Quantitative prediction of fluorescence quantum yields for tryptophan in proteins

Variation of intrinsic tryptophan (Trp) fluorescence intensity and lifetime in proteins is widely exploited to follow changes in protein structure such as folding/unfolding, substrate or ligand binding, and protein-protein interactions. Although a credible candidate for the source of weak Trp fluorescence in some proteins has long been believed to be electron transfer from excited Trp to an amide carbonyl group, recognition of when such a process should be exceptionally efficient has not been possible. Here we propose a reasonable basis for the 30-fold variation by the use of quantum mechanics-molecular mechanics simulations in which the energy of the lowest Trp ring-to-amide backbone charge transfer (CT) state is monitored during dynamics trajectories for 24 Trps in 17 proteins. The energy, fluctuations, and relaxation of high lying CT states are extremely sensitive to protein environment (local electric field direction and strength). Application of basic electron transfer theory with a single empirical electronic coupling reveals that the entire 30-fold-range is explainable from local electric field effects on electron transfer to a nearby amide, although other acceptors may sometimes be responsible for a low yield. A key new concept uncovered in this work is that charged groups near the Trp can have profound effects on fluorescence lifetime and quantum yield, but location is critical. Negative (positive) charge will decrease (increase) quantum yield if closer to the indole ring than to the electron acceptor because these arrangements stabilize (destabilize) the CT state. If the charge is closer to the acceptor, the opposite will be true. A semiquantitative prediction using only the average energy gap and variance was achieved by using a universal electronic coupling constant and energy offset.