IMARISC

In live fluorescence imaging, phototoxicity and fluorescent protein (FP) photobleaching strongly limit the excitation light dose, at the expense of image contrast, spatio-temporal resolution and experiment duration. It is generally accepted that these two processes involve as intermediates the triplet excited states of FPs and of some endogenous photosensitizers (PS) of the porphyrin and flavin families. Triplet states can indeed interact with oxygen to form singlet oxygen and other reactive oxygen species that oxidize fluorophores and cell components and induce unwanted stress responses. Our consortium of photochemists and microbiologists proposes to develop and validate a novel method to reduce phototoxicity and FP photobleaching in live fluorescence imaging. This method will exploit a photophysical process called reverse intersystem crossing (RISC) to depopulate triplet states with light. RISC is a widespread mechanism that has in particular been demonstrated in some porphyrins and flavins as well as in blue, green, yellow and red FPs. However, RISC has remained under-exploited in fluorescence imaging. In the project, we will induce it by co-illuminating the samples, during their excitation, at one or more wavelengths absorbed by the triplet states of FPs and PSs. The design of the illumination schemes will rely on a combination of time-resolved spectroscopy, dual-illumination microscopy and photophysical modelling approaches. We will start by measuring the triplet absorption spectra, yields, and lifetimes of PSs and FPs of all colours from blue to infrared by micro-millisecond transient absorption spectroscopy. RISC yields will also be determined in original 3-pulse femtosecond spectroscopy experiments. We will exploit the triplet spectra to study in vitro and then in live cells the effect of RISC induced with an appropriate wavelength on FP photobleaching and PS phototoxicity, over a wide range of light intensities. Phototoxicity will be quantified by the production of singlet oxygen, that will be detected by its near-infrared luminescence, as well as by physiological assays such as the growth rate of bacterial colonies. We will then build photophysical models integrating all the spectroscopic data and accounting for the effect of RISC on FP photobleaching and PS phototoxicity as well as its dependence on light intensities. These models will be used to optimize co-illumination for samples containing multiple PSs and/or FP colours. The new low-bleaching, low-toxicity imaging method will finally be validated in live cells by monitoring the fidelity of DNA replication in E. coli using wide-field, single and multi-colour time-lapse fluorescence microscopy. Our project should have a major impact in the field of (live) fluorescence imaging. The new method is indeed technically simple, potentially compatible with a large number of FPs and biological samples, and efficient according to our preliminary results. By allowing the use of higher doses of excitation light, it should lead to significant gains in contrast, spatio-temporal resolution and experiment duration and thus give access to a variety of new biological information. By reducing endogenous phototoxicity, that is particularly important in the blue spectral region, it could in addition revitalize the use of blue and cyan FPs, which would be a very important advance for multi-colour imaging.