Báo cáo khoa học: Comparative biochemical and functional studies of family I soluble inorganic pyrophosphatases from photosynthetic bacteria potx

Comparative biochemical and functional studies of family I soluble inorganic pyrophosphatases from photosynthetic bacteria Marı ´ aR.Go ´ mez-Garcı ´ a*, Manuel Losada and Aurelio Serrano Instituto de Bioquı ´ mica Vegetal y Fotosı ´ ntesis, Centro de Investigaciones Cientı ´ ficas Isla Cartuja, CSIC-Universidad de Sevilla, Spain Soluble inorganic pyrophosphatase (sPPase) (inorganic diphosphatase, EC 3.6.1.1) is an essential and ubiquit- ous metal-dependent enzyme that cleaves inorganic pyrophosphate (PP i ), producing inorganic orthophos- phate (P i ). Its role in metabolism is thought to be the removal of PP i , a byproduct of many vital anabolic reactions, especially those involved in the synthesis of polymers, making them thermodynamically irreversible [1]. sPPases belong to two nonhomologous families: family I, widespread in all types of organism [2], and family II, so far confined to a limited number of bacteria and archaea [3,4]. The families differ in many functional properties; for example, Mg 2+ is the pre- ferred cofactor for family I sPPases studied, whereas Mn 2+ confers maximal activity to family II sPPases [5,6]. Although no sequence or overall structural similarity is observed between these two protein classes, there is a striking conservation of key active Keywords anoxygenic photosynthetic bacteria; cyanobacteria; functional complementation; ppa; soluble pyrophosphatases Correspondence M. R. Go ´ mez-Garcı ´ a, Department of Biochemistry, Stanford University School of Medicine, Beckman Center B413, 300 Pasteur Dr., Stanford, CA, 94305-5307, USA Fax: +1 650 725 6044 Tel: +1 650 723 5348 E-mail: mrgomez@stanford.edu A. Serrano, Instituto de Bioquı ´ mica Vegetal y Fotosı ´ ntesis, CSIC-Univ. de Sevilla, Avda. Ame ´ rico Vespucio 49, 41092 - Sevilla, Spain Fax: +34 954460065 Tel: +34 954489524 E-mail: aurelio@ibvf.csic.es *Present address Department of Biochemistry, Stanford Uni- versity School of Medicine, Beckman Center B413, Stanford, CA, USA (Received 17 April 2007, revised 23 May 2007, accepted 8 June 2007) doi:10.1111/j.1742-4658.2007.05927.x Soluble inorganic pyrophosphatases (inorganic diphosphatases, EC 3.6.1.1) were isolated and characterized from three phylogenetically diverse cyano- bacteria ) Synechocystis sp. PCC 6803, Anabaena sp. PCC 7120, and Pseudanabaena sp. PCC 6903 – and one anoxygenic photosynthetic bacter- ium, Rhodopseudomonas viridis (purple nonsulfur). These enzymes were found to be family I soluble inorganic pyrophosphatases with c. 20 kDa subunits with diverse oligomeric structures. The corresponding ppa genes were cloned and functionally validated by heterologous expression. Cyano- bacterial family I soluble inorganic pyrophosphatases were strictly Mg 2+ - dependent enzymes. However, diverse cation cofactor dependence was observed for enzymes from other groups of photosynthetic bacteria. Immunochemical studies with antibodies to cyanobacterial soluble inor- ganic pyrophosphatases showed crossreaction with orthologs of other main groups of phototrophic prokaryotes and suggested a close relation- ship with the enzyme of heliobacteria, the nearest photosynthetic relatives of cyanobacteria. A slow-growing Escherichia coli JP5 mutant strain, containing a very low level of soluble inorganic pyrophosphatase activity, was functionally complemented up to wild-type growth rates with ppa genes from diverse photosynthetic prokaryotes expressed under their own promoters. Overall, these results suggest that the bacterial family I soluble inorganic pyrophosphatases described here have retained functional similar- ities despite their genealogies and their adaptations to diverse metabolic scenarios. Abbreviations A. 7120 sPPase, Anabaena sp. PCC 7120 soluble inorganic pyrophosphatase; Ec-sPPase, Escherichia coli soluble inorganic pyrophosphatase; P. 6903 sPPase, Pseudanabaena sp. PCC 6903 soluble inorganic pyrophosphatase; PCC, Pasteur culture collection; S. 6803 sPPase, Synechocystis sp. PCC 6803 soluble inorganic pyrophosphatase; sPPase, soluble inorganic pyrophosphatase. 3948 FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works site residues, a remarkable example of convergent enzyme evolution [5–10]. The two best-studied exam- ples of family I enzymes are the hexameric sPPase of Escherichia coli (Ec-sPPase) and the dimeric enzyme of Saccharomyces cerevisiae, prototypes of prokaryotic and eukaryotic family I sPPases, respectively [11]. Bac- terial and archaeal family I sPPases are usually homo- hexamers, whereas eukaryotic sPPases are homodimers or monomers [12]. The subunit size is generally 19– 22 kDa in prokaryotic sPPases and 30–34 kDa in their dimeric or monomeric eukaryotic counterparts [2,11,12]. In a previous study, it was shown that sPPases of photosynthetic plastids from microalgae and plants are eukaryotic family I enzymes, and it was suggested that during the evolutionary history that gave rise to these organelles, the prokaryotic sPPase of the ancestral cyanobacterial-like endosymbiont was functionally substituted by its host cell homolog [12]. In this context, the sPPases of photosynthetic bacteria, a polyphyletic and very diverse assembly of prokaryo- tes, are worth characterizing. An increasing body of biochemical and genetic evi- dence suggests that PP i plays an important role in the bioenergetics of many archaea, bacteria, and protists [13,14]. In these organisms, two types of inorganic pyrophosphatases, sPPases and proton-translocating PPases, H + -PPases, with different subcellular localiza- tions, hydrolyze PP i generated by cell anabolism, and replenish the P i pool needed for phosphorylation reac- tions. The widespread presence of these key enzymes of PP i metabolism in photosynthetic organisms, except cyanobacteria, strongly supports the ancestral nature of bioenergetics based on this simple energy-rich com- pound that may play an important role in survival under different biotic and abiotic stress conditions [13,15]. This work shows that cyanobacterial strains as well as diverse anoxygenic photosynthetic bacteria possess family I sPPases with different catalytic and physico- chemical properties (i.e. divalent cation