The Unique Cyanobacterial Protein OpcA Is an Allosteric Effector of Glucose-6-phosphate Dehydrogenase in Nostoc punctiformeATCC 29133*

Glucose-6-phosphate dehydrogenase (G6PD), encoded by zwf, is essential for nitrogen fixation and dark heterotrophic growth of the cyanobacterium Nostoc punctiforme ATCC 29133. In N. punctiforme,zwf is part of a four-gene operon transcribed in the orderfbp-tal-zwf-opcA. Genetic analyses indicated thatopcA is required for G6PD activity. To define the role ofopcA, the synthesis, aggregation state, and activity of G6PD in N. punctiforme strains expressing different amounts of G6PD and/or OpcA were examined. A single tetrameric form of G6PD was consistently observed for all strains, as well as for recombinantN. punctiforme His-G6PD purified from Escherichia coli, regardless of the quantity of OpcA present. However, His-G6PD and the G6PD of strain UCD 351, which lacks OpcA, had low affinities for glucose 6-phosphate (G6P) substrate (K m (app) = 65 and 85 mm, respectively) relative to wild-type N. punctiforme G6PD (K m (app) = 0.5 mm). Near wild-type affinities for G6P were observed for these enzymes when saturating amounts of His-OpcA- or OpcA-containing extract were added. Kinetic studies were consistent with OpcA acting as an allosteric activator of G6PD. A role in redox modulation of G6PD activity was also indicated, because thioredoxin-mediated inactivation and reactivation of His-G6PD occurred only when His-OpcA was present.

Cyanobacteria are members of a phylogenetically cohesive group of prokaryotes that perform oxygenic photosynthesis and assimilate CO 2 via the reductive pentose phosphate (Calvin) cycle. Excess photosynthate produced in the light is stored as glycogen, which is catabolized to provide maintenance energy during periods of darkness. Although most cyanobacteria are restricted to this photoautotrophic mode of growth, some, such as the filamentous heterocyst-forming isolate Nostoc punctiforme ATCC 29133, exhibit greater nutritional versatility. When provided with exogenous sugars such as fructose or glucose, N. punctiforme is capable of both photoheterotrophic growth in the light and heterotrophic growth in the dark.
The oxidative pentose phosphate (OPP) 1 cycle is assumed to be the major route for carbon catabolism in cyanobacteria, both in vegetative cells for the breakdown of glycogen or exogenously supplied sugars, and in nitrogen-fixing heterocysts, where it provides a reductant (NADPH) required by nitrogenase (1)(2)(3). Regulation of the activity of glucose-6-phosphate dehydrogenase (G6PD; E.C. 1.1.1.49), the enzyme controlling the entry of carbon into the cycle, is complex and poorly understood. Metabolites such as NADPH, ATP, glucose 6-phosphate (G6P), glutamine, and ribulose bisphosphate have been implicated in its regulation (4,5) as has thioredoxin, which may reductively inactivate and oxidatively reactivate the enzyme in the light and dark, respectively, to prevent futile cycling, which would occur if the Calvin and OPP cycles operated simultaneously (5)(6)(7)(8). G6PD has been purified from the filamentous, heterocyst-forming cyanobacteria Anabaena variabilis (5) and Anabaena sp. strain PCC 7120 (Anabaena 7120) (4,8). Although the A. variabilis enzyme is a 250-kDa tetramer (5), the Anabaena 7120 enzyme is found in multiple aggregation states. Three forms (M 1 , M 2 , and M 3 ; 120, 240, and 345 kDa, respectively) with different catalytic activities were initially reported for Anabaena 7120 G6PD (4); the 240-kDa M 2 form was by far the most prevalent, but the larger M 3 form appeared to be more active. More recently, the Anabaena 7120 enzyme was reported to exist in two aggregation states (250 and 750 kDa) (8). The equilibrium between the different forms of the Anabaena 7120 enzyme depends upon pH, the concentration of the enzyme in solution, and the presence or absence of substrates or other effectors. Multiple reactive bands have also been observed in extracts from Anabaena 7120 and some other cyanobacterial strains after native polyacrylamide gel electrophoresis (PAGE) and staining of the gels for G6PD activity (9). The sizes, numbers, and intensities of the bands vary with the strain and with growth conditions. A genetic analysis of the physiological role of G6PD and the OPP cycle in N. punctiforme was recently performed (10). In N. punctiforme, zwf, the gene encoding G6PD, is part of a fourgene operon (fbp-tal-zwf-opcA) 2 with multiple transcriptional start points and complex transcriptional regulation (11). Insertional inactivation of zwf resulted in essentially complete loss of G6PD activity and rendered the zwf mutant incapable of nitrogen fixation or dark heterotrophic growth, indicating that G6PD is required for both of these processes (10). Inactivation of opcA gave rise to a mutant with the same phenotype. The opcA mutant was fully complemented by opcA supplied in trans on a multicopy plasmid, whereas complementation of the zwf mutant with zwf on the same multicopy plasmid abolished the growth defects but restored G6PD activity to only 25% of wildtype levels. If both zwf and opcA were present on the multicopy plasmid, however, the G6PD activity of the complemented zwf mutant was increased 20-fold over wild-type levels. These results imply that the downstream gene opcA is required for the synthesis or activation of fully functional G6PD.
