The DNA binding stringency is crucial for the proper function of proteins that recognize specific DNA sequences. Our now extensive set of measurements of the dependence of the relative binding constants to specific and nonspecific sequences on osmotic pressure shows that water is a key factor that discriminates among different modes of binding. Specific sequence complexes sequester many fewer waters at the protein-DNA interface than noncognate or nonspecific complexes. The differences in number of associated waters between specific, noncognate, and nonspecific complexes seen using osmotic pressure are consistent with the differences in the crystal structures when available. Importantly, the large osmotic pressure dependence we have measured for EcoRV, BamHI, and EcoRI restriction endonucleases and for gal and cro repressors indicates that measurement of protein-DNA binding constants and specificity in dilute solution cannot be directly applied to the crowded environment of the cell. We have developed a novel single turn-over kinetics protocol utilizing osmotic stress to measure the numbers of Mg2+ ions, H+ protons, and water coupled to the cleavage kinetics of EcoRV. Type II restriction endonucleases require metal ions to specifically cleave DNA at canonical sites. Despite the wealth of structural and biochemical information, the number of Mg2+ ions used for cleavage by EcoRV, in particular, at physiological divalent ion concentrations is still not established. The number and identity of metal ions required for DNA cleavage by nucleases are most commonly inferred from x-ray structures. Somewhat unexpectedly, EcoRV has been crystallized with different numbers of metal ions varying from 1-3 per monomer. The dependence of kinetic rates or equilibrium binding constants on salt concentration or pH is routinely used to determine the numbers of ions or protons coupled to the cleavage or binding reactions. Kinetic measurements with EcoRV, however, have been plagued by the lack of binding specificity of the enzyme in the absence of Mg2+ and the fast dissociation rate of specifically bound protein. The equilibrium binding of Mg2+ by free enzyme greatly complicates the measurement of divalent ion binding coupled to cleavage. We found previously that at optimal conditions for enzymatic activity (pH 7.5 and 100 mM NaCl) the ratio between EcoRV binding constants to a DNA fragment with a specific recognition site and to a nonspecific 30 bp long oligonucleotide is only 56 in the absence of divalent ions. This is much smaller than measured for most other restriction nucleases. This specific-nonspecific binding ratio increases, however, to 450,000 in the presence of 4 osmolal betaine glycine. Once again, this large increase is because the specific EcoRV-DNA complex sequesters some 125 fewer water molecules from glycine betaine than the nonspecific complex, as also confirmed by the x-ray structures. Since the specific to nonspecific mode of protein binding is a necessary early step in dissociation of the specific complex, osmotic pressure also greatly slows the dissociation rate of the specific complex. We utilize osmotic stress in our single-turnover experimental protocol to ensure that the cleavage reaction is initiated with protein virtually stoichiometrically bound to a DNA fragment containing the recognition site in the absence of Mg2+ and that dissociation of the specific complex before cleavage is minimized. The cleavage reaction is initiated by adding Mg2+ and enough competitor oligonucleotide to prevent rebinding of any dissociated enzyme either before or after cleavage. In this way we specifically measure the numbers of Mg2+, H+, and water molecules coupled to the cleavage kinetics between the initial specific EcoRV-DNA complex formed without bound Mg2+ and the transition cleavage states. This is an extension of the self-cleavage assay we developed previously to measure specific binding of the restriction endonucleases through their cleavage with high precision. An additional feature of using osmotic pressure in the single-turnover kinetic experiments is that since the kinetic steps leading to cleavage also require water uptake, the kinetics are slowed to a time scale of minutes to hours, allowing convenient measurement. We find that the time dependence of the fraction DNA cleaved after incubation of EcoRV-DNA complexes with Mg2+ exhibits a lag phase and that the EcoRV cleavage kinetics can be well described by two consecutive reaction steps with double stranded cleavage only occurring after the second. The dependence of these rate constants on Mg2+ concentration, pH, and osmotic pressure gives a number of Mg2+ ions, protons, and water molecules coupled to each kinetic step of the cleavage reaction. The kinetic steps are remarkably similar, coupled to the binding of 2 Mg2+ ions, the release of 2 H+ protons, and the binding of 30 water molecules. The number of waters coupled to each step indicates the magnitude of the conformational change required to bind the extra Mg2+ ions. The 30 water molecules observed is much less than the 125 needed for dissociation suggesting a more subtle structural is needed. To the best of our knowledge, this is the first direct thermodynamic determination of the number of Mg2+ needed for EcoRV cleavage and was only possible using osmotic stress. Further experiments with supercoiled plasmids showed that each step is also associated with the cleavage of a DNA single strand. The first step leads to relaxation of the supercoil and the second to the formation of linear DNA. We suggest that each kinetic step reflects an independent, rate limiting conformational change of each monomer of the dimeric enzyme that allows Mg2+ ion binding and consequent single strand cleavage. These results indicate a plasticity of the enzyme that was not at all apparent from its crustal structure. This modified single turn-over protocol has general applicability.

Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Zip Code
Sidorova, Nina Y; Scott, Thomas; Rau, Donald C (2013) DNA concentration-dependent dissociation of EcoRI: direct transfer or reaction during hopping. Biophys J 104:1296-303