The Spermidine Transport System Is Regulated by Ligand Inactivation, Endocytosis, and by the Npr1p Ser/Thr Protein Kinase in Saccharomyces cerevisiae *

We have characterized the regulation of spermidine transport in yeast and identified some of the genes involved in its control. Disruption of the SPE2 gene encodingS-adenosylmethionine decarboxylase, which catalyzes an essential step in polyamine biosynthesis, up-regulated the initial velocity of spermidine uptake in wild-type cells as well as in the polyamine transport-deficient pcp1 mutants. Exogenous spermidine rapidly inactivated spermidine transport with a half-life of ≈10–15 min via a process that did not require de novoprotein synthesis but was accelerated by cycloheximide addition. Conversely, reactivation of spermidine influx upon polyamine deprivation required active protein synthesis. The stability of polyamine carrier activity was increased 2-fold in polyamine-depletedspe2 deletion mutants, indicating that endogenous polyamines also contribute to the down-regulation of spermidine transport. Ligand-mediated repression of spermidine transport was delayed in end3 and end4 mutants that are deficient in the initial steps of the endocytic pathway, and spermidine uptake activity was increased 4- to 5-fold in end3mutants relative to parental cells, although the stability of the transport system was similar in both strains. Disruption of theNPR1 gene, which encodes a putative Ser/Thr protein kinase essential for the reactivation of several nitrogen permeases, resulted in a 3-fold decrease in spermidine transport in NH4 +-rich media but did not prevent its down-regulation by spermidine. The defect in spermidine transport was more pronounced in NH4 +- than proline-grownnpr1 cells, suggesting that NPR1 protects against nitrogen catabolite repression of polyamine uptake activity. These results suggest that (a) the polyamine carrier is an unstable protein subject to down-regulation by spermidine via a process involving ligand inactivation followed by endocytosis and that (b) NPR1 expression fully prevents nitrogen catabolite repression of polyamine transport, unlike the role predicted for that gene by the inactivation/reactivation model proposed for other nitrogen permeases.

We have characterized the regulation of spermidine transport in yeast and identified some of the genes involved in its control. Disruption of the SPE2 gene encoding S-adenosylmethionine decarboxylase, which catalyzes an essential step in polyamine biosynthesis, upregulated the initial velocity of spermidine uptake in wild-type cells as well as in the polyamine transportdeficient pcp1 mutants. Exogenous spermidine rapidly inactivated spermidine transport with a half-life of Ϸ10 -15 min via a process that did not require de novo protein synthesis but was accelerated by cycloheximide addition. Conversely, reactivation of spermidine influx upon polyamine deprivation required active protein synthesis. The stability of polyamine carrier activity was increased 2-fold in polyamine-depleted spe2 deletion mutants, indicating that endogenous polyamines also contribute to the down-regulation of spermidine transport. Ligand-mediated repression of spermidine transport was delayed in end3 and end4 mutants that are deficient in the initial steps of the endocytic pathway, and spermidine uptake activity was increased 4-to 5-fold in end3 mutants relative to parental cells, although the stability of the transport system was similar in both strains. Disruption of the NPR1 gene, which encodes a putative Ser/Thr protein kinase essential for the reactivation of several nitrogen permeases, resulted in a 3-fold decrease in spermidine transport in NH 4 ؉ -rich media but did not prevent its down-regulation by spermidine. The defect in spermidine transport was more pronounced in NH 4 ؉ -than proline-grown npr1 cells, suggesting that NPR1 protects against nitrogen catabolite repression of polyamine uptake activity. These results suggest that (a) the polyamine carrier is an unstable protein subject to down-regulation by spermidine via a process involving ligand inactivation followed by endocytosis and that (b) NPR1 expression fully prevents nitrogen catabolite repression of polyamine transport, unlike the role predicted for that gene by the inactivation/reactivation model proposed for other nitrogen permeases.
Putrescine and the polyamines spermidine and spermine are ubiquitous molecules required for macromolecular biosynthesis (1,2), for posttranslational maturation of the essential protein eukaryotic initiation factor-5A (3), and for the control of various ion channel activities (4). Intracellular polyamine pools are actively regulated by de novo synthesis, degradation, excretion, and import from extracellular sources. Polyamine-specific carriers are widely distributed in prokaryotes and eukaryotes (5)(6)(7) and can replenish polyamine pools upon inhibition of the biosynthetic enzymes (8 -10). Mammalian polyamine transport activity is also acutely controlled by cell cycle events (11,12) and hormonal stimulation (5,9,10,13,14). Polyamine transport is a saturable, carrier-mediated and energy-dependent process (5,15,16). Although its physiological properties have been extensively studied, the molecular characteristics of the diamine and polyamine carrier proteins have only been elucidated in prokaryotes (7,(17)(18)(19).
