A Deoxyribonuclease Which Requires Nucleoside Triphosphate from Micrococcus lysodeikticus EXCHANGE REACTION CATALYZED BY THE ENZY~VJE PREPARATION*

Abstract The ADP-ATP exchange reaction catalyzed by the highly purified Micrococcus lysodeikticus DNase has been investigated in detail. The exchange activity is eluted together with the DNase activity from triethylaminoethylcellulose and the ratio between both activities is relatively constant in the peak fractions. The exchange reaction between [8-14C]ADP and ATP has an absolute requirement for ATP and divalent cations. The enzyme is not dependent on the presence of a DNA substrate. The enzyme does not catalyze a measurable exchange between ATP and orthophosphate. The optimum pH of the reaction is 8.5 in glycine-NaOH buffer. Mn2+ is the most effective metal ion but can be replaced, to some extent, by Mg2+, Co2+, or Ni2+. Of the nucleoside triphosphates tested, ATP is most effective in the presence of Mn2+, while in the presence of Mg2+ instead of Mn2+, ATP and dATP are most effective at almost the same level. In contrast to the ADP saturation curve which shows typical Michaelis-Menten kinetics, the ATP saturation curve is sigmoidal, suggesting multiple binding of ATP on the enzyme. These results suggest that the enzyme and ATP react to form the enzyme-phosphate intermediate as the first step of the over-all reaction of DNA hydrolysis.

The previous papers in this series (1, 2) described the purihat,ion and some of the properties of Micrococcus lysodeikticus DNase which has a specific requirement for the presence of nucleoside triphosphate.
The enzyme catalyzes the hydrolysis of DNA to a mixture of oligonucleotides terminated by 5'-phosphoryl group, with concomitant cleavage of nucleotide to equimolar amounts of nucleoside diphosphate and orthophosphate. Recently, similar DNase activity was found in extracts of Bacillus laterosporu9 (3), Diprococcus pneumoniae (4, 5), Escher-* This work has been supported by research grants from the National Institutes of Health, United States Public Health Service (GM 12052), Jane Coffin Childs Memorial Fund for Medical Research, and theMinistry of Education of Japan. A preliminary report of some of the experiments was presented at 43rd Annual Meeting of the Japanese Biochemical Society in October 1970. 1 ANAI, M., YAMANAKA, M., MIHARA, T., AND TAKAGI, Y., Abstracts of 19th symposium on enzyme chemistry, Japan, 1968, p. 78. ichia coli (6, 7), and Mycobacterium smegmatis (8,9), and it was suggested that this novel type of DNase may play some physiologically significant role in the cell such as genetic recombination (4, 7, 10, 11).
In the course of studies on the participation of nucleotide in the hydrolysis of DNA by this enzyme, it was found that the highly purified preparation of the DNase catalyzes an exchange reaction between ADP and ATP in the absence of a DNA substrate. This communication will be concerned with the exchange reaction catalyzed by the purified preparation of M. lysodeikticus DNase.
It has been suggested that the enzyme and ATP react to form an enzyme-phosphate complex as the first step of the over-all reaction of DNA hydrolysis.

EXPERIMENTAL PROCEDURE
Xaterials~M. lysodeikticus DNase was purified and assayed as reported previously (l), and the concentrated hydroxylapatite fraction was used in all the experiments to be described unless otherwise noted.
Unlabeled nucleotides were purchased from Sigma or Calbiochem.
[T-~*P]ATP was prepared by the method of Weiss,Live,and Richardson (12). Other chemicals were obtained as described in the previous papers (1, 2).
Dilutions of enzyme were made as described previously for the DNase assay (1). After incubation for 10 min at 35", the reaction mixture was applied, together with 0.1 pmole each of carrier ATP, ADP, and AMP, to Whatman No. 3MM filter paper.
Paper electrophoresis was carried out in 50 mM citrate buffer, pH 3.3, at a potential gradient of 22 volts per cm for 2 hours at 4'. After the paper was dried in an oven, the nucleotides were located under ultraviolet light, and cut out for determination of radioactivity, which was measured in the Beckman model LS-100 liquid scintillation system.

