1 Department of Genetics, Yale University School of Medicine, 333 Cedar Street, New Haven, CT, USA
2 Departments of Pediatrics and Human Genetics, Cedars-Sinai Med Ctr, University of California, Los Angeles, CA, USA
3 Department of Medicine (Hematology), Ullman 925, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA

Received: 23 December 2000 / Accepted: 8 February 2001
(c) Springer-Verlag New York Inc. 2001 Mammalian Genome 12, 462-465 (2001).

Correspondence to: J. Greally; E-mail: jgreally@aecom.yu.edu

New applications for mouse molecular cytogenetics are emerging, including the definition of transgene integration sites in epigenetics studies (Alami et al. 2000), the characterization of the mouse genome following increasingly sophisticated techniques for its engineering (Su et al. 2000) and the screening for cytogenetic ab-normalities in cell lines, such as embryonic stem cells (Longo et al. 1997). FISH mapping in mouse is complicated by the relative difficulty of mouse chromosome identification (karyotyping) by laboratories not familiar with mouse chromosome banding techniques.Fluorescent karyotyping can be facilitated by the use of counterstains (Korenberg et al. 1999), labeled repetitive probes (Boyle et al. 1990), or multiple reference probes (Speicher et al. 1996; Ried et al. 1998), but each approach requires specialized equipment and experience. For example, while the use of multiple reference probes yields readily interpretable data for nonspecialists, as many as six fluorescent dyes are required for human and mouse chromosome identification in standard multiplex FISH (M-FISH) techniques (Speicher et al. 1996). This requires microscopy and imaging facilities found only in specialized centers.

We have previously described a set of BACs that map close to the centromeric and distal chromosomal ends for each mouse chromosome (Korenberg et al. 1999). We now exploit the ready identification of the centromere by simple background fluorescence (or by counterstaining) to develop a simple, three-color approach that identifies all mouse chromosomes in a metaphase. This procedure allows researchers to assign an unknown genomic clone to a specific mouse chromosome in a single experiment, without the need for chromosome identification based on banding pattern recognition. The basis of our strategy is the ability to distinguish two different chromosomes using a pair of similarly colored DNA probes on each chromosome, if their orientation relative to the centromere differs. This strategy takes advantage of the acrocentric nature of mouse chromosomes. Two probes, one close to the centromere (proximal) and the other one close to the opposite end (distal), are labeled with different colors (for example, red and green). Using the centromere as a landmark, one chromosome could be identified by the pattern proximal-green + distal-red while a second chromosome could be distinguished by the proximal-red + distal-green pattern. This approach allows the combination of as few as three colors, either singly (green, red or blue) or in pairs (green/red, green/blue), to generate 25 different chromosome labeling patterns. These patterns are sufficient to distinguish all 19 mouse autosomes and the X chromosome, while a remaining red/blue combination can be reserved for mapping an unknown locus. We identify the Y chromosome by exclusion, as the chromosome carrying no reference probes in a metaphase. Our technique can be performed with a fluorescence microscope capable of detecting fluorophores in the visible spectrum (usually red, green, and blue, as commonly used for immunofluorescence techniques) and consumer graphics software (Photoshop) for image merging. With this common equipment, the usable fluors can be fluorescein isothiocyanate (FITC, green), rhodamine or Texas Red (red), and aminomethyl coumarin (AMCA, blue). While this study used a cooled CCD camera to photograph fluorescence microscope images, we have previously shown that even digital photographic cameras can be used to detect such fluorescent signals in the visible spectrum (Henegariu et al. 1999). The cost of this technique can also be minimized by the use of a previously described method for inexpensive nucleotide labeling (Henegariu et al. 2000) by a factor of 100- to 200-fold when compared with the use of commercial labeled nucleotides. With PCR labeling, there are about 12 separate labeling reactions needed at the same time (a maximum of four to different probes are labeled in the same reaction tube), and a total of about 200 ml of labeled PCR product is used for one hybridization. DNA sufficient for 40 tests can be labeled in one 96-well PCR plate. The total cost of the reagents per single analysis is about $0.20 when using custom-made nucleotides. Alternatively, nick-translation can be used for labeling (data not shown). It has the advantage that it requires only three labeling reactions (one for each fluor used), and the amount of DNA that can be labeled at one time is unlimited. The drawback is that new BAC DNA for the reporter probes has to be isolated periodically. Whichever the labeling approach, little optimization is required for the labeling protocol because all the probes are of a similar large size.

The labeling strategy is represented in Fig. 1 (iv). We used FITC-dUTP for green,carboxyrhodamine 6G-dUTP (R6G-dUTP) for red, and biotin-dUTP (BIO-dUTP) detected with avidin AMCA for blue. As AMCA and 48,6-diamidino-2-phenylindole (DAPI) have almost identical absorption and emission spectra, they cannot be used simultaneously. Therefore, after adding the avidin-AMCA, the slide was mounted with antifade without DAPI counterstaining. Images of several metaphases were captured and their coordinates recorded, by using the verniers of the microscope stage. As the overall fluorescence on the slide is strong, it was very easy to locate metaphases using the 40x or 60x objective [Fig. 1 (i) and (vi)]. As previously described (Henegariu et al. 1999, 2001), the cover slip was removed, the slides were counterstained with DAPI, mounted again in antifade, and images of the same metaphases were recaptured with the DAPI filter. For the purpose of this report, DAPI counterstaining is only used to easily identify the centromeric end of the chromosome [Fig. 1 (ii)]. However, DAPI can be avoided altogether: because of background fluorescence, chromosomes and nuclei are visible when overexposing the fluorescence image through the blue or green filters. Thus, an image showing the shape of all chromosomes can be simply recorded by overexposing the metaphase image 5- to 10-fold longer, using either the DAPI-filter or FITC-filter. This usually shows the position of the centromeric condensation. This overexposing step is done after the FISH signals are recorded through each of the three filters, to avoid photobleaching the specific signals. As a parallel method of confirming the centromeric end, the background image can be overlapped (in the merged image) with one, two, or all three other images captured. The centromeric BACs will show signals within the chromosomal mass as they are distal to the centromere, whereas the distal BACs will show signals at the very end or even the outside of the chromosomal shape.

