Constitutive Inorganic Pyrophosphatase of Escherichia coli I. PURIFICATION AND CATALYTIC PROPERTIES*

SUMMARY An intracellular, inorganic pyrophosphatase from Escherichia co2i K-12 3000(X) has been purified 500-fold to a state of apparent homogeneity. The enzyme catalyzes hydrolysis of inorganic pyrophosphate, tripolyphosphate, and tetrapoly-phosphate with relative velocities of 1.000, 0.016, and 0.007, respectively. No activity whatsoever was found with a variety of other phosphate esters. There was an absolute requirement for divalent cation in amounts suggesting stoichiometric combination with the substrate. pH 9.1, and ; Although no exchange reaction + PPi) demonstrated, a net reversal of the reaction was achieved by coupling through phosphoglucomutase oxidation of glucose 6-phosphate 6-phosphogluconic The stoichiometry of the reverse reaction indicated of 10% of the ‘2Pi (trapped as triphosphate).

chia co2i K-12 3000(X) has been purified 500-fold to a state of apparent homogeneity.
No activity whatsoever was found with a variety of other phosphate esters. There was an absolute requirement for divalent cation in amounts suggesting stoichiometric combination with the substrate.
Although no exchange reaction (32Pi + PPi) could be demonstrated, a net reversal of the reaction was achieved by coupling through thymidine diphosphate glucose pyrophosphorylase and phosphoglucomutase to oxidation of glucose 6-phosphate to 6-phosphogluconic acid. The stoichiometry of the reverse reaction indicated utilization of 10% of the '2Pi (trapped as thymidine triphosphate).
The enzyme was exceptionally stable.
In the presence of 0.01 M Mg2+, it withstood a temperature of 80' for 10 min.
Activities which hydrolyze inorganic pyrophosphate have been demonstrated in a wide variety of natural sources, and the enzyme purified from yeast has been extensively characterized (1,2). Although the role of these activities in over-all cell metabolism is not certain, inorganic pyrophosphatases (pyrophosphate phosphohydrolase, EC 3.6.1.1) may be of considerable importance as they influence the equilibria of several vital synthetic reaction sequences. As recently reviewed by Kornberg (3), PPi is a by-product of numerous important enzymatic syntheses, including the reactions of deoxyribo-and ribonucleic acid polymerization, coensyme synthesis, and amino acid and fatty acid activation.
Although some of these reactions are themselves exergonic, the free energy change often is not great, as, for example, in amino acid activation (4). However, by coupling these reactions with hydrolysis of concurrently produced PPi (AR" (standard free energy change at neutrality) = * This work was supported by Grant 10642 from the Institute of General Medical Sciences of the United States Public Health Service.
-5 kcal), the over-all equilibrium can be markedly shifted in favor of synthesis.
Because many of these synthetic steps have been most carefully studied in Escherichia co& it was of interest to investigate the pyrophosphatase activity present in that organism.
There are two distinct pyrophosphatase activities, operative at neutral to alkaline pH values, in E. coli. One is present only in phosphate-deprived cells (or in certain mutants) and is identical with the inducible, nonspecific alkaline phosphatase (orthophosphoric monoester phosphohydrolase, EC 3.1.1.1) first described by Horiuchi, Horiuchi, and Mizuno (5, 6) and by Torriani (7). The other is a constitutive enzyme of highly restricted specificity, the purification and properties of which are reported here and in the following paper.
A preliminary report of these studies has already been presented (8).
