) (Fig. 7). Adding glutamine to gtr1Δ and pib2Δ cells activated TORC1; however, activation was lower than that in wild-type cells, and no clear oscillation was observed (Fig. 7). Adding arginine activated TORC1 in pib2Δ cells, albeit less than wild-type cells after 20 min, but not in gtr1Δ cells (Fig. 7). Leucine stimulation also activated TORC1 in pib2Δ cells but not in gtr1Δ cells (data not shown). These results suggest that glutamine activates TORC1 via both the Pib2- and the EGO complex-dependent machinery, but arginine and leucine activate TORC1 only through the EGO complex-dependent machinery until they are metabolized to produce glutamine. These observations are consistent with our data from in vitro experiments showing that Pib2 is involved in glutamine-specific sensing machinery independent of the EGO complex.
The vps34Δ cell phenotype was more severe than that of pib2Δ cells. In vps34Δ cells, glutamine- as well as arginine-responsive activation of TORC1 was greatly impaired (Fig. 7), which may reflect the multiple roles of Vps34 besides control of the localization of Pib2 in amino acid sensing. In fact, vps34Δ cells had a severe growth defect, whereas pib2Δ cells did not (data not shown).
In this study, we reproduced amino acid-responsive activation of TORC1 kinase activity in vitro for the first time, using semi-intact cells and purified vacuoles as TORC1 sources. The activation was specific for l-glutamine and l-cysteine. The intracellular cysteine level (0.7 to 1.4 mM) is much lower than that of glutamine (15 to 35 mM) or the cysteine concentration employed in this in vitro study (17 mM) (30,–33). On the other hand, the glutamine concentration used in this in vitro study (0.14 to 55 mM) is comparable to the intracellular glutamine level (32). Therefore, glutamine probably plays a more dominant role than cysteine in the regulation of TORC1 in vivo under physiological conditions.
We also showed that glutamine-responsive activation of TORC1 in vitro was independent of the EGO complex, which is in agreement with a previous report that stimulation of yeast cells with high-end amino acids, which increase the intracellular level of glutamine, results in sustained TORC1 activation in an EGO complex-independent manner (11). Our in vitro assay did not show response to amino acids other than glutamine or cysteine. This is probably because some essential components of the EGO complex-dependent amino acid-sensing mechanism were lost during preparation of semi-intact cells or vacuolar purification, and thereby, only the EGO complex-independent mechanism remained.
In this study, we also attempted to identify the molecules involved in the EGO complex-independent glutamine sensing and found that Vps34-Vps15 and Pib2 are necessary for glutamine responsiveness in vitro and in vivo. In line with our findings, Pib2 was recently reported to be involved in activating TORC1 through an EGO complex-independent pathway (28). As Pib2 has a PI(3)P-interacting FYVE domain near the C terminus, localizes on the vacuolar membrane through the domain, and delocalizes to the cytoplasm in vps34Δ cells (28), it is conceivable that the role of Vps34-Vps15 in the glutamine responsiveness of TORC1 is to recruit Pib2 to the vacuolar membrane.
The Pib2-mediated TORC1 activation machinery primarily responded to glutamine. Monitoring the cellular level of glutamine as an indicator of nitrogen nutrient status seems reasonable, as glutamine is a nitrogen donor during nucleotide and amino acid biosynthesis, and glutamine replenishes tricarboxylic acid cycle intermediates (34). When cells accelerate anabolic processes to grow, they need higher levels of nitrogen and energy to synthesize proteins and nucleotides that can be converted from glutamine. Interestingly, some cancer cell lines are addicted to glutamine, indicating that the level of glutamine is a bottleneck for high proliferation under certain conditions (35).