dependence, oligomeric structure), and extends prior work on cyanobacterial counterparts [16], as no detailed com- parative studies of these enzymes from prokaryotic photosynthetic organisms have been performed so far. The only previous study on cyanobacterial sPPases reported that the enzyme from the unicellular cyano- bacterium Microcystis aeruginosa NIES-44 was a trimeric protein with a 28 kDa subunit [16], in contrast to the well-characterized hexameric structure of Ec-sPPase [2]. The characterization of photobacterial sPPases has also allowed us to establish phylogenetic and evolutionary relationships between prokaryotic enzymes and homologs from photosynthetic plastids. Results and Discussion Detection of sPPase activity in photosynthetic prokaryotes ) enzymatic features of isolated sPPases Family I is the most widespread and probably the most ancestral sPPase group [2,11]. Molecular phylo- genetic analyses indicate the existence of two divergent evolutionary lineages in this protein assembly: the ‘eukaryotic’ (fungi, plants, metazoa, and most pro- tists), and the ‘prokaryotic’ (bacteria, archaea, and photosynthetic eukaryotes) [12]. Family I is therefore an ancient conserved group of orthologs from evolu- tionarily very distant organisms. Cell-free protein extracts from all photosynthetic prokaryotes studied (Table 1) contain substantial levels of an alkaline Table 1. Strains used in this work. Section a Strain Cyanobacteria I Synechocystis sp. PCC 6803 b Synechococcus sp. PCC 7942 Microcystis aeruginosa NIES-44 c Microcystis aeruginosa HUB5-2-4 d II Dermocarpa sp. PCC 7437 III Pseudanabaena sp. PCC 6903 Phormidium laminosum (argardh)Gom.H-1pC11 e Spirulina sp. PCC 6313 IV Anabaena sp. ATCC 29413 f Anabaena sp. PCC 7120 Nostoc sp. PCC 7107 Calotrix sp. PCC 7601 V Fischerella sp. UTEX 1829 g Anoxygenic photosynthetic bacteria Purple nonsulfur bacteria Rhodopseudomonas palustris h Rhodopseudomonas viridis h Rhodospirillum rubrum S1 Rhodobacter sphaeroides DSM 158S i Rhodobacter capsulatus E1F1 Purple sulfur bacteria Amoebobacter roseus j Chromatium vinosum j Green sulfur bacteria Chlorobium limicola j Chlorobium tepidum ATCC 49652 Chlorobium phaeobacteroides j Heliobacteria Heliobacterium chlorum DSM 1132 a Sections in the classification of Rippka et al. [17]. b PCC, Pasteur Culture Collection. c NIES, National Institute of Environmental Stud- ies, Japan. d Humbodt University Berlin, Professor T. Bo ¨ rner. e Univ. Pais Vasco (Dr J. L. Serra). f ATCC, American Type Culture Collection. g UTEX, Culture Collection University of Texas. h Dr A. Verme ´ glio, CEA-Caradache, France. i DSM, Deutsche Sammlung von Microorganismen, Germany. j University of Girona, Spain, Pro- fessor Jordi Mas. M. R. Go ´ mez-Garcı ´ a et al. Pyrophosphatases from photosynthetic bacteria FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works 3949 sPPase activity (0.2–2.5 UÆmg )1 protein) that abso- lutely requires a divalent metal cation. Cyanobacterial sPPases from different taxonomic groups [17] (Table 1) are all strictly Mg 2+ -dependent enzymes with c. 22 kDa subunits (Table 2, and data below). They exhibit fairly constant specific activity levels (0.2– 0.4 UÆmg )1 protein). On the contrary, a marked vari- ability of cation dependence was found among anoxy- genic bacteria sPPases; other cations, such as Zn 2+ , Mn 2+ or Co 2+ , replace Mg 2+ efficiently in extracts of the purple, nonsulfur and sulfur (not shown) anoxy- genic bacteria studied (Table 2). Thus, Rhodospirillum rubrum and Rhodopseudomonas viridis enzymes are Zn 2+ -dependent, whereas Rhodobacter capsulatus sPPase is Mn 2+ -dependent. Interestingly, the green (sulfur and nonsulfur) photosynthetic bacteria and the Heliobacterium strain tested exhibit sPPase activity with a marked preference for Mg 2+ , being similar in this respect to their cyanobacterial counterparts (Table 2 and data not shown). On the whole, specific activity levels in extracts of anoxygenic bacteria (1.0– 2.5 UÆmg )1 protein) were higher than in cyanobacterial extracts. A purification procedure, similar to the one des- cribed for the isolation of sPPase isoforms from the unicellular alga Chlamydomonas reinhardtii [12], was used to isolate the sPPases from the cyanobacteria Synechocystis sp. PCC 6803 (S. 6803 sPPase), Anabae- na sp. PCC 7120 (A. 7120 sPPase), and Pseudanabaena sp. PCC 6903 (P. 6903 sPPase), and the purple bacter- ium Rhodop. viridis. The method yielded electrophoret- ically pure sPPases with specific activities in the range 120–300 UÆmg )1 protein and recovery levels of 20– 30%. In all cases, the analysis by SDS ⁄ PAGE and native PAGE of purified preparations showed only one protein band of 20–22 kDa (Figs 1 and 2; supple- mentary Fig. S1). Analytical gel filtration FPLC of S. 6803 sPPase revealed one active hexameric sPPase (native molecular mass of 110 ± 5 kDa), in accordance with the oligomeric state of the archetypal Ec-sPPase [2,18]. A subunit molecular mass of 19 187 Da ± 0.1% was determined by MALDI-TOF MS for S. 6803 sPPase, somewhat lower than but in fair agree- ment with the apparent molecular mass values estima- ted by indirect measurements (Fig. 1). Native PAGE showed small differences in the migration of purified S. 6803 sPPase and A. 7120 sPPase (Fig. 2), in accord- ance with the strong acidic character of the native S. 6803 sPPase (pI 4.70) determined by column chro- matofocusing (data not shown). The same oligomeric states were found for A. 7120 sPPase (not shown) and P. 6903 sPPase (supplementary Fig. S1) (native molecular masses of 114 ± 5 kDa and 120 ± 5 kDa, respectively). In all cases, both the subunit molecular masses and oligomeric states are similar to those described for Ec-sPPase [2,18]. Interestingly, the sPPase from the purple nonsulfur photobacterium Rhodop. viridis exhibits a clearly larger native molecu- lar mass (240 ± 15 kDa), suggesting a higher oligo- meric state (dodecameric) (Fig. 1B). Hexameric Ec-sPPase has been described as a dimer of trimers, and the formation of these structures involves residues such as H136, H140 and D143, which participate in strong ionic interactions mediated by Mg 2+ [2,19]. Changes in the residues involved in the interactions between subunits could explain the unusual oligomeric states found for some photosynthetic bac- teria