MJ0917 in Archaeon Methanococcus jannaschii Is a Novel NADP Phosphatase/NAD Kinase*

NAD kinase phosphorylates NAD+ to form NADP+. Conversely, NADP phosphatase, which has not yet been identified, dephosphorylates NADP+ to produce NAD+. Among the NAD kinase homologs, the primary structure of MJ0917 of hyperthermophilic archaeal Methanococcus jannaschii is unique. MJ0917 possesses an NAD kinase homologous region in its C-terminal half and an inositol-1-phosphatase homologous region in its N-terminal half. In this study, MJ0917 was biochemically shown to possess both NAD kinase and phosphatase activities toward NADP+, NADPH, and fructose 1,6-bisphosphate, but not toward inositol 1-phosphate. With regard to the phosphatase activity, kinetic values indicated that NADP+ is the preferred substrate and that MJ0917 would function as a novel NADP phosphatase/NAD kinase showing conflicting dual activities, viz. synthesis and degradation of an essential NADP+. Furthermore, in vitro analysis of MJ0917 showed that, although MJ0917 could supply NADP+, it prevented excess accumulation of NADP+; thus, it has the ability to maintain a high NAD+/NADP+ ratio, whereas 5′-AMP would decrease this ratio. The evolutionary process during which MJ0917 arose is also discussed.

key enzyme involved in regulating this balance is NADP phosphatase (NADPase); is catalyzes the reverse reaction of NAD kinase, i.e. the dephosphorylation of NADP ϩ to NAD ϩ . However, little information is available regarding NADPase because the enzyme has not yet been identified. NADPase activity has been detected in rat liver mitochondria (4) and in bacteria (5), and it was also found to correlate with the circadian rhythm of Euglena (6) and with seed dormancy in Avena sativa L. (7). Although NADPase activity has been detected in an outer membrane-associated acid phosphatase found in Hemophilus influenzae, this enzyme physiologically functions in the uptake and utilization of exogenous NADP ϩ (8). In contrast to NADPase, NAD kinase has been biochemically characterized in various organisms such as Gram-positive bacteria (9,10), Gram-negative bacteria (11,12), yeast (2,13,14), plants (15), and humans (16); however, NAD kinase has not yet been reported in Archaea. Presently, Ͼ273 NAD kinase homologs are found in available data bases such as the Kyoto Encyclopedia of Genes and Genomes (available at www.genome.jp/kegg/).
Among the NAD kinase homologs found in the Kyoto Encyclopedia of Genes and Genomes, the primary structures of only two archaeal proteins (MJ0917 of Methanococcus jannaschii and MMP1489 of Methanococcus maripaludis) are characteristic and consist of two clearly distinguishable regions, viz. a C-terminal NAD kinase region and an N-terminal inositol-1-phosphatase (Ins-1-Pase) region (see Fig. 1A). M. jannaschii and M. maripaludis are hydrogenotrophic methanogenic Archaea whose entire genomic sequences have been determined (17,18); they are thermophilic and mesophilic Archaea growing preferably at 85 and 37°C, respectively (17,19). Ins-1-Pase (EC 3.1.3.25) is known to dephosphorylate D-myo-inositol 1-phosphate, yielding myo-inositol, which is necessary for the synthesis of phosphatidylinositol (20).
The characteristic primary structures of MJ0917 and MMP1489, as well as the presence of Ins-1-Pase (MJ0109) in M. jannaschii, raised questions about the function of both MJ0917 and MMP1489. We hypothesized that MJ0917 and MMP1489 may be novel bifunctional NADPases/NAD kinases with the potential to generate intracellular NADP ϩ and to maintain a suitable balance of NAD ϩ /NADP ϩ . To confirm this hypothesis, we chose MJ0917 and analyzed its biochemical properties.
