Eur. J. Biochem. 98, 417-423 (1979)
Properties of Bacillus subtilis ATP-Dependent Deoxyribonuclease Michael F. SHEMYAKIN, Alexander A. GREPACHEVSKY, and Arsenii V. CHESTUKHIN Institute of Bioorganic Chemistry, U.S.S.R. Academy of Sciences, Moscow (Receivcd March 20, 1979)
A purification procedure described previously resulting in electrophoretically pure Bacillus subtilis ATP-dependent DNAase has now been modified by adding a fractionation stage with Polymin P to permit large-scale isolation of the enzyme. It has been found that the enzyme molecule ( M , = 300000) consists of two large subunits with M , 155000 and 140000. The purified enzyme has three activities: (1) DNAase on linear single-stranded and double-stranded DNAs, (2) DNAunwinding and (3) ATPase. Circular duplex DNAs were not affected by the enzyme. Study of the dependence of these activities on temperature, pH, and ATP and Mg2+ concentrations has revealed two different states of the enzyme. At low ATP concentrations and alkaline pH, it showed chiefly nuclease action, degrading considerable amounts of DNA to small fragments five residues long on average. At higher ATP concentrations and neutral pH (more physiological conditions) it predominantly unwound DNA. Simultaneously it cut preferentially one of the duplex strands to fragments more than 1000 residues in length. The results obtained suggest that the energy of the enzymecleaved ATP is mainly expended on unwinding rather than on degrading DNA molecules.
ATP-dependent DNAase has been found in many bacterial species. For some of them (Escherichia coli [l 1, Diplococcus pneumoniae [2], Bacillus suhtilis [3]) proof was obtained that the enzyme participates in genetic recombination. Early work on the properties of ATP-dependent DNAase established that under conditions optimal for DNAase action, namely at high pH and low (0.05 mM or less) ATP concentration, it vigorously cleaves linear duplex DNA to acid-soluble oligonucleotides five residues long on average. At the same time the E. coli enzyme does not digest circular DNA duplexes unless these contain gaps, but it nicks singlestranded circles [4]. These properties are difficult to reconcile with current concepts of the mechanism of genetic recombination which apparently involves rather large DNA fragments only. Therefore attempts were made recently to find conditions under which ATP-dependent DNAase does not cause considerable DNA degradation. It was shown that on brief incubation at neutral pH, lowered temperature and high ATP concentration [5,6] or low Mg2+ concentration [7], DNA degradation by E. coli or Haemophilus inpuenzae ATP-dependent DNAase decreases manifoldly. The resulting large single-stranded fragments _ _ -~ Enzymes. ATP-dependent deoxyribonuclease or exonuclease V
(EC 3.1.4.33); alkaline phosphatase (EC 3.1.3.1); lysozyme (EC 3.2.1.17); S1 nuclease (EC 3.1.4.21).
and duplexes with long single-stranded tails could be intermediates of recombination. Earlier we found ATP-dependent DNAase in B. subtilis cells, investigated some properties of the partially purified enzyme [3] and obtained data for its involvement in the recombination process [8]. Recently we were able to overcome the difficulties associated with the lability of B. subtilis ATP-dependent DNAase and to develop a novel procedure for its purification to electrophoretic homogeneity [9]. This has led us to a further study of its structure and properties.
MATERIALS AND METHODS
Materials DEAE-Sephadex A-25 and A-50 and Sephadex G-150 were obtained from Pharmacia Fine Chemicals; Tris from Merck; ATP, dithiothreitol, bovine serum albumin (fraction V) and S1 nuclease from Calbiochem; sodium dodecylsulfate and EDTA from Serva; 2-mercaptoethanol, acrylamide, N,N'-methylene-bisacrylamide, agarose, egg-white lysozyme and imidazole from Bio-Rad; calf-thymus DNA from P-L Biochemicals ; ['4C]ATP and [p3'P]ATP from Amersham ; polyethyleneimine thin-layer cellulose MN 300 from Brinkmann Instruments, Inc. Alkaline phos-
B. suhtilis ATP-Dependent DNAase
41 8
phatase (specific activity 20 units/mg protein) was isolated from E. coli [lo]. Isolation of D N A [14C]DNA of B. subtilis was isolated as described earlier [3]; unlabelled phage T7 DNA was prepared according to Richardson [l 11, 14C-labelled and j2P-labelled phage T7 DNAs according to Friedman and Smith [12]; unlabelled and 14C-labelled ColEl DNAs were isolated from the E. cofi strain K12 637 thy- [13]. The specific activity of B. subtilis [14C]DNA was 7 counts min-' pmol-', that of E. coli [32P]DNA 20 counts min-' pmol-', that of T7 [14C]DNA 1.5 counts min-' pmol-' and that of ColEl ['"CIDNA 9 counts min-' pmol-'. DNA concentrations are expressed as mononucleotide molar concentrations. Determination of Enzyme Activities DNAase activity was determined in a standard assay mixture (0.2 ml) containing 0.05 M Tris-HC1, pH 7.5-9.0, 0.01 M MgCL, 1 mM dithiothreitol, 0.5 mM EDTA, 0.2 M KCl, 6 nmol of labelled DNA, 0.05-3.0 mM ATP and about 1 unit of enzyme (fraction VI). After 20 rnin at 37 "C, samples were chilled and 0.05 ml of 1 "/, bovine serum albumin and 0.1 ml of 18 "/, trichloroacetic acid were added to each sample. The precipitates were removed by centrifugation. Radioactivity of the supernatant was determined in a Mark I1 counter after adding 10 ml of Triton X-100 scintillation fluid [14]. One unit of DNAase activity was defined as the amount of enzyme that made 1 nmol of DNA acid-soluble in 20 rnin at 37 "C. ATPase activity of the enzyme was determined in a reaction mixture (0.1 ml) containing 0.02 M TrisHCl, pH 7.5-9.0, 0.01 M MgC12, 1 mM dithiothreitol, 0.5 mM EDTA, 0.2 M KCl, 3 nmol of T7 DNA, 0.05 - 3.0 mM ATP, [14C]ATP (6 x lo3 counts/min) and 0.5 unit of enzyme. After incubation for 20 rnin at 37"C, samples were chilled and 0.01 ml of 0.5 M EDTA was added to each. From each sample a 0.01-ml aliquot was applied to a polyethyleneiminecellulose plate, with 5 pl of a mixture of ATP, ADP and AMP (10 mM each), followed by ascending chromatography in 4 M NaCl/l M CH3COOH (4/1). After drying, the ATP and ADP spots were detected in ultraviolet light, cut out and their radioactivity was counted. The amount of single-stranded DNA formed by the action of ATP-dependent DNAase was determined as the fraction of total DNA made acid-soluble by the single-strand-specific S1 nuclease. The reaction mixture for the ATP-dependent DNAase was diluted threefold into a solution containing 0.1 M sodium
acetate buffer, pH 4.5, 0.1 M NaCl, 1.3 mM ZnC11, 7 % glycerol, 0.7 % sodium dodecylsulfate and 400 nmol/ml of denatured calf thymus DNA. To each sample 0.05 unit of S1 nuclease was added followed by incubation for 45 rnin at 37 "C. The samples were chilled and 0.2 ml of calf thymus DNA solution (5 mg/ml) and 0.3 ml of 18 % trichloroacetic acid were added to each. The precipitate was removed by centrifugation, scintillation fluid was added and supernatant radioactivity was determined. Isolation and Characterization ATP-Dependent DNAase
of
