Alternative splicing produces an H-protein with better substrate properties for the P-protein of glycine decarboxylase Dirk Hasse 1 , Stefan Mikkat 2 , Martin Hagemann 1 and Hermann Bauwe 1 1 Department of Plant Physiology, University of Rostock, Germany 2 Core Facility Proteome Analysis, Medical Faculty, University of Rostock, Germany Introduction Several thousand plant genes are known to produce multiple transcripts, but the precise function of most of the alternatively encoded proteins is not known [1]. Alternative splicing has also been observed for H-pro- tein, a component of the ubiquitous multi-enzyme sys- tem glycine decarboxylase (GDC, EC 1.4.4.2) [2,3]. GDC is essential for photorespiratory and one-carbon metabolism, and the vital function of this enzyme complex is indicated by the fact that its inactivation is fatal to plants [4] and animals [5]. The GDC H-protein (GLDH) has no catalytic activ- ity itself but interacts via its lipoyl arm one after the other with the three other GDC subunits, P-, T- and L-protein. As a result of the GDC reaction cycle, which requires the coenzymes NAD + and tetrahydro- folate, methylene-tetrahydrofolate and NADH are generated and CO 2 and NH 3 are released [6]. Plant H-proteins are often encoded by small multi- gene families, and the individual H-proteins suppos- edly fulfil various functions in plant metabolism [7–9]. Some plants, however, harbour only one H-protein gene, and we have previously reported that C 4 species of the genus Flaveria produce two H-proteins in an organ-dependent manner by alternative 3¢ splice site selection [2,10]. The alternatively encoded H-protein, FtGLDH AA , harbours two additional alanine residues very close to its N-terminus, and is by far the most dominant H-protein type in leaf mitochondria of these plants. Because of photorespiratory metabolism, leaf mitochondria contain much more GDC than root Keywords alternative splicing; glycine decarboxylase; H-protein; photorespiration Correspondence H. Bauwe, Department of Plant Physiology, University of Rostock, Albert-Einstein- Straße 3, D-18059 Rostock, Germany Fax: +49 381 498 6112 Tel: +49 381 498 6110 E-mail: hermann.bauwe@uni-rostock.de Website: http://www.biologie.uni-rostock.de/ pflanzenphysiologie (Received 7 July 2009, revised 5 September 2009, accepted 25 September 2009) doi:10.1111/j.1742-4658.2009.07406.x Several thousand plant genes are known to produce multiple transcripts, but the precise function of most of the alternatively encoded proteins is not known. Alternative splicing has been reported for the H-protein subunit of glycine decarboxylase in the genus Flaveria. H-protein has no catalytic activity itself but is a substrate of the three enzymatically active subunits, P-, T- and L-protein. In C 4 species of Flaveria, two H-proteins originate from single genes in an organ-dependent manner. Here, we report on differ- ences between the two alternative H-protein variants with respect to their interaction with the glycine-decarboxylating subunit, P-protein. Steady-state kinetic analyses of the alternative Flaveria H-proteins and artificially pro- duced ‘alternative’ Arabidopsis H-proteins, using either pea mitochondrial matrix extracts or recombinant cyanobacterial P-protein, consistently dem- onstrate that the alternative insertion of two alanine residues at the N-ter- minus of the H-protein elevates the activity of P-protein by 20% in vitro, and could promote glycine decarboxylase activity in vivo. Abbreviations GDC, glycine decarboxylase; GLDH, GDC H-protein; GLDH AA, alternative H-protein; LplA, lipoate–protein ligase. FEBS Journal 276 (2009) 6985–6991 ª 2009 The Authors Journal compilation ª 2009 FEBS 6985 mitochondria [11,12], but the relevance of alternative splicing for