Chromosome-specific staining to detect genetic rearrangements associated with chromosome 3 and/or chromosome 17

Abstract

Methods and compositions for staining based upon nucleic acid sequence that employ nucleic acid probes are provided. Said methods produce staining patterns that can be tailored for specific cytogenetic analyses. Said probes are appropriate for in situ hybridization and stain both interphase and metaphase chromosomal material with reliable signals. The nucleic acid probes are typically of a complexity greater than 50 kb, the complexity depending upon the cytogenetic application. Methods and reagents are provided for the detection of genetic rearrangements. Probes and test kits are provided for use in detecting genetic rearrangements, particularly for use in tumor cytogenetics, in the detection of disease related loci, specifically cancer, such as chronic myelogenous leukemia (CML), retinoblastoma, ovarian and uterine cancers, and for biological dosimetry. Methods and reagents are described for cytogenetic research, for the differentiation of cytogenetically similar but genetically different diseases, and for many prognostic and diagnostic applications.

Description

RELATED APPLICATION

This application is
If a large amount of blocking nucleic acid, for example, 10 ug were used (according to the standard hybridization protocols exemplified in Section VI.B infra wherein the practical limit of total nucleic acid is on the order of 10 ug in a 10′ ul hybridization mixture) with the 2 ng of probe, then Q=3×10^−5×104 ng/2 ng=3/2×10⁻¹=0.15. Thus, Q is <1, and is so low that the blocking DNA cannot substantially interfere with the desired signal. Increasing the amount of labeled probe nucleic acid to speed the hybridization would further decrease Q. In practice, one would typically use 1 ug of blocking DNA for such a hybridization.

Case II. As the size of the target region is increased, the complexity of the probe necessarily is increased, and the amount of DNA in the hybridization mix needs to be increased in order to have a sufficient concentration of each portion of specific sequence to hybridize. Also, if one desires to decrease the hybridization time of the procedure, the probe concentration must be increased. In these situations, the increase in probe concentration results in an increase in the amount of shared sequences in the hybridization mixture, which in turn increases the amount of hybridization that will occur to the shared sequences in the target area or areas, thereby reducing the contrast ratio.

With very high complexity probes spanning several entire chromosomes, L/G can approach 1. In order to stain such a portion of the genome within a reasonable time, for example, overnight, the concentration of labeled nucleic acid needs to be increased, for example, 200 ng in 10 ul of hybridization mixture. Up to about 3000 ng of blocking DNA can be used and still keep Q≦5 [wherein the calculation is Q=5 =0.3 m_b/200 ng or m_b=1000 ng/0.3=3,333 ng]. In practice, staining 25% and more of the human genome (for example, human chromosomes 1, 3 and 4) can be accomplished with the blocking protocols described below, but the contrast is less than for that achieved with probes for smaller regions.

III. Labeling the Nucleic Acid Fragments of the Heterogeneous Mixture

Several techniques are available for labeling single- and double-stranded nucleic acid fragments of the heterogeneous mixture. They include incorporation of radioactive labels, e.g. Harper et al. Chromosoma, Vol 83, pgs. 431-439 (1984); direct attachment of fluorochromes or enzymes, e.g. Smith et al., Nucleic Acids Research, Vol. 13, pgs. 2399-2412 (1985), and Connolly et al., Nucleic Acids Research Vol. 13, pgs. 4485-4502 (1985); and various chemical modifications of the nucleic acid fragments that render them detectable immunochemically or by other affinity reactions, e.g. Tchen et al., “Chemically Modified Nucleic Acids as Immunodetectable Probes in Hybridization Experiments,” Proc. Natl. Acad. Sci., Vol 81, pgs. 3466-3470 (1984); Richardson et al., “Biotin and Fluorescent Labeling of RNA Using T4 RNA Ligase,” Nucleic Acids Research, Vol. 11, pgs. 6167-6184 (1983); Langer et al., “Enzymatic Synthesis of Biotin-Labeled Polynucleotides: Novel Nucleic Acid Affinity Probes,” Proc. Natl. Acad. Sci., Vol. 78, pgs. 6633-6637 (1981); Brigati et al., “Detection of Viral Genomes in Cultured Cells and Paraffin-Embedded Tissue Sections Using Biotin-Labeled Hybridization Probes,” Virology, Vol. 126, pgs. 32-50 (1983); Broker et al., “Electron Microscopic Visualization of tRNA Genes with Ferritin-Avidin: Biotin Labels,” Nucleic Acids Research, Vol. 5, pgs. 363-384 (1978); Bayer et al., “The Use of the Avidin Biotin Complex as a Tool in Molecular Biology,” Methods of Biochemical Analysis Vol. 26, pgs. 1-45 (1980) Kuhlmann, Immunoenzyme Techniques in Cytochemistry (Weinheim, Basel, 1984). Langer-Safer et al., PNAS USA, 79: 4381 (1982): Landegent et al., Exp. Cell Res., 153: 61 (1984); and Hopman et al., Exp. Cell Res., 169: 357 (1987).

