Detection of an L-amino acid dehydrogenase activity in Synechocystis sp. PCC 6803.

The protein Slr0782 from Synechocystis sp. PCC 6803, which has similarity to L-amino acid oxidase from Synechococcus elongatus PCC 6301 and PCC 7942, has been characterized in part. Immunoblot blot analysis showed that Slr0782 is mainly thylakoid membrane-associated. Moreover, expression of slr0782 mRNA and Slr0782 protein were analyzed and an activity assay was developed. Utilizing toluene-permeabilized cells, an L-arginine-stimulated O(2) uptake became detectable in Synechocystis sp. PCC 6803. Besides oxidizing the basic L-amino acids L-arginine, L-lysine, L-ornithine, and L-histidine, a number of other L-amino acids were also substrates, while D-amino acids were not. The best substrate was L-cysteine, and the second best was L-arginine. The L-arginine-stimulated O(2) uptake was inhibited by cations. The inhibition by o-phenanthroline and salicylhydroxamic acid suggested the presence of a transition metal besides FAD in the enzyme. Moreover, it is shown that inhibitors of the respiratory electron transport chain, such as KCN and 2,5-dibromo-3-methyl-6-isopropyl-p-benzoquinone, also inhibited the L-arginine-stimulated O(2) uptake, suggesting that Slr0782 functions as an L-arginine dehydrogenase, mediating electron transfer from L-arginine into the respiratory electron transport chain utilizing O(2) as electron acceptor via cytochrome oxidase. The results imply that Slr0782 is an additional substrate dehydrogenase being able to interact with the electron transport chain of the thylakoid membrane.

Introduction

In general, L-amino acid oxidases (L-Aoxs) are homodimeric flavoproteins containing a non-covalently bound FAD as cofactor in each subunit (Bright and Porter, 1975; Curti ; Meister, 1965). They catalyze the oxidative deamination of L-amino acids to produce the corresponding 2-keto acid, ammonium, and hydrogen peroxide by utilizing molecular oxygen as electron acceptor. L-Aoxs occur in insects (Ahn et al., 2000), snake venomes (Massey and Curti, 1967; Meister, 1965; Du and Clemetson, 2002; Torii et al., 1997), and are also found in bacteria (Coudert, 1975; Brearly ), fungi (Kusakabe ; DeBusk and Ogilvie, 1984; Niedermann and Lerch, 1990), algae (see below), snails (Ehara ), and mammals (Nakano ).

The crystal structure of the snake venom L-Aox from Calloselasma rhodostoma (Pawelek ) and Agkistrodon halys pallas (Zhang ), and of the bacterial L-Aox from Rhodococcus opacus (Faust ) have been published. A phylogenetic tree of the evolutionary distances of a number of L-Aoxs has recently been given (Gau ; Macheroux ; Nishizawa ).

Besides L-Aoxs, which utilize molecular O2 as electron acceptor leading to hydrogen peroxide formation, a few amino acid dehydrogenases have been described in the literature, which are associated with a respiratory electron transport chain and which donate electrons to the quinone pool. Subsequently, these electrons are transferred to a terminal cytochrome oxidase leading to the reduction of molecular oxygen to water. This reaction is inhibited by KCN. Such a membrane-bound amino acid oxidase is, for example, a D-amino acid dehydrogenase being associated with the respiratory electron transport chain in E. coli (Franklin and Venables, 1976; Olsiewski ; Jones and Venables, 1983; Anraku and Gennis, 1987). D-alanine was shown to be the most active substrate, and the enzyme is suggested to contain FAD and a non-heme (heme AE) iron.

Another Aox, which is associated with the respiratory electron transport chain, is a L-cysteine oxidase in the cytoplasmic membrane of Neisseria meningitides (Yu and DeVoe, 1981). This enzyme is inhibited by o-phenanthroline and salicylhydroxamic acid (SHAM) implying the presence of a metal cofactor besides the presence of FAD. The function of this enzyme in the metabolism is yet unclear. A membrane-associated L-Aox is also present in Proteus rettgeri (Duerre and Chakrabarty, 1975) and in Proteus mirabilis (Pelmont ). The Proteus rettgeri enzyme has a high specificity for basic L-amino acids. Both enzymes are suggested to interact with the respiratory electron transport chain.

Information on L-Aoxs in photosynthetic organisms is still rather scarce. Three eukaryotic marine phytoplankton species (Pleurochrysis, Prymnesium, and Amphidinium) have been shown to contain cell-surface located L-Aoxs for the utilization of L-amino acids as a nitrogen source (Palenik and Morel, 1990, b). These L-Aoxs have broad substrate specificity and low Km values in the μM range. In the unicellular green alga Chlamydomonas reinhardtii, either two L-Aoxs or two forms of the same L-Aox are present, which become induced when no primary nitrogen source is available to produce ammonia (Munoz-Blanco ; Piedras ; Vallon ). The marine red alga Gymnogongrus flabelliformis contains an L-Aox catalyzing the oxidative deamination of basic L-amino acids, L-citrulline and L-methionine, and of the non-proteinogenic amino acid L-gigartinine (Fujisawa ). An L-Aox with a somewhat broader substrate specificity is also present in the calcareous marine red alga Amphiroa crassissima Yendo (Ito ). Moreover, it has been shown that the red sea weed Chondrus crispus contains an L-Aox located in the apoplast, which has a function in pathogen defence (Weinberger ).

