Stabilization by ATP and ADP of Escherichia coli dnaB protein activity.

The effect of adenine ribonucleotides on the stability of Escherichia coli dnaB protein in cellular crude extracts was studied. Stabilization of dnaB protein by ATP or ADP, but not by AMP, was manifested in that (i) the activity and yield of wild type dnaB protein is enhanced in the presence of ATP, (ii) the dnaB protein of E. coli dnaB mutants, such as groPB and dnaB252/ColE1::dnaC+, which is inactive in a dnaB complementation assay, can be isolated in active form in the presence of ATP or aDP, (iii) ATP or ADP protect the dnaB protein of an E. coli dnaBts mutant from inactivation at 37 degrees C, and (iv) inactive groPB and dnaBts protein can be reactivated partially by ATP. Thus, the stabilizing effect of ATP and ADP can be exploited for the isolated of otherwise inactive or labile mutant dnaB proteins.

The dnaB protein of Escherichia coli is essential throughout the replication cycle of the bacterial chromosome (1). It interacts in vitro with dnaC (2) and other priming proteins in a "primosome" (3), thereby initiating 4x174 DNA complementary strand synthesis (4, 5). It participates in 4X RF multiplication (6), phage X (7), and ColEl DNA replication (8). The dnaB protein from wild type strains has been purified to homogeneity (9, 10) and it was shown that the pure protein possesses a ribonucleoside triphosphatase activity (10, 11).
It is expected that the analysis of dnaB protein of E. coli dnaB mutants with well defined biological defects will help to clarify the role of this multifunctional enzyme. A variety of mutant dnaB or dnaB analog proteins have been isolated until now by affinity chromatography on immobilized ATP (12-14) using a 4x174 DNA-dependent dnaB complementation assay. The dnaB protein binds to ATP-agarose and can be eluted by ATP or ADP. Following affinity chromatography, the recovery of dnaB activity often was 100% or even higher indicating a stabilizing and/or regenerating effect of ATP or ADP on the dnaB protein (12, 13). Therefore, we expected that the same nucleotides would also stabilize other mutant dnaB proteins whose isolation had been hindered SO far because the proteins were found to be inactive in a dnaB complementation assay (12,15). These are the dnaB proteins of (i) E. coli groPB, a subclass of dnaB mutants, which replicate certain h P mutants but not h wild type (16), and (ii) the initiation-defective E. coli dnaB252 mutant (12, 17).
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
As will be shown in this paper, dnaB activity, indeed, can be recovered in crude extracts from such dnaB mutant strains in the presence of ATP or ADP. Furthermore, it is found that the dnaB activity from an E. coli dnaBts mutant can be stabilized or even reactivated by the same nucleotides.

