Polyphosphate Kinase: Distribution, Someproperties and Its Application

Agric. Biol. Chem., 52 (6), 1471 - 1477, 1988 1471

Polyphosphate Kinase: Distribution, Some Properties and Its Application as an ATPRegeneration System Kousaku Murata,* Tomofumi Uchida, Jyoji Kato and Ichiro Chibata Research Laboratory of Applied Biochemistry, Tanabe Seiyaku Co., Ltd., 16-89 Kashima 3-chome, Yodogawa-ku, Osaka 532, Japan Received December 23, 1987

Thedistribution and someproperties ofpolyphosphate kinase, that catalyzes the formation of polyphosphate from ATP, were investigated. High enzymeactivity was found in Alcaligenes faecalis, Brevibacterium ammoniagenes, Escherichia coli, Micrococuss lysodeikticus and Pseudomonasaeruginosa. The enzymerequired Mg2+for maximumactivity and was activated by basic proteins, polyamines and phosphate polymers of low molecular weight. The enzymefrom E. coli B could catalyze the reverse reaction, and generated ATPfrom ADPand metaphosphate. The feasibility of the generation of ATPby the E. coli B enzyme was confirmed by means of the coupled reactions of polyphosphate kinase and hexokinase, which produces glucose-6-phosphate.

Wehave been studying phosphate polymer- polyphosphate kinase, that catalyzes the for- utilizing enzymes in microbial cells and have mation of highly polymerized phosphates from found that such enzymesare frequently found the terminal phosphate of adenosine-5'- in cells of Achromobacter, Brevibacterium and triphosphate (ATP) (Eq. I). This enzyme ac- Micrococcus species.1>2) Polyphosphate gluco- tivity has been found in various microbial kinase in Achromobacter butyri was used for strains such as Escherichia coli,6) Corynebac- the production of glucose-6-phosphate (G-6- terium xerosis1] and Enterobacter aerogenes.8) P) from glucose and metaphosphate.3) Although the reaction (Eq. I) is probably a Metaphosphate-dependent nicotinamide adenine dinucleotide (NAD) kinase in Brevi- ATP+ [Pi]n >ADP+ [PiL+1 (I) bacterium ammoniagenes4)was also useful for biosynthetic route, the enzyme in E. coli has the production of nicotinamide adenine dinu- been shown to catalyze the phosphorylation of cleotide phosphate (NADP)from NADand adenosine-S'-diphosphate (ADP) using phos- metaphosphate.5) The activities of poly- phate polymers6) (Eq.II). The results suggested phosphate fructokinase, polyphosphate man- that nokinase and enzymes catalyzing the phosphorylation of purine and pyrimidine nucleosides ADP+ [Pi]n >ATP+ [Pi]II_1 (II) to the corresponding nucleotides using meta- the polyphosphate kinase may be applicable to phosphate as a phosphoryl donor were also a bioreactor system as an ATPregeneration detected in cell extracts prepared from system, since substrates (phosphate polymers) Enterobacter aerogenes, Brevibacterium ammo- for the enzyme are readily available at low niagenes, Micrococcus lysodeikticus and other cost. strains (Murata et al, unpublished data). To determine the feasibility of the polyphos- Other than these phosphate polymer- phate kinase reaction as an ATPregeneration utilizing enzymes, microorganisms contain system, we investigated the distribution and

Present address: Research Institute for Food Science, Kyoto University, Uji, Kyoto 611, Japan. 1472 K. Murata et al. some properties of the enzyme in micro- fraction." The SI fraction contained 25mMacid labile organisms. phosphate, which was estimated by determining the inorganic phosphate concentration after hydrolysis of the phosphate polymers in 1.0n HC1 at 100°C for 7min. MATERIALS AND METHODS Inorganic phosphate was determined by the method of Fiske and Subbarow.10) Preparation of cell extracts. All microbial cells were grown in a medium containing 0.5% glucose, 1.0% yeast Assay for polyphosphate kinase. The activity as to the extract, 1.0% polypeptone, 0.2% meat extract, 0.5% forward reaction (Eq. I) of this enzyme was assayed in a KH2PO4 and 0.5% NaCl (pH 7.2). The cultures were reaction mixture (0.1 ml) containing 0.2fimo\ MgCl2, reciprocally shaken at 30°C for 16hr in 100ml medium in 100 fig protamine, 