The opcA gene has since been identified in all other cyanobacteria examined. Outside of the division Cyanobacteria, a derived amino acid sequence with significant similarity across the entire length of the polypeptide has not been observed. A limited similarity to restricted regions of the polypeptide can be detected in derived sequences from other microorganisms by BLAST analysis (available from the National Center for Biotechnology Information on the Web) (12). In Synechococcus sp. PCC 7942 (Synechococcus 7942) (13) and Anabaena 7120 (available from Kazusa DNA Research Institute on the Web), opcA is located directly downstream from zwf, whereas the two genes are unlinked in Synechocystis sp. strain PCC 6803 (also available from the National Center for Biotechnology Information on the Web). Insertional inactivation of opcA in Synechococcus 7942 resulted in a mutant that had lost G6PD activity and was no longer viable in the dark (9), a phenotype consistent with that observed for N. punctiforme (10). On the basis of G6PD activity staining of mutant and wild-type cell extracts on native PAGE gels, it has been proposed that OpcA catalyzes the formation of active G6PD oligomers in Synechococcus 7942 (9). We report here the results of further experiments undertaken with N. punctiforme that are consistent with OpcA acting as an allosteric activator of G6PD and not as a chaperone for the assembly of G6PD oligomers.

EXPERIMENTAL PROCEDURES
Strains, Culture Conditions, and Cell Lysis-Plasmids and bacterial strains used in this study are shown in Table I. N. punctiforme 29133 and Anabaena 7120 were cultured in modified Allen and Arnon medium (14) at 25°C in an illuminated (9 watts/m 2 ) New Brunswick incubator with shaking at 150 rpm. The medium was supplemented with 2.5 mM NH 4 Cl and 5 mM MOPS, pH 7.8, with 50 mM fructose or glucose, or with ampicillin (Ap) or neomycin (Nm) at 10 g/ml when appropriate. When darkness was required, culture flasks were wrapped in multiple layers of aluminum foil to exclude light. Cells were typically harvested after 7-10 days when culture densities reached 2-3 g of chlorophyll a (Chl a)/ml.
Extracts used for SDS and native PAGE and enzyme assays were prepared from cells grown as described above and harvested by centrifugation at 6000 ϫ g, 4°C. Cell pellets were washed once in 50 mM Tris maleate buffer (pH 6.5) containing 10 mM MgCl 2 , resuspended in the same buffer, and broken with several passages through a chilled French pressure cell at 15,000 p.s.i. Extracts were clarified by centrifugation for 5 min at 16,000 ϫ g, 4°C prior to use. Protein assays and determinations of the Chl a content of cyanobacterial strains were performed as described previously (15).
Strain Construction and Protein Purification-Routine DNA manipulations in Escherichia coli strain DH5␣-MCR were performed according to standard procedures (16). DNA-modifying enzymes were purchased from New England BioLabs and Life Technologies, Inc. and were used according to the manufacturers' instructions. Large scale plasmid preparations were obtained using a commercial kit (Qiagen). N. punctiforme strain UCD 466 (zwf::P psbA nptII) was constructed by digesting pSCR115 with PmlI to remove a 0.7-kb fragment from the zwf gene. This fragment was replaced with a 1.1-kb-blunted XbaI fragment from pRL448 containing P psbA nptII (encoding Nm resistance) to generate pSCR346, which was digested with DraI and XbaI. The 6.7-kb fragment containing the nptII-interrupted zwf gene was ligated to the sacB-containing conjugatable vector pRL271, which had been digested with Ecl136II and XbaI. In the resulting plasmid, pSCR347, P psbA nptII is in the same orientation as zwf and transcription of opcA occurs. Introduction of pSCR347 into wild-type N. punctiforme by conjugation, selection for Nm-and sucrose-resistant exconjugants, and confirmation of the insertional inactivation by Southern blotting was performed as described previously (10,14). Strain UCD 482 was constructed by introducing pSCR121 into wild-type N. punctiforme via electroporation as described (10). The construction of strains UCD 341, 348, 351, and 364 is described elsewhere (10,17).
Affinity-tagged G6PD and OpcA proteins were prepared using the pET system of Novagen or the pMAL protein fusion and purification system of New England BioLabs. DNA fragments containing the N. punctiforme zwf or opcA genes were obtained using the polymerase chain reaction (PCR) with pSCR115 as a template and the following primers: 5Ј-ATGGTTAGTCTGCTAGAAAATCCCTTGC-3Ј (His-G6PD forward) and 5Ј-GCTCTAGAGTCTGCGCCAGCGACGGCCATCC-3Ј (His-G6PD reverse), or 5Ј-ATGACAAAATCCAAAATCTCAAATCCAA-AACTTCC-3Ј (MBP-OpcA and His-OpcA forward) and 5Ј-GCTCTAGA-CGGTGTATACACAAGTCGTTAAAACCCG-3Ј (MBP-OpcA reverse) or 5Ј-CGCGCTCGAGGTTAGGTCTAAGTTAATTCG-3Ј (His-OpcA reverse). PCR amplifications were performed in a PerkinElmer Cetus thermal cycler with AmpliTaq polymerase. The 1.5-kb PCR product containing zwf was initially cloned into the EcoRV site of pBluescript II KS(ϩ). Digestion with BamHI and HindIII released a 1.5-kb fragment containing zwf, which was inserted into the BamHI and HindIII sites of the expression vector pET-28a(ϩ) to generate pSCR197. pSCR195, used to generate MBP-OpcA for antibody production, was constructed by cloning a 1.6-kb PCR fragment containing opcA into the EcoRV site of pBluescript II KS(ϩ). This fragment was removed from pBluescript II KS(ϩ) by digestion with SalI and PstI and was then ligated into the SalI and PstI sites of pMAL-c2. A 1.4-kb PCR fragment containing opcA was digested with XhoI and inserted into the Ecl136II and XhoI sites of pET-28a(ϩ) to generate pSCR352, which was used to produce His-OpcA. MBP-OpcA was soluble when expressed in E. coli DH5␣-MCR, whereas a portion of each of the His-tagged proteins was present in inclusion bodies when expressed in E. coli BL21(DE3). In each case, the soluble protein was purified by affinity chromatography according to the kit manufacturers' instructions, with phenylmethylsulfonyl fluoride (1 mM) added to all buffers used during purification. Purified proteins were stored at 4°C at concentrations of 0.5-1.5 mg of protein/ml. The His-tagged proteins tended to precipitate out of solution; precipitation occurred more readily when protein concentrations were higher. His-G6PD retained activity for 2-3 weeks when stored at 4°C in elution buffer (20 mM Tris-HCl, 500 mM NaCl, 1 M imidazole, pH 7.8). His-OpcA rapidly lost its ability to activate G6PD when it was stored in elution buffer; consequently, it was dialyzed in several changes of 50 mM Tris maleate buffer (pH 6.5) containing 10 mM MgCl 2 prior to storage at 4°C for 2-3 weeks. Neither protein was stable at Ϫ20°C when stored in buffers containing 0 -50% glycerol nor under any condition if diluted to less than 0.5 mg/ml. Losses of activity during purification and storage have been reported previously for G6PD (4,8,18,19).
