Troubleshooting

Fish Guide Selected References

1. Boggs, B. A., and Chinault, A. C., 1997, Analysis of DNA replication by fluorescence in situ hybridization,??Methods??13(3):259-70.

2. Chaffanet, M., Baens, M., Aerssens, J., Schoenmakers, E., Cassiman, J. J., and Marynen, P., 1995, Mapping of an ordered set of 14 cosmids to human chromosome 12p by two- color in situ hybridization,??Cytogenet Cell Genet??69(1-2):27-32.

3. Chang, S. S., and Mark, H. F., 1997, Emerging molecular cytogenetic technologies,??Cytobios90(360):7-22.

4. Chowdhary, B. P., de la Sena, C., Harbitz, I., Eriksson, L., and Gustavsson, I., 1995, FISH on metaphase and interphase chromosomes demonstrates the physical order of the genes for GPI, CRC, and LIPE in pigs,??Cytogenet Cell Genet??71(2):175-8.

5. de Meulemeester, M., Vink, A., Jakobs, M., Hermsen, M., Steenman, M., Slater, R., Dietrich, A., and Mannens, M., 1996, The application of microwave denaturation in comparative genomic hybridization,??Genet Anal??13(5):129-33.

6. du Manoir, S., Schrock, E., Bentz, M., Speicher, M. R., Joos, S., Ried, T., Lichter, P., and Cremer, T., 1995, Quantitative analysis of comparative genomic hybridization,??Cytometry19(1):27-41.

7. Flejter, W. L., Barcroft, C. L., Guo, S. W., Lynch, E. D., Boehnke, M., Chandrasekharappa, S., Hayes, S., Collins, F. S., Weber, B. L., and Glover, T. W., 1993, Multicolor FISH mapping with Alu-PCR-amplified YAC clone DNA determines the order of markers in the BRCA1 region on chromosome 17q12-q21,??Genomics??17(3):624-31.

8. 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

9. Henegariu O, Heerema NA, Bray-Ward P, Ward DC (1999). Colour-changing karyotyping: an alternative to M-FISH/SKY [letter]. Nat Genet??23:263-4

10. Henegariu O, Heerema NA, Wright LL, Bray-Ward P, Ward DC, Vance GH (2001). Improvements in cytogenetic slide preparation: controlled chromosome spreading, chemical aging and gradual denaturing. Cytometry??43: 101-109

11. Heng, H. H., and Tsui, L. C., 1993, Modes of DAPI banding and simultaneous in situ hybridization,??Chromosoma??102(5):325-32.

Heng, H. H., and Tsui, L. C., 1994, FISH detection on DAPI-banded chromosomes,??Methods Mol Biol??33:35-49.

12. Kallioniemi, O. P., Kallioniemi, A., Piper, J., Isola, J., Waldman, F. M., Gray, J. W., and Pinkel, D., 1994, Optimizing comparative genomic hybridization for analysis of DNA sequence copy number changes in solid tumors,??Genes Chromosomes Cancer??10(4):231-43.

13. Karhu, R., Kahkonen, M., Kuukasjarvi, T., Pennanen, S., Tirkkonen, M., and Kallioniemi, O., 1997, Quality control of CGH: impact of metaphase chromosomes and the dynamic range of hybridization,??Cytometry??28(3):198-205.

14. Kuukasjarvi, T., Tanner, M., Pennanen, S., Karhu, R., Visakorpi, T., and Isola, J., 1997, Optimizing DOP-PCR for universal amplification of small DNA samples in comparative genomic hybridization,??Genes Chromosomes Cancer??18(2):94-101.

15. Laan, M., Isosomppi, J., Klockars, T., Peltonen, L., and Palotie, A., 1996, Utilization of FISH in positional cloning: an example on 13q22,??Genome Res??6(10):1002-12.

16. Lundsteen, C., Maahr, J., Christensen, B., Bryndorf, T., Bentz, M., Lichter, P., and Gerdes, T., 1995, Image analysis in comparative genomic hybridization,??Cytometry??19(1):42-50.

17. Monier, K., Usson, Y., Mongelard, F., Szepetowski, P., Robert-Nicoud, M., and Vourch, C., 1996, Metaphase and interphase mapping by FISH: improvement of chromosome banding and signal resolution in interphase nuclei by means of iterative deconvolution,??Cytogenet Cell Genet72(2-3):200-4.

18. Moscone, E. A., Matzke, M. A., and Matzke, A. J., 1996, The use of combined FISH/GISH in conjunction with DAPI counterstaining to identify chromosomes containing transgene inserts in amphidiploid tobacco [corrected and republished with original paging, article originally printed in Chromosoma 1996 Oct;105(4):231-6],??Chromosoma??105(5):231-6.

19. Rao, P. H., Cigudosa, J. C., Ning, Y., Calasanz, M. J., Iida, S., Tagawa, S., Michaeli, J., Klein, B., Dalla-Favera, R., Jhanwar, S. C., Ried, T., and Chaganti, R. S., 1998, Multicolor spectral karyotyping identifies new recurring breakpoints and translocations in multiple myeloma,Blood??92(5):1743-8.

20. Roy, L., Sorokine-Durm, I., and Voisin, P., 1996, Comparison between fluorescence in situ hybridization and conventional cytogenetics for dicentric scoring: a first-step validation for the use of FISH in biological dosimetry,??Int J Radiat Biol??70(6):665-9.