B. subtilis ATP-dependent DNAase was isolated by the procedure [ 91 including the chromatography on the DEAE-Sephadex A-25 column ( 5 x 100 cm) loaded with preformed KCl gradient, ammonium sulfate fractionation, Sephadex G-150 chromatography, DEAE-Sephadex A-50 chromatography and DNA-Sepharose chromatography. The conditions of the DEAE-Sephadex chromatography do not allow one to use more than 50 g of frozen cells for a single isolation of the enzyme. To increase the scale of isolation an additional step was included, the Polymin P fractionation of crude extract, which removed 85 - 90 % of the protein and almost all the nucleic acid. Addition of this step enables one to fractionate the material from 400 g of frozen cells on the DEAESephadex A-25 column. This step is described below. 400 g of frozen B. subtilis cells were suspended in 800 ml of 0.05 M imidazole-HC1, pH 7.0, 10% glycerol, 1 mM EDTA, 5 mM 2-mercaptoethanol and 0.3 M KCl (buffer A) with added phenylmethylsulfonyl fluoride (20 pg/ml) and lysozyme (0.5 mg/ml) and incubated for 20 rnin at 37 "C. The lysate was chilled at 4"C, sonicated at 22 kHz and maximal input (UZDN-1 sonicator, USSR) without allowing it to heat above 8 "C, and centrifuged at 12000 rev./min for 1 h in a Beckman JA-14 rotor. The supernatant was diluted with buffer A to 10 mg protein/ml; 10% Polymin P (pH 7.5) was then added to a final concentration of 0.6 %,followed 10minlaterby centrifugation for 15 min at 8000 rev./min. The enzyme was eluted from the pellet with 800 ml of 0.2 M ammonium sulfate in buffer A and precipitated from the eluate with 60% ammonium sulfate. The precipitate was dissolved in 100 ml of buffer A and then used for further fractionation as previously described [9]. By increasing the isolation scale eightfold, it is possible to isolate five or six times more ATP-dependent DNAase, the inclusion of the Polymin P step giving rise to a slightly poorer yield (7 - 8 % instead of 8-10%). This method yields 0.5-0.6mg of the enzyme with a specific activity of about 4000 units/mg (zz 5000-fold purification). By 5 % polyacrylamide gel electrophoresis at pH 9 according to Davis [15], the
M. F. Shemyakin, A. A. Grepachevsky, and A. V. Chestukhin
final preparation showed only one Coomassie-stained band which contained more than 90 % of the adsorbed dye while less than 10% of the adsorbed dye was uniformly distributed along the gel. On centrifugation in a glycerol gradient, the ATPdependent DNAase sediments as a single peak with a sedimentation coefficient of 13- 14 S, which corresponds to a molecular weight of approximately 300000. Roughly the same molecular weight value is obtained by estimating the relative mobility of the enzyme in polyacrylamide gel according to Davis [15]. On electrophoresis by the procedure of Weber and Osborn [16], the enzyme protein separates into two bands of equal intensities with molecular weights of 155000 and 140000 [9], which suggests that the molecule of B. subtilis ATP-dependent DNAase consists of two large subunits of similar size. DNA was determined according to Spirin [17] and protein according to Lowry [18].
419 3
n
R
Migration
-
Fig. 1. Effect of ATP-dependent DNAase on linear undcircular C‘olEI DNAs. T o the standard assay mixture (pH 9.0) a mixture of linear (2), circular relaxed (1) and super-coiled ( 3 ) ColEl DNAs (0.5 pg of each) was added, followed by incubation at 37 ‘ C for 20 min. Samples were chilled and submitted to electrophoresis in 0.9 2, agarose gel at 0.8 V/cm fof 15 h at 22 ‘C. The gels were stained with ethidium bromide and subjected to densitometry. (A) Without ATP; (B) with ATP
2.5 7
RESULTS 2 .o
The Main Activities of B. subtilis ATP-Dependent DNAase
ATP-dependent DNAases from other bacteria are remarkable for their inability to degrade circular double-stranded DNAs, both covalently closed supercoiled ones and those relaxed by nick in one strand [4,19]. This property is shared by B. subtilis ATPdependent DNAase. As can be seen from Fig. 1, incubation with this enzyme of a ColEl DNA containing linear, circular super-coiled and relaxed molecules, leads to disappearance of linear molecules only. The enzyme degrades the double-stranded DNA much better than the single-stranded one: in the same conditions 34% of double-stranded and 5.2% of single-stranded DNA become acid-soluble. The products of duplex DNA digestion by the enzyme comprise a collection of oligonucleotides (1 - 20 residues in length, as determined by DEAEcellulose chromatography) with an average length of five residues (Fig. 2). Similarly sized degradation products have been reported for ATP-dependent DNAases from other organisms [4,20]. In common with the ATP-dependent DNAase of E. coli [21], or H. influenzae [6,7], the B. subtilis enzyme has a DNA-unwinding activity (Fig. 3 - 7) and a DNA-dependent ATPase activity producing ADP and Pi (Table 1). Thus, the B. subtilis ATP-dependent DNAase has three main activities : (a) linear duplex DNA degradation to acid-soluble products, (b) linear duplex DNA unwinding and (c) DNA-dependent degradation of ATP. The interrelationship of these activities under various conditions is described below.
6
-- 1.5 -aF
5 g c
O
2
1
4 :
.o
3 6 2
0.5 1
0
10
20
30
45
60
0
Incubation time (rnin)
Fig.2. Determination of average lengths oJ the products 41 T7 [ 3 Z P ] D N Adegradation by ATP-dependent DNAase. The standard assay mixture (pH 9.0) containing 0.05 m M ATP and 7 nmol of [32P]DNA was incubated at 37°C. Samples were taken at the indicated times, heated for 5 min at 90°C and chilled. From each sample a 0.05-ml aliquot was drawn to determine acid-soluble radioactivity. In the remaining 0.15 ml of the mixture, the amount of 32P resistant to alkaline phosphatase was measured [20]. (-0) Acid-soluble [32P]DNA;(-0) alkaline-phosphax ) oligonucleotide length Vase-sensitive 32P; ( x ~
Table 1. ATPase action of ATP-dependent DNAase 0.05 m M ATP, [14C]ATPor [ Y - ~ ~ P I A(6 T Px lo3 counts/min each) was added to the standard assay mixture; incubation was for 20 min at 37°C. The hydrolysates were assayed as described under Materials and Methods Incubation time
min 0 20
Difference
Amount of _______________~___ [I4C]ATP [14C]ADP [y-32P]ATP
32P1
nmol _________________ 4 60 4 59 0 41 4 69 0 19 0 51
0 40 4 43
- 4 08
+428
- 3 81
-
+403
420
B . subrilis ATP-Dependent DNAase
Interrelationship of the Enzyme Activities in Various Reaction Conditions
The dependence of DNAase activity on ATP concentration is not a simple one (Fig.3). The curve is bell-shaped with a maximum near 0.05mM ATP. The curve of DNA-unwinding activity versus ATP concentration is quite different. As the ATP concentration is raised this activity increases to reach a plateau at about 1 mM ATP. At this ATP concentration DNAase activity is only about one tenth of the maximal. The ATPase activity reaches a plateau at the same ATP concentration and the curve is similar to that for the unwinding, but not DNAase, activity. Fig. 4 shows that the temperature dependence of DNAase activity is different from that of unwinding activity which resembles the temperature dependence of ATPase activity, especially in the initial portion of the graph. The same is true of the graphs relating these three activities to pH (Fig. 5 ) and to Mg2+ concentration (Fig. 6). Again in both these cases the
ATPase activity curves resemble more the unwinding activity than the DNAase activity curves. This suggests that ATP energy is expended by the enzyme mainly on unwinding the DNA rather than on degrading it. Kinetics of A TP-dependent DNAase Activities under ‘Physiological’ and ‘Unphysiological’ Conditions
In the cases considered, the nuclease activity was at its maximum under overtly unphysiological conditions: alkaline pH, high temperature or low ATP concentration. It was therefore of interest to study in more detail the behaviour of the enzyme under less unphysiological conditions and to compare this behaviour with that under conditions optimal for nuclease activity. To create conditions optima1 for DNAase activity, a low ATP concentration (0.05 mM) and an alkaline pH (9.0) were used, and to mimic physiological conditions, a high ATP concentration (3 mM) and a neutral pH (7.5) were employed.
