Application of spectral imaging microscopy in cytomics and fluorescence resonance energy transfer (FRET) analysis

Background: Specific signal detection has been a fundamental issue in fluorescence microscopy. In the context of tissue samples, this problem has been even more pronounced, with respect to spectral overlap and autofluorescence. Methods: Recent improvements in confocal laser scanning microscopy combine sophisticated hardware to obtain fluorescence emission spectra on a single-pixel basis and a mathematical procedure called "linear unmixing" of fluorescence signals. By improving both the specificity of fluorescence acquisition and the number of simultaneously detectable fluorochromes, this technique of spectral imaging (SI) allows complex interrelations in cells and tissues to be addressed. Results: In a comparative approach, SI microscopy on a quantitative basis was compared to conventional bandpass (BP) filter detection, demonstrating substantial superiority of SI with respect to detection accuracy and dye combination. An eight-color immunofluorescence protocol for tissue sections was successfully established. Moreover, advanced use of SI in fluorescence resonance energy transfer (FRET) applications using enhanced green fluorescence protein (EGFP) and enhanced yellow fluorescence protein (EYFP) in a confocal set up could be demonstrated. Conclusions: This novel technology will help to perform complex multiparameter investigations at the cellular level by increasing the detection specificity and permitting simultaneous use of more fluorochromes than with classical techniques based on emission filters. Moreover, SI significantly extends the possibilities for specialized microscopy applications, such as the visualization of macromolecular interactions or conformational changes, by detecting FRET. (C) 2004 Wiley-Liss, Inc.