Nanoswitches based on DNA base pairs: why adenine-thymine is less suitable than guanine-cytosine

Substituted Watson-Crick guanine-cytosine (GC) base pairs were recently shown to yield robust three-state nanoswitches. Here, we address the question: Can such supramolecular switches also be based on Watson-Crick adenine-thymine (AT) base pairs? We have theoretically analyzed AT pairs in which purine-C8 and/or pyrimidine-C6 positions carry a substituent X=NH(-), NH(2), NH(3) (+) (N series), O(-), OH or OH(2) (+) (O series), using the generalized gradient approximation (GGA) of density functional theory at the BP86/TZ2P level. Thus, we explore the trend in geometrical shape and hydrogen bond strengths in AT pairs along a series of stepwise protonations of the substituents. Introducing a charge on the substituents leads to substantial and characteristic changes in the individual hydrogen bond lengths when compared to the neutral AT pair. However, the trends along the series of negative, neutral, and positive substituents are less systematic and less pronounced than for GC. In certain instances, internal proton transfer from thymine to adenine occurs. Our results suggest that AT is a less suitable candidate than GC in the quest for chemically controlled nanoswitches.