There was a report that PI 3,5-bisphosphate [PI(3,5)P2] directly binds to Kog1 and Sch9 and is necessary for the intact vacuolar localization of Kog1 and Sch9 (36). Therefore, we tested whether PI(3,5)P2 is necessary for TORC1 glutamine responsiveness in vitro. Fab1, which is the sole PI(3)P5-kinase in yeast, as well as Vac7 and Vac17, which are positive regulators of Fab1, was dispensable for glutamine-responsive activation of TORC1 in vitro (data not shown). This indicates that PI(3,5)P2 is not necessary for glutamine responsiveness and that the role of Vps34-Vps15 in vitro is not to provide a precursor of PI(3,5)P2. Furthermore, Kog1-GFP in vps34Δ cells was localized on the vacuolar membrane (Fig. 4D), suggesting that the contribution of the PI(3,5)P2-Kog1 interaction to vacuolar localization of TORC1 is limited.
It remains controversial whether Vps34 is involved in amino acid-responsive TORC1 activation in higher eukaryotes. Vps34 is necessary for the TORC1 amino acid responsiveness in mammals (37,–39). The involvement of Vps34 in Drosophila melanogaster is contradicted by the observation that the Vps34 null mutant does not have altered TORC1 activity (40). However, it is now clear that TORC1 is activated by several upstream pathways, so the disruption of a single pathway might be insufficient to monitor the contribution of the pathway in activating TORC1.
Vps34 may regulate TORC1 via different mechanisms in yeast and mammals. In mammals, Vps34 is suggested to activate TORC1 by recruiting phospholipase D (PLD) to the lysosomal membrane (37,–39). Lysosome-recruited PLD generates phosphatidic acid (PA) on the membranes, which is required for TORC1 activation. On the other hand, Spo14, which is the sole orthologue of PLD in yeast, was dispensable for the glutamine responsiveness of TORC1 in vitro (data not shown), indicating that Vps34 in yeast does not regulate the glutamine responsiveness of TORC1 via PLD activity.
We observed that TORC1 maintained vacuolar localization in pib2Δ cells (Fig. 4D, ,E,E, and andF).F). Furthermore, on the purified vacuolar membranes, l-glutamine, but not d-glutamine, induced Pib2-TORC1 interaction as well as TORC1 activation (Fig. 6 and and1E).1E). These results suggest that Pib2 is directly involved in glutamine-responsive TORC1 activation through its interaction with TORC1 rather than controlling vacuolar localization of TORC1.
Pib2 resides on the outer surfaces of vacuolar membranes via interactions between its FYVE domain and vacuolar PI(3)P (28), suggesting that the glutamine-sensing machinery is located and functions on the outer surface of the vacuolar membrane. Consistent with this hypothesis, neither CCCP nor nigericin, which are ionophores that inhibit amino acid transporters that reside in vacuolar membranes (41, 42), inhibited TORC1 activation in vitro (Fig. 3A). In addition, making pores in vacuolar membranes by treatment with nystatin did not impede TORC1 activation in vitro (Fig. 3B). These observations suggest that the uptake of glutamine into the vacuolar lumen is not necessary for sensing glutamine and support the hypothesis that sensing of glutamine occurs on the outer surfaces of vacuolar membranes.
Our present study demonstrated the existence of previously uncharacterized glutamine sensing machinery, in which vacuolar Pib2 is involved. Our current model for the EGO complex-independent glutamine-responsive TORC1 activation mechanism is as follows. Pib2 and TORC1 constitutively reside on the vacuolar membrane. Cytoplasmic glutamine is sensed by an unidentified sensor on the vacuolar membrane. The sensing event induces association between Pib2 and TORC1, which, in turn, somewhat promotes activation of TORC1. However, the identity of the glutamine sensor and how the sensor transmits the signal to TORC1 via Pib2 remain to be clarified. In addition, whether homologous machinery exists in higher eukaryotes remains unknown, although LAPF/phafin1 has been suggested to be a Pib2 orthologue in mammals (28). Further study is necessary to resolve these issues.
