- Email: [email protected]
Regulation of Glucose Partitioning by PAS Kinase and Ugp1 Phosphorylation
Molecular Cell
Article Regulation of Glucose Partitioning by PAS Kinase and Ugp1 Phosphorylation Tammy L. Smith1 and Jared Rutter1,* 1
Department of Biochemistry, University of Utah School of Medicine, Salt Lake City, UT 84112, USA *Correspondence: [email protected] DOI 10.1016/j.molcel.2007.03.025
SUMMARY
The ability of cells to recognize and respond to specific metabolic deficiencies is required for all forms of life. We have uncovered a system in the yeast S. cerevisiae that, in response to a perceived deficiency in cell wall glucan, alters partitioning of glucose toward glucan synthesis and away from glycogen synthesis. The paralogous yeast PAS kinases Psk1 and Psk2 phosphorylate UDP-glucose pyrophosphorylase (Ugp1), the primary producer of UDP-glucose, the glucose donor for glucan biosynthesis. Unexpectedly, phosphorylation of Ugp1 does not affect its catalytic activity but instead alters the terminal destination of the UDP-glucose it generates. Phosphorylated Ugp1 is required for intensive glucan production, and inability to phosphorylate Ugp1 is associated with a weak cell wall, decreased glucan content, and increased glycogen content. We provide data indicating that phosphorylation by Psk1 or Psk2 targets Ugp1 to the cell periphery, where the UDP-glucose it produces is in proximity to the site of glucan synthesis. We propose that regulation of glucose partitioning by altered enzyme and substrate localization is a rapid and potent response to metabolic deficiency.
INTRODUCTION Glucose is a versatile source of energy used by most living cells. Among other fates, internalized glucose can be used in glycolysis to create usable energy, in glycogen synthesis to store energy, and in the pentose-phosphate pathway to generate NADPH, or it can be used to glycosylate proteins or lipids. Misregulated flux partitioning can lead to an inappropriate surplus of product from one pathway and a deficit of product from another. An essential aspect of the adaptive behavior of cells is the ability to rapidly sense such a deficiency and alter flux to correct it, thus maintaining homeostasis. In multicellular organisms, this
process needs not only to respond to cellular deficiencies, but also to the needs of tissues and the organism as a whole. This is accomplished partially through the hormonal and neural coordination of metabolic supply and demand. Complex human diseases, such as obesity and diabetes mellitus, are fundamentally characterized by a cellular dysregulation of sugar and lipid metabolic flux and flux partitioning. Understanding how flux is managed within individual cells is essential to understanding and managing these organismal diseases. Individual Saccharomyces cerevisiae cells are essentially autonomous, providing a simpler system in which to study metabolic flux partitioning. As with most other cells, yeast cells rely upon appropriate metabolic adaptation to respond to constantly evolving demands. One example of such an adaptive response involves the partitioning of intracellular glucose between storage as glycogen and use in cell wall glucan synthesis (reviewed in Klis et al. [2002]). Actively dividing yeast cells have an enormous demand for cell wall glucan synthesis, as the extensive cell wall needs to be replicated during each round of division. As nutrients are depleted and division slows, yeast adapt glucose partitioning toward glycogen storage in preparation for impending starvation (Francois and Parrou, 2001). In addition to these normal physiological changes in demand, this partitioning is also responsive to other stressors, including heat shock and cell wall stress (Levin, 2005). We have discovered a system that is essential in yeast for appropriate glucose partitioning and for adaptation of that partitioning to changes in demand. The central components of this system are the enzyme UDP-glucose pyrophosphorylase (Ugp1) and its product UDP-glucose, which is the immediate substrate for both glucan and glycogen synthesis (Daran et al., 1995). Herein we describe two isoforms of Ugp1 that differ in the phosphorylation status of a single serine residue (S11). This phosphorylation event, catalyzed by the paralogous PAS kinases Psk1 and Psk2, leads to a conformational change in Ugp1 and is required for appropriate glucose partitioning toward glucan synthesis. We provide evidence that the phosphorylation-induced conformational change targets Ugp1 to the cell periphery, where its product UDP-glucose will be generated in proximity to the site of glucan synthesis (Montijn et al., 1999). This represents a previously undescribed mechanism for cells to direct glucose flux partitioning through localized metabolism of substrates.
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 491
Molecular Cell PAS Kinase and Glucose Partitioning
RESULTS Phosphorylation of Ugp1 at S11 by Psk2 Occurs In Vivo We previously demonstrated that Psk2 efficiently phosphorylates purified Ugp1 in vitro predominantly at serine 11 (S11) (Rutter et al., 2002). To determine the in vivo relevance of this phosphorylation event, we examined wild-type and PAS kinase-deficient (psk1 psk2) strains of Saccharomyces cerevisiae for phosphorylation of endogenous Ugp1 using an S11 Ugp1 phosphospecific antibody. While Ugp1 protein levels are identical in the two strains (Figure 1A, lower panel), only the wild-type strain contains S11-phosphorylated Ugp1 (upper panel). The psk1 psk2 strain is unable to efficiently grow on galactose-containing medium at 38 C (Galts phenotype) (Figure 1B, rows 1 and 2) (Rutter et al., 2002). To determine whether the inability of the psk1 psk2 strain to phosphorylate Ugp1 at S11 was responsible for the Galts phenotype, we generated a strain wherein the endogenous UGP1 allele was replaced with an unphosphorylatable S11A mutant allele of UGP1. The ugp1-S11A strain shows a Galts phenotype identical to the psk1 psk2 strain (Figure 1B, row 3), as does the strain wherein the psk1 psk2 mutations and the ugp1-S11A mutation are combined. Strains containing either a ugp1-S11D or ugp1-S11E allele behaved identically to the ugp1-S11A strain, suggesting that in this context, neither aspartate nor glutamate mimic phosphoserine. Together, these results strongly support the conclusion that phosphorylation of Ugp1 at S11 is biologically relevant and, at least for the Galts phenotype, is the major in vivo function of Psk1 and Psk2. While phosphorylation of Ugp1 S11 was necessary for viability under Galts and other (see below) conditions, the direct biochemical effect of phosphorylation was unknown. Phosphorylation has long been known to directly regulate the catalytic activity of other enzymes of glycogen metabolism, including glycogen synthase (Cohen, 1982; Hardy and Roach, 1993) and glycogen phosphorylase (Lerch and Fischer, 1975; Lin et al., 1995). We therefore sought to determine whether the effect of Ugp1 phosphorylation might be to directly affect the catalytic rate of the enzyme. To test this, we generated phospho-Ugp1 by incubation in the presence of both PAS kinase and ATP for 1 hr at 25 C. These phosphorylation conditions generate stoichiometrically phosphorylated Ugp1 as measured by 32 P incorporation using [g-32P]ATP. We also performed control incubations lacking PAS kinase, ATP, or both. We then measured the catalytic rate for each of these Ugp1 samples. As shown in Figure 2A, none of these treatments altered Ugp1 catalytic rate. We also examined Ugp1 catalytic rate in extracts from the same four strains described above (i.e., wild-type, psk1 psk2, ugp1-S11A, and psk1 psk2 ugp1-S11A). As with Ugp1 phosphorylated in vitro, Ugp1 catalytic rate in these strains was not significantly different (Figure 2B). These data support the unexpected conclusion that while phosphorylation of Ugp1 is biologically important, it does not exert its effects through a change in the catalytic rate of the enzyme.
Figure 1. The Psk Proteins Phosphorylate Ugp1 at Ser11 In Vivo (A) BHY10 (PSK1 PSK2) and JRY40 (psk1 psk2) strains were grown to log phase in synthetic complete medium. Total protein extract was separated and subjected to western blotting using either a Ser11Ugp1 phosphospecific antibody (top panel) or a panspecific Ugp1 antibody (bottom panel). (B) BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to saturation, and 5-fold serial dilutions were spotted on a synthetic galactose medium plate and grown for 5 days at 38 C (Galts conditions).
Phosphorylation Is Necessary and Sufficient for Conformational Change of Ugp1 As we observed that phosphorylation does not exert its biochemical effects through a change in Ugp1 catalytic rate, we tested whether it affected the isoform distribution that has been observed for Ugp1 previously (Dutra et al., 1996). We grew the four strains to log phase in standard minimal yeast medium and produced a clarified extract from each. This extract was separated using anion exchange chromatography, and the resultant fractions were assayed for Ugp1 activity. All strains showed a peak of activity between fractions 5 and 6, but only the wild-type strain showed a second peak in fraction 13 (Figure 3A). When the fractions from the wild-type strain were examined by western blotting, both peaks contained Ugp1 protein, but only the second peak contained S11phosphorylated Ugp1 (Figure 3B). Western blotting of fractions from the three mutant strains showed Ugp1 only in the first peak in correspondence with the enzyme
492 Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc.
Molecular Cell PAS Kinase and Glucose Partitioning
phosphorylation for 120 min (data not shown). We conclude that S11 phosphorylation of Ugp1 is sufficient to cause the transition to isoform 2. We also note that the observed bimodal distribution, most clearly evident in the 30 min time point, strongly suggests a concerted mechanism for this transition. This cooperativity implies a multimeric organization of Ugp1, and Ugp1 from S. cerevisiae is indeed an octamer (data not shown) as it is in most species (Turnquist and Hansen, 1973). To further characterize the nature of the transition from isoform 1 to isoform 2, we examined the gross structure of the two isoforms by limited proteolysis (see Figure S1 in the Supplemental Data available with this article online). Phosphorylated and unphosphorylated Ugp1 were subjected to digestion with either Pronase E or Proteinase K for various times. Two noticeable differences were apparent between the phosphorylated and unphosphorylated proteins, and these differences were observed with both proteases. First, the phosphorylated protein has increased general resistance to proteolysis, as observed by the presence of the intact protein after 40 min of digestion, by which time the unphosphorylated intact protein is gone. Secondly, the patterns of digestion products over the time course are significantly different. These results, and particularly their similarity with two different proteases, strongly suggest that the transition from isoform 1 to isoform 2 is characterized by a concerted change from one distinct conformation to another. Figure 2. Phosphorylation Does Not Affect Ugp1 Enzymatic Activity (A) Purified Ugp1 protein was incubated at 30 C for 1 hr in the presence or absence of purified PASK protein and ATP as indicated. Ugp1 activity was then assayed as described (n = 3). (B) BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to log phase, and a total protein extract was generated. This extract was then assayed for Ugp1 activity as described (n = 4). Error bars indicate standard deviation.
