Skip to main content

Development and validation of image analysis methodology for automated detection of chromosomal damage in lymphocytes

Objective


A critical first step in these activities involved development of an automated system that can automatically screen for metaphases in lymphocyte preparations using DAPI as counterstain. This system consists of a Macintosh IIfx computer, and an automated Ergolux microscope with a fluorescence illuminator, scanning stage, focus motor drive, computer controlled objective rotor and an automated filter wheel. As an image sensor, a Sony black and white CCD camera was used with capabilities for on-chip integration. Using this combination of equipment it was possible to automatically screen a single microscope slide for metaphases in a time span of about 70 minutes including a time period of 25 minutes for autofocussing. The false positive rate for metaphase finding was 7% whereas about 13% of metaphases were missed. These percentages correspond quite well with results from commercially available metaphase finders which mostly operate on Giemsa-stained slide preparations.
An important aspect of our metaphase finding system is that slides used for metaphase screening can subsequently be used for automated scoring of painted chromosomes and thus for translocation detection. The translocation detection starts with a relocation of preselected stored metaphases at high magnification using a dual band by-pass filter for FITC and DAPI. Initially single colour FISH was performed using a point probe specific for chromosome 4. This procedure was used to screen for translocations induced in lymphocytes that were given an in vitro exposure of 4 Gy of X-rays. In this pilot study 120 metaphases were screened manually and automatically for induced translocations. The result was that the automated device correctly identified translocations in 90% of the metaphases selected. In more recent experiments, dual colour FISH was employed using paint probes for chromosomes 4 and 8 labelled with spectrum green and spectrum orange respectively. In this case cells from three healthy donors were exposed to 0.1, 0.25, 0.5, 1.0 and 2.0 Gy so that dose-effect curves can be constructed both manually and automatically. Results from the manual analysis are already available and thousands of automatically selected metaphases have been stored in the computer for future automated detection of translocations. Future comparison of manual and automated findings will allow us to validate the success of the automated screening procedure.
In the mean time, a new detection system with improved efficiency and sensitivity is being tested. This promising system consists of a DM-XRA microscope, a Power Macintosh computer and a Xilix camera. This powerful combination of new equipment will increase the speed of image analysis by about 800%. At the moment, the development of software for the new equipment is almost completed. Apart from these technological improvements we have been able to work out a more sensitive staining procedure for translocation scoring. This procedure is called ratio-labelling. It essentially involves staining of a probe for an individual chromosome pair with either a particular fluorochrome (e.g. AMCA, FITC or TRITC) or with a combination of fluorochromes mixed in a well-defined ratio. In this way more colours can be obtained thereby making it possible to separate at least 6 differentially stained chromosome pairs within the same metaphase without taking any spatial information into account. Details about results obtained with the ratio-labeling procedure will soon be published.

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

Rijksuniversiteit Leiden
Address
72,Wassenaarseweg
2300 RA Leiden
Netherlands

Participants (2)

Universiteit Gent
Belgium
Address
4,Godshuizenlaan
9000 Gent
Universiteit Gent
Belgium
Address
35,K.l. Ledeganckstraat
9000 Gent