0.04^mol y-32P-ATP (0.4 fid), 5.0 fiinol a 2 1 Sakaguchi flask. The harvested cells were washed potassium phosphate buffer (pH 7.2) and cell extract once with a 0.85%saline solution and then resuspended in (10fig as protein). The reaction was carried out at 30°C for 5.0raM potassium phosphate buffer (pH 7.2) containing 10min and then aliquots (20/il) of the reaction mixture 0.5mM sodium deoxycholate. The cell suspension was were spotted onto 2cm squares ofToyo Filter Paper (No. ultrasonically treated at 90kHz for 5min in a Kubota 51). Chromatography was conducted at room temperature Model 200M Insonator and then the supernatant obtained by the descending method with 10% trichloroacetic acid on centrifugation at 25,000xg for 40min was dialyzed (TCA) as a solvent. After development, the paper was against 5.0mMpotassium phosphate buffer (pH 7.2) at dried and then the origins were cut out in 2cm squares for 4°C overnight. Protein was determined by the method of radioactivity determination. The activity as to the reverse Lowry et al.9) reaction (Eq. II) of this enzyme was determined in a reaction mixture (0.1 ml) containing 2.0/imol (as acid Partial purification ofpolyphosphate kinase from E. coli labile phosphate) SI fraction, 0.5/imol MgCl2, 0.2/imol B. To a cell suspension (25ml, 0.5g cells/ml) ofE. coli B, ADP, 5.0fimo\ potassium phosphate buffer (pH 7.2) and 2.0mg lysozyme was added. After incubation at 37°C for the partially purified polyphosphate kinase (10 fig as pro- 30min, 0.1mg DNase and 0.05mg RNase were added tein) from E. coli B. After incubation at 30°C for several successively to the mixture and then the incubation was hours, the reaction was terminated by immersing the test continued for a further 30min. To the mixture was then tube in boiling water for 1min, and then ATP in the added 1.0mg streptomycin sulfate, and the precipitated supernatant was determined by means of the Luciferase- material was collected by centrifugation at 25,000 x g for Luciferin assay. U) 30 min. The precipitate was dissolved in 5.0 mMpotassium phosphate buffer (pH 7.2) (25ml, 2.1 mg/ml protein) and Coupled reactions of polyphosphate kinase and hexo- then fractionated with ammonium sulfate. Solid am- kinase. To produce G-6-P from glucose, the ATP gener- moniumsulfate (7 g) was added to the streptomycin sulfate ation reaction (reverse reaction) catalyzed by polyphos- fraction, followed by incubation at 0°C for 30min. The phate kinase was coupled with the hexokinase reaction. precipitate was obtained by centrifugation at 25,000 x g The reaction mixture (0. 1 ml), consisting of2.0 /imol ADP, for 30min, dissolved in 2.5ml of 5.0mM potassium phos- 5.0/imol glucose, 2.0/imol (as acid labile phosphate) SI phate buffer (pH 7.2) and then dialyzed against the same fraction, 0.5 /imol MgCl2, 5.0 /imol potassium phosphate buffer at 4°C overnight. The dialysate thus obtained buffer (pH 7.2), 5.0 units of hexokinase and 10fig (as (3.2ml, 2.4mg protein) was used as the source of poly- protein) of the partially purified polyphosphate kinase phosphate kinase throughout this study. Through the from E. coli B, was incubated at 30°C for several hours. purification steps described above, an approximately 15- The reaction was terminated by boiling at 100°C for 1 min fold increase in the specific activity of polyphosphate and then G-6-P in the supernatant was determined en- kinase wasattained. zymatically by the method of Hohorst.12) Columnchromatography of metaphosphate. The com- Chemicals. Polyphosphate and metaphosphate were ponents in metaphosphate were fractionated by Dowex purchased from Katayama Chemicals Industry, Co., Ltd., 1 x2 (Cl~) column chromatography as described pre- Osaka, Japan. Ribonuclease, deoxiribonuclease, trimeta- viously.l) The metaphosphate solution (10%, pH 6.5) was phosphate and tetrametaphosphate were from Sigma applied to the column (5cm x 30cm) and then the column Chemical Co., St. Louis, MO. y-32P-Adenosine 5'-tri- was extensively washed with water. The absorbed phos- phosphate was purchased from New England Nuclear, phate polymers were then eluted with a linear gradient of Mass. LiCl, the concentration increasing from 0 (400ml) to 1.0 m (400ml) (pH 7.2). Fractions were collected at 7ml/tube/ RESULTS 5min at room temperature. The active fractions (Nos. 48 ~52), as a phosphoryl donor for polyphosphate kinase Assay conditions for polyphosphate kinase in E. coli B, were pooled (35ml) and designated the "SI The assay conditions for polyphosphate ki- ATPRegeneration by Polyphosphate Kinase Reaction 1473