Issue of May 10, 1966 J. Jose 1939 10 to 40 volumes of 0.02 N HCl and 0.10 M LiCl; Pd,r, 10 to 40 volumes of 0.02 N HCl and 0.20 M LiCl. Cyclic tri-and tetrametaphosphates eluted at 5 to 20 and 50 to 90 volumes, respectively, of 0.02 N HCl and 0.20 M LiCl; fortunately, the starting preparations of these materials were never contaminated with l'4.i. Eluted fractions were immediately neutralized with LiOH and appropriately pooled. Following adjustment of the pH to 9 and addition of barium acetate to 0.01 M, 2 volumes of ethanol were added. After 30 min the precipitate was collected by centrifugation at 5000 x g for 10 min and dissolved in 2 ml of ice-cold 0.001 N HCI. This was passed over a column of Dowex 50 (sodium form, 2% cross-linked, 200 to 400 mesh, 0.2 cm* x 6 cm) at 2', and the eluate plus water washings were immediately neutralized with NaOH. Over-all recoveries, based on total Pi equivalents, were greater than 907,. The purified poly-and metaphosphates were homogeneous by paper electrophoresis and cl~ron~atograpl~y (see "Methods") and were free of Pi as judged by analyses according to Lowry and Lopez (11). In each case treatment with 1 N 11a04 at 100" for 7 min converted more than 95% of the phosphate to Pi. 32P32Pi was prepared by heating Na&132P04 (Oak Ridge National Laboratory) at 400' for 1 hour and was purified by ion eschange chromatography as outlined above. Phosphorofluoridate and adenosine 5'-phosphorofluoridate were kindly clonated by Dr. D. Lipkin, Washington University. TDP-glucose and purified TDP-glucose pyrophosphorylase with specific activity of 40 units per mg (12) were gifts from Dr. L. Glaser, Washington University. Streptomycin sulfate was obtainecl from Pfizer, and DEAR-cellulose from Gallard-Schlesinger. 2-Amino-2-methyl-1,3-propanediol (Eastman) was recrystallized twice from acetone before use. All other enzymes were products of Worthington, a.nd other phosphate esters and chemica.ls were purchased from Sigma. rlssu~ oj Enzymes-Catalytic hydrolysis of PPr was measured by appearance of Pi as detected by the method of Fiske and SubbaRow (13). The incubation mixture (0.3 ml) contained 20 pmoles of 2-amino-2-methyl-l,3-propanediol chloride (pH 9.1), 0.4 pmole of M&12, 0.2 pmole of XadPPr, and sufficient enzyme (diluted in 0.01 M Tris-chloride, pH 7.5) to hydrolyze 0.01 to 0.10 pmole of substrate after 15 min at 37". The reaction was stopped by chilling to O", followed by addition of 0.7 ml of a mixture containing 0.1 ml of 5 N H$04, 0.1 ml of 2.5% ammonium molybdate (tetrahydrate), 0.1 ml of 3% NaHS03-1 7. p-methylaminophenol sulfate (Elon), and 0.4 ml of water. After 10 min at 18", optical density at 660 rnp was determined. This slight modification of the Fiske and SubbaRow procedure allowed full development of color due to Pi with undetectable acid hydrolysis of PPr and minimal (less than 3%) breakclown of more labile substrates such as P4.i. A unit of enzyme activity is that amount which will hydrolyze 10 pmoles of PPi in 15 min in the above assay. The rate of liberation of Pi in the assay was proportional to amount of enzyme up to 0.12 pmole of PPi hydrolyzed in 15 min. Reproducibility between duplicate assays was *370.
Alkaline phosphatase was assayed at 37' by the method of Garen and Levinthal (14).
Other Methods-Total phosphorus was measured by the method of Chen, Toribara, and Warner (15) after ashing with 10% Mg(NO3)3 in 95% ethanol. Acid-labile phosphorus was meas-urecl by the Fiske ancl SubbaRow procedure after treatment with 1 N H2SOa at 100" for '7 min. Protein was determined according to Lowry ei al. (16).
Paper electrophoresis was carriecl out at 25O by the general technique of Markham and Smith (17)  . When 32P-labeled compounds were present, the spray was not used, because of scintillation quenching by color, and ATP, localized with an ultraviolet lamp, was a useful reference marker. Radioactive spots, also localized with aid of a thin window monitor or Vanguard model 880 scamrer, or both, were cut into small squares and counted in a Packard Tri-Carb scintillation detector.
Starch gel electrophoresis was performed at 4' for 24 hours by the method of Smithies (20) with a gradient of 3 volts per cm. The electrode buffer was 0.05 31 soclium formate, pH 3.0, and the gel was buffered with O.Oi M Tris-chloride, pH 8.7.