21. Schrock, E., du Manoir, S., Veldman, T., Schoell, B., Wienberg, J., Ferguson-Smith, M. A., Ning, Y., Ledbetter, D. H., Bar-Am, I., Soenksen, D., Garini, Y., and Ried, T., 1996, Multicolor spectral karyotyping of human chromosomes [see comments],??Science??273(5274):494-7.

22. Shen, S. X., Weaver, Z., Xu, X., Li, C., Weinstein, M., Chen, L., Guan, X. Y., Ried, T., and Deng, C. X., 1998, A targeted disruption of the murine Brca1 gene causes gamma-irradiation hypersensitivity and genetic instability [In Process Citation],??Oncogene??17(24):3115-24.

23. Speicher, M. R., Gwyn Ballard, S., and Ward, D. C., 1996, Karyotyping human chromosomes by combinatorial multi-fluor FISH,??Nat Genet??12(4):368-75.

24. Telenius H, Carter NP, Bebb CE, Nordenskjold M, Ponder BA, Tunnacliffe A (1992). Degenerate oligonucleotide-primed PCR: general amplification of target DNA by a single degenerate primer. Genomics 13:718-723

25. Trask, B. J., Massa, H., Kenwrick, S., and Gitschier, J., 1991, Mapping of human chromosome Xq28 by two-color fluorescence in situ hybridization of DNA sequences to interphase cell nuclei,??Am J Hum Genet??48(1):1-15.

Multicolor Immunostaining

(Multicolor) fluorescent immunostaining??guidlines

Slide preparation, cell fixation, permeabilization and antibody incubations are performed similar to the “FISH + immunostaining” page, but omitting the denaturing and hybridization.??
– After placing the cells on slides (for example using a cytospin or simply by growing them on slides or coverslips), they are fixed in methanol or ethanol, either at room temperature for 1 hour, or by heating them 1 minute at 65 C in ethanol (chemical aging).??
– Cell permeabilization can be done by 1-5 minute incubation in a solution containing almost any detergent (SDS, saponin, Tween 20, NP-40) at concentrations of 0.05%-0.2% (lower for SDS, higher for the other !!) in PBS or other isotonic buffer.??
– All antibodies were diluted in??4xSSC??at??1:100 to 1:500 dilutions of 0.5-1 mg/ml stocks. All antibody incubations were done??10 minutes at 37 C.
– Between any primary and secondary antibody incubation, the slide is rinsed three times 5 minutes each, in a jar with 4xSSC or with 4xSSC/0.1% Tween20 (which decreases somewhat the background, and does not affect binding of the IgG antibody)

For multicolor immunostaining, as for multicolor FISH, one has to pay special attention to??potential cross-reactivity??among antibodies from different species. Therefore, antibodies should be added in a “logical” step-wise approach, to prevent these nteractions from taking place. The use of colors depends on the filter availability in your fluorescence microscope. With a good filter set, up to??eight different fluorophores??can be simultaneously discriminated on the same preparation. A more thorough discussion on fluorophores and colors is presented in the??multicolor FISH,??CCK??andnucleotide labeling??sections.

A few??examples??of immunostaining are shown in the figure below.All nuclei were stained blue (DAPI). The two images on the left show the distribution of the actin fibers in wild-type and mutant fibroblasts (mutation for a gene involved in phoshphorylation). The figure at the bottom right corner, shows the distribution of vimentin (part of the desmosomes) at the junctions between cells. The large image shows part of a fibroblasts in which the actin fibers were detected in red, whereas vinculin (part of the focal adhesions) was detected green.


Fish Immunostaining

This procedure was used for the simultaneous hybridization of a human centromeric probe (FISH) and detection of the centrioles (immunostaining). Hybridization was first, followed by antibody detection for both the FISH probe and gamma-tubulin. A brief description of the procedure was also published in: Mantel et al. 1999 Blood 93(4):1390-1398.

Note:??duration of antibody incubation (1:100 to 1:500 dilutions of a 1mg/ml stock) for both FISH and immunostaining is complete after??10 minutes at 37 C. Lomger incubations are not necessary. All antibodies were diluted in 4xSSC prior to placing them on the slide.

Slide preparation and fixation

Cells grown in culture, were placed on the slide using a cytospin. To permeabilize the cell membranes, slides were incubated in a 0.05% SDS in PBS solution for 2 1/2 minutes at room temperature. After incubation, slides were rinsed well in PBS, then passed for a few seconds each in 70% ethanol (to wash off salts) and 100 ethanol (dehydration) and then placed for a few seconds on a 75 C metal plate for quick drying.

Cells were fixed using the??“chemical aging”??approach, by keeping the slides for 30 seconds in ethanol at 95 C. After fixation, slides were air dried at room temperature.

Hybridization and detection

A digoxigenin-labeled DNA probe specific for the centromeric region of human chromosome 7 (Oncor, Inc) was mixed in Hybrisol VII (Oncor, Inc.) and was denatured 5 minutes at 75 C according to manufacturer???s protocol. The slides were denatured by placing them for 2 minutes in a 70% formamide/2x SSC solution at 75 C, followed by successive washes in 70%, 90% and 100% ethanol at room temperature.