assay data. Therefore, the first peak corresponds to unphosphorylated Ugp1 (isoform 1) and the second peak corresponds to phosphorylated Ugp1 (isoform 2). The in vivo analysis of the four strains suggests that phosphorylation at S11 is required for the transition from isoform 1 to isoform 2. We next sought to determine whether phosphorylation was sufficient for this transition. Purified Ugp1 was phosphorylated in vitro with purified PAS kinase. Samples were taken at various time points and chromatographically analyzed as before. Before phosphorylation (0 min) all of the Ugp1 is found in isoform 1, corresponding to the unphosphorylated enzyme as seen in Figure 3B. After 30 min, a 50% shift to isoform 2 is evident, and by 60 min this shift is essentially complete (Figure 4A). When the fractions from the 30 min time point were examined by western blotting as before, both peaks contained Ugp1, but as with the in vivo experiment only the second peak contained S11-phosphorylated Ugp1 (Figure 4B). Importantly, purified S11A mutant Ugp1 protein was never observed to shift to isoform 2 even after
Phosphorylation of Ugp1 Is Required for Cell Wall Maintenance at the Expense of Glycogen Synthesis Ugp1 catalyzes the conversion of glucose-1-phosphate and UTP to UDP-glucose, which is used in yeast primarily as a glucose donor for the synthesis of storage carbohydrates (such as glycogen) and structural carbohydrates (such as cell wall glucans) (Daran et al., 1995). We hypothesized that the two isoforms of Ugp1 might be differentially responsible for supplying UDP-glucose to these two distinct pathways. If so, an inability to phosphorylate Ugp1 and thereby create isoform 2 might cause inappropriate glucose partitioning toward glycogen at the expense of cell wall glucan synthesis. To test this hypothesis, we determined glycogen levels in the wild-type and mutant strains. As shown in Figure 5A, the three mutant strains showed an 8- to 10-fold increase in glycogen levels relative to the wild-type strain. To test for a coincident decrease in cell wall glucans, we tested the same four strains for sensitivity to K1 killer toxin, a pore-forming protein that kills sensitive cells by binding to b-(1,6)-glucan, resulting in membrane damage and cell death (Shahinian and Bussey, 2000). Yeast cells containing less cell wall b-(1,6)-glucan show decreased sensitivity to this toxin, as evidenced by a smaller kill zone in a plate test (Boone et al., 1990). As shown in Figure 5B, wild-type yeast were significantly more sensitive to K1 killer toxin in a plate test than the three mutant strains. We therefore conclude that inability to phosphorylate S11 of Ugp1
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 493
Molecular Cell PAS Kinase and Glucose Partitioning
Figure 3. S11 Phosphorylation Is Required for Ugp1 Chromatographic Shift BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to log phase, and a total protein extract was generated. This extract was fractionated by MonoQ ionexchange chromatography. (A) The resultant fractions were assayed for Ugp1 activity and are plotted. The data shown are representative of four independent experiments. (B) The fractions from the wild-type (BHY10) strain were separated and subjected to western blotting using either a panspecific Ugp1 antibody (top panel) or a Ser11-Ugp1 phosphospecific antibody (bottom panel).
causes glycogen hyperaccumulation at the expense of synthesis of a key cell wall component, b-(1,6)-glucan. Decreased cell wall b-(1,6)-glucan content could lead to a weaker cell wall and decreased cellular stability. We analyzed the cell wall integrity of the four strains by testing their viability under several conditions of cell wall stress. Congo red, caffeine, and calcofluor white have each been shown to interfere with assembly of various cell wall components (de Groot et al., 2001; Kopecka and Gabriel, 1992). Mutations in the cell wall glucan-producing machinery have been shown to cause altered sensitivity to these agents (Lussier et al., 1997). Consistent with Ugp1 isoform 2 being preferentially involved in cell wall maintenance, the three mutant strains that are unable to generate isoform 2 exhibit marked hypersensitivity to Congo red (Figure 6A) and calcofluor white (data not shown), and resistance to caffeine (data not shown).
As described previously, the mutant strains are unable to grow under Galts conditions. Such a growth phenotype could result from many different primary defects, one being inadequate cell wall glucan production. If the Galts phenotype of the mutant strains is due to inadequate cell wall synthesis, then it should be suppressed by inclusion of the osmotic stabilizer sorbitol. Indeed, sorbitol completely rescues the Galts defect of the mutant strains (Figure 6B). Combined, these data demonstrate that an inability to phosphorylate Ugp1 causes increased glycogen production and compromised cell wall integrity. These results suggest a model wherein phosphorylation of Ugp1 controls glucose partitioning by determining the fate of its product UDP-glucose. Specifically, unphosphorylated Ugp1 (isoform 1) produces UDP-glucose that is preferentially used for glycogen synthesis, and phosphorylated Figure 4. S11 Phosphorylation Is Sufficient for Ugp1 Chromatographic Shift Purified Ugp1 was incubated at 30 C for the times indicated in the presence of purified PASK protein and ATP. These reactions were stopped by the addition of EDTA and fractionated by MonoQ ion-exchange chromatography. (A) The Ugp1 activity of the resultant fractions was assayed and is plotted. The data shown are representative of three independent experiments. (B) The fractions from the 30 min time point were separated and subjected to western blotting using either a panspecific Ugp1 antibody (top panel) or a Ser11-Ugp1 phosphospecific antibody (bottom panel).
494 Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc.
Molecular Cell PAS Kinase and Glucose Partitioning
Figure 5. S11-Ugp1 Phosphorylation Is Required for Normal Glycogen and b-(1,6)-Glucan Synthesis (A) BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to log phase and assayed for glycogen content as described (Rutter et al., 2002). The average of three experiments ± SD is shown. (B) BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to saturation and seeded in a minimal galactose plate. Concentrated K1 killer toxin, purified from strain SK-S14a, was spotted, and plates were incubated at 18 C for 7 days. The zone of growth inhibition was measured, and the average of four experiments ± SD is shown.
Ugp1 (isoform 2) produces UDP-glucose that is preferentially used for glucan synthesis. Cell wall glucan synthesis occurs primarily at the plasma membrane, where the synthetic enzymes for the major cell wall glucans are located (Montijn et al., 1999; Qadota et al., 1996). Conversely, glycogen synthase is localized to the cytoplasm (Huh et al., 2003; Kumar et al., 2002). We hypothesized that the different isoforms of Ugp1 might determine the fate of their product UDP-glucose by being differentially localized to the site of glycogen synthesis (isoform 1) or glucan synthesis (isoform 2). Ideally we would be able to visualize this localization by immunofluorescence with an anti-Ugp1 antibody or anti-epitope antibody to tagged Ugp1 or through a Ugp1-GFP fusion. Unfortunately, many different anti-Ugp1 antibodies proved ineffective for immunocytochemistry and both N- and C-terminal GFP or epitope tags disrupted the proper folding of Ugp1. As an alternative test of the localization hypothesis, we engineered Ugp1 with three independent sequences designed to target it to the plasma
membrane in the absence of phosphorylation. Specifically, the targeting sequences used were the C-terminal transmembrane segment from synaptobrevin (Snc2) (Pryciak and Huntress, 1998), the C-terminal prenylation sequence from Ras2 (Bhattacharya et al., 1995; Boyartchuk et al., 1997; Dong et al., 2003), and the N-terminal myristoylation sequence from Sip2 (Lin et al., 2003). As shown in Figure 7A, each of these targeted Ugp1 variants is able to substantially complement the Galts phenotype of a psk1 psk2 mutant strain. Further, the presence of wildtype PSK1 and PSK2 genes does not increase growth under Galts conditions when the Ugp1 is ectopically targeted to the cell periphery. If the three artificially membrane-associated Ugp1 variants faithfully mimic phospho-Ugp1, then they should correct the glucose partitioning defects found in a psk1 psk2 mutant. Indeed, we found that psk1 psk2 mutant strains bearing any of the three targeted Ugp1 variants had significantly decreased glycogen relative to the psk1 psk2 mutant with wild-type Ugp1 (Figure 7B). Further, b-(1,6)glucan levels were restored to wild-type levels in a psk1 psk2 mutant strain with any of the three targeted Ugp1 variants (Figure 7C). In each of these three assays (Galts suppression, glycogen levels, and glucan levels), the PSK1 PSK2 UGP1:RAS2 and PSK1 PSK2 UGP1:SIP2 strains were similar to the PSK1 PSK2 UGP1:SNC2 strain (data not shown). Combined, these data provide evidence that Ugp1 phosphorylation leads to plasma membrane localization and that this localization is required for increased glucan synthesis at the expense of glycogen synthesis. This increased glucan is then necessary for growth under stress, such as in Galts conditions. DISCUSSION In order to survive, living cells need to constantly monitor their environment and make appropriate metabolic decisions in response to environmental changes. The ability of a cell to rapidly respond to changes in nutritional status and environmental stress confers growth and reproductive advantages (Francois and Parrou, 2001; Wullschleger et al., 2006). We have shown here that an important metabolic decision regarding the fate of intracellular glucose is enacted by phosphorylation of Ugp1 by PAS kinase. This phosphorylation event determines whether UDPglucose is preferentially used to make storage or structural carbohydrates. We made the surprising observation that, while phosphorylation of Ugp1 was essential under certain conditions, it did not affect the enzymatic activity of Ugp1 in vitro or in vivo. We found that phosphorylation, instead, caused a conformation change in Ugp1 that is easily detected using limited proteolysis. Further, unphosphorylated Ugp1 (isoform 1) and phospho-Ugp1 (isoform 2) are readily resolved using ion-exchange chromatography. Cells unable to make isoform 2, either through elimination of both PAS kinase genes or through mutation of the PAS kinase phosphorylation site on Ugp1, exhibit marked
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 495
Molecular Cell PAS Kinase and Glucose Partitioning
Figure 6. S11-Ugp1 Phosphorylation Is Required for Growth under Cell Integrity Stress Conditions BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to saturation, and serial dilutions were spotted on (A) a synthetic complete glucose medium plate with 40 mg/ml Congo red and grown for 4 days at 30 C or on (B) a synthetic galactose medium plate containing 1 M sorbitol and grown for 5 days at 38 C. These strains all grow identically on synthetic complete glucose medium lacking Congo red (Rutter et al., 2002). Compare (B) to growth under the same conditions except without sorbitol in Figure 1B.
