Search

The methods covered in this unit include flow cytometry of metaphase chromosomes, chromosome dissection, and the DOP-PCR amplification methods for reverse chromosome painting. Successful application in these areas requires care and attention to methodological details, and this unit is particularly comprehensive.

We have investigated the performance of two types of standard flow cell sorter instruments, a System 50 Cytofluorograph and a FACSTar PLUS cell sorter, for the on-line centromeric index (CI) analysis of human chromosomes. To optimize the results, we improved the detection efficiency for centromeres in two ways. A higher efficiency was obtained first by elongation of the chromosomes and second by introducing a high resolution lens system for laser beam focusing. In the two-parameter flow karyotype of CI and DNA content of human chromosomes, distinct peaks are produced not only by the larger chromosomes 1-8 and X, but by the smaller nonacrocentric chromosomes 9-12 and 16-20 as well. As the chromosomes 9-12 cannot be distinguished by other flow karyotyping methods, we discriminated and sorted chromosomes 12 and 10 from 9 and 11 to investigate the capacity for the separation of chromosomes in this group. A purity of at least 90% was achieved; in the isolated population the fraction chromosomes 12 was 55%; the remaining 45% were chromosomes 10 (40%) and unidentifiable chromosomes (5%).

This paper highlights two improvements of the on-line centromeric index (CI) analysis for the detection of chromosome abnormalities. On-line CI versus DNA content analysis of an EBV-transformed cell line, with a deletion (11)(p13p15.1), of a patient with aniridia and Wilms' tumour demonstrates the first improvement of the method of on-line CI analysis for flow karyotyping and sorting; a reciprocal translocation, insertion, or deletion can, when the cell type contains not more than a few of these types of abnormalities, be traced to the p-arm or q-arm of the relevant chromosome. On-line CI analysis was also performed with chromosomes isolated from a transitional cell carcinoma of the bladder. Cytogenetic analysis of this cell line showed numerous chromosomal abnormalities. Chromosomes of this cell line were also karyotyped by bivariate flow cytometry using a different set of parameters: Hoechst 33,258 fluorescence intensity (HOfl) versus chromomycin A3 fluorescence intensity (CAfl). A comparison of these results reveals the second improvement of the CI method for the detection of chromosome abnormalities; bivariate analysis of CI versus propidium fluorescence (PIfl) are complementary to bivariate analysis of HOfl versus CAfl. Chromosomes with distributions that fuse together in the HO/CA flow karyotype may be distinguished as individual peaks on the basis of their CI values.

Slit-scan flow cytometry provides a method to analyze large numbers of metaphase chromosomes in a relatively short time due to morphological features. The high detection rate requires fast computing for on-line analysis. Up to now, this has been achieved using special-purpose computers, parallel systems or other complex hardware. Here, we describe an algorithm that can be implemented on a general-purpose personal computer. Digitized chromosome profiles can be classified by several criteria especially for the detection of chromosome abnormalities in biological dosimetry. A data set of approximately 4600 profiles was used. Programming in assembler results in an average computing time of about 600 microseconds per profile. Thus on-line evaluation of slit-scanning data appears to become feasible for many flow cytometers running nowadays.

A method was developed to detect dicentric chromosomes by slit-scan flow cytometry. The two centromeres of dicentric chromosomes are represented by the two dips in the trimodal fluorescence profile. A trimodal profile can, however, also be generated by aggregates of chromosomes. We tested the effectiveness of slit-scan profile criteria that were applied to discriminate between trimodal profiles generated by dicentrics and trimodal profiles generated by artefacts. A Profile-Dip Counter (PDC) module was designed that can assess, in real time, the number of dips in slit-scan profiles. The PDC module was used in combination with a Cytofluorograph System 50 cell sorter for slit-scan sorting of chromosomes prepared from irradiated V79 cells. Chromosomes corresponding to trimodal profiles were sorted individually onto slides for subsequent visual inspection by fluorescence microscopy. The isolated chromosomes were stretched by treatment with trypsin to increase the efficiency for centromere detection. When fixed with glutaraldehyde, chromosomes could be sorted intact on slides. We found that trimodal profiles are generated by dicentric chromosomes as well as by monocentric and aggregated chromosomes. When stringent pulse-shape criteria were applied for the selection of profiles, the yield of dicentric chromosomes was 70% of the sorted chromosomes.

Bivariate flow karyotyping can be used for the detection of recurrent chromosome abnormalities in tumor cells. For this purpose 2 cell lines originally derived from Ewing's sarcomas and 4 cell lines from peripheral neuroectodermal tumors were used. The characteristic t(11;22) was known to be present in 5 cell lines. The remaining cell line was known to have a variant t(2;11;22;21) translocation. Metaphase chromosomes were stained with the fluorescent dyes Hoechst 33258 and Chromomycin A3 and analyzed subsequently using bivariate flow cytometry. The resulting bivariate flow karyotypes of the tumor cells were normalized by a standardized procedure using a computerized method and compared with a reference flow karyotype of normal chromosomes. In 5 cell lines two recurring abnormal chromosome peaks were identified at positions expected for the der(11) and der(22) chromosomes characteristic for the reciprocal t(11;22)(q24;q12). In the remaining cell line with the variant t(2;11;22;21), only the peak representing the der(22) was identifiable. It is concluded that bivariate flow karyotyping can be used for the semiautomated detection of recurrent translocations and the assessment of their variability among different tumors.