In the yeast Saccharomyces cerevisiae, exogenous polyamines are internalized by specific, saturable plasma membrane uptake system(s) (20 -24). Recently, we (22) and others (24) have identified the PTK2 (ϭ STK2) gene, which encodes a putative Ser/Thr protein kinase, as a major determinant of high affinity polyamine transport in yeast. The closely homologous PTK1 (ϭSTK1) gene, which directs a lower affinity, low capacity polyamine transport system, has also been cloned and characterized (22,23). However, the exact role of these kinases in the regulation of polyamine transport as well as the molecular identity of the putrescine and polyamine carrier(s) in yeast have not yet been determined.
In mammalian cells, polyamine transport activity is tightly regulated by negative feedback mechanisms that depend on intracellular polyamine levels. A marked induction of polyamine transport activity is triggered by polyamine depletion, such as that caused by inhibition of ornithine decarboxylase or S-adenosylmethionine decarboxylase (AdoMetDC) 1 (5, 6, 8 -10, 25-28), and rapidly repressed by the addition of exogenous polyamines (9, 26, 29 -31). The expression of ornithine decarboxylase antizyme, which is translationally regulated by polyamine levels (32), plays an important role in the acute downregulation of mammalian polyamine uptake (29 -31). However, an exhaustive search in the Saccharomyces Genome Data Base has disclosed no antizyme-like gene in yeast, and little is known on the regulation of diamine and polyamine transport in lower eukaryotes. A Ca 2ϩ -sensitive gene product has been reported to negatively regulate putrescine transport in Neuros-pora crassa, although its molecular identity is as yet unknown (33).
More recently, we reported that spermidine transport in yeast is subject to negative regulation by substrate availability (22). The aim of the present study was to gain insight into the components and mechanisms responsible for the regulation of polyamine transport in yeast. By using various mutant strains, we now provide evidence that spermidine transport is mediated by an unstable carrier protein that is down-regulated by intracellular polyamines of both endo-or exogenous origins via a mechanism involving ligand-mediated inactivation and endocytosis. In addition, we show that the NPR1 gene, which encodes a Ser/Thr protein kinase structurally related to Ptk1p and Ptk2p, and is required for the derepression of various nitrogen permeases upon transfer to poor nitrogen sources (34,35), plays an important role in the expression of high affinity spermidine uptake activity under conditions (e.g. high [NH 4 ϩ ]) that were previously thought to down-regulate Npr1p activity.

MATERIALS AND METHODS
Strains, Media, and Plasmids-The yeast strains used in this study are listed in Table I and were routinely grown in YPD medium (1% yeast extract, 2% peptone, 2% D-glucose). The minimal medium used to study the influence of nitrogen sources on polyamine transport was made of 2% D-glucose and 0.17% amino acid-free yeast nitrogen base (Difco) without (NH 4 ) 2 SO 4 and supplemented with the appropriate essential amino acids. This medium was supplemented with either 10 mM (NH 4 ) 2 SO 4 (minimal NH 4 ϩ medium) or 9 mM L-proline (minimal proline medium) as sole nitrogen-containing compounds. Polyamine-free YNB medium (H medium; 2% D-glucose, 0.17% amino acid-free yeast nitrogen base, 38 mM (NH 4 ) 2 SO 4 , and essential amino acids) and agar plates were prepared as described (36). Yeast transformations were carried out using the lithium acetate method (37). The single copy YCp-PTK2 expression vector carrying the 3.4-kilobase pair BamHI fragment encompassing the full-length PTK2 gene cloned into YCp50 has been previously described (22). The PTK2 gene, including 9 and 70 bp adjacent to the 5Ј-and 3Ј-ends of the coding sequence, was isolated as a HpaI-HindIII fragment from the PTK2-encoding insert of YCp-PTK2 and cloned into the PvuII and HindIII restriction sites of the pYES2.0 multicopy expression vector (Invitrogen). The wild-type strain RH144-3D and its RH266-1D (end3) and RH268-1C (end4) mutant derivatives were kindly provided by Dr. Howard Riezman (University of Basel, Switzerland). The npr1 (21994b) and npi1 (27038a) mutants and parental strain (33346c), and the pMV33 plasmid carrying the NPR1 gene as a 5.2-kilobase pair insert in the multicopy pFL1 plasmid (38), were generous gifts of Dr. Bruno André (Université Libre de Bruxelles, Belgium).
Isolation of a Polyamine Transport-deficient Mutant-Wild-type strain DBY747 was mutagenized with methyl methanesulfonate (0.1%, v/v) as described (39). The mutagenized cell suspensions were plated on YPD agar plates containing 20 mM methylglyoxal bis(guanylhydrazone) (MGBG), a cytotoxic polyamine analog that shares the polyamine transport system in mammalian cells (5,40). MGBG-resistant colonies thus obtained were screened for spermidine uptake activity (cf. below), and one polyamine transport-deficient isolate (DRY405) was selected for further characterization. Diploidy tests for dominance of the genetic defect present in DRY405 cells were carried out as described previously (22).