Previously, it has been shown that the two closely related mesophilic cyanobacteria Synechococcus elongatus PCC 6301 and Synechococcus elongatus PCC 7942 (subsequently named S. elongatus PCC 6301/PCC 7942) possess an L-Aox, which catalyzes the oxidative deamination of the basic L-amino acids L-Arg>L-Lys>L-Orn>L-His utilizing O2 as electron acceptor resulting in the formation of the corresponding 2-keto acid, ammonium, and hydrogen peroxide, an activity, which is strongly inhibited by cations (Pistorius ; Pistorius and Voss, 1980; Engels ; Gau ). The L-Aoxs of S. elongatus PCC 6301/PCC 7942 are in part located in the soluble protein fraction of the periplasm and in part in the spheroplast fraction, mainly precipitating with the membrane fraction (Bockholt ; S Schriek, unpublished results). The L-Aoxs of S. elongatus PCC 6301/PCC 7942 contain FAD as cofactor and are encoded by the aoxA genes (syc0596_c) (Gau ). The two L-Aoxs are 100% identical and encode proteins of 54 kDa with a calculated isoelectric point of 7.9. When the aoxA gene in S. elongatus PCC 7942 was insertionally inactivated, cells no longer grew with L-arginine as the sole N-source, suggesting that this enzyme is the only one that enables the cells to utilize extracellularly added L-arginine as a N-source (Bockholt ).

Both cyanobacteria contain an additional gene with similarity to the aoxA gene, which has been called aoxB, being syc1144_c for S. elongatus PCC 6301 and being synpcc7942_0369 for S. elongatus PCC 7942 (Gau ). The product of the aoxB gene has not yet been characterized biochemically. Whether AoxB is the thylakoid membrane-bound L-Aox, catalyzing the conversion of phenylalanine to phenylpyruvic acid (Loeffelhardt, 1977), can not yet be answered.

A similar L-Aox activity with high specificity for basic L-amino acids, with L-arginine being the best substrate, has also been detected in Synechococcus cedrorum PCC 6908 (Gau ). Moreover, a bioinformatic analysis of 24 cyanobacterial genomes revealed the presence of one or two gene(s) with similarity to AoxA in 10 other cyanobacterial species (Gau ; Schriek, 2008; Schriek et al., 2007).

In the present paper, our interest was focused on the AoxA-similar protein Slr0782 in Synechocystis sp. PCC 6803 for two main reasons. Although L-Aox activity could easily be detected in S. elongatus PCC 6301/PCC 7942 by measuring L-arginine-stimulated O2 uptake, which is inhibited by cations such as CaCl2, in cell suspensions as well as in cell-free extracts of Synechocystis sp. PCC 6803, we had not been able to detect such activity. The second reason is related to the observation that a complex interrelationship seems to exist between L-arginine catabolism and photosynthesis/respiration in Synechocystis sp. PCC 6803, especially when the light intensity during growth was set to 200 μmol photons m−2 s−1 (Schriek et al. 2008; Schriek 2008; Stephan ). Previous results have shown that a PsbO-free Synechocystis sp. PCC 6803 mutant was much better able to utilize L-arginine as the sole N-source than the wild type (WT), suggesting that a change on the donor side of photosystem II (PSII) has a substantial influence on L-arginine catabolism. Therefore, we were interested in finding conditions under which an L-Aox activity was detectable in Synechocystis sp. PCC 6803 and to see whether such results might help to explain why a PsbO-free mutant is better able than WT to utilize L-arginine as the sole N-source when the light intensity during growth corresponds to approximately 200 μmol photons m−2 s−1.

Materials and methods

Cyanobacterial strains, growth conditions, and cell harvest

The cyanobacterial strain Synechocystis sp. PCC 6803 WT was obtained from the Pasteur Culture Collection of Cyanobacterial Strains, Paris, France. The PsbO-free mutant was the same as described earlier (Engels ). Cyanobacteria were cultivated in 250 ml gas wash bottles (16 cm in height and 4.5 cm in diameter), which were placed in a water bath (size 40×70×25 cm) set at 30 °C. Illumination was with six beams (Philips Cool Spot, angle 12°, 150 W) placed 36 cm above the culture bottles. The ambient illumination at the outside of the culture bottles corresponded to 200 μmol photons m−2 s−1. The light intensity was determined with a LI-250 light meter with a LI-190SZ quantum sensor (Li-Cor, Lincoln, Nebraska, USA) measuring photons between 400–700 nm. Since the cultures revealed a higher growth rate at 200 μmol photons m−2 s−1 than at an illumination of 50 μmol photons m−2 s−1 (not shown), this light regime obviously did not cause substantial photoinhibitory damage. The BG11 medium was continuously bubbled via two manometers (Porter Instruments, Hatfield, USA) with 20 l h−1 2% CO2-enriched air. Growth was performed with a slightly modified BG11 medium (Stephan ) with nitrate as the N-source (17.7 mM sodium nitrate) or with nitrate-free BG11 medium containing 5 mM L-arginine-HCl to which 50 mM EPPS-NaOH pH 7.5 was added to prevent acidification. After 48 h of growth with nitrate or L-arginine the pH of the CO2-aerated BG11 medium was between 7.0 and 7.5 (Nodop ). Growth of the PsbO-free mutant was performed as described above, except that the growth medium contained kanamycin sulphate (7.5 mg l−1). The standard culture inoculum corresponded to an absorbance of 0.3 at 750 nm (OD750 nm). Growth rates were determined by measuring the OD750 nm of cell suspensions. The chlorophyll content was determined according to previously published protocols (Grimme and Boardman, 1972).

