5Dip. di Fisica, Università di Firenze, Sezione INFM di Firenze.

Classical microscopy techniques provide a bidimensional view of the objects observed. Cells are semitransparent and, in a volume of approximately 500 µ³, contain a great number of substructures whose composition and disposition is related to the function carried out. Bidimensional view of a cell results in a quite confuse image were very few focused structures are superimpose to the light scattering from its neighbouring slices, one above and one below. In histology and cytology staining methods have been developed in order to evidence particular structures and to have a more clear vision. In consequence of the artificial (counterfeit) effects produced by staining and bidimensional flattening, models of cellular structure have been constructed far from reality.

The main techniques allowing a tridimentional view are electron scanning microscopy and confocal microscopy. However, both of them produce heavy alterations of the native cell conditions. In addition, electron scanning microscopy performs the investigation of the cell surface only, while confocal microscopy has several disadvantages: low light efficiency because of the rejection of more than 95% of the fluorescent light; limited choice of fluorescent dyes available for confocal imaging, which results in a limited selection of cell components that can be imaged, severe photo-bleaching.

There is another approach to performing 3D optical sectioning microscopy. The availability of high sensitivity, low noise CCD cameras make possible the detection of low-quantum yield autofluorescence images not only as a noisy background but as high quality images, comparable to that obtained with high-quantum yield exogenous markers. Accurate images of fluorescence in cells can be acquired with minimal light exposure and a conventional wide-field epifluorescence microscope. It is possible to acquire stacks of images at different Z positions, and, under computer control, the instrument can takes optical section at sequential focal planes inside the cell. These images can be processed using an image-restoration algorithm to extract information from out-of-focus light collected in each image. The obtained 3D data set is volume rendered as a translucent substance of varying opacity, the opacity representing the density of the imaged cell. Volume rendering allows the entire 3D data field to be viewed at once.

The possibility to know the true tridimensional structure of cells and to investigate cell surface, the morphology of intracellular organelles, their disposition and reciprocal interaction is of great importance in order to understand the correlation between morphology and function.