Preferential Site Binding of Monovalent Cations With the Random Coil Form of DNA

Abstract

Collins and Rogers (Chem. Biol. Interactions 19 197 1977) have suggested that monovalent tetraalkylammonium (TXA+) ions diminish the melting temperature of duplex DNA by preferential binding to the random coil form. We have investigated this suggestion using a known 16 base hairpin as the test oligomer and electrophoretic mobility in free solution as the analytical parameter. Measurements were made in the range 15-60° using background electrolyte solutions (BGE) containing 0.01-0.9 M monovalent cation (M+) and diethylmalonate as the buffering cation at pH 7.3. A 16 base all-T oligonucleotide served as a reference random coil to correct for changes in hairpin mobility contributed by changes in the physical properties of the BGE. Thermal transitions were observed as a decrease in the mobility ratio of the hairpin/coil to 1.00 as the temperature was increased at a fixed [M+]. The Tm of the hairpin increased hyperbolically from 37° as the [M+] increased linearly. The increase in Tm observed with Na+ is exactly that predicted by the DINAMelt algorithm. The span of the increase in Tm has the order Na=Li=K>NH4>TMA>Tris>TEA>TPA>TBA. Cation site binding was analyzed as a change in the mobility ratio with increasing [M+] at 20° where the hairpin conformation predominates. The mobility ratio increases hyperbolically with increasing [M+], consistent with the view that M+ preferentially site binds with the coil form. The magnitude of the increase in mobility ratio from the binding measurements at 20° correlates with the dTm/d(log M+) from the melting measurements, suggesting that the two phenomena are related. We propose that the effect of M+ on DNA melting is the net result of two opposing contributions, preferential stabilization of the counterion cloud about the structured form and preferential site binding with the unstructured random coil form.