cellular metabolism in these organs is not known. An inspection of published structural data [13,14] revealed that the N-terminus of the H-protein is close to its lipoyl arm (Fig. S1). Hence, structural alterations to this region could possibly provide GDC variants that are fine-tuned for different metabolic environments. In this study, we examined whether the changes to the N-terminus of Flaveria H-protein alter its reactivity with P-protein, the actual glycine-decarboxylating sub- unit of GDC. These experiments revealed that P-pro- tein showed significantly higher activity with the alternative H-protein than with the normal H-protein. This was also found to occur with P-protein from Pisum sativum (pea) leaf mitochondrial matrix extracts as a source of native plant P-protein and with recom- binant P-protein of the cyanobacterium Synechocystis sp. strain PCC 6803 (referred to simply as Synecho- cystis below). Further support came from experiments with an engineered ‘alternative’ H-protein from Ara- bidopsis thaliana,aC 3 plant. This artificially modified H-protein also showed higher activity with P-protein in comparison with the native H-protein. Hence, both natural and artificial N-terminal insertion of two ala- nyl residues result in higher P-protein activities in vitro, and could increase GDC activity in vivo. Results and Discussion The two alternative H-proteins from the C 4 plant Flaveria trinervia, FtGLDH and FtGLDH AA , were overexpressed in Escherichia coli and obtained as apparently pure recombinant proteins (Fig. 1). To be functional, H-protein must be lipoylated at a specific lysine residue [12,15]. While complete lipoylation can be achieved by the addition of lipoic acid to the E. coli growth medium in some expression systems [16], we observed that lipoylation of H-protein was still incom- plete in our system (Fig. S2). Therefore, all recombi- nant H-proteins used in our study were treated with E. coli lipoate–protein ligase A (LplA, EC 2.7.7.63) [17]. Complete lipoylation was verified by native poly- acrylamide electrophoresis in combination with use of lipoate-specific antibodies, and further confirmed by MALDI-TOF MS (Fig. 2). Leaf mitochondria contain large amounts of all GDC subunits and provide a convenient source of native P-protein. Under in vitro assay conditions, P-protein dissociates from the other GDC subunits and its activity becomes sensitive to the addition of extra H-protein [18,19]. Likewise, addition of 50 lm recombinant FtGLDH to assays containing matrix extracts prepared from purified pea leaf mitochondria resulted in an approximately 50-fold stimulation of the glycine–bicarbonate exchange rate [20] relative to the rate measured in the absence of exogenous H-protein (Table 1). Similar to previously reported data [21,22], the glycine saturation kinetics were hyperbolic, with a K m of 6 mm (Fig. S3). We next examined the reactivity of the two F. triner- via H-proteins in this system (Table 1). H-protein satu- ration followed Michaelis–Menten kinetics for both FtGLDH and FtGLDH AA . The K m values for H-pro- tein were between 16 and 20 lm, and hence somewhat larger than published values for H-proteins from other sources [18,23]. This difference could be due, at least in part, to the small amounts of pea H-protein intro- duced into the assay with the mitochondrial extracts. Notably, the V max values were significantly higher with FtGLDH AA than they were with FtGLDH, indicating that alternative splicing could produce a more efficient H-protein. This possibility was examined in a more precisely defined assay using recombinant cyanobacte- rial P-protein. As eukaryotic P-proteins cannot yet be produced by recombinant DNA technology, we used the P-protein from the cyanobacterium Synechocystis, which