The total variation of chromosome peak positions, in bivariate distributions of Hoechst 33258 and chromomycin A3 fluorescence of 19 healthy individuals, was compared with the experimental variation, determined from 23 bivariate distributions of chromosomes prepared separately from a single cell lineage. The experimental variation in Hoechst and chromomycin fluorescence and the relative chromosomal DNA content were determined from experiments performed over several days. The additional variance contributed by time was the same as the daily variance. The accuracy by which the relative chromosomal DNA content can be calculated from bivariate peak positions was investigated. A least squares method was used to fit the distributions of relative DNA content, obtained, respectively, from mono- and bivariate flow analyses of chromosomes from the same cell lineage. In general the DNA contents match quite well, but for a few chromosomes a difference was found, statistically discernible at the 5% level. The average relative chromosomal DNA content of the chromosomes from the 19 normal individuals, calculated from bivariate peak positions, showed a linear relation with the estimates published by other investigators.

In this paper we describe an indirect fluorescence double staining procedure for the simultaneous detection of IdUrd and CldUrd in the same cell nucleus. Two commercially available antibodies were selected for this purpose. A rat anti-BrdUrd monoclonal antibody from Sera-lab was found to bind specifically to CldUrd and BrdUrd. A mouse monoclonal anti-BrdUrd antibody from Becton Dickinson used in a 1:2 dilution binds to all halogenated deoxyuridines but, when the cells were extensively washed with Tris buffer with a high salt concentration, almost no binding to CldUrd was observed. An immunofluorescence procedure was developed, based on these primary antibodies, raised in different species (rat and mouse), in combination with highly purified second antibodies: FITC conjugated goat antirat and Texas-Red conjugated goat antimouse.

Using slit-scan flow cytometry, the shape of human metaphase chromosomes, as expressed in their centromeric index (CI), and the DNA content of the chromosomes have been used as parameters in bivariate flow karyotyping. The resolution of the DNA vs CI flow karyogram of the larger chromosomes up to chromosome 13 is much higher than the resolution obtained in the DNA-based monovariate flow karyogram. Chromosome length appears to be an important factor in the resolution of the DNA vs CI-based flow karyogram. A method has been developed to obtain chromosomes in suspension that are long enough for adequate analysis. Several chromosomes that cannot be distinguished or are difficult to discriminate in the DNA-based karyogram can now be distinguished as individual peaks, e.g., chromosomes 1 and 2. The peak of chromosomes 9-12 can be separated into two peaks formed by chromosomes 9 and 11, and 10 and 12, respectively. The advantage of the system applied in this study is that the DNA vs CI analysis is performed on-line, allowing chromosomes to be sorted on the bases of their CI. Pulse shapes of the selected chromosomes can be recorded simultaneously with the transmission of the sorting command. The purity of the sorted fraction can be estimated from the off-line inspection of these pulse shapes. Fractions of chromosome 1 have been sorted out on the basis of the CI information, centrifuged on slides, fixed and subsequently banded with trypsin and Giemsa or hybridized with the chromosome 1 specific probe, pUC 1.77. The observed purity under the selected conditions ranges from 80%-99% and is in accordance with the estimates of the purities made on the basis of the simultaneously recorded pulse shapes. Fixation of the chromosome suspension prior to flow cytometric analysis and sorting appears to be essential for the preservation of their morphology and has no adverse influence on the resolution of Giemsa banding or on the quality of in situ hybridization.

An instrument is described which analyses shapes of fluorescence profiles generated by particles passing through the focussed laser beam of a flow cytometer. The output signal of this pulse-shape analyzer is used as input for the signal processing electronics of a commercial flow cytometer system. The instrument detects dips in pulse-profiles; a shape parameter named Pulse Dip Index (PDI) is defined as the ratio of the integrated signal from the beginning of the pulse until the first dip, relative to the integrated signal of the complete profile. This PDI is similar to the Centromeric Index of chromosomes. The composition of aggregates in mixtures of fluorescent particles of different sizes was evaluated by PDI analysis. In our experiments the PDI was determined within 30 microseconds from the onset of the pulse-profile and particles with a specified morphology of interest were selected for on-line registration of their profiles as digitized pulse-shapes. In a cell sorter system, the PDI can be used as a parameter for sorting.

A method is described by which metaphase chromosomes are isolated from cells of a transplantable rat sarcoma. The chromosomes are derived from cells in a suspension prepared by trypsinisation of tumours from rats that have been treated with vindesine 24 h before excision in order to accumulate cells in mitosis. Histograms obtained for the chromosomes of the solid tumour are compared with flow karyotypes of cells cultured for 20 h or for several generations in vitro.