Figure 1 shows representative results of a hybridization using a batch of labeled probes. We found that, while the PCR-labeled BAC clones gave very robust signals, the results could be further improved with a secondary detection by using antibodies to amplify all primary signals. We used goat anti-FITC and donkey anti-goat-FITC to amplify the signal from the FITC-labeled DNA and rabbit anti-R6G-CY3 to amplify the R6G signal. As previously mentioned, the red/blue color combination was not used in the reference BAC panel, but was reserved for labeling probes that required mapping. To test the sensitivity of the procedure, three different probes, annotated as A-C in Fig 1(i), were mapped at the same time (magenta signals). Test probe A is readily apparent, located in an interstitial region of Chr 14, physically separated from the reference probes at both ends of the chromosome. Test probe B is closer to the centromeric reference probe for that chromosome, but the hybridization foci are clearly distinct. Test probe C is very close to the centromeric end of the respective chromosome, and identifies a transgene integrated into the genome of that mouse. The mouse was heterozygous for this transgene, accounting for its presence in only one of the two Chr 5 shown. Even when the probe to be mapped is very close to a reference probe, the color combination algorithm used always picks up the position of the probe to be mapped.

Although our technique is primarily designed for simple mapping in laboratories not expert in mouse cytogenetics, the reference BAC probe set lends itself to more advanced applications also. With a microscope equipped with four filters, DAPI counterstaining can be added to the analysis, replacing AMCA with another fluorophore (for example, CY5). If a fluorescence microscope equipped with six to eight filters is used (Henegariu et al. 2000), addition of further colors allows multiple mapping studies to be performed in a single hybridization. Figure 1 (vi) shows an example of such an experiment. In this case, probe A was labeled with FITC-dUTP and dinitrophenol (DNP)-dUTP, probe B with digoxigenin (DIG)-dUTP, and probe D with CY3-dUTP (detected in the same channel as the R6G signals). The DNP was detected with rat anti-DNP-CY3.5 and the DIG with sheep anti-digoxigenin- DEAC, followed by donkey anti-sheep-DEAC. Three unknown probes were mapped in the one experiment shown, al-though many more could be mapped by the combinatorial use of extra colors.

This triple-color strategy is not optimal for detecting translocations, as a translocation may change the pattern of similarly labeled reference probes and thus escape detection. However, by performing sequential three-color hybridizations or by increasing the number of fluorescent colors used to five, the same reference panel and techniques can be used for this purpose also. The five-color approach gives each chromosome a unique fluorescent signature (manuscript in preparation) and allows detection of most balanced and unbalanced translocations. In summary, this BAC panel is a versatile resource for both simple and complex mouse cytogenetic applications, which can be made very cost-efficient by using the techniques we have described here and previously.

Acknowledgments.

This work was supported by DOE grant DE-FC03-96ER62294 and NIH/NICHD grants P01 HD17449 and PO#600641 to J.R.Korenberg; NIH grant HG00272 to D.C. Ward; and NIH/NIDDK grants DK02467 and DK45676 to J.M. Greally. J.R. Korenberg holds the Geri and Richard Brawerman Chair in Molecular Genetics. We thank Drs. Eric Bouhassira, Frank Ruddle, and Richard Lifton for mice and test clones.

Alami R, Greally JM, Tanimoto K, Hwang S, Feng Y-Q et al. (2000) b-Globin YAC transgenes exhibit uniform expression levels but position effect variegation in mice. Hum Mol Genet 9, 631-636

Boyle A, Ballard G, Ward D (1990) Differential distribution of long and short interspersed element sequences in the mouse genome: chromosome karyotyping by fluorescence in situ hybridisation. Proc Nat Acad Sci USA 87, 7757-7761

Henegariu O, Heerema NA, Bray-Ward P, Ward DC (1999) Colour-changing karyotyping: an alternative to M-FISH/SKY. Nat Genet 23, 263-264

Henegariu O, Bray-Ward P, Ward DC (2000) Custom fluorescent-nucleotide synthesis as an alternative method for nucleic acid labeling. Nat Biotechnol 18, 345-348

Korenberg JR, Chen XN, Devon KL, Noya D, Oster-Granite ML, Birren BW (1999) Mouse molecular cytogenetic resource: 157 BACs link the chromosomal and genetic maps. Genome Res 9, 514-523

Longo L, Bygrave A, Grosveld F, Pandolfi P (1997) The chromosome make-up of mouse embryonic stem cells is predictive of somatic and germ cell chimaerism. Transgenic Res 6, 321-328

Ried T, Schrock E, Ning Y, Wienberg J (1998) Chromosome painting: a useful art. Hum Mol Genet 7, 1619-1626

Speicher MR, Gwyn Ballard S, Ward DC (1996) Karyotyping human chromosomes by combinatorial multi-fluor FISH. Nat Genet 12, 368-375

Su H, Wang X, Bradley A (2000) Nested chromosomal deletions induced with retroviral vectors in mice. Nat Genet 24, 92-95