Urmylation Controls Nil1p and Gln3p-Dependent Expression of Nitrogen-Catabolite Repressed Genes in Saccharomyces Cerevisiae

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

FEBS Letters 581 (2007) 541–550

Urmylation controls Nil1p and Gln3p-dependent expression of nitrogen-catabolite repressed genes in Saccharomyces cerevisiae

Marta Rubio-Texeira Whitehead Institute, Massachusetts Institute of Technology, 9 Cambridge Center 653, Cambridge, MA 02142, United States

Received 10 November 2006; revised 22 December 2006; accepted 9 January 2007

Available online 18 January 2007

Edited by Horst Feldmann

TORC1 promotes cell growth by positively regulating protein Abstract Urm1 is a modifier protein that is conjugated to substrate proteins through thioester formation with the E1-like synthesis and other anabolic processes, and by negatively reg- enzyme, Uba4. Here is shown that the lack of urmylation causes ulating mechanisms of stress-response and catabolic processes, derepression of the GAP1 gene (encoding a nitrogen-regulated such as autophagy and the expression of genes involved in the broad-spectrum amino acid-scavenging permease) in the presence utilization of lower-quality nutrients. TORC2 seem to primar- of rich nitrogen sources, and simultaneous inhibition of the ily regulate spatial aspects of cell growth such as actin/cyto- expression of CIT2, a TCA-cycle gene involved in the production skeleton organization and cell polarity [1,2]. Although of glutamate and glutamine. This effect is dependent on the sharing several components, one of the most striking differ- TORC1- and nutrient-regulated transcriptional factors, Nil1p ences between these two complexes is that TORC1 activity is and Gln3p. Evidence is provided that, in the absence of urmyla- sensitive to rapamycin whereas TORC2 is not [6]. tion, nuclear/cytosolic shuffling of both transcriptional factors is New components of the TOR-dependent regulatory path- altered, ultimately leading to inability to repress GAP1 gene in the presence of a rich nitrogen source. Altogether, the data pre- ways are continuously being identified. In particular, recent sented here indicate an important role of the urmylation pathway studies have suggested a potential link between the yet poorly in regulating the expression of genes involved in sensing and con- understood role of a newly identified ubiquitin (Ub)-like con- trolling amino acids levels. jugation system, urmylation, and the TOR pathways [7]. 2007 Federation of European Biochemical Societies. URM1 (URM, for Ub-related modifier) is one of the five dif- Published by Elsevier B.V. All rights reserved. ferent Ub-like modifier proteins (Ubls) until date identified in yeast [8]. Urmylation has a unique resemblance to ATP-depen- Keywords: Urmylation; GAP1; CIT2; GLN3; NIL1 dent cofactor sulfuration reactions involved in the prokaryotic synthesis of thiamin and molibdopterin, which has provided an important clue for the origin of ATP-dependent protein conjugation systems [9]. Although the nature of the putative targets for this type of regulation remains obscure, URM1 must have 1. Introduction an important role in eukaryotic cells since this gene has been evolutionarily conserved from yeast to mammalians, including Sensing of the amount and quality of nutrients available in humans [9,10]. Only one substrate for urmylation has until the environment is crucial for cell survival and has been found now been identified in yeast: Ahp1, a protein with a role in oxi- to be largely controlled in eukaryotic cells by the molecular dative-stress response [11]. This finding potentially links urmy- target of rapamycin, TOR (latest reviewed in [1,2]). Rapamy- lation to the control of stress responses. Moreover, additional cin is an antitumour and immunosuppressant macrocyclic work from Sprague Jr. and collaborators has shown that lactone that has the ability to bind the peptidylisoprolyl isom- defects in urmylation lead to rapamycin sensitivity that can be erase FK506-binding protein (FKBP12) encoded in yeast by rescued by the rapamycin-resistant allele TOR2-1, and has also FPR1 [3,4]. When bound to this protein, rapamycin causes reported interactions between this protein-modification system inactivation of TOR proteins which in turn triggers a series and invasive/pseudohyphal growth, which is usually elicited by of events within cells that closely mimic those elicited in prep- nutrient starvation conditions such as nitrogen limitation [7]. aration for survival and adaptation to different types of stress The quality and quantity of nitrogen sources are carefully and nutrient starvation [1,2]. Yeasts have two separate TOR assessed by yeast cells through monitoring of the relative levels genes, TOR1 and TOR2, but only TOR2 is found to be essen- of intracellular glutamine (Gln) and glutamate (Glu) [12,13]. tial [5]. Tor proteins belong to a group of kinases known as the Gln and Glu are central precursors for the synthesis of tricar- phosphatidylinositol kinase-related kinase (PIKK) family, boxylic acids (TCA) cycle intermediates, their relative levels characterized by the presence of a carboxy-terminal serine/ playing a key role in the generation of energy as well as in ana- threonine protein kinase domain that resembles the catalytic bolic reactions, such as the biosynthesis of amino acids and domain of phosphatidylinositol 3-kinases (PI3Ks) and PI4Ks other nitrogen-containing metabolites [12]. Hence, Gln and [2]. In yeast, Tor proteins control different aspects of cell Glu are the preferred nitrogen environmental sources for cells. growth by forming part of two distinct, although partially External absence of these two nitrogen sources and/or reduced redundant, protein complexes (TORC1 and TORC2; [1,2]). levels of amino acids in the environment trigger degradation of most of the highly-specific amino acid permeases, and promote the synthesis and sorting to the plasma membrane of lower- E-mail address: [email protected] (M. Rubio-Texeira). quality nitrogen-scavenging permeases such as the general

