Regulation of purine nucleotide biosynthesis: in yeast and beyond

Transcription

1 786 Biochemical Society Transactions (2006) Volume 34, part 5 Regulation of purine nucleotide biosynthesis: in yeast and beyond R.J. Rolfes 1 Department of Biology, Reiss Science Building 406, Georgetown University, Washington, DC , U.S.A. Abstract Purine nucleotides are critically important for the normal functioning of cells due to their myriad of activities. It is important for cells to maintain a balance in the pool sizes of the adenine-containing and guaninecontaining nucleotides, which occurs by a combination of de novo synthesis and salvage pathways that interconvert the purine nucleotides. This review describes the mechanism for regulation of the biosynthetic genes in the yeast Saccharomyces cerevisiae and compares this mechanism with that described in several microbial species. Introduction Purine nucleotides occupy a central position in cellular metabolism. These nucleotides are constituents of the genetic material, participate as cofactors in enzymatic reactions, serve as intracellular and extracellular signals, function as phosphate donors and are the major carriers of cellular energy. Thus they are critical for virtually every aspect of cellular life; however, they are expensive to synthesize and imbalances between the different nucleotide pools can perturb the normal functioning of the cell. Purine biosynthesis Purine nucleotides are maintained in intracellular pools through a combination of de novo synthesis and salvage pathways (Scheme 1). A ten-step de novo biosynthetic pathway converts PRPP (5-phosphoribosyl-α-1-pyrophosphate) into the first purine nucleotide, IMP (inosine monophosphate). IMP is a branch point in the biosynthetic pathway with two additional steps necessary to produce AMP and GMP. The salvage pathways exist to take up extracellular purine bases converting them into nucleotides, to re-utilize nucleosides and nucleotides produced by degradation and to interconvert the pools of adenine-containing nucleotides with guaninecontaining nucleotides. In microbes such as Escherichia coli and Saccharomyces cerevisiae, the inability to biosynthesize purine nucleotides leads to auxotrophy [1,2]. In Drosophila, purine nucleotide synthesis is required for development and metamorphosis [3 6]. In plants such as Arabidopsis and tobacco (Nicotiana tabacum), synthesis of nucleotides is developmentally regulated [7], whereas in the tropical legumes it plays an additional important role in nitrogen storage [7,8]. In humans, disorders in the purine nucleotide biosynthetic and salvage pathways Key words: adenine, BAS1, PHO2, purine nucleotide biosynthesis, transcription, yeast. Abbreviations used: AICAR, 5-amino-4-imidazolecarboxamide ribonucleoside; ChIP, chromatin immunoprecipitation; IMP, inosine monophosphate; PRPP, 5-phosphoribosyl-α-1-pyrophosphate; PRAT, PRPP amidotransferase; SAICAR, 5-amino-4-imidazole-N-succinocarboxamide ribonucleoside. 1 rolfesr@georgetown.edu have devastating consequences, leading to disorders such as SCIDS (severe combined immunodeficiency syndrome), Lesch Nyhan syndrome, mental retardation and autism [9]. Thus the ability to maintain nucleotide pools is extremely important to cells. Regulation of AMP synthesis in S. cerevisiae In yeast, the ADE genes ADE1, ADE2, ADE4, ADE5,7, ADE6, ADE8, ADE12, ADE13, ADE16 and ADE17 encode thede novo pathway leading to IMP and the branch to AMP [2,10], as shown in Scheme 2. The branch from IMP to GMP is encoded by three IMD genes and GUA1 [2,11]. Regulation of the biosynthetic pathway occurs at both the enzymatic and genetic levels. At the enzymatic level, the first enzyme of the de novo pathway glutamine PRAT (PRPP amidotransferase) is feedback-inhibited by the end-products ATP and ADP [12]. At the genetic level, expression of the ADE genes is repressed when cells are grown in an excess of extracellular purines [13 18]; limitation for purines results in transcriptional de-repression of gene