glycogen hyperaccumulation. These same strains also show a decrease in at least one major cell wall component, b-(1,6)-glucan, and exhibit growth defects under cell integrity stress. We hypothesize, therefore, that isoform 1 makes UDP-glucose preferentially used for storage carbohydrate synthesis while isoform 2 makes UDPglucose bound for structural carbohydrates. A simple homeostatic prediction would be that this phosphorylation event might be stimulated by conditions of cell wall insufficiency. Indeed, when wild-type cells growing in rich medium are subjected to a cell integrity stress (i.e., treatment with SDS, calcofluor white, or chlorpromazine) their Ugp1 composition switches from isoform 1 to predominantly isoform 2 (J.H. Grose, T.L.S., and J.R., unpublished data). This phosphorylated Ugp1 then contributes to an increased production of cell wall, enabling adaptive resistance to the environmental stress. We hypothesize that phosphorylation-regulated UDPglucose partitioning is mediated by changes in the subcellular localization of Ugp1. In yeast, the glycogen production machinery is found diffusely within the cytoplasm (Huh et al., 2003; Kumar et al., 2002), while enzymes involved in cell wall production are localized primarily at the cell periphery (Cid et al., 1995; Montijn et al., 1999). If phosphorylated Ugp1 were preferentially trafficked to the cell periphery, it would be brought into proximity with the enzymes necessary for cell wall production. UDP-glucose produced at this site might then be preferentially encountered by the glucan-producing enzymes relative to glycogen synthase. This preferential interaction may involve direct substrate channeling between Ugp1 and the enzymes involved with cell wall synthesis, wherein UDP-glucose is passed directly from one to the other. Interestingly, channeling of newly synthesized UDP-glucose has also been previously described in liver homogenates, where its synthesis is directly coupled to transport into Golgi vesicles (Persat et al., 1984).
How this phenomenon relates to that described in the present work is unknown, but it will be interesting to determine whether it is regulated in response to metabolic demand. In yeast, chitin synthase 3 (Chs3) has been shown to traffic to the cell membrane in response to various forms of stress, raising the possibility that chitin synthesis might also be controlled by regulated enzyme localization (Valdivia and Schekman, 2003). Substrate channeling has also been demonstrated or proposed in many other biochemical pathways (Ovadi and Srere, 2000). Importantly, this product fate determination is not complete. At any given time, a subset of isoform 1 will be found near the cell membrane. This is apparently adequate to support basal cell wall production, as shown by the normal growth of psk1 psk2 or ugp1-S11A strains under standard laboratory conditions. However, the generation of isoform 2 and its peripheral localization is required to meet the increased demand for structural carbohydrate production due to cell wall damage, as demonstrated by the same strains exhibiting a Galts phenotype. Furthermore, when the Ugp1 of psk1 psk2 strains is artificially targeted to the membrane as shown in Figure 7, the growth defect is ameliorated. We have shown that the phosphorylation of Ugp1 by PASK is important for proper yeast glucose partitioning. PASK is conserved from yeast to man, and although the details are different, PASK is also important for appropriate mammalian glucose partitioning, both at the organismal and cellular level (da Silva Xavier et al., 2004; Wilson et al., 2005). In mammals, glucose partitioning is not only regulated by cell-autonomous signals and responses but is coordinated throughout the organism by hormones such as insulin. We have shown that PASK functions as a glucose sensor in pancreatic b cells, and upon activation by elevated glucose, stimulates preproinsulin gene expression (da Silva Xavier et al., 2004). PASK may also control cellular glucose partitioning in mammals directly
496 Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc.
Molecular Cell PAS Kinase and Glucose Partitioning
Figure 7. Targeting Ugp1 to the Membrane Is Sufficient to Confer Growth under Galts Conditions (A) BHY10 (PSK genotypes as indicated), JRY40, JRY352 (the UGP1:SNC2[85–115] allele encodes Ugp1 with the C-terminal transmembrane segment [residues 85–115] of the yeast synaptobrevin Snc2 genetically fused to its C terminus), JRY356, JRY359 (the UGP1: RAS2[302–322] allele encodes Ugp1 with the C-terminal palmitoylation/prenylation sequence [residues 302–322] of Ras2 genetically fused to its C terminus), and JRY361 (the UGP1:SIP2[1–36] allele encodes Ugp1 with the N-terminal myristoylation sequence [residues 1–36] of Sip2 genetically fused to its N terminus) strains were grown to saturation, and serial dilutions were spotted on a synthetic galactose medium plate and grown for 5 days at 38 C (Galts conditions). The image shown is representative of four independent experiments. (B) The strains shown in (A) were grown to early log phase and assayed for glycogen content as described (Rutter et al., 2002). The average of three experiments ± SD is shown. (C) The strains shown in (A) were grown to saturation and seeded in a minimal galactose plate. Concentrated K1 killer toxin, purified from strain SK-S14a, was spotted, and plates were incubated at 18 C for 7 days. The zone of growth inhibition was measured, and the average of four experiments ± SD is shown.
through phosphorylation and inhibition of glycogen synthase (Wilson et al., 2005). The regulated intracellular trafficking of proteins is widely recognized and appreciated. We have shown in this work that the regulated trafficking of a metabolic enzyme can have profound effects on flux partitioning between competing metabolic pathways. This regulation is part of a homeostatic response system that is initiated in response to, and serves to ameliorate, a cellular defi-
ciency. We think it is likely that this mode of regulation will be found to be important in many aspects of metabolic flux control. EXPERIMENTAL PROCEDURES Yeast Cells, Culture Media, and Materials Saccharomyces cerevisiae strain BHY10 (MATa leu2-3,112::CPY-Inv LEU2 ura3-52 his3-D200 lys2-801 trp1-D901 suc2-D9) (Horazdovsky
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 497
Molecular Cell PAS Kinase and Glucose Partitioning
et al., 1994) was obtained from B. Horazdovsky. Strain JRY40 (BHY10; psk1::HIS3 psk2::NEO) was previously described (Rutter et al., 2002). Strains JRY232 and JRY234 were generated by homologous recombination replacement of the endogenous UGP1 allele with PCR product containing the Ugp1-S11A mutant. The UGP1:SNC2(85–115) allele was generated by genetic fusion of residues 85–115 of Snc2 to the C terminus of Ugp1. The UGP1:RAS2(302–322) allele was generated by genetic fusion of residues 302–322 of Ras2 to the C terminus of Ugp1. The UGP1:SIP2(1–36) allele was generated by genetic fusion of residues 1–36 of Sip2 to the N terminus of Ugp1. Each of these fusions was integrated at the native UGP1 locus. Integrity of recombination was verified by PCR and sequencing. All medium was prepared as described (Sherman, 1991), and cultures were maintained at 30 C unless otherwise stated. Cultures were harvested at mid-log phase (OD600 0.6) for all analyses described. Protein Expression and Purification PASK protein was expressed and purified from baculovirus-infected Sf-9 insect cells as previously described (Rutter et al., 2001). Ugp1 protein was expressed and purified from E. coli as previously described (Rutter et al., 2002). Preparation of Cell-free Extracts Extracts for enzyme assays and ion-exchange chromatography were prepared by pelleting 50 ml of log-phase cells and resuspending the pellet in 10 ml XWA (20 mM HEPES, 10 mM KCl, 1.5 mM MgCl2, 1 mM EDTA, 1 mM EGTA, and 1 mM DTT [pH 7.4]), 50 mM NaCl, 1:300 Protease Inhibitor Cocktail (Sigma) and 1:300 Phosphatase Inhibitor Cocktail 1 (Sigma). Four grams of glass beads were added, and the cells were intermittently vortexed for a total of 3 min. The resulting lysate was centrifuged at 20,000 3 g for 20 min, filtered through a 0.2 mm filter, and assayed immediately or loaded onto an anion exchange column for fractionation. Ion-Exchange Chromatography Cell-free extract or in vitro phosphorylation samples were prepared as described and immediately loaded onto a MonoQ 5/50 GL chromatographic column (Amersham Pharmacia) using an Amersham Pharmacia AKTA FPLC system, equilibrated in XWA + 50 mM NaCl. The column was washed in 5 column volumes of XWA + 50 mM NaCl, and sample was eluted into 0.5 ml fractions over 20 ml at a flow rate of 1 ml/min using a linear gradient of 200–350 mM NaCl in XWA buffer. In Vitro Phosphorylation Purified UGP1 (30 mg) and PAS kinase (0.2 mg) were incubated in the presence of 2 mM ATP, 100 mM KCl, 10 mM MgCl2, and 25 mM HEPES (pH 7.0) at 25 C for the times indicated in the figure. Reactions were stopped by addition of 10 ml XWA + 50 mM NaCl, filtered through a 0.2 mm filter, and run on the FPLC as described above.
EDTA. These samples were then treated with 170 ng of either Pronase E or Proteinase K for the various times as indicated. Following proteolysis, the reactions were precipitated with 1 volume 100% TCA and 8 volumes ice-cold acetone. The precipitant was washed with 100% acetone, dried, resuspended in SDS sample buffer, and separated by SDS-PAGE on 12% gels. The gels were stained by Coomassie brilliant blue and scanned. K1 Killer Toxin Test K1 killer toxin was prepared from killer strain SK-S14a, generously provided by the Bussey lab, as described (Brown et al., 1994). Sensitivity to K1 killer toxin was tested as described (Brown et al., 1994). Briefly, yeast strains to be tested were grown in SD media for 48 hr to saturation. Cells were pelleted and resuspended in SGal medium, then seeded into 1X Halvorson buffered medium at pH 4.7 at a dilution of 1:2000 with 2% galactose as the sole carbon source. Plates were left at room temperature for 3 hr to dry, and 5 uL of concentrated K1 killer toxin was spotted directly onto the surface of the solidified medium. Plates were incubated at 18 C for 4 days, after which 5 uL of fresh concentrated toxin was added and plates were incubated for 3 more days at 18 C. Kill zones were measured as a radius from the outside of the toxin spot to edge of the kill zone. Supplemental Data Supplemental Data include one figure and can be found with this article online at http://www.molecule.org/cgi/content/full/26/4/491/DC1/. ACKNOWLEDGMENTS We thank Steve McKnight, Janet Lindsley, and David Stillman for many helpful discussions and comments on the manuscript. We are grateful to Howard Bussey for provision of the SK-S14a yeast strain, and Alan Tong for technical assistance. This work was supported by a Sara and Frank McKnight Foundation fellowship (J.R.) and a Searle Scholars Award (J.R.). Received: January 2, 2007 Revised: February 26, 2007 Accepted: March 30, 2007 Published: May 24, 2007 REFERENCES Bhattacharya, S., Chen, L., Broach, J.R., and Powers, S. (1995). Ras membrane targeting is essential for glucose signaling but not for viability in yeast. Proc. Natl. Acad. Sci. USA 92, 2984–2988. Boone, C., Sommer, S.S., Hensel, A., and Bussey, H. (1990). Yeast KRE genes provide evidence for a pathway of cell wall beta-glucan assembly. J. Cell Biol. 110, 1833–1843.