Denatured centromeric probe was placed on slide, covered with a coverslip, sealed with rubber cement and incubated for 30 minutes to 2 hours at 37 C in a moist chamber.

After hybridization, slides were washed 3×5 minutes each in 50% formamide/2x SSC and 1xSSC solutions at 45 C.

For double staining, slides were incubated with a combination of antibodies against digoxigenin and gamma-tubulin as follows:

  • in the first detection step, sheep-antidigoxigenin-FITC (Boehringer) and mouse-antitubulin (Sigma) were mixed together at 1:100 dilution each in a 4xSCC solution. 100 ul of this antibody solution was placed on each slide, covered with plastic coverslips and incubated 15 minutes at 37 C in a moist chamber. After incubation, the coverslips were removed and the slides were washed 3×5 minutes in 4xSSC/0.1% Tween 20 at 45 C.
  • in the second step, a horse antimouse-TexasRed (Vector Laboratories) was used at 1:100 dilution in 4xSSC to detect the mouse antitubulin antibody. After 15 minutes incubation, slides were rinsed again 3×5 minutes in 4xSSC/0.1% Tween 20 at 45 C, and then briefly rinsed in distilled water to wash off the salt. Slides were air-dried at room temperature.
  • Slides were covered with a DAPI antifade solution, covered with coverslips and examined under a fluorescent microscope (Leica Aristoplan) equipped with a cooled CDD camera (Photometrics) and image acquisition and processing software (Vysis, Inc) using a set of appropriate filters (Chroma Technologies).

An??example??of this procedure is depicted in the figure below (normal cells). A DIG-labeled probe for the human chromosome 7 centromere was hybridized for 2 hours. Posthybridization washes were followed by the antibody detection step, using an antidigoxigenin antibody labeled with FITC (green) and an primari antibody against gammatubulin. A secondary antibody labeled with TexasRed (red) was used to detect the centrosome signals.

Fish On Paraffin Embedded Tissue Pet

1. Using DNA probes

  • In our??experience, probably the most important step for FISH on PET was??slide pretreatment with proteases. Paraffin embedded material required much higher concentrations of??pepsin(0.4-5%) compared to 0.005% for regular cytogenetic preparations. Protease digestion can be done between 2 and 20 minutes, depending on how the tissue was fixed. Hybridizaton results are influenced by the original fixation procedure, and some paraffin sections may not hybridize well.

  • Protocol:??alternatively, and probably with even better success, one can use??Proteinase K??to treat the deparaffinized sections prior to hybridizatoin. Briefly, the slide with the paraffin section is placed in a jar with xylene for 10-15 minutes, then in a jar with 1:1 xylene:ethanol for 10 minutes, followed by incubation only in 100% ethanol for 10 minutes. A ProteinaseK digestion solution is prepared by mixing 20ml PBS + 100??l 10%SDS + 200??l 20mg/ml Proteinase K. We do the digestions in a small plastic slide holder, which accomodates 20 ml liquid. Other jars can be used as well. The jar with the proteinse solution is kept in a waterbath at 45-50 C. The deparaffinated slides are incubated in the proteinase K solution 8-25 minutes,depending on how the paraffinization and fixation was done. The optimal incubation time needs to be identified by parallel experiments. After digestion, the slide is rinsed 1-3 minutes in PBS, followd by 5 minutes incubations in 70% and 100% ethanol. No further fixation is needed. Slides can be denatured and used.??This protocol worked equally well for DNA and RNA probes !!.
  • Hybridizations on PET also benefit from??longer denaturing times??(5-6 minutes) and

  • higher denaturing temperatures??(up to 82-85 C).Denaturing is done in 70%FA/2xSSC.

  • Another parameter is the??thickness of the section. Whereas for RNA hybridizations 2-3 ??m sections were ideal, DNA hybridizations worked best on 10??m thick sections. This correlates with the RNA being found primarily in the cytoplasm, where it is readily available to hybridizations, whereas the DNA is in the nucleus. Sections that are too thin may include only thin areas/slices of the nuclei, which may lack the targeted sequences and thus will provide false negative results.

In FISH on PET it is very important that the??DNA probe used is cut??in small fragments (200bp average). Too long DNA fragments will not penetrate the tissue and will not reach the target DNA.Posthybridization washes??should be gentle, as the probe fragments are small and the tissue is “hard” to hybridize onto. Too stringent washes may wash away specific signals.

An example of triple color FISH with DNA painting probes on PET is provided in??Fig.11l(courtesy of Dr. Ronald Honchel) where three 12p specific paint probes were used on a PET section of a germ cell tumor tissue. Many nuclei show more than two groups of signals, indicating an increase in 12p copy-number , a common feature of germ cell tumors.

2. Using RNA probes (to target gene expression)

Slide preparation and preatreatment protocol is the same as above. We used??riboprobes??anddouble-stranded DNA probes??(gene exons generated and labeled by PCR) with equal success. However, it is important to mention that the majority of RNA signals were too weak to be visualized using fluor-dUTP or fluor-UTP labeled probes. Instead, we labeled the riboprobes or DNA probes with??haptenes??(BIO-, DIG-, DNP-dUTPs or -UTPs) and detected them with fluor-labeled antibodies. This approach was always more efficient. Also successful was the use of antibodies against fluorophores (antibodies against fluorescein and rhodamine derivatives can be used to amplify the signals as well).