PAGE and Immunoblots-SDS and native PAGE and Coomassie Blue staining were performed according to standard protocols (20) using either the Bio-Rad Mini-Protean II system or the larger Bio-Rad Protean system. Rabbit polyclonal antisera were raised against purified His-G6PD or MBP-OpcA by standard techniques (21) and were used at a dilution of 1:400. Immunoblotting was performed as described previously (22). Bound primary antibody was detected either with horseradish peroxidase-conjugated goat ␣-rabbit secondary antibody and a p-iodophenol/Luminol chemiluminescent protocol (22) or with alkaline phosphatase-conjugated goat ␣-rabbit secondary antibody (1:50,000 dilution) and Amersham Pharmacia Biotech ECF substrate. When the latter detection method was used, detection and quantitation of chemifluorescent bands was performed using a Storm PhosphorImager with ImageQuaNT software (Molecular Dynamics). Immunoprecipitations were performed according to standard methods (21) with 100 l of N. punctiforme extract (10 -20 mg of protein/ml) and 5 l of ␣-G6PD or ␣-OpcA antiserum or preimmune serum. After 1 h on ice, 100 l of 10% (v/v) protein A-Sepharose CL-4B beads was added, and the mixture was incubated 1 h at 4°C with gentle rocking. The beads were washed four times with ice-cold lysis buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% Nonidet P-40). Electrophoresis of the immune complexes was performed using 10% polyacrylamide SDS-PAGE gels with ␤-mercaptoethanol omitted from the sample buffer. Immunoprecipitated proteins were detected by Western blotting.
The molecular masses of His-G6PD and G6PDs in cyanobacterial extracts were determined by Ferguson analysis using 5, 6, and 7% acrylamide native PAGE gels. Standards used for native PAGE were obtained from Sigma Chemical Co. and included the following: bovine serum albumin (66 and 132 kDa), ␤-amylase (200 kDa), apoferritin (450 kDa), urease (545 kDa), and thyroglobulin (669 kDa). The gels were stained with Coomassie Blue or G6PD activity stain (4) to identify standards and unknowns, respectively. Average R F values for each protein at each gel concentration were determined by measuring the distances from the center of the bands to the dye front on duplicate gels.
G6PD Activity Assay-G6PD activity was assayed by a modification of the method of Schaeffer and Stanier (4), in which the standard assay NADP concentration was increased from 0.5 to 2.0 mM. The standard assay mixture also contained 50 mM Tris maleate, pH 7.5, 10 mM MgCl 2 , and 5 mM G6P (dipotassium salt). Where indicated, the G6P and NADP concentrations were varied for the determination of kinetic parameters. The reaction was initiated by the addition of cell extract or purified protein, and the reduction of NADP at room temperature was followed by monitoring the increase in absorbance at 340 nm, typically for 1 min. To ensure that initial velocities were measured, all assays were determined to be in the linear range with respect to protein concentration, time, and substrate utilization. With crude cell extracts, in which the enzyme concentration varied, at least three total protein concentrations were surveyed for linearity before extended assays were done at the selected protein concentration. For purified His-G6PD, activity in standard assays was linear between at least 0.2 and 2.0 g of protein and 0.5 g was typically used. The quantity of G6P and NADP substrate consumed was calculated at all substrate, protein, and effector concentrations used. For determinations of K m (app) for NADP, less than 5% of the initial concentrations of NADP or G6P were consumed and for determinations of K m (app) for G6P, the amounts of G6P and NADP consumed were always less than 2% of the initial concentrations. In all experiments, at all concentrations of protein, substrates, and effector, changes in A 340 were linear with time, implying that any potential end product inhibition by NADPH or 6-phosphogluconolactone was inconsequential. Kinetic data were analyzed using Kaleida-Graph v. 3.08. Spirulina thioredoxin and the chemical oxidants and reductants used in redox modulation experiments were obtained from Sigma Chemical Co.

RESULTS
Growth Condition-dependent Activity and Aggregation State of G6PD-To determine whether growth conditions affected the activity or aggregation state of N. punctiforme G6PD, cells were cultured with ammonium in the light (photoautotrophic conditions), without ammonium in the light (photoautotrophic, nitrogen-fixing conditions), or in the dark in the absence of ammonium and presence of 50 mM fructose (heterotrophic, nitrogen-fixing conditions). Cell extracts prepared from each culture were first assayed for G6PD activity and were then separated by electrophoresis on 10% acrylamide native PAGE gels, blotted to nitrocellulose membrane, and incubated with ␣-G6PD or ␣-OpcA antibodies. The G6PD activity clearly varied with growth conditions. In the experiment shown in Fig. 1, the activity of cells grown with ammonium in the light was 0.035 mol⅐min Ϫ1 ⅐mg of protein Ϫ1 . G6PD activity increased by 1.6-fold in nitrogen-fixing cells grown in the light (0.055 mol⅐min Ϫ1 ⅐mg of protein Ϫ1 ) and by 7.6-fold (0.265 mol⅐min Ϫ1 ⅐mg of protein Ϫ1 ) in nitrogen-fixing cells grown heterotrophically in the dark with fructose. The increases in G6PD activity were accompanied by increased synthesis of G6PD and OpcA, but the aggregation states of G6PD and OpcA were unchanged (Fig. 1).
