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

provided by Elsevier - Publisher Connector

FEBS Letters 582 (2008) 3025–3028

Only plant-type (GLYK) glycerate produce D-glycerate 3-phosphate

Oliver Bartsch, Martin Hagemann, Hermann Bauwe* Department of Plant Physiology, University of Rostock, Albert-Einstein-Straße 3, D-18059 Rostock, Germany

Received 8 July 2008; revised 23 July 2008; accepted 23 July 2008

Available online 7 August 2008

Edited by Ulf-Ingo Flu¨gge

majority of cyanobacteria including all filamentous strains, Abstract D-Glycerate kinases (GK) occur in three phylogenet- ically distinct classes. Class II GKs produce glycerate 2-phos- which are supposed to be close relatives of the endosymbiontic phate, while both class I GK and class III GK (GLYK) are plastid ancestor [14]. thought to produce glycerate 3-phosphate. We report on the Here, we report on the identification of a bacterial-type class identification of a bacterial-type class I GK in the unicellular I GK in the unicellular cyanobacterium Synechocystis sp. cyanobacterium Synechocystis sp. strain PCC 6803 and of a strain PCC 6803 (hereafter Synechocystis) and of a plant-type plant-type GLYK in the filamentous cyanobacterium Nostoc GLYK in the filamentous cyanobacterium Nostoc sp. strain sp. strain PCC 7120. The comparison with other prokaryotic PCC 7120 (hereafter Nostoc). The comparison with other and eukaryotic GKs of both classes shows that glycerate 3-phos- GKs of these two classes shows that 3PGA is produced only phate is produced only by the GLYKs, but, in contrast to current by the GLYKs, but, in contrast to current thinking, not by thinking, not by any of the examined class I . any of the examined class I enzymes. 2008 Federation of European Biochemical Societies. Pub- lished by Elsevier B.V. All rights reserved.