100 80
2.5
-
2.0
60
3>
1.5 1.0
a
0.01
0.1 ATP (rnM)
1
G
1.5
5 z Q
0
-
V
-
0
.o
PH
Fig.3. Eifect of A TP (,onwntrution on A TP-dependent DNAasu activities. DNAase ( x x ), ATPase (0-0) and DNAunwinding (0-0) activities were determined in standard assay ATP cleavage mixtures (pH 9.0). (A--A)
28 40 Temperature (“C)
+-
m 2
-s?
40
20
-E,
52
Fig. 5. eflect of p H on ATP-dependent DNAase activities. DNAase (x x ) , ATPdse ( 6 - 0 ) and DNA-unwinding (+o) activities were determined at the indicated pH values in standard assay mixtures (0.05 mM ATP) ~~
0
60
Fig. 4. Effect of temperature on ATP-dependent DNAase activities. DNAase ( x -- x), ATPase (0-0) and DNA-unwinding (0-0) activities were determined at the indicated temperatures in standard assay mixtures (pH 9.0, 0.05 mM ATP)
M g C b (mM)
Fig.6. Effect of MgClz c,onceniration on ATP-dependent DNAase uctivities. DNAase (x-x), ATPase (-0) and DNAunwinding (O--O) activities were determined at the indicated MgClz concentrations in standard assay mixtures (pH 9.0,0.05 mM ATP)
42 1
M. F. Shemyakin, A. A. Grepachevsky, and A. V. Chestukhin
Fig.7 shows the kinetics of the three main activities of the enzyme under the above conditions. In either type of conditions the enzyme rapidly unwinds DNA, with the curves reaching a plateau level in the first 2- 3 min of incubation. At pH 9.0 and 0.05 mM ATP (Fig.7A) it degrades about half of the DNA' whereas under physiological conditions (Fig. 7 B) DNAase activity is virtually absent. DNAase activity likewise takes a short time (5 min) to attain a plateau level. Unwinding and DNAase activities coincide during the first 1.5 min of incubation. This coincidence suggests that as it moves down the DNA molecules the enzyme preferentially degrades one strand. The rate of ATP cleavage considerably slows towards the end of incubation at the low ATP concentration and remains almost unchanged at the high ATP concentration. The amount of ATP cleaved in the latter case is 5 - 10-fold greater than in the former case. At the low concentration, ATP is degraded almost completely at 15 min of incubation. Since under the physiological conditions the enzyme produces no acid-soluble material, its DNAase activity was analysed by centrifugation of digested DNA in glycerol gradient. The results are presented in Fig.8. It can be seen that at pH 9.0 and 0.05 mM ATP the enzyme rapidly degrades a proportion of the DNA to small fragments while leaving much of it intact and producing no fragments of intermediate size. At pH 7.5 and 3 mM ATP there are also seen two peaks following the action of DNAase: one of these coincides with the intact DNA and contains no single-stranded nicks (alkaline gradient) while the other consists of rather large fragments with molecular weights of about lo6. The first peak of the neutral gradient (physiological conditions) contains about 30 % of single-stranded material and the second peak about 50 % of such material. Judging from the second peak position in the alkaline gradient, the average length of cleaved single-stranded fragments is several thousand residues. Therefore under physiological conditions the nuclease action of ATP-dependent DNAase is highly limited.
DISCUSSION In a previous paper [9] we described a new procedure for the isolation of B. subtilis ATP-dependent DNAase. Here we present a modification of that procedure which enables one to increase the scale of isolation eightfold and to obtain five or six times more of the enzyme. In both methods we were able to overcome the extreme instability of the enzyme and to obtain its stable and electrophoretically homogeneous preparations with a satisfactory yield (7- 10 %) and a rather high specific activity ( z 4000 units/mg) [9]. The actual specific activity of the B. subtilis enzyme
4
3
2
1 ; 3 3
-
Incubation time (min)
2-
0
3. s 0
U
L
20 - 60 ~
50
a
t
15
- 40 - 30 - 20
10
5
- 10
0.5 1.5
3
5 10 Incubation time (min)
15
Fig. 7. Eflect of' p H und A T P c'o~c'entrutioii 011 ATP-riiyxwdiwr DNAase activities. DNAase ( x x), ATPase (6 0 ) and DNA-unwinding (0-0) activities were determined in standard assay mixtures. (A-A) The ratio of degraded ATP to unwound DNA. (A) 0.05 mM ATP, pH 9.0 (unphysiological conditions); (B) 3 mM ATP, pH 7.5 (physiological conditions) ~
was much higher as in our assay conditions the DNA degradation reaction was far from linear (Fig.7A); in fact, judging from the initial velocity of the reaction, it was about tenfold higher and approximated to that of the E. coli ATP-dependent DNAase [4,21]. According to our data, the molecule of B. suhtilis ATP-dependent DNAase is composed of two subunits with molecular weights of 155000 and 140000. According to Doly and Anagnostopoulos [22] it consists of five subunits with molecular weights ranging from 81 000 to 42 500, yet our molecular weight estimate for the active enzyme coincides with that of the above authors (E 300000). We suggest that the large number of polypeptides found in the enzyme by Doly and Anagnostopoulos are the result of the action of proteinases during the early stages of purification since the activity of the latter in B. subtilis is known to be very high. This could explain the instability of the B. subtilis ATP-dependent DNAase isolated by these authors and the consequent low yield, specific activity and degree of purification of their preparations. High instability of the enzyme, both on isolation and on storage, compelled us to abandon completely the purification procedure which we proposed earlier [3] and
422
B. subtilis ATP-Dependent DNAase
bottom
Fraction number
top
Fig. 8. Glycerol gradient centrifugation of [14C]DNA treated with A TP-dependent DNAase ut various p H and A T P concentration values. Standard assay mixtures (0.2 ml) having (A, B) unphysiological conditions (pH 9, 0.05 mM ATP) or physiological (C, D) conditions (pH 7.5, 3 mM ATP) were incubated at 37' C. At the times indicated, samples were chilled and either (A, C) KCI and EDTA were added to final concentrations of 0.7 M and 5 m M respectively (neutral gradient) or (B, D) NaOH was added to a final concentration of 0.25 M (alkaline gradient). The mixtures were layered onto 10-35% linear glycerol gradients containing 0.05 M Tris-HC1, pH 7.6, 1 mM EDTA and 0.7 M KCI (neutral gradient) or 0.25 M NaOH (alkaline gradient). Centrifugation was done at 44000 rev./min for 3.5 h at 15°C in a Spinco L5-50 SW 50 rotor. Five-drop fractions were collected from the bottom of each tube into 24 vials and radioactivity was measured. (A) Neutral gradient, unphysiological conditions; (B) alkaline gradient, unphysiological conditions; (C) neutral gradient, physiological conditions; (D) alkaline gradient, physiological conditions. Incubation: 0 min (-a), 1.5 min (+O), 5 min ( x x) ~
