Hydrolysis of Nucleoside Triphosphates Catalyzed by the recA Protein of Escherichia coli

The DNA-dependent ATPase activity of the recA protein of Escherichia coli shows a complex dependence on ATP concentration. With a single-stranded (SS) DNA cofactor, the Hill coefficient for ATP is 3.3 at pH 8.1 and 1.4 at pH 6.2. With a double-stranded (DS) DNA cofactor, the Hill coefficient is 3.3 at pH 6.2 (no reaction is detectable at pH 8.1). In the presence of SS DNA, the K,,, for ATP is 20 PM, independent of pH, while with DS DNA at pH 6.2, KmATP is 100 e. These and other observations indicate that the interaction of recA protein with ATP is influenced by both pH and DNA cofactor. ADP, UTP, dlTP, and GTP are competitive inhibitors of the ATPase activity of recA protein, indicating that there is a single binding site for nucleoside triphosphates. Nucleoside triphosphates, but not ADP, reduce the Hill coefficient for ATP hydrolysis and thus can contribute to the cooperative effect of ATP.

ATP is required for binding of recA protein to DS DNA. Although not required for SS DNA binding, ATP also influences the stability of recA protein-SS DNA complexes (6, 7). Furthermore, ATP has an effect on the oligomeric structure of the recA protein (6, 9). Thus, ATP serves a number of functions in the complex series of events which make up the hybridization reactions.
To gain further insight into the relationship of these effects to the mechanism of ATP hydrolysis and the DNA hybridization reactions, we have performed a steady state kinetic analysis of ATP hydrolysis. We have found that hydrolysis has an exponential dependence on ATP Concentration. This dependence is complex in that it is affected both by pH and the DNA factor. These factors account for the different sensitivities to ADP of SS and DS DNA-dependent ATP hydrolysis that we had observed previously (8). In addition, we have found that UTP, which is also hydrolyzed by the recA protein, is a competitive inhibitor of the ATPase activity, consistent with the notion of a single (or overlapping) active site for nucleoside triphosphate hydrolysis. Finally, we have observed that the pH, previously found to affect DS DNAdependent ATP hydrolysis, also affects the dependence on ATP and DNA concentration of SS DNA-dependent ATP hydrolysis.

EXPERIMENTAL PROCEDURES
All reagents and assays were as described in the previous paper (8). All velocities were determined by measuring the time course of the reaction.

Dependence of A T P Hydrolysis on A T P Concentration-
Both SS DNA-dependent ATP hydrolysis at pH 8.1 and DS DNA-dependent ATP hydrolysis at its optimum, pH 6.2, exhibited a complex dependence on ATP concentration ( Fig.  1) with Hill coefficients of about 3.3 at low ATP concentrations (Fig. 2). This finding suggests that there is cooperative binding of ATP to the recA protein. Above 50 PM ATP, both reactions had Hill coefficients of 1, indicating that the enzyme was nearly saturated with ATP. Surprisingly, when hydrolysis was examined at pH 6.2 with SS DNA, there was a much simpler ATP dependence (Fig. l a ) , with a Hill coefficient of 1.4 at low ATP concentrations (Fig. 2). Thus, the pH, previously seen to affect binding of recA protein to DS DNA (6), also affects the ATP dependence of SS DNA-dependent hydrolysis.
The apparent KmATP values for ATP hydrolysis catalyzed by recA protein are given in Table I. These values were determined above 50 PM ATP, where the Hill Coefficient was unity, using an Eadie-Hofstee plot. Despite the pH dependence of the Hill coefficient, KmATP for SS DNA-dependent hydrolysis was about 20 PM at 30 "C at both pH 6.2 and 8.  In the absence of DNA, the reciprocal plot for the ATP dependence of hydrolysis was essentially linear at both pH 6.2 and 8.1 (Fig. 3). KmATP was about 60 ~L M at pH values above 7 and 120 p~ at pH 6.2, while V,,,/E was 0.016 and 0.13 mol of ADP formed/min/mol of recA protein at high and low pH,   Fig. 4 were replotted as in Fig. 2. Vmax was determined from an Eadie-Hofstee plot. The "No ADP" curves are from Fig. 2.