expression by 3- to 18-fold [14,19]. The transcriptional activators Bas1 and Pho2 Two trans-acting factors, Bas1 and Pho2 (which is also known as Bas2 and Grf10), are necessary for ADE gene derepression [14,15]. Bas1, a myb-domain-containing protein [20], appears to be a dedicated regulator of the ADE regulon, whereas the homeodomain-containing transcription factor Pho2 participates with Pho4 in the regulation of several PHO genes in the phosphate utilization pathway and it participates with Swi5 to up-regulate the expression of HO [21,22]. A schematic representation of these proteins is shown in Figure 1. In addition to nine ADE genes (the ADE genes listed above except ADE16), Bas1 and Pho2 together also promotethetranscriptionofthehis1, HIS4 and HIS7 genes of the histidine biosynthesis pathway and GLN1, SHM2 and

2 Information Processing and Molecular Signalling 787 Scheme 1 Biosynthetic and salvage pathways for purine nucleotide synthesis The de novo biosynthetic pathway to make IMP consists of ten reactions (1 10), the branch to AMP consists of two reactions (8 and 11) and the branch to GMP consists of two reactions (12 13). The salvage pathway consists of reactions Enzymes catalysing the reactions are numbered as follows: (1) glutamine phosphoribosylpyrophosphate amidotransferase (ADE4); (2) glycinamide ribotide synthase (ADE5); (3) glycinamide ribotide transformylase (ADE8); (4) formylglycinamide synthase (ADE6); (5) aminoimidazole ribotide synthase (ADE7); (6) aminoimidazole ribotide carboxylase (ADE2); (7) succinylaminoimidazolecarboxamide ribotide synthase (ADE1); (8) adenylosuccinate lyase (ADE13); (9) aminoimidazole carboxamide ribotide transformylase (ADE16 and ADE17); (10) IMP cyclohydrolase (ADE16, and ADE17); (11) adenylosuccinate synthase (ADE13); (12) IMP dehydrogenase (IMD2, IMD3 and IMD4); (13) GMP synthase (GUA1); (14) AMP deaminase (AMD1); (15) adenine phosphoribosyltransferase (APT1); and (16) hypoxanthine-guanine phosphoribosyltransferase (HPT1). AIR, 5 -phosphoribosyl-5-aminoimidazole; CAIR, 5 -phosphoribosyl-5-aminoimidazole-4-carboxylate; FAICAR, 5 -phosphoribosyl-4-carboxamide-5- formamidoimidazole; FGAM, 5 -phosphoribosyl-n-formylglycinamidine; FGAR, 5 -phosphoribosyl-n-formylglycinamide; GAR, 5-phosphoribosylglycinamide; PRA, 5-phospho-β-d-ribosylamine; SAMP, adenylosuccinate; THF, tetrahydrofolate. MTD1 genes involved in the synthesis of glutamine, glycine and 10-formyltetrahydrofolate respectively [15,23,24]. Expression of the genes in the ADE regulon is basal and not able to be de-repressed in response to purine limitation in strains with bas1 and pho2 mutations [14], indicating that the regulators act positively to up-regulate transcription. Since transcription of the genes encoding the Bas1 and Pho2 proteins is not regulated by adenine levels [20,25], adenine repression occurs by antagonizing an activity of the Bas1 or Pho2 proteins necessary for forming an activation-competent complex. The intracellular signal Regulation of ADE gene expression is linked to flux through the biosynthetic pathway via the intermediates AICAR (5-amino-4-imidazolecarboxamide ribonucleoside) and SAICAR (5-amino-4-imidazole-N-succinocarboxamide ribonucleoside) [12,26]. These intermediates are thought to indirectly reflect levels of the nucleotide pools, as their concentrations are determined by activity of the de novo pathway that is itself regulated by ATP and ADP (Scheme 2). Under repressing conditions when purines are abundant, activity of PRAT is feedback-inhibited by the end-products ATP and ADP [12], decreasing flux through the de novo pathway and generating a low level of the signalling intermediates. When intracellular ATP and ADP levels decrease, feedback inhibition is relieved, flux through the biosynthetic pathway increases, and the intracellular signal SAICAR/AICAR is generated, leading to ADE gene expression [12]. Thus the intermediates act as small molecule inducers that potentiate the transcriptional activity of the Bas1 and Pho2 