is structurally very similar to eukaryotic P-proteins. Moreover, recombinant Synechocystis P-protein is enzymatically active and can be produced in large enough quantities (lane 6 in Fig. 1). The enzymatic activities obtained with recombinant plant H-proteins as substrates were in a similar range to those measured with Synechocystis H-protein, which served as an inter- nal control (Table 1). This extends an earlier report on the use of chicken H-protein as a substrate for the P-protein from Arthrobacter globiformis [3,24]. In Fig. 1. SDS–PAGE with recombinant H-proteins from Flave- ria trinervia, Arabidopsis thaliana and Synechocystis, and recombi- nant P-protein from Synechocystis. M, size markers (kDa); lane 1, FtGLDH; lane 2, FtGLDH AA ; lane 3, AtGLDH1; lane 4, AtGLDH1 AA ; lane 5, Synechocystis H-protein; lane 6, Synechocystis P-protein. Alternative splicing of glycine decarboxylase H-protein D. Hasse et al. 6986 FEBS Journal 276 (2009) 6985–6991 ª 2009 The Authors Journal compilation ª 2009 FEBS comparison with Synechocystis H-protein, the plant H-proteins showed lower affinities for the prokaryotic P-protein; however, similar to our experiments with mitochondrial matrix extracts, the P-protein activity with saturating FtGLDH AA was about 20% higher than that with saturating normal FtGLDH. In order to determine whether this finding can be generalized beyond Flaveria H-proteins, we engineered a splice variant of the H-protein AtGLDH1 from the C 3 plant A. thaliana (AtGLDH1 AA ). This artificial ‘alternative’ H-protein differs from the naturally occur- ring Arabidopsis H-protein only by the insertion of two additional alanyl residues near the N-terminus. Kinetic analysis of the these two H-proteins fully con- firmed our findings with the two Flaveria H-proteins, i.e. the AtGLDH1 AA -saturated specific reaction rate was significantly higher than that with the normal AtGLDH1 (Table 1). This corroborated our view that the alternatively spliced Flaveria H-protein is a better substrate for P-protein, and indicated that this improvement is exclusively caused by a slight modifica- tion to its N-terminus. It should be noted that our method for producing recombinant H-proteins leaves a small extension of four amino acid residues at the N-terminus of each recombinant protein after removal of the His tag (Fig. S4). The kinetic differences between the normal and alternative H-proteins, either from Flaveria or from Arabidopsis, were very consistent. Nonetheless, to fully exclude the possibility that this artificial extension could bias our results, all four plant H-proteins were re-cloned into the tagless expression vector pET-3a. The biochemical parameters obtained with these tagless H-proteins were almost identical to those for the respective thrombin-cleaved H-proteins (Table 1). This confirmed that the higher V max values measured Fig. 2. Correct structure of H-proteins was confirmed by MALDI- TOF MS, with differences between theoretical and measured molecular masses always below 1 Da. The signal at m ⁄ z 14 230 corresponds to the singly charged ion of lipoylated mature Flave- ria trinervia H-protein GLDH (mass shift of 188 Da relative to unli- poylated H-protein). H-protein FtGLDH AA shows the expected mass increase of 142 Da, which is due to the insertion of two extra alanyl residues. Lipoylated mature Arabidopsis thaliana H-protein GLDH1 also shows the expected signal at m ⁄ z 14 417 Da. The Arabidopsis H-protein variant AtGLDH1 AA was produced by mutagenesis, and the two extra alanyl residues result in the calculated mass increase. Small satellite peaks and shoulders are artefacts that originated during sample preparation or the desorption ⁄ ionization process. Table 1. Kinetic parameters for Pisum sativum and