which differs from that of Doly and Anagnostopoulos only in lacking the electrophoresis step. That the B. subtilis ATP-dependent DNAase is composed of' two subunits is further supported by the two-subunit composition of this DNAase in E. coli where the subunits were estimated to have molecular weights of 140000 and 128000 by Goldmark and Linn [4] and 130000 and 120000 by Eickler and Lehman [21]. We have found the properties of highly purified B. subtilis ATP-dependent DNAase to be very similar to those of E. coli [5,21] and H . influenzae [6,7] ATPdependent DNAases. The enzyme displays three main activities : ATP-dependent degradation of linear singlestranded and double-stranded DNAs to acid-soluble products averaging five residues in length, ATPdependent unwinding of the linear duplex and DNAdependent cleavage of ATP to yield ADP and Pi. Study of these activities of B. subtilis ATP-dependent DNAase in relation to the main reaction parameters has revealed some interesting relationships. Thus the curves of DNA-unwinding activity versus
ATP concentration, temperature and ambient pH are all very similar to the corresponding curves for ATPase activity but are sharply different from those for DNAase activity (see Fig. 3 - 5). Cleavage of ATP and unwinding of DNA both reach a maximum by 1 mM Mg2+ whereas DNA degradation is at its maximum only at 6-8 mM Mg2+. An appreciable depression of unwinding activity at Mg2+ concentrations above 2 mM is probably due to stabilization of the DNA double helix by divalent cation. All these results indicate that ATP energy is apparently used to unwind DNA rather than to degrade it. This assumption is also supported by the relatively slight temperature dependence of unwinding activity (see Fig. 6). Consideration of the three activities in relation to pH, temperature, and ATP and Mg2+ concentrations has disclosed two different functional states of the B. subtilis ATP-dependent DNAase. At high pH values and low ATP concentrations (i.e. under 'unphysiological' conditions), the enzyme degrades more than 40% of the DNA to acid-soluble material, and this appears to be preceded by the unwinding of more
423
M. F. Shemyakin, A. A. Grepachevsky, and A. V. Chestukhin
than 60% of the DNA under such conditions. At neutral pH and high ATP concentrations (‘physiological’ conditions) [23,24], the enzyme retains its ability to unwind DNA but is quite incapable of degrading it to acid-soluble products. It must be emphasized that in contrast to the ATPdependent DNAases of E. coli and H. influenzae [5,7], the B. subtilis enzyme fails to exhibit exonuclease action under physiological conditions even after a relatively long incubation with DNA at 37 “C. It can be calculated that early in the reaction about one molecule of ATP is used up on unwinding each DNA base pair under unphysiological conditions and about two molecules under physiological ones. We thus have a reasonably good agreement, for the molar energy yield is 29 kJ (7 kcal) in the reaction ATP +ADP Pi and 33 kJ (8 kcal) in the DNA denaturing reaction [25]. It remains, however, unclear why prolongation of the incubation, particularly at a high ATP concentration, results in a manifold increase in the molar ratio of cleaved ATP to unwound DNA. The large differences between ‘physiological’ and ‘unphysiological’ conditions as regards the amount of ATP used up by the enzyme and the kinetic curves of ATPase and unwinding activities, could be attributed to the fact that large amounts of energy are required to maintain the long DNA strands in the unwound state. It is known, however, that the ATPdependent DNAases of E. coli [26] and H. influenzae [27], while being unable to cleave DNA molecules with cross links, hydrolyze ATP no less vigorously than on incubation with intact DNA. It is therefore conceivable that much of the ATP is cleaved by the enzyme ‘idly’ also in the reaction with a normal DNA devoid of cross links. The most important function of ATP-dependent DNAase seems to be the ATP-dependent unwinding of DNA accompanied by cleavage of large fragments from one of the strands. The big enhancement of DNAase activity which seems to occur at low ATP concentration may be a consequence of an insufficient speed of enzyme movement along the DNA molecule during unwinding because of an ATP deficit. Finally it should be added that under physiological conditions, as compared to unphysiological ones, the properties of ATP-dependent DNAase can meet the
+
demands made on it as a participant in genetic recombination much better.
REFERENCES 1 . Buttin, G. & Wright, M. (1968) Cold Spring Hurhor Syinp. Quant. Biol. 33, 259 - 269. 2. Vovis, Q. F. & Buttin, G. (1970) Biochim. Biophys. Actu, 224, 42 - 54. 3. Chestukhin, A. V., Shemyakin, M. F., Kalinina, N . A. Sr Prozorov, A. A. (1972) FEBS Lett. 24, 121 - 125. 4. Goldmark, P. J. SC Linn, S . (1972) J . Biol. Chem. 247, 1849 1860. 5. Mackay, V. & Linn, S. (1974) J . Biol. Chem. 249, 4286-4295. 6. Friedman, E. A. & Smith, H. 0. (1973) Nut. New Biol. 241, 54- 58. 7. Wilcox, K . W. & Smith, H. 0. (1976) J . Biol. Chem. 251, 6121-6134. 8. Prozorov, A. A,, Kalinina, N. A,, Naumov, L. S., Chestukhin, A. V. & Shemyakin, M. F. (1972) Genetiku, 5, 142-148. 9. Chestukhin. A. V.. GreDachevskv. A. A.. Prozorov. A. A. & Shemyakin, M. F. (1977) Bioorian. Chem. 3, 981 -988. 10. Garen, A. & Levinthal, C. (1960) Biochim. Biophy.7. Actu, 38. 470 - 483. 11. Richardson, C. C. (1966) J . Mol. Bid. 15, 49-61. 12. Friedman, E. A. & Smith, H. 0. (1972) J . Biol. Chem. 247, 2846-2853. 13. Clewell, D. B. & Helinski, D. R. (1969) Proc. Nail Acad. Sci. U.S.A. 62, 1159-1166. 14. Smith, H. 0.& Wilcox, K. W. (1970) J . Mol. Biol.51,379-391. 15. Davis, B. S. (1964) Ann. N . Y. Acud. Sci. I21, 404-427. 16. Weber, K. & Osborn, M. (1969) J . Biol. Chem. 244,4406-4412. 17. Spirin, A. S. (1958) Biokhirnzya, 23, 656-661. 18. Lowry, O., Rosebrough, N. J., Farr, A. & Randall, R. (1951) J . Biol. Chem. 193, 265-275. 19. Friedman, E. A. & Smith, H. 0. (1972) J . Bid. Chem. 247, 2859 - 2865. 20. Anai, M., Hirahashi, T. &Tokagi, Y . (1970) J . Bid. Chrm. 245, 767 - 174. 21. Eickler, D. C. & Lehman, I. R . (1977) J . Bid. (’hem. 252, 499 - 503. 22. Doly, J . & Anagnostopoulos, K. (1976) Eur. J . Biochem. 71, 309-316. 23. Mizushima, S., Machida, Y . Sr Kitahara, K . (1963) J . Bucteviol. 86,1295 - 1300. 24. Lowry, 0. H., Carter, J., Ward, J. B. & Glaser, L. (1971) J . Biol. Chem. 246, 651 1 - 6521. 25. Bunville, L., Geiduschek, E., Rawitscher, M. & Sturtevant, J. (1965) Biopolymers, 3, 213-240. 26. Karu, A. E. & Linn, S. (1972) Proc. Natl Acad. Sci. U . S . A . 69, 2855 - 2859. 27 Orlosky, M. & Smith, H. 0. (1976) J . Bid. Chrm. 251, 61176121.