RESULTS
The specific activities (units of pyrophosphatase activity per mg of soluble protein) present in sonic extracts of three strains of E. co2i grown to the stationary phase in Media I of different phosphate concentrations are given in Table I. Corresponding specific activities of alkaline phosphatase in these extracts are also shown. Horiuchi et al. (5,6) and Torriani (7) have established that an alkaline phosphatase is induced in B. co2i under conditions of Pi deprivation, and the work of Heppel, Harkness, and Hilmoe (21) has shown that this enzyme is active with PI'i as well as a variety of phosphomonoesters. Mutant El5 cannot be induced for alkaline phosphatase because of a defect in the structural gene for the enzyme (22), and mutant CW3747, which has a regulator gene defect as well as multiple copies of the structural gene, is constitutive for the enzyme. It is clear from Table I that there is at least one pyrophosphatase activity in B. coli in addition to the inducible alkaline phosphatase, and that this activity is not influenced by Pi concentration in the media. In fact, the assay for pyrophosphatase is hardly influenced by the presence or absence of alkaline phosphatase in the cell extract. Experiments with purified enzymes have shown that this is the result of (a) unfavorable conditions for alkaline phosphatase in the pyrophosphatase assay (3% of the phosphate ester-cleaving activity which is obtained in the alkaline phosphatase assay), and (6) higher (25-fold) pyrophosphatehydrolyzing capacity per mg of protein of inorganic pyrophosphatase compared to alkaline phosphatase, even when the latter is assayed under optimal conditions. It is also apparent that the two pyrophosphatases have distinctly different coding sites in the bacterial chromosome, since mutant E15, in which a large portion of the alkaline phosphatase genome is deleted, has a fully active inorganic pyrophosphatase. The data in Table II indicate that the two enzymes occur in different parts of the cell. Malamy and Horecker (23) have shown that when E. coli are converted to spheroplasts by treatment with EDTA and lysozyme, alkaline phosphatase is released from the cells and is consequently external to the cell membrane. As indicated in Table II, E. coli CW3747 cells, constitutive for alkaline phosphatase, released more than 88% of this activity into the spheroplast medium while retaining more than 827, of inorganic pyrophosphatase and more than 95% of glucose 6-ph0sphat.e dehydrogenase.
Like the last named cnzyme, inorganic pyrophosphatase is therefore an "interior" enzyme; n'eu and Heppel have obtained similar results (25).
The relatively constant specific activity of inorganic pyrophosphatase in cell extracts illustrated in Table I  when E. coli K-12 3000(X) is maximally induced for both alkaline phosphatase (by phosphate deprivation) and fl-galactosidase (by use of isopropylthio-,&n-galactopyranoside at 0.5 mM in the medium) (27).

Purijcation of Enzyme
All operations were carried out at O-5" unless otherwise indicated; centrifugations were performed at 15,000 x g for 15 min. Tris buffer refers to 0.05 M Tris-chloride, pH 7.5. The procedure is summarized in Table III.
All subsequent studies employed the most purified material (Fraction VII).
Growth of Cells-E. coli K-12 3000(X) were grown overnight at 37" with sparger aeration in carboys containing 30 liters of Medium II. Cells were harvested with a continuous flow Sharples centrifuge.
Generally the yield was 5 g per liter of  at 75% of maximum speed for four 5-min intervals with intermittent cooling to 15"; Tris buffer (140 ml) was added, and the mixture was stirred at slow speed for an additional 5 min. After the glass beads had settled, the liquid was decanted, and the beads were washed with 100 ml of Tris buffer. The combined extract and wash (240 ml) were centrifuged, and the supernatant fluid was diluted with Tris buffer to 10 mg of protein per ml (Fraction I).
Streptomycin Precipitation of Inactive Material-Streptomycin sulfate (6 liters of a 5% solution) was added with stirring to 15 liters of combined Fractions I. After an additional 30 min of stirring, the mixture was centrifuged, and the supernatant liquid was collected (Fraction II).
Heat Treatment-Fraction II, 1 liter, in a g-liter Erlenmeyer flask was mixed with 10 ml of 1 M MgClz and placed in a steam bath for 4 min, by which time the temperature of the liquid had reached 70-75".
The flask was immediately immersed and swirled in an 85" water bath for an additional 6 min, the temperature of the contents by then having reached 82-83". After Fraction VII (30 ~1 of a 5 mg per ml solution) was run in each of the two right-hand slots, and mother liquor from the crystallization step (40 ~1 of a 3 mg per ml solution) in each of the two left-hand slots, After electrophoresis at 4" for 24 hours as described in "Methods," one half-thickness of gel was stained with Amido black (28). .The appropriate portions of the other half-thickness were cut out. and ground to a fine suspension in 20-ml portions of 0.01 M Tris-chloride, pH 7.5; after freezing and thawing of the suspension and centrifugation at 10,000 X g for 20 min, the supernatant liquid was assayed for enzymatic activity. the suspension was cooled to 5" and centrifuged, the supernatant fluid was retained (Fraction III).
Ammonium Sulfate Fractionation-Solid ammonium sulfate (5.8 kg) was added with stirring over a 15-min period to 15 liters of combined Fractions III.