regulators. Two complementary studies using protein fusions provided insights into the mechanism for activation [25,27]. In the first study, Bas1 and Pho2 were fused to the LexA DNAbinding domain [25], and in the second study, the proteins were fused to the VP16 activation domain [27]. LexA-Bas1 responded to extracellular adenine and required native Pho2 to promote transcription. Likewise, Pho2-VP16 protein responded to adenine and was dependent on native Bas1. However, both the LexA-Pho2 and Bas1-VP16 proteins promoted constitutive transcription, requiring neither the co-regulator nor the inducer signal. These results provided evidence that the regulatory signal affects the ability of Pho2 to bind DNA and the ability of Bas1 to transactivate. Pho2 has a weak affinity for DNA and requires interaction with co-regulators to bind to DNA. Co-operative DNA binding between Pho2 and Pho4 was detected at PHO5 [28] and between Pho2 and Swi5 was seen at HO [22]. Pho2-binding sites have been mapped immediately adjacent to Bas1- binding sites at HIS4 and ADE5,7 [20,29], but co-operativity was not detected in vitro [12,20]. Although co-operativity was not demonstrated, several lines of evidence indicate that direct protein protein interaction occurs between Bas1 and Pho2: (i) the LexA-Bas1 protein, which requires native Pho2 for adenine-dependent transactivation, could also transactivate in conjunction with mutant Pho2 proteins that lacked functional DNA binding domains [25]; (ii) epitopetagged Bas1 and Pho2 proteins were able to co-immunoprecipitate [30]; (iii) a large internal region of Pho2, from amino acid 112 to 404, was mapped as important for interaction with Bas1 using two-hybrid experiments [30]; and (iv) a cluster of mutations was identified in the PHO2 gene, corresponding to amino acids (Figure 1), that prevented transactivation with one or more of its three co-regulators [31]. Bas1 protein has an internal activation and regulation domain that was mapped by deletion mapping to a region from amino acid 596 to 664 [25,27]. ChIP (chromatin immunoprecipitation) assays demonstrated that Bas1 binds

3 788 Biochemical Society Transactions (2006) Volume 34, part 5 Scheme 2 Regulation of gene expression is linked to metabolic activity of the biosynthetic pathway through the intermediates SAICAR and AICAR Activityof the first enzyme of the de novo pathway is feedback-inhibited by the end products ATP and ADP. Increased levels of the biosynthetic intermediates SAICAR and AICAR signal conditions for increased expression of the ADE genes and other members of the ADE regulon. Figure 1 Schematic representation of the Bas1 and Pho2 proteins The functional domains of the Bas1 and Pho2 proteins are as shown: the DNA-binding domains of the proteins are in a stippled pattern, the activation domains in stripes, and the mapped regions of interactionareshowninblack. constitutively to its cognate binding sites near the core promoters [32]. Under repressing conditions, the activation domain of Bas1 is masked to prevent it from promoting transcription. Formation of the transactivation-competent protein requires interaction of Bas1 with Pho2, which unmasks the activation domain in Bas1. Pho2 also provides its own activation domain [30]; hence, activation of transcription is triggered by the presence of the activation domains from both proteins. Reception of the regulatory signal The biosynthetic intermediates SAICAR and AICAR provide the regulatory signal that increases the ability of Bas1 and Pho2 to interact [25,27]. Although the details of this molecular mechanism remain unclear, two generalized mechanisms for regulation are considered here. In the first case, Bas1 and Pho2 do not directly interact with the signalling intermediates, but instead another protein recognizes the signal. This unidentified regulatory protein could bind or modify Bas1 or Pho2 in a manner that would modulate their ability to form a complex. It is unlikely that a novel regulator would bind DNA, as a