RESULTS
Requirements for Reaction-The requirements for the exchange reaction between [%X!]ADP and nonradioactive ATP are shown in Table I. The reaction has an absolute dependence on the pres-Nucleotide-requiring DNase. III Vol. 246,No. 21  The assay conditions were those as described under "Experimental Procedure" with the exception that the enzyme concentration was varied up to 30 pg of protein as indicated. B, time course of the exchange reaction.
The assay conditions were those as described under "Experimental Procedure" with 20 pg of the enzyme, and the reaction mixtures were incubated for the periods as indicated. 85% of the maximum activity was obtained (Fig. 1). At pH 9.4 in glycine-NaOH buffer, the optimum pH for ATP-dependent DNase activity, only 607$ of maximum exchange activity was observed.
E$ect of Divalent Cations on Exchange Reaction-The most effective metal ion was Mn2f at 1 X IO+ M and at higher concpntrations considerable inhibition was observed (Table II).
Thus the cation requirement for the exchange reaction is different from that for ATP-dependent DNase reaction, in which Mg2+ is more effective than Mn2+.
EJect of Enzyme Concentration on Exchange Reaction-The exchange rate was directly proportional to the amount of enzyme added, up to 30 pg of protein, when a lo-min incubation period was used (Fig. 2A).
The enzyme (20 pg) catalyzed the exchange reaction, showing linear response up to about 15 min (Fig. 2B) 3. Effect of ATP concentrations on ADP-ATP exchange. A. the assav was carried out as described under "Exnerimental Procedure""with the exception that the ATP concentration was varied as shown. B, experimental conditions were as described in A with the exception that MnClz was replaced by 5.0 X Wa M MgCl,.
Znset shows a Lineweaver-Burk plot. 4. Effect of ADP concentrations on exchange reaction. A, the exchange assay was carried out as described under "Experimental Procedure" with the exception that the ADP concentration was varied as indicated.
B, the experimental conditions were as described in A with the exception that Mn2+ was replaced by 5.0 X 1O-3 M Mg* and incubation time was 30 min. Inset shows a Lineweaver-Burk plot.
other various nucleoside triphosphates added in place of ATP were compared for the effect on the exchange reaction (Table III). In the presence of Mn2f, ATP was most effective among nucleoside triphosphates investigated. dATP was far less effective than ATP, while other nucleoside triphosphates were more feeble. However, in the presence of Mg2+ instead of MI?+, ATP and dATP were most effective at almost the same level, and other nucleoside triphosphates showed lower activity. The pattern of the action of nucleotides obtained in the presence of Mgz+ is similar to that found for DNA-hydrolyzing reaction. Effect of ATP Concentrations on Exchange Reaction-The relationship between the concentration of ATP and the rate of the ADP-ATP exchange reaction in the presence of Mn2+ or Mgzf is shown in Fig. 3, A and B, respectively.
The enzyme was saturated with almost the same concentration of ATP in both cases, but in the presence of Mn 2+ the maximum velocity of the reaction was apparently greater by several fold than that in the presence of Mg"+.
Thus Mn2f exerts its stimulatory effect on the ADP- The Sephadex G-200 fraction was adsorbed and eluted from TEAE-cellulose.as described previously (1). Fractions were collected and assayed for the DNase and ADP-ATP exchange activities as described under "Experimental Procedure." ATP exchange by altering the I/;nax for ATP. In both reactions, the ATP saturation curve was sigmoidal, and, therefore, the Lineweaver-Burk plot gave a concave curve. Effect of ADP Concentrations on Exchange Reaction-In contrast to the ATP saturation curve, the relationship between the concentration of ADP and the velocity of ADP-ATP exchange reaction showed typical Michaelis-Menten kinetics, with the K, value of 3.3 X 10e6 M or 4.0 X 10es M in the presence of Mn2+ or Mg2f, respectively (Fig. 4, A and B) . Chromatography of Exchange and DNA Hydrolysis Activities-The ADP-ATP exchange activity and ATP-dependent DNase activity behave similarly during the routine purification procedures.
When the Sephadex G-200 fraction was subjected to the second TEAE-cellulose2 column chromatography, the first peak of ADP-ATP exchange activity eluted from the column coincided with the peak of DNase activity, and the ratio of exchange activity was relatively constant in the most active six fractions (Fig.  5), whereas the second peak of ADP-ilTP exchange activity which could be separated from the first peak was devoid of DNase activity.