An example of RNA in situ detection is shown in the??figure??below.??(1)??RNA detection.??DIG-labeled riboprobes of the rat surfactant protein B (SPB) were detected red (upper left image) in the type II pneumocytes in the lung alveoli. On a different slide (upper right image), SPB expression was also detected (green) in the cells lining the lumen of small bronchioli.??(2)??RNA and DNA detection. Using a mouse chromosome Y-specific probe (labeled red) and two small DNA probes (labeled green) for three exons of the mouse gene for albumin, we showed that a female mouse, recipient of a bone marrow transplant from a mouse male donor, carried male cells fully differentiated into hepatocytes. In all images, the DNA was counterstained with DAPI (blue). Thered arrow??shows the Y-specific centromere signal (red) in one of the nuclei, whereas the??green arrows??show the abundent hybridization of the two exon-type DNA sequences (each ~100 bp long ) in the cytoplasm of most hepatocytes, and also in their nuclei (the nuclear signals are not from the genomic DNA sequence of the albumin gene, but rather from the mRNA transcription centers). In multinuclear hepatocytes, 4-8 such strong nuclear green signals were clearly visible. The??short blue arrow??points to a nucleus which contains both the Y-signal and the albumin transcription centers, and is surrounded by abundedent mRNA signal in the cytoplasm.

??

Dna Fibers

Length measurements of DNA probes.

DNA stretching used in this laboratory is a 15-20 minutes procedure, which can be used to calculate the length of probes from 10 kb to hundreds of kilobases. Briefly, isolated DNA of any probe is diluted to 1-5ng/ul in 10 mM AMP solution, pH-8.2-8.4. From this solution, 6.5-7 ul are pipetted on one end of a silanized slide (Sigma). The end of a 24×40 coverslip is placed at an angle, close to the DNA drop on the slide, so it touches the slide and then, in a quick move, that end of the coverslip is pressed against the slide with the thumb or a clamp. This makes the fluid spread between the slide and coverslip, from one end of the coverslip to the other. The flow of the solution is the driving force stretching the DNA. After 30-60 seconds the coverslip is removed. The slide is allowed to air -dry, and is placed for a minute each in jars with 70% and 100 % ethanol (fixation). The slide is covered with antifade/mounting medium containing 10-7 YOYO stain (Molecular Probes), then kept 3-5 minutes in the dark and examined at the microscope. Slide can be visualized and images taken using a regular green filter (for example FITC filter). Calculations indicated that, at least in our hands, the DNA is stretched to about 130% of its theoretical value, which is 34 uM for 1000bp (1kb). Details about the procedure will be further added after the publication of the submitted manuscript.

??

FISH on stretched DNA.

FISH on DNA fibers is useful in assessing the length of DNA probes, and to map probes relative to one another (Fig. 10j), as it can reveal even their degree of overlap. Thus, fiber FISH has superior mapping resolution compared to interphase FISH. There are many protocols for??fiber-FISH??orhalo-FISH, and a discussion of all is beyond the purpose of this report, but several useful observations will be briefly mentioned. As a hybridization example, the figure below shows the extensive overlap between two PAC clones (red and green labeled, left side). On the right side, four cosmids (spanning a physical region of about 220kb) show a small overlapping area between the two clones (red and green) in the middle. A cosmid has an average size of about 35-40kb, whereas a PAC is usually 100-150 kb.

Why does hybridization resemble “beads on a string”?

If stretched DNA is stained with YOYO and observed through the FITC filter at the fluorescence microscope, immediately after exposure to light, the DNA fibers start breaking at many locations. From each break, the two free ends of DNA??spring back to the nearest attachment point and coil, leaving a gap between them. These ends resemble little beads, and are thicker than the rest of the fiber. It is well known that when FISH is performed on stretched DNA/DNA fibers, hybridization always looks like an array of signals, or like “beads on a string” (Fig. 10j). The length of the array of dots is proportional to the physical length of the DNA probe. To determine precisely where the labeled DNA fragments bind on the DNA fibers, and where the “beads” are located, a simple experiment was done, in which a labeled probe was detected with a red fluorophore and the fibers were counterstained with a green fluorescent dye. We tested both YOYO (binds double-stranded DNA) and SYBR-Green II (binds single and double stranded DNA). In such experiments, performed with repeated or with single copy probes, we noticed that every red “dot” of the hybridized probe was also stained with green from either one of the counterstain dyes, but not every green dot was accompanied by a hybridization signal. The larger the hybridization signal the fainter the green color and vice-versa. When the DNA denatures, physical constraints within the double helix should allow some areas to become single stranded but at the expense of other regions, which should become overcoiled. Graphically, this would look similar to a chain of replication bubbles linked together. A double stranded DNA stain like YOYO should provide an image similar to beads on a string,??but the fiber should be contiguous !!. Thus, SYBR-green II should show a contiguous signal for any DNA fiber. Experiments show that both dyes produce the same staining pattern, and that the DNA is broken during denaturing.

These observations suggest the following??hybridization scenario??in fiber FISH: the stretched DNA, when subjected to denaturing conditions (70%FA/2xSSC) at 75?? to 95?? C, actually breaks at many points. At each break, the two loose ends snap back towards the nearest attachment point and coil in a “dot-like” structure. The number and density of these dots depends on how many attachment points the initial DNA fiber has had to the slide, which, in turn, depends on the fixation procedure and the denaturing conditions. It is the DNA in the “dot-like” structures where denaturing takes place and where the labeled DNA probe binds. The stronger the hybridization, the more denaturing took place, and the less intense the YOYO signal. This explains easily the beads-on-a-string hybridization pattern.