* This work was supported in part by Grant-in-aid 15780212 from the Ministry of Education, Culture, Sports, Science, and Technology of Japan and by the Program for Promotion of Basic Research Activities for Innovative Biosciences. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental TABLES 1S and 2S. 1 To whom correspondence should be addressed.

EXPERIMENTAL PROCEDURES
Assays-ATP-dependent NAD kinase activity was assayed in a 1.0-ml reaction mixture (5.0 mM NAD ϩ , 5.0 mM ATP, 20 mM MgCl 2 , and 100 mM Tris-HCl (pH 8.5)) at 85°C (12). The reaction was stopped by the addition of 0.2 ml of 200 mM EDTA. The NADP ϩ formed was determined using glucose-6-phosphate dehydrogenase as described (12). Less than 30 l of the enzyme solution was added to the reaction mixture. When NAD ϩ was replaced with NADH, NADPH was determined using glucose-6-phosphate dehydrogenase after oxidation of NADPH to NADP ϩ by glutamate dehydrogenase (EC 1.4.1.3) (12). Phosphatase activity was assayed in a 0.10-ml reaction mixture (1.0 mM phosphorylated compound, 20 mM MgCl 2 , and 100 mM Tris-HCl (pH 8.5)) at 85°C. The reaction was stopped by the addition of 0.1 ml of 10% SDS (5,26), and the released P i was determined as described previously (27). In the presence of inorganic polyphosphate (poly(P)) or ATP, P i was assayed by another method (26). p-Nitrophenylphosphatase (pNPPase) activity was also assayed by measuring p-nitrophenol liberated at A 405 (⑀ ϭ 18.8 mmol Ϫ1 cm Ϫ1 ) in a 0.5-ml reaction mixture of the phosphatase assay (27). Non-enzymatically released P i or p-nitrophenol was subtracted from the total P i or p-nitrophenol released. Less than 10 l (0.1-ml reaction) or 50 l (0.5-ml reaction) of the enzyme solution (after 10-fold dilution with 10 mM Tris-HCl (pH 8.5) at 4°C) was routinely added to the reaction mixture. In all assays for NAD kinase and phosphatase, reactions were initiated by the addition of the enzyme to reaction mixtures preincubated at 85°C for 3 min. During purification, NAD kinase and pNPPase activities were assayed by measuring the amount of p-nitrophenol liberated in the presence of 5.0 and 10 mM MgCl 2 , respectively. Assays were linear with respect to time (at least 30 s and 1 min) and enzyme concentration (three points). One unit of enzyme activity is defined as 1.0 mol of product formed in 1 min at 85°C, and specific activity is expressed in units/mg of protein. Protein concentrations were determined using Bradford reagent (Sigma) with bovine serum albumin as a standard. The amount of NAD ϩ was enzymatically determined using alcohol dehydrogenase and semicarbazide (28). TLC was performed using a solvent system of 5:3 (v/v) isobutyrate and 500 mM NH 4 OH (5). V max and K m values were calculated from a Hanes-Woolf plot. A Hill plot was used to determine the Hill coefficient (n H ) and the apparent substrate concentration (S 0.5 ) that gives one-half of V max . These kinetic values are presented as the means Ϯ S.D. for more than three determinations.
Cloning and Expression-Genes encoding MJ0917 (full-length, residues 1-574), MJ0917-N (N-terminal half, residues 1-297), and MJ0917-C (C-terminal half, residues 298 -574) were amplified by PCR using genomic DNA from M. jannaschii (ATCC 43067D, American Type Culture Collection) as a template and inserted into the NdeI/XhoI sites of pET-21b (Novagen, Darmstadt, Germany). Ile-298 (an initial residue of MJ0917-C) was changed to Met in the cloned gene encoding MJ0917-C. The integrity of the cloned gene was confirmed by DNA sequencing. The resultant plasmids and only pET-21b were introduced into Escherichia coli Origami B(DE3) (Novagen), yielding MK1206 for the expression of MJ0917, MK1205 for the expression of MJ0917-N, MK1204 for the expression of MJ0917-C, and MK257 for the control. These strains were cultivated in LB medium according to the pET system manual (Novagen) at 18°C for 108 h after the addition of isopropyl 1-thio-␤-D-galactopyranoside (final concentration of 25 M). Soluble MJ0917 and MJ0917-C were expressed under these conditions. Compared with the use of BL21(DE3) (Novagen) (data not shown), the use of Origami B(DE3) increased the amount of soluble MJ0917 and MJ0917-C. In Origami B(DE3) and BL21(DE3), MJ0917-N was expressed as an insoluble protein (data not shown).