Glycogen Determination and Assay of Purified and Extract Ugp1 UGPase activity was determined by the rate of formation of glucose-1phosphate from UDP-glucose in an NADP-linked glucose-6-phosphate dehydrogenase assay as previously described (Daran et al., 1995). The reaction mixture contained 50 mM Tris (pH 8.0), 10 mM dithiothreitol, 10 mM MgCl2, 0.2 mM NADP, 10 mM glucose-1,6biphosphate, 2 mM UDP-glucose, 0.6 U phosphoglucomutase, 0.5 U glucose-6-phosphate dehydrogenase, 10 mM Na-pyrophosphate, and sample (FPLC fraction, whole-cell lysate, or purified protein ± phosphorylation). Assay was read at 340 nm in a 1 cm path cuvette in an Ultraspec 2000 spectrophotometer (Amersham Pharmacia). Glycogen content was assayed as previously described (Rutter et al., 2002).
Boyartchuk, V.L., Ashby, M.N., and Rine, J. (1997). Modulation of Ras and a-factor function by carboxyl-terminal proteolysis. Science 275, 1796–1800.
Partial Proteolysis Purified Ugp1 (30 mg) was subjected to phosphorylation conditions, as described above, either with or without 0.2 mg PAS kinase. After incubation for 1 hr at 25 C, the reaction was stopped by addition of 10 mM
Daran, J.M., Dallies, N., Thines-Sempoux, D., Paquet, V., and Francois, J. (1995). Genetic and biochemical characterization of the UGP1 gene encoding the UDP-glucose pyrophosphorylase from Saccharomyces cerevisiae. Eur. J. Biochem. 233, 520–530.
Brown, J.R.T., Lussier, M., Sdicu, A.-M., and Bussey, H. (1994). The K1 killer toxin: molecular and genetic applications to secretion and cell surface assembly. In Molecular Genetics of Yeast: A Practical Approach, J. Johnston, ed. (Oxford: IRL Press), pp. 217–232. Cid, V.J., Duran, A., del Rey, F., Snyder, M.P., Nombela, C., and Sanchez, M. (1995). Molecular basis of cell integrity and morphogenesis in Saccharomyces cerevisiae. Microbiol. Rev. 59, 345–386. Cohen, P. (1982). The role of protein phosphorylation in neural and hormonal control of cellular activity. Nature 296, 613–620.
498 Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc.
Molecular Cell PAS Kinase and Glucose Partitioning
da Silva Xavier, G., Rutter, J., and Rutter, G.A. (2004). Involvement of Per-Arnt-Sim (PAS) kinase in the stimulation of preproinsulin and pancreatic duodenum homeobox 1 gene expression by glucose. Proc. Natl. Acad. Sci. USA 101, 8319–8324. Published online May 17, 2004. 10.1073/pnas.0307737101. de Groot, P.W., Ruiz, C., Va´zquez de Aldana, C.R., Duenas, E., Cid, V.J., del Rey, F., Rodrı´quez-Pen˜a, J.M., Pe´rez, P., Andel, A., Caubı´n, J., et al. (2001). A genomic approach for the identification and classification of genes involved in cell wall formation and its regulation in Saccharomyces cerevisiae. Comp. Funct. Genomics 2, 124–142. Dong, X., Mitchell, D.A., Lobo, S., Zhao, L., Bartels, D.J., and Deschenes, R.J. (2003). Palmitoylation and plasma membrane localization of Ras2p by a nonclassical trafficking pathway in Saccharomyces cerevisiae. Mol. Cell. Biol. 23, 6574–6584.
myces cerevisiae by affecting cellular histone kinase activity, recombination at rDNA loci, and silencing. J. Biol. Chem. 278, 13390–13397. Published online January 31, 2003. 10.1074/jbc.M212818200. Lussier, M., White, A.M., Sheraton, J., di Paolo, T., Treadwell, J., Southard, S.B., Horenstein, C.I., Chen-Weiner, J., Ram, A.F., Kapteyn, J.C., et al. (1997). Large scale identification of genes involved in cell surface biosynthesis and architecture in Saccharomyces cerevisiae. Genetics 147, 435–450. Montijn, R.C., Vink, E., Muller, W.H., Verkleij, A.J., Van Den Ende, H., Henrissat, B., and Klis, F.M. (1999). Localization of synthesis of beta1,6-glucan in Saccharomyces cerevisiae. J. Bacteriol. 181, 7414–7420. Ovadi, J., and Srere, P.A. (2000). Macromolecular compartmentation and channeling. Int. Rev. Cytol. 192, 255–280.
Dutra, M.B., Silva, J.T., Mattos, D.C., and Panek, A.D. (1996). Regulation of UDPG-pyrophosphorylase isoforms in Saccharomyces cerevisiae and their roles in trehalose metabolism. Biochim. Biophys. Acta 1289, 261–269.
Persat, F., Azzar, G., Martel, M.B., and Got, R. (1984). Evidence for coupling between transport of UDP-glucose and its synthesis by membrane-bound pyrophosphorylase in Golgi apparatus of cat liver. Biochim. Biophys. Acta 769, 377–380.
Francois, J., and Parrou, J.L. (2001). Reserve carbohydrates metabolism in the yeast Saccharomyces cerevisiae. FEMS Microbiol. Rev. 25, 125–145.
Pryciak, P.M., and Huntress, F.A. (1998). Membrane recruitment of the kinase cascade scaffold protein Ste5 by the Gbetagamma complex underlies activation of the yeast pheromone response pathway. Genes Dev. 12, 2684–2697.
Hardy, T.A., and Roach, P.J. (1993). Control of yeast glycogen synthase-2 by COOH-terminal phosphorylation. J. Biol. Chem. 268, 23799–23805. Horazdovsky, B.F., Busch, G.R., and Emr, S.D. (1994). VPS21 encodes a rab5-like GTP binding protein that is required for the sorting of yeast vacuolar proteins. EMBO J. 13, 1297–1309. Huh, W.K., Falvo, J.V., Gerke, L.C., Carroll, A.S., Howson, R.W., Weissman, J.S., and O’Shea, E.K. (2003). Global analysis of protein localization in budding yeast. Nature 425, 686–691. Klis, F.M., Mol, P., Hellingwerf, K., and Brul, S. (2002). Dynamics of cell wall structure in Saccharomyces cerevisiae. FEMS Microbiol. Rev. 26, 239–256. Kopecka, M., and Gabriel, M. (1992). The influence of congo red on the cell wall and (1–3)-beta-D-glucan microfibril biogenesis in Saccharomyces cerevisiae. Arch. Microbiol. 158, 115–126. Kumar, A., Agarwal, S., Heyman, J.A., Matson, S., Heidtman, M., Piccirillo, S., Umansky, L., Drawid, A., Jansen, R., Liu, Y., et al. (2002). Subcellular localization of the yeast proteome. Genes Dev. 16, 707–719. Lerch, K., and Fischer, E.H. (1975). Amino acid sequence of two functional sites in yeast glycogen phosphorylase. Biochemistry 14, 2009– 2014. Levin, D.E. (2005). Cell wall integrity signaling in Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 69, 262–291.
Qadota, H., Python, C.P., Inoue, S.B., Arisawa, M., Anraku, Y., Zheng, Y., Watanabe, T., Levin, D.E., and Ohya, Y. (1996). Identification of yeast Rho1p GTPase as a regulatory subunit of 1,3-beta-glucan synthase. Science 272, 279–281. Rutter, J., Michnoff, C.H., Harper, S.M., Gardner, K.H., and McKnight, S.L. (2001). PAS kinase: an evolutionarily conserved PAS domainregulated serine/threonine kinase. Proc. Natl. Acad. Sci. USA 98, 8991–8996. Rutter, J., Probst, B.L., and McKnight, S.L. (2002). Coordinate regulation of sugar flux and translation by PAS kinase. Cell 111, 17–28. Shahinian, S., and Bussey, H. (2000). beta-1,6-Glucan synthesis in Saccharomyces cerevisiae. Mol. Microbiol. 35, 477–489. Sherman, F. (1991). Getting started with yeast. In Guide to Yeast Genetics and Molecular Biology, G.R. Fink, ed. (San Diego, CA: Academic Press), pp. 3–21. Turnquist, R.L., and Hansen, R.G. (1973). Uridine diphosphoryl glucose pyrophosphorylase. In The Enzymes, P.D. Boyer, ed. (New York: Academic Press), pp. 51–71. Valdivia, R.H., and Schekman, R. (2003). The yeasts Rho1p and Pkc1p regulate the transport of chitin synthase III (Chs3p) from internal stores to the plasma membrane. Proc. Natl. Acad. Sci. USA 100, 10287– 10292. Published online August 19, 2003. 10.1073/pnas.1834246100.
Lin, K., Hwang, P.K., and Fletterick, R.J. (1995). Mechanism of regulation in yeast glycogen phosphorylase. J. Biol. Chem. 270, 26833– 26839.
Wilson, W.A., Skurat, A.V., Probst, B., de Paoli-Roach, A., Roach, P.J., and Rutter, J. (2005). Control of mammalian glycogen synthase by PAS kinase. Proc. Natl. Acad. Sci. USA 102, 16596–16601. Published online November 7, 2005. 10.1073/pnas.0508481102.
Lin, S.S., Manchester, J.K., and Gordon, J.I. (2003). Sip2, an N-myristoylated beta subunit of Snf1 kinase, regulates aging in Saccharo-
Wullschleger, S., Loewith, R., and Hall, M.N. (2006). TOR signaling in growth and metabolism. Cell 124, 471–484.