Synechocystis P-protein with H-protein from Flaveria trinervia, Arabidopsis thaliana or Synechocystis as substrate. H-proteins carrying an N-terminal extension of four amino acids, which remain after thrombin cleav- age of the His tag, are labelled ‘+4’. Values are means ± SE from three (mitochondrial matrix P-protein) or four (purified Synechocys- tis P-protein) independent experiments. K m (H-protein) (lM) V max (nmolÆCO 2 Æmg )1 Æmin )1 ) Mitochondrial P-protein FtGLDH +4 20.4 ± 2.1 9.4 ± 0.1 FtGLDH AA +4 15.9 ± 1.6 11.2 ± 0.4 AtGLDH1 + 4 23.5 ± 3.1 9.8 ± 0.5 AtGLDH1 AA +4 26.7 ± 3.0 12.2 ± 0.6 No exogenous protein – 0.2 ± 0.03 Synechocystis P-protein SyGLDH +4 1.6 ± 0.1 5.5 ± 0.12 FtGLDH +4 106 ± 17 3.4 ± 0.11 FtGLDH AA +4 93 ± 15 4.1 ± 0.07 AtGLDH1 + 4 71 ± 4.5 4.8 ± 0.15 AtGLDH1 AA +4 55 ± 5.4 6.0 ± 0.26 FtGLDH 140 ± 9.6 3.5 ± 0.20 FtGLDH AA 125 ± 10.4 4.2 ± 0.23 AtGLDH1 51 ± 4.8 4.9 ± 0.16 AtGLDH1 AA 60 ± 2.2 5.8 ± 0.07 D. Hasse et al. Alternative splicing of glycine decarboxylase H-protein FEBS Journal 276 (2009) 6985–6991 ª 2009 The Authors Journal compilation ª 2009 FEBS 6987 with the alternative H-proteins, both the naturally occuring FtGLDH AA and the artificial AtGLDH1 AA , are indeed exclusively due to the insertion of two additional alanyl residues into the N-terminus of the H-protein. In conclusion, the substrate properties of the alter- natively encoded Flaveria H-protein FtGLDH AA in the P-protein reaction of GDC differ significantly from those of the normal Flaveria H-protein. This was found in two independent assay systems, using either mitochondrial matrix extracts or recombinant cyano- bacterial P-protein. It was further substantiated by analysis of an engineered ‘alternative’ Arabidopsis H-protein. Hence, several lines of independent experi- mental evidence consistently demonstrated that alternative 3¢ splice site selection results in a more efficient H-protein. As fas as the physiological relevance of this effect is concerned, several other important variables are still unknown and require further research. From a methodical point of view, the GDC concentration is extremely high in leaf mitochondria but very much lower in mitochondria from heterotrophic tissue. Although root mitochondria have not yet been exam- ined for this feature, H-protein is present at levels of approximately 1 mm in the matrix of pea leaf mito- chondria, which is 7-fold higher than the concentration of P-protein [12] and 40-fold higher then the K m values determined in vitro. This could indicate that H-protein is possibly saturating in the leaves of C 3 plants, but unfortunately in vitro assays are very difficult to per- form at such high protein concentrations. Our assay conditions may be closer to the situation in C 4 leaf mitochondria, which contain distinctly less GDC, and even closer to that of root mitochondria. Furthermore, it is not known whether alternative splicing of H-protein pre-mRNA is a particular feature of Flaveria C 4 species or also occurs in other C 4 plants. The genomes of maize (Zea mays, http://www. maizesequence.org) and Sorghum bicolor [25] do not indicate the presence of alternative H-proteins in these monocot C 4 plants, and sequence data for other dicot- yledonous C 4 plants are not yet available. Our data show that the alternative H-protein enhances P-protein activity by about 20% in vitro. Although this is not a very large effect, it could have significance for both C 3 and C 4 plants in vivo. For C 3 plants, this may be assumed from effects on photosynthesis observed after antisense suppression of P-protein in potato, which indicated co-limitation of leaf metabolism by GDC activity [26]. In C 4 plants, the photorespiratory carbon flux is low, but nevertheless essential [27]. C 4 plants hence