RNA isolation and slot-blot RNA–DNA hybridization analysis

Isolation of total RNA was performed as described previously (Michel ). The isolation protocol was improved by an on-column DNase digestion step with the RNase-free DNase set from Qiagen (Qiagen, Hilden, Germany). RNA quantification was performed with a NanoDrop ND-1000 spectrometer (PeqLab, Erlangen, Germany), and the integrity and quality of RNA samples were checked with an Agilent Bioanalyser chip. RNA was stored at –80 °C. In slot-blot RNA–DNA hybridization experiments, 2 μg total RNA were denatured for 10 min at 68 °C in a formaldehyde/formamide-containing buffer and transferred to HybondN+ membranes (GE HealthCare, Freiburg, Germany) with a Bio-Rad Dot-blot SF Microfiltration Apparatus (Bio-Rad, Munich, Germany). RNA was UV-cross-linked to the membrane, and samples were probed with different PCR-derived digoxygenin-dUTP (Dig-dUTP) labelled gene-specific DNA probes (Table 1). Detection was performed with the CDP-Star ready-to-use system (Roche, Mannheim, Germany) according to the manufacturer's recommendation. The rnpB probe was used to ensure equal loading.

Table 1.

Oligonucleotides used for slot-blot RNA-DNA hybridzation

Primer	Name	Amplified product	DNA sequence 5′→3′ direction
slr0782	slr0782-F	1325 bps	CCATCCTCGTCCTGTGATTG
	slr0782-R		CCAGTACGAATTGCACCATC
sll1077	speB2-F	1054 bps	CAGCAGGAGGTTGACCAAGG
	speB2-R		CAGCATGGATATAGGCCGGT
slr1022	argD-F	1224 bps	GTTGTTGAATCCGTCGAAGC
	argD-R		TTCTGCTTCCGTCACCACTA
slr0370	gabD-F	895 bps	GCCGAGGAATACTTAGCCGA
	gabD-R		GGTTAGTTGTCCATGCACTG
rnpB	slr1469-F	599 bps	GCGGCCTATGGCTCTAATCA
	slr1469-R		TTGACAGCATGCCACTGGAC

Preparation of cell-free extracts, SDS PAGE, and immunoblotting

Preparation of cell-free extracts, SDS PAGE, and immunoblotting were performed as previously described (Tölle ). Cells were harvested by centrifugation, resuspended in 50 mM TRIS-HCl, pH 7.4, containing 5 mM EDTA-NaOH, pH 7.4, 1 mM benzamidine, 0.1 mM PMSF, and broken in a Hybaid Ribolyser (FastPrep Instrument, Q-biogene, Heidelberg, Germany). The extract was centrifuged for 1 min at 16,200 g at 4 °C, and the supernatant was used. Samples corresponding to 15 μg proteins were denatured for 20 min at 60 °C using a 20 mM dithiothreitol- and 4% SDS-containing buffer and subjected to TRIS-glycine SDS PAGE. Subsequently, the proteins were transferred to BA-85S nitrocellulose membranes (Schleicher and Schuell, Dassel, Germany) as described previously (Michel ). The antisera used were the anti-PsbA (dilution 1:2,000) (Engels ) and anti-PetA (dilution 1:1,500) (Michel ; Pietsch et al., 2007). The anti-Slr0782 antiserum was raised in rabbits against heterologously expressed and affinity-purified Slr0782 (Pineda Antikörper-Service, Berlin, Germany) (Schriek, 2008). The second antibody was a peroxidase-coupled swine anti-rabbit IgG (dilution 1 to 250–500) or an alkaline phosphatase-conjugated goat anti-rabbit IgG (dilution 1:500, DAKO A/S, Glostrup, Denmark). Immunoblot staining was performed either with the ECL Detection Kit (GE Healthcare, Munich, Germany) or with nitrotetrazolium blue/X-phosphate as substrate for the alkaline phosphatase.

Immunocytochemistry and sucrose density centrifugation

Synechocystis sp. PCC 6803 cells were grown and harvested as described above. Cell pellets were washed three times with EM buffer (50 mM KH2PO4-Na2HPO4, pH 7). Embedding and cutting of ultra-thin sections was performed according to previously described protocols (Engels ). Immunocytochemistry with the anti-Slr0782 antiserum at a dilution of 1:100 in TBS buffer was carried out according to Stephan . Sucrose density centrifugation was carried out as previously described (Omata and Murata, 1983, 1984).

Toluene treatment of cells and measurements of L-arginine-stimulated oxygen uptake

Permeabilization of Synechocystis sp. PCC 6803 cells was carried out as described before (Quintero ). After organic solvent treatment, cells were kept on ice for 30 min in the dark. Afterwards, cells corresponding to 5 μg Chl were tested for their L-arginine-stimulated O2 uptake activity in a Clark-type electrode. The assay consisted of a total volume of 3 ml containing 5 mM Tricine-NaOH pH 8.5, 0.5 mM MgCl2, and 5 mM L-arginine. Inhibitors or other amino acids were added as indicated.