0014-5793/$32.00 2007 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2007.01.019 542 M. Rubio-Texeira / FEBS Letters 581 (2007) 541–550 amino acid permease, Gap1p, a broad-spectrum amino acid sources, the upregulation of certain genes regulated by Nil1p permease able to internalize all naturally occurring amino triggered by an intermediate-quality nitrogen source such as acids [14–18]. Lower-quality nitrogen sources internalized by ammonia, also favor the production of glutamine from gluta- Gap1p are then used to synthesize Glu and Gln. mate. Genes directly and indirectly involved in the synthesis of TORC1 is inactivated in the presence of rapamycin because Glu/Gln are tightly regulated transcriptionally and post-trans- glutamine production, which posits the signal for nitrogen ex- criptionally. In a simple picture, enzymatic activities acting in cess, is inhibited so that such signal no longer reaches Tor [23]. the earlier steps of the TCA cycle are mainly responsible for Inactivation of TORC1 by rapamycin or in response to nutri- the generation of glutamate and glutamine. CIT2 encodes a ent starvation causes Gln3p dissociation from Ure2p, which in citrate synthase involved in the first catalytic step for the gen- turn allows Gln3p entry to the nucleus, where it activates the eration of a-ketoglutarate from acetyl-CoA [12]. GDH1 en- expression of nitrogen-catabolite repressed genes [27]. Interest- codes a NADP+-linked glutamate dehydrogenase that uses ingly, the lack of Ure2p alters the pattern of urmylated ammonia derived from the internalized nitrogen sources to substrates in a Gln3p-dependent manner, suggesting that convert a-ketoglutarate into glutamate, and GLN1 encodes a urmylation may be used as an additional modification system glutamine synthetase, which also uses ammonia to further con- to regulate proteins related to the Gln3p–Ure2p pathway [7]. vert glutamate into glutamine [12]. GAP1 is positively correg- In the present work, evidence is presented that urmylation ulated with GDH1 and GLN1 to couple the internalization of affects nuclear-cytosolic localization of the transcriptional fac- low-quality nitrogen sources through Gap1p with their use in tors Nil1p and Gln3p, thereby constituting an additional level the production of Glu and Gln. Increasing levels of Glu and of regulation in the expression of genes involved in the synthe- Gln ultimately lead to downregulation of the expression of sis of Glu and Gln. This finding provides further evidence these genes and activity of their products through a negative showing that urmylation has an important role in the control feedback loop [12,13,19]. of nutrient sensing in TOR-related pathways. Genes involved in the earliest steps of the TCA cycle, such as CIT2, are positively regulated by the Rtg 1/3 transcription factors, and negatively regulated by Mks 1p (in combination 2. Materials and methods with Bmh1, and Bmh2), and the TOR-associated protein Lst8, in response to the intracellular levels of Glu/Gln (retro- 2.1. Strains, plasmids, and media grade (RTG) pathway; reviewed in [20]). The transcription of All the yeast strains used in this study (listed in Table 1) are in the S288C background, which displays high Gap1p expression and activity GAP1, GDH1, and GLN1, is controlled by the two GATA when ammonia is the nitrogen source [28]. All complete gene deletions transcriptional factors, Nil1p and Gln3p [19,21–23]. Gln3p is described here were obtained by replacement of the functional ORF of repressed on glutamine or ammonia medium, whereas Nil1p the corresponding gene by homologous recombination with either a is repressed on glutamate or glutamine medium [12]. Gln3p- kanMX4/6 or a natMX4 cassette [29,30] in the wild type strain CKY835 [18] or in strains derived from the same genetic background. dependent transcription is repressed by a physical interaction Plasmids used in this study are pMS29, a PGAP1-LacZ fusion at with the negative regulator Ure2p, which sequesters Gln3p in codon 53 of GAP1 in the URA3-CEN vector pBL101 [31]; pEC261, the cytosol when glutamine is present [24,25]. A similar regula- a PCIT2-LacZ fusion after codon 3 of CIT2 in pRS316 [32]; and tion exerted on Nil1p has not yet been demonstrated. Recent pMRT14, which contains the constitutive promoter PADHI followed evidence has shown that although the sets of genes responsive by 3·FLAG sequence, URM1 ORF fused in frame to 3·FLAG, and kanMX6 marker cloned in XbaI/EcoRI sites present in the multi clon- to Nil1p and Gln3p regulation overlap, the activation of Nil1p ing site of pRS316. is mainly aimed at the generation of energy via production of Minimal medium was prepared using yeast nitrogen base without TCA cycle intermediates from the precursor glutamate, amino acids and without ammonium sulfate (Difco, Detroit, MI), whereas the activation of Gln3p is involved in the collection 2% glucose, and a nitrogen source: 0.1% glutamine, 0.1% glutamate, 0.5% ammonium sulfate, or urea 0.2%. All growth experiments were of alternative nitrogen sources to synthesize glutamine [26]. carried out at 24 C except when indicated. For the rapamycin treat- Although the main role of Nil1p seems to be the maintenance ment, YPD standard plates were prepared containing 10 or 50 nM of proper levels of glutamate in the presence of poor carbon of rapamycin (Sigma-Aldrich, St. Louis, MO).

Table 1 Yeast strains Strain Genotype Reference CKY835 MATa ura3-52 [18] MRTY305 MATa urm1D::kanMX4 ura3-52 This study MRTY306 MATa uba4D::kanMX4 ura3-52 This study CKY834 MATa GAP1-GFP::kanMX6 [18] MRTY382 MATa urmlD::kanMX4 GAP1-GFP::kanMX6 ura3-52 This study MRTY383 MATa uba4D::kanMX4 GAP1-GFP::kanMX6 ura3-52 This study CKY777 MATa ure2D::kanMX6 ura3-52 [32] CKY778 MATa g1n3D::kanMX6 [32] CKY888 MATa nil1D::kanMX6 ura3-52 C.A. Kaiser MRTY378 MATa urmlD::kanMX4 nil1D::kanMX6 ura3-52 This study MRTY380 MATa urmlD::kanMX4 ure2D::kanMX6 ura3-52 This study MRTY384 MATa urmlD::kanMX4 gln3D::kanMX6 ura3-52 This study MRTY404 MATa NIL1-HA::kanMX6 ura3-52 This study MRTY405 MATa GLN3-myc::kanMX6 ura3-52 This study MRTY406 MATa urmlD::kanMX4 NIL1-HA::kanMX6 ura3-52 This study MRTY407 MATa urmlD::kanMX4 GLN3-myc::kanMX6 ura3-52 This study M. Rubio-Texeira / FEBS Letters 581 (2007) 541–550 543

2.2. Assays for b-galactosidase reporter expression and amino acid leads to the accumulation of a 131-kDa Urm1p-modified spe- uptake cies that, interestingly, disappears when the GLN3 gene is b-Galactosidase activity was measured by the permeabilized cell simultaneously knocked-out [7]. method [33]. Assays of the rate of uptake of radio labeled amino acids were performed as described previously [34]. In order to ascertain whether the urmylation pathway further affects the signaling mediated through these proteins, 2.3. Fluorescence microscopy and immunofluorescence the level of expression of the Gln3p-dependent gene GAP1 Immunofluorescence was performed using standard protocols [33] was analyzed in the presence of null mutations in UBA4 or with the modifications described previously [32]. Antibodies used were URM1. Expression of TCA cycle enzymes is regulated by the mouse anti-HA monoclonal 16B12 (BabCO, Richmond, CA); mouse transcription factors Rtg1/3p, which in turn are negatively by anti-myc, monoclonal 9E10 (Santa Cruz Biotechnology, Santa Cruz, CA); and Alexa 488-conjugated goat anti-mouse IgG (Molecular Mkslp and the TOR-associated Lst8p in response to high lev- Probes, Eugene, OR). els of Glu/Gln [20,32]. Since misregulation in the levels of For GFP localization studies, Gap1p-GFP expressing cells were Gln3p-dependent genes caused by the lack of urmylation grown to exponential phase in liquid cultures with ammonia or gluta- may lead to altered levels of the TCA cycle enzymes due to mine as nitrogen sources. Cells were washed and suspended in 1 M alteration of Glu/Gln levels, the expression of CIT2 (involved Tris, pH 8.0, 5% NaN3 as described previously [18]. Images were cap- tured with a Nikon E800 microscope (Melville, NY) equipped with a in the generation of a-ketoglutarate, the carbon skeleton from Hamamatsu digital camera (Bridgewater, NJ). The chroma filter set which Glu/Gln are derived) was also analyzed. The pattern of at 41012 was used for GFP detection. Improvision OpenLabs 2.0 soft- expression of CIT2 and GAP1 genes was analyzed through ware (Lexington, MA) was used to process images. quantification of the levels of b-galactosidase activity in cells expressing the PCIT2-LacZ or PGAP1-LacZ reporters (Fig. 1). The lack of urmylation genes led to dramatic increases (6–7- 3. Results fold) in the levels of GAP1 expression in conditions in which this gene should be totally or partially repressed, such as in 3.1. Deletion of genes involved in the urmylation pathway alter cells growing in the presence of glutamine or glutamate (the the pattern of expression of GAP1 and CIT2 genes in latter mostly in the case of urm1D) as only nitrogen source. different nitrogen sources The levels were also notably increased (3-fold) in cells grown Sprague Jr. and colleagues found that the lack of Ure2p, the in ammonia as the only nitrogen source. The expression of TOR-regulated repressor of the transcriptional factor Gln3p, GAP1 was not altered when growth conditions in which the

