Biophys. J. 106, 639–648 (2014)

Credit: ELSEVIER

Fluorescence-based biosensors exist for monitoring numerous cellular processes, including visualization of single neuronal action potentials in vivo. Fluorescence-based biosensors of membrane voltage (Vm) exist but use intensity-based measurements, which cannot report an absolute numerical value. Also, these biosensors suffer from the inability to measure slow shifts in resting voltage, as, for instance, that which occurs during apoptosis. To explore potential alternatives for voltage sensing, Hou et al. sought a biosensor based on the light-powered proton pump Arachaerhodopsin 3 (Arch) in which voltage controls the equilibrium between a fluorescent state, protonated at the retinal Schiff base, and a nonfluorescent deprotonated state. The counterion to the Schiff base, Asp95, has a key role in modulating protonation. Screening a library of Arch mutants at Asp95 found that the D95H variant showed the best voltage-sensitive fluorescence. Reasoning that a change in illumination wavelength at constant voltage would change the photostationary distribution of the ArchD95H conformational ensemble in time, the authors monitored the trajectory of the relaxation from a fluorescence distribution induced by blue wavelength light ('pump') to a new distribution at orange wavelength ('probe') and concluded that, indeed, the relaxation between photostationary conformations could report on Vm. Once the pump-probe parameters were optimized, the authors used principal component analysis to parameterize voltage-dependent changes in fluorescence transients and concluded that their ArchD95H biosensor had an accuracy of ∼10 mV. Therefore, the time-based biosensor could prove useful in cases where Vm shifts are minimal or slow.