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 499
Article Regulation of Glucose Partitioning by PAS Kinase and Ugp1 Phosphorylation Tammy L. Smith1 and Jared Rutter1,* 1
Department of Biochemistry, University of Utah School of Medicine, Salt Lake City, UT 84112, USA *Correspondence: [email protected] DOI 10.1016/j.molcel.2007.03.025
SUMMARY
The ability of cells to recognize and respond to specific metabolic deficiencies is required for all forms of life. We have uncovered a system in the yeast S. cerevisiae that, in response to a perceived deficiency in cell wall glucan, alters partitioning of glucose toward glucan synthesis and away from glycogen synthesis. The paralogous yeast PAS kinases Psk1 and Psk2 phosphorylate UDP-glucose pyrophosphorylase (Ugp1), the primary producer of UDP-glucose, the glucose donor for glucan biosynthesis. Unexpectedly, phosphorylation of Ugp1 does not affect its catalytic activity but instead alters the terminal destination of the UDP-glucose it generates. Phosphorylated Ugp1 is required for intensive glucan production, and inability to phosphorylate Ugp1 is associated with a weak cell wall, decreased glucan content, and increased glycogen content. We provide data indicating that phosphorylation by Psk1 or Psk2 targets Ugp1 to the cell periphery, where the UDP-glucose it produces is in proximity to the site of glucan synthesis. We propose that regulation of glucose partitioning by altered enzyme and substrate localization is a rapid and potent response to metabolic deficiency.
INTRODUCTION Glucose is a versatile source of energy used by most living cells. Among other fates, internalized glucose can be used in glycolysis to create usable energy, in glycogen synthesis to store energy, and in the pentose-phosphate pathway to generate NADPH, or it can be used to glycosylate proteins or lipids. Misregulated flux partitioning can lead to an inappropriate surplus of product from one pathway and a deficit of product from another. An essential aspect of the adaptive behavior of cells is the ability to rapidly sense such a deficiency and alter flux to correct it, thus maintaining homeostasis. In multicellular organisms, this
process needs not only to respond to cellular deficiencies, but also to the needs of tissues and the organism as a whole. This is accomplished partially through the hormonal and neural coordination of metabolic supply and demand. Complex human diseases, such as obesity and diabetes mellitus, are fundamentally characterized by a cellular dysregulation of sugar and lipid metabolic flux and flux partitioning. Understanding how flux is managed within individual cells is essential to understanding and managing these organismal diseases. Individual Saccharomyces cerevisiae cells are essentially autonomous, providing a simpler system in which to study metabolic flux partitioning. As with most other cells, yeast cells rely upon appropriate metabolic adaptation to respond to constantly evolving demands. One example of such an adaptive response involves the partitioning of intracellular glucose between storage as glycogen and use in cell wall glucan synthesis (reviewed in Klis et al. [2002]). Actively dividing yeast cells have an enormous demand for cell wall glucan synthesis, as the extensive cell wall needs to be replicated during each round of division. As nutrients are depleted and division slows, yeast adapt glucose partitioning toward glycogen storage in preparation for impending starvation (Francois and Parrou, 2001). In addition to these normal physiological changes in demand, this partitioning is also responsive to other stressors, including heat shock and cell wall stress (Levin, 2005). We have discovered a system that is essential in yeast for appropriate glucose partitioning and for adaptation of that partitioning to changes in demand. The central components of this system are the enzyme UDP-glucose pyrophosphorylase (Ugp1) and its product UDP-glucose, which is the immediate substrate for both glucan and glycogen synthesis (Daran et al., 1995). Herein we describe two isoforms of Ugp1 that differ in the phosphorylation status of a single serine residue (S11). This phosphorylation event, catalyzed by the paralogous PAS kinases Psk1 and Psk2, leads to a conformational change in Ugp1 and is required for appropriate glucose partitioning toward glucan synthesis. We provide evidence that the phosphorylation-induced conformational change targets Ugp1 to the cell periphery, where its product UDP-glucose will be generated in proximity to the site of glucan synthesis (Montijn et al., 1999). This represents a previously undescribed mechanism for cells to direct glucose flux partitioning through localized metabolism of substrates.
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 491
Molecular Cell PAS Kinase and Glucose Partitioning
RESULTS Phosphorylation of Ugp1 at S11 by Psk2 Occurs In Vivo We previously demonstrated that Psk2 efficiently phosphorylates purified Ugp1 in vitro predominantly at serine 11 (S11) (Rutter et al., 2002). To determine the in vivo relevance of this phosphorylation event, we examined wild-type and PAS kinase-deficient (psk1 psk2) strains of Saccharomyces cerevisiae for phosphorylation of endogenous Ugp1 using an S11 Ugp1 phosphospecific antibody. While Ugp1 protein levels are identical in the two strains (Figure 1A, lower panel), only the wild-type strain contains S11-phosphorylated Ugp1 (upper panel). The psk1 psk2 strain is unable to efficiently grow on galactose-containing medium at 38 C (Galts phenotype) (Figure 1B, rows 1 and 2) (Rutter et al., 2002). To determine whether the inability of the psk1 psk2 strain to phosphorylate Ugp1 at S11 was responsible for the Galts phenotype, we generated a strain wherein the endogenous UGP1 allele was replaced with an unphosphorylatable S11A mutant allele of UGP1. The ugp1-S11A strain shows a Galts phenotype identical to the psk1 psk2 strain (Figure 1B, row 3), as does the strain wherein the psk1 psk2 mutations and the ugp1-S11A mutation are combined. Strains containing either a ugp1-S11D or ugp1-S11E allele behaved identically to the ugp1-S11A strain, suggesting that in this context, neither aspartate nor glutamate mimic phosphoserine. Together, these results strongly support the conclusion that phosphorylation of Ugp1 at S11 is biologically relevant and, at least for the Galts phenotype, is the major in vivo function of Psk1 and Psk2. While phosphorylation of Ugp1 S11 was necessary for viability under Galts and other (see below) conditions, the direct biochemical effect of phosphorylation was unknown. Phosphorylation has long been known to directly regulate the catalytic activity of other enzymes of glycogen metabolism, including glycogen synthase (Cohen, 1982; Hardy and Roach, 1993) and glycogen phosphorylase (Lerch and Fischer, 1975; Lin et al., 1995). We therefore sought to determine whether the effect of Ugp1 phosphorylation might be to directly affect the catalytic rate of the enzyme. To test this, we generated phospho-Ugp1 by incubation in the presence of both PAS kinase and ATP for 1 hr at 25 C. These phosphorylation conditions generate stoichiometrically phosphorylated Ugp1 as measured by 32 P incorporation using [g-32P]ATP. We also performed control incubations lacking PAS kinase, ATP, or both. We then measured the catalytic rate for each of these Ugp1 samples. As shown in Figure 2A, none of these treatments altered Ugp1 catalytic rate. We also examined Ugp1 catalytic rate in extracts from the same four strains described above (i.e., wild-type, psk1 psk2, ugp1-S11A, and psk1 psk2 ugp1-S11A). As with Ugp1 phosphorylated in vitro, Ugp1 catalytic rate in these strains was not significantly different (Figure 2B). These data support the unexpected conclusion that while phosphorylation of Ugp1 is biologically important, it does not exert its effects through a change in the catalytic rate of the enzyme.
Figure 1. The Psk Proteins Phosphorylate Ugp1 at Ser11 In Vivo (A) BHY10 (PSK1 PSK2) and JRY40 (psk1 psk2) strains were grown to log phase in synthetic complete medium. Total protein extract was separated and subjected to western blotting using either a Ser11Ugp1 phosphospecific antibody (top panel) or a panspecific Ugp1 antibody (bottom panel). (B) BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to saturation, and 5-fold serial dilutions were spotted on a synthetic galactose medium plate and grown for 5 days at 38 C (Galts conditions).
Phosphorylation Is Necessary and Sufficient for Conformational Change of Ugp1 As we observed that phosphorylation does not exert its biochemical effects through a change in Ugp1 catalytic rate, we tested whether it affected the isoform distribution that has been observed for Ugp1 previously (Dutra et al., 1996). We grew the four strains to log phase in standard minimal yeast medium and produced a clarified extract from each. This extract was separated using anion exchange chromatography, and the resultant fractions were assayed for Ugp1 activity. All strains showed a peak of activity between fractions 5 and 6, but only the wild-type strain showed a second peak in fraction 13 (Figure 3A). When the fractions from the wild-type strain were examined by western blotting, both peaks contained Ugp1 protein, but only the second peak contained S11phosphorylated Ugp1 (Figure 3B). Western blotting of fractions from the three mutant strains showed Ugp1 only in the first peak in correspondence with the enzyme
492 Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc.
Molecular Cell PAS Kinase and Glucose Partitioning
phosphorylation for 120 min (data not shown). We conclude that S11 phosphorylation of Ugp1 is sufficient to cause the transition to isoform 2. We also note that the observed bimodal distribution, most clearly evident in the 30 min time point, strongly suggests a concerted mechanism for this transition. This cooperativity implies a multimeric organization of Ugp1, and Ugp1 from S. cerevisiae is indeed an octamer (data not shown) as it is in most species (Turnquist and Hansen, 1973). To further characterize the nature of the transition from isoform 1 to isoform 2, we examined the gross structure of the two isoforms by limited proteolysis (see Figure S1 in the Supplemental Data available with this article online). Phosphorylated and unphosphorylated Ugp1 were subjected to digestion with either Pronase E or Proteinase K for various times. Two noticeable differences were apparent between the phosphorylated and unphosphorylated proteins, and these differences were observed with both proteases. First, the phosphorylated protein has increased general resistance to proteolysis, as observed by the presence of the intact protein after 40 min of digestion, by which time the unphosphorylated intact protein is gone. Secondly, the patterns of digestion products over the time course are significantly different. These results, and particularly their similarity with two different proteases, strongly suggest that the transition from isoform 1 to isoform 2 is characterized by a concerted change from one distinct conformation to another. Figure 2. Phosphorylation Does Not Affect Ugp1 Enzymatic Activity (A) Purified Ugp1 protein was incubated at 30 C for 1 hr in the presence or absence of purified PASK protein and ATP as indicated. Ugp1 activity was then assayed as described (n = 3). (B) BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to log phase, and a total protein extract was generated. This extract was then assayed for Ugp1 activity as described (n = 4). Error bars indicate standard deviation.