Purification-E. coli cells were suspended in TE buffer (10 mM Tris-HCl (pH 8.5) and 1.0 mM EDTA) at 4°C and disrupted by sonication using an Insonator 201M sonicator (Kubota Corp., Tokyo, Japan). The clear supernatant that was obtained following centrifugation of the sonicated cell suspension at 20,000 ϫ g for 10 min was used as the cell extract. For purification of MJ0917, the activities of pNPPase and ATPdependent NAD kinase were measured. The cell extract containing MJ0917 was heated at 85°C for 2.5 min. After removing aggregated proteins by centrifugation at 20,000 ϫ g for 10 min, the supernatant was applied to a Toyopearl AF-Blue HC-650 M column (1.0 ϫ 3.2 cm; Tosoh Corp., Tokyo) equilibrated with TE buffer containing 10 mM MgCl 2 at 4°C and eluted with a linear gradient of NaCl (0 -3000 mM) in the same buffer (40 ml). The active fractions were combined and used as purified MJ0917. MJ0917-C was partially purified by heating the cell extract containing MJ0917-C at 85°C for 2.5 min and then removing the aggregated proteins as described above.
Other Analytical Methods-SDS-PAGE was conducted with a 12.5% gel as described (29). Proteins in the gel were visualized with Coomassie Brilliant Blue R-250. The molecular mass of the enzyme was calculated by gel filtration chromatography on a Superdex 200 prep grade column (1.6 ϫ 60 cm; Amersham Biosciences) with an Ä KTA purifier (Amersham Biosciences) as recommended by the manufacturer using 5.0 mM potassium phosphate (pH 7.0) containing 150 mM NaCl as the elution buffer and the standards in a high molecular weight calibration kit (Amersham Biosciences). To determine the N-terminal amino acid sequence, the purified enzyme was directly analyzed with a Procise 492 protein sequencing system (Applied Biosystems). The DNA sequence was determined using an automated DNA sequencer (Model 377, Applied Biosystems). BLAST homology analysis (30) was conducted on the GenomeNet web site (available at blast.genome.jp/) against the Kyoto Encyclopedia of Genes and Genomes GENES plus DGENES Databases. Cluster analysis was performed on the Kyoto Encyclopedia of Genes and Genomes web site (available at www.genome.jp/kegg/).

RESULTS
Expression and Purification-MJ0917 (full-length, residues 1-574), MJ0917-N (N-terminal half, residues 1-297), and MJ0917-C (C-terminal half, residues 298 -574) were expressed in E. coli as recombinant proteins. SDS-PAGE of the E. coli cell extracts showed that MJ0917 (Fig.  1B, lane 3) and MJ0917-C (data not shown) were expressed as soluble proteins, but that MJ0917-N was expressed only as an insoluble protein (data not shown). The cell extract containing only the vector (pET-21b) did not show pNPPase or ATP-dependent NAD kinase activity, probably due to the high reaction temperature employed (85°C). The cell extract containing soluble MJ0917 (full-length) showed both pNPPase and ATP-dependent NAD kinase activities (TABLE ONE). In contrast, the cell extract containing soluble MJ0917-C showed ATP-dependent NAD kinase activity, but no pNPPase activity (data not shown). Thus, MJ0917 was considered to be a fusion protein consisting of a phosphatase and an NAD kinase located in its N-and C-terminal halves, respectively. The pNPPase activity of MJ0917-N could not be confirmed because it could not be expressed as a soluble protein.
