Exploring DNA-Mediated Charge Transport with Fast Radical Traps

Abstract

The π-stack of DNA is competent for mediating charge transport (CT), both by single-step and multi-step mechanisms. The yield of long-range single-step CT from photoexcited 2-aminopurine to guanine across adenine tracts has a shallow, periodic distance dependence, with increasing amplitude and decreasing slope with temperature. To measure total CT yield, herein we employ the fast radical traps N 2 -cyclopropylguanine ( CP G), and N 6 -cyclopropyladenine ( CP A), which are similar to the unmodified bases, but undergo rapid decomposition upon oxidation. We find that decomposition of CP G by a photoexcited rhodium intercalator across an adenine tract has similar periodic distance dependence to quenching of 2-aminopurine by guanine, and the same temperature dependence as well. In contrast, decomposition of CP G by photoexcited 2-aminopurine is monotonic with respect to adenine tract length, and also competes with back electron transfer (BET). Eliminating BET by separating 2-aminopurine from the adenine tract with three high-potential inosines restores the non-monotonic distance dependence. We also determined decomposition of CP A along adenine tracts by photoexcited rhodium, and found the CT yield to be distance-independent, demonstrating that the periodicity associated with guanine oxidation is with respect to adenine tract length, not donor-acceptor separation. This length-dependent periodicity, and the associated temperature dependence, support a model of conformational gating in the formation of CT-active domains along the DNA. DNA-mediated electrochemistry is facile in self-assembled monolayers on electrodes, and redox-active dyes are reduced through the DNA π-stack at potentials far lower than those of the individual bases. Since cytosine is the most readily reduced base, we incorporated CP C into DNA monolayers to assay for bridge occupation, and CP C decomposition was not observed. To explore the relative contributions of single-step and multi-step mechanisms to CT yield across adenine tracts, we compared quantum yields previously collected from 2-aminopurine fluorescence quenching experiments to those of CP G decomposition. For seven or eight intervening adenines, single-step CT accounts for the entire CT yield, while for four to six adenines, multi-step CT is the dominant mechanism. We interrupted multi-step CT by substituting CP A for an adenine on the bridge, and found the total CT yield across five or six intervening adenines is lowered to the single-step CT yield. Blocking single-step CT by replacing the terminal guanine with redox-inactive inosine does not affect CP A decomposition on the bridge. These results imply that single-step and multi-step CT processes are not in direct competition for these assemblies, consistent with the model of conformationally gated CT-active states.