assay data. Therefore, the first peak corresponds to unphosphorylated Ugp1 (isoform 1) and the second peak corresponds to phosphorylated Ugp1 (isoform 2). The in vivo analysis of the four strains suggests that phosphorylation at S11 is required for the transition from isoform 1 to isoform 2. We next sought to determine whether phosphorylation was sufficient for this transition. Purified Ugp1 was phosphorylated in vitro with purified PAS kinase. Samples were taken at various time points and chromatographically analyzed as before. Before phosphorylation (0 min) all of the Ugp1 is found in isoform 1, corresponding to the unphosphorylated enzyme as seen in Figure 3B. After 30 min, a 50% shift to isoform 2 is evident, and by 60 min this shift is essentially complete (Figure 4A). When the fractions from the 30 min time point were examined by western blotting as before, both peaks contained Ugp1, but as with the in vivo experiment only the second peak contained S11-phosphorylated Ugp1 (Figure 4B). Importantly, purified S11A mutant Ugp1 protein was never observed to shift to isoform 2 even after
Phosphorylation of Ugp1 Is Required for Cell Wall Maintenance at the Expense of Glycogen Synthesis Ugp1 catalyzes the conversion of glucose-1-phosphate and UTP to UDP-glucose, which is used in yeast primarily as a glucose donor for the synthesis of storage carbohydrates (such as glycogen) and structural carbohydrates (such as cell wall glucans) (Daran et al., 1995). We hypothesized that the two isoforms of Ugp1 might be differentially responsible for supplying UDP-glucose to these two distinct pathways. If so, an inability to phosphorylate Ugp1 and thereby create isoform 2 might cause inappropriate glucose partitioning toward glycogen at the expense of cell wall glucan synthesis. To test this hypothesis, we determined glycogen levels in the wild-type and mutant strains. As shown in Figure 5A, the three mutant strains showed an 8- to 10-fold increase in glycogen levels relative to the wild-type strain. To test for a coincident decrease in cell wall glucans, we tested the same four strains for sensitivity to K1 killer toxin, a pore-forming protein that kills sensitive cells by binding to b-(1,6)-glucan, resulting in membrane damage and cell death (Shahinian and Bussey, 2000). Yeast cells containing less cell wall b-(1,6)-glucan show decreased sensitivity to this toxin, as evidenced by a smaller kill zone in a plate test (Boone et al., 1990). As shown in Figure 5B, wild-type yeast were significantly more sensitive to K1 killer toxin in a plate test than the three mutant strains. We therefore conclude that inability to phosphorylate S11 of Ugp1
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 493
Molecular Cell PAS Kinase and Glucose Partitioning
Figure 3. S11 Phosphorylation Is Required for Ugp1 Chromatographic Shift BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to log phase, and a total protein extract was generated. This extract was fractionated by MonoQ ionexchange chromatography. (A) The resultant fractions were assayed for Ugp1 activity and are plotted. The data shown are representative of four independent experiments. (B) The fractions from the wild-type (BHY10) strain were separated and subjected to western blotting using either a panspecific Ugp1 antibody (top panel) or a Ser11-Ugp1 phosphospecific antibody (bottom panel).
causes glycogen hyperaccumulation at the expense of synthesis of a key cell wall component, b-(1,6)-glucan. Decreased cell wall b-(1,6)-glucan content could lead to a weaker cell wall and decreased cellular stability. We analyzed the cell wall integrity of the four strains by testing their viability under several conditions of cell wall stress. Congo red, caffeine, and calcofluor white have each been shown to interfere with assembly of various cell wall components (de Groot et al., 2001; Kopecka and Gabriel, 1992). Mutations in the cell wall glucan-producing machinery have been shown to cause altered sensitivity to these agents (Lussier et al., 1997). Consistent with Ugp1 isoform 2 being preferentially involved in cell wall maintenance, the three mutant strains that are unable to generate isoform 2 exhibit marked hypersensitivity to Congo red (Figure 6A) and calcofluor white (data not shown), and resistance to caffeine (data not shown).
As described previously, the mutant strains are unable to grow under Galts conditions. Such a growth phenotype could result from many different primary defects, one being inadequate cell wall glucan production. If the Galts phenotype of the mutant strains is due to inadequate cell wall synthesis, then it should be suppressed by inclusion of the osmotic stabilizer sorbitol. Indeed, sorbitol completely rescues the Galts defect of the mutant strains (Figure 6B). Combined, these data demonstrate that an inability to phosphorylate Ugp1 causes increased glycogen production and compromised cell wall integrity. These results suggest a model wherein phosphorylation of Ugp1 controls glucose partitioning by determining the fate of its product UDP-glucose. Specifically, unphosphorylated Ugp1 (isoform 1) produces UDP-glucose that is preferentially used for glycogen synthesis, and phosphorylated Figure 4. S11 Phosphorylation Is Sufficient for Ugp1 Chromatographic Shift Purified Ugp1 was incubated at 30 C for the times indicated in the presence of purified PASK protein and ATP. These reactions were stopped by the addition of EDTA and fractionated by MonoQ ion-exchange chromatography. (A) The Ugp1 activity of the resultant fractions was assayed and is plotted. The data shown are representative of three independent experiments. (B) The fractions from the 30 min time point were separated and subjected to western blotting using either a panspecific Ugp1 antibody (top panel) or a Ser11-Ugp1 phosphospecific antibody (bottom panel).
494 Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc.
Molecular Cell PAS Kinase and Glucose Partitioning
Figure 5. S11-Ugp1 Phosphorylation Is Required for Normal Glycogen and b-(1,6)-Glucan Synthesis (A) BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to log phase and assayed for glycogen content as described (Rutter et al., 2002). The average of three experiments ± SD is shown. (B) BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to saturation and seeded in a minimal galactose plate. Concentrated K1 killer toxin, purified from strain SK-S14a, was spotted, and plates were incubated at 18 C for 7 days. The zone of growth inhibition was measured, and the average of four experiments ± SD is shown.
Ugp1 (isoform 2) produces UDP-glucose that is preferentially used for glucan synthesis. Cell wall glucan synthesis occurs primarily at the plasma membrane, where the synthetic enzymes for the major cell wall glucans are located (Montijn et al., 1999; Qadota et al., 1996). Conversely, glycogen synthase is localized to the cytoplasm (Huh et al., 2003; Kumar et al., 2002). We hypothesized that the different isoforms of Ugp1 might determine the fate of their product UDP-glucose by being differentially localized to the site of glycogen synthesis (isoform 1) or glucan synthesis (isoform 2). Ideally we would be able to visualize this localization by immunofluorescence with an anti-Ugp1 antibody or anti-epitope antibody to tagged Ugp1 or through a Ugp1-GFP fusion. Unfortunately, many different anti-Ugp1 antibodies proved ineffective for immunocytochemistry and both N- and C-terminal GFP or epitope tags disrupted the proper folding of Ugp1. As an alternative test of the localization hypothesis, we engineered Ugp1 with three independent sequences designed to target it to the plasma
membrane in the absence of phosphorylation. Specifically, the targeting sequences used were the C-terminal transmembrane segment from synaptobrevin (Snc2) (Pryciak and Huntress, 1998), the C-terminal prenylation sequence from Ras2 (Bhattacharya et al., 1995; Boyartchuk et al., 1997; Dong et al., 2003), and the N-terminal myristoylation sequence from Sip2 (Lin et al., 2003). As shown in Figure 7A, each of these targeted Ugp1 variants is able to substantially complement the Galts phenotype of a psk1 psk2 mutant strain. Further, the presence of wildtype PSK1 and PSK2 genes does not increase growth under Galts conditions when the Ugp1 is ectopically targeted to the cell periphery. If the three artificially membrane-associated Ugp1 variants faithfully mimic phospho-Ugp1, then they should correct the glucose partitioning defects found in a psk1 psk2 mutant. Indeed, we found that psk1 psk2 mutant strains bearing any of the three targeted Ugp1 variants had significantly decreased glycogen relative to the psk1 psk2 mutant with wild-type Ugp1 (Figure 7B). Further, b-(1,6)glucan levels were restored to wild-type levels in a psk1 psk2 mutant strain with any of the three targeted Ugp1 variants (Figure 7C). In each of these three assays (Galts suppression, glycogen levels, and glucan levels), the PSK1 PSK2 UGP1:RAS2 and PSK1 PSK2 UGP1:SIP2 strains were similar to the PSK1 PSK2 UGP1:SNC2 strain (data not shown). Combined, these data provide evidence that Ugp1 phosphorylation leads to plasma membrane localization and that this localization is required for increased glucan synthesis at the expense of glycogen synthesis. This increased glucan is then necessary for growth under stress, such as in Galts conditions. DISCUSSION In order to survive, living cells need to constantly monitor their environment and make appropriate metabolic decisions in response to environmental changes. The ability of a cell to rapidly respond to changes in nutritional status and environmental stress confers growth and reproductive advantages (Francois and Parrou, 2001; Wullschleger et al., 2006). We have shown here that an important metabolic decision regarding the fate of intracellular glucose is enacted by phosphorylation of Ugp1 by PAS kinase. This phosphorylation event determines whether UDPglucose is preferentially used to make storage or structural carbohydrates. We made the surprising observation that, while phosphorylation of Ugp1 was essential under certain conditions, it did not affect the enzymatic activity of Ugp1 in vitro or in vivo. We found that phosphorylation, instead, caused a conformation change in Ugp1 that is easily detected using limited proteolysis. Further, unphosphorylated Ugp1 (isoform 1) and phospho-Ugp1 (isoform 2) are readily resolved using ion-exchange chromatography. Cells unable to make isoform 2, either through elimination of both PAS kinase genes or through mutation of the PAS kinase phosphorylation site on Ugp1, exhibit marked
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 495
Molecular Cell PAS Kinase and Glucose Partitioning
Figure 6. S11-Ugp1 Phosphorylation Is Required for Growth under Cell Integrity Stress Conditions BHY10, JRY40, JRY232, and JRY234 (genotypes as indicated) were grown to saturation, and serial dilutions were spotted on (A) a synthetic complete glucose medium plate with 40 mg/ml Congo red and grown for 4 days at 30 C or on (B) a synthetic galactose medium plate containing 1 M sorbitol and grown for 5 days at 38 C. These strains all grow identically on synthetic complete glucose medium lacking Congo red (Rutter et al., 2002). Compare (B) to growth under the same conditions except without sorbitol in Figure 1B.