Exemplary labeling means include those wherein the probe fragments are biotinylated, modified with N-acetoxy-N-2-acetylaminofluorene, modified with fluorescein isothiocyanate, modified with mercury/TNP ligand, sulfonated, digoxigenenated or contain T-T dimers.

The key feature of “probe labeling” is that the probe bound to the target be detectable. In some cases, an intrinsic feature of the probe nucleic acid, rather than an added feature, can be exploited for this purpose. For example, antibodies that specifically recognize RNA/DNA duplexes have been demonstrated to have the ability to recognize probes made from RNA that are bound to DNA targets [Rudkin and Stollar, Nature, 265:472-473 (1977)]. The RNA used for such probes is unmodified. Probe nucleic acid fragments can be extended by adding “tails” of modified nucleotides or particular normal nucleotides. When a normal nucleotide tail is used, a second hybridization with nucleic acid complementary to the tail and containing fluorochromes, enzymes, radioactivity, modified bases, among other labeling means, allows detection of the bound probe. Such a system is commerically available from Enzo Biochem (Biobridge Labeling System; Enzo Biochem Inc., New York, N.Y.).

Another example of a means to visualize the bound probe wherein the nucleic acid sequences in the probe do not directly carry some modified constituent is the use of antibodies to thymidine dimers. Nakane et al., 20 (2):229 (1987), illustrate such a method wherein thymine-thymine dimerized DNA (T-T DNA) was used as a marker for in situ hybridization. The hybridized T-T DNA was detected immunohistochemically using rabbit anti-T-T DNA antibody.

All of the labeling techniques disclosed in the above references may be preferred under particular circumstances. Accordingly, the above-cited references are incorporated by reference. Further, any labeling techniques known to those in the art would be useful to label the staining compositions of this invention. Several factors govern the choice of labeling means, including the effect of the label on the rate of hybridization and binding of the nucleic acid fragments to the chromosomal DNA, the accessibility of the bound probe to labeling moieties applied after initial hybridization, the mutual compatibility of the labeling moieties, the nature and intensity of the signal generated by the label, the expense and ease in which the label is applied, and the like.

Several different high complexity probes, each labeled by a different method, can be used simultaneously. The binding of different probes can thereby be distinguished, for example, by different colors.

IV. In Situ Hybridization

Application of the heterogeneous mixture of the invention to chromosomes is accomplished by standard in situ hybridization techniques. Several excellent guides to the technique are available, e.g., Gall and Pardue, “Nucleic Acid Hybridization in Cytological Preparations,” Methods in Enzymology. Vol. 21, pgs. 470-480 (1981); Henderson, “Cytological Hybridization to Mammalian Chromosomes,” International Review of Cytology, Vol. 76, pgs. 1-46 (1982); and Angerer, et al., “In Situ Hybridization to Cellular RNAs,” in Genetic Engineering: Principles and Methods Setlow and Hollaender, Eds., Vol. 7, pgs. 43-65 (Plenum Press, New York, 1985). Accordingly, these references are incorporated by references.

