Permeant cationic fluorescent probes are shown to be selectively accumulated by the mitochondria of living cells. Mitochondria-specific interaction of such molecules is apparently dependent on the high trans-membrane potential (inside negative) maintained by functional mitochondria. Dissipation of the mitochondrial trans-membrane and potential by ionophores or inhibitors of electron transport eliminates the selective mitochondrial association of these compounds. The application of such potential-dependent probes in conjunction with fluorescence microscopy allows the monitoring of mitochondrial membrane potential in individual living cells. Marked elevations in mitochondria-associated probe fluorescence have been observed in cells engaged in active movement. This approach to the analysis of mitochondrial membrane potential should be of value in future investigations of the control of energy metabolism and energy requirements of specific biological functions at the cellular level.

Pt K2 rat kangaroo epithelial cells and Rat-I fibroblasts were grown on conductive glass discs, fixed, and permeabilized, and the cytoskeletal elements actin, keratin, and vimentin were visualized by indirect immunofluorescence. After the fluorescence microscopy, the cells were postfixed and dehydrated for photoelectron microscopy. The contrast in these photoelectron micrographs is primarily topographical in origin, and the presence of fluorescent dyes at low density does not contribute significantly to the material contrast. By comparison with fluorescence micrographs obtained on the same individual cells, actincontaining stress fibers, keratin filaments, and vimentin filaments were identified in the photoelectron micrographs. The apparent volume occupied by the cytoskeletal network in the cells as judged from the photoelectron micrographs is much less than it appears to be from the fluorescence micrographs because the higher resolution of photoelectron microscopy shows the fibers closer to their true dimensions. Photoelectron microscopy is a surface technique, and the images highlight the exposed cytoskeletal structures and suppress those extending along the substrate below the nuclei. The results reported here show marked improvement in image quality of photoelectron micrographs and that this technique has the potential of contributing to higher resolution studies of cytoskeletal structures.Photoelectron microscopy (photoemission electron microscopy or PEM) has recently been introduced into the study of whole cells (1, 2) although the origins of this technique are old, predating both transmission electron microscopy and scanning electron microscopy (for review, see ref. 3 for physics and refs. 4 and 5 for biological applications). Photoelectron microscopy differs significantly from the established techniques of transmission and scanning electron microscopy even though the image is formed by electrons. The photoelectron microscope can be considered to be the electron optics analogue of the fluorescence microscope. UV light from a short arc lamp is focused on the specimen as in fluorescence microscopy but, instead of imaging the emitted fluorescent light with a light optics system, emitted electrons are accelerated and imaged with an electron lens system. Photoelectron microscopy has several advantages, including high sensitivity to topographic detail (3, 6), a new source of contrast based on the photoelectric effect (7-9), and an unusually short depth of information (10). The increase in image quality during the development of the photoelectron microscope over the past few years has been substantial. Although the basic mechanisms by which the photoelectron microscopy image arises are understood, the interpretation of photoelectron microscopy images of biological specimens is the focus of current research. Here we report the comparison of photoelectron micrographs with fluorescence micrographs of the same cells that have been labeled by indirect immunofluorescence techniques with antibodies specific fo...