DISCUSSION
A unique feature of DN.4 degradation by the action of M. lysodeilcticus DNase is the absolute participation of nucleoside triphosphate in the hydrolysis reaction. Details of the mechanism of nucleotide requirement for the reaction are presently undefined.
The DNA-dependent cleavage of ATP to ADP and inorganic orthophosphate, as reported in the preceding paper (2), shows that the nucleotide does not act as an allosteric effector. A similar type of activation of enzyme by ATP has been described for succinyl-CoA synthetase (13), y-glutamylcysteine synthetase (14), glutathione synthetase (15), ATP-citrate lyase (16, 17), or pyruvate carboxylase (18). These enzymes catalyze an exchange of phosphate between ATP and ADP, and in this type of activation ATP interacts with the enzyme, and enzyme-phosphate complex is formed as an activated intermediate (16,17).
III Vol. 246,No. 21 In this paper, evidence was presented that the highly purified preparation of M. lysodeikticus DNase can also catalyze an exchange of ADP with ATP in the absence of DNA.
The properties of the exchange were shown to be similar to those of ATPdependent DNA hydrolysis reaction catalyzed by the same preparation, but several differences between the two activities were observed.
The DNase preparation used in these experiments is a highly purified one, but not homogeneous as judged by polyacrylamide gel electrophoresis. However, no detectable radioactive ATP is formed from [W]ADP by the final enzyme preparation in the absence of ATP (Table I) Furthermore, the exchange activity was eluted together with the DNase from TEAIZ-cellulose and the ratio between both activities is relatively constant in the peak fractions.
Therefore, the data presented in this paper provide the evidence that the exchange reaction and the ilTP-requiring DNase reaction are catalyzed by a single enzyme protein.
It may imply that the exchange reaction represents a portion of over-all reaction of the DNA hydrolysis. Thus, the over-all reaction may involve the following reactions.
Enzyme + ATP F) enzyme-Pi + ADP (1) Enzyme-Pi + DNA + oligonucleotide + enzyme + Pi (2) It is possible, therefore, to rationalize that the difference in some properties of the exchange reaction and of the over-all reaction could be due to the influence of the property of Reaction 2 to that of the over-all reaction.
Several attempts have been made to isolate the predicted enzyme-phosphate complex directly by incubating the enzyme and Mn2+ or Mg2+ with [T-"P]ATP and separating the radioactive complex from free [32P]ATP by the Sephadex gel filtration, but so far all have been unsuccessful.
It appears, therefore, that the binding of ATP to the enzyme is freely reversible.
Another interesting feature of the exchange reaction is t,hat the ATP saturation curve is sigmoidal, and the Lineweaver-Burk plot gives a concave curve suggesting multiple binding of ATP on the enzyme.
These data, together with the previous finding that the ratio of the amount of ATP hydrolyzed versus the number of cleaved phosphodiester bonds during the DNA degradation by M. Zysodeikticus DNase is calculated to be approximately 3 : 1 instead of 1: 1 (2)) seem to shed light on the role of nucleotide in the reaction mechanism.
More precise kinetic studies on this line are now in progress in our laboratory, in order to settle the question.