??

Variability of signal-array length

Techniques for stretching genomic DNA, often expose the cellular material to detergents, chemicals or enzymes to release the DNA, immediately before stretching. The reason is the necessity to produce contiguous, long DNA fibers, important for FISH mapping. If fiber FISH is performed on such preparations, the arrays of signals obtained from the same probe may have different length. This is due to the variable degree of stretching of the genomic DNA, which may be only partial partially released from its chromatin structure. For a better assessment of the size of any probe, the longest array of hybridization signals should be the closest ot the real size.

Alternatively, genomic DNA can be freed from proteins through various treatment steps, while trapped in low melting point agarose blocks, which prevents the DNA fibers from breaking. The agarose is melted and DNA released right before use.

??

Tips for improving fiber-FISH hybridizations

  • Fixation??increases the number of attachment points between the DNA fiber and the slide surface. Chemical aging, 10-30 seconds at 94?? C in 100% ethanol provides an easy and convenient way to fix the fibers. Alternatively, a simple incubation of the slides, for 1-2 minutes each, in 70% and 100% ethanol at room temperature may provide sufficient fixation. Baking the slides is also an alternative approach.
  • Both simultaneous and separate??denaturing??protocols work for fiber FISH. Denaturing in a 70% FA/2xSSC solution works well, and higher temperatures (95?? C) appear to denature the stretched DNA better than lower temperatures (70-75?? C). A sudden chilling of the slide after the high denaturing temperature also improves the signals.
  • For optimal FISH signals, the??amount of labeled probe??should be about twice or three times higher in fiber FISH compared with regular FISH.
  • Blocking the repetitive sequences is also very important, otherwise hybridization signals will not be identifiable from all other background signals. Thus, a higher than usual (at least double)??Cot1 amount??should be used, compared to regular FISH.
  • For the same reason, very??stringent posthybridization washes??should be used, to remove non-specifically bound, repetitive fragments. The charged slide surface on which the DNA is stretched may have higher affinity for some fluorescent dyes. To decrease the nonspecific binding of a dye or of an antibody to the slide surface, a 10 minute incubation of the slide in 0.1-1mg/ml??BSA or other blocking reagent??can be used.

Cost Reduction Of Fish With Commercial Probes

The “lower stringency” approach

Although commercial probes are provided as kits with ready-to-use solutions, they should not be regarded as unchangeable entities. All DNA probes behave like any other DNA probes, and protocols can be modified and improved. Moreover, probes from different kits can be combined.

For example, a cosmid or BAC probe (usually washed in 0.2xSSC at 65?? C post- hybridization) can be combined with a painting probe (usually washed in 2xSSC at 37?? C). In such cases, the second posthybridization wash (in SSC) will be performed at the temperature used for the painting probe (less stringent wash). If background from the BAC is a concern, 10-15 ??g competitor DNA can be added to the cosmid-paint probe mixture prior to hybridization, followed by ethanol precipitation, resuspension in hybridization buffer, denaturing and hybridization. The extra Cot1-DNA will block repetitive sequences better.

??

The “low amount of probe” approach.

Commercial probes kits provide sufficiently large amounts of labeled probe, so as to allow hybridization even in less optimal conditions. If the FISH procedure is efficiently, DNA probes purchased in kits can be used for 3-5 times more hybridizations than indicated, thus reducing the cost for each hybridization. This was tested in this laboratory numerous times. For example, in??Fig. 6, three different commercial probe kits were used. However, for each hybridization 1/4 or 1/5 of the recommended amount of probe was used (2 ul of the provided probe was mixed with 8-9 ul hybridization buffer). Simultaneous or separate denaturing worked just as well. Moreover, as shown in Fig. 6, good hybridization results were obtained after only 3-5 hours hybridization. Overnight hybridizations did not improve the signals in any ways and did not seem necessary. This shows that, even the reduced amount of each probe used, could be further decreased.

??

The “replacement” approach

If fluorescence-antibodies are necessary, they can be purchsed either non-labeled or labeled (but in higher amounts) from other vendors. Non-labeled antibodies can be easily labeled in any laboratory. Besides, the fluorophore used for any probe detection can be replaced with any other one at will, the only limitation being the number of filters and the quality of the filters with which a fluorescence microscope is equipped. A summary of some of the fluorophores used in this laboratory is presented??here??or??here??(Please note that this list, although it covers the major groups of fluorophores, is by no means complete. There are many other fluors available comercially).

??