and photosynthetic eukaryote sPPases [12], so the trimeric structure reported by Kang & Ho for Mi. aeru- ginosa NIES-44 sPPase [16] is probably due to the pres- ence of trimers in solution. All determinations of native molecular masses reported here were performed with excess of Mg 2+ in solution to allow the interactions involved in hexamer formation. The Rhodop. viridis sPPase (Fig. 1B) shows differences in the oligomeric state (a dodecameric structure), probably due also to changes in the residues involved in the formation of the dimer of trimers. Although these proposals require Table 2. Cation dependence of sPPase enzymes. The level 100 is assigned to the activity determined with Mg 2+ in each case. Assays were performed in the presence of 4 m M divalent cation using purified enzyme (partially purified, in the cases of Rhodob. capsulatus, Rhodop. palustris and Chlorob. tepidum) in the range 20–30 UÆmL )1 . Cation Synechocystis sp. PCC 6803 Anabaena sp. PCC 7120 Pseudanabaena sp. PCC 6903 Rhodos. rubrum Rhodop. viridis Rhodop. palustris Rhodob. capsulatus Chlorob. tepidum E. coli Mg 2+ 100 100 100 100 100 100 100 100 100 Mn 2+ 3 3 1 115 97 97 180 13 13 Cu 2+ 4 2 2 42 94 105 50 20 3 Fe 2+ 3 1 5 150 134 131 21 9 4 Zn 2+ 13 1 2 165 140 40 90 15 10 Co 2+ 2 3 2 154 125 41 84 17 10 No cation 0 0 0 0 0 0 0 0 0 Pyrophosphatases from photosynthetic bacteria M. R. Go ´ mez-Garcı ´ a et al. 3950 FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works further study, it should be noted that all sPPase sequences from photosynthetic bacteria found in data- bases (see below) show nonconservative substitutions in most residues corresponding to those six involved in Ec-sPPase hexamer stabilization [2,19] (data not shown). The catalytic properties of photosynthetic bacterial sPPases studied in this work are shown in Table 3. Similar to Ec-sPPase [20], the cyanobacterial enzymes exhibit a high affinity for the substrate; however, the anoxygenic bacteria sPPases show K m values one order of magnitude higher than the cyanobacterial counter- parts. The catalytic efficiencies (estimated as k cat ⁄ K m ratios) of the sPPases from purple photosynthetic bac- teria are in the same range as that found for Ec-sPPase and their homologs from photosynthetic eukaryotes [12], but the cyanobacterial enzymes show values one order of magnitude higher, due to their lower K m . sPPases among diverse cyanobacteria and anoxygenic photosynthetic bacteria Western blot analyses performed using a monospecific polyclonal antibody against S. 6803 sPPase with sol- uble protein extracts from cyanobacteria belonging to all different taxonomic sections [17] revealed that the product of the ppa gene was present in all of the strains tested (Fig. 3A). All of them exhibit Mg 2+ -dependent sPPase activity (Table 2 and data not shown) and possess 20–22 kDa polypeptides, which strongly cross- reacted with the antibody to S. 6803 sPPase, suggesting that cyanobacteria have tightly related prokaryotic fam- ily I sPPases. P. 6903 sPPase, in accordance with the greater length of its ppa gene, and orthologs in other strains of section III (see Table 1), showed an immuno- detected band with a higher molecular mass (Fig. 3A, middle). Two different strains of Mi. aeruginosa Elution volume (ml) ALD BSA OVA CYT 2 1 CAT 9. 10. 8. 3 Log molecular mass 0 0.01 0.02 0 0246810 5 10 15 20 25 30 sPPase activity (U/ml) ( ) 35 0 5 10 15 20 25 30 sPPase activity (U/ml) ( ) 35 A B Elution volume (ml) 02 4 6 8 10 12 Elution volume (ml) * 45 29 20 66 kDa Absorbance at 280 nm 0 0.01 0.02 Absorbance at 280 nm Mass/Charge Relative intesity 3000010000 20000 40000 M M 2+ M 3+ 2M + 100 50 + 45 29 20 14 kDa Synechocystis sPPase 5. 6. 7. 4 Elution volume (ml) Log molecular masss 5 FER CAT ALD LPD BSA OVA CYT 6 Rps. viridis sPPase * Fig. 1. Gel filtration FPLC of purified native Synechocystis sp. PCC 6803 (A) and Rho- dop. viridis (B) sPPases. Aliquots (0.5 mL) of S. 6803 sPPase purified preparations were applied to a Superose 12HR 10 ⁄ 30 column. Isocratic elution was performed at a flow rate of 1 mLÆmin )1 , and 0.2 mL fractions were collected. The Coomassie Blue-stained SDS ⁄ PAGE gels of the indicated fractions around the activity peaks (highest activity fraction marked with an asterisk) show a single 22 kDa protein that coeluted with sPPase activity in both cases. The positions and molecular masses of protein standards are indicated on the left side of the gels. (A) The upper right inset shows column calibra- tion protein standards (CAT, catalase; ALD, aldolase; BSA, bovine seroalbumin; OVA, ovoalbumin; CYT, cytochrome c) and the positions of the cyanobacterial sPPase peak (black circle), which corresponds to a native molecular mass of c. 110 kDa. The MALDI- TOF MS profile of S. 6803 sPPase is shown on the left. (B). The upper inset shows col- umn calibration with protein standards (FERR, ferritine; CAT, catalase; ALD, aldo- lase; LPD, lipoamide dehydrogenase; BSA, bovine seroalbumin; OVA, ovoalbumine; CYT, cytochrome c). A native molecular mass of 240 kDa was estimated for the Rhodop. viridis sPPase. M. R. Go ´ mez-Garcı ´ a et al. Pyrophosphatases from photosynthetic bacteria FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works 3951 (NIES-44 and HUB5-2-4) also showed an immunode- tectable band of c. 22 kDa and exhibited Mg 2+ - dependent activity (data not shown), suggesting that their sPPase subunits might be similar to those of typical cyanobacterial homologs (Fig. 3A, right). This is in disagreement with previously reported data on Mi. aeruginosa NIES-44 sPPase, where a larger 28 kDa subunit was estimated by SDS ⁄ PAGE and a trimeric structure was found by gel filtration [16]. The antibody against S. 6803 sPPase also crossreact- ed with soluble extracts from nearly all anoxygenic photosynthetic bacteria tested, which belong to differ- ent taxonomic groups (Table 1). They display, how- ever, greater heterogeneity in the molecular mass of the detected protein band, as is also the case for metal cation dependence, as the sPPase activity of many of these bacteria can efficiently use other divalent cations as cofactors, e.g. Zn 2+ ,Co 2+ and Fe 2+ (Table 2). The enzymes of rhodospirillacean species Rhodos. rubrum, Rhodop. palustris and Rhodop. viridis had 22 kDa subunits, suggesting