6th ESACP Congress, Heidelberg, April 7-11, 1999

A095
DETECTION OF FLUORESCENT FOCI AND EVALUATION OF SPATIAL RELATIONSHIPS IN 3D-FLUORESCENCE MICROSCOPY IMAGES OF MAMMALIAN CELLS
Linnman C *, Bengtsson E *, Ekholm-Jensen S **, Zetterberg A **

* Centre for Image Analysis, Uppsala University, Uppsala, **Dept. of Oncology-Pathology, Div. of Tumorcytology, Karolinska Institute, Stockholm, Sweden

The high specificity of molecular in situ techniques in combination with new 3D techniques in fluorescence microscopy opens new possibilities for investigating structural and functional molecular and cell biology. Since 3D volume images are difficult to evaluate quantitatively by visual methods, computer assisted methods are needed. We have developed a two step method for studying the spatial relationships between fluorescent foci. First we find fluorescent foci in 3D images by locating local intensity maxima. This method in combination with 3D neighborhood mean filtering can be used to detect fluorescent foci also in noisy and low contrast images. The method only takes the nearest neighborhood into consideration when detecting foci and is therefor independent of variations in fluorescence intensity due to uneven illumination and light attenuation. After detecting the foci we create a distance map by applying a 3D distance transform to the detected foci of one wavelength channel. The map gives direct information about the shortest distance between foci of the wavelength channel where the distance map was created and foci in other wave length channels. This two step method is currently being used for evaluation of the spatial relationships between newly synthesized DNA and replication complexes. Cells are double stained for BrdU, a thymidine analogue, and PCNA, an auxiliary protein of DNA polymerase d. Foci for BrdU and PCNA co-localize at sites of DNA replication when cells are fixed directly after exposure to BrdU. This result will be compared with cells incubated with fresh medium and cultured for 2 hours after the BrdU incorporation, a so called pulse-chase experiment. By varying the pulse-chaise time, this method could be used to measure the speed of movement for DNA replication units during S-phase.