Quantitative 3D Imaging to Analyze Developmental Variability in Pregastrula Drosophila Embryos
David W. Knowles*, Soile Keranen+, Suman Dasgupta*, Mark D. Biggin+,
Damir Sudar*
Departments of Cell and Molecular Biology* and Genome Sciences+, Life
Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California
94720
The first 14 mitotic cycles in Drosophila result in a single layer of
nuclei surrounding a yolk sac in the pregastrula embryo. Despite this morphological
simplicity, the transcriptional network controlling pregastrula development
is complex and little is know about what aberrations occur or the responses
they evoke. To understand the role of biological variability during development,
we are producing novel optical imaging techniques to map the relative position
and gene expression of every nucleus within the embryo. Total DNA and specific
gene products are imaged in 3D using confocal fluorescence microscopy.
Innovative algorithms allow automatic segmentation of the DNA-image and
produce an enumerated mask defining individual nuclear boundaries. The
developmental stage of each embryo is classed by mitotic division number
and determined from the total nuclei number; then the morphological mask
is used to quantify gene-product on a per nuclei basis. What results is
a map of pregastrular development showing the variability of nuclear number,
relative nuclear packing density, and the expression pattern of specific
gene products. This work illustrates the power of quantitative optical
imaging and is an initial step towards uncovering the rules determining
how patterns of gene expression are generated.