MJ0917 (full-length) was purified by measuring the activities of pNPPase and NAD kinase (TABLE ONE). Purified MJ0917 had a molecular mass of 64 kDa, which is in accordance with the calculated value (64,118 Da) (Fig. 1B). The N-terminal amino acid sequence of purified MJ0917 was determined as MVIMEGFK, which is the same as the N-terminal sequence deduced from the nucleotide sequence of the MJ0917 gene. Upon gel filtration chromatography, purified MJ0917 eluted as a single peak with an approximate molecular mass of 290 kDa, suggesting that MJ0917 is a homotetramer.
MJ0917 pNPPase and ATP-dependent NAD kinase activities were optimum under alkaline conditions ( Fig. 2A) and were completely dependent on Mg 2ϩ . pNPPase and NAD kinase showed maximum activities at 20 and 50 mM Mg 2ϩ , respectively (Fig. 2B). MJ0917 pNPPase and ATP-dependent NAD kinase showed the highest activity at 100°C and were inactive below 60°C (Fig. 2C). M. jannaschii preferably grows at 85°C (17). Therefore, the phosphatase and NAD kinase activities of MJ0917 were hereafter determined at 85°C in the presence of 20 mM Mg 2ϩ at pH 8.5 unless stated otherwise.
Substrate Specificities of MJ0917-The substrate specificity (phosphoryl donor and acceptor) of MJ0917 NAD kinase was determined (TABLE TWO). As a phosphoryl donor, MJ0917 utilized nucleoside triphosphates and poly(P) as substrates, but not ADP or 5Ј-AMP, which is reminiscent of Mycobacterium tuberculosis NAD kinase (9). How- NAD kinase activity was assayed in the presence of 5.0 mM Mg 2ϩ . The pH at 85°C is indicated and was adjusted with Tris. The activity was assayed for 1 and 2 min. Each of the relative activities at pH 8.5 (pNPPase) and pH 7.7 (NAD kinase) was taken as 100%. B, pNPPase and ATP-dependent NAD kinase activities were assayed in a manner similar to that described for A at various concentrations of Mg 2ϩ . Each of the relative activities at 20 mM Mg 2ϩ (pNPPase) and 50 mM Mg 2ϩ (NAD kinase) was taken as 100%. C, pNPPase and ATP-dependent NAD kinase activities were assayed in a manner similar to that described for A at different reaction temperatures, with the exception that assay times were 30 s and 1 min at 100°C. Each of the relative activities at 100°C (pNPPase and NAD kinase) was taken as 100%.
a Phosphoryl donors (Sigma) were used at a concentration of 1.0 mM, but 0.1% (1.0 mg/ml) hexametaphosphate (Wako, Osaka, Japan) was used. The relative activity of NAD kinase toward ATP was taken as 100%. b Activity was not detected (A 340 Ͻ 0.01 for a 1-min reaction) in the presence of 0.40 units of enzyme. c Activity was assayed in the presence of 5.0 mM Mg 2ϩ because precipitates formed under standard conditions (20 mM Mg 2ϩ ). The relative activity of NAD kinase toward ATP at 5.0 mM Mg 2ϩ was taken as 100%. d NAD ϩ and NADH were used at 1.0 mM. The relative activity of NAD kinase toward NAD ϩ was taken as 100%. ever, poly(P) is not considered to be the true phosphoryl donor in M. jannaschii because orthologs of the genes for poly(P)-synthesizing enzymes (poly(P) kinase-1 and -2), which occur in M. tuberculosis, are not found in M. jannaschii (31). MJ0917 used NADH as a phosphoryl acceptor but to a lesser extent, suggesting that NADH is not the physiological substrate. The substrate specificity of the MJ0917 phosphatase was determined (TABLE THREE). MJ0917 exhibited high activity toward NADP ϩ , NADPH, and 2Ј-AMP (analog of NADP ϩ and NADPH). Enzymatic formation of NAD ϩ from NADP ϩ was confirmed by TLC (data not shown). MJ0917 also showed high phosphatase activity toward fructose 1,6-bisphosphate (Fru-1,6-P 2 ). The phosphatase activities toward NADP ϩ , NADPH, and Fru-1,6-P 2 were lower at pH 6.5 than at pH 8.5 (Ͻ10% activity) and preferred a high reaction temperature (optimum at 100°C in a 1-min reaction). They also had an absolute requirement for Mg 2ϩ (maximum activity at ϳ20 mM Mg 2ϩ ). Partially purified MJ0917-C showed NAD kinase activity, but not NADPase activity, indicating that the NADPase activity of full-length MJ0917 is due to the phosphatase activity of the N-terminal Ins-1-Pase region (MJ0917-N) and not due to the reverse reaction of the C-terminal NAD kinase region (MJ0917-C).