Molecular Mass and Monomer Composition of Native G6PD-N. punctiforme G6PD consistently migrated as a single band with a molecular mass that appeared to be between 200 and 545 kDa based on the positions of native PAGE standards, whereas OpcA migrated as a single band near the 66-kDa marker (Fig. 1). Because the predicted molecular masses of the N. punctiforme zwf and opcA gene products are 58 and 51 kDa, respectively, these results implied that the G6PD band was an oligomer and the OpcA band was a monomer. G6PD and OpcA from N. punctiforme extracts were never observed to comigrate on native PAGE gels (e.g. compare lanes 1-3 with lanes 4 -6 in Fig. 1). To confirm the implication that OpcA was not associated with the G6PD oligomer in N. punctiforme cell extracts, immunoprecipitation experiments were performed using either ␣-G6PD or ␣-OpcA antibodies. Each protein was immunoprecipitated with its respective antibody, but the two proteins did not coimmunoprecipitate, and neither cross-immunoprecipitation nor immunoprecipitation with pre-immune serum occurred (data not shown).
Ferguson analysis was used to more precisely determine the molecular mass of the G6PD oligomer. Samples containing N. punctiforme cell extract, Anabaena 7120 cell extract (control), or N. punctiforme His-G6PD purified from E. coli were loaded on a series of native PAGE gels (5, 6, and 7% acrylamide) along with native protein standards. After electrophoresis, the proteins were visualized with Coomassie Blue (standards) or G6PD activity stain (samples containing G6PD). A single strong band that appeared within 15 min of G6PD activity staining was observed for His-G6PD and in N. punctiforme and Anabaena 7120 extracts (Fig. 2). A faint lower molecular mass band was occasionally detected in Anabaena 7120 extract after several hours of staining; no additional bands were observed with His-G6PD or N. punctiforme extracts after prolonged staining, nor were additional bands detected with extracts from cells grown with glucose or fructose (data not shown). R F values were calculated for each protein, and the molecular masses of the cyanobacterial G6PDs were determined as shown in Fig. 3  (A and B). The molecular masses of the G6PDs in N. punctiforme and Anabaena 7120 cell extracts were 200 and 263 kDa, respectively, whereas the molecular mass of His-G6PD was 229 kDa. There was no change in the migration of His-G6PD upon the addition of His-OpcA, nor was the aggregation state of His-G6PD altered by increasing its concentration in solution to 2.5 mg/ml, by changing the buffer pH from 7.9 to 6.5 or by incubating it with G6P or NADP (data not shown).
Effect of the Stoichiometry Between OpcA and G6PD on G6PD Activity and Aggregation State-To determine whether differences in the amount of OpcA present in the cell affected the quantity, aggregation state, or activity of N. punctiforme G6PD, G6PD assays and quantitative Western blots were performed with extracts from several strains that produced amounts of OpcA and G6PD that differed from those normally found in wild-type N. punctiforme. The results are summarized in Table II. Wild-type N. punctiforme extract typically contained 1 or 2 mg of OpcA and G6PD monomers per g of cell protein when the cells were cultured with ammonium or dinitrogen, respectively. The aggregation state of G6PD did not change when the stoichiometry between OpcA and G6PD was altered (Table II); in all strains that carried one or more intact copies of the zwf gene, G6PD monomers were synthesized and assembled into a single aggregation state that corresponded to the 200-kDa form found in wild-type N. punctiforme. The G6PD activity varied, however. In strain UCD 351, which produced 15 times more G6PD than the wild-type, but lacked OpcA, G6PD activity was nearly 9-fold lower than in wild-type. Activity was at baseline levels in strains UCD 466 and UCD 341, which lacked G6PD but differed in their OpcA content. In strains that produced both G6PD and OpcA, G6PD activity depended on the total quantity of G6PD present and on the stoichiometry between OpcA and G6PD: In strain UCD 482, which synthesized normal amounts of G6PD in the presence of elevated amounts of OpcA, G6PD activity was similar to that of wild-type. Near-wild-type G6PD activity was also observed for strain UCD 364, which overproduced G6PD in the presence of wild-type amounts of OpcA. A large (23-fold) increase in G6PD activity was observed only for strain UCD 348, which overproduced both OpcA and G6PD.  1 and 4), photoautotrophically with dinitrogen (lanes 2 and 5), or heterotrophically in the dark with glucose and dinitrogen (lanes 3 and 6) as described under "Experimental Procedures." Aliquots of the cell extracts containing 70 g of total protein were loaded on a 10% acrylamide gel, and the proteins were separated by native PAGE. The experiments described above implied that G6PD synthesized in the absence of OpcA was assembled into its native aggregation state but was still relatively inactive. To determine whether the activity of this inactive G6PD could be restored by the addition of OpcA, extract from strain UCD 351 (G6PD overproduced, no OpcA) was mixed with extract from strain UCD 466 (no G6PD, OpcA overproduced). In the experiment shown in Fig. 4, the amount of strain UCD 351 extract was held constant, and increasing amounts of strain UCD 466 extract were added until no further change in G6PD activity was observed. The G6PD activity of the strain UCD 351 extract increased nearly 20-fold, from 0.007 mol⅐min Ϫ1 ⅐mg of protein Ϫ1 in the absence of strain UCD 466 extract to 0.127 mol⅐min Ϫ1 ⅐mg of protein Ϫ1 when strain UCD 466 extract was saturating ( Fig. 4 and Table III). The increase in G6PD activity upon the addition of strain UCD 466 extract occurred within 30 s after the mixing of the extracts and remained constant after incubation of the mixture on ice for more than 1 h. There was no increase in G6PD activity when strain UCD 341 extract (no G6PD or OpcA) was added to strain UCD 351 extract, nor was there any change in the G6PD activity of wild-type N. punctiforme extract when extract from strain UCD 466 was added (data not shown). Quantitative Western blotting indicated that the strain UCD 351 extract used in the experiment shown in Fig. 4 contained 10.3 ng of G6PD/g of total protein, whereas the strain UCD 466 extract contained 179 ng of OpcA/g of total protein. Therefore, the concentration of G6PD in the assay was 13.0 nM, and the concentration of OpcA was varied between 0 and 360 nM. When the data were fitted to the Hill equation and plotted (Fig. 4, inset), the OpcA concentration at half-maximal velocity (S 0.5 ) was determined to be 11.0 nM. Additional experiments performed with G6PD and OpcA from different sources are summarized in Table III. Although the concentrations of UCD 351 G6PD and His-G6PD used in these experiments were similar, the concentration of OpcA required for half-maximal activation was lower for UCD 351 extract G6PD than for His-G6PD.