Three factors influence the staining sensitivity of the hybridization probes: (1) efficiency of hybridization (fraction of target DNA that can be hybridized by probe), (2) detection efficiency (i.e., the amount of visible signal that can be obtained from a given amount of hybridization probe), and (3) level of noise produced by nonspecific binding of probe or components of the detection system.

Generally in situ hybridization comprises the following major steps: (1) fixation of tissue or biological structure to be examined, (2) prehybridization treatment of the biological structure to increase accessibility of target DNA, and to reduce nonspecific binding, (3) hybridization of the heterogeneous mixture of probe in the DNA in the biological structure or tissue; (4) posthybridization washes to remove probe not bound in specific hybrids, and (5) detection of the hybridized probes of the heterogeneous mixture. The reagents used in each of these steps and their conditions of use vary depending on the particular situation.

The following comments are meant to serve as a guide for applying the general steps listed above. Some experimentation may be required to establish optimal staining conditions for particular applications.

In preparation for the hybridization, the probe, regardless of the method of its production, may be broken into fragments of the size appropriate to obtain the best intensity and specificity of hybridization. As a general guideline concerning the size of the fragments, one needs to recognize that if the fragments are too long they are not able to penetrate into the target for binding and instead form aggregates that contribute background noise to the hybridization; however, if the fragments are too short, the signal intensity is reduced.

Under the conditions of hybridization exemplified in Section VI.B wherein human genomic DNA is used as an agent to block the hybridization capacity of the high copy shared repetitive sequences, the preferred size range of the probe fragments is from about 200 bases to about 600. The preferred hybridization temperature is about 30° C. to about 45° C., more preferably about 35° C. to about 40° C., and still more preferably about 37° C.; preferred washing temperature range is from about 40° C. to about 50° C., more preferably about 45° C.

The size of the probe fragments is checked before hybridization to the target; preferably the size of the fragments is monitored by electrophoresis, more preferably by denaturing agarose gel electrophoresis.

Fixatives include acid alcohol solutions, acid acetone solutions, Petrunkewitsch's reagent, and various aldehydes such as formaldehyde, paraformaldehyde, glutaraldehyde, or the like. Preferably, ethanol-acetic acid or methanol-acetic acid solutions in about 3:1 proportions are used to fix the chromosomes in metaphase spreads. For cells or chromosomes in suspension, a fixation procedure disclosed by Trask, et al., in Science, Vol. 230, pgs. 1401-1402 (1985), is useful. Accordingly, Trask et al., is incorporated by reference. Briefly, K₂CO₃ and dimethylsuberimidate (DMS) are added (from a 5×concentrated stock solution, mixed immediately before use) to a suspension containing about 5×10⁶ nuclei/ml. Final K₂CO₃ and DMS concentrations are 20 mM and 3 mM, respectively. After 15 minutes at 25° C., the pH is adjusted from 10.0 to 8.0 by the addition of 50 microliters of 100 mM citric acid per milliliter of suspension. Nuclei are washed once by centrifugation (300 g, 10 minutes, 4° C. in 50 mM kCl, 5 mM Hepes buffer, at pH 9.0, and 10 mM MgSO₄).

A preferred fixation procedure for cells or nuclei in suspension is disclosed by Trask et al., Hum. Genet. 78:251-259 (1988), which article is herein incorporated by reference. Briefly, nuclei are fixed for about 10 minutes at room temperature in 1% paraformaldehyde in PBS, 50′ mM MgSO₄, pH 7.6 and washed twice. Nuclei are resuspended in isolation buffer (IB) (50 mM KCl, 5 mM HEPES, 10 mM MgSO₄, 3 mM dithioerythritol, 0.15 mg/ml RNase, pH 8.0)/0.05% Triton X-100 at 10⁸/ml.

Frequently before in situ hybridization chromosomes are treated with agents to remove proteins. Such agents include enzymes or mild acids. Pronase, pepsin or proteinase K are frequently used enzymes. A represent- ative acid treatment is 0.02-0.2 N HCl, followed by high temperature (e.g., 70° C.) washes. Optimization of deproteinization requires a combination of protease concentration and digestion time that maximizes hybridization, but does not cause unacceptable loss of morphological detail. Optimum conditions vary according to tissue types and method of fixation. Additional fixation after protease treatment may be useful. Thus, for particular applications, some experimentation may be required to optimize protease treatment.