Mixing of multiple commercial and custom probes

When purchasing them from vendors, commercial probes usually arrive in hybridization buffer, denatured, re-annealed and combined with competitor DNA. If probes from three or four kits or sources are to be used together,??in the same hybridization, the final volume may be higher than the 10-12 ??l accomodated under a standard 22×22 mm coverslip. To solve this problem, all probes can be co-precipitated together. That the probes may have been previously denatured has no importance. For example, if the total volume of the mixed probes is 25-30 ??l, they can be resuspended in 5-10 volumes of TE or water (to approximately 200 ??l) followed by ethanol precipitation. If desired, 15-20 ??l competitor DNA can also be added. The pellet is resuspended in 10-12 ??l hybridization buffer and used. An example of this approach is shown in??Fig. 10d. A commercial, dual-color probe (red=rhodamine and green=FITC) was mixed and used together with a biotinylated commercial paint probe for a chromosome in group F. As recommended by the vendors, 10 ??l of each probe were mixed in the same vial, 20 ??g Cot 1 DNA was added, and the DNA was precipitated, resuspended in hybridization buffer and hybridized. The biotinylated probe was detected with a third color (avidin Cy5, available in our laboratory), and the chromosomes were counterstained with DAPI. All three probes could be visualized, although they were not used according to the protocols provided by the two vendors.

Fish With Complex Probes

General observations

For probes in this category (microdissected probes, whole chromosome paint probes and whole genomic probes) one of the most important parts of the FISH protocol is slide preparation. Cell suspensions used for slide preparation need to be as free of cytoplasmic residua as possible. This requires optimum hypotonic treatment and proper fixation. Metaphases should be spread well enough so that chromosomes do not cross each other, but spreading should not be excessive, so the entire metaphase can fit in a single microscope field (100x magnification). Slide preparation is detailed elsewhere. Chemical aging offers very good results with all complex FISH probes, and dry heat aging of the slide should be avoided. Slides should be prepared fresh and used immediately for hybridization. Slide pretreatment using a protease (usually pepsin) always improved hybridization results. If a better DAPI banding is required (for example in CGH), immediately after the enzymatic pretreatment, the slide can be kept no longer than 15-20 seconds in 1% formaldehyde in isotonic buffer, and then passed through an ethanol series as usual. Too long formaldehyde incubation decreases hybridization. Separate and simultaneous denaturing provide identical results.

Complex probes are conveniently labeled by??PCR using degenerate primers. As usual, labeled fragments should be cut to 200-300bp average length before hybridization.??Nick translation??can also be used for labeling. Complex probes can be mixed with one another, with single-copy probes or with repetitive probes, depending on the purpose. For example, microdissected probes for chromosome 12p, 12q and the 12 centromere probe (Fig. 10k) and microdissected probes for the three bands of 12p were hybridized and detected in triple color FISH (Fig. 10l).

??

Hybridization time

When using chemical aging, hybridization does not need to proceed longer than overnight (14-16 hours) no matter what probes are used. In our hands, CGH results were identically good after 16 (one day) and 36 hours (2 days) but worse after 70 hours (3 days). M-FISH results were very good after 14-16 hours hybridization. Shorter times were not tested.

??

Comparative genomic hybridization (CGH)

CGH principle??(see figure below). CGH is used to detect differences in DNA copy number between a test sample (for example tumor DNA) and a control sample (normal genomic DNA). The tumor DNA is labeled with a red fluorophore and the normal DNA with a green fluorophore (or vice-versa), and the two samples are mixed together, followed by overnight hybridizen ontonormal??metaphases. After hybridization and DAPI staining, a computer program is used to calculate the ratio of fluorescence between the red and the green colors along the axis of every chromosome. Whenever the tumor DNA shows DNA amplification in some chromosomal areas, there will be more red-labeled copies of DNA hybridizing for thaose regions, so the hybridization color (and the graph) will shift toward red. Whenever the tumor DNA displays loss of DNA for some chromosomal areas, there will be proportionally more green-labeled fragments hybridizing on those regions, thus the fluorescent color (and graph of ratio) will shift toward green. CGH is particularly important, as it allows to test DNA from archived samples (paraffin embedded material), and does not require cell culture of every test sample.

The quality of slides??used for CGH is vital, thus the cell suspension should always be tested. A good cell suspension can be used for over a year, if stored at -20?? C in fixative.??DAPI banding??can be improved by:??(1)??10-30 seconds incubation in 1% formaldehyde after pretreatment.??(2)??gradual slide denaturing, using either the temperature gradient of a thermocycler block or several jars with denaturing solution at various temperatures.??(3)??DAPI staining in a jar with a DAPI solution, and not by adding DAPI to the antifade solution. DNA labeling can be done by nick translation or PCR (with a degenerate primer). Nick translated probes require, in general, a somewhat larger amount of competitor DNA to compete out repetitive sequences.??Fig. 11a??shows CGH results when an insufficient amount of competitor DNA was used. The centromeres of almost all chromosomes show bright signals, indicating that repetitive sequences were not competed out. Hybridization of repetitive sequences prevents proper CGH analysis, as they smooth out the differences/ratios between the two colors. In??Fig.11b, the DNA was labeled by PCR. Although the DNA fragments were digested by DNase to below 500bp, hybridization was more “grainy”. In both Fig. 11a and 11b, posthybridization washing conditions were very stringent (20 minutes at 65?? C in 0.2% SSC). Our experiments showed that complex probes do not require high stringency washes. Because repetitive sequences may become a concern, their binding can be avoided by increasing the amount of Cot1 DNA during hybridization. If the hybridization signals along the chromosomes are not smooth, the ratio analysis between the two colors will be less accurate. A good CGH hybridization is depicted in??Fig. 11c.