The suspension was stirred for an additional 15 min and then centrifuged. The supernatant fluid (17.2 liters) was treated with 3.6 kg of ammonium sulfate, which was added over 15 min with stirring.
After an additional hour of stirring followed by centrifugation, the precipitate was collected and dissolved in 250 ml of Tris buffer (Fraction IV).
The supernatant liquid was then passed through a column (13 cm2 X 20 cm) of DEAE-cellulose which had been equilibrated with the same buffer, and was washed into the resin with an additional 30 ml of buffer. A linear gradient of elution from 0 to 0.5 M KC1 was applied to the column, both limiting solutions containing 0.02 M potassium phosphate buffer, pH 6.5. The total gradient volume was 3.2 liters, and the flow rate was 80 ml per hour. Fractions of 40 ml were collected.
The enzyme was eluted as a skewed peak in the middle third of the gradient.
Active fractions were pooled (Fraction V) and dialyzed overnight against 20 volumes of 0.02 M potassium phosphate buffer, pH 8.
Second DEAE-cellulose Chromatography-This was conducted exactly as outlined for the first chromatography except that dialyzed Fraction V (600 ml) was applied to a column (7 cm2 X 20 cm) of resin, total gradient volume was 1.6 liters, flow rate was 40 ml per hour, and fraction volume was 20 ml. The enzyme eluted as before, but there was now constant specific activity across the peak (Fig. 1). Active fractions were pooled, concentrated to 5 ml by pervaporation in a Schleicher and Schuell collodion bag filtration apparatus, and equilibrated against 0.01 M KCl.
If insoluble material was present, this was removed by centrifugation, and the supernatant liquid was collected (Fraction VI).
Crystallization-To 5 ml of Fraction VI (protein concentration, 10 to 15 mg per ml), 0.25 ml of 1 M sodium acetate buffer, pH 5.0, and 3.2 ml of a saturated solution of ammonium sulfate (3") were added.
The solution was gently mixed by inversion and stored in an immobile state at O-2". In 24 hours, crystals began to appear, and by 4 to 5 days, the process was usually complete.
Crystals took the form of cubes of varying sizes (Fig. 2). They were harvested by centrifugation and dissolved in water or Tris buffer (Fraction VII).
The yield of activity from the crystallization step varied between 60 and 90% with different preparations; storage as a pH 5 suspension for periods in excess of 7 days caused gradual lowering of activity.
Purity of Final Fraction-Starch gel electrophoresis of a large amount of Fraction VII yielded a single, very sharp band which was enzymatically active (Fig. 3). It is clear that the crystallization step removed traces of contaminating protein since, as also shown in Fig. 3, the mother liquor has a second Amid0 blackstaining band of lower mobility. 3 Preliminary physical studies of Fraction VII indicate only a single protein species; these will be fully described in a later report.
Stability after any step and the fraction stored for at least 1 year without loss of activity.
Although storage was usually done at -15", this was only to prevent microbial growth; t,he enzyme was equally stable at room temperature.

Properties of Enzyme
Effect of pll on Rate of Reaction-The enzyme had a pH optimum of 9.1 with a relatively broad range of activity (Fig. 4). Although the ionization state of the PPi substrate is no doubt important, this alone does not entirely explain the high pH optimum.
Titration of a solution containing 0.67 mM NarPPi and 1.33 mM MgCl? (same as assay mixture but without buffer) indicated that ionization to PPi4-was more than 95% complete at pH 8.5. As noted below, the pH optimum in assay mixtures containing Zn2+ or Co2+ in place of MS"+ was lower (pH 7.5).
&$ects of Ions and Inhibitors-There leas no detectable activity in the absence of divalent cation.
Of the several salts tested, only those with Mg2f, R/In*, Zn*+, or Co* permitted enzymatic activity (Table IV).
The anion was of no consequence so long as the salt remained soluble, but, as described above and documented in Table IV, there were definite pH effects. The optimal concentration of cation was of approximately the same level as that of substrate (0.67 mM), suggesting that metal ion was necessary for stoichiometric combination with PPi anion; in the following paper this is shown to be the case for Mg* at pH 9.1. When more than one of these cations were present together, there was augmentation of activity if both were at less than 0.5 mM; however, the level of activity seldom was as great as that obtained with either single cation alone at twice its original concentration.