YEp352-YPK2(D239A)-HA was a kind gift from Y. Kamada (43). pTS200 (p416-TOR1D330-3×GFP) was constructed via gap repair cloning from strain VA66 (a kind gift from M. N. Hall ). The yeast strains used are listed in Table 1. All strains used were in the S288C background except where noted otherwise. Strain MH971 was constructed by crossing strain WW135 with strain TS270. Strains MH1008, MH1010, MH1023, MH1028, MH1059, MH1061, and MH1064 were constructed by replacing EGO1 or EGO3 of strain TS013, VPS34 of TM141 or TS171, and PIB2 of TM141, TS171, or MH1038 with the corresponding kanMX4 cassettes that were amplified by PCR using genomic DNA of the corresponding knockout collection strains as the templates. MH1016, MH1019, and MH1017 were constructed by crossing TS890 with MH1010 and TS890 with MH1008. MH974 and MH980 were constructed by replacing GTR1 of TS270 and AVO3 of TS270 carrying the YPD2(D239A)-expressing multicopy plasmid YEp352-YPK2(D239A)-HA with the hphNT1 cassette that was amplified by PCR as described previously (44). MH1036 and MH1038 were constructed as follows. TS559 and TM101 were mated, VPS34 of the resultant diploid strain was replaced with a kanMX4 cassette, the obtained strain was sporulated, and the tetrad was dissected.
4EBP1, carbonyl cyanide m-chlorophenylhydrazone (CCCP), nigericin, and nystatin were purchased from Sigma-Aldrich (St. Louis, MO). Rapamycin was purchased from LC Laboratories (Woburn, MA). Torin-1 was obtained from ChemScene LCC (Monmouth Junction, NJ). FM4-64 was purchased from Molecular Probes-Invitrogen (Eugene, OR). The antihemagglutinin (anti-HA) (12CA5) antibody was kindly provided by Y. Hoshikawa. The anti-Flag M2 antibody was purchased from Sigma-Aldrich. The anti-Tor1 antibody yN-15 was from Santa Cruz Biotechnology (Dallas, TX). The anti-Vph1 antibody was from Abcam (Cambridge, UK). The anti-phospho-4EBP1 (T37/46) antibody 9459 and anti-4EBP1 antibody 9452 were from Cell Signaling Technology (Danvers, MA). The anti-glutathione S-transferase (anti-GST) antibody 4C10 was purchased from Abcam (Cambridge, UK). The anti-phospho-Sch9 (T737) antibody was described previously (45). IRdye800-conjugated anti-rabbit antibody was from Rockland (Limerick, PA), and Alexa Fluor 680-conjugated anti-mouse antibody was from Thermo Fisher Scientific (Waltham, MA).
Spheroplasts were prepared as described previously (20) with a slight modification as follows. Logarithmically growing cells in 50 ml of yeast extract-peptone-dextrose (YPD) medium were harvested by centrifugation and washed with MilliQ water containing 2 mM phenylmethylsulfonyl fluoride (PMSF). The cells were suspended in 1 ml of 0.1 M Tris-HCl (pH 9.4) with 10 mM dithiothreitol (DTT) and incubated for 10 min at room temperature. Then, the cells were collected by centrifugation, suspended in 1 ml of spheroplasting buffer (0.7 M sorbitol, 10 mM Tris-HCl [pH 7.5], 1 mM DTT, 20 mM NaN3, and 0.1 mg/ml Zymolyase-100T), and incubated for 20 to 30 min at room temperature. The spheroplasted cells were centrifuged at 1,000 × g for 2 min at 4°C. Then, the cells were washed once with 1 ml of cold sorbitol buffer (1 M sorbitol, 150 mM K acetate, 5 mM Mg acetate, and 20 mM HEPES-KOH [pH 6.6]), resuspended in the same buffer to an optical density at 600 nm (OD600) of 100/ml, and stored at −80°C in small aliquots. Semi-intact cells were prepared from the spheroplasts by washing out the cytosol as described previously (20).