M. F. Shemyakin, A. A. Grepachevsky, and A. V. Chestukhin, lnstitut Bioorganicheskoj Khimii imeni M . M. Shemyakina, Akademiya Nauk S.S.S.R., Vavilova ulitsa 32, Moskva, U.S.S.R. 117312
Properties of Bacillus subtilis ATP-Dependent Deoxyribonuclease Michael F. SHEMYAKIN, Alexander A. GREPACHEVSKY, and Arsenii V. CHESTUKHIN Institute of Bioorganic Chemistry, U.S.S.R. Academy of Sciences, Moscow (Receivcd March 20, 1979)
A purification procedure described previously resulting in electrophoretically pure Bacillus subtilis ATP-dependent DNAase has now been modified by adding a fractionation stage with Polymin P to permit large-scale isolation of the enzyme. It has been found that the enzyme molecule ( M , = 300000) consists of two large subunits with M , 155000 and 140000. The purified enzyme has three activities: (1) DNAase on linear single-stranded and double-stranded DNAs, (2) DNAunwinding and (3) ATPase. Circular duplex DNAs were not affected by the enzyme. Study of the dependence of these activities on temperature, pH, and ATP and Mg2+ concentrations has revealed two different states of the enzyme. At low ATP concentrations and alkaline pH, it showed chiefly nuclease action, degrading considerable amounts of DNA to small fragments five residues long on average. At higher ATP concentrations and neutral pH (more physiological conditions) it predominantly unwound DNA. Simultaneously it cut preferentially one of the duplex strands to fragments more than 1000 residues in length. The results obtained suggest that the energy of the enzymecleaved ATP is mainly expended on unwinding rather than on degrading DNA molecules.
ATP-dependent DNAase has been found in many bacterial species. For some of them (Escherichia coli [l 1, Diplococcus pneumoniae [2], Bacillus suhtilis [3]) proof was obtained that the enzyme participates in genetic recombination. Early work on the properties of ATP-dependent DNAase established that under conditions optimal for DNAase action, namely at high pH and low (0.05 mM or less) ATP concentration, it vigorously cleaves linear duplex DNA to acid-soluble oligonucleotides five residues long on average. At the same time the E. coli enzyme does not digest circular DNA duplexes unless these contain gaps, but it nicks singlestranded circles [4]. These properties are difficult to reconcile with current concepts of the mechanism of genetic recombination which apparently involves rather large DNA fragments only. Therefore attempts were made recently to find conditions under which ATP-dependent DNAase does not cause considerable DNA degradation. It was shown that on brief incubation at neutral pH, lowered temperature and high ATP concentration [5,6] or low Mg2+ concentration [7], DNA degradation by E. coli or Haemophilus inpuenzae ATP-dependent DNAase decreases manifoldly. The resulting large single-stranded fragments _ _ -~ Enzymes. ATP-dependent deoxyribonuclease or exonuclease V
(EC 3.1.4.33); alkaline phosphatase (EC 3.1.3.1); lysozyme (EC 3.2.1.17); S1 nuclease (EC 3.1.4.21).
and duplexes with long single-stranded tails could be intermediates of recombination. Earlier we found ATP-dependent DNAase in B. subtilis cells, investigated some properties of the partially purified enzyme [3] and obtained data for its involvement in the recombination process [8]. Recently we were able to overcome the difficulties associated with the lability of B. subtilis ATP-dependent DNAase and to develop a novel procedure for its purification to electrophoretic homogeneity [9]. This has led us to a further study of its structure and properties.
MATERIALS AND METHODS
Materials DEAE-Sephadex A-25 and A-50 and Sephadex G-150 were obtained from Pharmacia Fine Chemicals; Tris from Merck; ATP, dithiothreitol, bovine serum albumin (fraction V) and S1 nuclease from Calbiochem; sodium dodecylsulfate and EDTA from Serva; 2-mercaptoethanol, acrylamide, N,N'-methylene-bisacrylamide, agarose, egg-white lysozyme and imidazole from Bio-Rad; calf-thymus DNA from P-L Biochemicals ; ['4C]ATP and [p3'P]ATP from Amersham ; polyethyleneimine thin-layer cellulose MN 300 from Brinkmann Instruments, Inc. Alkaline phos-
B. suhtilis ATP-Dependent DNAase
41 8
phatase (specific activity 20 units/mg protein) was isolated from E. coli [lo]. Isolation of D N A [14C]DNA of B. subtilis was isolated as described earlier [3]; unlabelled phage T7 DNA was prepared according to Richardson [l 11, 14C-labelled and j2P-labelled phage T7 DNAs according to Friedman and Smith [12]; unlabelled and 14C-labelled ColEl DNAs were isolated from the E. cofi strain K12 637 thy- [13]. The specific activity of B. subtilis [14C]DNA was 7 counts min-' pmol-', that of E. coli [32P]DNA 20 counts min-' pmol-', that of T7 [14C]DNA 1.5 counts min-' pmol-' and that of ColEl ['"CIDNA 9 counts min-' pmol-'. DNA concentrations are expressed as mononucleotide molar concentrations. Determination of Enzyme Activities DNAase activity was determined in a standard assay mixture (0.2 ml) containing 0.05 M Tris-HC1, pH 7.5-9.0, 0.01 M MgCL, 1 mM dithiothreitol, 0.5 mM EDTA, 0.2 M KCl, 6 nmol of labelled DNA, 0.05-3.0 mM ATP and about 1 unit of enzyme (fraction VI). After 20 rnin at 37 "C, samples were chilled and 0.05 ml of 1 "/, bovine serum albumin and 0.1 ml of 18 "/, trichloroacetic acid were added to each sample. The precipitates were removed by centrifugation. Radioactivity of the supernatant was determined in a Mark I1 counter after adding 10 ml of Triton X-100 scintillation fluid [14]. One unit of DNAase activity was defined as the amount of enzyme that made 1 nmol of DNA acid-soluble in 20 rnin at 37 "C. ATPase activity of the enzyme was determined in a reaction mixture (0.1 ml) containing 0.02 M TrisHCl, pH 7.5-9.0, 0.01 M MgC12, 1 mM dithiothreitol, 0.5 mM EDTA, 0.2 M KCl, 3 nmol of T7 DNA, 0.05 - 3.0 mM ATP, [14C]ATP (6 x lo3 counts/min) and 0.5 unit of enzyme. After incubation for 20 rnin at 37"C, samples were chilled and 0.01 ml of 0.5 M EDTA was added to each. From each sample a 0.01-ml aliquot was applied to a polyethyleneiminecellulose plate, with 5 pl of a mixture of ATP, ADP and AMP (10 mM each), followed by ascending chromatography in 4 M NaCl/l M CH3COOH (4/1). After drying, the ATP and ADP spots were detected in ultraviolet light, cut out and their radioactivity was counted. The amount of single-stranded DNA formed by the action of ATP-dependent DNAase was determined as the fraction of total DNA made acid-soluble by the single-strand-specific S1 nuclease. The reaction mixture for the ATP-dependent DNAase was diluted threefold into a solution containing 0.1 M sodium
acetate buffer, pH 4.5, 0.1 M NaCl, 1.3 mM ZnC11, 7 % glycerol, 0.7 % sodium dodecylsulfate and 400 nmol/ml of denatured calf thymus DNA. To each sample 0.05 unit of S1 nuclease was added followed by incubation for 45 rnin at 37 "C. The samples were chilled and 0.2 ml of calf thymus DNA solution (5 mg/ml) and 0.3 ml of 18 % trichloroacetic acid were added to each. The precipitate was removed by centrifugation, scintillation fluid was added and supernatant radioactivity was determined. Isolation and Characterization ATP-Dependent DNAase
of