␣-OpcA antibodies were used to examine the interaction between G6PD and OpcA. Aliquots (40 l) of ␣-OpcA or preimmune serum were added either to His-OpcA (4.4 g, 80 pmol)

TABLE II Activity and aggregation state of G6PD in N. punctiforme strains expressing different amounts of G6PD and OpcA protein
Cell extracts were prepared from N. punctiforme strains grown photoautotrophically with ammonium. The G6PD and OpcA monomer content relative to wild-type N. punctiforme was determined using SDS-PAGE and quantitative Western blotting with ␣-G6PD and ␣-OpcA antibodies. G6PD activity was assayed using standard assay conditions. G6PD activity is reported as mean Ϯ S.E. with the number of trials given in parentheses, and the units of G6PD activity are mol of NADP reduced ⅐ min Ϫ1 ⅐ mg protein Ϫ1 . The aggregation state of G6PD was determined for each strain by comparing the migration of its G6PD to that of wild-type N. punctiforme and to protein standards using native PAGE and Western blotting. alone or to a mixture containing an identical amount of His-OpcA plus His-G6PD (0.5 g, 8 pmol). The samples were incubated on ice for 1 h. The serum-treated His-OpcA samples were then mixed with His-G6PD (0.5 g, 8 pmol) to test their ability to enhance G6PD activity, whereas the serum-treated samples containing the mixture of His-OpcA and His-G6PD were assayed directly for G6PD activity. His-OpcA that had been incubated with ␣-OpcA retained only 3% of its initial ability to activate His-G6PD; no loss in the ability to activate G6PD was observed when His-OpcA was incubated with preimmune serum. Inactivation of His-OpcA also occurred when it was mixed with His-G6PD prior to the addition of ␣-OpcA; only 15% of the initial activity of the His-G6PD plus His-OpcA protein mixture remained after incubation with ␣-OpcA, whereas 87% of the initial activity of the mixture remained after incubation with preimmune serum. Kinetic Constants of G6PD Measured in the Absence and Presence of OpcA-Additional experiments were performed with cell extracts and the purified proteins to determine the nature of the defect in N. punctiforme G6PD synthesized in the absence of OpcA. A comparison of kinetic parameters revealed that the affinity of G6PD for G6P substrate was ϳ30-fold lower when OpcA was absent. When the amount of cell extract was held constant and the concentration of G6P was varied, a sigmoidal rate versus substrate concentration curve was obtained with strain UCD 351 extract, whereas a hyperbolic curve was obtained with wild-type N. punctiforme extract (Fig.  5A). A sigmoidal curve was also obtained with His-G6PD (Fig.  5B). The apparent K m for G6P (S 0.5 ) was determined to be 0.5 Ϯ 0.1 mM for wild-type N. punctiforme G6PD, 85 mM for strain UCD 351 G6PD, and 65 Ϯ 10 mM for His-G6PD (Table IV). Near-wild-type affinities for substrate were restored to His-G6PD and strain UCD 351 G6PD when saturating amounts of His-OpcA were added to the assay (Fig. 5B and Table IV). As illustrated in Fig. 5B, the shift in apparent K m was unaccompanied by a change in V max . The kinetics of NADP cofactor binding were also examined (Table IV). At saturating G6P concentrations, NADP binding curves were nearly hyperbolic. In the absence of OpcA, the apparent K m for NADP was 48 M for strain UCD 351 extract G6PD and 49 Ϯ 4.2 M for His-G6PD. Upon the addition of His-OpcA, a 2-fold reduction in the apparent K m for NADP occurred and K m (app) values nearly identical to those of wild-type N. punctiforme extract were obtained.
Effect of Redox Modulators on OpcA and G6PD Activity-The activity of cyanobacterial G6PD is reported to depend upon its oxidation-reduction state; therefore, the potential for redox regulation of His-G6PD was examined. First, chemical oxidizing or reducing agents were added to His-G6PD samples to a final concentration of 5 or 50 mM. The samples were incubated 30 min on ice and assayed for G6PD activity. Neither the oxidizing agents sodium tetrathionate (STT), diamide, hydrogen peroxide, or oxidized glutathione nor the reducing agents dithiothreitol (DTT) or reduced glutathione affected the activity of His-G6PD that had not been previously activated by His-OpcA; however, when DTT was added at high concentration (50 mM) to a mixture containing both His-G6PD and His-OpcA, the G6PD activity of the mixture declined 77% after 5 min and 96% after 30 min. No activation of His-G6PD occurred when His-OpcA was incubated with 50 mM DTT for 30 min prior to the mixing of the two proteins.