Isolation of AdoMetDC-deficient Mutants-The spe2 mutants carrying a chromosomal deletion of the SPE2 gene encoding AdoMetDC were obtained by a one-step disruption technique (36). Wild-type yeast strain DBY747 and its polyamine transport-deficient derived strain DRY405 (ϭpcp1) were transformed with a 2829-bp NcoI-SphI fragment encompassing the SPE2 gene region in which the first 833 bp of the SPE2 coding sequence have been replaced with a 1970-bp HpaI-NarI fragment from YEp351 containing the full-length LEU2 gene (36), and Leu ϩ transformants were selected. Deletion-insertion in the chromosomal SPE2 gene in the clones thus obtained (DBY747spe2⌬ ϭPCP1spe2⌬ and DRY405spe2⌬ ϭ pcp1spe2⌬) was confirmed by PCR (data not shown).
Polyamine Analysis-For intracellular polyamine determination, cells were washed three times with ice-cold 100 mM sodium citrate (pH 5.5), and after centrifugation, cell pellets were resuspended in 0.5 ml of 5% (w/v) trichloroacetic acid and stored at Ϫ20°C until chromatographic analysis. After thawing, cell lysates were sonicated in an ultrasonic water bath for 3 cycles of 45 s in an ice-water slurry and microcentrifuged for 5 min at room temperature. Supernatants were filtered through 0.45-m cellulose acetate filters, and polyamine concentrations were determined using ion pairing reversed-phase high pressure liquid chromatography with post-column derivatization with o-phthaldialdehyde (41). Protein pellets were redissolved in 0.2 N NaOH and quantitated by the method of Bradford (42).
Spermidine and Citrulline Uptake Assays-[terminal methylene-3 H]Spermidine trihydrochloride (2.65 ϫ 10 4 Ci/mol) was obtained from NEN Life Science Products (Lachine, Quebec, Canada), and L-[ureido-14 C]citrulline (57.8 Ci/mol) was purchased from Amersham Corp. Spermidine uptake assays were performed as described previously (22), using 10 M [ 3 H]spermidine (50 Ci/mol) as substrate, unless otherwise indicated. Analysis of spermidine uptake characteristics was carried out according to Michaelis-Menten kinetics as described (22). Citrulline transport was measured by incubating 100 l of yeast cell suspension (5 ϫ 10 7 cells/ml) in citrate/glucose buffer (22) containing 20 M L-[ 14 C]citrulline. At predetermined times, the reaction was stopped by adding 1 ml of ice-cold stop buffer (citrate/glucose buffer containing 2 mM unlabeled L-citrulline). Cell suspensions were layered on cellulose acetate filters (pore size ϭ 0.45 m) held on a manifold filter vacuum apparatus pre-washed twice with ice-cold stop buffer, and mild vacuum was applied. Filters were washed 3 times with stop buffer, and radioactivity on the filters was then measured by liquid scintillation spectrometry.

Isolation of Polyamine Transport-deficient Yeast Mutants-
Previously characterized polyamine transport-deficient yeast mutants from this laboratory (22) and other laboratories (23,24) share mutations in the PTK2 gene as a common genetic defect. Such mutants are most likely regulatory mutants, since PTK2 encodes a putative Ser/Thr protein kinase of unknown function and is devoid of transmembrane domains (22,24). To generate polyamine transport mutants of a different genotype, wild-type DBY747 cells were chemically mutagenized and selected for their ability to grow in the presence of 20 mM MGBG, a toxic polyamine analog (22,40). One isolate from this selection (DRY405) exhibited a marked defect in spermidine transport (Fig. 1A) as well as a complete lack of specific putrescine uptake (data not shown). Kinetic analysis showed that the deficiency in spermidine uptake activity resulted from a Ϸ50fold increase in K m with no change in the V max for transport ( Fig. 1B), consistent with a structural mutation in the polyamine permease gene. Moreover, spermidine transport deficiency and growth sensitivity to MGBG could not be restored in DRY405 cells after transformation with the cloned PTK2 and NPR1 genes that are both required for normal polyamine transport (cf. below) (22). 2 Taken together, these observations suggest that DRY405 cells might express a dysfunctional poly- amine carrier, and the locus of the defective gene was tentatively called PCP1 (for polyamine carrier protein). The pcp1 mutation was found to be recessive, as diploids produced in crosses with the EI457 strain were not resistant to 20 mM MGBG (data not shown).

Spermidine Transport Is under Feedback Regulation by Intra-and Extracellular
Polyamines-In a previous study, we determined that preincubation with extracellular spermidine leads to a profound repression of spermidine but not putrescine uptake activity in yeast (22). To assess the relative contribution of endo-and exogenous polyamines in the control of spermidine uptake activity, the SPE2 gene encoding AdoMetDC, which provides aminopropyl group donors for both spermidine and spermine synthesis, was deleted by a one-step disruption method (36) in the wild-type DBY747 strain as well as in the pcp1 mutants. The resulting ⌬spe2-5::LEU2 mutants can be depleted from intracellular polyamines after incubation for a few generation times in polyamine-free medium (cf . Table II) (36). The effect of intracellular polyamine depletion and of exogenous spermidine on spermidine transport was then examined. After incubation in the presence of spermidine, a condition known to repress polyamine transport (22), spermidine uptake activity was identical in wild-type and PCP1spe2⌬ cells (Table II). However, deletion of SPE2 up-regulated spermidine uptake velocity following incubation in polyamine-free medium suggesting that endogenous polyamines negatively regulate spermidine transport. On the other hand, suppression of polyamine biosynthesis slightly enhanced the rate of spermidine accumulation in pcp1 mutants grown in the presence of spermidine (data not shown), although the initial velocity of uptake remained unaffected (Table II). Moreover, both the initial uptake velocity and rate of spermidine accumulation were markedly enhanced in pcp1spe2⌬ double mutants grown in polyamine-free medium. Thus, the pcp1 mutation still allows the polyamine-regulated expression of a functional permease with a strongly impaired affinity.