B. subtilis ATP-dependent DNAase was isolated by the procedure [ 91 including the chromatography on the DEAE-Sephadex A-25 column ( 5 x 100 cm) loaded with preformed KCl gradient, ammonium sulfate fractionation, Sephadex G-150 chromatography, DEAE-Sephadex A-50 chromatography and DNA-Sepharose chromatography. The conditions of the DEAE-Sephadex chromatography do not allow one to use more than 50 g of frozen cells for a single isolation of the enzyme. To increase the scale of isolation an additional step was included, the Polymin P fractionation of crude extract, which removed 85 - 90 % of the protein and almost all the nucleic acid. Addition of this step enables one to fractionate the material from 400 g of frozen cells on the DEAESephadex A-25 column. This step is described below. 400 g of frozen B. subtilis cells were suspended in 800 ml of 0.05 M imidazole-HC1, pH 7.0, 10% glycerol, 1 mM EDTA, 5 mM 2-mercaptoethanol and 0.3 M KCl (buffer A) with added phenylmethylsulfonyl fluoride (20 pg/ml) and lysozyme (0.5 mg/ml) and incubated for 20 rnin at 37 "C. The lysate was chilled at 4"C, sonicated at 22 kHz and maximal input (UZDN-1 sonicator, USSR) without allowing it to heat above 8 "C, and centrifuged at 12000 rev./min for 1 h in a Beckman JA-14 rotor. The supernatant was diluted with buffer A to 10 mg protein/ml; 10% Polymin P (pH 7.5) was then added to a final concentration of 0.6 %,followed 10minlaterby centrifugation for 15 min at 8000 rev./min. The enzyme was eluted from the pellet with 800 ml of 0.2 M ammonium sulfate in buffer A and precipitated from the eluate with 60% ammonium sulfate. The precipitate was dissolved in 100 ml of buffer A and then used for further fractionation as previously described [9]. By increasing the isolation scale eightfold, it is possible to isolate five or six times more ATP-dependent DNAase, the inclusion of the Polymin P step giving rise to a slightly poorer yield (7 - 8 % instead of 8-10%). This method yields 0.5-0.6mg of the enzyme with a specific activity of about 4000 units/mg (zz 5000-fold purification). By 5 % polyacrylamide gel electrophoresis at pH 9 according to Davis [15], the
M. F. Shemyakin, A. A. Grepachevsky, and A. V. Chestukhin
final preparation showed only one Coomassie-stained band which contained more than 90 % of the adsorbed dye while less than 10% of the adsorbed dye was uniformly distributed along the gel. On centrifugation in a glycerol gradient, the ATPdependent DNAase sediments as a single peak with a sedimentation coefficient of 13- 14 S, which corresponds to a molecular weight of approximately 300000. Roughly the same molecular weight value is obtained by estimating the relative mobility of the enzyme in polyacrylamide gel according to Davis [15]. On electrophoresis by the procedure of Weber and Osborn [16], the enzyme protein separates into two bands of equal intensities with molecular weights of 155000 and 140000 [9], which suggests that the molecule of B. subtilis ATP-dependent DNAase consists of two large subunits of similar size. DNA was determined according to Spirin [17] and protein according to Lowry [18].
419 3
n
R
Migration
-
Fig. 1. Effect of ATP-dependent DNAase on linear undcircular C‘olEI DNAs. T o the standard assay mixture (pH 9.0) a mixture of linear (2), circular relaxed (1) and super-coiled ( 3 ) ColEl DNAs (0.5 pg of each) was added, followed by incubation at 37 ‘ C for 20 min. Samples were chilled and submitted to electrophoresis in 0.9 2, agarose gel at 0.8 V/cm fof 15 h at 22 ‘C. The gels were stained with ethidium bromide and subjected to densitometry. (A) Without ATP; (B) with ATP
2.5 7
RESULTS 2 .o
The Main Activities of B. subtilis ATP-Dependent DNAase
ATP-dependent DNAases from other bacteria are remarkable for their inability to degrade circular double-stranded DNAs, both covalently closed supercoiled ones and those relaxed by nick in one strand [4,19]. This property is shared by B. subtilis ATPdependent DNAase. As can be seen from Fig. 1, incubation with this enzyme of a ColEl DNA containing linear, circular super-coiled and relaxed molecules, leads to disappearance of linear molecules only. The enzyme degrades the double-stranded DNA much better than the single-stranded one: in the same conditions 34% of double-stranded and 5.2% of single-stranded DNA become acid-soluble. The products of duplex DNA digestion by the enzyme comprise a collection of oligonucleotides (1 - 20 residues in length, as determined by DEAEcellulose chromatography) with an average length of five residues (Fig. 2). Similarly sized degradation products have been reported for ATP-dependent DNAases from other organisms [4,20]. In common with the ATP-dependent DNAase of E. coli [21], or H. influenzae [6,7], the B. subtilis enzyme has a DNA-unwinding activity (Fig. 3 - 7) and a DNA-dependent ATPase activity producing ADP and Pi (Table 1). Thus, the B. subtilis ATP-dependent DNAase has three main activities : (a) linear duplex DNA degradation to acid-soluble products, (b) linear duplex DNA unwinding and (c) DNA-dependent degradation of ATP. The interrelationship of these activities under various conditions is described below.
6
-- 1.5 -aF
5 g c
O
2
1
4 :
.o
3 6 2
0.5 1
0
10
20
30
45
60
0
Incubation time (rnin)
Fig.2. Determination of average lengths oJ the products 41 T7 [ 3 Z P ] D N Adegradation by ATP-dependent DNAase. The standard assay mixture (pH 9.0) containing 0.05 m M ATP and 7 nmol of [32P]DNA was incubated at 37°C. Samples were taken at the indicated times, heated for 5 min at 90°C and chilled. From each sample a 0.05-ml aliquot was drawn to determine acid-soluble radioactivity. In the remaining 0.15 ml of the mixture, the amount of 32P resistant to alkaline phosphatase was measured [20]. (-0) Acid-soluble [32P]DNA;(-0) alkaline-phosphax ) oligonucleotide length Vase-sensitive 32P; ( x ~
Table 1. ATPase action of ATP-dependent DNAase 0.05 m M ATP, [14C]ATPor [ Y - ~ ~ P I A(6 T Px lo3 counts/min each) was added to the standard assay mixture; incubation was for 20 min at 37°C. The hydrolysates were assayed as described under Materials and Methods Incubation time
min 0 20
Difference
Amount of _______________~___ [I4C]ATP [14C]ADP [y-32P]ATP
32P1
nmol _________________ 4 60 4 59 0 41 4 69 0 19 0 51
0 40 4 43
- 4 08
+428
- 3 81
-
+403
420
B . subrilis ATP-Dependent DNAase
Interrelationship of the Enzyme Activities in Various Reaction Conditions
The dependence of DNAase activity on ATP concentration is not a simple one (Fig.3). The curve is bell-shaped with a maximum near 0.05mM ATP. The curve of DNA-unwinding activity versus ATP concentration is quite different. As the ATP concentration is raised this activity increases to reach a plateau at about 1 mM ATP. At this ATP concentration DNAase activity is only about one tenth of the maximal. The ATPase activity reaches a plateau at the same ATP concentration and the curve is similar to that for the unwinding, but not DNAase, activity. Fig. 4 shows that the temperature dependence of DNAase activity is different from that of unwinding activity which resembles the temperature dependence of ATPase activity, especially in the initial portion of the graph. The same is true of the graphs relating these three activities to pH (Fig. 5 ) and to Mg2+ concentration (Fig. 6). Again in both these cases the
ATPase activity curves resemble more the unwinding activity than the DNAase activity curves. This suggests that ATP energy is expended by the enzyme mainly on unwinding the DNA rather than on degrading it. Kinetics of A TP-dependent DNAase Activities under ‘Physiological’ and ‘Unphysiological’ Conditions
In the cases considered, the nuclease activity was at its maximum under overtly unphysiological conditions: alkaline pH, high temperature or low ATP concentration. It was therefore of interest to study in more detail the behaviour of the enzyme under less unphysiological conditions and to compare this behaviour with that under conditions optimal for nuclease activity. To create conditions optima1 for DNAase activity, a low ATP concentration (0.05 mM) and an alkaline pH (9.0) were used, and to mimic physiological conditions, a high ATP concentration (3 mM) and a neutral pH (7.5) were employed.