To determine whether similar results could be obtained with thioredoxin, Spirulina thioredoxin that had been reduced or oxidized with 2 mM DTT or 2 mM STT, respectively, was added to His-G6PD or His-OpcA. Like the chemical-oxidizing and -reducing agents, reduced or oxidized thioredoxin had no effect on the activity of His-G6PD when it was added in the absence of His-OpcA (data not shown). However, when reduced thioredoxin was added to His-OpcA, His-G6PD added subsequently to the His-OpcA plus reduced thioredoxin mixture was no longer activated (Fig. 6). The G6PD-activating ability of His-OpcA was restored when STT was added to the His-OpcA/thioredoxin mixture at a concentration that was twice the DTT concentration (Fig. 6), and this reduction/oxidation cycle of inactivation/ activation could be repeated in the same sample. When thioredoxin was omitted from the His-OpcA/thioredoxin mixture, the low concentrations of DTT and STT appeared to have little effect, and activation of His-G6PD by His-OpcA was always observed (Fig. 6). DISCUSSION The opcA gene was previously shown to be required for G6PD activity in N. punctiforme (10) and Synechococcus 7942 (9), and it has been proposed that OpcA is involved in the assembly of active G6PD oligomers (9). Using N. punctiforme strains that lack OpcA or express amounts of G6PD and OpcA that differ from wild-type, we have shown that the assembly of G6PD into its native tetrameric form occurs in N. punctiforme even in the absence of OpcA or when the stoichiometry between OpcA and G6PD is greatly altered. Although N. punctiforme G6PD is assembled correctly in the absence of OpcA, it is essentially inactive, because it has a very low affinity for G6P substrate. Our kinetic studies with G6PD and OpcA indicate that OpcA may function as a positive allosteric effector that shifts the K m for G6P substrate to a physiologically relevant value. OpcA may also be involved in redox regulation of G6PD activity, because redox modulators affected G6PD activity only when OpcA was present.
During nitrogen fixation and dark heterotrophic growth, the cellular demand for reductant and, ultimately, energy from the OPP pathway is high, and increased transcription of the zwf and opcA genes of N. punctiforme has been documented (11). In extracts from N. punctiforme grown under these conditions, G6PD activity was elevated above the levels detected in cells grown photoautotrophically with ammonium. Although there was increased synthesis of G6PD and OpcA, no change in the aggregation state of G6PD was detected (Fig. 1). Thus, rather than producing more highly aggregated and catalytically more active forms of G6PD when carbon flow through the OPP path- way increased, N. punctiforme simply synthesized more OpcA and additional G6PD oligomers of a single aggregation state (Fig. 1).
G6PD purified from prokaryotic and eukaryotic organisms most often exists as a homodimer or homotetramer, although larger oligomers have been described (18). Enzyme concentration, pH, ionic strength, and the presence of divalent cations, G6P, or NADP have all been reported to affect the activity and aggregation state of the enzyme in vitro (18). Several aggregation states that differ in their kinetic properties have been described in Anabaena 7120 (4, 8), and multiple bands have been detected in extracts from several cyanobacterial strains on native PAGE gels stained for G6PD activity (9). Ferguson analysis in our laboratory revealed that the molecular masses of the G6PD in N. punctiforme and Anabaena 7120 cell extracts were 200 and 263 kDa respectively, whereas N. punctiforme His-G6PD synthesized in and purified from E. coli was 229 kDa (Fig. 3B). The value of 263 kDa we obtained for the enzyme in Anabaena 7120 extract is in good agreement with the 240-kDa value obtained by Schaeffer and Stanier (4) for the predominant M 2 form of the enzyme, as well as with the 250-kDa value obtained for the smaller of the two forms observed by Gleason TABLE III Determination of the OpcA concentration required for half-maximal activation of G6PD from different sources The G6PD activity of strain UCD 351 extract (70 g of total protein; 13 pmol of G6PD) or His-G6PD (1.0 g; 16 pmol) was first determined in the absence of added OpcA using standard assay conditions (2 mM NADP and 5 mM G6P) and is reported in the "No Added OpcA" column. The G6PD concentration was then held constant at 13 nM for UCD 351 G6PD and 16 nM for His-G6PD, and increasing amounts of His-OpcA or strain UCD 466 extract were added until G6PD activity no longer increased and OpcA became saturating. The maximum G6PD activity achieved is shown in the "ϩ OpcA" column. The units of G6PD activity are mol of NADP reduced ⅐ min Ϫ1 ⅐ mg UCD 351 extract or His-G6PD Ϫ1 , and activity is reported as mean Ϯ S.E. with the number of trials given in parentheses. Values of n, the Hill coefficient, and S 0.5 , the concentration of OpcA required for half-maximal activation of G6PD, were determined by fitting the data to the Hill equation.

Source of G6PD
Source of OpcA  5. The affinity of N. punctiforme G6PD for G6P substrate was increased in the presence of OpcA. A, activity versus G6P substrate concentration curves for G6PD in strain UCD 351 extract (open circles) or wild-type N. punctiforme extract (closed circles). Strain UCD 351 overproduces G6PD but lacks OpcA, whereas wild-type N. punctiforme contains approximately equal amounts of both proteins. Here the amount of each extract was held constant at 70 g of total protein, NADP was held constant at 2 mM, and the G6P concentration was varied. Because strain UCD 351 contains more G6PD per unit total protein than wild-type N. punctiforme, G6PD activity for each strain is expressed as a percentage of the maximum activity obtained for that strain. B, activity versus G6P substrate concentration curves for purified His-G6PD in the absence (open squares) or presence (closed squares) of saturating His-OpcA. His-G6PD (0.5 g; 8 pmol) was assayed alone or was mixed with His-OpcA (4.4 g; 80 pmol). The NADP concentration was held constant at 2 mM. Kinetic parameters were determined by fitting the data to the Hill equation and are V max ϭ 80.4 mol⅐min Ϫ1 ⅐mg His-G6PD, S 0.5 ϭ 58.9 mM, and n ϭ 2.4 for His-G6PD in the absence of His-OpcA and V max ϭ 77.7 mol⅐min Ϫ1 ⅐mg His-G6PD, S 0.5 ϭ 1.8 mM, and n ϭ 0.9 for His-G6PD in the presence of His-OpcA. Inset shows detail of the region between 0 and 10 mM G6P.