In some cases pretreatment with RNase may be desirable to remove residual RNA from the target. Such removal can be accomplished by incubation of the fixed chromosomes in 50-100 microgram/milliliter RNase in 2×SSC (where SSC is a solution of 0.15M NaCL and 0.015M sodium citrate) for a period of 1-2 hours at room temperature.

The step of hybridizing the probes of the heterogeneous probe mixture to the chromosomal DNA involves (1) denaturing the target DNA so that probes can gain access to complementary single-stranded regions, and (2) applying the heterogeneous mixture under conditions which allow the probes to anneal to complementary sites in the target. Methods for denaturation include incubation in the presence of high pH, low pH, high temperature, or organic solvents such as formamide, tetraalkylammonium halides, or the like, at various combinations of concentration and temperature. Single-stranded DNA in the target can also be produced with enzymes, such as, Exonuclease III [van Dekken et al., Chromosoma (Berl) 97:1-5 (1988)]. The preferred denaturing procedure is incubation for between about 1-10 minutes in formamide at a concentration between about 35-95 percent in 2×SSC and at a temperature between about 25-70° C. Determination of the optimal incubation time, concentration, and temperature within these ranges depends on several variables, including the method of fixation and type of probe nucleic acid (for example, DNA or RNA).

After the chromosomal DNA is denatured, the denaturing agents are typically removed before application of the heterogeneous probe mixture. Where formamide and heat are the primary denaturing agents, removal is conveniently accomplished by several washes with a solvent, which solvent is frequently chilled, such as a 70%, 85%, 100% cold ethanol series. Alternatively the composition of the denaturant can be adjusted as appropriate for the in situ hybridization by addition of other constituents or washes in appropriate solutions. The probe and target nucleic acid may be denatured simultaneously by applying the hybridization mixture and then heating to the appropriate temperature.

The ambient physiochemical conditions of the chromosomal DNA and probe during the time the heterogeneous mixture is applied is referred to herein as the hybridization conditions, or annealing conditions. Optimal hybridization conditions for particular applications can be adjusted by controlling several factors, including concentration of the constituents, incubation time of chromosomes in the heterogeneous mixture, and the concentrations, complexities, and lengths of the nucleic acid fragments making up the heterogeneous mixture. Roughly, the hybridization conditions must be sufficiently close to the melting temperature to minimize nonspecific binding. On the other hand, the conditions cannot be so stringent as to reduce correct hybridizations of complementary sequences below detectable levels or to require excessively long incubation times.

The concentrations of nucleic acid in the hybridization mixture is an important variable. The concentrations must be high enough so that sufficient hybridization of respective chromosomal binding sites occurs in a reasonable time (e.g., within hours to several days). Higher concentrations than that necessary to achieve adequate signals should be avoided so that nonspecific binding is minimized. An important practical constraint on the concentration of nucleic acid in the probe in the heterogeneous mixture is solubility. Upper bounds exist with respect to the fragment concentration, i.e., unit length of nucleic acid per unit volume, that can be maintained in solution and hybridizes effectively.

In the representational examples described in Section VI.B (infra), the total DNA concentration in the hybridization mixture had an upper limit on the order of 1 ug/ul. Probe concentrations in the range of 1-20 ng/ul were used for such whole chromosome staining. The amount of genomic blocking DNA was adjusted such that Q was less than 5. At the low end of probe concentration, adequate signals were obtained with a one hour incubation, that is, a time period wherein the probe and blocking DNA are maintained together before application to the targeted material, to block the high-copy sequences and a 16 hour hybridization. Signals were visible after two hours of hybridization. The best results (bright signals with highest contrast) occurred after a 100 hour hybridization, which gave the low-copy target-specific sequences more opportunity to find binding sites. At the high end of the probe concentration, bright signals are obtained after hybridizations of 16 hours or less; the contrast was reduced since more labeled repetitive sequences were included in the probe.