require less GDC than C 3 plants. However, the photorespiratory flux can considerably increase under conditions of low water supply and high temperature [28]. Under such conditions, which occur intermittently or may even prevail in many habitats typical for C 4 plants, the use of alternative H-protein could counter- act limitations to photosynthetic ⁄ photorespiratory metabolism. Possible enhancement effects on T- and L-protein remain to be examined, but could add to a larger overall activity enhancement. Finally, alternative splicing of H-protein pre-mRNA is regulated organ-dependently in C 4 Flaveria species [2]. It appears that C 4 plants very often harbor only one H-protein gene (Zea mays,C 4 Flaveria) or two H-protein genes (Sorghum bicolor), whereas up to four H-protein genes occur in C 3 plants [7,9]. Interestingly, transcript analyses revealed that at least one member of these gene families is preferentially expressed in photosynthesizing tissues [7,8]. This is why some researchers differentiate between two classes of H-pro- teins: class I H-proteins, which are preferentially expressed in tissue with high photorespiration, and class II H-proteins, which are more strongly involved in one-carbon metabolism of non-photorespiratory tis- sues [9]. Similar to this, alternative H-proteins could possibly allow fine adjustment of GDC to the parti- cular metabolic requirements in various organs of C 4 plants. Experimental procedures Overexpression constructs Overexpression constructs for Synechocystis P-protein (SwissProt P74416) and H-protein (Swissprot P73560) have been described previously [29]. Coding regions of the mature F. trinervia H-proteins [2] FtGLDH and FtGLDH AA (SwissProt P46485) were ligated into the expression vectors pET-28a (tagged) and pET-3a (tagless) (Merck, Darmstadt, Germany) via the NdeI and BamHI restriction sites. For the A. thaliana H-proteins AtGLDH1 (At2g35370, SwissProt P25855) and AtGLDH1 AA , cDNA was prepared from 2.5 lg of purified leaf RNA (NucleospinÔ RNA plant kit, Macherey-Nagel, Du ¨ ren, Germany) using the Revert- AidÔ H Minus cDNA synthesis kit (MBI Fermentas, St Leon-Rot, Germany). The coding region of the mature H-protein AtGLDH1 was PCR-amplified using the gene- specific primers (with underlined restriction sites) 5¢- CATAT GTCCACAGTTTTGGA-3¢ (sense, for native AtGLDH1), 5¢- CATATGTCCACAGCTGCAGTTTTGGA-3¢ (sense, for the artificial AtGLDH1 AA ) and 5¢-GGATTCCCTAGTGA GCAGCATCT-3¢ (antisense), and the ElongaseÔ enzyme mix (Invitrogen, Karlsruhe, Germany). The PCR product Alternative splicing of glycine decarboxylase H-protein D. Hasse et al. 6988 FEBS Journal 276 (2009) 6985–6991 ª 2009 The Authors Journal compilation ª 2009 FEBS was ligated via pGEM-T (Promega, Mannheim, Germany) into pET-28a, and recloned into pET-3a as described above for F. trinervia H-protein overexpression constructs. The E. coli lplA gene (SwissProt P32099) [17] was ampli- fied by PCR (sense primer, 5¢- CATATGTCCACATTAC GCCTGCT-3¢; antisense primer, 5¢- AAGCTTCTACCTTA CAGCCCCCG-3¢). The sense and antisense primers were extended by NdeI and HindIII sites (underlined), respec- tively. After initial cloning of the PCR amplificates into pGEM-T, coding sequences were excised and ligated into corresponding cloning sites of pET-28a. Recombinant proteins Synechocystis P-protein was produced essentially as described previously [29]. Major modifications to this earlier procedure comprised addition of 200 lm pyridoxal- phosphate and 15 mm 2-mercaptoethanol to all buffers. E. coli strain BL21 (DE3) cells overexpressing H-pro- teins were induced using 1 mm isopropyl-b-d-thiogalacto- pyranoside for 16 h at room temperatures in 2YT medium supplemented with 0.4 mm lipoic acid, harvested, and soni- cated in 20 mm Bis ⁄ Tris (pH 6). Cleared lysates were then loaded onto a Q-SepharoseÔ column (GE Healthcare, Munich, Germany), and eluted using a linear buffered 0–0.5 m NaCl gradient. Fractions containing H-protein were pooled, concentrated, and further purified on a Seph- acrylÔ S-100 column (GE Healthcare) equilibrated in the same buffer. E. coli LplA was overexpressed and purified by metal affinity chromatography similar to Synechocystis P-protein but with 10% glycerol added to all buffers. The enzyme was stored at )20 °C in a buffer containing 10 mm Tris ⁄ 10 mm Mops (pH 7.5), 0.5 mm MgSO 4 , 50% glycerol and 5 mm 2-mercaptoethanol. Prior to use, it was activated for at least 30 min by the addition of 200 mm 2-mercaptoethanol. For complete lipoylation, recombinant H-proteins (at 0.6 mgÆmL )1 , corresponding to approximately 50 lm) were incubated at 30 °C for 3 h in 10 mm Tris ⁄ 10 mm Mops (pH 7.5), 5 mm MgSO 4 ,5mm ATP, 5 lm E. coli LplA and 200 lm lipoic acid (which is saturating for LplA). LplA was removed by chromatography through Q-SepharoseÔ and SephacrylÔ S-100 columns as described above. Purified H-proteins were used immediately for kinetic assays, or flash frozen in liquid nitrogen after addition of 20% v ⁄ v glycerol and stored at )80 °C. MALDI-TOF MS The purity of recombinant proteins was assessed by SDS– PAGE [30], and the molecular identity of all H-proteins including completeness of lipoylation was further verified by MALDI-TOF MS. Sample aliquots were desalted using ZipTip pipette tips containing C 18 reverse-phase medium (Millipore Corp., Bedford, MA, USA) and applied to a polished steel MALDI target by mixing with an equal volume of saturated a-cyano-4-hydroxycinnamic acid in 35% acetonitrile ⁄ 0.1% trifluoroacetic acid. Data were acquired in linear mode using a Reflex III mass spectro- meter (Bruker Daltonics, Bremen, Germany). For exact determination of the mean mass, the protein sample was mixed with an aliquot of protein standard I (Bruker Daltonics) to allow internal calibration, resulting in a mean mass error of < ±1 Da at the investigated mass range. Matrix extracts from pea leaf mitochondria Mitochondria were prepared from young garden pea plants and purified as described previously [31] on self-generating density gradients. Matrix extracts were prepared by lysis in four freeze–thaw cycles, followed by centrifugation at 44 000 g for 30 min to remove mitochondrial membranes. The supernatant was concentrated using VivaspinÔ col- umns with a 10 kDa exclusion limit (Sartorius, Go ¨ ttingen, Germany), and adjusted to a protein concentration of 3.5 mgÆmL )1 . All steps were performed at 4 °C. Determination of P-protein activity The recombinant P- and H-proteins were equilibrated in 20 mm sodium phosphate (pH 7.5) using pre-packed PD-10 columns (SephadexÔ G-25 M; Pharmacia, Freiburg, Germany). After determination of protein concentration [32], P-protein activity was assayed using the glycine–bicar- bonate exchange reaction [20]. Assays with Synechocystis P-protein contained 100 mm sodium phosphate (pH 6.0), 0.1 mm pyridoxalphosphate, 2 mm dithiothreitol, 20 mm glycine, 30 mm NaH 14 CO 3 (2.5 lCi), 2.5 lg P-protein (0.04 lm P-protein dimer) and 0–250 lm H-protein in a total volume of 300 lL [29]. Assays with mitochondrial matrix extracts contained 50 mm sodium phosphate (pH 6.0), 0.1 mm pyridoxalphosphate, 2 mm dithiothreitol, 30 mm glycine, 30 mm NaH 14 CO 3 (2.5 lCi), 9 lg matrix protein and 0–100 lm H-protein in a total volume of 150 lL. H-protein concentrations varied as indicated in the figures. Reactions were run at 30 °C in the linear response range for 20 min (varying glycine) or 30 min (varying H-protein). Rates obtained without substrate were sub- tracted, and kinetic parameters were calculated by nonlin- ear