detailed analysis of the regulatory region of the ADE5,7 gene failed to identify new activator- or repressor-binding sites [29]. Two of the co-regulators with Pho2, Pho4 and Swi5, are regulated by phosphorylation [33,34]. In both cases, however, the phosphorylation affects the nuclear localization of the factors, a situation that does not occur with Bas1. An alternative indirect model, similar to the mechanism for the regulation of Gal4 by Gal80 [35], could involve a regulator that does not bind DNA but does bind Bas1 or Pho2. However, extensive genetic screens have not yielded new regulatory proteins ([36]; J.L. Urbanowski and R.J. Rolfes, unpublished work). In contrast, the regulatory signal could interact directly with Bas1, Pho2 or at an interface between the factors. In this mechanism, the small molecule could initiate a conformational change in one of the activators that would allow recognition by the co-regulator or it could provide critical stabilizing contacts necessary to form the complex. Several examples of direct, small molecule interaction with transcription factors have been described in yeast [37]. Regulation in other organisms The regulation of purine nucleotide biosynthesis has been described for a few other organisms, and they will be briefly

4 Information Processing and Molecular Signalling 789 considered here for contrasting with the yeast mechanism. For several microbial species, the regulatory signals and the protein factors have been described in detail. There is a surprisingly wide variety of mechanisms and signalling molecules used in these systems. Finally, this mini review cannot do justice to the large amount of data in these systems and readers are referred to several excellent review papers for additional information [1,38,39]. E. coli and Salmonella Typhimurium The details of the regulatory mechanism for purine nucleotide synthesis was first described in E. coli, and the mechanism is conserved in the closely related species S. Typhimurium (reviewed in [1]). The pur genes, which encode the biosynthetic enzymes, are scattered throughout the genome in small operons or as single genes. Regulation of the pur genes requires the repressor protein PurR, a member of the LacI family of repressors [1,40]. PurR binds to an operator sequence to prevent promoter recognition by RNA polymerase under conditions of sufficient cellular purine pools [1,40]. Association of the PurR repressor with DNA requires the purine bases hypoxanthine or guanine [41,42]. Bacillus subtilis The regulatory mechanism in the Gram-positive bacterium B. subtilis is different from that in E. coli (reviewed in [38]). In B. subtilis, thepur genes that encode the enzymes to synthesize IMP are primarily organized in a single operon with isolated genes encoding the branches to AMP and GMP. Transcription of the operon is regulated by two separate mechanisms, repression and termination anti-termination. Repression of transcription initiation requires the PurR repressor, a protein that is not homologous with the E. coli PurR protein [43]. PurR responds to the intracellular levels of PRPP. PRPP is a substrate in both the first step of the de novo pathway as well as in the uptake/salvage pathway, and its synthesis is feedback-inhibited by ADP. Thus its accumulation is an indication of decreased ADP and exogenous purine bases. Interaction of PRPP with PurR promotes the dissociation of the repressor from DNA, allowing transcription to initiate [43]. A mechanism of termination/anti-termination that responds specifically to guanine nucleotide levels [44] is also found in Bacillus. In an interesting, protein-free mechanism, the 5 -untranslated leader sequence of the pur operon binds the purine base guanine directly [45]. Binding of the base alters secondary structures formed in the 5 -untranslated region of the RNA leading to transcriptional termination prior to the first open reading frame. The existence of the two mechanisms together allows very tight control on gene