glycogen hyperaccumulation. These same strains also show a decrease in at least one major cell wall component, b-(1,6)-glucan, and exhibit growth defects under cell integrity stress. We hypothesize, therefore, that isoform 1 makes UDP-glucose preferentially used for storage carbohydrate synthesis while isoform 2 makes UDPglucose bound for structural carbohydrates. A simple homeostatic prediction would be that this phosphorylation event might be stimulated by conditions of cell wall insufficiency. Indeed, when wild-type cells growing in rich medium are subjected to a cell integrity stress (i.e., treatment with SDS, calcofluor white, or chlorpromazine) their Ugp1 composition switches from isoform 1 to predominantly isoform 2 (J.H. Grose, T.L.S., and J.R., unpublished data). This phosphorylated Ugp1 then contributes to an increased production of cell wall, enabling adaptive resistance to the environmental stress. We hypothesize that phosphorylation-regulated UDPglucose partitioning is mediated by changes in the subcellular localization of Ugp1. In yeast, the glycogen production machinery is found diffusely within the cytoplasm (Huh et al., 2003; Kumar et al., 2002), while enzymes involved in cell wall production are localized primarily at the cell periphery (Cid et al., 1995; Montijn et al., 1999). If phosphorylated Ugp1 were preferentially trafficked to the cell periphery, it would be brought into proximity with the enzymes necessary for cell wall production. UDP-glucose produced at this site might then be preferentially encountered by the glucan-producing enzymes relative to glycogen synthase. This preferential interaction may involve direct substrate channeling between Ugp1 and the enzymes involved with cell wall synthesis, wherein UDP-glucose is passed directly from one to the other. Interestingly, channeling of newly synthesized UDP-glucose has also been previously described in liver homogenates, where its synthesis is directly coupled to transport into Golgi vesicles (Persat et al., 1984).
How this phenomenon relates to that described in the present work is unknown, but it will be interesting to determine whether it is regulated in response to metabolic demand. In yeast, chitin synthase 3 (Chs3) has been shown to traffic to the cell membrane in response to various forms of stress, raising the possibility that chitin synthesis might also be controlled by regulated enzyme localization (Valdivia and Schekman, 2003). Substrate channeling has also been demonstrated or proposed in many other biochemical pathways (Ovadi and Srere, 2000). Importantly, this product fate determination is not complete. At any given time, a subset of isoform 1 will be found near the cell membrane. This is apparently adequate to support basal cell wall production, as shown by the normal growth of psk1 psk2 or ugp1-S11A strains under standard laboratory conditions. However, the generation of isoform 2 and its peripheral localization is required to meet the increased demand for structural carbohydrate production due to cell wall damage, as demonstrated by the same strains exhibiting a Galts phenotype. Furthermore, when the Ugp1 of psk1 psk2 strains is artificially targeted to the membrane as shown in Figure 7, the growth defect is ameliorated. We have shown that the phosphorylation of Ugp1 by PASK is important for proper yeast glucose partitioning. PASK is conserved from yeast to man, and although the details are different, PASK is also important for appropriate mammalian glucose partitioning, both at the organismal and cellular level (da Silva Xavier et al., 2004; Wilson et al., 2005). In mammals, glucose partitioning is not only regulated by cell-autonomous signals and responses but is coordinated throughout the organism by hormones such as insulin. We have shown that PASK functions as a glucose sensor in pancreatic b cells, and upon activation by elevated glucose, stimulates preproinsulin gene expression (da Silva Xavier et al., 2004). PASK may also control cellular glucose partitioning in mammals directly
496 Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc.
Molecular Cell PAS Kinase and Glucose Partitioning
Figure 7. Targeting Ugp1 to the Membrane Is Sufficient to Confer Growth under Galts Conditions (A) BHY10 (PSK genotypes as indicated), JRY40, JRY352 (the UGP1:SNC2[85–115] allele encodes Ugp1 with the C-terminal transmembrane segment [residues 85–115] of the yeast synaptobrevin Snc2 genetically fused to its C terminus), JRY356, JRY359 (the UGP1: RAS2[302–322] allele encodes Ugp1 with the C-terminal palmitoylation/prenylation sequence [residues 302–322] of Ras2 genetically fused to its C terminus), and JRY361 (the UGP1:SIP2[1–36] allele encodes Ugp1 with the N-terminal myristoylation sequence [residues 1–36] of Sip2 genetically fused to its N terminus) strains were grown to saturation, and serial dilutions were spotted on a synthetic galactose medium plate and grown for 5 days at 38 C (Galts conditions). The image shown is representative of four independent experiments. (B) The strains shown in (A) were grown to early log phase and assayed for glycogen content as described (Rutter et al., 2002). The average of three experiments ± SD is shown. (C) The strains shown in (A) were grown to saturation and seeded in a minimal galactose plate. Concentrated K1 killer toxin, purified from strain SK-S14a, was spotted, and plates were incubated at 18 C for 7 days. The zone of growth inhibition was measured, and the average of four experiments ± SD is shown.
through phosphorylation and inhibition of glycogen synthase (Wilson et al., 2005). The regulated intracellular trafficking of proteins is widely recognized and appreciated. We have shown in this work that the regulated trafficking of a metabolic enzyme can have profound effects on flux partitioning between competing metabolic pathways. This regulation is part of a homeostatic response system that is initiated in response to, and serves to ameliorate, a cellular defi-
ciency. We think it is likely that this mode of regulation will be found to be important in many aspects of metabolic flux control. EXPERIMENTAL PROCEDURES Yeast Cells, Culture Media, and Materials Saccharomyces cerevisiae strain BHY10 (MATa leu2-3,112::CPY-Inv LEU2 ura3-52 his3-D200 lys2-801 trp1-D901 suc2-D9) (Horazdovsky
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 497
Molecular Cell PAS Kinase and Glucose Partitioning
et al., 1994) was obtained from B. Horazdovsky. Strain JRY40 (BHY10; psk1::HIS3 psk2::NEO) was previously described (Rutter et al., 2002). Strains JRY232 and JRY234 were generated by homologous recombination replacement of the endogenous UGP1 allele with PCR product containing the Ugp1-S11A mutant. The UGP1:SNC2(85–115) allele was generated by genetic fusion of residues 85–115 of Snc2 to the C terminus of Ugp1. The UGP1:RAS2(302–322) allele was generated by genetic fusion of residues 302–322 of Ras2 to the C terminus of Ugp1. The UGP1:SIP2(1–36) allele was generated by genetic fusion of residues 1–36 of Sip2 to the N terminus of Ugp1. Each of these fusions was integrated at the native UGP1 locus. Integrity of recombination was verified by PCR and sequencing. All medium was prepared as described (Sherman, 1991), and cultures were maintained at 30 C unless otherwise stated. Cultures were harvested at mid-log phase (OD600 0.6) for all analyses described. Protein Expression and Purification PASK protein was expressed and purified from baculovirus-infected Sf-9 insect cells as previously described (Rutter et al., 2001). Ugp1 protein was expressed and purified from E. coli as previously described (Rutter et al., 2002). Preparation of Cell-free Extracts Extracts for enzyme assays and ion-exchange chromatography were prepared by pelleting 50 ml of log-phase cells and resuspending the pellet in 10 ml XWA (20 mM HEPES, 10 mM KCl, 1.5 mM MgCl2, 1 mM EDTA, 1 mM EGTA, and 1 mM DTT [pH 7.4]), 50 mM NaCl, 1:300 Protease Inhibitor Cocktail (Sigma) and 1:300 Phosphatase Inhibitor Cocktail 1 (Sigma). Four grams of glass beads were added, and the cells were intermittently vortexed for a total of 3 min. The resulting lysate was centrifuged at 20,000 3 g for 20 min, filtered through a 0.2 mm filter, and assayed immediately or loaded onto an anion exchange column for fractionation. Ion-Exchange Chromatography Cell-free extract or in vitro phosphorylation samples were prepared as described and immediately loaded onto a MonoQ 5/50 GL chromatographic column (Amersham Pharmacia) using an Amersham Pharmacia AKTA FPLC system, equilibrated in XWA + 50 mM NaCl. The column was washed in 5 column volumes of XWA + 50 mM NaCl, and sample was eluted into 0.5 ml fractions over 20 ml at a flow rate of 1 ml/min using a linear gradient of 200–350 mM NaCl in XWA buffer. In Vitro Phosphorylation Purified UGP1 (30 mg) and PAS kinase (0.2 mg) were incubated in the presence of 2 mM ATP, 100 mM KCl, 10 mM MgCl2, and 25 mM HEPES (pH 7.0) at 25 C for the times indicated in the figure. Reactions were stopped by addition of 10 ml XWA + 50 mM NaCl, filtered through a 0.2 mm filter, and run on the FPLC as described above.
EDTA. These samples were then treated with 170 ng of either Pronase E or Proteinase K for the various times as indicated. Following proteolysis, the reactions were precipitated with 1 volume 100% TCA and 8 volumes ice-cold acetone. The precipitant was washed with 100% acetone, dried, resuspended in SDS sample buffer, and separated by SDS-PAGE on 12% gels. The gels were stained by Coomassie brilliant blue and scanned. K1 Killer Toxin Test K1 killer toxin was prepared from killer strain SK-S14a, generously provided by the Bussey lab, as described (Brown et al., 1994). Sensitivity to K1 killer toxin was tested as described (Brown et al., 1994). Briefly, yeast strains to be tested were grown in SD media for 48 hr to saturation. Cells were pelleted and resuspended in SGal medium, then seeded into 1X Halvorson buffered medium at pH 4.7 at a dilution of 1:2000 with 2% galactose as the sole carbon source. Plates were left at room temperature for 3 hr to dry, and 5 uL of concentrated K1 killer toxin was spotted directly onto the surface of the solidified medium. Plates were incubated at 18 C for 4 days, after which 5 uL of fresh concentrated toxin was added and plates were incubated for 3 more days at 18 C. Kill zones were measured as a radius from the outside of the toxin spot to edge of the kill zone. Supplemental Data Supplemental Data include one figure and can be found with this article online at http://www.molecule.org/cgi/content/full/26/4/491/DC1/. ACKNOWLEDGMENTS We thank Steve McKnight, Janet Lindsley, and David Stillman for many helpful discussions and comments on the manuscript. We are grateful to Howard Bussey for provision of the SK-S14a yeast strain, and Alan Tong for technical assistance. This work was supported by a Sara and Frank McKnight Foundation fellowship (J.R.) and a Searle Scholars Award (J.R.). Received: January 2, 2007 Revised: February 26, 2007 Accepted: March 30, 2007 Published: May 24, 2007 REFERENCES Bhattacharya, S., Chen, L., Broach, J.R., and Powers, S. (1995). Ras membrane targeting is essential for glucose signaling but not for viability in yeast. Proc. Natl. Acad. Sci. USA 92, 2984–2988. Boone, C., Sommer, S.S., Hensel, A., and Bussey, H. (1990). Yeast KRE genes provide evidence for a pathway of cell wall beta-glucan assembly. J. Cell Biol. 110, 1833–1843.