100 80
2.5
-
2.0
60
3>
1.5 1.0
a
0.01
0.1 ATP (rnM)
1
G
1.5
5 z Q
0
-
V
-
0
.o
PH
Fig.3. Eifect of A TP (,onwntrution on A TP-dependent DNAasu activities. DNAase ( x x ), ATPase (0-0) and DNAunwinding (0-0) activities were determined in standard assay ATP cleavage mixtures (pH 9.0). (A--A)
28 40 Temperature (“C)
+-
m 2
-s?
40
20
-E,
52
Fig. 5. eflect of p H on ATP-dependent DNAase activities. DNAase (x x ) , ATPdse ( 6 - 0 ) and DNA-unwinding (+o) activities were determined at the indicated pH values in standard assay mixtures (0.05 mM ATP) ~~
0
60
Fig. 4. Effect of temperature on ATP-dependent DNAase activities. DNAase ( x -- x), ATPase (0-0) and DNA-unwinding (0-0) activities were determined at the indicated temperatures in standard assay mixtures (pH 9.0, 0.05 mM ATP)
M g C b (mM)
Fig.6. Effect of MgClz c,onceniration on ATP-dependent DNAase uctivities. DNAase (x-x), ATPase (-0) and DNAunwinding (O--O) activities were determined at the indicated MgClz concentrations in standard assay mixtures (pH 9.0,0.05 mM ATP)
42 1
M. F. Shemyakin, A. A. Grepachevsky, and A. V. Chestukhin
Fig.7 shows the kinetics of the three main activities of the enzyme under the above conditions. In either type of conditions the enzyme rapidly unwinds DNA, with the curves reaching a plateau level in the first 2- 3 min of incubation. At pH 9.0 and 0.05 mM ATP (Fig.7A) it degrades about half of the DNA' whereas under physiological conditions (Fig. 7 B) DNAase activity is virtually absent. DNAase activity likewise takes a short time (5 min) to attain a plateau level. Unwinding and DNAase activities coincide during the first 1.5 min of incubation. This coincidence suggests that as it moves down the DNA molecules the enzyme preferentially degrades one strand. The rate of ATP cleavage considerably slows towards the end of incubation at the low ATP concentration and remains almost unchanged at the high ATP concentration. The amount of ATP cleaved in the latter case is 5 - 10-fold greater than in the former case. At the low concentration, ATP is degraded almost completely at 15 min of incubation. Since under the physiological conditions the enzyme produces no acid-soluble material, its DNAase activity was analysed by centrifugation of digested DNA in glycerol gradient. The results are presented in Fig.8. It can be seen that at pH 9.0 and 0.05 mM ATP the enzyme rapidly degrades a proportion of the DNA to small fragments while leaving much of it intact and producing no fragments of intermediate size. At pH 7.5 and 3 mM ATP there are also seen two peaks following the action of DNAase: one of these coincides with the intact DNA and contains no single-stranded nicks (alkaline gradient) while the other consists of rather large fragments with molecular weights of about lo6. The first peak of the neutral gradient (physiological conditions) contains about 30 % of single-stranded material and the second peak about 50 % of such material. Judging from the second peak position in the alkaline gradient, the average length of cleaved single-stranded fragments is several thousand residues. Therefore under physiological conditions the nuclease action of ATP-dependent DNAase is highly limited.
DISCUSSION In a previous paper [9] we described a new procedure for the isolation of B. subtilis ATP-dependent DNAase. Here we present a modification of that procedure which enables one to increase the scale of isolation eightfold and to obtain five or six times more of the enzyme. In both methods we were able to overcome the extreme instability of the enzyme and to obtain its stable and electrophoretically homogeneous preparations with a satisfactory yield (7- 10 %) and a rather high specific activity ( z 4000 units/mg) [9]. The actual specific activity of the B. subtilis enzyme
4
3
2
1 ; 3 3
-
Incubation time (min)
2-
0
3. s 0
U
L
20 - 60 ~
50
a
t
15
- 40 - 30 - 20
10
5
- 10
0.5 1.5
3
5 10 Incubation time (min)
15
Fig. 7. Eflect of' p H und A T P c'o~c'entrutioii 011 ATP-riiyxwdiwr DNAase activities. DNAase ( x x), ATPase (6 0 ) and DNA-unwinding (0-0) activities were determined in standard assay mixtures. (A-A) The ratio of degraded ATP to unwound DNA. (A) 0.05 mM ATP, pH 9.0 (unphysiological conditions); (B) 3 mM ATP, pH 7.5 (physiological conditions) ~
was much higher as in our assay conditions the DNA degradation reaction was far from linear (Fig.7A); in fact, judging from the initial velocity of the reaction, it was about tenfold higher and approximated to that of the E. coli ATP-dependent DNAase [4,21]. According to our data, the molecule of B. suhtilis ATP-dependent DNAase is composed of two subunits with molecular weights of 155000 and 140000. According to Doly and Anagnostopoulos [22] it consists of five subunits with molecular weights ranging from 81 000 to 42 500, yet our molecular weight estimate for the active enzyme coincides with that of the above authors (E 300000). We suggest that the large number of polypeptides found in the enzyme by Doly and Anagnostopoulos are the result of the action of proteinases during the early stages of purification since the activity of the latter in B. subtilis is known to be very high. This could explain the instability of the B. subtilis ATP-dependent DNAase isolated by these authors and the consequent low yield, specific activity and degree of purification of their preparations. High instability of the enzyme, both on isolation and on storage, compelled us to abandon completely the purification procedure which we proposed earlier [3] and
422
B. subtilis ATP-Dependent DNAase
bottom
Fraction number
top
Fig. 8. Glycerol gradient centrifugation of [14C]DNA treated with A TP-dependent DNAase ut various p H and A T P concentration values. Standard assay mixtures (0.2 ml) having (A, B) unphysiological conditions (pH 9, 0.05 mM ATP) or physiological (C, D) conditions (pH 7.5, 3 mM ATP) were incubated at 37' C. At the times indicated, samples were chilled and either (A, C) KCI and EDTA were added to final concentrations of 0.7 M and 5 m M respectively (neutral gradient) or (B, D) NaOH was added to a final concentration of 0.25 M (alkaline gradient). The mixtures were layered onto 10-35% linear glycerol gradients containing 0.05 M Tris-HC1, pH 7.6, 1 mM EDTA and 0.7 M KCI (neutral gradient) or 0.25 M NaOH (alkaline gradient). Centrifugation was done at 44000 rev./min for 3.5 h at 15°C in a Spinco L5-50 SW 50 rotor. Five-drop fractions were collected from the bottom of each tube into 24 vials and radioactivity was measured. (A) Neutral gradient, unphysiological conditions; (B) alkaline gradient, unphysiological conditions; (C) neutral gradient, physiological conditions; (D) alkaline gradient, physiological conditions. Incubation: 0 min (-a), 1.5 min (+O), 5 min ( x x) ~