Fig. 1. Chromatographic Separation of 32P-ATP from 32P-polyphosphate. A: A chromatogram of a 20/il aliquot of the polyphosphate kinase reaction mixture incubated at 30°C for lOmin. B: Achromatogram of the corresponding control Fig. 2. Chromatographic Analysis of the Reaction using the boiled enzyme solution. C: A chromatogram of Products. the reaction mixture incubated in the absence of Mg2+. The polyphosphate kinase reaction (forward) was carried After development, the chromatographic paper was dried out as described under Materials and Methods. A: An and then cut into 1 cm sections except for the origins which aliquot (10/il) of the reaction mixture was spotted on to werecut out in 2cm squares for radioactivity determi- Toyo Filter Paper (No. 51) and then developed with Ebel's nation. Other conditions for the reaction and chromatog- solvent at room temperature. Ebel's solvent contained raphy are given under Materials and Methods. 350ml iso-propanol, 20g TCA, 2.5ml 25% NH4OHand 150ml of water. The chromatogram (ascending) was nase activity (Eq. I: forward reaction) were developed for approximately 20 hr. To locate the positions of authentic phosphate compounds, the chromatographic invetsigated using cell extract of Arthrobacter papers were sprayed with acid molybdate solution, heated atrocyaneus ATCC13752 showing poly- at 80°C for lOmin and then exposed to ultraviolet ra- phosphate kinase activity. Figure 1 shows diation. For the determination of radioactivity, the chro- the chromatographic patterns of the reaction matographic papers were dried and cut into 0.5cm sections, and then the radioactivity in each section was products. In the complete system containing determined. B: An aliquot (40^1) of the reaction mixture ATP, Mg2+ and protamine (Fig. 1A), a large was diluted with an equal volume of2.On HC1, and then amount of radioactivity was found at the the mixture was incubated at 100°C for 7min. An aliquot origin. Onthe other hand, the radioactivity at (40jA) of the sample after treatment was spotted, chroma- the origin was low whenthe boiled extract was tographed and then analysed under the same conditions as above. The arrows in A and B indicate the positions used (Fig. IB). Omission of Mg2+ from the of (a) orthophosphate, (b) pyrophosphate and (c) tripoly- reaction mixture resulted in a decrease in the phosphate. radioactivity at the origin (Fig. 1C). In the complete system (Fig. 1A), the radioactivity at the origin increased with reaction time (up to 2B). With this treatment, the radioactivity at lOmin) and protein concentration (up to the origin disappeared and was found at the 0.3 mg/ml). sameposition as that of orthophosphate. The reaction mixture for the complete system was chromatographed with Ebel's sol- Effects of basic proteins and polyamines vent.13) Radioactivity was again found at the The effects of various proteins and poly- origin, as wasobserved in the case of develop- amines on the polyphosphate kinase reaction ment with 10% TCA. On the chromatogram, were investigated (Table I). Of the proteins intermediate products of phosphate polymers tested, basic proteins such as protamine and of low molecular weight were not detected histamine stimulated the enzyme reaction. during the reaction (Fig. 2A). The reaction Polyamines such as putrescine, spermine and mixture for the complete system was treated spermidine were also effective as to stimulation with 1.0n HC1 for 7min at 100°C and then of the polyphosphate kinase reaction. chromatographed with Ebel's solvent (Fig. 1474 K. Murata et al.