The observation that MJ0917 did not show Ins-1-Pase activity was unexpected. The phosphatase was inert toward the substrates of NAD kinase, i.e. NAD ϩ , NADH, ATP, and poly(P), although NADH and poly(P) are not supposed to be true substrates of NAD kinase as mentioned above. This demonstrates that the phosphatase activity of MJ0917 never interferes with the NAD kinase activity by degrading its substrates. The phosphatase in MJ0917 was also inactive toward ADP, 5Ј-AMP, and nicotinamide mononucleotide (NMN ϩ ). The inactivity of MJ0917 toward bis(p-nitrophenyl) phosphate, NAD ϩ , NADH, and poly(P) also indicated that the enzyme was inactive in cleaving a phosphodiester bond in diphosphate. Fru-1,6-Pase dephosphorylates the phosphate group at C-1 of Fru-1,6-P 2 (32). When considered collectively, it was concluded that the phosphatase in MJ0917 acts on the phosphate group at C-1 of Fru-1,6-P 2 and on the terminal phosphate group at C-2Ј of adenosine in 2Ј-AMP, NADP ϩ , and NADPH, but not on the phosphate group at C-5Ј of adenosine in 5Ј-AMP, ADP, ATP and at nicotinamide ribose in NMN ϩ .
The NAD kinase activity of MJ0917 gave hyperbolic saturation curves for NAD ϩ and ATP, whereas the phosphatase activities exhibited sigmoidal kinetics toward NADP ϩ , NADPH, and Fru-1,6-P 2 (Fig. 3). Kinetic values were determined (TABLE FOUR). NADP ϩ , NADPH, and Fru-1,6-P 2 had n H values of ϳ2.0, in good agreement with the positive cooperative sigmoidal behaviors of the phosphatase activities. The K m value for the NAD kinase activity of NAD ϩ was higher than that of ATP and the S 0.5 values for the phosphatase activities. Among the kinetic values for the phosphatase activities toward NADP ϩ , NADPH, and Fru-1,6-P 2 , the S 0.5 and k cat /S 0.5 values for NADP ϩ were the smallest and the highest, respectively. This indicates that NADP ϩ is the preferred substrate for the phosphatase in MJ0917 and that MJ0917 functions as a novel NADPase/NAD kinase exhibiting conflicting dual activities, viz. synthesis and degradation of NADP ϩ .