Results

The cyanobacterium Synechocystis sp. PCC 6803 contains the gene slr0782 encoding a protein with similarity to the aoxA gene encoding an L-Aox in S. elongatus PCC 7942/PCC 6301 (Gau ; Schriek ). In the present paper, a partial characterization of the Slr0782 enzyme in Synechocystis sp. PCC 6803 is presented.

Bioinformatic analysis of the slr0782 gene and the Slr0782 protein in Synechocystis sp. PCC 6803

The gene slr0782 of Synechocystis sp. PCC 6803 is located on the chromosome between sll0750 encoding the histidine kinase. Genes sasA and sll0751 both encode yet uncharacterized proteins. The gene slr0782 starts with a GTG start codon and consists of 1415 bps. It encodes a protein of 471 amino acid residues with a deduced molecular weight of 51, 404 kDa. Slr0782 contains 52 negatively and 38 positively charged amino acid residues resulting in a calculated isoelectric point of 5.19 (Expasy, ProtParam, http://www.expasy.ch/cgi-bin/protparam). Evaluation of the primary amino acid sequence with different software packages (DAS membrane prediction tool: http://www.sbc.su.se/∼miklos/DAS/) revealed that Slr0782 contains one putative transmembrane helix at the N-terminus and another one at the C-terminus of the protein. Slr0782 contains a typical GXGXXG dinucleotide-binding motif at its N-terminal end (Wierenga ; Macheroux ) (Fig. 1). The protein shares high similarity to the aoxA encoded gene product of S. elongatus PCC 6301/PCC 7942 (syc0596_c and synpcc7942_0946: 21% identical, 23% similar, and 13% weakly similar amino acid residues). The AoxA protein has previously been characterized on a biochemical basis and shown to be an L-amino acid oxidase (Pistorius and Voss, 1980; Engels ; Gau ). A slightly higher similarity of Slr0782 than with AoxA is observed with AoxB of S. elongatus PCC 6301/PCC 7942 (syc1144_c and synpcc7942_0369: 31% identical, 22% strongly similar, and 13% weakly similar amino acid residues), which has not yet been characterized. Slr0782 has also similarity to FAD-containing monoamine oxidases (InterProScan, http://www.ebi.ac.uk/Tools/InterProScan/) (Gau ). Although the similarity of Slr0782 to AoxA and AoxB is relatively high, especially in the FAD-binding domain, there also exists a major difference between the protein sequences. Slr0782 in Synechocystis sp. PCC 6803 does not contain the N-terminal twin-arginine motif of AoxA and AoxB in S. elongatus PCC 6301/PCC 7942 (Gau ). A twin-arginine motif is part of a translocation pathway signal (Berks, 1996).

Fig. 1.

Sequence alignment of the extended putative dinucleotide-binding fold of Slr0782 from Synechocystis sp. PCC 6803 with the dinucleotide-binding fold consensus sequence (Macheroux ) and to the dinucleotide-binding sequences of AoxA and AoxB from Synechococcus elongatus PCC 7942. A full alignment of Slr0782 with AoxA has been published previously (Schriek ). The term majority refers to the sequence being present in the majority of the investigated proteins (Macheroux ).The presence of FAD has been experimentally proven for AoxA (Pistorius and Voss, 1980; Engels ). The alignment was performed with the ClustalW software.

Expression of Slr0782 on mRNA and protein level in Synechocystis sp. PCC 6803

The steady-state transcript pool of the slr0782 mRNA and the concentration of the Slr0782 protein were investigated in Synechocystis sp. PCC 6893 WT and a PsbO-free mutant, when grown either with nitrate or L-arginine as the sole N-source and with an illumination of 200 μmol photons m−2 s−1 for 24 h, 48 h, and 72 h. The PsbO-free mutant was included in the investigation, since previous results provided evidence that the PsbO-free mutant was better able than WT to grow with L-arginine as the sole N-source, when the light intensity corresponded to 200 μmol photons m−2 s−1 (Stephan ; Schriek , Schriek 2008). This suggests a complex interrelationship between L-arginine catabolism and photosynthesis/respiration, since lack of the Mn/Ca2+-stabilizing PsbO protein of PSII has a substantial effect on the ability of Synechocystis sp. PCC 6803 to utilize L-arginine.

Slot-blot RNA–DNA hybridization experiments showed that the slr0782 mRNA level increased, when cells were grown with L-arginine as compared to the growth of cells with nitrate (Fig. 2). The increase was substantially higher in the PsbO-free mutant than in WT. The same trend was observed for the Slr0782 protein in immunoblot experiments utilizing an antiserum raised against the recombinant Slr0782 protein. As documented in Fig. 3, the Slr0782 concentration was higher in the mutant than in WT. It was also higher in L-arginine-grown mutant cells after 48 h of growth compared to nitrate-grown mutant cells. Thus, Slr0782 expression was substantially higher in the mutant than in WT. In the L-arginine oxidase/dehydrogenase pathway, as being characterized in Pseudomonas putida (Miller and Rodwell, 1971; Vanderbilt ; Cunin ; Lu, 2006), L-arginine is degraded to succinate by the enzymes L-arginine oxidase/dehydrogenase, 4-guanidine butyrase, 4-aminobutyrate transaminase, and succinate semialdehyde dehydrogenase. The transcripts encoding these enzymes were also up-regulated in the PsbO-free mutant of Synechocystis sp. PCC 6803, when grown with L-arginine (Fig. 2). This result suggests that in the PsbO-free mutant L-arginine can effectively be metabolized to succinate.