that they should be members of family I (Figs 1B and 3B), in agreement with the genome database sequences available (see below). However, the sPPase of the closely related species Rhodob. capsulatus shows Mn 2+ -dependent activity and is not recognized by the antibody raised against S. 6803 sPPase, as expected, because Rhodob. capsula- tus and Rhodop. sphaeroides sPPases have already been described as family II enzymes [21]. It can be specula- ted that this could be a singular case of horizontal gene transfer in photosynthetic prokaryotes, as it has been shown that marine unicellular cyanobacteria pos- sess two paralogous ppa genes of different phylogeny; one of them, similar to the proteobacterial homologs, was probably obtained by horizontal gene transfer [22] (see below). Highly degenerate family I sPPase-enco- ding pseudogenes are also present in the genomes of a number of prokaryotes from diverse taxonomic groups with functional family II sPPase genes, thus illustrating functional substitution by nonhomologous sPPases in a context of gene degradation and displacement, which is proposed to be of major importance in microbial genome evolution [22,23]. The clearly larger size of the Chromatium vinosum immunodetected protein band (60 kDa) (Fig. 3B) is an exception among photosyn- thetic bacteria; nevertheless, there are no data in the literature regarding sPPases from purple sulfur photo- synthetic bacteria, or genome sequence projects of any organisms of this phylogenetic group, that could sug- gest that its sPPases form a subfamily with distinctive features. The high variability in cation dependence and oligo- meric states found for the anoxygenic bacteria may reflect adaptations to specific metabolic scenarios. This proposal is supported by the striking differential inhibition of enzymatic activity by phosphorylated metabolites shown by S. 6803 sPPase and the Rho- dop. viridis sPPase: the purple bacterial enzyme was strongly inhibited by fructose 1,6-bisphosphate or 2-phosphoglycerate in the assays (70% and 40%, respectively; 1 mm in the assay); however, its cyano- bacterial counterpart was not affected at all; ATP was also inhibitory (c. 50%; 1 mm in the assay) to both photobacterial enzymes, probably by virtue of its 321 AB 3'2'1' Anabaena Synechocystis Fig. 2. Native PAGE analysis of S. 6803 sPPase and A. 7120 sPPase. (A) Coomassie Blue-stained nondenaturing PAGE of puri- fied S. 6803 sPPase (lanes 1 and 2) and A. 7120 sPPase (lane 3) resolved in 7% polyacrylamide gel. (B) In situ sPPase activity assay of the same preparations in a native gel run in parallel (1¢,2¢ and 3¢). Six micrograms of protein was loaded per lane. Table 3. Kinetic parameters of sPPase enzymes. Data are an average of at least three independent determinations. Synechocystis sp. PCC 6803 Synechocystis sp. PCC 6803 recombinant Anabaena sp. PCC 7120 Pseudanabaena sp. PCC 6903 Rhodop. viridis Rhodos. rubrum a Rhodop. palustris a E. coli b K m (lM) 2.8 3.1 3 2.9 27.4 25.2 30.1 4.5 K cat (s )1 ) 916 800 850 850 750 750 800 200 K cat ⁄ K m 327 258 283 293 27 30 27 44 a Data obtained from partially purified preparation. b Salminen et al. [20]. Pyrophosphatases from photosynthetic bacteria M. R. Go ´ mez-Garcı ´ a et al. 3952 FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works chelating properties. It is interesting to note in this respect that the pioneering work of Klemme and Guest in the early 1970s already identified two classes of sPPase in rhodospirillaceae with different biochemical and metabolite-dependent regulatory characteristics, which may correspond to the currently identified fam- ily I and II enzymes [24]. It is noteworthy that antibody to S. 6803 sPPase showed a strong crossreaction with a 20 kDa protein band in the extract of Heliobacterium chlorum, a mem- ber of the only group of photosynthetic Gram-positive bacteria known so far (heliobacteria) (Fig. 3B, right). This is in agreement with the close relationship between cyanobacteria and heliobacteria determined by phylogenetic analysis of photosynthetic genes [25]. No protein bands were immunodetected in cell extracts from E. coli K12 and DH5a or nonphotosynthetic pro- tists (data not shown). Genomic DNA from strains representative of all cyanobacterial taxonomic groups (Chroococcales, Synechocystis sp. PCC 6803 and Syn- echococcus sp. PCC 7942; Oscillatoriales, Pseudanabae- na sp. PCC 6903; Nostocales, Nostoc sp. PCC 7107 and Calothrix sp. PCC 7601; Stigonematales, Fischerel- la sp. UTEX182) and the green anoxygenic photobac- teria Chlorobium tepidum and Chlorob. limicola were found to possess homologous ppa genes by Southern blot analysis using the Synechocystis sp. PCC 6803 ppa gene as a probe (data not shown). Hence, ppa genes and their products (family I sPPases) are widely distri- buted among diverse photosynthetic prokaryotes. Functional complementation studies sPPase appears to be essential for cell anabolism, and it has not been possible to generate a mutant totally lacking this activity in E. coli [26] or Synechocystis sp. PCC 6803 [27]. However, some reconstitution studies have been performed with a thermosensitive E. coli mutant [28]. Here, we used an E. coli JP5 strain [29] obtained by chemical mutagenesis, lacking c. 90% of its native sPPase activity, as a host for in vivo comple- mentation experiments to test the functionality of ppa genes cloned from Synechocystis sp. PCC 6803, Anab- aena sp. PCC 7120, Pseudanabaena sp. PCC 6903 and Chlorob. tepidum, using their native promoters. As can be observed from the growth phenotype (Fig. 4), the photobacterial sPPases produced from pRGS, pRGA, pRGP and pRGCT plasmids restored normal E. coli PCC 6903 PCC 7437 PCC 7601 PCC 7120 PCC 6803 PCC 7942 UTEX 1829 PCC 6803 PCC 6903 PCC 6313 Phormidium laminosum PCC 6803 NIES-44 HUB5-2-4 C.7601 Rsp . rubrum Rsp . palustris Rsp . viridis Rb . capsulatus Chr . vinosum Am. r oseus H b . chlorum Chl . limicola Chl . tepidum Chl . phaeobacter 24 kDa A B 22 kDa 24 kDa 22 kDa 22 kDa 22 kDa Fig. 3. Western blot analysis of sPPases in cell-free extracts from diverse cyanobacteria and anoxygenic photosynthetic bacteria. (A) Western blots probed with a monospecific polyclonal antibody to S. 6803 sPPase showing crossreaction with sPPase orthologs of phylogenetically diverse cyanobacteria. A