Table I. Effects of Basic Proteins and Polyamines Table II. Effects of Phosphate Polymers on the ON THE POLYPHOSPHATE KlNASE ACTIVITY POLYPHOSPHATE KlNASE ACTIVITY of A. atrocyaneus of A. atrocyaneus The forward reaction was carried out in the presence of The forward reaction was carried out in the presence of basic proteins or polyamines. The assay conditions are various phosphate polymers. The relative activity is given given under Materials and Methods. The relative in parenthesis. activity is given in parenthesis. Phosphate polymer him Activity (x 10 3cpm) Addition Concentration Activity ( x 10 3 cpm) None 2.12 (1.00) None 1.82 (1.00) Pyrophosphate 2.20 (1.04) Protamine l.O mg/ml 2.96 (1.63) 2.31 (1.09) Histamine l.Omg/ml 2.47 (1.36) 0.91 (0.54) Casein l.O mg/ml 1.77 (0.97) Tripolyphosphate 2.22 (1.05) 3.ll (1.47) Bovine serum l.Omg/ml 1.68 (0.92) albumin 2.05 (0.97) Putrescine l.OmM 2.76 (1.52) Tetrapolyphosphate 5.81 (2.74) Spermine l.OmM 2.91 (1.60) 6.25 (2.95) Spermidine l.OmM 3.10 (1.70) 3.62 (1.71) Trimetaphosphate 1.90 (0.98) 3.50 (1.65) Effects of phosphate polymers 2.83 (1.35) The effects of phosphate polymers of low Tetrametaphosphate 4.21 (1.99) molecular weight were tested as to their ability 5.12 (2.42) to act as "templates" for polyphosphate syn- 4.65 (2.19) thesis. Of the phosphate polymers tested, tetrapolyphosphate, trimetaphosphate and tetra- Table III. Distribution of Polyphosphate Kinase metaphosphate significantly increased the Activity amongBacteria polyphosphate synthesis (Table II). The forward reaction was carried out using crude cell extracts, as described under Materials and Methods. Activity is expressed as nmole of32P incorporated per mg Distribution ofpolyphosphate kinase in bacteria of protein per min. Other reaction conditions are given To select the most suitable strains for the under Materials and Methods. application of the polyphosphate kinase reaction as an ATP regeneration system, the Strain Activity distribution of the enzyme activity among Acetobacter xylinum 2.01 bacteria was investigated (Table III). The en- Achromobacter butyri 1.84 zyme activity (Eq. I: forward reaction) was Alcaligenes faecalis 13.2 Arthrobacter atrocyaneus 1.32 found in all the bacterial strains tested, being Bacillus subtHis 1.30 higher in A. faecalis, B. ammoniagenes, E. coli Brevibacterium ammoniagenes 6.10 B, M. lysodeikticus and P. aeruginosa. The 3.21 reverse reaction (Eq. II) of the enzyme was CorynebacteriumBrevibacterium flavumsepedonicum 1.30 investigated amongthe strains listed in Table Enterobacter aerogenes 2.77 III, but only E. coli B cell extract showed Escherichia coli B 7.20 Elavobacterium arborescens 0.ll detectable activity as to the formation of ATP 4.30 from ADP and the SI fraction. E. coli B was Micrococcus flavuslysodeikticus ll.4 Pseudomonas aeruginosa 10.2 selected as the source ofpolyphosphate kinase Proteus mirabilis 0.23 for ATP regeneration. Serratia marcescens 0.42 PhosphoryI donor in metaphosphate Metaphosphateis a mixture of various ring donor in metaphosphate for the polyphos- phosphate polymers with different chain phate kinase reaction, metaphosphate was lengths. To obtain the intrinsic phosphoryl fractionated by column chromatography and ATPRegeneration by Polyphosphate Kinase Reaction 1475

Fig. 3. Elution Pattern of the Phosphoryl Donor in Metaphosphate. Fig. 4. ATPRegeneration through the Polyphosphate Kinase Reaction and Its Application to G-6-P Production. A 10% metaphosphate solution (pH 6.5) was applied to a Dowex 1 x2 (Cl~) column and then eluted with a linear A: The ATPregeneration reaction was carried out in the absence ( ) and presence ( ) of the SI fraction as LiCl gradient as described under Materials and Meth- described under Materials and Methods. At prescribed ods. The phosphoryl donor activity in each fraction was determined using the partially purified polyphosphate times, an aliquot of the reaction mixture wastaken, and