Fig. 2.

Transcript analysis of genes encoding enzymes of the L-amino acid dehydrogenase pathway in Synechocystis sp. PCC 6803. Slot-blot RNA–DNA hybridization was performed with total RNA extracted from Synechocystis sp. PCC 6803 WT and the PsbO-free mutant, and with Dig dUTP-labelled gene-specific probes. Cells were grown for 24, 48, or 72 h with nitrate or L-arginine as the sole N-source at a light intensity of 200 μmol photons m−2 s−1. After cell harvest, total RNA was extracted and 2 μg of RNA each were hybridized with gene-specific probes against the L-amino acid dehydrogenase slr0782 transcript, the 4-guanidinobutyrase sll1077 transcript, the 4-aminobutyrate transaminase sll1022 transcript, and the succinate semialdehyde dehydrogenase slr0370 transcript. The rnpB probe was used to assure equal loading of RNA.

Fig. 3.

Expression of the L-amino acid dehydrogenase Slr0782 in Synechocystis sp. PCC 6803 WT and the PsbO-free mutant under selected growth conditions. Cells were grown for 24, 48, or 72 h with nitrate or L-arginine as the sole N-source at a light intensity of 200 μmol photons m−2 s−1. Cell-free extracts corresponding to 50 μg protein were resolved on SDS PAGE, transferred on to nitrocellulose membranes, and probed with the anti-Slr0782 antiserum (dilution 5,000). The anti-Slr0782 antiserum recognizes a 51 kDa band corresponding to the calculated molecular mass of Slr0782. A slower moving band of approximately 70 kDa is most likely due to an association of part of Slr0782 with lipids and/or cations altering the electrophoretic mobility during SDS PAGE. Immunoblot detection of AtpA (CF1 of ATP synthase) is given as an additional reference for equal protein loading.

Localization of Slr0782 in Synechocystis sp. PCC 6803

Immunocytochemical investigation with the anti-Slr0782 antiserum and a gold-conjugated anti-rabbit-IgG as second antibody provided evidence that Slr0782 is almost exclusively located on the thylakoid membranes in the PsbO-free mutant of Synechocystis sp. PCC 6803 (Fig. 4A). Sucrose density centrifugation to obtain subcellular fractions confirmed the presence of a substantial amount of the Slr0782 protein in the thylakoid membrane fraction (Fig. 4B). Moreover, differential detergent membrane solubilization showed that part of the Slr0782 protein was found in a PSII-enriched fraction, which had been solubilized by simultaneous treatment of membranes with octylglucoside and β-D-dodecyl maltoside according to the procedure given in Burnap . This suggests that either the Slr0782 protein is located close to the PSII complex in the thylakoid membrane or PSII and Slr0782 are co-solubilized by the detergent concentration used.

Fig. 4.

Localization of the L-amino acid dehydrogenase Slr0782 in the PsbO-free Synechocystis sp. PCC 6803 mutant. Cells were grown for 48 h with nitrate as sole N-source. (A) Immunocytochemistry was carried out with the anti-Slr0782 antiserum and a gold-conjugated anti-rabbit IgG as second antibody (dilution 1:100). (B) Fractions of sucrose density gradient centrifugation containing the thylakoid membrane fraction or the soluble protein fraction were probed with the anti-PsbA- and the anti-Slr0782 antiserum. In this experiment, the anti-Slr0782 antiserum detected an additional third band of Slr0782, which is running slightly faster than the 71 kDa band of Slr0782. This result might be due to a specific modification of Slr0782 with lipids or degradation processes of the protein during the isolation procedure in spite of the presence of protease inhibitors.

Partial characterization of the Slr0782 activity in Synechocystis sp. PCC 6803

In S. elongatus PCC 6301/PCC 7942 an L-arginine-stimulated O2 uptake, which is cation-inhibited, as for example, by CaCl2, could easily be detected in cell suspensions as well as in cell-free extracts (Pistorius and Voss, 1980; Engels ; Gau ). This activity has been shown to be due to the aoxA gene product (Gau ). Since Slr0782 has substantial similarity to AoxA, it was rather puzzling that so far we had not been able to detect such an L-arginine-stimulated O2 uptake in Synechocystis sp. PCC 6803. Therefore, substantial efforts were made to find conditions under which such an L-arginine-stimulated O2 uptake becomes detectable in Synechocystis sp. PCC 6803. This was not possible either with cell suspensions or with cell-free extracts. However, when toluene-permeabilized cells were used, an L-arginine stimulated O2 uptake finally became detectable. For these experiments, Synechocystis sp. PCC 6803 WT and the PsbO-free mutant were cultivated with L-arginine and an illumination of 200 μmol photons m−2 s−1 for 48 h. The cells were harvested and permeabilized with toluene according to the protocol given by Quintero . The permeabilized cells were kept on ice for at least 30 min prior to use.