Keywords: Glycerate ; Bacteria; Cyanobacteria; Fungi; 2. Materials and methods Plants 2.1. Expression vectors Escherichia coli strain TG1 was used for all routine DNA manipula- tions. The GK genes were amplified by PCR using cDNA (AtGLYK and OsGLYK, RevertAid cDNA synthesis kit, FERMENTAS) or genomic DNA (ScGLYK, NosGLYK, SynGK, EcoGK1, and EcoGK2), 1. Introduction respectively, with a combination of proof-reading Pfu (FERMENTAS) and Taq-PCR Master Mix Kit (QUIAGEN) and gene-specific primers as listed in Table 1. For AtGLYK and OsGLYK, The diversity of D-glycerate-forming pathways is reflected in sense primers were designed to exclude sequences encoding the chloro- the occurrence of three phylogenetically distinct classes of gly- plast-targeting peptides. PCR-products were first cloned into pGEM-T cerate kinases (GK) [1]. Class I GKs, with only one reported (PROMEGA), excised and then ligated via the introduced flanking exception [2], are thought to act as glycerate 3-kinases [3,4], restriction sites into the expression vector pBAD/His-A (INVITRO- GEN) or, in case of SynGK, pASK-IBA7plus (IBA). generating glycerate 3-phosphate (3PGA) in bacterial gluca- rate and glycolate [5]. Class II GKs form glycerate 2.2. Production of recombinant proteins and isolation of AtGLYK 2-phosphate (2PGA). These enzymes contribute to the use of Expression was induced in E. coli strains LMG194 (pBAD/His-A, 1 1 one-carbon compounds via the cycle of methylotrophic 200 mg l L-arabinose) or BL21 Gold (pASK-IBA7, 0.2 mg l anhy- bacteria [6,7], sugar degradation via the non-phosphorylating drotetracycline). After 5 h, cells were harvested by centrifugation, branch of the Entner–Doudoroff pathway of Archaea [8,9], resuspended in ice-cold 20 mM Na-phosphate buffer (pH 7.8), and son- icated (six 10 s bursts, 90 W, ice-cooling). Supernatants obtained after and serine catabolism in animals [10]. The 3PGA-forming class centrifugation (12000 rpm, 30 min, 4 C) were used for tag-purification III GKs (GLYKs) have been purified from Saccharomyces of the recombinant enzymes according to the provided manufacturer cerevisiae [11] and a number of plants (for example [12]), but protocols. All buffers used for purification, storage, and biochemical identified on the molecular level only recently [1]. GLYK is analyses of SynGK contained 5 mM DTT. Native Arabidopsis GLYK an indispensable of the photorespiratory cycle in was prepared from 200 g leaves as described before but using a 1.6 · 15 cm Blue Sepharose CL-6B column (GE HEALTHCARE) plants. However, photorespiration is not unique to plants for the last purification step [1]. Purity was examined in Tricine–SDS but also occurs in cyanobacteria [13]. Interestingly, some uni- gels [15]. cellular cyanobacteria, such as Synechocystis sp. strain PCC6803, seem to harbour class I GKs. In contrast, putative 2.3. Enzyme assays and kinetic parameters GLYK-homologous enzymes have been predicted in the Enzyme activities were determined at 30 C by coupling turnover to oxidation of NADH in three different assays [12], all initi- ated by the addition of 2 mM (varying concentrations for Km) D-gly- *Corresponding author. Present address: Universita¨t Rostock, Institut cerate in a total volume of 1 ml. Standard assay 1 (ADP formation) fu¨r Biowissenschaften, Abteilung Pflanzenphysiologie, Albert-Ein- contained 100 mM Tricine–KOH (pH 7.8), 5 mM ATP, 6.25 mM stein-Straße 3, D-18051 Rostock, Germany. Fax: +49 (0) 3814986112. MgCl2, 0.2 mM NADH, 2.5 mM phosphoenolpyruvate (PEP), 5 U E-mail address: [email protected] (H. Bauwe). each of (PK) and lactate dehydrogenase (LDH; all en- zymes from SIGMA–ALDRICH). Standard assay 2 (2PGA forma- Abbreviations: 2PGA, glycerate 2-phosphate; 3PGA, glycerate 3- tion) contained 100 mM Tricine–KOH (pH 7.8), 5 mM ATP, phosphate; GK, glycerate kinase; GLYK, class III GK 6.25 mM MgCl2, 0.2 mM NADH, 5 U each of enolase, PK and

0014-5793/$34.00 2008 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2008.07.038 3026 O. Bartsch et al. / FEBS Letters 582 (2008) 3025–3028

Table 1 Oligonucleotides used for PCR amplification (restriction sites underlined) of GK genes from Arabidopsis thaliana (AtGLYK, At1g80380), Oryza sativa (OsGLYK, Os01g0682500), Nostoc sp. PCC 7120 (NosGLYK, CYORF alr2873), Saccharomyces cerevisiae (ScGLYK, YGR205W), Synechocystis sp. PCC 6803 (SynGK, slr1840), and E. coli K12 (EcoGK1, BAE77171; EcoGK2, BAE76292) Gene Direction Sequence AtGLYK Forward 50-CTCGAGACAGTGGATGTCTCTTCGGTGTCAG-30 Reverse 50-CTCGAGTTAGTTTGCGAGTATCGGGTTCCT-30

OsGLYK Forward 50-CTCGAGATCTCCTCCGTCCAGGAC-30 Reverse 50-GGTACCTCATCTACCCCACATAGGATTCCTC-30

ScGLYK Forward 50-CTCGAGCCTTCCTCTTCTTATTTATCCTCCAAG-30 Reverse 50-GGTACCCTATTCAATACATCTCGTTTTCGTAG-30

NosGLYK Forward 50-CTCGAGATGCAAGTTGGGCAGAAAGAAGCG-30 Reverse 50-GGTACCTTACCAATTCGGTTGATAAACTTGAC-30

SynGK Forward 50-GCTCGAGATGCGTCAAATTTTGAT-30 Reverse 50-CGAATTCTCAGTTTTGACTTTGGA-30