CIT2 GAP1 A 800 160 700 140 600 120 500 100 expression

expression 400

80 (% of wt) LacZ

- 300 (% of wt) 60 LacZ - 200 40 GAP1 CIT2 P 100

P 20 0 0 WT uba4Δ urm1Δ WT uba4Δ urm1Δ B 140 250 120 200 100 80 150 expression expression 60

(% of wt) 100 LacZ LacZ (% of wt) - 40 -

CIT2 50 GAP1 P 20 P 0 0 WT uba4Δ urm1Δ WT uba4Δ urm1Δ C 140 350 120 300 100 250 80 200 expression expression 60 150 (% of wt)

LacZ (% of wt) LacZ - 40 - 100

CIT2

GAP1 P 20 50 P 0 0 WT uba4Δ urm1Δ WT uba4Δ urm1Δ

Fig. 1. Lack of urmylation pathway causes derepression of GAP1 and hyper-repression of CIT2. (Right panel) Wild-type (CKY835), urmlD (MRTY305), and uba4D (MRTY306) strains were transformed with PGAP1-LacZ (pMS29) and assayed for b-galactosidase activity after growth at 24 C in minimal medium with glutamine (A), glutamate (B), or ammonia (C) as nitrogen source. (Left panel) The same strains were transformed with PCIT2-LacZ (pEC261) and assayed for b-galactosidase activity after growth within the same conditions as in right panel. 544 M. Rubio-Texeira / FEBS Letters 581 (2007) 541–550

1400 GAP1 evidence indicate that cells tightly sense reductions in the quality of the nitrogen source, responding to this perturbation by 1200 synthesizing a certain pool of the permease, in order to prepare 1000 themselves for an eventual lack of nutrients. Once in the pres-

800 ence of poor nitrogen sources and/or low levels of amino acids,

expression Gap1p is synthesized and targeted primarily to the plasma 600 membrane to display its activity.

LacZ (% of wt) - 400 The lack of urmylation could aﬀect the regulation of Gap1p

GAP1

P activity not only at the level of expression but also at the level 200 of sorting. Changes in the level of Gap1p activity caused by 0 alterations on its sorting may also indirectly affect the internal WT WT WT urm1Δ Gln shift Urea Gln levels of Glu/Gln and amino acids [13]. In order to test this Rap possibility, the levels of Gap1p activity were analyzed by mea- Fig. 2. Comparative levels of GAP1 expression between wild type and suring the uptake of radiolabeled citrulline, which specifically urmlD mutant cells. Wild-type (CKY835), and urmlD (MRTY305) cells enters cells through Gap1p (as previously described, [34]). expressing b-galactosidase from PGAP1-LacZ (pMS29) were grown to No significant changes were detected in the uba4D and urm1D exponential phase at 24 C in minimal medium with glutamine, and mutants respect to the wild type strain on any nitrogen source assayed for b-galactosidase activity. Levels of GAP1 expression between the wild type and the mutant are compared to those of wild (glutamine, glutamate, ammonia, or urea; data not shown). type cells treated for 30 min with 200 nM rapamycin after growth in The localization of genomically expressed Gap1p-GFP in glutamine, or wild type cells grown to exponential phase in minimal wild type, uba4D and urm1D cells growing on either glutamine medium with urea as nitrogen source. (no expression conditions) or ammonia (expression and sorting to plasma membrane) was also analyzed. Whereas wild type cells show practically no GFP signal for cells growing on glu- gene is normally induced (e.g. presence of poor nitrogen tamine, uba4D and urm1D cells displayed intense GFP signal sources such as urea) were used (data not shown). GAP1 levels localized at the vacuolar lumen (Fig. 3). The pattern in ammo- of expression in an urm1D mutant exponentially growing in nia looked more similar to that of the wild type strain with glutamine were also 3-fold higher than those of wild type most of the protein being localized at the plasma membrane. cells growing in the same nitrogen source after being exposed However, an important amount of signal was also observed to 200 nM rapamycin for 30 min to 1 h period (Fig. 2). The at the vacuolar lumen of the two mutants, reflecting an excess urm1D levels were 60% of those present in wild type cells of the protein being synthesized and targeted to the vacuole. growing exponentially in urea as only nitrogen source. Overall, These results are in agreement with the previous observa- these results indicate that the lack of urmylation totally or par- tions made for the expression of GAP1 gene in urmylation mu- tially abrogates the repression of Gln3/Nil1p-transcribed genes tants growing in these different nitrogen sources (Figs. 1 and like GAP1 in the presence of rich- to intermediate-quality 2). It appears that, without urmylation pathway, cells are nitrogen sources. unable to repress the expression of GAP1 but that regulation Expression of CIT2 was analyzed in identical conditions as at the level of sorting remains unaltered so that the excess of for GAP1 expression (Fig. 1), showing instead 10-fold repres- the protein being synthesized in the presence of rich/intermedi- sion over wild type levels in the presence of glutamine and ate nitrogen sources in urmylation mutants is still targeted for ammonia, or glutamate in the case of urm1D. In glutamate, vacuolar degradation. the repression was not as dependent on the lack of UBA4 as it was on that of URM1. Similarly, derepression of GAP1 in the uba4D mutant was less pronounced than in urm1D. This may suggest the existence of alternative proteins involved in ammonia glutamine the conjugation of Urm1p to substrates under such conditions. As seen for GAP1, when the expression of CIT2 was tested in WT poorer nitrogen sources, no changes were observed over those of a wild type strain (data not shown). The simultaneous derepression of the GAP1 gene and hyper- repression of the CIT2 gene observed in mutants impaired for urm1Δ urmylation grown in glutamine indicates that the urmylation pathway is necessary to transduce the signal that maintains genes for the use of alternative nitrogen sources repressed in the presence of a rich nitrogen source. uba4Δ

3.2. Regulation of Gap1p expression but not sorting is altered in Gap1p- DIC Gap1p- DIC cells with defective urmylation pathway GFP GFP GAP1 is regulated both transcriptionally and post-transcriptionally, particularly at the level of sorting [12,17,18]. In the Fig. 3. Gaplp-GFP is normally sorted but abnormally expressed in presence of glutamine, the expression of GAP1 is almost response to the nitrogen source in the absence of urmylation. Fluorescence microscopy images of wild type cells (CKY834) and totally repressed. In the presence of glutamate, expression of the mutants urmlD (MRTY382), and uba4D (MRTY383), expressing GAP1 takes place, although, under this condition, Gap1p is genomic GFP-tagged Gap1p and exponentially growing in minimal primarily targeted for vacuolar sorting and degradation. Most ammonia or glutamine medium. M. Rubio-Texeira / FEBS Letters 581 (2007) 541–550 545