An elevated initial velocity of spermidine transport was generally correlated with a lower spermidine content, either after preincubation in polyamine-free medium or disruption of the AdoMetDC gene (Table II). SPE2 deletion also resulted in putrescine accumulation, as expected from the blockade in polyamine biosynthesis and from the relief of negative feedback inhibition of ornithine decarboxylase expression (20,43). The latter phenomenon also likely accounts for the decrease in putrescine levels accompanying accumulation of exogenous spermidine. Moreover, intracellular spermidine levels in cells with wild-type polyamine transport activity (PCP1) grown in the presence of spermidine reached a similar value whether endogenous biosynthesis was functional (SPE2) or not (spe2), suggesting that the size of the spermidine pool is adjusted by a common mechanism when an extracellular source of the polyamine is available. However, incubation with exogenous spermidine failed to significantly enlarge (p Ͼ 0.05) spermidine content in pcp1SPE2 cells, despite a detectable reduction in putrescine content. In fact, although the size and composition of the putrescine and polyamine pools were not affected by the pcp1 mutation, steady-state spermidine accumulation was 3-6-fold lower in these mutants upon SPE2 inactivation, consistent with a chronic repression of a defective spermidine uptake system by low levels of accumulated spermidine. Putrescine addition had no effect on spermidine transport velocity (data not shown), indicating that fluctuations in putrescine content resulting from the different incubation conditions and genetic backgrounds did not influence spermidine uptake activity. These data suggest that influx of exogenous spermidine strongly down-regulates spermidine transport, whereas endogenous levels of polyamines, but not putrescine, also exert a significant, but much less stringent negative control of spermidine transport. Moreover, marked down-regulation of polyamine uptake activity by exogenous spermidine was accompanied with only minor (e.g. in pcp1spe2⌬ cells) or even without  detectable changes (e.g. in pcp1SPE2 cells) in total spermidine content, suggesting that only a minor fraction of the total spermidine pool is required for feedback transport inhibition. The latter assumption was further confirmed by determining the concentrations of spermidine required to repress polyamine uptake activity (Fig. 2). Following overnight preincubation with increasing concentrations of spermidine, subsequent spermidine uptake activity was inactivated with an IC 50 ϭ 2.4 Ϯ 0.7 M, i.e. a value close to the K m of spermidine for transport in S. cerevisiae (4 to 5 M) (20,22). This observation is again consistent with a close connection between substrate internalization and the mechanism of transport inactivation triggered by spermidine.
Down-regulation of Spermidine Transport Does Not Require de Novo Protein Synthesis-Substrate-mediated down-regulation of spermidine transport could either result from decreased synthesis, increased turnover, or posttranslational inactivation of the polyamine permease. Thus, the effect of exogenous spermidine on the half-life (t 0.5 ) of polyamine uptake activity was examined in wild-type cells in the presence of 10 g/ml cycloheximide (CHX). Spermidine (100 M) was added to cells preincubated under polyamine-free conditions to fully derepress spermidine transport, and the rate of decay of spermidine transport activity was then determined. As shown in Fig. 3A, spermidine triggered an immediate and rapid loss of spermidine uptake activity, with a t 0.5 Ϸ10 -15 min, until a new equilibrium value was reached after 1 h. After CHX addition, spermidine uptake activity decayed much more slowly, with a Ϸ35% decrease in velocity observed after 3 h. Furthermore, protein biosynthesis inhibition enhanced the extent of substrate-induced inactivation of spermidine transport without affecting the t 0.5 of the decay process. Down-regulation of spermidine uptake was not prevented when 5 mM NaN 3 was added prior to spermidine addition (data not shown), demonstrating that ligand-mediated inactivation of spermidine transport did not directly depend on metabolic energy.
The time course of derepression of spermidine transport was examined by preincubating wild-type cells in the presence of spermidine and then transferring cultures to spermidine-free conditions in the presence or absence of CHX. Spermidine uptake activity increased in a time-dependent manner following transfer to polyamine-free medium, with a 2.5-fold increase observed after 3 h, and a significant derepression already detectable 30 min after the onset of polyamine deprivation (Fig.  3B). Derepression of spermidine uptake absolutely depended on de novo protein biosynthesis since it was completely abol-ished by CHX. Moreover, a continuous supply of exogenous spermidine was required to fully destabilize the transport system, as shown by the slower rate of decay of spermidine uptake activity noted upon transfer to spermidine-free conditions in the presence of CHX.