EcoGK1 Forward 50-CTCGAGATGAAAATCGTAATCGC-30 Reverse 50-GAATTCTCACCCCGCGTTGCGC-30

EcoGK2 Forward 50-CTCGAGATGAAGATTGTCATTGC-30 Reverse 50-GAATTCTTAGTTTTTAATTCCC-30

LDH. Standard assay 3 (3PGA formation) contained 100 mM Tri- zymes were obtained in good purity (Fig. 1). Only the recom- cine–KOH (pH 7.8), 5 mM ATP, 6.25 mM MgCl2, 0.2 mM NADH, binant putative NosGLYK co-eluted with a second protein, 2.5 mM PEP, 5 U each of 3PGA kinase, glyceraldehyde 3-phosphate which was mass spectrometrically identified as the E. coli dehydrogenase and PK. Addition of PK prevented inhibition of the 60 kDa chaperonin Cpn60 (GroEL, data not shown). 3PGA kinase reaction by ADP. Km values were determined by non-lin- ear regression analysis (GraphPad Prism) from the rates of ADP for- Using assay 1 (ADP formation), which does not discriminate mation (assay 1). between 2PGA or 3PGA formation, respectively, we deter- mined high GK activity for all enzymes including the putative Synechocystis, Nostoc, and rice GKs (Table 2). The products 3. Results and discussion of the respective genes in the genomes of Synechocystis (slr1840), Nostoc (alr2873), and rice (Os01g0682500) are cur- We have earlier identified the GLYKs from Arabidopsis rently annotated as hypothetical proteins or putative phospho- (AtGLYK) and S. cerevisiae (ScGLYK) as members of a novel ribulokinase/uridine kinase-related proteins. We therefore kinase family [1]. To extend these studies, we now examined tested ADP formation with ribulose 5-phosphate and uridine, putative GLYK orthologs from the filamentous cyanobacte- respectively, but could not detect any enzymatic activity even rium Nostoc and the monocot plant Oryza sativa (NosGLYK with relatively high concentrations of these two compounds and OsGLYK, respectively). For functional identification and (data not shown). Our data hence identify these three proteins comparison, respectively, with class I GKs, we included the as functional class I GK (SynGK) and GLYKs (NosGLYK putative GK from the unicellular cyanobacterium Synechocys- and OsGLYK). tis (SynGK) and the two reported GK isoforms from the pro- For SynGK, the initially measured specific GK activity was teobacterium E. coli (EcoGK1, gene glxK; EcoGK2, gene rather low (1.02 lmol min1 mg1 protein), but could be more garK) [16]. The PCR-amplified genes were expressed in than 10-fold elevated by addition of 5 mM dithiothreitol E. coli and the recombinant enzymes tag-purified. Most en- (DTT) to the enzyme assay. This is similar to observations with the phylogenetically unrelated GLYK from maize [17] and indicates that SynGK activity may be affected by the redox status of the cell. In contrast, the addition of DTT had no ef- fect on any of the other examined enzymes. Since the early reports, it is generally assumed that class I GKs and GLYKs produce 3PGA [3,4], while class II GKs form 2PGA [8,18]. However, there is also one report about a 2PGA-forming class I GK in E. coli [2]. In order to possibly clarify these conflicting views, we used two additional assays to differentiate between 2PGA or 3PGA formation. Assay 2 selectively converts 2PGA to phosphoenolpyruvate, while as- say 3 phosphorylates 3PGA to glycerate 1,3-bisphosphate. Fig. 1. Affinity-purified recombinant GKs separated by SDS–PAGE. The respective rates were then compared with ADP formation (M) Molecular size markers; (1) AtGLYK; (2) OsGLYK; (3) rates measured in assay 1. For all GLYKs (AtGLYK, Os- ScGLYK; (4) NosGLYK; (5) SynGK; (6) EcoGK1; (7) EcoGK2. GLYK, ScGLYK, and NosGLYK), the formation of 3PGA O. Bartsch et al. / FEBS Letters 582 (2008) 3025–3028 3027