CGH can be performed with DNA probes labeled with fluorescent nucleotides (FITC and TRITC,Fig. 11d-f) or haptene-labeled nucleotides (biotin and digoxigenin,??Fig. 11g-i). Both procedures work equally well if labeling is good, DNA fragments are small and slides are properly prepared.

??

Multicolor karyotyping (M-FISH, SKY and CCK)

Multicolor karyotyping labels and identifies each human chromosome pair by a different fluorescent color, using combinatorial labeling. Two groups (E. Schroeck and M. Speicher), using two different experimental strategies, have pioneered the procedure using five or six different fluors (SKYand M-FISH, respectively). Several M-FISH labeling schemes can be found??here. Another group has achieved color karyotyping (Color Changing Karyotyping =??CCK) using only three fluorophores, without the need for any specialized equipment or software.

In M-FISH, five fluorophores are used to acquire the desired 24 combinations. Individual chromosome painting libraries and many individual fluorophores were tested, either separately or in groups (Fig. 11j). DNA amount for the various painting probes were adjusted, to bring the signals to comparable intensities. DNA template for all chromosomes labeled with the same dye was mixed together, resulting in five different pools. Five separate labeling reactions (nick translation or PCR) were performed, one on each pool, using the corresponding dye. After labeling, different amounts of each pool were combined and the labeled DNA was precipitated in the presence of Cot1 and resuspended in hybridization buffer. Any five fluorophores can be used for M-FISH, provided that the microscope is equipped with the appropriate set of excitation and emission filters. For example, the original M-FISH procedure used FITC, Cy3, Cy3.5, Cy5 and Cy7. As Cy7 was not available in this laboratory, it was replaced with AMCA (blue side of the spectrum) or Cy5.5 (far-red). Because AMCA and DAPI have similar spectral characteristics, DAPI cannot be used at the same time with AMCA. M-FISH can be performed without DAPI, and various software packages (PSI) can be used to analyze the images even in the absence of DAPI (Fig. 11k). However, if the microscope is equipped with only 3-5 filters and AMCA was used in the analysis, DAPI images can still be captured using the following approach: M-FISH or CCK images of the best metaphases are captured first, and the position of these metaphases recorded using the verniers of the microscope stage. , The coverslip is removed, the slide is rinsed 15 minutes in 4xSSC buffer at room temperature, DAPI staining is performed and images of the same metaphases are captured using the DAPI filter. Depending on the amount of labeled probe used and the cost of individual reagents from a variety of commercial sources, the total cost of one M-FISH hybridization is in the range of $250-300.

However, by custom-preparing fluor-labeled dUTPs (for protocol details click??here) and custom labeling the chromosome painting probes , the overall cost per analysis can be reduced to less than $1.

M-FISH examples

1. Detection of balanced or non-balanced translocations.

2. Detection of complex translocation in tumors

??

Short Dna Probes

Plasmids

Short DNA plasmid probes (from 100 to 3000 bp) can be labeled using either nick translation or PCR. For very short probes, PCR is probably more efficient, whereas for longer fragments both labeling technique will work. This is illustrated in Fig. 9f and 9g, in which a 3kb plasmid probe was labeled using either PCR with vector primers (Fig. 9f) or nick translation (Fig. 9g). Labeling and hybridization worked well with both procedures, but the FISH signal appeared somewhat stronger when using the PCR-labeled probe. However, as the DNase treatment of the PCR fragments was minimal and the labeled fragments remained somewhat larger (500-2000 bp), there was a relatively large amount of background on the slide after hybridization (Fig. 9f). By comparison, the nick translated DNA fragments yielded a much cleaner hybridization. This underscores again the importance of using labeled DNA fragments around 100-300 bp average.

??

PCR products

FISH can be performed with unique, very short DNA fragments (primarily PCR products). In??Fig. 9h, two loci on chromosome Y (sY14 =??475bp on Yp and sY81 =??206bp on Yq) were amplified and labeled with DIG by PCR. The products were mixed together and hybridized onto normal male chromosomes. In some metaphases, two pairs of signals could be detected, the Yp ones (sY14) usually brighter by comparison. However, given the overall background on the slide, it would have been difficult to locate the signals without knowing their position. Many metaphases showed only one signal and signals were often not paired.

The reason for this variability is probably the reduced size of the labeled DNA fragments. The success of their hybridization depends on the chance that the corresponding DNA sequence is denatured and accessible for hybridization not only in every chromosome but also in each chromatid of the same chromosome. Denaturing is a more or less random process and the availability of any particular DNA stretch to hybridization also depends on the spatial position of the target sequence (whether it is towards the outside or is buried inside the chromatin). Therefore, hybridization of short sequences will result in double FISH signals (both chromatids) only occasionally.

By comparison, hybridization of a larger probe (a 200kb BAC) will result in a hybridization signal on almost every chromatid of every chromosome available on the slide. The labeled DNA fragments hybridizing to the target loci on the two chromosomes of the same metaphase are probably from different parts of the BAC. The reason is that on every chromatid, only random parts of the 200 kb target DNA will be denatured and available to hybridization with the labeled DNA probe fragments.