3.3. Epistatic relationship between URM1 and genes involved in [26,36]). However, this growth rescue effect was not as pro- the expression of nitrogen-catabolite repressed genes nounced as that of a gln3D in higher (lethal) concentrations Binding of rapamycin to a protein of poorly understood of rapamycin (data not shown). The lack of nil1D caused a function, FKBP12, causes inhibition of TORC1 [3]. Among slight growth rescue in the double mutant nil1D urm1D when other effects, sensitivity to rapamycin in rich medium derives compared to urm1D but this effect was less pronounced than from the unregulated expression of genes involved in the utili- that caused by the simultaneous loss of Gln3 function. zation of alternative nutrient sources, which should otherwise Although a certain degree of overlap exists, Nil1p and Gln3p be repressed by the TOR kinases through negative control of control functionally different sets of genes [26]. This result then their transcriptional factors. The deletion of GLN3 therefore shows that, overall, unregulated expression of the subset of confers partial rapamycin resistance whereas the deletion of Gln3p-dependent genes has a more important contribution its negative regulator URE2 confers hypersensitivity to rapa- in the level of sensitivity to rapamycin in rich medium than mycin [35]. It has recently been shown that the lack of urmyla- that of Nil1p-dependent genes. tion also causes sensitivity to rapamycin [7]. This sensitivity The same single and double mutant strains were also ana- can be rescued by overexpression of a rapamycin-resistant al- lyzed for their patterns of P-galactosidase activity expression lele of TOR2, which suggests that the sensitivity to rapamycin from the PCIT2-LacZ or PGAP1-LacZ reporters when cell cul- conferred by defects in urmylation is dependent upon an tures were grown in the presence of glutamine, glutamate, or upstream signal that is mediated by a certain type of TOR ammonia as only nitrogen source (Fig. 5). In glutamine, complex. either the lack of NIL1 or GLN3 genes reduces GAP1 expres- As shown above, the lack of urmylation significantly dere- sion below wild type levels indicating that, even in the pres- pressed the gene GAP1 in the presence of a rich nitrogen ence of the richest nitrogen source, there must be a certain source. To ascertain whether the sensitivity to rapamycin level of basal expression that is dependent upon both tran- caused by the lack of urmylation is dependent upon the activ- scriptional factors (Fig. 5A). As expected, lack of the ity of Gln3p and Nil1p, transcription factors for GAP1 and URE2 repressor abrogates repression of the GAP1 gene in other nitrogen-responsive genes [31],anurm1D mutation was glutamine. In accordance with previous observations [12,31], combined with a mutation in ure2D, nil1D,orgln3D, and the most GAP1 expression in glutamate was primarily dependent behavior of the different strains in the presence of a range of on the presence of Gln3p, whereas in ammonia the major sublethal concentrations of rapamycin was analyzed (Fig. 4). contributor to GAP1 expression is Nil1p. NIL1 expression Interestingly, the lack of the Ure2p repressor further enhanced is indirectly downregulated by Gln3p through Gln3p-depen- the sensitivity of the urm1D mutant. However, the deletion of dent expression of its negative regulator Dal80p [12]. This GLN3 in an urm1D mutant resulted epistatic to the mutation may explain why the deletion of Gln3p further induces in URM1, rescuing the ability of this mutant to grow in the Nil1p-dependent GAP1 expression in ammonia (Fig. 5C). In presence of rapamycin. This result suggests that the lack of glutamate, although most of the expression is dependent on regulation via urmylation causes sensitivity to rapamycin by Gln3p, a certain contribution of Nil1p must also exist, since a mechanism that depends on active Gln3p. A single nil1D a nil1D mutant displays lower levels of expression than the mutation also caused a certain level of resistance to sublethal wild type. However, NIL1 expression is also positively regu- concentrations of rapamycin, as previously reported (Fig. 4; lated by Gln3p so that, when GLN3 is deleted, Nil1p-dependent expression of GAP1 in glutamate can no longer be detected and GAP1 expression is abrogated ([12]; Fig. 5B). YPD Rapamycin Rapamycin Although the overall level of GAP1 expression in glutamine 10 nM 50 nM is about 10–20-fold lower than in ammonia [31], it seems that the major contributor to the basal levels of expression of Wild type GAP1 under such conditions is, as in ammonia, NIL1 gln3Δ (Fig. 5A). Very low levels of DAL80 under these conditions nil1Δ may, at least partially, account for this particular effect, although under such conditions autoactivation of NIL1 Δ ure2 expression is still restricted by the constitutive repressor Nil2p urm1Δ [12]. The derepression of GAP1 caused by the lack of URM1 on urm1Δ gln3Δ either of the three nitrogen sources is completely dependent on urm1Δ nil1Δ the presence of Nil1p and Gln3p since the double mutants urm1Δ ure2Δ nil1D urm1D and gln3D urm1D displayed reductions in the levels of GAP1 expression on each of the nitrogen sources that Δ + urm1 [URM1] were very similar to those caused by the lack of either NIL1 or GLN3 alone. The only exception was in ammonia Fig. 4. Simultaneous deletion of GLN3 and NIL1 reduces sensitivity of (Fig. 5C), where the double mutant gln3D urm1D had much urmylation mutants to rapamycin. A concentration of 0.5 OD600/ml and two subsequent 10-fold serial dilutions of cells exponentially higher levels of GAP1 expression than the gln3D or urm1D sin- growing in minimal medium with ammonia as nitrogen source from the gle mutants. These levels were comparable to those of an ure2D wild-type (CKY835), gln3D (CKY778), nil1D (CKY888), ure2D urm1D strain. This might again be related to the lack of (CKY777), urm1D (MRTY305), urm1D gln3D (MRTY384), urm1D DAL80 negative regulation of NIL1 expression in a gln3D nil1D (MRTY378), urm1D ure2D (MRTY380), and urm1D (MRTY305, expressing wild-type URM1 from pMRT14) strains, were spotted onto background, which may additionally contribute to the effect YPD or YPD + 10 nM or 50 nM (sublethal concentrations) of caused by the lack of negative regulation that is normally ex- rapamycin-containing plates and incubated at 30 C for three days. erted via URM1. 546 M. Rubio-Texeira / FEBS Letters 581 (2007) 541–550

GAP1 CIT2 A 10000 D 300

250 1000 200

expression 150 100 expression (% of wt) LacZ (% of wt) 100 LacZ - - 10

CIT2 50 GAP1 P P 1 0 WT uba4Δ urm1Δ nil1Δ gln3Δ ure2Δ urm1Δ urm1Δ urm1Δ WT uba4Δ urm1Δ nil1Δ gln3Δ ure2Δ urm1Δ urm1Δ urm1Δ nil1Δ gln3Δ ure2Δ nil1Δ gln3Δ ure2Δ

B 450 E 300 400 250 350 300 200

expression 250 expression 150 200 (% of wt) LacZ (% of wt) LacZ - 150 - 100

CIT2 GAP1 100

P 50 P 50 0 0 WT uba4Δ urm1Δ nil1Δ gln3Δ ure2Δ urm1Δ urm1Δ urm1Δ WT uba4Δ urm1Δ nil1Δ gln3Δ ure2Δ urm1Δ urm1Δ urm1Δ nil1Δ gln3Δ ure2Δ nil1Δ gln3Δ ure2Δ

C 10000 F 160 140 1000 120 100 expression

100 expression 80 60 (% of wt)

LacZ (% of wt)