Table 2 Kinetic parameters and formation 1 1 Enzyme Km (D-glycerate) [mM] Km (ATP) [mM] Vmax [lmol min mg ] Product Class I EcoGK1 0.086 ± 0.001 0.121 ± 0.016 465 ± 10 2PGA EcoGK2 0.056 ± 0.006 0.200 ± 0.014 70.5 ± 9.5 2PGA SynGK 0.043 ± 0.011 0.212 ± 0.033 14.5 ± 2.9 2PGA

Class III NosGLYK 0.104 ± 0.011 0.206 ± 0.005 42.6 ± 10 3PGA ScGLYK 0.208 ± 0.010 0,890 ± 0.043 268 ± 16 3PGA OsGLYK 0.450 ± 0.054 0.812 ± 0.054 295 ± 65 3PGA AtGLYK 0.266 ± 0.003 0.834 ± 0.138 364 ± 54 3PGA AtGLYKnat 0.264 ± 0.036 0.783 ± 0.059 Not determined 3PGA Values show mean ± S.D. from two or three experiments with independent protein preparations. Rates were measured twice at each substrate nat concentration. Kinetic parameters of SynGK were determined after activation with 5 mM DTT. Km values for the native AtGLYK served as a control to exclude effects of the N-terminal His-tag (one experiment only, statistic parameters from non-linear regression). Product 3PGA or 2PGA means that formation of 2PGA (assay 2) or 3PGA (assay 3), respectively, was undetectable with a 20-fold excess (0.8 lmol min1) of the indicated enzyme.