expression. Lactococcus lactis The pur genes in the lactic acid bacterium L. lactis are found in small operons (reviewed in [39]), in contrast with the gene organization in B. subtilis. L. lactis regulates the de novo synthesis of purine nucleotides using a regulatory protein PurR that is homologous with the B. subtilis PurR protein [46]. The L. lactis PurR protein also binds PRPP; however, it acts as a transcriptional activator rather than as a repressor, in a manner opposite to that of the homologous B. subtilis protein. L. lactis PurR binds DNA independently of the PRPP levels, but interaction with RNA polymerase requires the PurR PRPP complex. Higher eukaryotes The mechanisms that regulate purine nucleotide synthesis in complex eukaryotes, such as Drosophila, plants and humans, have not yet been determined. Experiments indicate that expression of the biosynthetic enzymes is under housekeeping control in most tissues and is up-regulated in response to specific developmental signals in particular tissues. As the microbial systems have demonstrated, there is a wide range of signals and mechanisms that can be used to sense purine nucleotide levels and regulate gene expression. Conclusions and future directions in Saccharomyces In the S. cerevisiae system, positive-acting regulatory proteins Bas1 and Pho2 as well as the biosynthetic intermediates SAICAR and/or AICAR are required to de-repress expressionoftheade genes. SAICAR and AICAR were identified as regulators through genetic experiments that blocked their formation or prevented their further metabolism to nucleotides [12,26]. It is likely that one or both of these molecules serve as the intracellular signal; however, it is possible that these intermediates are converted into a different molecule that is recognized instead. Furthermore, it will be important to demonstrate what protein is actually receiving the regulatory signal, whether it is the transcription factors or another protein. Bas1 and Pho2 have been shown to co-regulate a number of genes involved in purine and histidine synthesis as well as in one-carbon metabolism [13 19,23,24]. Two genome-wide studies have investigated the binding of the Bas1 protein to DNA microarrays after ChIP [47,48]. Most of the identified binding sites were the same in the two studies, although there were some differences. It is important to determine if Bas1 is bound and constitutively active at these loci or whether its activity is regulated. If its activity is regulated, is Pho2 the co-regulator or is there another factor that interacts with Bas1? This question is particularly important in considering the genes involved in one-carbon metabolism. Two DNA microarray studies investigated the transcriptional response to glycine involved in one-carbon metabolism [49,50]. The identified genes, including GCV1, GCV2 and GCV3 genes and SHM2, are up-regulated in response to the addition of glycine through a decrease in the cytosolic 5,10-CH 2 - tetrahydrofolate levels [51]. These genes were identified as binding Bas1 in the ChIP-chip experiments [47,48]. One of the studies found that Bas1 is important for the glycine response [50], whereas the other study found no dependence [49]. Subramanian et al. [50] concluded that a novel

5 790 Biochemical Society Transactions (2006) Volume 34, part 5 co-regulator potentiates the transactivation function of bound Bas1 and both groups appear to be trying to identify this protein [49,50]. In a mechanism reminiscent of that proposed for Bas1 with Pho2, transcriptional activation of the one-carbon metabolism regulon depends on Bas1 with the novel co-regulator and a metabolic intermediate, 5,10- CH 2 -tetrahydrofolate, which might directly interact with the transcription factors. References 1 Zalkin, H. and Nygaard, P. (1996) in Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd edn (Neidhard, F.C., ed.), pp , American Society for Microbiology Press, Washington, DC 2 Jones, E.W. and Fink, G.R. (1982) in The Molecular Biology of the Yeast Saccharomyces cerevisiae: Metabolism and Gene Expression (Strathern, J.N., Jones, E.W. and Broach, J.R., eds.), pp , Cold Spring Harbor Laboratory Press, Plainview 3 Ji, Y. and Clark, D.V. (2006) Genetics 172, Tiong, S.Y.K. and Nash, D. (1990) Genetics 124, Tiong, S.Y.K. and Nash, D. (1993) Genome 36, O Donnell, A.F., Tiong, S., Nash, D. and Clark, D.V. (2000) Genetics 154, Zrenner, R., Stitt, M., Sonnewald, U. and Boldt, R. (2006) Annu. Rev. Plant Biol. 57, Smith, P.M.C. and Atkins, C.A. (2002) Plant Physiol. 