Computer Programs That Allow Fast Acquisition, Visualization and Overlap Quantitation of Fluorescent 3D Microscopic Objects by Using Nearest-Neighbor Deconvolution Algorithm

Abstract

Analysis of complex multicolor immunofluorescence staining patterns at high magnification is still a challenge for many laboratories working with relatively weak fluorescence signals. To facilitate the analysis of very fine speckled staining patterns in the eukaryotic nucleus, we tested the feasibility of generating tailor-made computer algorithms with the help of a very high level graphical programming language that is able to produce complex routines without the need of code writing. The graphical programming system ISee is a language that uses logical symbols (p-nodes) that are connected into executable flowchart-like networks. We generated programs that perform rapid automatic acquisition of three-color fluorescent images from multiple focal planes followed by rapid deblurring by using the nearest-neighbor deconvolution algorithm. Here we present five programs that carry out the following tasks in a single continuous series: acquisition and projection of an image stack, either as a single RGB image (TROOPER3) or as a stereo-rendered RGB image pair (STEREOTROOPER); acquisition and generation of combined mosaic image from different focal planes (MOSAIC), followed by quantitation of the extent of overlap between the red and green colors (CALCULATE-OVERLAP); and finally quantitation of red and green signals representing different nuclear proteins that localize to areas of the nucleus with different DNA densities (CHROMATIN). We found that a wide-field illumination fluorescent microscope equipped with cold CCD camera and controlled by the described programs provides an excellent and very flexible alternative to laser confocal microscopes for routine work. Source codes of the programs are available at ftp://laszlo.mtc.ki.se/ISee.