Inhibition of MJ0917-The inhibitory effects of several compounds on MJ0917 NADPase, NADPH phosphatase (NADPHase), and Fru-1,6-Pase activities were examined (TABLE FIVE and supplemental Table  1S) at substrate concentrations (NADP ϩ , NADPH, and Fru-1,6-P 2 ) that were similar to those of their S 0.5 values (0.2, 0.3, and 0.5 mM, respectively) (TABLE FOUR). The phosphate-containing compounds examined (NAD ϩ , NADH, ATP, ADP, 5Ј-AMP, and NMN ϩ ) were resistant to the phosphatase activity of MJ0917 (TABLE THREE). Nicotinamide, NMN ϩ , and adenosine exhibited little inhibitory effect (supplemental Table 1S). Among the other compounds, the inhibitory effects of 5Ј-AMP, NAD ϩ , and ATP increased in the order 5Ј-AMP Ͼ NAD ϩ Ͼ  Substrate specificity of MJ0917 for phosphatase activity Activity at pH 8.5 (85°C) is indicated by considering the relative activity for p-nitrophenyl phosphate (pNPP) as 100%. Less than 0.050 units of MJ0917 was used in the 30-s and 1-min reactions. Substrates were used at a concentration of 1.0 mM, but 0.050% poly(P) (hexametaphosphate) was used. The activity toward poly(P) was assayed in the presence of 5.0 mM Mg 2ϩ because precipitates formed under standard conditions (20 mM Mg 2ϩ ).

Substrate
Relative activity  NOVEMBER 25, 2005 • VOLUME 280 • NUMBER 47 ATP, whereas ADP and NADPH exhibited only a slight inhibitory effect (TABLE FIVE). The absence of inhibition by NMN ϩ and adenosine, as well as the slight inhibitory effect of ADP, indicated that the structures of NAD ϩ and 5Ј-AMP are important for inhibition. Of the phosphatase activities, Fru-1,6-Pase activity was the most sensitive to the inhibitory compounds (supplemental Table 1S). In the presence of both NAD ϩ and ATP, a slight synergistic inhibitory effect was observed on Fru-1,6-Pase activity, but not on NADPase or NADPHase activity (supplemental TABLE 1S). The effects of ADP, 5Ј-AMP, NADP ϩ , and NADPH on MJ0917 NAD kinase activity were examined (TABLE FIVE) at substrate concentrations that were similar to those of their respective K m values (3.0 mM for NAD ϩ and 0.35 mM for ATP) (TABLE FOUR). 5Ј-AMP activated the NAD kinase activity to a slight extent. ADP produced only a slight inhibitory effect, but NADPH produced no effect. Notably, even at 0.1 mM, NADP ϩ produced an inhibitory effect, which was considerably greater at 0.3 mM; however, this inhibitory effect was relieved when the concentration of ATP was increased to 3.0 mM. This inhibitory effect of NADP ϩ at a low ATP concentration (0.35 mM) and relief of this effect at a high ATP concentration (3.0 mM) were also observed with MJ0917-C, indicating that the ATP-dependent inhibitory effect of NADP ϩ is a property of the MJ0917 NAD kinase region. However, the activating effect of 5Ј-AMP was not detected with MJ0917-C, implying that the activating effect of 5Ј-AMP observed with MJ0917 probably results from inhibition of the NADPase activity rather than activation of the NAD kinase activity.
Stoichiometry-The biochemical data presented above were obtained by measuring the initial rate of production of NADP ϩ (in the case of NAD kinase) or P i (in the case of phosphatase) in reactions that mostly lasted for up to 1 min. We also followed the reaction catalyzed by MJ0917 for longer incubation times by determining the amounts of NADP ϩ , P i , and NAD ϩ produced in the reaction mixture (Fig. 4, A and  B, upper panels) or by TLC analysis (lower panels). After a 10-min incubation at 85°C, ϳ75% of the input NAD ϩ remained, whereas ATP was comparatively more stable (Fig. 4, A and B, lower panels, ϪMJ0917). After a 15-min incubation, only ϳ50% of the NAD ϩ remained (data not shown). Hence, we selected a reaction time of 10 min.