128, Van den Berghe, G. (2000) Adv. Exp. Med. Biol. 486, Tibbetts, A.S. and Appling, D.R. (1997) Arch. Biochem. Biophys. 340, Escobar-Henriques, M. and Daignan-Fornier, B. (2001) J. Biol. Chem. 276, Rébora, K., Desmoucelles, C., Borne, F., Pinson, B. and Daignan-Fornier, B. (2001) Mol. Cell. Biol. 21, Mäntsälä, P. and Zalkin, H. (1984) J. Biol. Chem. 259, Daignan-Fornier, B. and Fink, G.R. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, Denis, V., Boucherie, H., Monribot, C. and Daignan-Fornier, B. (1998) Mol. Microbiol. 30, Gedvilaite, A. and Sasnauskas, K. (1994) Curr. Genet. 25, Giani, S., Manoni, M. and Breviario, D. (1991) Gene 107, Stotz, A., Muller, P.P. and Linder, P. (1993) Curr. Genet. 24, Rolfes, R.J. and Hinnebusch, A.G. (1993) Mol. Cell. Biol. 13, Tice-Baldwin, K., Fink, G.R. and Arndt, K.T. (1989) Science 246, Oshima, Y. (1982) in Molecular Biology of the Yeast Saccharomyces cerevisiae: Metabolism and Gene Expression (Strathern, J.N., Jones, E.W. and Broach, J.R., eds.), pp , Cold Spring Harbor Laboratory Press, Plainview 22 Brazas, R.M. and Stillman, D.J. (1993) Mol. Cell. Biol. 13, Arndt, K.T., Styles, C. and Fink, G.R. (1987) Science 237, Denis, V. and Daignan-Fornier, B. (1998) Mol. Gen. Genet. 259, Zhang, F., Kirouac, M., Zhu, N., Hinnebusch, A.G. and Rolfes, R.J. (1997) Mol. Cell. Biol. 17, Rébora, K., Laloo, B. and Daignan-Fornier, B. (2005) Genetics 170, Pinson, B., Kongsrud, T.L., Ording, E., Johansen, L., Daignan-Fornier, B. and Gabrielsen, O.S. (2000) Nucleic Acids Res. 28, Barbaric, S., Münsterkötter, M., Svaren, J. and Hörz, W. (1996) Nucleic Acids Res. 24, Rolfes, R.J., Zhang, F. and Hinnebusch, A.G. (1997) J. Biol. Chem. 272, Hannum, C., Kulaeva, O.I., Sun, H., Urbanowski, J.L., Wendus, A., Stillman, D.J. and Rolfes, R.J. (2002) J. Biol. Chem. 277, Bhoite, L.T., Allen, J., Garcia, E., Thomas, L.R., Gregory, I.D., Voth, W.P., Whelihan, K., Rolfes, R.J. and Stillman, D.J. (2002) J. Biol. Chem. 277, Som, I., Mitsch, R.N., Urbanowski, J.L. and Rolfes, R.J. (2005) Eukaryot. Cell 4, Komeili, A. and O Shea, E. (1999) Science 284, Moll, T., Tebb, G., Surana, U., Robitsch, H. and Nasmyth, K. (1991) Cell 66, Leuther, K.K. and Johnston, S.A. (1992) Science 256, Guetsova, M.L., Lecoq, K. and Daignan-Fornier, B. (1997) Genetics 147, Sellick, C.A. and Reece, R.J. (2005) Trends Biochem. Sci. 30, Zalkin, H. (1991) Res. Microbiol. 142, Kilstrup, M., Hammer, K., Jensen, P.R. and Martinussen, J. (2005) FEMS Microbiol. Rev. 29, Rolfes, R.J. and Zalkin, H. (1988) J. Biol. Chem. 263, Rolfes, R.J. and Zalkin, H. (1990) J. Bacteriol. 172, Meng, L. and Nygaard, P. (1990) Mol. Microbiol. 4, Weng, M., Nagy, P.L. and Zalkin, H. (1995) Proc. Natl. Acad. Sci. U.S.A. 92, Ebbole, D.J. and Zalkin, H. (1987) J. Biol. Chem. 262, Mandel, M., Boese, B., Barrick, J., Winkler, W. and Breaker, R. (2003) Cell 113, Kilstrup, M. and Martinussen, J. (1998) J. Bacteriol. 180, Lee, T.I., Rinaldi, N.J., Robert, F., Odom, D.T., Bar-Joseph, Z., Gerber, G.K., Hannett, N.M., Harbison, C.T., Thompson, C.M., Simon, I. et al. (2002) Science 298, Mieczkowski, P., Domisnska, M., Buck, M., Gerton, J., Lieb, J. and Petes, T. (2006) Mol. Cell. Biol. 26, Gelling, C.L., Piper, M.D.W., Hong, S.P., Kornfeld, G.D. and Dawes, I.W. (2004) J. Biol. Chem. 279, Subramanian, M., Qiao, W.-B., Khanam, N., Wilkins, O., Der, S.D., Lalich, J.D. and Bognar, A.L. (2005) Mol. Microbiol. 57, Piper, M.D., Hong, S.P., Ball, G.E. and Dawes, I.W. (2000) J. Biol. Chem. 275, Received 1 August 2006