which differs from that of Doly and Anagnostopoulos only in lacking the electrophoresis step. That the B. subtilis ATP-dependent DNAase is composed of' two subunits is further supported by the two-subunit composition of this DNAase in E. coli where the subunits were estimated to have molecular weights of 140000 and 128000 by Goldmark and Linn [4] and 130000 and 120000 by Eickler and Lehman [21]. We have found the properties of highly purified B. subtilis ATP-dependent DNAase to be very similar to those of E. coli [5,21] and H . influenzae [6,7] ATPdependent DNAases. The enzyme displays three main activities : ATP-dependent degradation of linear singlestranded and double-stranded DNAs to acid-soluble products averaging five residues in length, ATPdependent unwinding of the linear duplex and DNAdependent cleavage of ATP to yield ADP and Pi. Study of these activities of B. subtilis ATP-dependent DNAase in relation to the main reaction parameters has revealed some interesting relationships. Thus the curves of DNA-unwinding activity versus
ATP concentration, temperature and ambient pH are all very similar to the corresponding curves for ATPase activity but are sharply different from those for DNAase activity (see Fig. 3 - 5). Cleavage of ATP and unwinding of DNA both reach a maximum by 1 mM Mg2+ whereas DNA degradation is at its maximum only at 6-8 mM Mg2+. An appreciable depression of unwinding activity at Mg2+ concentrations above 2 mM is probably due to stabilization of the DNA double helix by divalent cation. All these results indicate that ATP energy is apparently used to unwind DNA rather than to degrade it. This assumption is also supported by the relatively slight temperature dependence of unwinding activity (see Fig. 6). Consideration of the three activities in relation to pH, temperature, and ATP and Mg2+ concentrations has disclosed two different functional states of the B. subtilis ATP-dependent DNAase. At high pH values and low ATP concentrations (i.e. under 'unphysiological' conditions), the enzyme degrades more than 40% of the DNA to acid-soluble material, and this appears to be preceded by the unwinding of more
423
M. F. Shemyakin, A. A. Grepachevsky, and A. V. Chestukhin
than 60% of the DNA under such conditions. At neutral pH and high ATP concentrations (‘physiological’ conditions) [23,24], the enzyme retains its ability to unwind DNA but is quite incapable of degrading it to acid-soluble products. It must be emphasized that in contrast to the ATPdependent DNAases of E. coli and H. influenzae [5,7], the B. subtilis enzyme fails to exhibit exonuclease action under physiological conditions even after a relatively long incubation with DNA at 37 “C. It can be calculated that early in the reaction about one molecule of ATP is used up on unwinding each DNA base pair under unphysiological conditions and about two molecules under physiological ones. We thus have a reasonably good agreement, for the molar energy yield is 29 kJ (7 kcal) in the reaction ATP +ADP Pi and 33 kJ (8 kcal) in the DNA denaturing reaction [25]. It remains, however, unclear why prolongation of the incubation, particularly at a high ATP concentration, results in a manifold increase in the molar ratio of cleaved ATP to unwound DNA. The large differences between ‘physiological’ and ‘unphysiological’ conditions as regards the amount of ATP used up by the enzyme and the kinetic curves of ATPase and unwinding activities, could be attributed to the fact that large amounts of energy are required to maintain the long DNA strands in the unwound state. It is known, however, that the ATPdependent DNAases of E. coli [26] and H. influenzae [27], while being unable to cleave DNA molecules with cross links, hydrolyze ATP no less vigorously than on incubation with intact DNA. It is therefore conceivable that much of the ATP is cleaved by the enzyme ‘idly’ also in the reaction with a normal DNA devoid of cross links. The most important function of ATP-dependent DNAase seems to be the ATP-dependent unwinding of DNA accompanied by cleavage of large fragments from one of the strands. The big enhancement of DNAase activity which seems to occur at low ATP concentration may be a consequence of an insufficient speed of enzyme movement along the DNA molecule during unwinding because of an ATP deficit. Finally it should be added that under physiological conditions, as compared to unphysiological ones, the properties of ATP-dependent DNAase can meet the
+
demands made on it as a participant in genetic recombination much better.
REFERENCES 1 . Buttin, G. & Wright, M. (1968) Cold Spring Hurhor Syinp. Quant. Biol. 33, 259 - 269. 2. Vovis, Q. F. & Buttin, G. (1970) Biochim. Biophys. Actu, 224, 42 - 54. 3. Chestukhin, A. V., Shemyakin, M. F., Kalinina, N . A. Sr Prozorov, A. A. (1972) FEBS Lett. 24, 121 - 125. 4. Goldmark, P. J. SC Linn, S . (1972) J . Biol. Chem. 247, 1849 1860. 5. Mackay, V. & Linn, S. (1974) J . Biol. Chem. 249, 4286-4295. 6. Friedman, E. A. & Smith, H. 0. (1973) Nut. New Biol. 241, 54- 58. 7. Wilcox, K . W. & Smith, H. 0. (1976) J . Biol. Chem. 251, 6121-6134. 8. Prozorov, A. A,, Kalinina, N. A,, Naumov, L. S., Chestukhin, A. V. & Shemyakin, M. F. (1972) Genetiku, 5, 142-148. 9. Chestukhin. A. V.. GreDachevskv. A. A.. Prozorov. A. A. & Shemyakin, M. F. (1977) Bioorian. Chem. 3, 981 -988. 10. Garen, A. & Levinthal, C. (1960) Biochim. Biophy.7. Actu, 38. 470 - 483. 11. Richardson, C. C. (1966) J . Mol. Bid. 15, 49-61. 12. Friedman, E. A. & Smith, H. 0. (1972) J . Biol. Chem. 247, 2846-2853. 13. Clewell, D. B. & Helinski, D. R. (1969) Proc. Nail Acad. Sci. U.S.A. 62, 1159-1166. 14. Smith, H. 0.& Wilcox, K. W. (1970) J . Mol. Biol.51,379-391. 15. Davis, B. S. (1964) Ann. N . Y. Acud. Sci. I21, 404-427. 16. Weber, K. & Osborn, M. (1969) J . Biol. Chem. 244,4406-4412. 17. Spirin, A. S. (1958) Biokhirnzya, 23, 656-661. 18. Lowry, O., Rosebrough, N. J., Farr, A. & Randall, R. (1951) J . Biol. Chem. 193, 265-275. 19. Friedman, E. A. & Smith, H. 0. (1972) J . Bid. Chem. 247, 2859 - 2865. 20. Anai, M., Hirahashi, T. &Tokagi, Y . (1970) J . Bid. Chrm. 245, 767 - 174. 21. Eickler, D. C. & Lehman, I. R . (1977) J . Bid. (’hem. 252, 499 - 503. 22. Doly, J . & Anagnostopoulos, K. (1976) Eur. J . Biochem. 71, 309-316. 23. Mizushima, S., Machida, Y . Sr Kitahara, K . (1963) J . Bucteviol. 86,1295 - 1300. 24. Lowry, 0. H., Carter, J., Ward, J. B. & Glaser, L. (1971) J . Biol. Chem. 246, 651 1 - 6521. 25. Bunville, L., Geiduschek, E., Rawitscher, M. & Sturtevant, J. (1965) Biopolymers, 3, 213-240. 26. Karu, A. E. & Linn, S. (1972) Proc. Natl Acad. Sci. U . S . A . 69, 2855 - 2859. 27 Orlosky, M. & Smith, H. 0. (1976) J . Bid. Chrm. 251, 61176121.
M. F. Shemyakin, A. A. Grepachevsky, and A. V. Chestukhin, lnstitut Bioorganicheskoj Khimii imeni M . M. Shemyakina, Akademiya Nauk S.S.S.R., Vavilova ulitsa 32, Moskva, U.S.S.R. 117312