- LacZ - 10 40

GAP1 CIT2 P P 20

1 0 WT uba4Δ urm1Δ nil1Δ gln3Δ ure2Δ urm1Δ urm1Δ urm1Δ WT uba4Δ urm1Δ nil1Δ gln3Δ ure2Δ urm1Δ urm1Δ urm1Δ nil1Δ gln3Δ ure2Δ nil1Δ gln3Δ ure2Δ

Fig. 5. Lack of the transcriptional factors Nil1p and Gln3p counteract the eﬀect of urmylation mutations in the expression of GAP1 and have an epistatic eﬀect in the expression of CIT2. Wild-type (CKY835), urm1D (MRTY305), uba4D (MRTY306) ure2D (CKY777), gln3D (CKY778), nil1D (CKY888), urm1D nil1D (MRTY378), urm1D g1n3D (MRTY384), and urm1D ure2D (MRTY380) strains were transformed with PGAP1-LacZ (pMS29) and assayed for b-galactosidase activity after growth at 24 C in minimal medium with glutamine (A), glutamate (B), or ammonia (C) as nitrogen source. The same experiment was carried out with the strains transformed with PCIT2-LacZ (pEC261). (D)–(F), respectively designate growth in glutamine, glutamate, or ammonia.

Interestingly, in all three nitrogen sources the double mutant a reduction in the expression of CIT2 is observed in this ure2D urm1D shows higher levels of GAP1 expression than mutant control to which the effect of urm1D closely resembles. either of the two single mutants, ure2D or urm1D. The double A reduction in the levels of CIT2 expression in the ure2D mutant ure2D urm1D also shows the highest level of sensitivity mutant grown in glutamine was not observed. Growth in glu- to rapamycin (Fig. 4). This additive effect points to urmylation tamine may already provide with sufficient levels of internal acting as a parallel signal rather than exerting its repressive ac- Glu/Gln so that effects in the localization of Gln3p derived tion directly through Ure2p. from the lack of Ure2p do not further influence intracellular When the levels of CIT2 expression were analyzed and as ex- levels of Glu/Gln in a way that can be detected through pected from the previous observations made (Fig. 5D–F), the changes in the levels of CIT2 expression. However, is interest- effects were in average opposite to those observed on GAP1 ing to note that this is not the case in an urm1D ure2D double expression. Lack of Nil1p or Gln3p caused significant in- mutant, in which CIT2 remains as repressed as in urm1D. This creases of CIT2 expression, as observed for the gln3D mutant epistatic effect of the mutation urm1D over the mutation ure2D, in glutamate (conditions in which GDH1 and GLN1 are mostly along with data shown in Fig. 4, further supports the idea of controlled by Gln3p) and for the nil1D mutant in ammonia urmylation acting in a parallel rather than in an Ure2p-depen- (conditions in which GDH1 and GLN1 are mostly controlled dent pathway. Lack of urmylation could cause more desensiti- by Nil1p), or both mutants in glutamine. In order to maintain zation to the repressible nitrogen signal at the level of Gln3p/ the levels of Glu/Gln, an enhancement of a-ketoglutarate pro- Nil1p-dependent gene expression than lack of Ure2p, in such duction is very likely to occur to compensate for the reduction way that the Gdh1/Gln1-dependent pathway for Glu/Gln pro- in the levels of expression of GDH1 and GLN1 caused in the duction would improperly be turned on even though in the absence of either of these GATA transcription factors. The presence of external glutamine as only nitrogen source. lack of ure2D enhances the expression of GDH1 and GLN1 be- Interestingly, the lack of URM1 was epistatic to the lack of cause Gln3p nuclear localization is no longer impaired, there- either Gln3p in glutamate or Nil1p in ammonia, since the dou- by increasing the production of Glu/Gln [12]. For such reason, ble mutants showed levels of CIT2 expression below those of M. Rubio-Texeira / FEBS Letters 581 (2007) 541–550 547

A B CD

urm1Δ

WT, Rap

Gln3-myc DAPI Gln3-myc DAPI Nil1-HA DAPI Nil1-HA DAPI

Fig. 6. Lack of urmylation alters the localization pattern for Gln3p and Nil1p in different nitrogen sources. The wild-type (MRTY405), and urmlD (MRTY407) strains expressing integrated GLN3-myc, or the wild-type (MRTY404), and urmlD (MRTY406) strains expressing integrated NIL1-HA, were grown in ammonia (A or C), or glutamate (B or D) medium at 24 C. A final concentration of 200 nM of rapamycin was added to the indicated sample for 30 min before fixation of cells for immunofluorescence. All images are shown at the same magnification.

the wild type strain (Fig. 5E and F). This effect could be ex- 4. Discussion plained if the lack of urmylation caused derepression of Nil1p activity in cells grown in glutamate and derepression of Gln3p In this study, evidence is presented that urmylation acts as a activity in cells grown in ammonia. For example, if Nil1p nu- regulatory signal in the control of the expression of nitrogen- clear levels were improperly increased in glutamate, because of sensitive genes. Furthermore, the results presented here clearly the lack of URM1, that would override induction of CIT2 even indicate that this posttranslational regulation is required for in the absence of Gln3p. Similarly, increased localization of nitrogen-regulated intracellular localization of the transcrip- Gln3p in the nucleus in urmylation defective cells growing in tional factors Nil1p and Gln3p. ammonia, would also override the induction of CIT2 even The molecular function of the URM1 pathway as well as the though in the absence of Nil1p. identity of the Urm1p-conjugates remains to be uncovered. Sprague Jr. and colleagues have shown evidence that this path- 3.4. Nitrogen-regulated nuclear localization of Nil1p and Gln3p way has implications in vegetative and invasive growth in depends on urmylation S. cerevisiae [5]. Furthermore, they showed that loss of the One possibility to explain all the results observed above is urmylation pathway causes cells to be sensitive to rapamycin that the control of the nuclear-cytosolic shuffling of Nil1p and that such sensitivity is overridden by expression of the rap- and Gln3p may be dependent upon urmylation. In order to amycin-resistant allele TOR2-1, which suggests a connection discern this possibility, immunofluorescence of strains express- between this type of protein modification and the TOR signaling genomically tagged Gln3p-myc or Nil1p-HA was analyzed ing pathway. Intriguingly, in this same work they also showed in glutamate (conditions of nuclear localization for Gln3p and that the pattern of urmylated species is altered in the presence cytosolic localization of Nil1p) or ammonia (conditions of nu- of different combinations of the ure2D and gln3D mutant back- clear localization for Nil1p and cytosolic localization of grounds. Gln3p) to see whether under these conditions Gln3p and Nil1p To investigate a potential role of urmylation in the control were improperly localized. of nitrogen-regulated genes, the pattern of expression of two Wild type cells exposed to ammonia showed nuclear locali- genes involved in amino acid biosynthesis, GAP1 and CIT2, zation of Nil1p-HA and cytosolic localization of Gln3p-myc, was analyzed. Interestingly, the elimination of the urmylation and only when simultaneously exposed to rapamycin did they pathway leads to a loss of nitrogen-dependent regulation of the show Gln3p-myc also localized in the nucleus (Fig. 6A and C). expression of these two genes. Under such conditions, GAP1 In contrast, urm1D cells in the presence of ammonia showed expression (which is co-regulated with that of GDH1 and nuclear localization not only of Nil1p-HA but also of GLN1; [12]) becomes derepressed in the presence of a rich Gln3p-myc (Fig. 6A and C). Similarly, while wild type cells nitrogen source, whereas CIT2 expression becomes simulta- grown in glutamate showed nuclear localization of Gln3p- neously hyper-repressed. This phenomenon is consistent with myc and Nil1p-HA only localized in this compartment when a simultaneous derepression of GDH1 and GLN1 leading to cells were exposed to rapamycin, the mutant urm1D cells also an increased production of Glu/Gln, which in turn would turn showed an intensified signal of Nil1p-HA localized within off the expression of genes acting upstream in the pathway for the nucleus (Fig. 6B and D). Glu/Gln biosynthesis, such as CIT2. This result directly correlates urmylation to the control of Further experiments shown here demonstrate that the effect nuclear-cytosolic shuffling of the transcriptional factors Nil1p in the expression of GAP1 and CIT2 observed in the absence of and Gln3p in response to the quality of the nitrogen source. urmylation is dependent on functionally active Gln3p and 548 M. Rubio-Texeira / FEBS Letters 581 (2007) 541–550