kinase from E. coli B. #. phosphoryl donor activity; then the concentrations of ATP (#) and acid-labile phos- O, inorganic phosphate; O, acid-labile phosphate; phate (O) were determined. B: The ATP regeneration , LiCl concentration. reaction above was coupled with the hexokinase reaction, and then G-6-P produced in the absence or presence of the SI fraction was determined. The concentration of the SI the phosphoryl donor activity in each frac- fraction was 0 (O), 5 (#) and 10 (O) mM. Other tion was determined using the partially puri- conditions for the reactions in A and B were as described fied polyphosphate kinase of E. coli B (Fig. under Materials and Methods. 3). Only one active peak was eluted, with 0.5m LiCl. The active fractions were pooled reaction. The ATPformation observed in the and used as the substrate for polyphosphate absence of the SI fraction was due to the kinase, as described under Materials and adenylate kinase activity contaminating in the Methods. This substrate fraction, designated polyphosphate kinase preparation. as the "SI fraction," contained 25niM acid- The polyphosphate kinase (partially pu- labile phosphate, and a blue color was pro- rified) reaction, as an ATPregeneration sys- duced when the SI fraction was mixed with tem, was coupled with the hexokinase reaction Fiske-Subbarow reagent, indicating that the for the production of G-6-P (Fig. 4B). These phosphate polymer in the SI fraction is a ring two kinds of reactions were coupled well, and phosphate polymer, since ring phosphate G-6-P was found to be produced using ATP polymers react with Fiske-Subbarow reagent regenerated from ADP and the SI fraction to give a blue color.1} through the polyphosphate kinase reaction. G- 6-P formation in the absence of the SI fraction A TP regeneration through the polyphosphate waspresumably due to the adenylate kinase kinase reaction reaction, as mentioned above. The partially To regenerate ATPfrom ADPand phos- purified polyphosphate kinase preparation phate polymers, the partially purified poly- showed no polyphosphate glucokinase ac- phosphate kinase from E. coli B was incu- tivity, which phosphorylates glucose to G-6-P bated with ADPin the presence or absence using metaphosphate.1 * of the SI fraction. Although the reaction was not stoichiometric, ATPwas found to be DISCUSSION synthesized in association with the degra- dation of phosphate polymers in the SI frac- The polyphosphate kinase catalyzing the tion (Fig. 4A), which was indicated by the polymerization of phosphorus was found in all decrease in acid-labile phosphate during the the bacterial strains listed in Table III, and E. 1476 K. Murata et al. coli B was selected as the source of the enzyme, ing the synthesis of the high molecular weight since the enzymein this organism could cat- phosphate polymers (Fig. 2A). alyze the reverse reaction (Eq. II: ATP- The structure and molecular size of the synthesizing activity, from ADPand phos- intrinsic phosphoryl donor in metaphosphate phate polymers). The products of the enzyme is of great interest as to the reverse reaction reaction were established to be phosphate (Eq. II) of polyphosphate kinase. Chroma- polymers from the paper-chromatographic tographic separation of metaphosphate mobility and from the generation of ortho- showed the existence of only one substrate phosphates on hydrolysis (Fig. 2A, B). (phosphoryl donor) for the reverse reac- The E. coli B enzyme reaction was stimu- tion of polyphosphate kinase (Fig. 3). This lated in the presence of basic proteins, poly- substrate, which was eluted with 0.5m LiCl, amines and phosphate polymers of low mo- seemed to be different from those for poly- lecular weight (Tables I and II). Basic proteins phosphate glucokinase and metaphosphate- and polyamines can bind to phosphate poly- dependent NADkinase. In case of polyphos- mers and may protect the phosphate poly- phate glucokinase, at least three kinds ofphos- mers from attack