EXPERIMENTAL PROCEDURES
Materials and methods, unless otherwise indicated, were as previously described (14,15,18). E. coli strains used are listed in Tables I  and 11. Adenine ribonucleotides were from Boehringer. Stock solutions (0.5 M) were prepared by dissolving the disodium salts of ATP, ADP, and AMP in bidistilled water and adjusting the pH to 7.5 by 1 M Tris-base. PE1'-cellulose (20 X 20-cm plates) were from Schleicher and Schull. Buffer A is 20 mM Tris-HC1, pH 7.5, 0.1 mM EDTA, 0.1 mM dithiothreitol, 1 mM MgC12, 50 mM NaCl, 10% (w/v) glycerol. It is supplemented with ATP, ADP, or AMP and/or additional MgC12 as indicated in Tables I and 11. Cellular Crude Extract-Bacteria were grown in TY medium and harvested at about 4 X lo8 cells/ml as described (13). The growth temperature was 30 "C if not otherwise noted. About 2 g of wet cell paste were obtained from 1 liter of culture. Bacteria were lysed by the lysozyme-spermidine procedure (19). The crude extract contained about 20 mg of protein/ml (about 50 mg of protein/g of wet cell paste) and was treated with streptomycin sulfate as described (13). Ammonium Sulfate Fraction-Crude extract contained (final concentrations in parentheses) Tris-HC1, pH 7.5 (40 mM), EDTA (1 mM), dithiothreitol(1.5 mM), NaCl(70 mM), and residual amounts of sucrose part of extract, ammonium sulfate (0.243 g/ml) was slowly added and (55%). spermidine ( 5 7 mM), and streptomycin sulfate (54%). To one stirred for 30 min, and the suspension was kept overnight at 0 "C. The precipitate was collected by centrifugation (17,000 X g, 20 min), washed with Buffer A containing ammonium sulfate (0.263 g/ml), collected again by centrifugation, and dissolved in Buffer A. The material was dialyzed by rotary shaking against Buffer A for 4 h and repeated changes of Buffer A, centrifuged once again (17,000 X g, 10 min) in order to remove insoluble material, and frozen in liquid nitrogen. The other part of extract was supplemented with ATP, ADP, or AMP, and/or MgCL (final concentrations are indicated in Tables I and 11). Incubation was for 2 min at 0 "C before ammonium sulfate was added. All the following operations were as described above except that the nucleotide and MgC12 concentration in Buffer A was the same as at the beginning. The materiai (about 20 mg of protein/ml) was stored at -70 "C (Fraction I).
Complementation Assays-dnaB-and dnaC-complementing activity was determined by using Fraction I (100 to 150 pg of protein, prepared without ATP) from strain BT1071 dnaBts (13) and PC22 dnaCts (20), respectively. Fraction I from BT1071 was heated for 1 min at 37 "C in order to inactivate residual dnaBts protein. Assays were performed as described previously (13). One unit of dnaB-and dnaC-complementing activity represents the incorporation of 1 nmol of dTMP.
Thin Layer Chromatography-Turnover of ATP in Fraction I was determined by one-dimensional lithium chloride Chromatography on PEI-cellulose layers as described (21). This system separates ATP, ADP, AMP, and adenosine from each other and from the corresponding deamination products. Fraction I was deproteinized by the addition of an equal volume of trichloroacetic acid (20%). After centrifugation, the supernatant was extracted repeatedly with H20-saturated diethyl ether in order to remove the trichloroacetic acid. Ten to 60 1.1 of deproteinized Fraction I was applied onto a PEI-cellulose plate tides were run in parallel as markers. Spots seen under the UV light (20 X 20 cm). Adenosine, inosine, and the corresponding ribonucleowere marked, scraped off the plate, and extracted by 0.1 M HCl at room temperature for 16 h. After centrifugation, the nucleosides and nucleotides were determined quantitatively by UV spectrophotometry.
The abbreviation used is: PEI, polyethyleneimine.  (15). However, when the preparations were done in the presence of ATP and MgClZ (10 mM each), dnaB activity was recovered ( Table I). More detailed studies were undertaken with the mutant B612 which is temperature-sensitive in its groPB phenotype (16). Ten mM ATP alone (without MgCM but not MgC12 alone (without ATP) was sufficient to yield dnaB activity (Table I). When 2 or 0.1 mM ATP (in the presence of 10 mM MgCM was used, no activity was found. ADP is as effective as ATP, but AMP is not (Table I). The recovery of dnaB activity was not affected when Fraction I was prepared in the presence of 0.4 M NaCl in addition to ATP and MgCk When a preparation of Fraction I with ATP and MgC12 (10 mM each) once did not yield dnaB activity, a subsequent supplementation with 10 mM ATP and repeated dialysis led to the recovery of activity (0.5 units dnaB/mg of Fraction I). In the presence of ATP and MgClz, dnaB activity of groPB612 (0.26 units/mg of Fraction I) was also recovered when the temperature of the growing culture was shifted to 40 "C for 10 min before the bacteria were harvested. dnaB activity from the wild type strain (B') was 1.8-fold higher when Fraction I was prepared in the presence of ATP and MgC12 ( Table I).
Effect of ATP on dnaBts and dnaCts Protein Activity- The dnaBts protein of E. coli BT1071 is active in dnaB complementation when Fraction I is prepared in Buffer A.
Again, the activity of the protein increased nearly 2-fold when the preparation was done in the presence of ATP and MgCL ( However, when the dnaB252 mutant strain harbors the multicopy hybrid plasmid ColEl::dnaC, dnaB activity was found provided that ATP and MgC12 were present and the strain had been grown before at 30 "C (Table 11). Surprisingly, under conditions where the dnaB mutation is completely suppressed in vivo (i.e. at 40 "C growth temperature, Ref. 23), no activity was found in vitro in the presence of ATP and MgC12, although dnaB protein was detected immunologically ( Table 11). The dnaC activity of strain BT1071 and RS185 roughly correlates with the corresponding dnaB activity, indicating complex formation between the two proteins (2). An attempt to stabilize the dnaCts protein of strain PC22 by the addition of ATP and MgClz was, however, negative. Only the corresponding dnaB activity increased 2-fold under these conditions ( Table   11).
An increase in the dnaB activity by ATP and MgC12 is accompanied by a corresponding increase in the amount of dnaB protein recovered immunologically (Table 11). Specific activities of 1300 to 2900 units/mg of dnaB protein were calculated for strain BT1071 and PC22, respectively. These values are similar to those found for the purified dnaB protein of other strains (13,14). On the other hand, the specific activity of dnaB252 protein of strain RS185 is only about 300 (Table 11).
Stabilization and Regeneration by ATP and ADP of dnaBts Protein Activity-Fraction I of strain BT1071 was prepared in Buffer A (-nucleotide), supplemented afterwards by ATP, ADP, or AMP, and incubated at 37 "C. dnaB activity remained stable for more than 2 min in the presence of ATP and ADP, but not with AMP or without nucleotides (Fig. 1). Moreover, when an inactivated sample of Fraction I was supplemented by ATP and MgClz and dialyzed for 4 h at 0 "C, nearly 60% of the dnaB activity was recovered. MgC12 alone had no effect (Table 111) Aliquots were then supplemented with ATP, ADP, AMP (10 mM final concentration each), or a corresponding volume of Buffer A ('"nucleotide").
After 5 min at 0 "C, the temperature was raised to 37 "C and incubation continued for a total of 2.5 min. Samples were taken a t intervals and dnaB activity was determined. 100% corresponds to 0.59 (+ATP), 0.55 (+ADP), 0.44 (+AMP), and 0.46 (-nucleotide) units dnaB/mg of Fraction I.