regression analysis using the software package graphpad prism (GraphPad Software, San Diego, CA, USA). All kinetic data are means ± SD from three or four analyses over the full substrate range. Acknowledgements Financial support to D.H. by the Landesgraduier- tenfo ¨ rderungsprogramm Mecklenburg-Vorpommern is gratefully acknowledged. D. Hasse et al. Alternative splicing of glycine decarboxylase H-protein FEBS Journal 276 (2009) 6985–6991 ª 2009 The Authors Journal compilation ª 2009 FEBS 6989 References 1 Reddy ASN (2007) Alternative splicing of pre-messen- ger RNAs in plants in the genomic era. Annu Rev Plant Biol 58, 267–294. 2 Kopriva S, Cossu R & Bauwe H (1995) Alternative splicing results in two different transcripts for H-protein of the glycine cleavage system in the C 4 species Flave- ria trinervia. Plant J 8, 435–441. 3 Kikuchi G, Motokawa Y, Yoshida T & Hiraga K (2008) Glycine cleavage system: reaction mechanism, physiological significance, and hyperglycinemia. Proc Jpn Acad Ser B Phys Biol Sci 84 , 246–263. 4 Engel N, van den Daele K, Kolukisaoglu U ¨ , Morgen- thal K, Weckwerth W, Pa ¨ rnik T, Keerberg O & Bauwe H (2007) Deletion of glycine decarboxylase in Arabid- opsis is lethal under non-photorespiratory conditions. Plant Physiol 144, 1328–1335. 5 Applegarth DA & Toone JR (2006) Glycine encepha- lopathy (nonketotic hyperglycinemia): comments and speculations. Am J Med Genet 140A, 186–188. 6 Kikuchi G & Hiraga K (1982) The mitochondrial gly- cine cleavage system. Unique features of the glycine decarboxylation. Mol Cell Biochem 45, 137–149. 7 Kopriva S & Bauwe H (1995) H-protein of glycine decarboxylase is encoded by multigene families in Flave- ria pringlei and F. cronquistii (Asteraceae). Mol Gen Genet 248, 111–116. 8 Wang YS, Harding SA & Tsai CJ (2004) Expression of a glycine decarboxylase complex H-protein in non- photosynthetic tissues of Populus tremuloides. Biochim Biophys Acta 1676, 266–272. 9 Rajinikanth M, Harding SA & Tsai CJ (2007) The glycine decarboxylase complex multienzyme family in Populus. J Exp Bot 58, 1761–1770. 10 Kopriva S, Chu CC & Bauwe H (1996) H-protein of the glycine cleavage system in Flaveria: alternative splic- ing of the pre-mRNA occurs exclusively in advanced C 4 species of the genus. Plant J 10, 369–373. 11 Oliver DJ, Neuburger M, Bourguignon J & Douce R (1990) Interaction between the component enzymes of the glycine decarboxylase multienzyme complex. Plant Physiol 94, 833–839. 12 Douce R, Bourguignon J, Neuburger M & Rebeille F (2001) The glycine decarboxylase system: a fascinating complex. Trends Plant Sci 6, 167–176. 13 Cohen-Addad C, Pares S, Sieker L, Neuburger M & Douce R (1995) The lipoamide arm in the glycine decarboxylase complex is not freely swinging. Nat Struct Biol 2, 63–68. 14 Pares S, Cohen-Addad C, Sieker LC, Neuburger M & Douce R (1995) Refined structures at 2 and 2.2- -A ˚ ngstrom resolution of 2 forms of the H-protein, a lipoamide-containing protein of the glycine decarboxyl- ase complex. Acta Crystallogr D51 , 1041–1051. 15 Fujiwara K, Okamura-Ikeda K & Motokawa Y (1986) Chicken liver H-protein, a component of the glycine cleavage system. Amino acid sequence and identification of the N’-lipoyllysine residue. J Biol Chem 261, 8836– 8841. 16 Macherel D, Bourguignon J, Forest E, Faure M, Cohen-Addad C & Douce R (1996) Expression, lipoylation and structure determination of recombinant pea H-protein in Escherichia coli. Eur J Biochem 236, 27–33. 17 Morris TW, Reed KE & Cronan JE Jr (1994) Identifi- cation of the gene encoding lipoate–protein ligase A of Escherichia coli. Molecular cloning and characterization of the lplA gene and gene product. J Biol Chem 269, 16091–16100. 