Intracellular Polyamine Depletion Stabilizes the Spermidine Transport System-Since endogenous polyamine pools also exert a down-regulatory effect on polyamine transport, we determined the decay of spermidine uptake activity after CHX addition in spe2⌬ mutants preincubated in polyamine-free medium. As shown in Fig. 4, the spermidine transport system was more stable in the spe2⌬ mutants than in the parental strain. Thus, the increase in spermidine transport activity observed upon depletion of endogenous polyamines is clearly associated with a decreased rate of its inactivation, in a manner similar to deprivation of exogenous polyamines.
Polyamine-induced Down-regulation of Spermidine Transport Partly Depends on Endocytosis-Several plasma membrane transporters in yeast are cleared from the cell surface by endocytic internalization (43)(44)(45)(46)(47)(48). The END3 (49) and END4 genes (50, 51) encode proteins that are involved in actin cytoskeleton organization, and their expression is required at early steps of the endocytic pathway for the internalization of several membrane proteins (52). To assess whether endocytosis is involved in the inactivation and/or degradation of the poly- amine carrier, we analyzed the stability of this transporter in end3 and end4 mutants that are thermosensitive for growth (52). Whereas the end3 mutants also exhibit a defect in endocytosis at the permissive temperature (25°C), the genetic lesion of end4 cells is more stringently temperature-dependent (52).
At the permissive temperature, spermidine transport under fully derepressing (i.e. polyamine-free) conditions was 4-to 5-fold higher in end3 cells than in the parental strain (Fig. 5). As expected, no significant difference in spermidine uptake was observed between end4 mutants and wild-type cells at 25°C, and spermidine transport velocity in wild-type cells was considerably lower than at 30°C, consistent with a requirement for metabolic energy for the uptake process. To assess the effect of impaired endocytosis on the turnover of the polyamine uptake system, we then determined the t 0.5 of spermidine transport inactivation in end3 and end4 mutants grown at 25°C under fully derepressing conditions and then transferred to a non-permissive temperature (37°C) in the presence of CHX. Fig. 6A shows that following deprivation of exogenous polyamines, the initial rate of decay of spermidine uptake activity was similar in end3 mutants and in the parental strain (t 0.5 ϭ 60 Ϯ 12 and 55 Ϯ 9 min, respectively) but was notably accelerated in end4 mutants (t 0.5 ϭ 21 Ϯ 6 min). Thus, the increased spermidine uptake activity observed in end3 mutants is clearly not associated with a lower rate of inactivation of the polyamine transport system, and in fact, the defect in endocytosis present in the end4 mutants decreases the stability of this carrier.
It has been shown that substrate-mediated inactivation and degradation of yeast plasma membrane carriers such as the copper transporter Ctr1p (53) occur via a mechanism that does not require prior endocytic internalization. To determine whether spermidine can inactivate the polyamine carrier system independently from endocytosis, we first deprived end3 and end4 mutants and wild-type cells from exogenous polyamines at 25°C, then transferred cultures for 10 min into spermidine-containing medium at the non-permissive temperature to initiate the inactivation process, and finally transferred cells to polyamine-free medium in the presence of CHX while continuously monitoring the time course of spermidine transport velocity. Spermidine rapidly triggered the decay of spermidine uptake activity in the parental strain, and spermidine removal slowed down but did not prevent the time-de-pendent inactivation of spermidine transport (Fig. 6B). Moreover, although the initial rate of inactivation of spermidine transport was significantly lower in the end3 and end4 mutants immediately after spermidine addition, a steady decay of sper- FIG. 4. Spermidine depletion stabilizes the polyamine transport system. DBY747 (wild-type, wt; E) and their spe2⌬ knockout derivatives (q) were grown to mid-exponential stage in polyamine-free medium. At time ϭ 0, CHX (10 g/ml) was added to the medium, and [ 3 H]spermidine uptake was measured for a 10-min period at the indicated times. Data are expressed as a percentage of initial activities and represent the mean Ϯ S.D. of triplicate determinations.
FIG. 5. Spermidine uptake activity is up-regulated by the end3 mutation at a permissive temperature. The wild-type RH144-3D strain (END3END4, E) and its end3 (end3END4, q) and end4 (END3end4, Ⅺ) mutant derivatives were grown to mid-exponential stage at 25°C in polyamine-free medium, after which the time course of [ 3 H]spermidine uptake was determined. Data are the mean Ϯ S.D. of triplicate determinations.

FIG. 6. Stability and spermidine-induced down-regulation of the spermidine transport system in endocytosis-deficient mutants.
