Development of a fast TCSPC FLIM-FRET imaging system

Forster/Fluorescence resonant energy transfer (FRET) has become an extremely important technique to explore biological interactions in cells and tissues. As the non-radiative transfer of energy from the donor to acceptor occurs typically only within 1-10nm, FRET measurement allows the user to detect localisation events between proteinconjugated fluorophores. Compared to other techniques, the use of time correlated single photon counting (TCSPC) to measure fluorescence lifetime (FLIM) has become the gold standard for measuring FRET interactions in cells. The technique is fundamentally superior to all existing techniques due to its near ideal counting efficiency, inherent low excitation light flux (reduced photobleaching and toxicity) and time resolution. Unfortunately due to its slow acquisition time when compared with other techniques, such as Frequency-domain lifetime determination or anisotropy, this makes it impractical for measuring dynamic protein interactions in cells. The relatively slow acquisition time of TCSPC FLIMFRET is simply due to the system usually employing a single-beam scanning approach where each lifetime (and thus FRET interaction) is determined individually on a voxel by voxel basis. In this paper we will discuss the development a microscope system which will parallelize TCSPC for FLIM-FRET in a multi-beam multi-detector format. This will greatly improve the speed at which the system can operate, whilst maintaining both the high temporal resolution and the high signal-to-noise for which typical TCPSC systems are known for. We demonstrate this idea using spatial light modulator (SLM) generated beamlets and single photon avalanche detector (SPAD) array. The performance is evaluated on a plant specimen.