Nil1p. Previous studies have shown that the expression of phenomenon is in fact caused because, in npr1D mutants, GAP1, GDH1, and GLN1, is mostly dependent upon Gln3p ammonia transport is downregulated by ubiquitination and in the presence of glutamate or upon Nil1p in the presence consequent vacuolar sorting of Mep ammonia transporters of ammonia (reviewed in [12]). Interestingly, sensitivity to rap- due to the lack of Npr1p-dependent phosphorylation of the amycin caused by defects in urmylation is largely abrogated by nitrogen-scavenging permeases [46,47]. All these evidence the lack of functional Gln3p and moderately abrogated by the makes it less likely for Npr1 and/or the ubiquitination machin- lack of Nil1p. This epistatic effect of the GLN3 and NIL1 null ery to be the direct targets of urmylation responsible for the mutations over that of URM1 deletion suggests that urmyla- control of nuclear translocation of the transcriptional factors tion normally protects cells from sensitivity to rapamycin in Gln3p and Nil1p, although the possibility remains to be tested. a background of unregulated Gln3- and Nil1-dependent Interestingly, an increasing amount of data is showing that expression, likely by working upstream from these activities. components of the TORC kinase complexes (including the Furthermore, here is shown that the effect observed in the Tor kinases themselves; [48]) are most likely acting in associa- absence of urmylation results from a loss of nitrogen-depen- tion to membranes. For example, Gln3p nuclear translocation dent control of the intracellular localization of Nil1p and in response to the nitrogen signal, but not in response to rap- Gln3p. Nuclear-cytosolic shuffling of these two transcriptional amycin, requires an intact cytoskeleton, which has led to the factors is regulated in response to nutrients. In the presence of suggestion that the Ure2p-Gln3p complex may be in fact asso- a rich nitrogen source, Gln3p is trapped in the cytosol through ciated to a vesicular complex [49,50]. Members of the recently the formation of a complex with the repressor protein Ure2 discovered EGO (exit from rapamycin-induced growth arrest) [24]. Nutrient signaling mediated through the TOR kinases is complex, acting as putative signal inducers for TORC1 and as thought to alter the pattern of phosphorylation of both Gln3 a recycling complex specifically involved in Gap1p retrieval and Ure2 proteins, such that Gln3p is released from its inter- from the vacuolar targeting pathway, exert their function in action with Ure2p and able to migrate into the nucleus where association to membranes [51,52]. In the absence of Pmr1, a it activates gene expression. Whether Nil1p establishes a simi- Golgi Ca2+/Mn2+-ATPase, Npr1 function seems to be upregu- lar interaction with Ure2p is still the object of discussion. One lated, phenomenon that places yet another membrane compo- possibility is that Gln3p, and/or Nil1p, are regulated by being nent as a regulator of a TOR-related pathway [53]. These and direct substrates for urmylation. Unfortunately, immunopre- other related findings suggest that association of the main cipitation of Nil1p-HA and Gln3p-myc and further analysis members of the TOR pathway with vesicular compartments by Western blot using anti-FLAG to detect a functional N- may in fact be the way in which cells are able to couple sensing terminal FLAG-tagged version of Urm1p (expressed from of the nutrient signal with immediate regulation of both the the ADH1 promoter in the CEN-URA3 plasmid, pMRT14) gene expression and protein sorting of the different functions failed to detect urmylated forms of the transcriptional factors, involved in nutrient detection and uptake. excluding this possibility (data not shown). Taking the latter facts into account, urmylation could either Another possibility is that urmylation affects nuclear trans- affect components of the signal at the level of nuclear-cytosolic location of Gln3 and Nil1 proteins through modification of shuffling as mentioned above, and/or alternatively affect the other proteins involved. For instance, nuclear import and status of proteins associated to the membrane-bound TOR export of Gln3p has been shown to be dependent on the and interacting protein complexes. For instance, one of the lat- yeast karyopherins a/Srp1p (importin) and Crm1p (exportin) ter possibilities could be that urmylation regulates the interac- [37,38]. Coupling of the exportin with Ran-GTP within the tion of the transcriptional factors with a repressor, such as the nucleus promotes shuffling back to the cytosol. Until now, interaction of Ure2p that keeps Gln3p inactivated. Since it is evidence has shown that phosphorylation plays a key role in known that Gln3–Ure2p also interacts physically with TORC1 the control of the binding of the transcriptional factors to these [54], the activation of Gln3p could be somehow a membrane karyopherins. However, different studies have shown a lack of (e.g. perinuclear ER)-associated process regulated by urmyla- correlation between Gln3p levels of phosphorylation and its tion. Although by different mechanisms, examples of transcrip- nuclear translocation [39–41]. An additional level of regulation tional factors being regulated through their association to could, therefore, involve nitrogen-sensitive urmylation. Other membranes have previously been described [55]. In spite of Ubl systems such as sumoylation are known to play important being an attractive hypothesis, this latter possibility is not sup- roles in the regulation of nuclear import/export events [8]. ported by the epistatic analysis between the urm1D and ure2D Alternatively, urmylation could modify the activity of other mutations shown in Figs. 4 and 5. The effects of both muta- putative regulators of Gln3- and Nil1-regulated genes, such as tions in GAP1 and CIT2 expression seem to be rather additive, Npr1p. Npr1p is a Ser/Thr kinase known to positively regulate suggesting that urmylation is unlikely to be a direct effector at the level of sorting the activity of low-quality nitrogen perm- regulating the interaction between Gln3 and Ure2 proteins. eases such as Gap1p and the ammonia permeases encoded by An additional alternative is that urmylation could perhaps the MEP genes [42,43]. Npr1p-dependent phosphorylation more directly affect the synthesis of Glu/Gln through control prevents Rsp5/Bul1/Bul2 ubiquitin-E3 ligase complex-depen- of proteins such as Lst8, Mks1, and/or the transcriptional fac- dent ubiquitination of these permeases, which would otherwise tors Rtg1p and Rtg3p [20,32]. This possibility is also not highly target them for vacuolar degradation [17,42,44]. Based primar- supported by the data shown here, because there seems to be a ily on the finding that Gln3p is nuclear in an npr1D growing on more direct effect in the localization of Gln3p and Nil1p. How- ammonia, it was recently proposed that Npr1p negatively reg- ever, since changes in the levels of activity of all these regula- ulates Gln3p nuclear translocation and that Rsp5/Bul1/Bul2 tors would have potential effects in the intracellular levels of antagonizes this effect [45]. However, later evidence has shown Glu/Gln, and these could affect the localization of these tran- that the Npr1 kinase effect on Gln3p is rather indirect and scriptional factors, this is a possibility that should also be restricted to the use of ammonia as a nitrogen source. This tested. M. Rubio-Texeira / FEBS Letters 581 (2007) 541–550 549