At concentrat,ions above 0.5 mM there was usually antagonism, with lower activity than with either ion Stoichiometry of Reaction--The data in Table V show that. alone. disappearance of PPi could be entirely accounted for by ap-Enzymatic activity in the routine assay (5 mpg of Fraction pearance of Pi. There was no detectable condensation to form VII) was inhibited by NaF (9y0 activity at 1 mM, less than P3,i or P4,i. 0.01% activity at 10 mM) and guanidine-HCl (36% activity Reversal of Reaction-Cohn (29) has convincingly demonat 1 M, 3% activity at 2 M), but not by KCN (100% activity strated exchange of Pi (both l*O-labeled l'i and "Pi) into PPi at 0.1 rnq 87% activity at 10 mnf) or p-hydroxymercuribenzoate with yeast inorganic pyrophosphatase, and Robbins and Lip-(100% activity at 0.1 mM, 67 % activity at 10 mM). mann (30), using yeast inorganic pyrophosphatase and coupling to the adenosine triphosphate sulfurylasc reaction, have shown  Activity was measured in the routine assay with the indicated concentrations of substrates and cations (as chloride or acetate salts), and 2-amino-2-methyl-1,3-propanediol chloride, pH 9.1, as buffer when Mg*+ was present or Tris-chloride, pH 7.5, for Zn2+. The respective amounts of enzyme (Fraction VII) added were: PPi, 5 mpg; Ps.i, 200 mpg; Pd,i, 800 mpg. The activity values obtained for P3.i and Pd.i, based upon appearance of Pi, were extrapolated to the levels expected for 5 mpg of enzyme.
(It is assumed that intermediates in the enzymatic hydrolyses of Pa. i and Pt. i to Pi do not accumulate in significant amounts.) The concentra-with time or with increasing amounts of enzyme, provided that adequate concentrations of substrate and cation were present. Total hydrolysis of each to Pi was easily obtained. As with PI'i (Table IV), the requirement for divalent cation was absolute, and activity with Mg2+ or Mn2f was maximal at pH 9.1, while that with Zn2+ or Co* was highest at pH 7.5. 6. Cochromatography of pyro-, tripoly-, and tetrapolyphosphatase activities on DEAE-cellulose. Chromatography was conducted ss described for the first DEAE-cellulose chromatography step under "Purification of Enzyme." Fractions were assayed for the respective activities at pH 9.1 by the usual method, with the substrates and Mg2+ concentrations listed in Table VI. One unit of activity is that amount which hydrolyzes 10 pmoles of substrate in 15 min at 37".
with Mg2+, Mn*, Zn2+, or Co* at either pH 9.1 or pH 7.5: Substrate Speci$city-A wide variety of phosphate esters were (a) ribo-or deoxyribonucleoside mono-, di-, or triphosphates; tested for activity with the enzyme, as measured by release of (b) ribo-or deoxyribopolynucleotides; (c) nucleotide coenzymes Pi. Only those listed in Table VI  T im e at SO' (min) Fro. 7. Inactivation kinetics of pyrophosphatase (0), tripolyphosphatase (o), and tetrapolyphosphatase activities (0) at 90". Aliquots (0.1 ml) of Fraction VII (0.05 mgperml of 0.01 M Tris-chloride, pH 7.5) with and without 0.01 Y MgC12 were dispensed into glass-stoppered, lo-ml test tubes and equilibrated at 70". At zero time they were plunged into a 90" water bath and gently swirled; under these conditions the temperature of the solutions rose to 90" in 15 sec. Every 2 min, a tube was removed and rapidly chilled at 0". The heated aliquots were then assayed for the three activities as described in the legend of Fig. 6. metaphosphates; and (j) phosphorofluoridates (inorganic phosphorofluoridate, adenosine 5'-phosphorofluoridate) .4 The presence of tri-and tetrapolyphosphatase activity in purified E. coli inorganic pyrophosphatase raised the question whether these were properties of the same protein or of contaminating enzymes. For example, yeast inorganic pyrophosphatase does not act upon Pa,i (2, 33); there is a separate tripolyphosphatase in that organism (10). Three lines of evidence indicate that hydrolysis of PPi, Ps,i and Pd,i by Fraction VII are all properties of one enzyme. 1. Copurification of the three activities: The data in Table   4 The various compounds (at 0.67 mM and 3.33 mrd) were tested in the routine assay with 2.67 mM Mgz+, Mnzf, Zn*, and Coz+, separately.
(Several of the compounds of Group a as obtained commercially were grossly contaminated with Pi and PPi; these were first purified by adsorption to charcoal or by ion exchange chromatography (31) .) Compounds of Groups 6 and c, containing phosphodiester linkages, were additionally tested with human semen phosphomonosterase (31,32) after incubation with E. coli inorganic pyrophosphatase.