Kinetic values of MJ0917 for NAD kinase and phosphatase activities
The kinetic values are presented as the means Ϯ S.D. for four (in the case of NAD kinase) and three (in the case of phosphatase) determinations.  a The NADPase and NAD kinase activities were assayed in the presence of NADP ϩ at around its S 0.5 value (0.2 mM) and substrates at their respective K m values (3.0 mM for NAD ϩ and 0.35 mM for ATP) as well as in the presence of the indicated compounds at the indicated concentrations. Each relative activity in the absence of the compound was taken as 100%. b Activity was assayed in the presence of 3.0 mM NAD ϩ and 3.0 mM ATP. The relative activity that was assayed under this condition without NADP ϩ was taken as 100%.
When 1.0 mM NAD ϩ and 1.0 mM ATP were used as substrates, both NADP ϩ and P i were produced (Fig. 4A). During the initial period (until ϳ1 min), the amount of NADP ϩ was higher than that of P i , and the production rate of NADP ϩ was linear. However, during the additional incubation period, the amount of NADP ϩ gradually decreased, whereas that of P i increased continuously. The amount of NAD ϩ decreased during the initial period (until ϳ1 min), but then was maintained at a high level during the remaining part of the incubation. TLC analysis showed that ATP was rapidly converted into ADP, whereas NAD ϩ remained comparatively unchanged (Fig. 4A). These behaviors of NADP ϩ , P i , and NAD ϩ , as well as those of ATP and ADP, were attributed to the dual activities of MJ0917: the NAD kinase activity that converts NAD ϩ into NADP ϩ by utilizing ATP and the NADPase activity that converts NADP ϩ into NAD ϩ . The behaviors of NADP ϩ , P i , and NAD ϩ were not significantly affected when the substrate concentrations were increased (3.0 mM NAD ϩ and 3.0 mM ATP) (Fig. 4B), although the levels of both NADP ϩ and P i were higher.
The effects of 5Ј-AMP were examined in the presence of 1.0 mM NAD ϩ and 1.0 mM ATP (Fig. 4C). 5Ј-AMP suppressed the production of P i and increased the production of NADP ϩ during incubation; this demonstrated that the activating effect of 5Ј-AMP is due to the suppression of the NADPase activity (TABLE FIVE) rather than activation of the NAD kinase activity.
Cluster analysis, which detects operon-like structures, for each of the 314 and 273 orthologs of MJ0917-N and MJ0917-C classified them into 57 and 41 cluster groups, respectively. (Cluster groups were not observed in the 71 and 48 orthologs of MJ0917-N and MJ0917-C.) Of the cluster groups, only orthologs of MJ0917-N and MJ0917-C in Euryarchaeota (Methanobacterium thermoautotrophicum, Methanosarcina acetivorans, Methanosarcina mazei, A. fulgidus, and Methanopyrus kandleri) form a cluster, i.e. the phosphatase ortholog and the NAD kinase ortholog in these Euryarchaea form an operon-like structure (Fig. 5).
NADPase, NADPHase, and Fru-1,6-Pase activities exhibited positive cooperative sigmoidal behaviors (Fig. 3) that were in good agreement with the n H values (ϳ2.0) for NADP ϩ , NADPH, and Fru-1,6-P 2 (TABLE  FOUR). The crystal structures of MJ0109 and AF2372 (Ins-1-Pase/Fru-1,6-Pases, homologs of MJ0917-N) are dimers and contain one substrate-binding site/subunit (22,23); this implies that one substratebinding site exists in one molecule of the MJ0917-N portion. Based on this, it is possible that two substrate molecules bind in a positive cooperative manner to the two MJ0917-N portions in tetrameric MJ0917, enabling sigmoidal kinetics with n H values of ϳ2.0.
