Interaction of recA protein with a photoaffinity analogue of ATP, 8-azido-ATP: determination of nucleotide cofactor binding parameters and of the relationship between ATP binding and ATP hydrolysis

The binding and cross-linking of the ATP photoaffinity analogue 8-azidoadenosine 5′-triphosphate (azido-ATP) with recA protein have been investigated, and through cross-linking inhibition studies, the binding of other nucleotide cofactors to recA protein has also been studied. The azido-ATP molecule was shown to be a good ATP analogue with regard to recA protein binding and enzymatic function by three criteria: first, the cross-linking follows a simple hyperbolic binding curve with a K_d of 4 μM and a cross-linking efficiency ranging from 10% to 70% depending on conditions; second, ATP, dATP, and adenosine 5′-O-(3-thiotriphosphate) (ATP-γ-S) specifically inhibit the cross-linking of azido-ATP to recA protein; third, azido-ATP is a substrate for recA protein ATPase activity. Quantitative analysis of the cross-linking inhibition studies using a variety of nucleotide cofactors as competitors has shown that the binding affinity of adenine-containing nucleotides for recA protein decreases in the following order: ATP-γ-S > dATP > ATP > adenylyl β,γ-imidodiphosphate (AMP-PNP) ≫ adenylyl β,γ-methylenediphosphate (AMP-PCP) ≈ adenine. Similar competition studies also showed that nearly all of the other nucleotide triphosphates also bind to recA protein, with the affinity decreasing in the following order: UTP > GTP ≈ dCTP > dGTP > CTP. In addition, studies performed in the presence of single-stranded DNA demonstrated that the affinity of ATP, dATP, ATP-γ-S, and AMP-PNP for recA protein is significantly increased. These results are discussed in terms of the reciprocal effects that nucleotide cofactors have on the modulation of recA protein-single-stranded DNA binding affinity and vice versa. In addition, it is demonstrated that nucleotide and DNA binding are necessary though not sufficient conditions for ATPase activity. The significance of this result in terms of the possible requirement of critically sized clusters of 15 or more recA protein molecules contiguously bound to DNA for ATPase activity is discussed.