by the phosphate polymers phoryl donors were found in metaphosphate, from attack by polyphosphate-degrading en- and they were all very unstable.1} On the other zymes. The interaction of phosphate polymers hand, only one kind ofphosphoryl donor was with basic proteins or polyamines is of interest found for NADkinase, which was eluted with as to the physiological significance of phos- 0.1m LiCl from Dowex 1x2 (Cl~) column phate polymers. A recent study of Offen- (Murata et al., unpublished data). It is, there- bacher and Kline provided evidence that cer- fore, striking that each enzyme could use tain nuclear histone (basic) proteins (NHPs) phosphate polymers by discriminating the de- are strongly and covalently associated with gree of condensation of phosphorus, although phosphate polymers.14) Furthermore, a pre- nothing is known of the shape of the poly- vious investigation demonstrated that free phosphate molecule in solution, nor of its polyphosphate causes destabilization of chro- tendency to form aggregates. matin and enhancement of transcription in The reaction of polyphosphate kinase in E. vitro.15) Therefore, the existence of poly- coli B was coupled with the hexokinase re- phosphate-containing NHPsmay indicate the action for the production of G-6-P. The ATP roles of polyphosphates as positive modi- regeneration reaction catalyzed by polyphos- fier in gene expression. phate kinase worked well when both the par- The phosphate polymers of low molecular tially purified enzyme and the substrate- weight also stimulated the polyphosphate ki- enriched SI fraction were used (Fig. 4). nase reaction. This was presumably due to the Therefore, the polyphosphate kinase reaction "template effect" of these phosphate polymers seems to be promising as an ATPregeneration on polyphosphate synthesis. It has been sug- system for the production of useful com- gested that phosphate polymers may be syn- pounds. However, for the utilization of the thesized on a "template," which might be an polyphosphate kinase reaction for practical enzymesurface or someother macromolecule processes, partial purification of the enzyme such as RNA.14) The high energy phosphates and isolation of the intrinsic phosphoryl donor could becomeattached to the "template" sur- in metaphosphate seems to be indispensable, face and fairly large molecules synthesized since the intracellular activity of the enzyme without any intermadiates being formed. This and the phosphoryl donor content of meta- is similar to the mechanism proposed for the phosphate are extremely low. synthesis of proteins from amino acids and Acknowledgment. Wewish to thank Ms. M. Hosoda polynucleotides. This would explain the in- for her technical assistance. ability to detect any soluble intermediates dur- ATPRegeneration by Polyphosphate Kinase Reaction 1477

REFERENCES F. M. Harold, J. BacterioL, 86, 216 (1963). O. H. Lowry, N. J. Rosebrough,A. L. Farrand R. L. 1) K. Murata, J. Kato and I. Chibata, Agric. Biol. Randall, /. Biol. Chem., 193, 265 (1951). C. H. Fiske and Y. Subbarow, J. Biol. Chem., 66, 375 Chem., 42, 2221 (1980). (1925). 2) K. Murata, T. Uchida, K. Tani, J. Kato and I. De Luca and W. D. McElrog, Methods in Chibata, Agric. Biol. Chem., 44, 61 (1980). Enzymology, 57, 1 (1978). 3) K. Murata, T. Uchida, K. Tani, J. Kato and I. J. H. Hohorst, "Methods of Enzyme Analysis," 1st Chibata, Eur. J. Appl. Microbiol. Biotechnol., 7, 45 Ed., ed. by H. U. Bergmeyer, Verlag Chemie, (1979). Wienheim, p. 134. 4) K. Murata, T. Uchida,J. Katoand I. Chibata, Agric. J. P. Ebel, Bull. Soc. Chim. (France), 20, 991 (1953). Biol. Chem., 44, 1165 (1980). S. Offenbacher and E. S. Kline, Arch. Biochem. 5) K. Murata, J. Kato and I. Chibata, Biotech. Bioeng., Biophys., 231, 1 14 (1984). 21, 887 (1979). 6) S. R. Kornberg, J. Biol. Chem., 233, 159 (1958). L. J. Kleinsmith and V. G. Allfrey, Biochim. Biophys. Ada, 175, 136 (1969). 7) A. Muhammed, Biochim. Biophys. Acta, 59, 121 (1961).