Fluorescence In Situ Hybridisation (F.I.S.H.)

History.- The ideas of using tritiated thymidine probes to mark the cellular position of DNA, in situ
hybridisation to metaphases as target DNA, and probe detection using autoradigraphy, were exploited from the 1960's. Beautiful experiments of this nature showed among other things;

that DNA replication in the chromatid is semi-conservative,
that it is not synchronous along the length of the chromatid,
established the frequent ocurrence of sister chromatid exchanges in somatic cells, and
that in the mammalian female, one of the two X chromosomes replicates late in the cell cycle.

Throughout the 1970's and 80's, a large number of genes were physically mapped to their karyotype
loci by hybridising their thymidine-labelled DNA to banded metaphases. These experiments require a
length of at least 1Kb of continuos DNA sequence available for the labelling reaction, since the signal
from shorter lenghths would not be visible. Developing times for autorads can be very long (days or
weeks) if the signal is weak.

Principles of F.I.S.H.- Tritiated thymidine can now be replaced by fluorochrome-labelled
nucleotides, which can be used as fluorescent probes and visualised by fluorescence microscopy. A simplified view of the principles of FISH including the random priming technique for labelling the probes can be seen at the Fish site from the Pasteur Institute. Three types of probes have been developed to facilitate the rapid diagnosis of different types of chromosomal aberrations in interphase
or metaphase;

Centromeric probes are usually alpha-satellite repetitive probes that mark the centromeres of human chromosomes. They are particularly useful for diagnosis of aneuploidies on interphase nuclei. Several chromosomes can be checked simultaneously.
Whole chromosome painting probes are prepared from FACS-sorted suspensions of single whole chromosomes from which DNA is PCR-amplified using degenerate oligo probes. They are very useful to confirm the identity of the chromosomes involved in a rearrangement even in difficult metaphases, such as are often seen in tumours.
Locus-specific probes, spaning relatively constant breakpoints, as in the case of many translocations seen in leukaemias and other tumours, can be used to diagnosed known translocations in interphase nuclei. Locus-specific probes can also be used to confirm microdeletions difficult to visualise under the microscope with conventional cytogenetic techniques. This would include a fourth type of probe,
Telomeric probes. Read Carl Fratter's exam-oriented essay on the clinical use and laboratory aspects of this latest type of FISH probe.

Clinical applications of FISH.- The websites for Cytocell and Vysis are worth exploring, as they
illustrate diagnostic uses of each of these products for several clinical conditions. We also include a
link to one of the pages from the Dynacare Laboratories for the same reason, but here you will also
find a variety of constitutional black and white karyotypes. Further essays in the use of FISH probes in clinical cytogenetics can be found here.

Multicolour karyotypes.- Up to the late 1990's there were just three types of fluorochrome available, giving signals in the red, green or blue regions of the visible spectrum respectively. Using chromosome-specific sequences labelled with each of those fluorochromes, up to three chromosomes could be painted and visualised simultaneously. Obviously a two-colour scheme could be used to diagnose reciprocal translocations even in poor quality metaphases where banding would be difficult.

The obvious next step is to paint all chromosomes, each pair in a different colour. This would detect any type of rearrangement, particularly translocations, within the size limitations of the resolution of the technique (about 1Mb). In the absence of 24 distinct fluorochromes, we are reduced to mixtures of the three types of fluorochrome in different proportions that would be specific for each pair of homologues. These mixtures could be prepared to maximise the ease of separation by the eye. Two approaches can be envisaged; one is to capture the coloured image 'in bulk' and analyse the spectrum pixel by pixel, and from the proportion of each colour in the image and from the initial table of colour mixtures, to classify the homologue. To facilitate visual discrimination of the image, an artificial colour could be assigned to each homologue. This is the approach in 'Spectral Karyotyping', or SKY. A similar effect is illustrated on your screen using 'Pixie', an internet tool for colour analysis. The other approach, m-FISH, is a gradual capture of the image, colour by colour, to build up the total image. Results are similar to SKY but without the clarity of the artificial colour. One such karyotype is shown below;

You can now determine each colour quantitatively by using a colour picker such as 'Pixie" (see the tools page). Another colour karyotyping approach resulting in similar images is m-FISH. where the image is capture gradually though a set of colour-specific filters, and then the final image is built up as the sum of the individual images.