The fact that MJ0917 showed both NAD kinase and NADPase activities, wherein NADP ϩ was considered to be the most preferred substrate based on kinetic values (TABLE FOUR), indicates that MJ0917 would function as a novel NADPase/NAD kinase with conflicting dual activities, viz. synthesis and degradation of NADP ϩ . It should be noted that NADPase had not yet been identified. MJ0917 is the sole NAD kinase ortholog found in M. jannaschii, and NAD kinase is reported to be essential in M. tuberculosis and Bacillus subtilis (43,44). How can essential NADP ϩ be supplied by MJ0917 in M. jannaschii? The in vitro analysis of MJ0917 not only answered this question, but also suggested that only MJ0917 has the ability to prevent excess NADP ϩ accumulation, thereby maintaining a high NAD ϩ /NADP ϩ ratio. NADP ϩ is produced by MJ0917, but never accumulates continuously to excess levels, resulting in low NADP ϩ and high NAD ϩ levels (Fig. 4, A and B). NAD-Pase activity, as well as the product inhibition of NAD kinase activity at a low ATP level (TABLE FIVE), should contribute to preventing excess NADP ϩ accumulation. It should be noted that, during the initial reaction period, the production rate of NADP ϩ was higher than that of P i (the conversion rate of NADP ϩ into NAD ϩ ) (Fig. 4, A and B). The sigmoidal kinetics of the NADPase reaction (Fig. 3) and the inhibitory effects of substrates (NAD ϩ and ATP) of the NAD kinase reaction on the NADPase reaction (TABLE FIVE) should contribute to this lower production rate of P i . Furthermore, 5Ј-AMP suppressed the NADPase reaction (TABLE FIVE) and allowed MJ0917 to produce higher levels of NADP ϩ , thus decreasing the NAD ϩ /NADP ϩ ratio (Fig. 4C) and suggesting that 5Ј-AMP regulates the ratio of NAD ϩ to NADP ϩ . Despite the suppressed production rate of P i , it might be supposed that NADP ϩ should be completely depleted through a continuous conversion of NADP ϩ into NAD ϩ (Fig. 4). However, NADP ϩ would not be completely depleted if it were assumed that ATP is supplied continuously in vivo because a sufficient amount of NAD ϩ would be produced due to the NADPase reaction. Collectively, MJ0917 would maintain a high NAD ϩ /NADP ϩ ratio, wherein 5Ј-AMP would act to decrease the ratio.
The NAD ϩ /NADP ϩ ratio in living cells is regarded as being high, e.g. the ratio is 5.3 in E. coli and 8.3 in S. cerevisiae (45,46). Although the physiological NAD ϩ /NADP ϩ ratio in archaeal cells has not been reported, the MJ0917 NADPase reaction would be advantageous in maintaining a high ratio in M. jannaschii. To maintain a high ratio in other cells, a high NADPase activity and/or sufficient inhibition of NAD kinase activity should be required. However, no potent inhibitors for NAD kinase activity have yet been found among biochemically characterized NAD kinases (with the exception of E. coli NAD kinase, which is strongly inhibited by NADH and NADPH) (2, 9 -16). Hence, we propose that the NADPase reaction in other cells should also prevent the excess accumulation of NADP ϩ and should help maintain a high NAD ϩ /NADP ϩ ratio.
How did MJ0917 or MMP1489 arise during the evolutionary process? Cluster analysis provided insights into this question (Fig. 5). With the exception of the clustered orthologs of M. acetivorans, the 3Ј-regions of the clustered MJ0917-N (phosphatase) orthologs overlap with the 5Ј-regions of the MJ0917-C (NAD kinase) orthologs, thereby resulting in different reading frames. In the orthologs (41 bases overlapping) of M. mazei, an insertion of 2 bases or a deletion of 1 base from a region around the junction of the two orthologs can easily result in the fusion of the two orthologs. In the other orthologs (4 or 28 bases overlapping), an insertion of 1 base or a deletion of 2 bases can also result in the fusion of the two orthologs. Thus, we suggest that MJ0917 and MMP1489 could have been generated through an overlap of the two genes, which still occur separately in the present genomic sequences of several Euryarchaea shown in Fig. 5. In this context, we also propose that at least the clustered MJ0917-N orthologs should be NADPase genes.