corresponded to the formation of ADP, while no 2PGA for- In conclusion, our data provide first direct evidence for the mation was observed. On the other hand, ADP formation presence of GKs in cyanobacteria, a class I GK in the unicel- matched the production of 2PGA with all examined class I lular cyanobacterium Synechocystis and a GLYK in the fila- GKs (SynGK, EcoGK1, and EcoGK2) and 3PGA was not mentous, heterocyst-forming cyanobacterium Nostoc. The produced by these enzymes. This indicates that only GLYKs formation of 3PGA from D-glycerate and ATP appears as form 3PGA, whereas class I and II GKs are probably all gly- an exclusive property of GLYKs. It is tempting to speculate cerate 2-kinases. that the absence of 2PGA-forming GKs in combination with The examination of kinetic properties (Table 2) shows that the presence of a 3PGA-forming GLYK in advanced photo- the Km values for D-glycerate and ATP for the two recombi- trophic organisms is related to general requirements of photo- nant E. coli GKs, EcoGK1 and EcoGK2, match published synthetic–photorespiratory metabolism. Chloroplasts do not data [2,16] and are very similar to the respective parameters contain phosphoglycerate mutase, hence lacking a complete measured for the Synechocystis GK, SynGK. In contrast to glycolytic pathway and the capacity to convert 2PGA to EcoGK1, EcoGK2 is a relatively labile enzyme [16]. In accor- 3PGA [23]. Absence of this enzyme in chloroplasts may be dance with this earlier observation, we also found that the essential to avoid withdrawal of carbon from the Calvin cycle. activity of EcoGK2 strongly decreased after storage at 4 C The presence of a 3PGA-forming GK in advanced photoauto- for 1 day or prolonged storage at 20 C and could only par- trophic organisms matches this specific metabolic situation. tially be restored by extended incubation with 5 mM DTT. The occurrence of phosphoglycerate mutase in Synechocystis This indicates irreversible modification under these conditions explains the occurrence of a class I GK in this cyanobacterial and may be related the lower apparent specific activity. strain. The affinity of the cyanobacterial GKs to D-glycerate was distinctly higher than that of the plant GKs. This might reflect Acknowledgements: We thank Klaudia Michl for technical assistance the relatively low carbon flux through photorespiratory metab- and Dr. Stefan Mikkat (Medical Faculty, Core Facility Proteome Ana- lytics, University of Rostock) for help in Cpn60 identification. Finan- olism in cyanobacteria, which could result in lower cellular gly- cial support by the Deutsche Forschungsgemeinschaft (Grant BA cerate levels in comparison with plant leaves. Similarly, the Km 1177/5-1) is gratefully acknowledged. values for ATP are much higher for the plant and yeast GKs in comparison with the bacterial GKs. As discussed before for glycerate, this might reflect adaptation to the different cellular References ATP levels. Chloroplasts contain about 1 mM ATP in the light and 0.2 mM in the dark [19]. This range corresponds to the [1] Boldt, R., Edner, C., Kolukisaoglu, U¨ ., Hagemann, M., Weckw- determined K (ATP) for the plant GKs and to the higher erth, W., Wienkoop, S., Morgenthal, K. and Bauwe, H. (2005) D- m Glycerate 3-kinase, the last unknown enzyme in the photorespi- photorespiratory carbon flux in illuminated leaves. In illumi- 1 ratory cycle in Arabidopsis, belongs to a novel kinase family. nated Synechocystis cells, ATP levels of 250 nmol mg chloro- Plant Cell 17, 2413–2420. phyll were detected [20]. Darkened cells contain about 30% less [2] Hubbard, B.K., Koch, M., Palmer, D.R., Babbitt, P.C. and Gerlt, [21], which translates into saturating ATP concentrations of J.A. (1998) Evolution of enzymatic activities in the enolase about 0.5 mM in the light and 0.35 mM in the dark (assuming superfamily: characterization of the (D)-glucarate/galactarate 1 catabolic pathway in Escherichia coli. Biochemistry 37, 14369– 2 mg Chl ml cell volume, unpublished data from our labora- 14375. tory). Distinctly higher ATP concentrations prevail in yeast [3] Doughty, C.C., Hayashi, J.A. and Guenther, H.L. (1966) Puri- cells under carbon-saturated conditions (1.5 mM) and even fication and properties of D-glycerate 3-kinase from Escherichia more (4.5 mM) under carbon-limited conditions, where yeast coli. J. Biol. Chem. 241, 568–572. [4] Aghaie, A., Lechaplais, C., Sirven, P., Tricot, S., Besnard- cells require GK presumably for glycerol degradation [22]. Gonnet, M., Muselet, D., de Berardinis, V., Kreimeyer, A., This also corresponds to the somewhat higher Km (ATP) of Gyapay, G., Salanoubat, M. and Perret, A. (2008) New insights ScGLYK in comparison with NosGLYK. 3028 O. Bartsch et al. / FEBS Letters 582 (2008) 3025–3028