Exercise-:Using the Red Green Blue (RGB) palette, and values from 0 to 255 for each colour, prepare a table of colours for the visual identification of the 23 pairs of homologues. A possible such table is suggested below, but you could make your own. Then using the painting tool of the photo editor and one of the normal metaphases, simulate a 'sky' experiment to simultaneously paint all the chromosomes in the metaphase. Once you are finished, you could use 'Pixie' simultaneously with the photo editor and your colour table to 'analyse' the colours, i.e. to determine the identity of each painted chromosome.

Chromosome No.	Colour Intensity
	Red	Green	Blue
01	255	0	0
02	0	0	255
etc.

Rx-FISH.- Yet another approach to the individual classification of each of the 23 chromosomes is Rx-FISH, in which chromosome-specific probes developed for one species of primate are hybridised to the chromosomes of another species (Humans). In this case, a banding pattern is generated that reflects, not the standard G- or R-banding, but the pattern of translocations and inversions that reflect the evolutionary relationship between both species. Further explanations and pictures can be seen here.

Comparative Genomic Hybridisation (CGH).- Another cytogenetic technique developed in the late 1990's is 'Comparative Genomic Hybridisation', or CGH. Briefly, DNA is extracted from a problem sample, say a solid tumor from which metaphases may not be available. This DNA is labelled by a standard method (nick translation) using a red fluorochrome (TRITC). Good quality normal metaphases are hybridised simultaneously to the red-labelled abnormal DNA and to green-labelled (FITC) normal DNA, and then the ratio red signal : green signal is measured along the whole chromosome from pter to qter. If the tumor had a deletion in a region of the karyotype, that region will show a relative decrease in the red : green ratio. If the tumor had a duplication in certain region, the ratio red : green will be relatively higher. Red:Green ratio measurements along several chromosomes are then averaged, and the tumor characterised by a pattern of regional DNA gains and losses. A useful database of CGH characterisation of human tumors can be seen in this site from the Charite Hospital, Humboldt University Berlin.

Exercise 1.- After clicking on the link above (Charite Hospital), click on the 'CGH on line tumour Database' and read the instructions carefully. Select from the list on the right side of the screen one of the collectives, for instance, the collection of brain tumours secondary to lung cancer. Then select one particular chromosome (say #10) and observe the profile of this chromosome in several cases, one by one. Are there consensus regions of gain/loss in most or all of the cases? See if you find some pattern characteristic of these tumours with metatstases in the brain as opposed to tumours with metastases in other sites or with no metastases. What would you do with this information?

'Fiction'.- This is an acronym for Fluorescence Immunophenotyping and interphase Cytogenetics as a Tool for the Investigation of Neoplasms. The technique was originally developed by Klaus Weber-Matthiesen et al. in 1992 at the University of Kiel, and later perfected by Brigitte Schlegelberger, Reiner Siebert and Ignacio Martin. This set of techniques attempts to identify details of the cellular immunophenotype (protein surface markers, the Cluster of Differentiation or CD proteins) and karyotype simultaneously on the same leukaemic cell. Different CD proteins characterise the surface of human leukocytes as they progress through the differentiation stages in the process of haemopoiesis, so this technique identifies in one experiment the karyotype abnormality known to be characteristic of the malignant clone, and the differentiation stage where the lesion occurred. Details of genes and sequences for the human CD proteins can be seen under the 'Acquired' button, or in PROW, the Human Leukocyte Differentiation Antigens database. Using a multicolour FICTION technique, the authors mentioned above have verified that the Reed Stenberg cells characteristic of Hodgkin's Lymphoma are positive for anti CD-30 MoAb and showed for the first time evidence for IGH class switching in 13/18 primary cases.

Summary - Cytogenetics contributes to diagnosis (establishing the clinical entity) and prognosis (because from previous statistics we may know the likely course of the disease, including reaction to treatments, from other patients with exactly the same disease entity/karyotype). Cytogenetics has made an essential contribution to gene mapping, by providing the physical map, and by pointing out regions in the karyotype where genes involved in pathology were likely to be found.