Metaphase Interphase Mapping

Technique for chromosome gene mapping

Often, G-banded slides can be subsequently used for FISH (Fig. 9i), a technique particularly useful in mapping DNA probes on chromosomes. A few G-banded metaphases are photographed using a light microscope. Immersion oil is then removed by rinsing the slide in HemoDE, xylene or ethanol, for approximately 5-10 minutes. Slides are kept 15-30 minutes in PBS, 2xSSC or isotonic buffer at room temperature (rehydration) and then rinsed briefly in 70% and 100% ethanol (the alcohol will also remove the Giemsa stain). Afterwards, slides are pretreated with pepsin and used for FISH.

The main problems with this approach are the aging technique of the slide and the type of slide storage. Usually, G-banded slides are aged overnight at 65?? C, then treated with trypsin and stained. The dry heat aging always impairs subsequent DNA hybridization. If either non-banded or G-banded slides are stored for years at room temperature (or in a dessicator), the material will become very “hard” and hybridazation will not be possible.

Chromosome position shift.

Chromosomes can shift and change shape and position during the denaturing process when subjected to the large temperature variations of the solutions. In??Fig. 9i,??the??arrows point to a chromosome that clearly changes its shape and position after denaturing and hybridization.

??

Chromosome mapping of multiple DNA probes.

A simple procedure used for mapping multiple probes (cosmids, BACs) on chromosomes is triple color FISH. In this approach, probes are divided in “triplets” (groups of three) and one probe in each group is labeled with one of three different color (red, green and blue). Each triplet is hybridized at the same time on chromosomes, and the position of the probes relative to one another is recorded (see??figure??below). Later on, three, four or even more such triplets (depending on the color order showed by each triplet) are mixed together and hybridized at the same time. In the figure below, the drawings to the left show the principle of the procedure. In the middle, three “triplets” (image top) were hybridized together with two more probes (11 probes total) on the same chromosomes (image bottom). On the right side, 18 different cosmids were simultaneously hybridized and mapped using the same principle.

??

Interphase FISH mapping (on interphase nuclei)

When the physical distance between DNA probes is shorter than 500-1000 kb, their position relative to one another on metaphase chromosomes becomes more difficult to assess. To map DNA probes located less than 400-500 kb apart, one can use interphase FISH mapping. A convenient approach uses triple color FISH (see figure below). The probes are again grouped in triplets, and each probe labeled with a different color (red =??R, green =??G, blue =??B). In the scheme at the top (step A), two triplets hybridize on nuclei. Results show that the red probe (R1) is in the “middle position in one of the triplets ( R1 is in the middle because in 50-100 nuclei, the distance between R1-G1 and R1-B1 is shorter than G1-B1. One can also measure that R1-G1 is shorter than R1-B1. In the other triplet, the green probe G2 is in the middle, and closer to B2 (G2-B2 is shorter than G2-R2). Because one does not know the relative position of the probes in triplet 1 from the probes in triplet 2, another hybridization is necessary to detect that relationship. In the second hybridization (step B), one probe from triplet 1 is mixed with probes from triplet 2 and vice-versa. After hybridization, in this hypothetical example, R2 is located between G1 and B1 but is closer to B1 than two G1 (in one hybridization) and B1 is located between G2 and R1, but closer to G2. If one combines this data together, the order depicted on the right side of the diagram is found. Such interphase hybridizations are depicted at the bottom of the figure. Because there are two chromosomes/nucleus, two such triplets can be scored in a nucleus. On the left, arrows depict the interphase hybridization pattern on two nuclei (nucl), showing the red probe between the green and the blue. On a rare chromosome (chr), we confirmed that pattern, and found that the blue probes was telomeric.In the example on the right, the chromosome shows a proximal green signal, with the position of the red and the blue probes impossible to differentiate. In a nucleus, one triplet shows blue in the middle, the other triplet shows red in the middle. This is the reason why one needs to count 50-100 such nuclei, to statistically determine which probes are closer to one another.

Cell Suspension Quality

We studied how the nuclear or chromosomal morphology of two non-related cell suspensions responds to the various steps of the FISH protocol (Fig. 2g-l). Simple Giemsa staining was used to quickly test the quality of the slides, as it was faster and easier than DAPI staining or hybridization. One cell suspension, probably sub-optimally treated with hypotonic buffer, showed more cytoplasmic residua around metaphases and nuclei. Slides from this suspension, were subjected to various times of ethanol aging, 10 seconds in??Fig. 2g, 50 seconds in??Fig. 2h??and 2 minutes in??Fig. 2i, and were denatured, treated with trypsin and G-banded. The G-banded pattern, although not very sharp, improved somewhat along with increasing times of ethanol aging. In general, the use of protease after denaturing was not very useful.

A second cell suspension, showing much less cytoplasmic residua was used to prepare another set of slides. One of these (Fig. 2j) was subjected to 2 minutes ethanol aging followed by trypsin and Giemsa staining, but without denaturing. The second slide (Fig. 2k) was ethanol aged 2 minutes, denatured and Giemsa stained. The third slide (Fig. 2l) was subjected to 60 minutes dry heat (94 C) followed by Giemsa staining. A chromosome banding pattern could be distinguished on all slides, but banding was best when dry heat was used (Fig. 2l), and worse after denaturing (2j better than 2k).

In??Fig. 2i and 2k, slides prepared from the two suspensions were treated identically and were compared. Chromosomes in??Fig. 2k, which belonged to the cells with less cytoplasmic residua, showed better morphology and banding pattern. Thus, proper preparation of cell suspensions and optimal hypotonic buffer pretreatment are important for both banding and FISH procedures.