Toluene-treated Synechocystis sp. PCC 6803 cells have a low O2 uptake based on the presence of endogenous substrate or substrates (Fig. 5). When L-arginine was added to toluene-treated Synechocystis cells, a substantial increase of O2 uptake was observed in the PsbO-free mutant cells, while WT showed no or only a minor increase in oxygen consumption. When a low concentration of MgCl2 (0.5 mM) was added to the reaction mixture, the O2 uptake rate upon addition of L-arginine was stimulated 1.7-fold in the PsbO-free mutant cells and corresponded to 22.5 μmol O2 taken up mg−1 Chl h−1 (Fig. 5) Under such conditions WT also showed a low L-arginine-stimulated O2 uptake of 5.6 μmol O2 taken up mg−1 Chl h−1. The slight enhancement of the O2 uptake activity was seen with MgCl2 but not with other alkaline earth metal ions such as Ca2+ or Sr2+. The higher activity in the PsbO-free mutant as compared to WT is in good agreement with the higher Slr0782 protein concentration in the mutant as compared to WT (Fig. 3; 48 h of growth).

Fig. 5.

Activity slopes of L-arginine-stimulated oxygen consumption in toluene-treated Synechocystis sp. PCC 6803 WT and the PsbO-free mutant cells. Cells were grown for 3 d with 5 mM L-arginine as the sole N-source in nitrate-free BG11 medium, which was buffered with 50 mM EPPS-NaOH, pH 7.5. After cell harvest and toluene-treatment, the permeabilized cells corresponding to 5 μg Chl were tested for their L-arginine-stimulated oxygen consumption in a Clark-type oxygen electrode. The assay consisted of a total volume of 3 ml containing 5 mM Tricine-NaOH pH 8.5, 0.5 mM MgCl2, and cells corresponding to 5 μg Chl. 5 mM L-arginine was added at the time point marked by the black arrow after a constant O2 uptake rate due to endogenous substrate(s) was observed for at least 3 min (filled symbols). The corresponding control slopes were recorded for assays with cells to which no L-arginine was added (open symbols).

Since the L-arginine-stimulated O2 uptake was substantially higher in the PsbO-free mutant than in WT, all subsequent experiments were performed with toluene-permeabilized PsbO-free mutant cells grown with L-arginine as the sole N-source and with an illumination of 200 μmol photons m−2 s−1 for 48 h. Moreover, in all assay mixtures 0.5 mM MgCl2 was added to optimize the O2 uptake activity. As shown in Table 2, the enzyme acts stereospecifically on L-arginine as substrate. No activity was measurable with D-arginine as substrate. The basic L-amino acids, L-arginine, L-lysine, L-ornithine, and L-histidine, are substrates for the enzyme with L-arginine being the best substrate (Table 2), and the Km value for L-arginine is 3.2 mM. In addition to the tested basic L-amino acids, several other exogenously added L-amino acids were able to stimulate O2 uptake in toluene-permeabilized cells. Among these, L-cysteine was very active and caused an O2 uptake that was about 50% higher than that with L-arginine. Boiled toluene-treated cells did not cause an O2 uptake, which proved that the measured O2 uptake with L-cysteine is of an enzymatic nature and is not due to an autoxidation phenomenon. This is also supported by the fact that D-cysteine is not a substrate. Thus, the Slr0782 enzyme of Synechocystis sp. PCC 6803 oxidizes basic L-amino acids as does the AoxA enzyme of Synechococcus elongatus PCC 6301/PCC 7942, but Slr0782 has a broader substrate specificity than AoxA.

Table 2.

Oxygen uptake of toluene-treated PsbO-free Synechocystis mutant cells after addition of L- or D-amino acids

Substrate	Relative O2-consumption (%)
L-Arginine	100%
L-Lysine	65%
L-Ornithine	44%
L-Histidine	44%
L-Cysteine	150%
L-Leucine	52%
L-Phenylalanine	51%
L-Tyrosine	51%
L-Valine	40%
L-Threonine	40%
L-Alanine	37%
L-Glutamate	29%
L-Methionine	29%
L-Tryptophane	28%
L-Proline	27%
L-Glycine	13%
L-Isoleucine	12%
L-Glutamine	9%
L-Aspartate	nda
L-Asparagine	nda
L-Serine	nda
D-Arginine	nda
D-Cysteine	nda

Cells were grown for 2 d with L-arginine as the sole N-source and a light intensity of 200 μmol photons m−2 s−1. The assay contained cells corresponding to 5 μg Chl in a total volume of 3 ml 50 mM Tricine-NaOH pH 8.5, 0.5 mM MgCl2, and L- or D-amino acids (5 mM) as indicated. Measurements of oxygen consumption were carried out in a Clark-type electrode. The activity value obtained with 5 mM L-arginine was set to 100%. The given values for relative O2-consumption are representative of five independent experiments.

nd, not detectable.