TABLE IV K m (app) values for N. punctiforme G6PD and for strain UCD 351 G6PD and His-G6PD assayed in the absence and presence of His-OpcA
All assay mixtures contained 70 g of cell extract or 0.5 or 1.0 g of His-G6PD. Assay mixtures used for the determination of K m (app) for G6P contained 2 mM NADP and 0.1-500 mM G6P, whereas those used for the determination of K m (app) for NADP contained 500 mM G6P and 5-2000 M NADP. When required, His-OpcA was added at 10-or 20-fold molar excess relative to the amount of G6PD monomer present. Values for S 0.5 and the Hill coefficient, n, a measure of the cooperativity of substrate binding, were determined by fitting the data to the Hill equation. For hyperbolic curves n is 1, indicating non-cooperativity of substrate binding. S 0.5 and n are reported as mean Ϯ S.E. with the number of trials given in parentheses.
[NADP]  (8). There was no indication that OpcA comigrated with native G6PD (Fig. 1), and the two proteins were not coimmunoprecipitated from N. punctiforme extracts; therefore, it is unlikely that the G6PD oligomer contains OpcA or that any stable protein complex containing both OpcA and G6PD exists in N. punctiforme. From sequence analysis, the G6PD monomers from both N. punctiforme and Anabaena 7120 are predicted to be ϳ58 kDa, whereas the His-G6PD monomer is 62 kDa. We conclude from these data that the N. punctiforme, and most likely the Anabaena 7120 and A. variabilis, native enzymes are homotetramers with slightly different migratory properties. As discussed below, we speculate that the higher aggregation states that have been observed in vitro do not reflect the native state of the protein in vivo. Western blots showed that G6PD was assembled into its characteristic tetrameric 200-kDa form in every N. punctiforme strain that contained an intact zwf gene (Table II). N. punctiforme G6PD produced as a His-tagged protein in E. coli was found in the same aggregation state (Fig. 3, A and B), although E. coli appears by BLAST analysis to lack any protein with similarity to OpcA (12). There was also no change in the aggregation state of purified His-G6PD after activation by His-OpcA. Collectively, these results strongly imply that OpcA is not required for assembly of the N. punctiforme G6PD homotetramer. On native PAGE gels stained for G6PD activity, an increase in the intensity of the highest molecular mass G6PD band was observed upon mixing of extract from wild-type Synechococcus 7942 with extract from a Synechococcus 7942 strain lacking OpcA (see Fig. 5 in Ref. 9). In light of the results presented here, this increase in staining intensity is most likely to have been due to the activation of previously assembled, catalytically inactive tetrameric G6PD by OpcA rather than to OpcA-mediated assembly of inactive G6PD monomers into catalytically active oligomers, as was proposed (9).
OpcA was clearly required for the optimal catalytic activity of N. punctiforme G6PD, and its effect was to increase the affinity of G6PD for the G6P substrate ϳ30-fold to a physiologically relevant value (Table IV). OpcA also affected NADP cofactor binding, because a 2-fold increase in the affinity for NADP occurred for strain UCD 351 G6PD or His-G6PD in the presence of saturating His-OpcA. G6PD activity assays of extracts from strains that overproduced G6PD (Table II, strains UCD 351, UCD 364, and UCD 348) showed that, when G6PD was present in excess, it was the amount of OpcA that determined G6PD activity. As illustrated in Fig. 4 and Table III, the G6PD in strain UCD 351 was activated by the addition of OpcA from strain UCD 466 extract. Both strain UCD 466 extract and purified His-OpcA were able to activate strain UCD 351 G6PD and His-G6PD (Table III), indicating that it was OpcA that was responsible for the activation of G6PD and not some other component in N. punctiforme extract whose activity depended on the presence of OpcA. For strain UCD 351 G6PD, the OpcA concentration required for half-maximal activation of G6PD was much lower than that required for half-maximal activation of His-G6PD (Table III). The His-G6PD tetramer may have interacted less effectively with OpcA than the native G6PD tetramer if the His tag made it less accessible to OpcA.
The affinity of cyanobacterial G6PD for G6P in vitro is reported to vary with the pH, with the concentration of the enzyme in solution, and with its redox state and aggregation state (4,5,8,23). We were able to alter the affinity of purified His-G6PD for substrate only by adding OpcA, but the shift in apparent K m that occurred after OpcA was added ( Fig. 5B and Table IV) resembles the shift that others have observed when dilute purified cyanobacterial G6PD was concentrated in solution (4,8). Most recently, for example, Gleason (8) reported that G6PD purified from Anabaena 7120 cultured with nitrate was in a "high activity" form when concentrated in solution (Ն2 mg/ml) and in a "low activity" form when it was diluted. Both forms of the enzyme exhibited sigmoidal kinetics, but the apparent K m (S 0.5 ) for G6P at pH 7 to 8 was 2.2-7.2 mM for the concentrated enzyme and 49 -54 mM for the dilute enzyme. OpcA was most likely absent from this purified Anabaena 7120 G6PD preparation: A typical specific activity of 2-6 mol⅐min Ϫ1 ⅐mg of protein Ϫ1 was reported for the Anabaena 7120 enzyme, when it was assayed at 20 mM G6P (8). This value is slightly higher than the average specific activity of 0.9 Ϯ 0.1 mol⅐min Ϫ1 ⅐mg of protein Ϫ1 we typically obtained for purified His-G6PD preparations at 5 mM G6P, and is comparable to the specific activity of 8.0 mol⅐min Ϫ1 ⅐mg of protein Ϫ1 we observed for His-G6PD at 20 mM G6P (Fig. 5B). It is more than an order of magnitude lower than the highest specific activity (86 mol⅐min Ϫ1 ⅐mg His-G6PD protein Ϫ1 at 5 mM G6P) we obtained for the OpcA-activated enzyme (Table III). From Western blots, microscopic measurements, and protein and Chl a assays we have determined that the average wild-type N. punctiforme cell is 2.5 m x 4 m and contains ϳ1 mg of G6PD monomer and 20 mg of Chl a per g of total protein when cultured with ammonium. Assuming the cell is a cylinder, we calculate a cell volume of 2 ϫ 10 Ϫ11 ml, and if a cell contains 1.75 ϫ 10 Ϫ13 g of Chl a (14), we estimate that the cellular G6PD concentration is about 0.4 mg/ml. If G6PD is present at a similar concentration in Anabaena 7120 cultured with nitrate, at 2 mg/ml or greater, purified "high activity" G6PD could be as much as 5 times more concentrated in vitro than it is in the cell. Purified cyanobacterial G6PD has the unusual in vitro properties of inactivation and precipitation/activation in dilute and concentrated suspension, respectively (see "Experimental Procedures" and ref. 8). We speculate that nonphysiologically high concentrations of the enzyme, or of effectors such as G6P or NADP, and physicochemical factors such as low pH, might stabilize purified G6PD in the absence of OpcA, keeping it in an activated or partially activated conformation that resembles the conformation that is characteristically found in the cell when OpcA is present.