18 Walker JL & Oliver DJ (1986) Glycine decarboxylase multienzyme complex – purification and partial charac- terization from pea leaf mitochondria. J Biol Chem 261, 2214–2221. 19 Bourguignon J, Neuburger M & Douce R (1988) Resolution and characterization of the glycine cleavage reaction in pea leaf mitochondria. Properties of the forward reaction catalysed by glycine decarboxylase and serine hydroxymethyltransferase. Biochem J 255, 169–178. 20 Klein SM & Sagers RD (1966) Glycine metabolism. I. Properties of the system catalyzing the exchange of bicarbonate with the carboxyl-group of glycine in Peptococcus glycinophilus. J Biol Chem 241, 197–205. 21 Hiraga K & Kikuchi G (1980) The mitochondrial glycine cleavage system. Purification and properties of glycine decarboxylase from chicken liver mitochondria. J Biol Chem 255, 11664–11670. 22 Sarojini G & Oliver DJ (1983) Extraction and partial characterization of the glycine decarboxylase multien- zyme complex from pea leaf mitochondria. Plant Physiol 72, 194–199. 23 Fujiwara K & Motokawa Y (1983) Mechanism of the glycine cleavage reaction. Steady state kinetic studies of the P-protein-catalyzed reaction. J Biol Chem 258, 8156–8162. 24 Motokawa Y, Kikuchi G, Narisawa K & Arakawa T (1977) Reduced level of glycine cleavage system in the liver of hyperglycinemia patients. Clin Chim Acta 79, 173–181. 25 Paterson AH, Bowers JE, Bruggmann R, Dubchak I, Grimwood J, Gundlach H, Haberer G, Hellsten U, Mitros T, Poliakov A et al. (2009) The Sorghum bicolor genome and the diversification of grasses. Nature 457, 551–556. 26 Heineke D, Bykova N, Gardestro ¨ m P & Bauwe H (2001) Metabolic response of potato plants to an antisense reduction of the P-protein of glycine decar- boxylase. Planta 212, 880–887. Alternative splicing of glycine decarboxylase H-protein D. Hasse et al. 6990 FEBS Journal 276 (2009) 6985–6991 ª 2009 The Authors Journal compilation ª 2009 FEBS 27 Zelitch I, Schultes NP, Peterson RB, Brown P & Brutnell TP (2009) High glycolate oxidase activity is required for survival of maize in normal air. Plant Physiol 149, 195–204. 28 Osmond CB & Harris B (1971) Photorespiration dur- ing C 4 photosynthesis. Biochim Biophys Acta 234, 270–282. 29 Hasse D, Mikkat S, Thrun HA, Hagemann M & Bauwe H (2007) Properties of recombinant glycine decarboxyl- ase P- and H-protein subunits from the cyanobacterium Synechocystis sp. strain PCC 6803. FEBS Lett 581, 1297–1301. 30 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 31 Douce R, Bourguignon J, Brouquisse R & Neuburger M (1987) Isolation of plant mitochondria: general principles and criteria of integrity. Methods Enzymol 148, 403–420. 32 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal Biochem 72, 248–254. Supporting information The following supplementary material is available: Fig. S1. Three-dimensional structure of pea H-protein showing the N-terminus and the position of the lipoyl arm. Fig. S2. Overexpression and effects on H-protein of E. coli lipoate–protein ligase A. Fig. S3. Glycine saturation kinetics of pea mitochon- drial matrix P-protein. Fig. S4. N-terminal sequences of thrombin-cleaved recombinant H-proteins and mass spectrometric verifi- cation of the recombinant Arabidopsis H-proteins. This supplementary material can be found in the online version of this article. Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. D. Hasse et al. Alternative splicing of glycine decarboxylase H-protein FEBS Journal 276 (2009) 6985–6991 ª 2009 The Authors Journal compilation ª 2009 FEBS 6991 . Alternative splicing produces an H-protein with better substrate properties for the P-protein of glycine decarboxylase Dirk Hasse 1 , Stefan Mikkat 2 ,. response of potato plants to an antisense reduction of the P-protein of glycine decar- boxylase. Planta 212, 880–887. Alternative splicing of glycine decarboxylase