Wild-type (RH144-3D) strain (END3END4, E), and its end3 (end3END4, q) and end4 (END3end4, Ⅺ)-derived mutants were grown to mid-exponential stage at 25°C in polyamine-free medium and then resuspended for 10 min in medium pre-warmed to 37°C. A, at time ϭ 0, CHX (10 g/ml) was added to the medium, and [ 3 H]spermidine uptake velocity was measured for 10-min periods at the indicated times. B, at time ϭ 0, spermidine (100 M) was added to the medium; after 10 min, spermidine was removed, and cells were transferred back to spermidine-free medium pre-warmed to 37°C and containing 10 g/ml CHX. [ 3 H]Spermidine uptake velocity was measured for 10-min periods at the indicated times. Results are expressed as a percentage of initial transport activity and represent the mean Ϯ S.D. of triplicate determinations. midine transport activity was still observed following removal of the polyamine, at a rate comparable with that of wild-type cells. These results suggest that spermidine internalization rapidly initiates an irreversible process of polyamine transport inactivation that does not require prior endocytosis of the polyamine permease nor continuous protein synthesis.
The NPR1 Gene Is Involved in the Regulation of Spermidine Transport Activity-The NPR1 gene encodes a putative Ser/ Thr protein kinase that plays an important role in the regulated expression of several amino acid permeases (34,35,38,54), as well as of the NH 4 ϩ transporter Mep1p (34,55,56), which are under nitrogen catabolite repression (NCR). According to one model (34,35), NPR1 expression is required to counteract the constitutive inactivation/degradation of nitrogen permeases by the Npi1p ubiquitin-protein ligase in cells growing on poor nitrogen sources (e.g. low NH 4 ϩ , proline, etc.) via a posttranslational mechanism (57). Interestingly, the PTK1 and PTK2 genes that are required for expression of diand polyamine transport activity (22,24) belong to the same homology class of Ser/Thr protein kinases as NPR1 (58). The role of Npr1p as a reactivator of amino acid transport and the similarities between the predicted primary structures of Npr1p, Ptk1p, and Ptk2p prompted us to determine whether Npr1p might play a role in the derepression of spermidine uptake activity observed upon deprivation of exogenous polyamines.
As shown in Fig. 7A, spermidine uptake activity was decreased by 2-to 3-fold in npr1 mutants as compared with the parental strain after preincubation under polyamine-free conditions. Moreover, exogenous spermidine down-regulated spermidine transport in both wild-type and npr1 cells. The defect in spermidine transport in npr1 mutants was clearly related to the lack of NPR1 expression, as spermidine transport activity was nearly restored to wild-type levels upon transformation with a multicopy NPR1 expression vector (Fig. 7B). Thus, these data clearly demonstrate that Npr1p expression is indeed involved in the regulated expression of polyamine transport activity but not in the ligand-dependent repression/derepression of spermidine transport.
Polyamine Transport Activity Is Not Subject to NCR-An unexpected finding from the above experiments was that the lack of NPR1 expression in a NH 4 ϩ -rich medium (i.e. 38 mM (NH 4 ) 2 SO 4 ) strongly depressed spermidine transport activity, although under the same conditions, Gap1p activity is reduced to almost undetectable levels in an NPR1-independent manner (34,35,54). We indeed verified that Gap1p activity (as measured by the rate of [ 14 C]citrulline transport) was dramatically up-regulated upon transfer from NH 4 ϩ to proline as sole nitrogen sources for growth (data not shown). We next determined the effect of NCR on spermidine uptake activity in wild-type and npr1 cells preincubated in polyamine-free minimal medium containing either NH 4 ϩ or proline as sole nitrogen sources. Spermidine transport in wild-type cells was clearly not repressed by growth on NH 4 ϩ and was in fact reproducibly higher than that measured in proline-grown cells (Fig. 8). Interestingly, the differential effect of the npr1 mutation on spermidine uptake was much more pronounced in NH 4 ϩ -than prolinegrown cells. These results indicate that the spermidine transport system is resistant to NCR and that NPR1 expression might in fact protect this transporter against an NH 4 ϩ -dependent inactivating mechanism.
One such inactivating mechanism could obviously be the action of the ubiquitin-protein ligase Npi1p, which targets several yeast membrane transporters for degradation through the endocytosis-vacuolar pathway (34, 57, 59 -61). Although NPI1 is an essential gene (57), partly Npi1p-deficient mutants exhibit reduced rates of endocytosis and ubiquitination of target proteins such as the uracil permease Fur4p (60) and a pleiotropic up-regulation of amino acid permease activity under NH 4 ϩ -repressing conditions (34,61). Thus, if spermidine transport is under NH 4 ϩ -induced repression by Npi1p in a manner similar to Gap1p, polyamine uptake activity should be upregulated in npi1 mutants. However, there was no difference in spermidine transport between wild-type and npi1 cells under conditions (i.e. high [NH 4 ϩ ]) that lead to a marked derepression of Gap1p activity in NPI1-deficient cells (Fig. 9) (34, 61). Moreover, the ability of spermidine to down-regulate polyamine transport was not decreased by a deficiency in NPI1 expression, suggesting that Npi1p is not a major determinant in the spermidine-mediated inactivation of the polyamine permease.