Vacuoles were isolated as described previously (24) with the following minor changes. Cells were cultured in 500 ml of YPD medium to an OD600 of 1.2 to 1.8. The cells were collected by centrifugation, suspended in 20 ml of 0.1 M Tris-HCl (pH 9.4) with 10 mM DTT, and incubated for 7 min at 30°C. Then, the cells were collected by centrifugation, resuspended in 15 ml of SB (50 mM K phosphate [pH 7.5], 600 mM sorbitol, 0.16× YPD, and 1 mg/ml Zymolyase-100T), and incubated for 20 to 30 min at 30°C. The cells were collected by centrifugation, washed with 15 ml of SB, and resuspended in 2.5 ml of 15% Ficoll 400 in PS buffer [10 mM piperazine-N,N′-bis(2-ethanesulfonic acid) (PIPES)–KOH (pH 6.8), 200 mM sorbitol, 1 mM phenylmethylsulfonyl fluoride (PMSF), 40 μg/ml aprotinin, 10 μg/ml pepstatin A, and 20 μg/ml leupeptin]. After 300 μl of freshly made 0.4-mg/ml DEAE-dextran in 15% Ficoll PS buffer was added and gently mixed, the tubes were placed on ice for 2 min. Then, the tubes were incubated for 90 s in a 30°C water bath and placed on ice again. The cell lysate was transferred into a polyallomer centrifuge tube (14 by 89 mm) (Beckman-Coulter, Fullerton, CA) and overlaid with 3 ml of 8% Ficoll in PS buffer, 3 ml of 4% Ficoll in PS buffer, and 1.5 ml of PS buffer. After 75-min centrifugation at 150,000 × g in an SW40 Ti rotor (Beckman-Coulter) at 4°C, the vacuoles were harvested from the interphase between 0 and 4% Ficoll.
Semi-intact cells were resuspended in import buffer (0.4 M sorbitol, 150 mM K acetate, 5 mM Mg acetate, and 20 mM HEPES-KOH [pH 6.6]). In a typical experiment, semi-intact cells corresponding to OD600 of 0.7 × ml or 5 μg of vacuoles were suspended in 18 μl of reaction buffer (0.4 M sorbitol, 150 mM K acetate, 5 mM Mg acetate, 20 mM HEPES-KOH [pH 6.6], 40 mM creatine phosphate, 200 ng/μl creatine kinase, 1 mM Pefabloc SC, 4 μg/ml aprotinin, 1 μg/ml pepstatin A, 2 μg/ml leupeptin, and 0.3 μg/ml GST-4EBP1  or 4EBP1), and the reactions were started by adding 2 μl of ATP-amino acid mix solution (5 mM ATP and 2% amino acids unless otherwise indicated) and incubated for 10 min at 30°C. TOR inhibitors and other drugs indicated were added immediately before the addition of the ATP-amino acid mix solution. Kinase reactions were stopped by adding 20 μl of 2× Laemmli sample buffer and boiled for 4 min. The samples were centrifuged at 9,000 × g for 2 min, and the supernatant was subjected to Western blotting with the anti-phospho-4EBP1 antibody and either anti-GST or anti-4EBP1 antibody. IRdye800-conjugated anti-rabbit antibody and Alexa Fluor 680-conjugated anti-mouse antibody were used as secondary antibodies. Signals were detected using the Odyssey infrared imaging system (LI-COR, Lincoln, NE) according to the manufacturer's instructions. Statistical analyses were performed using a Student t test.
Logarithmically growing cells in YPD or synthetic complete medium lacking uracil (SC-Ura) (0.17% yeast nitrogen base, 0.5% ammonium sulfate, 2% glucose, and 0.13% uracil dropout powder ) to preserve the plasmids were collected, resuspended in YPD containing 32 nM FM4-64, and incubated for 20 min at 30°C. Then, the cells were washed once with YPD, resuspended in YPD, and incubated for 70 min at 30°C. The cells were observed microscopically after they were washed in SC medium. Images were acquired with the Olympus FV1200 microscope (Tokyo, Japan) with a 100× oil immersion objective.