TABLE I11
Recovery by A T P of dnaB activity from E . coli BT1071 dnaBts Fraction I was prepared in Buffer A (without ATP) and incubated for 2 min a t 37 "C. The sample was subsequently divided into two parts. One part was supplemented with ATP and MgCl2, the other part with MgCL only (IO mru final concentration each). Both samples were dialyzed by rotary shaking for 4 h a t 0 "C against the corresponding buffer (see "Experimental Procedures"). When the deproteinized Fraction I was analyzed by thin layer chromatography for the presence of nucleotides, it was found that more than 99% of the ATP had been degraded. In a typical analysis, about 80% of the original amount of ATP (10 m) was recovered in the form of ADP (1.2 mM), AMP (0.7 mM), IMP (2.8 mM), and inosine (3.2 mM) as the major reaction products. The amount of these products vaned considerably from experiment to experiment, but the concentration of ATP and ADP never exceeded 0.1 and 2 m, respectively. These results suggest that the nucleotide concentration required for stabilization of dn.aB protein activity may be much less than 10 mM. On the contrary, if the degradation of ATP and ADP is too extensive, then no dnaB activity may be recovered. Certainly, this was the case when 2 and 0.1 mM ATP was unsuccessfully used to recover groPB dnaB activity (see above). An extensive degradation of ATP also may have been the reason why we failed once (out of six preparations of Fraction I from groPB612) to recover dnaB activity even when the starting concentration of ATP was 10 mM.

DISCUSSION
Stabilization by ATP and ADP of dnaB protein activity in crude extracts can be exploited for the isolation of otherwise inactive mutant dnaB proteins. For example, the protein of E. coZigroPB mutants exhibits dnaB complementing activity in the presence of ATP and ADP, respectively (Table I). Thus, the dnaB protein of groPB612 which in the absence of ATP was shown to be smaller than the corresponding dnaB' protein (15) recovers the size ( M , -250,000) of the wild type protein in the presence of ATP.2 Obviously, ATP or ADP reenforces in vitro the formation of the dnaB multimer structure from the inactive and smaller, probably monomer, form of the groPB612 protein. This interpretation is supported by the recent finding that in the presence of Mg2+ both ATP and ADP can form a binary complex with dnaB+ protein (22). The fact that the groPB612 dnaB protein regains activity in the presence of ATP but without the addition of MgC12 (Table   I) must not mean that Mg2+ ions are not required for dnaB protein stabilization, because Fraction I may still contain Mg2+ ions derived from within the bacterial cells.
In contrast to the groPB mutants, ATP and MgC12 are not sufficient to recover dnaB complementing activity from crude extracts of the dnaB252 mutant. Obviously, an excess of dnaC protein has to be present in addition (Table 11). It remains to be shown whether on further purification the dnmB252 protein retains activity per se or only in the form of a dnaB-dnaC complex. Neither dnaB nor dnaC activity was found in Fraction I from RS185 grown at 40 "C although dnaB protein was detected (Table 11). Therefore, it appears questionable that suppression of the dnaB mutant by an excess of dnaC protein in vivo (23) is brought about simply by a stabilization of dnaB252 protein in a dnaB. dnaC complex.
ATP increases the dnaB activity in Fraction I from dnaB' and dnaBts strains. The increase in activity is due to a corresponding increase in the amount of dnaB protein recovered. Thus, the presence of ATP during the preparation can improve the yield of dnuB protein. Stabilization of dnaB' protein by ATP during the purification procedure has also been observed by others (24).