TABLE I1
Inhibition by ADP of SS and DS DNA-dependent ATPase activities of the recA protein Reactions were performed as described in Fig. 4. Vmax and KiADP were determined from Eadie-Hofstee plots and the Hill coefficients from Fig. 5 respectively (Table I). Thus, both KmATP and V,../E were pH dependent in the absence of DNA. Inhibition of Hydrolysis by ADP-ADP was an inhibitor of all 3 DNA-dependent ATP hydrolytic reactions (Fig. 4

TABLE I11
Inhibition of ATPase activity of recA protein by nucleoside triphosphates SS DNA-dependent reactions were as described in Fig. 6. Vms and K, were determined from an Eadie-Hofstee plot as in Fig. 6 and Hill coefficients as in Fie.  Reactions were performed as in Fig. 7 p H 6.2 and 7.5 were not greatly affected by ADP (Fig. 5, a  and b, Table 11). However, ADP affected the Hill coefficient for ATP in DS DNA-dependent ATP hydrolysis (Fig. 5c). In the presence of ADP, the HU coefficient was not constant but decreased gradually with increasing ATP concentration. At the lowest ATP concentrations, the Hill coefficient was greater than 4. In addition, as the ADP concentration increased, higher concentrations of ATP were required to attain a Hill coefficient of 1. These results imply that ADP does not contribute to and may, in fact, disrupt the cooperative interaction of ATP with recA protein.
Inhibition of Hydrolysis by Other Nucleoside Triphosphates-Both UTP, which is hydrolyzed by recA protein, and dTTP, which is not, behaved as inhibitors of the SS DNAdependent ATPase reaction at pH 8 (Fig. 6). Inhibition was competitive since Vmax was unaffected (Table 111)  These results suggest that other nucleoside triphosphates can contribute to the cooperative interaction of ATP with recA protein.
Dependence of ATP Hydrolysis on DNA Concentration-Titration of DNA in the various hydrolytic reactions (Fig. 7) indicated that the KmDNA values increased with increasing recA protein concentration (Table IV). This result may reflect the fact that DNA is not present in large excess over enzyme. As seen in Fig. 7, the reaction reached maximal velocity at 5 to 10 nucleotides/recA protein monomer, near the saturation value observed in DNA-binding studies (7).
Binding of DS DNA to recA protein at pH 8 requires that SS DNA be present (4) and is optimal at about 5 SS DNA nucleotides/recA monomer. Below this value, SS DNA is limiting, as observed for ATP hydrolysis. DS DNA (in excess) did not stimulate ATP hydrolysis when SS DNA was limiting and had little effect on the requirement for SS DNA (Fig. 8). This finding suggests that the initial interaction with SS DNA is rate-determining.

DISCUSSION
Two striking aspects of ATP hydrolysis catalyzed by the recA protein are the differences in KmATp values and Hill coefficients observed with SS and DS DNA cofactors. One explanation for this behavior is that there exist different ATP binding sites on the enzyme, specific for the different ATPase reactions. However, since all reactions show the same nucleotide specificity (8) and the hydrolyzable substrates (ATP and UTP) competitively inhibit each other, it is likely that the active site for hydrolysis is the same in all ATPase reactions. The differences in the kinetic parameters must, therefore, reflect differences in the mechanism by which hydrolysis is stimulated at this active site.
The differences in the KrnATP values obtained in the presence of SS and DS DNA are difficult to interpret for complex mechanisms. However, it is noteworthy that KIA'", which, for competitive inhibition, reflects the ADP binding constant, parallels the KmATP. Binding studies show that recA protein binds SS, but not DS, DNA in the absence of ATP (6). It is, therefore, plausible that ATP binds to recA protein-SS DNA complexes in the SS DNA-dependent reaction and to free recA protein in the DS DNA-dependent reaction. Thus, the different KmATP values might reflect a difference in ATP affinity or dissociation of these forms of the recA protein.
Since the Hill coefficient is greater than 1, more than 1 ATP molecule is required/hydrolytic cycle for the maximum rate of hydrolysis. This suggests a cooperative process in which the initial binding of ATP stimulates further binding of ATP. At pH 6.2, the reduction in the Hill coefficient seen with SS DNA suggests that 1 or more of these binding steps may be bypassed. Since there is probably only a single ATP binding site per recA protein monomer (13), the observed values of the Hill coefficient imply that the active form of the enzyme is at least a trimer, consistent with earlier observations (9).
Other factors may also contribute to these characteristics. The nonlinear dependence on enzyme concentration for DS DNA-dependent hydrolysis could result in a nonlinear dependence on ATP concentration. Furthermore, binding of ATP could be important for the dissociation of ADP from the enzyme, especially in the DS DNA-dependent reaction where the extent is also affected by the ADP concentration (8).
We had previously found that, at pH 6.2, ADP is a more potent inhibitor of ATP hydrolysis in the presence of DS DNA than in the presence of SS DNA (8). Although the KmATP values suggest weaker binding of ATP in the DS DNAdependent reaction, it is unlikely that this can account for the differential inhibition since KIADP is also greater in the pres-ence of DS DNA. Rather, the sensitivity of DS DNA-dependent hydrolysis appears to be due to the nonlinear ATP dependence of this reaction as compared with SS DNA-dependent hydrolysis (compare Fig. 4, b and c ) . Because of the greater exponential dependence on ATP concentration with DS DNA, a reduction in bound ATP due to competition by ADP causes a disproportionate reduction in velocity when compared with SS DNA-dependent hydrolysis.