single 22 kDa protein band was immunodetected in all unicellular and filamentous strains of sections I, II, IV and V [5] tested, including the unicellular strain Mi. aeruginosa NIES-44. Note that the three section III strains tested, namely Pseudanabaena sp. PCC 6903, Spirulina sp. PCC 6313 and Phormidium laminosum, showed one band of slightly higher apparent molecular mass (c. 24 kDa). The Synechocystis sp. PCC 6803 (section I) sPPase was used as an internal standard in all blots (left-hand lanes). Strains are identified by their bacterial collection numbers. About 40 lg was loaded per lane. (B) Western blots probed with the monospecific antibody to S. 6803 sPPase showing crossreaction with sPPases in soluble protein extracts from diverse anoxygenic photosynthetic bacteria. A single 22–24 kDa band was immunodetected in purple nonsulfur (Rhodospirillaceae) and sulfur (Chromatiaceae) and in green sulfur (Chlorobiaceae) strains, as well as in Helio. chlorum (Heliobacteriaceae). Remarkably, Chr. vinosum (purple sulfur) showed a protein band of c . 60 kDa, and no band was detected in Rhodob. capsulatus, which has a family II sPPase. About 80 lg of protein was loaded per lane, except for Helio. chlorum, Rhodos. rubrum, Rhodop. palustris and Rhodop. viridis extracts, when 40 lg was loaded. M. R. Go ´ mez-Garcı ´ a et al. Pyrophosphatases from photosynthetic bacteria FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works 3953 growth rates and sPPase activity levels (Table 4). The antibody against S. 6803 sPPase also recognized the sPPases expressed in the mutant (Fig. 4). This expres- sion is similar to that found by Lahti et al. [29] with a thermosensitive E. coli mutant [30]. The complementa- tion studies demonstrate that photobacterial sPPases are functionally equivalent to that of the host organ- ism, and that the promoters seem to be regulated by the same factors. Analysis of the promoter regions of these bacterial ppa genes may be helpful for under- standing their regulation in future studies. Sequence and phylogenetic analysis Knowledge of the N-terminal sequences of the three cyanobacterial sPPases purified and characterized in this work allowed us to identify the correct first codon of transcription of the Synechocystis sp. PCC 6803 ppa gene, as in the genome sequence of this cyanobacterium [31], it was assigned an ATG situated upstream of the real GTG used, actually encoding a formyl-Met, which was determined by Edman degradation sequen- cing of the N-terminal region of the native pro- tein (MDLSRIPAQP KAGLINVLIE IPAGSKNKYE FDKDMNNFAL DRV). A few ppa genes from Syn- echocystis sp. PCC 6803 and also Anabaena sp. PCC 7120 share this feature [32]. The encoded 170 amino acid polypeptide has a predicted molecular mass of 19 216 Da and a pI of 4.69, in good agreement with MALDI-TOF MS (Fig. 1A) and chromatofocusing data, respectively. The other three N-terminal sequences determined in this work, of Anabaena sp. PCC 7120 (MDLSRIPAQP KPGVINILIE IAG) (with an initial formyl-Met also encoded by a GTG codon), Pseudanabaena sp. PCC 6903 (MDLSRIPPQP KAGILNVLIE IPAG), and Rhodop. viridis (MRIDA IDXA), and that of Mi. aeruginosa NIES-44 (MDL SRKPAQP IPGLKNVLVE TAGSINIT) [16], show a high degree of similarity with other cytosolic sPPases and conservation of residues localized in the active site (shown in bold) and involved in catalysis in Ec-sPPase [2,11]. In all cases, the molecular mass determined by SDS ⁄ PAGE and ⁄ or MS is in good agreement with values estimated from the DNA sequence. The heterogeneity of the sPPases from photosyn- thetic prokaryotes is clearly in accordance with the phylogenetic analysis presented in Fig. 5. Two well- defined groups of family I sPPases cluster on the phylogenetic tree shown: on one side, cytosolic and organellar eukaryotic sPPases, and on the other side, 0 1.0 2.0 0 1.0 2.0 0 1.0 2.0 0 0 200 400 600 800 1000 1200 1.0 2.0 ppa Chl.tep. JP5 DH AB 5 ! Time (min) Absorbance at 600 nm C ppa S.6903 ppa A.7120 ppa S.6803 P. 6903 22 kDa 22 kDa 22 kDa 24 kDa Fig. 4. Functional complementation of E. coli JP5 mutant with pRGS, pRGA, pRGP and pRGCT plasmids. (A) Growth curves, checked by absorbance at 600 nm, of E. coli DH5a (control, C), E. coli JP5 mutant and E. coli JP5 expressing the Chlorob. tepidum ppa gene, and the cyanobacterial Synechocystis sp. PCC 6803, Anabaena sp. PCC 7120 and Pseudanabaena sp. PCC 6903 ppa genes. Growth of the complemented E. coli JP5 mutant recovered rates up to those of the wild type. (B) Western blot analysis of E. coli JP5 transformed with an empty plasmid (C, control) and plasmids expressing the Chlorob. tepidum, Synechocystis sp. PCC 6803, Anabaena sp. PCC 7120 and Pseudanabaena sp. PCC 6903 ppa genes. The recombinant photobacterial sPPases were immunodetected in cell-free extracts from the transformed clones. Forty micrograms of protein was loaded per lane, except in the case of the Chlorob. tepidum sPPase clone, when 70 lg was loaded. Table 4. sPPase specific activities of E. coli JP5 strains. Data are means ± standard errors of three independent determinations. Strain (plasmid) Specific activity (UÆmg )1 ) DH5a 4.05 ± 0.15 JP5 0.32 ± 0.10 JP5 (pRGS) 4.25 ± 0.10 JP5 (pRGA) 3.85 ± 0.15 JP5 (pRGP) 4.30 ± 0.10 JP5 (pRGCT) 5.30 ± 0.15 JP5 (pBS SK + ) 0.35 ± 0.05 Pyrophosphatases from photosynthetic bacteria M. R. Go ´ mez-Garcı ´ a et al. 3954 FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works the prokaryotic (bacterial and archaeal) sPPases and prokaryotic-type homologs of photosynthetic eukaryo- tes [12] (Fig. 5). Typical studied cyanobacteria, such as Synechocystis sp. PCC 6803, Anabaena sp. PCC 7120, or Pseudanabaena sp. PCC 6903, form a compact group that is different from other clusters of photosynthetic bacterial sPPases. As we previously reported, Prochlorococcus marinus MED4 and Syn- echococcus WH8102 have two ppa genes in their genomes: PPA1 codes for an inactive sPPase that 0.1 Mycoplasma genitalium Mycoplasma pneumoniae Gloeobacter violaceus Pseudanabaena PCC 6903 Synechocystis PCC 6803 Thermosynechococcus elongatus. BP-1 Trichodesmium erythraeum IMS101 Nostoc punctiforme Anabaena PCC 7120 Microcystis aeruginosa NIES44 B a c i l l u s s t e a r ot h e r m o p h i l u s B a c i l l u s h a l o d u r a n s Helicobacter pylori Sulfolobus acidocaldarius. Aquifex aeolicus Rickettsia prowazekii Escherichia coli Vibrio cholerae Caulobacter crescentus Rhodopseudomonas palustris Rhodospirillum rubrum Thermus thermophilus Dehalococcoides ethenogenes Streptomyces coelicolor Mycobacterium leprae Mycobacterium tuberculosis Chloroflexus aurantiacus Halobacterium NCR1 Thermoplasma acidophilum Methanobacterium thermophilus. Thermococcus litoralis Pyrococcus horikoshii Pyrococcus furiosus Chlorobium tepidum C h l a m y d o mo n a s r e i n h a r d t i i ( s P P a s e I I ) Arabidopsis thaliana Oryza sativa Zea mays Solanum tuberosum Chlamydia pneumoniae Chlamydia trachomatis Chlamydomonas reinhardtii CHLOR.