To prevent simultaneous operation of the oxidative and reductive pentose phosphate cycles, cyanobacterial and chloroplast G6PDs are active in the dark when they are in an oxidized state and are inactive in the light when they are in a reduced state, whereas the converse is true of key enzymes of the reductive pentose phosphate cycle (6,24). The inactivation of G6PD is mediated by thioredoxin (5,7,8,25) and occurs via covalent redox modification of specific regulatory cysteine residues (19). The location of the conserved cysteine residues differs between the cyanobacterial and chloroplast enzymes (19,26), and plants appear to lack any protein with similarity to cyanobacterial OpcA (BLAST search) (12), therefore, regulation of the two enzymes may be quite different. Because the oxidized and reduced forms of G6PD have been shown to differ in their affinities for G6P (19,23,27), we initially considered the possibility that strain UCD 351 G6PD and His-G6PD were in a reduced and inactive form and that the shift in apparent K m we observed was due to oxidative reactivation of the enzyme upon the addition of OpcA. Two lines of evidence now make this hypothesis unlikely: First, we were unable to oxidatively activate His-G6PD either with oxidized thioredoxin or with chemical oxidants such as STT, diamide, oxidized glutathione, and hydrogen peroxide, which have previously been reported to activate G6PD (7,27); thus, oxidation by OpcA also seems unlikely. Second, the experiments with ␣-OpcA antibody indicated that repeated interactions with His-OpcA were necessary to keep His-G6PD in an activated state. If activation required only a single interaction between the two proteins, as would be expected if OpcA activated G6PD by oxidation, then addition of ␣-OpcA after activation had occurred should have had no effect on G6PD activity. The decline in G6PD activity we observed after the addition of ␣-OpcA to the His-G6PD plus His-OpcA mixture implies that further interaction between the two proteins was required to keep the G6PD in an activated state and that binding of ␣-OpcA to His-OpcA either physically prevented this interaction or rendered the His-OpcA inactive.
When His-OpcA was incubated with reduced thioredoxin before it was mixed with His-G6PD, no activation of His-G6PD occurred (Fig. 6); activation of His-G6PD by His-OpcA occurred only after excess sodium tetrathionate was added to the His-OpcA plus reduced thioredoxin mixture. There are two possible explanations for these results: The first is that the His-OpcA was unaffected by reduced thioredoxin, but once His-G6PD had been activated by His-OpcA, it immediately became sensitive to redox modulation by thioredoxin; the second is that His-OpcA itself was reversibly inactivated by thioredoxin and could not activate His-G6PD when it was in a reduced state. Because the activation of His-G6PD by His-OpcA occurred very rapidly, we could not distinguish between these two possibilities by catalytic assays. Like other cyanobacterial G6PDs, N. punctiforme G6PD contains cysteine residues in conserved positions that may be involved in redox modulation of enzyme activity (26). The N. punctiforme OpcA sequence also contains cysteine residues; of the nine present, those at positions 183, 195, 396, 401, and 407 are conserved in five other cyanobacterial sequences with homology to OpcA and therefore appear to be the most likely targets for thioredoxin-mediated redox modulation of OpcA activity, if it does occur. (The N. punctiforme, Synechococcus 7942, and Synechocystis 6803 OpcA sequences are available on the National Center for Biotechnology Information web site; the Prochlorococcus marinus and marine Synechococcus sp. strain WH8102 OpcA sequences are both available at the U. S. Department of Energy Joint Genome Institute web site; and the Anabaena 7120 OpcA sequence is available at the Kazusa DNA Research Institute web site.) Detailed kinetic studies and site-directed mutagenesis of the cysteine residues in G6PD and OpcA should allow us to determine whether reduced thioredoxin inactivates one or both of these proteins.
The physiological role of OpcA and the exact nature of its interaction with G6PD remain to be determined. N. punctiforme strains lacking OpcA have low levels of G6PD activity and exhibit defects in nitrogen fixation and dark heterotrophic growth. The kinetic data presented here are most consistent with a role for OpcA as an allosteric activator of cyanobacterial G6PD that shifts the equilibrium between the T-and R-states of the enzyme toward the higher affinity R-state. We have also shown, however, that, for wild-type N. punctiforme cells, changes in growth conditions did not lead to disproportionate changes in the amounts of OpcA and G6PD protein. If OpcA is an allosteric effector of G6PD, it is unclear how it would function effectively if its concentration in the cell does not change significantly under conditions where regulation of enzyme activity is required. One possibility is that OpcA itself is modified in some manner, such that all or a portion of it is unable to interact with G6PD under conditions when low enzyme activity is desirable. The redox modulation experiments we have performed are consistent with the idea that OpcA is reductively inactivated by thioredoxin, although other forms of protein modification can also be considered.