Although the L-arginine-stimulated O2 uptake of Synechocystis sp. PCC 6803 is stimulated by low MgCl2 concentrations of 0.5 mM, increased MgCl2 concentrations had an inhibitory effect. 2 mM MgCl2 gave 50% inhibition and 10 mM MgCl2 completely inhibited O2 uptake (Table 3). Besides Mg2+, other divalent as well as monovalent cations inhibited the activity of Slr0782. The concentration necessary for a 50% inhibition of L-arginine-stimulated oxygen uptake for a selection of cations is given in Table 3. Moreover, the O2 uptake was inhibited by o-phenanthroline, SHAM (salicylhydroxamic acid), KCN, and DBMIB (2,5-dibromo-3-methyl-6-isopropyl-p-benzoquinone). Since the AoxAs of Synechococcus elongatus PCC 6301/PCC 7942 are inhibited by cations as well as by o-phenanthroline, it is highly likely that these substances also inhibit Slr0782 in Synechocystis sp. PCC 6803. Most likely, the inhibition by SHAM is also due to a direct inhibition of Slr0782, since a membrane-associated L-cysteine oxidase in Neisseria meningitides has been shown to be inhibited by SHAM (Yu and DeVoe, 1981). By contrast, the inhibition of the L-arginine-stimulated O2 uptake by DBMIB points to a participation of the cytochrome b6/f complex, and the inhibition by KCN points to a participation of a cytochrome oxidase in the L-arginine-stimulated O2 uptake. Thus, the results suggest that the Slr0782 enzyme feeds electrons from L-arginine oxidation and also from the oxidation of other L-amino acids into the respiratory electron transport chain and does not directly interact with molecular oxygen. Therefore, it is concluded that the Slr0782 enzyme in Synechocystis sp. PCC 6803 is an L-arginine dehydrogenase and not an L-arginine oxidase. This would explain the great difficulties, which we had to detect this activity in the past. An activity of Slr0782 can obviously only be detected, when the enzyme is still connected to an intact respiratory electron transport chain, which does not seem to be the case in cell-free extracts. The slight stimulatory effect of a low MgCl2 concentration is most likely due to Mg2+ acting as a bridging cation to optimize the interaction of proteins involved in the overall reaction mediating electron transport from Slr0782 to the cytochrome oxidase. The absence of a detectable activity in cell suspensions is most likely due to the inability of the cells to take up L-arginine under our assay conditions. Thus, it can be concluded that toluene-permeabilized cells retain a functional interaction of the Slr0782 enzyme with an intact respiratory chain, and these permeabilized cells enable the exogenously added L-arginine to reach the enzyme in the thylakoid membrane.

Table 3.

Inhibition of L-arginine-stimulated oxygen uptake in toluene-treated PsbO-free Synechocystis mutant cells

Inhibitors	50% inhibition
NaCl	8.00 mM
MgCl2	1.66 mM
SrCl2	0.13 mM
CaCl2	0.09 mM
ZnCl2	0.20 mM
CoCl2	0.11 mM
NiCl2	0.09 mM
o-Phenanthroline	0.05 mM
Dipyridyl	0.01 mM
SHAM	0.02 mM
DBMIB	0.03 mM
KCN	0.10 mM
NaN3	0.10 mM

Cells were grown for 2 d with L-arginine as the sole N-source and a light intensity of 200 μmol photons m−2 s−1. The assay contained cells corresponding to 5 μg Chl in a total volume of 3 ml 50 mM Tricine-NaOH pH 8.5, 0.5 mM MgCl2, and the inhibitors as indicated. Concentrations for 50% inhibition of activity were calculated using linear regression analysis for data from at least five different concentrations. Inhibitor solvents other than water were tested for non-specific side-effects on L-arginine-stimulated oxygen uptake. The given values are representative of three independent experiments.

Are L-arginine oxidation via the Slr0782 enzyme and photosynthetic water oxidation via PSII alternative reactions feeding electrons into the thylakoid membrane embedded electron transport chain?

As shown in Figs 2 and 3, when the PsbO-free mutant was grown with L-arginine as the sole N-source, the steady-state level of slr0782 mRNA and also the Slr0782 protein were up-regulated. If L-arginine oxidation and water oxidation represent alternative reactions feeding electrons into the plastoquinone pool, water oxidation should be down-regulated. As shown in Fig. 6, especially in the early phase of growth with L-arginine, the protein level for PsbA (PSII reaction AE center protein) was much lower in L-arginine-grown cells than in nitrate-grown cells. No large difference was seen for the cytochrome b6/f complex, which is required for photosynthetic as well as respiratory electron transport activity. Measurements of the fluorescence yield by pulse-amplified modulation (PAM) (Table 4) provided evidence that the PSII activity was highly reduced in cells grown with L-arginine as the sole N-source as compared to cells grown with nitrate, which is in agreement with the lower amounts of PSII reaction AE centre protein PsbA.

Fig. 6.

Immunoblot analysis of selected electron transport-associated proteins in Synechocystis sp. PCC 6803 WT and the PsbO-free mutant. Cells were grown for 24, 48, or 72 h with nitrate or L-arginine as the sole N-source at a light intensity of 200 μmol photons m−2 s−1. Cell-free extracts corresponding to 50 μg protein were resolved on SDS PAGE, transferred on to nitrocellulose membranes, and probed with the antisera raised against PsbA (reaction center protein D1 of PSII; dilution 1:2,000) and PetA (Cytf subunit of the Cyt b6/f complex; dilution 1:1,1500).

Table 4.

PAM measurements of photosynthetic yield of Synechocystis sp. PCC 6803 WT and the PsbO-free Synechocystis mutant

Synechocystis sp. PCC 6803	WT		PsbO-free mutant
N-source	NO3-	L-Arg	NO3-	L-Arg
	Photosynthetic yield (rel. units)		Photosynthetic yield (rel. units)
Growth for 24 h	0.36	0.08	0.16	0.03
Growth for 48 h	0.49	0.01	0.30	0.04
Growth for 72 h	0.49	0.11	0.32	0.01

Cells were grown for 24 h, 48 h or 72 h under constant illumination with 200 μmol photons m−2 s−1 either with nitrate or L-arginine as the sole N-source. Prior to PAM measurements, cells were washed once and concentrated to give a Chl concentration of 1 mg ml−1 (see Materials and methods). The given activity values are representative of three independent experiments.