(sPPase I) Arabidopsis thaliana CHLOR. Oryza sativa CHLOR. S. cerevisiae (PPA1) Mus musculus MIT. Bos taurus Mus musculus Synechococcus WH8102 (PPA2) Prochlorococcus marinus MED4 (PPA2) Haemophilus influenzae Neisseria meningitidis Eukaryotic Eukaryotic Family I Family I sPPases sPPases Synechococcus WH8102 (PPA1) Prochlorococcus marinus MED4 (PPA1) C y a n o b a c t e r i a C y a n o b a c t e r i a * * Ba c t e r ia l Ba c t e r ia l - - l i k e l i k e p l a n t p l a n t s P P a s e s s P P a s e s A r c h a e a A r c h a e a S. cerevisiae (PPA2) MIT 1000 * * 989 1000 1000 760 823 651 518 925 923 Purple Purple non non - - sulfur sulfur phot phot . . bact bact . . Green Green non non - - sulfur sulfur phot phot . . bact bact . . Green sulfur Green sulfur phot phot . . bact bact . . Prokaryotic Prokaryotic Family I Family I sPPases sPPases Fig. 5. Molecular phylogenetic analysis of family I sPPases of photosynthetic prokaryotes. Amino acid sequences deduced from prokaryotic sPPases were aligned using CLUSTALX. Most sequences have all the amino acids reported to be functionally important for sPPase activity and the PROSITE motif of family I sPPases. Numbers in nodes are bootstrap values (1000 replicates) supporting representative groups. Asterisks indicate the pairs of sPPase paralogs present in marine unicellular cyanobacteria [32]. The circled P indicates a partial N-terminal sequence. The 0.1 bar represents amino acid substitutions per site. Accession numbers for the sequences are (reading clockwise): Chlamydo. rein- hardtii CHLOR sPPase I, AJ298231; Arabidopsis thaliana CHLOR sPPase I, Atg09650; Oryza sativa CHLOR, BAD 16934; Sa. cerevisiae PPA1 (cytosolic), YBR011C; Sa. cerevisiae PPA2 MIT (mitochondrial), YMR267W; Mus musculus MIT (mitochondrial), Q91VM9; Bos taurus, P37980; Mus musculus, BAB25754; Synechococcus WH8102 PPA2, CAE08303; Pr. marinus MED4 PPA2, CAE18953; Hae. influenzae, P44529; Neisseria meningitidis, F81175; Mycoplasma genitalicum, P47593; Mycop. pneumoniae, P75250; Gloeobacter violaceus, grl4227; Pseudanabaena PCC 6903, P80898; Synechocystis PCC 6803, P80507; Thermosynechococcus elongatus BP-1, BAC09435; Trichodesmium erythraeum IMS101, ABG50803; Nostoc punctiforme, ZP_00112287; Anabaena PCC 7120, P80562; Pr. marinus MED4 PPA1, CAE18970; Synechococcus WH8102 PPA1, CAE08284; Mi. aeruginosa NIES44, 29 amino acid partial sequence, AAB19891; Bacillus stearothermophilus, O05724; Ba. halodurans, AP001512; Helicobacter pylori, AE001439; Sulfolobus acidocaldarius, P50308; Aquifex aeolicus, O67501; Rickettsia prowazekii, CA15034; E. coli, P17288; Vibrio cholerae, AAF95686; Caulobacter crescentus, AE005679; Rhodop. palustris, CAE25855; Rho- dos. rubrum, AAF21981; Chlorob. tepidum TLS, AAM72059; Chlorof. auranticus, EAO59327; Thermus thermophilus, P38576; Dehalococco- ides ethenogenes, AAW40363; Streptomyces coelicolor, CAB42762; Mycobacterium leprae, O69540; Mycob. tuberculosis, O06379; Halobacterium NCR, AAG18854.1; Thermoplasma acidophilum, P37981; Methanobacterium thermoautotrophicus, O26363; Thermococcus litoralis, P77992; Pyrococcus horikoshii, O59570; Py. furiosus, Q8U438; Chlamydo. reinhardtii (sPPase II), AJ298232; Ar. thaliana prokaryotic- like, At2g18230; O. sativa prokaryotic-like, O22537; Zea mays prokaryotic-like, O48556; Solanum tuberosum prokaryotic-like, CAA85362; Chlamydia pneumoniae, AAD19056; Chlamydia trachomatis, AAC68367. M. R. Go ´ mez-Garcı ´ a et al. Pyrophosphatases from photosynthetic bacteria FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works 3955 clusters with the ‘typical’ sPPases from freshwater cyanobacteria, whereas PPA2 codes for an isoform that is, so far, characteristic of marine unicellular cyanobacteria, and constitutes a second cyanobacteri- al sPPase group that is closely related to several non- photosynthetic proteobacteria (Haemophilus influenzae and Neisseria spp.) and the members of which are expressed as active sPPases [22]. It remains to be clar- ified whether PPA2 sPPases result from horizontal gene transfer or represent an ancestral cyanobacterial enzyme that was lost during the evolutionary history of the cyanobacterial lineage. The sPPases from nonsulfur purple bacteria (Rho- dos. rubrum, Rhodop. palustris) that show different cation dependence and oligomeric structure [Rho- dos. rubrum sPPase is a tetramer of c. 90 kDa (data not shown), and Rhodop. viridis sPPase appears to be a dodecamer; see Fig. 1B] are clustered with the main proteobacterial assembly. It can be speculated that these peculiar properties of the enzymes from photo- synthetic proteobacteria may result from functional adaptations to specific metabolic scenarios. The sPPases from the two classes of green photosyn- thetic bacteria associate with different branches, in agreement with their different molecular genealogies. The enzyme of the green nonsulfur bacterium Chlorofl- exus auranticus clusters weakly with archaeal homologs but it is also Mg 2+ -dependent. sPPases of the green sulfur bacteria Chlorob. tepidum and Chlorob. limicola are clearly Mg 2+ -dependent enzymes, and unexpect- edly cluster with the bacterium-like Mg 2+ -dependent sPPases of photosynthetic eukaryotes (Fig. 5), suggest- ing an interesting evolutionary relationship between the plant sPPase subfamily and a possible counterpart from a bacterial ancestor. Experimental procedures Organisms and growth conditions The photosynthetic bacteria used in this study are summar- ized in Table 1. Cyanobacteria were cultured at 30 °Cin BG11 liquid medium [17] supplemented with 1 g sodium bicarbonateÆL )1 and bubbled with 1.5% (v ⁄ v) CO 2 in air under continuous white light (25 WÆm )2 ). Nonsulfur purple bacteria were grown in modified Hutner medium [33], pur- ple and green sulfur bacteria were grown in Tru ¨ per and Pfenning medium [34], and Helio. chlorum was grown in heliobacteriaceae medium [35]. The E. coli JP5 strain [36], containing only c. 10–15% of wild-type sPPase activity levels, was cultured in LB medium supplemented with ampicillin when necessary (100 lgÆmL )1 ), and was grown at 37 °C with continuous shaking at 200 r.p.m. Competent E. coli JP5 cells were obtained using the protocol described previously [37]. Protein techniques Assay for sPPase activity sPPase was assayed by the colorimetric determination of P i produced by the enzymatic hydrolysis of PP i at 22 °C [12,38] with PP i as a substrate. The reaction mixtures con- tained 50 mm Tris ⁄ HCl (pH 7.5), 4 mm cation salt (MgCl 2 ) and 2 mm Na 4 PP i (standard assay conditions). The reaction was started by the addition of enzyme, and the PP i released after 10 min was determined. When the efficiencies of other divalent metal cations as cofactors were tested, the corres- ponding chloride salts were used in the assays instead of the Mg 2+ salt. Mg 2 PP i was utilized as substrate for kinetic parameter estimations. Reaction rates are expressed in terms of lmol P i generated per minute. SDS ⁄ PAGE, native PAGE (12% w ⁄ v) and Bradford pro- tein estimations were performed as described previously [39,41]. SDS ⁄ PAGE gels loaded with purified samples of sPPase were stained for activity as described by Kang & Ho [16]. Purification of the sPPases from Synechocystis sp. PCC 6803, Pseudanabaena sp. PCC 6903, Anabaena sp. PCC 7120 and Rhodop. viridis A purification protocol similar to the one described for the isolation of eukaryotic sPPases [12] was used for the native proteins from photobacteria. The recombinant S. 6803 sPPase was isolated from E. coli XL1blue trans- formed with pRGS plasmid and cultured in LB medium supplemented with ampicillin (100 lgÆmL )1 ) using the same protocol; anion exchange chromatography was essen- tial for the separation of the overexpressed sPPase and the native Ec-sPPase. Gel filtration FPLC (Amersham Phar- macia, Uppsala, Sweden) was used as an analytical tech- nique for purified enzymes. Column chromatofocusing on a Polybuffer Exchanger (PBE94) bed was performed according to the manufacturer’s instructions (Amersham Pharmacia). N-terminal sequences of purified sPPases from Synecho- cystis sp. PCC 6803, Anabaena sp. PCC 7120, Pseudanabae- na sp. PCC 6903 and Rhodop. viridis were obtained in this work by the Edman degradation method, using an automa- tic sequencer, at the protein analysis facilities of the Vienna Biocenter (University of Vienna, Austria). Immunochemical techniques A rabbit was injected with 500 lg of pure S. 6803 sPPase water ⁄ Freund’s coadjuvant (1 : 1). Antibodies were obtained as described previously [12,40]. Immunoblot Pyrophosphatases from photosynthetic bacteria M. R. Go ´ mez-Garcı ´ a et al. 3956 FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works assays (western blots) of protein samples were carried out after electrophoresis in SDS ⁄ PAGE (12%). Cloning and DNA manipulation The reaction mixture for PCR amplification of the ppa gene from Synechocystis sp. PCC 6803 (50 lL) contained 50 pmol of each of the following two pairs of oligonucleo- tide primers: ipyrpro,5¢-CGCTTAAGTTAAAAGCCTTT-3¢; and ipyrend,5¢-GCGAAGCTTATTTACCGGTTCA GTTAGCT-3¢. The PCR product contains the Synecho- cystis sp. PCC 6803 ppa gene and 375 bp of the upstream sequence containing the promoter region, and was cloned in pBS SK + (PRGS). For amplification of ppa from Chlo- rob. tepidum, the oligonucleotides used were: 5¢-ppa300chlor 5¢-TCGGGAAAGTGGCTCTG-3¢ and 3¢-ppa120chlo 5¢- CTCAGTCCTTGTCCACGGC-3¢. The PCR product was cloned in pBS SK + (pRGCT). A BclI–HindIII fragment of Synechocystis sp. PCC 6803 ppa was (460 bp) used as a probe for screening a genomic library of Anabaena sp. PCC 7120 [42]. The genomic lib- rary was plated at a dilution of 2000 colonies per plate. After incubation for 12 h at 37 °C, plates were replicated on 0.45 lm sheets, and the filters were then treated as described before [33] and hybridized overnight at 55 °C with the 32 P-labeled probe. One clone containing a 4.6 kb plasmid was isolated. Restriction analysis with XmnI and ClaI (pRGA) identified the ppa gene with an extra sequence at the 5¢-end corresponding to the promoter region (243 bp). This plasmid was used in heterologous expression experiments. The same methodology allowed us to screen a genomic library of Pseudanabaena sp. PCC 6903 [43]; one clone was obtained with a 7.2 kb plasmid (pRGP1) after hybridization. The plasmid was subjected to restriction analysis with HincII and EcoRI, and subcloned in pBS SK + obtaining pRGP, which con- tains Pseudanabaena sp. PCC 6903 ppa and 233 bp of the upstream region. The plasmids used for heterologous expression experiments (pRGS, pRGA, pRGP and pRGCT) containing ppa ORFs and the corresponding upstream regions were sequenced to ensure that ppa genes were cloned in the opposite orientation to the lacZ vec- tor’s promoter and that the expression was achieved under their own native promoters. Chromosomal DNA was isolated from bacterial cells as previously described [44]. For DNAÆDNA hybridization (Southern blotting), the method of Ausubel et al. was used [41]. Samples of bacterial genomic DNA were completely digested with dif- ferent restriction enzymes, run in 0.7% (w ⁄ v) agarose gels, and blotted onto nylon membranes (Zetaprobe; Biorad, Richmond, CA). The filter was hybridized using the pro- tocol described by Church & Gilbert [45] at 55 °C for heterologous hybridization. The nylon filters were then exposed to films (Kodak X-100 310S, Racine, Chicago, IL) at ) 80 °C and eventually developed. Nucleotide and N-terminal protein sequence accession numbers The EMBL ⁄ GenBank database accession number for the Synechocystis sp. PCC 6803 ppa gene is AJ252207. The accession number in the SwissProt database for both the natural and recombinant N-terminal protein sequences is P80507. The EMBL ⁄ GenBank database accession number for Anabaena sp. PCC 7120 ppa is AJ252206, and the accession number in the SwissProt database is P80562. The EMBL ⁄ GenBank database accession num- ber for Pseudanabaena sp. PCC 