Vacuoles were isolated as described above from logarithmically growing TS270 or MH1100 cells. About 200 μg of the isolated vacuoles was suspended in 135 μl of import buffer, and 15 μl of ATP-amino acid mix solution (5 mM ATP and 2% amino acids) was added and incubated for 5 min at 30°C. Then, 17 μl of 25 mM 3,3′-dithiobis(sulfosuccinimidylpropionate) (DTSSP) (Thermo Fisher Scientific) was added to the solution, and the mixture was incubated for 2 h on ice. The cross-linking reaction was stopped by adding 5 μl of 1 M Tris-HCl (pH 7.5), and the mixture was incubated for 15 min on ice. The membranes were solubilized by adding 15 μl of 1% SDS, 1.1 ml of immunoprecipitation (IP) buffer (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1 mM EDTA, 0.5% NP-40, 1 mM PMSF, 40 μg/ml aprotinin, 10 μg/ml pepstatin A, and 20 μg/ml leupeptin) was added, and the mixture was further incubated for 15 min on ice. After 10-min centrifugation at 21,800 × g at 4°C, the supernatant was subjected to immunoprecipitation using the anti-HA antibody-conjugated agarose beads (catalog no. A2095; Sigma-Aldrich). The beads were washed once with IP buffer and twice with wash buffer (20 mM Tris-HCl [pH 7.5], 1 M NaCl, 0.1% NP-40), then suspended in Laemmli sample buffer, and incubated for 15 min at 65°C to reverse the cross-link. The supernatant was subjected to Western blotting using the anti-myc antibody 9E10, the anti-HA antibody (Y-11), or the anti-FLAG M2 antibody.
Samples were prepared as described previously (45) with the following minor modifications. Prototrophic cells for the amino acids and nucleotide bases were prepared by introducing pRS416 and pTS44 (45). The cells were pregrown overnight in SC lacking uracil and tryptophan (SC-Ura Trp) medium, diluted in SC-Ura Trp medium to an OD600 of 0.2 to 0.35, and grown to an OD600 of 0.6 to 1.0. The cells were washed once with nitrogen-free medium, resuspended, and incubated in nitrogen-free medium for 40 min for starvation of nitrogen sources. Glutamine, arginine, or leucine (each 0.2%) was added from 15-fold stock solutions, and the cells were incubated for the indicated times. Trichloroacetic acid (final concentration, 6%) was added to the cell cultures, and the mixture was kept on ice for 15 min. Cells collected by centrifugation were suspended in urea buffer (50 mM Tris-HCl [pH 7.5], 5 mM EDTA, 6 M urea, 1% SDS, 50 mM NaF, 10 mM β-glycerophosphate, and 1 mM PMSF), disrupted by vortexing with glass beads, and incubated at 65°C for 10 min. The cell lysates were cleared by centrifugation, and total protein concentration was determined using the DC protein assay (Bio-Rad, Hercules, CA). The cell lysates were incubated with Laemmli SDS sample buffer at 65°C for 15 min and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting with anti-phospho-T737-Sch9 and anti-HA (12CA5) antibodies.
We thank Michael N. Hall, Yoshiaki Kamada, and Yutaka Hoshikawa for reagents used in this study, Terunao Takahara for his constructive comments on the manuscript, Takeshi Noda for sharing unpublished data, and all members of Maeda lab and Hiroyuki Tachikawa lab for help and discussions. We also thank the University of Tokyo IMCB Olympus Bioimaging Center (TOBIC) for access to the Olympus FV1200 microscope.
This work was supported in part by Japan Society for the Promotion of Science (JSPS) KAKENHI grants 25291042 and 17H03802 and grants from the Uehara Memorial Foundation and from the Noda Institute for Scientific Research (all to T.M.).