Quantitative image analysis for structural genomics

Abstract

The last decade the use of fluorescence microscopy in combination with DNA specific probes to produce three-(or four-) dimensional data has grown into a mature state. The improvements in imaging microscopy methods (hard- and software) and specific labeling methods (wetware) enables us to get a better understanding of the genome structure. These improvements have led to the research of structural genomics, where structure and function of the genome are connected. This is not new; the organization of the interphase nucleus has been studied since the late 19th century. It is now well accepted that the position of chromosomes in the nucleus plays an important role in gene regulation.

A recent point of interest is the study of telomeres, the ends of DNA-strands. The telomeres prevent the ends of DNA sticking together and also prevent DNA degradation during DNA syntheses. We found the structure of positions of the telomeres (the telomeric territory) to be re-organized during the cell cycle[1]. While in G0/G1 and S phase the telomeres are in a volume that is sphere-like, in G2 they have moved into a volume which is more disk-like (the telomeric disk). We also observed that this organization in G2 is lost is cancerous cases. Furthermore it looks like the telomeres tend to clump together, forming aggregates, in these cancerous cases.

The data acquisition (the recording of images) is usually based on digital imaging, producing enormous sets of data to be analyzed. Therefore the need for quantitative image analysis methods and algorithms in this field of cancer research has also grown. These algorithms have to consider the whole procedure that is being used to acquire the data, including optical properties of the microscope, nature of the probes and acquisition parameters.

We are in the process of developing tools for the quantitative analysis of digital microscopy images of the genome. This includes: loading and display of 3D images, automatic segmentation of regions of interest (e.g. nuclei, telomeres, chromosomes) and 3D image analysis algorithms (e.g. calculation of the distribution parameter, ?T, describing the flatness of the disk in which the telomeres are distributed[2]).

[1] Chuang, T.C.Y., et al., The three-dimensional organization of telomeres in the nucleus of mammalian cells. BMC Biology, 2004. 2(12).
[2] Vermolen, B.J., et al., Characterizing the Three-Dimensional Organization of Telomeres. Cytometry, 2005. in print.