6903 ppa is AJ252205, and the accession number in the SwissProt database is P80898. Protein sequence comparisons and phylogenetic analyses A multiple amino acid sequence alignment of the sPPases from photosynthetic prokaryotes and other selected prok- aryotic family I sPPases was performed using the clustalx v.1.8 program [46]. This alignment was used to construct a phylogenetic distance tree (neighbor-joining method, BLO- SUM matrix) with the same program. Sequence data from public databases or unfinished microbial genome projects were obtained by similarity searches using blast algorithms [47] against websites of the National Center of Biotechno- logy Information (NCBI), USA (http://www.ncbi.nih.gov/ PMGifs/Genomes/allorg.html), the Joint Genomic Institute (JGI), USA (http://spider.jgi-psf.org/JGI_microbial/html/), the Sanger Institute, UK (http://www.sanger.ac.uk/Projects/), or the Institute for Genomic Research (TIGR), USA (http://www.tigr.org/tdb/mdb/mdb.html). Acknowledgements The authors gratefully thank Dr N. N. Rao and Dr J. Josse for critical review of the manuscript, and Profes- sor W. Lo ¨ ffelhardt (University of Vienna, Austria) for his assistance in N-terminal protein sequencing and some MALDI-TOF MS analyses. This work was supported by research grants from the Spanish (BMC2001-563 and BFU2004-00843, MEC) and Andalusian Regional (PAI group CVI-261) Adminis- trations, funded in part by the EU FEDER program. References 1 Kornberg A (1962) On the metabolic significance of phosphorolytic and pyrophosphorolytic reactions. In Horizons in Biochemistry (Kasha M & Pullman D, eds), pp. 251–254. Academic Press, New York, NY. 2 Cooperman BS, Baykov AA & Lahti R (1992) Evolutionary conservation of the active site of soluble M. R. Go ´ mez-Garcı ´ a et al. Pyrophosphatases from photosynthetic bacteria FEBS Journal 274 (2007) 3948–3959 Journal compilation ª 2007 FEBS. No claim to original US government works 3957 [...]... complementation analysis of an inorganic pyrophosphatase from Bartonella bacilliformis Can J Microbiol 43, 734–743 Lahti R, Perala M, Heikinheimo P, Pitkaranta T, Kukko-Kalske E & Heinonen J (1991a) Characterization of the 5¢ flanking region of the Escherichia coli ppa gene encoding inorganic pyrophosphatase: mutations in the ribosome-binding site decrease the level of ppa mRNA J Gen Microbiol 137, 2517–2523... T, Uchiumi T, Yonezawa T, Salminen A, Baykov AA, Lahti R & Hachimori A (1998) Cloning and expression of a unique inorganic pyrophosphatase from Bacillus subtilis: evidence for a new family of enzymes FEBS Lett 439, 263–266 Parfenyev AN, Salminen A, Halonen P, Hachimori A, Baykov AA & Lahti R (2001) Quaternary structure and metal ion requirement of family II pyrophosphatases from Bacillus subtilis, Streptococcus... J, Goldman A, Heinonen J, Baykov AA, Cooperman BS & Lahti R (1995) Structure and function analysis of Escherichia coli inorganic pyrophosphatase: is a hydroxide ion the key to catalysis? Biochemistry 34, 782–791 21 Celis H, Franco B, Escobedo S & Romero I (2003) Rhodobacter sphaeroides has a family II pyrophosphatase: comparison with other species of photosynthetic bacteria Arch Microbiol 179, 368–376... Heinonen J (1985) Inorganic pyrophosphate and inorganic pyrophosphatase in Escherichia coli Int J Biochem 17, 575–580 Kaneko T, Sato S, Kotani H, Tanaka A, Asamizu E, Nakamura Y, Miyajima N, Hirosawa M, Sugiura M, Sasamoto S et al (1996) Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp PCC 6803 PCC 6803 II Sequence determination of the entire genome and assignment of. .. 11–16 Lahti R, Pikaranta T, Valve E, Ilta I, Kukko-Kalse E & Heinonen J (1988) Cloning and characterization of the gene encoding inorganic pyrophosphatase of Escherichia coli K-12 J Bacteriol 170, 5901–5907 Hanahan D (1983) Studies on transformation of Escherichia coli with plasmids J Mol Biol 166, 557–580 Fiske CH & Subarow Y (1925) The colorimetric determination of phosphorus J Biol Chem 66, 375–400... Expression studies of two paralogous ppa genes encoding distinct family I pyrophosphatases in marine unicellular cyanobacteria reveal inactivation of the typical cyanobacterial gene Biochem Biophys Res Commun 295, 890–897 23 Lawrence JG, Hendris RW & Casjens S (2001) Where are the pseudogenes in bacterial genomes? Trends Microbiol 9, 535–540 24 Klemme JH, Klemme B & Gest H (1971) Catalytic properties and. .. coli soluble inorganic pyrophosphatase ˚ at 2.7A resolution Protein Eng 7, 823–830 19 Efimova IS, Salminen A, Pohjanjoki P, Lapinniemi J, Magretova NN, Cooperman BS, Goldman A, Lahti R & Baykov AA (1999) Directed mutagenesis studies of the metal binding site at the subunit interface of Escherichia coli inorganic pyrophosphatase J Biol Chem 274, 3294–3299 20 Salminen A, Kapyla TJ, Heikinheimo P, Kankare... rapid and sensitive method for the quantitation of microgram quantities of protein Pyrophosphatases from photosynthetic bacteria 40 41 42 43 44 45 46 47 utilizing the principle of protein-dye binding Anal Biochem 72, 248–254 Harlow E & Lane D (1988) Antibodies: a Laboratory Manual Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, NY Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman... potential proteincoding regions DNA Res 3, 109–136 Gingrich JC, Gasparich GE, Sauer K & Bryant DA (1989) Nucleotide sequence and expression of the two genes encoding D2 protein and the single gene encoding the CP43 protein of photosystem II in the cyanobacterium Synechococcus sp PCC 7002 Photosynthesis Res 24, 137–150 Cohen-Bazire G, Sistrom WR & Stanier RY (1957) Kinetic studies of pigment synthesis... Losada M & Serrano A (2006) A novel subfamily of monomeric inorganic pyrophosphatases in photosynthetic eukaryotes Biochem J 395, 211– 221 Perez-Castineira JR, Gomez-Garcia R, Lopez-Marques RL, Losada M & Serrano A (2001) Enzymatic systems of inorganic pyrophosphate bioenergetics in photosynthetic and heterotrophic protists: remnants or metabolic cornerstones? Int Microbiol 4, 135–142 3958 ´ ´ M R Gomez-Garcıa . Comparative biochemical and functional studies of family I soluble inorganic pyrophosphatases from photosynthetic bacteria Marı ´ aR.Go ´ mez-Garcı ´ a*,. families: family I, widespread in all types of organism [2], and family II, so far confined to a limited number of bacteria and archaea [3,4]. The families