DISCUSSION
Like for mammalian polyamine transport (5, 6, 8 -10, 25-31), high affinity spermidine uptake activity in yeast is clearly under negative control by polyamines of both endogenous and exogenous origins. However, extracellular spermidine quantitatively exerts a more stringent feedback inhibition on its own high affinity transport than the endogenous pool. Likewise, deprivation of exogenous polyamines had a disproportionately large stimulatory effect on spermidine transport as compared with polyamine depletion due to a genetic block in endogenous synthesis, as evidenced in pcp1spe2⌬ double mutants. The weaker feedback inhibition of spermidine transport exerted by endogenous polyamines could be due to sequestration of de novo synthesized polyamines in the vacuole, as well as to binding to macromolecular polyanions. Although subcellular polyamine compartmentalization has yet to be investigated in S. cerevisiae, it has been estimated that at least 25% of total cellular spermidine is confined to the vacuole in N. crassa (62), and a vacuolar, H ϩ -ATPase-dependent active polyamine transport system has been described in yeast (63). Furthermore, due to their mixed polycationic/aliphatic character, intracellular polyamines may largely exist as complexes bound to nucleic acids and free nucleotides, a factor that could reduce their free fraction to less than 10% (64). Thus, the more efficient downregulation exerted by exogenous spermidine might indicate that the underlying mechanism involves a rapid and direct allosteric interaction with the polyamine permease, without mixing with the endogenous pool. This interpretation would account for the closely similar values of the K m of spermidine for high affinity uptake and the IC 50 of exogenous spermidine for transport repression. The present data cannot rule out the possibility that the regulatory site responsible for carrier destabilization by spermidine is identical to the substrate binding site or lies in a different, extracellular domain of the polyamine permease. However, the fact that intra-and extracellular polyamines affect polyamine permease stability in a qualitatively similar manner suggests that the allosteric site for ligand inactivation might lie on the endofacial side of the carrier.
In mammalian cells, polyamine depletion up-regulates poly-amine uptake as a result of decreased antizyme expression (26, 29 -31), as well as via antizyme-and protein synthesis-independent mechanisms (9). Because antizyme is a short-lived negative regulator of polyamine uptake, studies using CHX have not yielded clear estimates of the t 0.5 of the mammalian polyamine transporter(s) due to the relief of uptake activity that accompanies antizyme depletion. In contrast, the present data strongly suggest that the yeast polyamine permease is a rather unstable protein and that no antizyme-like system participates to its feedback inhibition, as predicted by the lack of sequences homologous to either bacterial and mammalian antizyme in the yeast genome. Some features of the rapid repression elicited by spermidine on polyamine transport are reminiscent of the ligand inactivation mechanism recently proposed for the down-regulation of the inositol permease Itr1p (45) and the copper transporter Ctr1p by their substrate (53). The rate of degradation of these carriers is dramatically enhanced upon exposure to their substrate via a feedback mechanism that likely involves direct ligand binding to an allosteric site on the exo-or endofacial side of the carrier (45,53). However, spermidine-mediated downregulation of the polyamine permease exhibits features that are intermediary between the ligand inactivation mechanisms respectively described for the Itr1p and Ctr1p carriers. Endocytosis has been shown to form an integral part of the inactivation/degradation mechanism elicited by inositol binding to Itrp1, since end3 and end4 mutants are spared from ligandmediated inactivation of Itr1p after removal of inositol from the medium (45). On the other hand, Ctr1p undergoes copperinduced degradation while still inserted in the plasma membrane and does not require endocytic internalization for its down-regulation (53). In contrast, we have observed that spermidine-mediated inactivation of polyamine transport is only delayed by defective endocytosis, suggesting that spermidine triggers the destabilization process independently from internalization of the polyamine permease and that ligand inactivation might in fact accelerate the endocytic degradation of the transporter. Moreover, deficient endocytosis did not decrease the basal rate of turnover of the polyamine uptake system, the t 0.5 of spermidine uptake activity being even shorter in end4 mutants. A role for endocytosis in the regulation of polyamine transport is nonetheless suggested by the fact that spermidine uptake was strongly up-regulated in end3 mutants grown at the permissive temperature. The marked instability of the yeast polyamine uptake system would predict that a high rate of insertion and endocytic removal of carrier molecules into and from the plasma membrane is required to maintain adequate transport activity. Since the t 0.5 of the transport system was not affected by the end3 mutation, increased spermidine uptake activity in end3 cells could result from defective recycling of functional polyamine permease molecules at the endocytic internalization step, resulting in a steady-state increase in the number of active spermidine carriers in the plasma membrane.