Whether urmylation affects either of the steps proposed cellular trafficking of the general amino acid permease. J. Cell above deserves indeed further investigation. Given the numer- Biol. 153, 649–662. ous possibilities now unveiled, future efforts in the identifica- [18] Rubio-Texeira, M. and Kaiser, C.A. (2006) Amino acids regulate retrieval of the yeast general amino acid permease from the tion of new substrates for urmylation will indeed prove vacuolar targeting pathway. Mol. Biol. Cell 17, 3031–3050. useful to elucidate the role of this poorly understood modifica- [19] Daugherty, J.R., Rai, R., el Berry, H.M. and Cooper, T.G. (1993) tion system in nutrient signaling processes. Regulatory circuit for responses of nitrogen catabolic gene expression to the GLN3 and DAL80 proteins and nitrogen Acknowledgements: Preliminary stages of this work were carried out in catabolite repression in Saccharomyces cerevisiae. J. Bacteriol. the laboratory of Dr. Chris A. Kaiser at the Biology Department of 175, 64–73. MIT and supported by a National Institute of General Medical Sci- [20] Butow, R.A. and Avadhani, N.G. (2004) Mitochondrial signaling: ences Grant (GM56933). I thank the Kaiser lab for gifts of plasmids, the retrograde response. Mol. Cell 14, 1–15. antibodies, and strains. [21] Stanbrough, M., Rowen, D.W. and Magasanik, B. (1995) Role of the GATA factors Gln3p and Nil1p of Saccharomyces cerevisiae in the expression of nitrogen regulated genes. Proc. Natl. Acad. Sci. USA 92, 9450–9454. References [22] Coffman, J.A., Rai, R., Cunningham, T., Svetlov, V. and Cooper, T.G. (1996) Gat1p, a GATA family protein whose production is [1] De Virgilio, C. and Loewith, R. (2006) Cell growth control: little sensitive to nitrogen catabolite repression, participates in tran- eukaryotes make big contributions. Oncogene 25, 6392–6415. scriptional activation of nitrogen-catabolic genes in Saccharomy- [2] Wullschleger, S., Loewith, R. and Hall, M.N. (2006) TOR ces cerevisiae. Mol. Cell Biol. 16, 847–858. signaling in growth and metabolism. Cell 124, 471–484. [23] Crespo, J.L., Powers, T., Fowler, B. and Hall, M.N. (2002) The [3] Heitman, J., Movva, N.R. and Hall, M.N. (1991) Targets for cell TOR-controlled transcription activators GLN3, RTG1, and cycle arrest by the immunosuppressant rapamycin in yeast. RTG3 are regulated in response to intracellular levels of gluta- Science 253, 905–909. mine. Proc. Natl. Acad. Sci. USA 99, 6784–6789. [4] Koltin, Y., Faucette, L., Bergsma, D.J., Levy, M.A., Cafferkey, [24] Courchesne, W.E. and Magasanik, B. (1988) Regulation of R., Koser, P.L., Johnson, R.K. and Livi, G.P. (1991) Rapamycin nitrogen assimilation in Saccharomyces cerevisiae: roles of the sensitivity in Saccharomyces cerevisiae is mediated by a peptidyl- URE2 and GLN3 genes. J. Bacteriol. 170, 708–713. prolyl cis–trans isomerase related to human FK506-binding [25] Kulkarni, A.A., Abul-Hamd, A.T., Rai, R., El Berry, H. and protein. Mol. Cell. Biol. 11, 1718–1723. Cooper, T.G. (2001) Gln3p nuclear localization and interaction [5] Helliwell, S.B., Wagner, P., Kunz, J., Deuter-Reinhard, M., with Ure2p in Saccharomyces cerevisiae. J. Biol. Chem. 276, Henriquez, R. and Hall, M.N. (1994) TOR1 and TOR2 are 32136–32144. structurally and functionally similar but not identical phospha- [26] Kuruvilla, F.G., Shamji, A.F. and Schreiber, S.L. (2001) Carbon- tidylinositol kinase homologues in yeast. Mol. Biol. Cell 5, 105– and nitrogenquality signaling to translation are mediated by 118. distinct GATA-type transcription factors. Proc. Natl. Acad. Sci. [6] Loewith, R., Jacinto, E., Wullschleger, S., Lorberg, A., Crespo, USA 98, 7283–7288. J.L., Bonenfant, D., Oppliger, W., Jenoe, P. and Hall, M.N. [27] Beck, T. and Hall, M.N. (1999) The TOR signalling pathway (2002) Two TOR complexes, only one of which is rapamycin controls nuclear localization of nutrient-regulated transcription sensitive, have distinct roles in cell growth control. Mol. Cell 10, factors. Nature 402, 689–692. 457–468. [28] Courchesne, W.E. and Magasanik, B. (1983) Ammonia regulation [7] Goehring, A.S., Rivers, D.M. and Sprague Jr., G.F. (2003) Mol. of amino acid permeases in Saccharomyces cerevisiae. Mol. Cell. Biol. Cell 14, 4329–4341. Biol. 3, 672–683. [8] Hochstrasser, M. (2000) Biochemistry. All in the ubiquitin family. [29] Longtine, M.S., McKenzie, A., Demarini, D.J., Shah, N.G., Nat. Cell Biol. 2, E153–E157. Wach, A., Brachat, A., Philippsen, P. and Pringle, J.R. (1998) [9] Furukawa, K., Mizushima, N., Noda, T. and Ohsumi, Y. (2000) Additional modules for versatile and economical PCR-based gene A protein conjugation system in yeast with homology to deletion and modification in Saccharomyces cerevisiae. Yeast 14, biosynthetic enzyme reaction of prokaryotes. J. Biol. Chem. 953–961. 275, 7462–7465. [30] Goldstein, A.L. and McCusker, J.H. (1999) Three new dominant [10] Singh, S., Tonelli, M., Tyler, R.C., Bahrami, A., Lee, M.S. and drug resistance cassettes for gene disruption in Saccharomyces Markley, J.L. (2005) Three-dimensional structure of the cerevisiae. Yeast 15, 1541–1553. AAH26994.1 protein from Mus musculus, a putative eukaryotic [31] Stanbrough, M. and Magasanik, B. (1995) Transcriptional and Urm1. Protein Sci. 14, 2095–2102. posttranslational regulation of the general amino acid permease [11] Goehring, A.S., Rivers, D.M. and Sprague Jr., G.F. (2003) of Saccharomyces cerevisiae. J. Bacteriol. 177, 94–102. Attachment of the ubiquitin-related protein Urm1p to the [32] Chen, E.J. and Kaiser, C.A. (2003) LST8 negatively regulates antioxidant protein Ahp1p. Eukaryotic Cell 2, 930–936. amino acid biosynthesis as a component of the TOR pathway. J. [12] Magasanik, B. and Kaiser, C.A. (2002) Nitrogen regulation in Cell Biol. 161, 333–347. Saccharomyces cerevisiae. Gene 290, 1–18. [33] Adams, A., Gottschling, D. and Kaiser, C. (1996) Methods in [13] Chen, E.J. and Kaiser, C.A. (2002) Amino acids regulate the Yeast Genetics: a Laboratory Course Manual, Cold Spring intracellular trafficking of the general amino acid permease of Harbor Laboratory Press, New York. Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 99, 14837– [34] Roberg, K.J., Bickel, S., Rowley, N. and Kaiser, C.A. (1997) 14842. Control of amino acid permease sorting in the late secretory [14] Grenson, M., Hou, C. and Crabeel, M. (1970) Multiplicity of the pathway of Saccharomyces cerevisiae by SEC13, LST4, LST7 and amino acid permeases in Saccharomyces cerevisiae. IV. Evidence LST8. Genetics 147, 1569–1584. for a general amino acid permease. J. Bacteriol. 103, 770–777. [35] Chan, T.F., Carvalho, J., Riles, L. and Zheng, X.F. (2000) A [15] Jauniaux, J.C. and Grenson, M. (1990) GAP1, the general amino chemical genomics approach toward understanding the global acid permease gene of Saccharomyces cerevisiae. Nucleotide functions of the target of rapamycin protein (TOR). Proc. Natl. sequence, protein similarity with the other bakers yeast amino Acad. Sci. USA 97, 13227–13232. acid permeases, and nitrogen catabolite repression. Eur. J. [36] Xie, M.W., Jin, F., Hwang, H., Hwang, S., Anand, V., Duncan, Biochem. 190, 39–44. M.C. and Huang, J. (2005) Insights into TOR function and [16] Roberg, K.J., Rowley, N. and Kaiser, C.A. (1997) Physiological rapamycin response: chemical genomic profiling by using a high- regulation of membrane protein sorting late in the secretory density cell array method. Proc. Natl. Acad. Sci. USA 102, 7215– pathway of Saccharomyces cerevisiae. J. Cell Biol. 137, 1469– 7220. 1482. [37] Carvalho, J., Bertram, P.G., Wente, S.R. and Zheng, X.F. (2001) [17] Helliwell, S.B., Losko, S. and Kaiser, C.A. (2001) Components of Phosphorylation regulates the interaction between Gln3p and the a ubiquitin ligase complex specify polyubiquitination and intra- nuclear import factor Srplp. J. Biol. Chem. 276, 25359–25365. 550 M. Rubio-Texeira / FEBS Letters 581 (2007) 541–550