Discussion

The results presented here suggest that Slr0782 in Synechocystis sp. PCC 6803 is an L-amino acid dehydrogenase (L-amino acid:plastoquinone oxidoreductase) feeding electrons from L-amino acid oxidation into the respiratory electron transport chain. This implies that the measured L-arginine-stimulated oxygen consumption is mediated by a cytochrome oxidase and not directly by Slr0782. Thus, Slr0782 is not an L-amino acid oxidase, which donates electrons directly to O2 to give hydrogen peroxide. To our knowledge, Slr0782 of Synechocystis sp. PCC 6803 is the first L-amino acid dehydrogenase in a cyanobacterium, for which an association with the thylakoid membrane-associated respiratory electron transport chain is shown.

Although the enzyme has a broad substrate specificity, it is likely that it functions mainly as an L-arginine dehydrogenase. The enzyme has a Km value in the mM range, and such a substrate concentration will, most likely, only be reached with L-arginine. This amino acid is taken up effectively by Synechocystis sp. PCC 6803 (Montesinos ), is the end-product of an alternative CO2 fixation pathway (Linko ; Tabita, 1987, 1994), and might also accumulate at such a concentration intracellularly, when cyanophycin (multi-L-arginyl-poly-L-aspartate) (Simon, 1971, 1987; Allen, 1984; Kolodny ; Maheswaran ) becomes degraded. The presence of an L-arginine dehydrogenase as an alternative substrate dehydrogenase sheds new light on the special role of L-arginine in the cyanobacterial metabolism and might contribute to a better understanding of the complex dynamic metabolism of cyanophycin as a N- and also C-reservoir (Simon, 1971, 1987; Allen, 1984, 1988; Mackerras , b; Berg ; Maheswaran ) and of the complex interrelationship of L-arginine catabolism with photosynthesis/ respiration (Schriek 2008, Schriek et al., 2007, Schriek et al., 2008, Stephan et al.., 2000)

Our model (Fig. 7) implies that under conditions, under which the L-arginine concentration reaches a threshold level, L-arginine via Slr0782 and water via PSII are alternative electron donors to the electron transport chain. L-arginine oxidation only proceeds well, when a reduction of photosynthetic water oxidation exists as being the case in the PsbO-free mutant. This model could explain why such a big phenotypical difference between Synechocystis sp. PCC 6803 WT and the PsbO-free mutant had previously been observed when grown with L-arginine under continuous light of about 200 μmol photons m−2 s−1 (Stephan ; Schriek et al., 2008). In contrast to the PsbO-free mutant, L-arginine oxidation via the respiratory electron transport chain can not proceed effectively in WT, because water oxidation is optimized in the presence of PsbO. PsbO has been suggested to have a regulatory function for photosynthetic water oxidation (Spetea ; Sherman ; Tucker ; De Las Rivas and Barber, 2004; Heide ; Lundin ). The signals that determine the ratio of the photosynthetic to the respiratory electron transport in the light is still mainly unknown (Vermaas, 2001).

Fig. 7.

Model of the function of the Slr0782 protein in Synechocystis sp. PCC 6803 as an alternative substrate dehydrogenase of the thylakoid membrane-associated electron transport chain. In this model, water via PSII (water:plastoquinone oxidoreductase) and L-arginine via Slr0782 (L-arginine:plastoquinone oxidoreductase) are alternative electron donors for the thylakoid membrane-associated electron transport system of Synechocystis sp. PCC 6803. Under high light conditions, as for example, growth with 200 μmol photons m−2 s−1, L-arginine oxidation via Slr0782 seems to proceed effectively only when a limitation in PSII exists, as being the case when PsbO is lacking. Moreover, CaCl2 has an antagonistic effect on water oxidation (Ca2+ and Cl– are stimulatory) and on L-arginine oxidation (Ca2+ and also other cations are inhibitory).

Although our results provide evidence that genes encoding the enzymes of the entire L-arginine dehydrogenase pathway (L-arginine to succinate) for the utilization of L-arginine as a N- as well as the C-source are present in Synechocystis sp. PCC 6803 (Fig. 2), WT as well as the PsbO-free mutant were not able to grow with L-arginine in the presence of DCMU (not shown). DCMU is known to bind to the PsbA protein of PSII (Ort and Yocum, 1996; Barber and Kuhlbrandt, 1999).

Since the L-arginine-stimulated O2 uptake in Synechocystis sp. PCC 6803 was not inhibited by DCMU, this observation suggests that a small amount of a functional PSII reaction centre has to be present for effective growth with L-arginine. If this is not the case, an imbalance of C- to N-metabolites or of ATP to NADPH might occur. No growth in the presence of DCMU was also reported for Aphanocapsa sp. PCC 6308 when cultivated with L-arginine (Weathers ).

To elucidate further the complex interrelationship of this enzyme with the electron transport chain and with the overall L-arginine metabolism, the L-arginine dehydrogenase-containing multiprotein complex has to be isolated from the thylakoid membrane to allow for a more detailed characterization of this complex and the Slr0782 enzyme.