Our results are thus consistent with a model where spermidine controls the rate of degradation of the polyamine permease via a posttranslational, ligand inactivation mechanism that destabilizes the carrier. The polyamine permease appears to turn over rapidly, and up-regulation of spermidine transport clearly depends on de novo protein biosynthesis, either because derepression reflects an increase in the number of polyamine permease molecules or because a spermidine-repressed, essential reactivator is induced following polyamine deprivation. Upon its binding to an allosteric site, spermidine accelerates the rate of endocytic internalization of functional transporters, apparently without an involvement of the Npi1p ubiquitinprotein ligase that controls the endocytic degradation of several yeast transporters (57,60). The observed kinetics of repression/ derepression of polyamine transport is consistent with a predominantly posttranslational control, as is the case for the regulation of yeast ornithine decarboxylase by spermidine (20,43). Without appropriate molecular tools, we cannot currently discriminate between the possibilities that ligand inactivation of the polyamine permease involves its proteolytic degradation, conformational changes that irreversibly inhibit its activity as a result of allosteric interaction, and/or spermidine-regulated posttranslational modifications.
In connection with the latter hypothesis, we have explored the possibility that polyamine transport might be regulated by the NPR1-dependent inactivation/reactivation system governing the posttranslational control of several nitrogen permeases (34,35). We have indeed found that NPR1 expression promotes spermidine transport, but its inactivation does not abolish high affinity spermidine uptake nor prevent its ligand-induced down-regulation, as noted for ptk2::lacZ disruption mutants (22). However, inactivation of both PTK1 and PTK2 was sufficient to fully suppress high affinity spermidine transport and feedback transport repression (22). Thus, although Npr1p, Ptk1p, and Ptk2p all encode putative, homologous Ser/Thr protein kinases, Npr1p action clearly requires functional PTK1 and/or PTK2 genes. Surprisingly, spermidine transport was strongly depressed in npr1 mutants under conditions (i.e. high [NH 4 ϩ ]) where Npr1p activity has been predicted to be downregulated by NCR (34,35). In fact, the npr1 mutation does not further decrease the already low activity of the Gap1p or Put4p permeases in NH 4 ϩ -rich media, although it up-regulates their expression at the mRNA level (34,38,54,65,66). Moreover, our data clearly demonstrate that the spermidine transport system is not subject to NCR, and if anything, NH 4 ϩ would rather be a weak inducer of spermidine transport.
Obviously, the model of inactivation/reactivation of nitrogen permeases by NCR as originally proposed (34,65) is not compatible with the observed mechanism of regulation of polyamine transport. In that model, NPR1-regulated permeases such as Gap1p are constitutively degraded via a ubiquitin-dependent pathway involving the NPI1 and NPI2 genes, and rich nitrogen sources such as NH 4 ϩ and L-glutamine down-regulate Npr1p expression at a posttranscriptional level (34,35,38). This level of regulation by nitrogen catabolites was suggested to involve the increased turnover of Npr1p, which, like Ptk1p and Ptk2p (22), has strong PEST sequences (67) that might signal its rapid degradation. However, no report is yet available on the regulation of Npr1p protein levels nor catalytic activity by NCR. Our data show that NH 4 ϩ certainly does not impair NPR1 expression as pertaining to spermidine transport regulation. If Npr1p activity per se is not down-regulated by NH 4 ϩ , reactivation of nitrogen permeases by NPR1 should thus involve some downstream, NCR-antagonized target of Npr1p. The simplest case would be that NPR1-dependent permeases require phosphorylation for their activity and/or stability and that NCR involves the posttranslational activation of a phosphatase directed to phosphorylated, active permeases. In support of this hypothesis, L-glutamine down-regulates Gap1p through its rapid dephosphorylation, a modification that inactivates the permease and promotes its degradation (68). There are precedents for a role of Ser/Thr protein phosphatases in the control of membrane transport in yeast. For instance, protein phosphatase 2B (or calcineurin) causes the switch of the Trk1p K ϩ transporter from a low to a high affinity mode of K ϩ transport upon Na ϩ stress (69).
Although spermidine transport is regulated by Npr1p, it is clearly resistant to NCR and is not inactivated via an NPI1catalyzed pathway. The Mep1p permease responsible for low affinity, high capacity NH 4 ϩ uptake is another NPR1-dependent, NPI1-independent nitrogen permease but is nonetheless subject to NCR (55,56). Thus, there is no obligatory link between NPR1, NCR, and the NPI1-dependent pathway of ubiquitination/degradation in the control of nitrogen permease activity. That Npr1p deficiency had a substantially larger effect on spermidine transport in NH 4 ϩ -than proline-based medium suggests that the putative kinase could in fact protect the polyamine transport system against NCR. Why would Npr1p expression confer resistance of spermidine transport to NCR but not for other Npr1p-dependent permeases? One possibility is that Npr1p regulates the expression and/or activity of intermediary factors specifically targeting the spermidine uptake system such as Ptk1p and Ptk2p, and that NPR1 function requires intact PTK1 and PTK2 genes. We have indeed found that ptk2::lacZ disruption mutants exhibit phenotypic responses similar to npr1 cells in response to nitrogen sources. 2 In fact, any of the Ser/Thr protein kinases known to affect spermidine transport could either activate the polyamine permease through phosphorylation, transcriptional induction, and/or inactivation of a putative permease phosphatase. The ongoing molecular identification of the PCP1 gene should help to elucidate the unique features of spermidine transport regulation in yeast.