into the alternative D-glucarate degradation pathway. J. Biol. M. (2006) The plant-like C2 glycolate pathway and the bacterial- Chem. 283, 15638–15646. like glycerate cycle cooperate in phosphoglycolate metabolism in [5] Cusa, E., Obradors, N., Baldoma, L., Badia, J. and Aguilar, J. cyanobacteria. Plant Physiol. 142, 333–342. (1999) Genetic analysis of a chromosomal region containing genes [14] Deusch, O., Landan, G., Roettger, M., Gruenheit, N., Kowallik, required for assimilation of allantoin and linked K.V., Allen, J.F., Martin, W. and Dagan, T. (2008) Genes of glyoxylate metabolism in Escherichia coli. J. Bacteriol. 181, cyanobacterial origin in plant nuclear genomes point to a 7479–7484. heterocyst-forming plastid ancestor. Mol. Biol. Evol. 25, 748–761. [6] Yoshida, T., Fukuta, K., Mitsunaga, T., Yamada, H. and Izumi, [15] Scha¨gger, H. (2006) Tricine–SDS–PAGE. Nat. Protoc. 1, 16–22. Y. (1992) Purification and characterization of glycerate kinase [16] Ornston, M.K. and Ornston, L.N. (1969) Two forms of D- from a serine-producing methylotroph, Hyphomicrobium methy- glycerate kinase in Escherichia coli. J. Bacteriol. 97, 1227–1233. lovorum GM2. Eur. J. Biochem. 210, 849–854. [17] Kleczkowski, L.A. and Randall, D.D. (1985) Light and thiol [7] Chistoserdova, L., Chen, S.W., Lapidus, A. and Lidstrom, M.E. activation of maize leaf glycerate kinase – the stimulating effect of (2003) Methylotrophy in Methylobacterium extorquens AM1 from reduced thioredoxins and ATP. Plant Physiol. 79, 274–277. a genomic point of view. J. Bacteriol. 185, 2980–2987. [18] Lamprecht, W., Heinz, F. and Diamantstein, T. (1962) Phos- [8] Kehrer, D., Ahmed, H., Brinkmann, H. and Siebers, B. (2007) phorylation of D-glyceric acid to 2-phospho-D-glyceric acid with Glycerate kinase of the hyperthermophilic archaeon Thermopro- glycerate kinase in the liver. II. Identification of the reaction teus tenax: new insights into the phylogenetic distribution and products by paper chromatography. Hoppe-SeylerÕs Z. Physiol. physiological role of members of the three different glycerate Chem. 328, 204–206. kinase classes. BMC Genom. 8, 301. [19] Kaiser, W.M. and Bassham, J.A. (1979) Light–dark regulation of [9] Reher, M., Bott, M. and Scho¨nheit, P. (2006) Characterization of starch metabolism in chloroplasts. II. Effect of chloroplastic glycerate kinase (2-phosphoglycerate forming), a key enzyme of metabolite levels on the formation of ADP-glucose by chloroplast the nonphosphorylative Entner–Doudoroff pathway, from the extracts. Plant Physiol. 63, 109–113. thermoacidophilic euryarchaeon Picrophilus torridus. FEMS [20] Hisbergues, M., Jeanjean, R., Joset, F., de Marsac, N.T. and Microbiol. Lett. 259, 113–119. Bedu, S. (1999) Protein PII regulates both inorganic carbon and [10] Guo, J.H., Hexige, S., Chen, L., Zhou, G.J., Wang, X., Jiang, nitrate uptake and is modified by a redox signal in Synechocystis J.M., Kong, Y.H., Ji, G.Q., Wu, C.Q., Zhao, S.Y. and Yu, L. PCC 6803. FEBS Lett. 463, 216–220. (2006) Isolation and characterization of the human D-glyceric [21] Lubberding, H.J. and Schroten, W. (1984) The ATP level in the acidemia related glycerate kinase gene GLYCTK1 and its alter- thermophilic cyanobacterium Synechococcus 6716 during light– natively splicing variant GLYCTK2. DNA Seq. 17, 1–7. dark transition and in the presence of some specific inhibitors. [11] Black, S. and Wright, N.G. (1956) Enzymatic formation of FEMS Microbiol. Lett. 22, 93–96. glyceryl and phosphoglyceryl methylthiol esters. J. Biol. Chem. [22] Larsson, C., Nilsson, A., Blomberg, A. and Gustafsson, L. (1997) 221, 171–180. Glycolytic flux is conditionally correlated with ATP concentration [12] Kleczkowski, L.A., Randall, D.D. and Zahler, W.L. (1985) The in Saccharomyces cerevisiae: a chemostat study under carbon- or substrate specificity, kinetics, and mechanism of glycerate-3- nitrogen-limiting conditions. J. Bacteriol. 179, 7243–7250. kinase from spinach leaves. Arch. Biochem. Biophys. 236, 185– [23] Stitt, M. and Ap Rees, T. (1979) Capacities of pea chloroplasts to 194. catalyse the oxidative pentose phosphate pathway and glycolysis. [13] Eisenhut, M., Kahlon, S., Hasse, D., Ewald, R., Lieman-Hurwitz, Phytochemistry 18, 1905–1911. J., Ogawa, T., Ruth, W., Bauwe, H., Kaplan, A. and Hagemann,