Binding of 9-aminoacridine to deoxydinucleoside phosphates of defined sequence: Preferences and stereochemistry

The effect of sequence on the binding of 9-aminoacridine to DNA has been investigated by studying its interaction with deoxydinucleoside phosphates of different sequences using proton nuclear magnetic resonance. Quantitative binding information can be obtained by comparison of the proton chemical shift behavior of 9-aminoacridine upon addition of dinucleoside phosphate to various models for the interaction using least-squares computer fitting procedures. The simplest model that fits the data includes (1) dimerization of 9-aminoacridine and (2) a mixture of 1:1 and 2:1 (dinucleoside phosphate/9-aminoacridine) complexes. The computed parameters allow comparison of binding constants and stereochemistry for different sequences. The 1:1 complexes seem to involve interaction of the ring nitrogen with the backbone phosphate and stacking of one or both chromophores on the acridine; preference in binding is observed for alternating (purine-pyrimidine or pyrimidine-purine) over non-alternating (purine-purine) dinucleoside phosphates. The 2:1 complexes involve intercalation of the acridine between two complementary dinucleoside phosphate strands with weak sequence preferences in binding. The stereochemistry of intercalation differs between non-alternating purine-purine sequences and the alternating pyrimidine-purine or purine-pyrimidine sequences in having the 9-aminoacridine stacked with the purines of one strand rather than straddling the purines on opposite strands. The difference in stereochemistry could possibly be a determining factor in frameshift sequence specificity.