Glycogen Determination and Assay of Purified and Extract Ugp1 UGPase activity was determined by the rate of formation of glucose-1phosphate from UDP-glucose in an NADP-linked glucose-6-phosphate dehydrogenase assay as previously described (Daran et al., 1995). The reaction mixture contained 50 mM Tris (pH 8.0), 10 mM dithiothreitol, 10 mM MgCl2, 0.2 mM NADP, 10 mM glucose-1,6biphosphate, 2 mM UDP-glucose, 0.6 U phosphoglucomutase, 0.5 U glucose-6-phosphate dehydrogenase, 10 mM Na-pyrophosphate, and sample (FPLC fraction, whole-cell lysate, or purified protein ± phosphorylation). Assay was read at 340 nm in a 1 cm path cuvette in an Ultraspec 2000 spectrophotometer (Amersham Pharmacia). Glycogen content was assayed as previously described (Rutter et al., 2002).
Boyartchuk, V.L., Ashby, M.N., and Rine, J. (1997). Modulation of Ras and a-factor function by carboxyl-terminal proteolysis. Science 275, 1796–1800.
Partial Proteolysis Purified Ugp1 (30 mg) was subjected to phosphorylation conditions, as described above, either with or without 0.2 mg PAS kinase. After incubation for 1 hr at 25 C, the reaction was stopped by addition of 10 mM
Daran, J.M., Dallies, N., Thines-Sempoux, D., Paquet, V., and Francois, J. (1995). Genetic and biochemical characterization of the UGP1 gene encoding the UDP-glucose pyrophosphorylase from Saccharomyces cerevisiae. Eur. J. Biochem. 233, 520–530.
Brown, J.R.T., Lussier, M., Sdicu, A.-M., and Bussey, H. (1994). The K1 killer toxin: molecular and genetic applications to secretion and cell surface assembly. In Molecular Genetics of Yeast: A Practical Approach, J. Johnston, ed. (Oxford: IRL Press), pp. 217–232. Cid, V.J., Duran, A., del Rey, F., Snyder, M.P., Nombela, C., and Sanchez, M. (1995). Molecular basis of cell integrity and morphogenesis in Saccharomyces cerevisiae. Microbiol. Rev. 59, 345–386. Cohen, P. (1982). The role of protein phosphorylation in neural and hormonal control of cellular activity. Nature 296, 613–620.
498 Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc.
Molecular Cell PAS Kinase and Glucose Partitioning
da Silva Xavier, G., Rutter, J., and Rutter, G.A. (2004). Involvement of Per-Arnt-Sim (PAS) kinase in the stimulation of preproinsulin and pancreatic duodenum homeobox 1 gene expression by glucose. Proc. Natl. Acad. Sci. USA 101, 8319–8324. Published online May 17, 2004. 10.1073/pnas.0307737101. de Groot, P.W., Ruiz, C., Va´zquez de Aldana, C.R., Duenas, E., Cid, V.J., del Rey, F., Rodrı´quez-Pen˜a, J.M., Pe´rez, P., Andel, A., Caubı´n, J., et al. (2001). A genomic approach for the identification and classification of genes involved in cell wall formation and its regulation in Saccharomyces cerevisiae. Comp. Funct. Genomics 2, 124–142. Dong, X., Mitchell, D.A., Lobo, S., Zhao, L., Bartels, D.J., and Deschenes, R.J. (2003). Palmitoylation and plasma membrane localization of Ras2p by a nonclassical trafficking pathway in Saccharomyces cerevisiae. Mol. Cell. Biol. 23, 6574–6584.
myces cerevisiae by affecting cellular histone kinase activity, recombination at rDNA loci, and silencing. J. Biol. Chem. 278, 13390–13397. Published online January 31, 2003. 10.1074/jbc.M212818200. Lussier, M., White, A.M., Sheraton, J., di Paolo, T., Treadwell, J., Southard, S.B., Horenstein, C.I., Chen-Weiner, J., Ram, A.F., Kapteyn, J.C., et al. (1997). Large scale identification of genes involved in cell surface biosynthesis and architecture in Saccharomyces cerevisiae. Genetics 147, 435–450. Montijn, R.C., Vink, E., Muller, W.H., Verkleij, A.J., Van Den Ende, H., Henrissat, B., and Klis, F.M. (1999). Localization of synthesis of beta1,6-glucan in Saccharomyces cerevisiae. J. Bacteriol. 181, 7414–7420. Ovadi, J., and Srere, P.A. (2000). Macromolecular compartmentation and channeling. Int. Rev. Cytol. 192, 255–280.
Dutra, M.B., Silva, J.T., Mattos, D.C., and Panek, A.D. (1996). Regulation of UDPG-pyrophosphorylase isoforms in Saccharomyces cerevisiae and their roles in trehalose metabolism. Biochim. Biophys. Acta 1289, 261–269.
Persat, F., Azzar, G., Martel, M.B., and Got, R. (1984). Evidence for coupling between transport of UDP-glucose and its synthesis by membrane-bound pyrophosphorylase in Golgi apparatus of cat liver. Biochim. Biophys. Acta 769, 377–380.
Francois, J., and Parrou, J.L. (2001). Reserve carbohydrates metabolism in the yeast Saccharomyces cerevisiae. FEMS Microbiol. Rev. 25, 125–145.
Pryciak, P.M., and Huntress, F.A. (1998). Membrane recruitment of the kinase cascade scaffold protein Ste5 by the Gbetagamma complex underlies activation of the yeast pheromone response pathway. Genes Dev. 12, 2684–2697.
Hardy, T.A., and Roach, P.J. (1993). Control of yeast glycogen synthase-2 by COOH-terminal phosphorylation. J. Biol. Chem. 268, 23799–23805. Horazdovsky, B.F., Busch, G.R., and Emr, S.D. (1994). VPS21 encodes a rab5-like GTP binding protein that is required for the sorting of yeast vacuolar proteins. EMBO J. 13, 1297–1309. Huh, W.K., Falvo, J.V., Gerke, L.C., Carroll, A.S., Howson, R.W., Weissman, J.S., and O’Shea, E.K. (2003). Global analysis of protein localization in budding yeast. Nature 425, 686–691. Klis, F.M., Mol, P., Hellingwerf, K., and Brul, S. (2002). Dynamics of cell wall structure in Saccharomyces cerevisiae. FEMS Microbiol. Rev. 26, 239–256. Kopecka, M., and Gabriel, M. (1992). The influence of congo red on the cell wall and (1–3)-beta-D-glucan microfibril biogenesis in Saccharomyces cerevisiae. Arch. Microbiol. 158, 115–126. Kumar, A., Agarwal, S., Heyman, J.A., Matson, S., Heidtman, M., Piccirillo, S., Umansky, L., Drawid, A., Jansen, R., Liu, Y., et al. (2002). Subcellular localization of the yeast proteome. Genes Dev. 16, 707–719. Lerch, K., and Fischer, E.H. (1975). Amino acid sequence of two functional sites in yeast glycogen phosphorylase. Biochemistry 14, 2009– 2014. Levin, D.E. (2005). Cell wall integrity signaling in Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 69, 262–291.
Qadota, H., Python, C.P., Inoue, S.B., Arisawa, M., Anraku, Y., Zheng, Y., Watanabe, T., Levin, D.E., and Ohya, Y. (1996). Identification of yeast Rho1p GTPase as a regulatory subunit of 1,3-beta-glucan synthase. Science 272, 279–281. Rutter, J., Michnoff, C.H., Harper, S.M., Gardner, K.H., and McKnight, S.L. (2001). PAS kinase: an evolutionarily conserved PAS domainregulated serine/threonine kinase. Proc. Natl. Acad. Sci. USA 98, 8991–8996. Rutter, J., Probst, B.L., and McKnight, S.L. (2002). Coordinate regulation of sugar flux and translation by PAS kinase. Cell 111, 17–28. Shahinian, S., and Bussey, H. (2000). beta-1,6-Glucan synthesis in Saccharomyces cerevisiae. Mol. Microbiol. 35, 477–489. Sherman, F. (1991). Getting started with yeast. In Guide to Yeast Genetics and Molecular Biology, G.R. Fink, ed. (San Diego, CA: Academic Press), pp. 3–21. Turnquist, R.L., and Hansen, R.G. (1973). Uridine diphosphoryl glucose pyrophosphorylase. In The Enzymes, P.D. Boyer, ed. (New York: Academic Press), pp. 51–71. Valdivia, R.H., and Schekman, R. (2003). The yeasts Rho1p and Pkc1p regulate the transport of chitin synthase III (Chs3p) from internal stores to the plasma membrane. Proc. Natl. Acad. Sci. USA 100, 10287– 10292. Published online August 19, 2003. 10.1073/pnas.1834246100.
Lin, K., Hwang, P.K., and Fletterick, R.J. (1995). Mechanism of regulation in yeast glycogen phosphorylase. J. Biol. Chem. 270, 26833– 26839.
Wilson, W.A., Skurat, A.V., Probst, B., de Paoli-Roach, A., Roach, P.J., and Rutter, J. (2005). Control of mammalian glycogen synthase by PAS kinase. Proc. Natl. Acad. Sci. USA 102, 16596–16601. Published online November 7, 2005. 10.1073/pnas.0508481102.
Lin, S.S., Manchester, J.K., and Gordon, J.I. (2003). Sip2, an N-myristoylated beta subunit of Snf1 kinase, regulates aging in Saccharo-
Wullschleger, S., Loewith, R., and Hall, M.N. (2006). TOR signaling in growth and metabolism. Cell 124, 471–484.
Molecular Cell 26, 491–499, May 25, 2007 ª2007 Elsevier Inc. 499