[38] Carvalho, J. and Zheng, X.F. (2003) Domains of Gln3p nteract- [47] Feller, A., Boeckstaens, M., Marini, A.M. and Dubois, E. (2006) ing with karyopherins, Ure2p, and the target of rapamycin Transduction of the nitrogen signal activating G1n3-mediated protein. J. Biol. Chem. 278, 16878–16886. transcription is independent of Npr1 kinase and Rsp5-Bull/2 [39] Cox, K.H., Kulkarni, A., Tate, J.J. and Cooper, T.G. (2004) Gln3 ubiquitin ligase in Saccharomyces cerevisiae. J. Biol. Chem. 281, phosphorylation and intracellular localization in nutrient limita- 28546–28554. tion and starvation differ from those generated by rapamycin [48] Kunz, J., Schneider, U., Howald, I., Schmidt, A. and Hall, M.N. inhibition of Tor 1/2 in Saccharomyces cerevisiae. J. Biol. Chem. (2000) HEAT repeats mediate plasma membrane localization of 279, 10270–10278. Tor2p in yeast. J. Biol. Chem. 275, 37011–37020. [40] Tate, J.J., Rai, R. and Cooper, T.G. (2005) Methionine sulfox- [49] Cox, K.H., Tate, J.J. and Cooper, T.G. (2004) Actin cytoskeleton imine treatment and carbon starvation elicit Snf1-independent is required for nuclear accumulation of G1n3 in response to phosphorylation of the transcription activator Gln3 in Saccharo- nitrogen limitation but not rapamycin treatment in Saccharomy- myces cerevisiae. J. Biol. Chem. 280, 27195–27204. ces cerevisiae. J. Biol. Chem. 279, 19294–19301. [41] Kulkarni, A., Buford, T.D., Rai, R. and Cooper, T.G. (2006) [50] Magasanik, B. (2005) The transduction of the nitrogen regulation Differing responses of Gat1 and Gln3 phosphorylation and signal in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA localization to rapamycin and methionine sulfoximine treatment 102, 16537–16538. in Saccharomyces cerevisiae. FEMS Yeast Res. 6, 218–229. [51] Dubouloz, F., Deloche, O., Wanke, V., Cameroni, E. and [42] Dubois, E. and Grenson, M. (1979) Methylamine/ammonia DeVirgilio, C. (2005) The TOR and EGO protein complexes uptake systems in Saccharomyces cerevisiae: multiplicity and orchestrate microautophagy in yeast. Mol. Cell 19, 15–26. regulation. Mol. Gen. Genet. 175, 67–76. [52] Gao, M. and Kaiser, C.A. (2006) A conserved GTPase-containing [43] De Craene, J.O., Soetens, O. and Andre, B. (2001) The Npr1 complex is required for intracellular sorting of the general amino- kinase controls biosynthetic and endocytic sorting of the yeast acid permease in yeast. Nat. Cell Biol. 8, 657–667. Gap1 permease. J. Biol. Chem. 276, 43939–43948. [53] Devasahayam, G., Ritz, D., Helliwell, S.B., Burke, D.J. and [44] Soetens, O., DeCraene, J.O. and Andre, B. (2001) Ubiquitin is Sturgill, T.W. (2006) Pmr1, a Golgi Ca2+/Mn2+-ATPase, is a required for sorting to the vacuole of the yeast general amino acid regulator of the target of rapamycin (TOR) signaling pathway in permease, Gap1. J. Biol. Chem. 276, 43949–43957. yeast. Proc. Natl. Acad. Sci. USA 103, 17480–17485. [45] Crespo, J.L., Helliwell, S.B., Wiederkehr, C., Demougin, P., [54] Bertram, P.G., Choi, J.H., Carvalho, J., Ai, W., Zeng, C., Chan, Fowler, B., Primig, M. and Hall, M.N. (2004) NPR1 kinase and T.F. and Zheng, X.F. (2000) Tripartite regulation of G1n3p by RSP5-BUL1/2 ubiquitin ligase control GLN3-dependent tran- TOR, Ure2p, and phosphatases. J. Biol. Chem. 275, 35727–35733. scription in Saccharomyces cerevisiae. J. Biol. Chem. 279, 37512– [55] Hoppe, T., Matuschewski, K., Rape, M., Schlenker, S., Ulrich, 37517. H.D. and Jentsch, S. (2000) Activation of a membrane-bound [46] Tate, J.J., Rai, R. and Cooper, T.G. (2006) Ammonia-specific transcription factor by regulated ubiquitin/proteasome-dependent regulation of G1n3 localization in Saccharomyces cerevisiae by processing. Cell 102, 577–586. protein kinase Npr1. J. Biol. Chem. 281, 28460–28469.