No Pi was liberated after this dual incubation, indicating that there is no phosphodiesterase activity in the pyrophosphatase.
VII show that the three activities maintain roughly similar relationships to one another during the course of purification.
The slightly higher relative rates of Pa,i and Pd,i hydrolysis in Fractions I and II leave open the possibility that very small amounts of one or more additional enzymes which act upon these substrates may be present in E. coli. 2. Cochromatography of the three activities on DEAE-cellulose: The skewed elution profile of Fraction V (first DEAEcellulose) in terms of pyrophosphatase activity is given in Fig. 6. The tri-and tetrapolyphosphatase activities of the fractions are also plotted. There is virtual coincidence.
3. Heat inactivation kinetics of the three activities: Like many inorganic pyrophosphatases, the E. coli enzyme is relatively resistant to heat treatment.
For example, Fraction VII (0.05 mg per ml of 0.01 M Tris-chloride, pH 7.5) withstood a temperature of 70" for 10 min with 95 to 100% retention of activity.
At temperatures above 70", Mg* at 0.01 M stabilized the enzyme. In the presence of Mg* there was 95 to 100% retention of activity after 80" for 10 min, whereas without it only 157, of the activity remained. (Protection was not provided by ZG+.) At 90", there was decay of activity even with Mg*, but, as shown in Fig. 7, the inactivation was much slower than without the cation. Both curves followed typical first order kinetics. The inactivation in all cases was irreversible; no recovery of activity was detected after incubation at 5' or 37" for 2 to 14 days. The tri-and tetrapolyphosphatase activities of heated enzyme are also plotted in Fig. 7; the inactivation kinetics of the three activities are indistinguishable.

DISCUSSION
In view of the widespread occurrence of inorganic pyrophosphatases in nature, it was not surprising to find that extracts of E. coli were potent in this activity. The enzyme described in this report appears to account for the vast majority of extract activity, as judged by the relatively high recoveries obtained throughout purification, although small amounts of other pyrophosphate-hydrolyzing enzymes have not been excluded, and the situation at acid pH values has not been extensively investigated. Of the total soluble protein of E. co& this enzyme comprises 0.2% (estimated from 500-fold purification and assuming near purity of Fraction VII), and, as described earlier, this percentage is invariant under a variety of conditions. The synthesis of this protein therefore is not sensitive to any of the known forms of regulatory control, and the enzyme may properly be termed constitutive.
The substrate specificity of the enzyme is of some interest and distinguishes it from inorganic pyrophosphatases previously described. Although velocity diminishes progressively as chain length increases, the E. coli enzyme catalyzes hydrolysis of P3.i and P4,i; presumably, with greater amounts of enzyme, hydrolysis of higher polyphosphates might also be detected. No other tri-or tetrapolyphosphatase activity is apparent in E. coli, except when alkaline phosphatase is induced. By way of contrast, there is a quite different situation in yeast, where the inorganic pyrophosphatase shows no activity with Pa,i, and there is a separate and distinct tripolyphosphatase (2,10,33). In other respects the E. coli enzyme shows stricter specificity than other pyrophosphatases. For example, there is no cleavage of nucleoside di-and triphosphates in any of the pH or ionic environments tested, whereas yeast inorganic pyrophosphatase readily hydrolyzes these substrates in the presence of Zn2+ or Co2+ (34). A more extreme contrast is apparent on comparison with the interesting pyrophosphatase of rat liver microsomes, which has several different activities, synthetic as well as hydrolytic (35,36), or with the inducible alkaline phosphatase of E. coli, which has virtually no specificity among pyrophosphates or phosphomonoesters (21). ;1s discussed in the introductory section, the physiological role of the enzyme is conjectural.
On general grounds, one would expect that an enzyme that constitutes 0.2% of the total cell protein and that remains at a constant level under a variety of growth conditions would play some important role in metabolism. We are in the process of isolat,ing mutant cells with altered levels of enzyme in an effort to investigate this question.
We hope that E. coli inorganic pyrophosphatase will be a useful tool for studies of enzyme mechanism and protein structure. It is easily purified by routine methods from a standard bacterial source, and the final fraction is very nearly pure, if not in fact homogeneous.
The enzyme is extremely stable, can be quickly assayed, and shows rigid specificity in catalyzing a very simple chemical reaction.
Comparative studies of altered enzymes from mutant cells should be of additional interest.