Subcellular visualization of light microscopic specimens by laser scanning microscopy and computer analysis: a new application of image analysis.

To identify subcellular organelles or to observe their pathological changes in sections prepared for light microscopy, immuno- and/or enzyme histochemical staining for the marker substances or enzymes of those subcellular organelles are frequently employed. With conventional light microscopes (CLM), however, it is hardly possible to determine whether or not the target organelles are properly stained and to confirm their fine structure. In the present study, the laser scanning microscope (LSM) was employed to obtain highly contrasted images of histochemically stained subcellular organelles at the limit of resolution in light microscopy. To refine or characterize those images, images built up as electronic signals in LSM were further processed in the Image Analysis System (IAS) with pipeline. Thus, the approximate figures of subcellular organelles such as microtubules, endoplasmic reticula, secretory granules, and mitochondria were visualized in brightfield on sections prepared for light microscopy (paraffin, frozen sections and cultured living cells). The validity of the images obtained by LSM or LSM-IAS was confirmed by immunoelectron microscopy when possible. The LSM images of histochemically stained suborganelles of various cells were definitely improved (refined and/or strengthened) by processing them with IAS.