The fixed target object can be treated in several ways either during or after the hybridization step to reduce non-specific binding of probe DNA. Such treatments include adding nonprobe, or “carrier”, DNA to the heterogeneous mixture, using coating solutions, such as Denhardt's solution (Biochem. Biophys. Res. Commun., Vol. 23, pgs. 641-645 (1966), with the heterogeneous mixture, incubating for several minutes, e.g., 5-20, in denaturing solvents at a temperature 5-10° C. above the hybridization temperature, and in the case of RNA probes, mild treatment with single strand RNase (e.g., 5-10 micrograms per millileter RNase) in 2×SSC at room temperature for 1 hour).

V. Chromosome-Specific Staining Reagents Comprising Selected Single-Copy Sequences

V.A. Making and Using a Staining Reagent Specific to Human Chromosome 21

V.A.1. Isolation of Chromosome 21 and Construction of a Chromosome 21-Specific Library

DNA fragments from human chromosome-specific libraries are available from the National Laboratory Gene Library Project through the American Type Culture Collection (ATCC), Rockville, Md. DNA fragments from chromosome 21 were generated by the procedure described by Fuscoe et al., in “Construction of Fifteen Human Chromosome-Specific DNA Libraries from Flow-Purified Chromosomes,” Cytogenet. Cell Genet., Vol. 43, pgs. 79-86 (1986), which reference is incorporated by reference. Briefly, a human diploid fibroblast culture was established from newborn foreskin tissue. Chromosomes of the cells were isolated by the MgSO₄ method of van den Engh et al., Cytometry, Vol. 5, pgs. 108-123 (1984), and stained with the fluorescent dyes—Hoechst 33258 and Chromomycin A3. Chromosome 21 was purified on the Lawrence Livermore National Laboratory high speed sorter, described by Peters et al., Cytometry, Vol. 6, pgs. 290-301 (1985).

After sorting, chromosome concentrations were approximately 4×10⁵/ml. Therefore, prior to DNA extraction, the chromosomes (0.2-1.0×10⁶) were concentrated by centrifugation at 40,000×g for 30 minutes at 4° C. The pellet was then resuspended in 100 microliters of DNA isolation buffer (15 mM NaCl, 10 mM EDTA, 10 mM Tris HCl pH 8.0) containing 0.5% SDS and 100 micrograms/ml proteinase K. After overnight incubation at 37° C., the proteins were extracted twice with phenol:chloroform: isoamyl alcohol (24:24:1) and once with chloroform:isoamyl alcohol (24:1). Because of the small amounts of DNA, each organic phase was reextracted with a small amount of 10 mM Tris pH 8.0, 1 mM EDTA (TE). Aqueous layers were combined and transferred to a Schleicher and Schuell mini-collodion membrane (#UHO20/25) and dialyzed at room temperature against TE for 6-8 hours. The purified DNA solution was then digested with 50 units of HindIII (Bethesda Research Laboratories, Inc.) in 50 mM NaCl, 10 mM Tris HCl pH 7.5, 10 mM MgCl₂, 1 mM dithiothreitol. After 4 hours at 37°, the reaction was stopped by extractions with phenol and chloroform as described above. The aqueous phase was dialyzed against water overnight at 4° C. in a mini-collodion bag and then 2 micrograms of Charon 21A arms cleaved with HindIII and treated with calf alkaline phosphatase (Boehringer Mannheim) were added. This solution was concentrated under vacuum to a volume of 50-100 microliters and transferred to a 0.5 ml microfuge tube where the DNA was precipitated with one-tenth volume 3M sodium acetate pH 5.0 and 2 volumes ethanol. The precipitate was collected by centrifugation, washed with cold 70% ethanol, and dissolved in 10 microliters of TE.

After allowing several hours for the DNA to dissolve, 1 microliter of 10×X ligase buffer (0.5M Tris HCl pH 7.4, 0.1 M MgCl₂, 0.1M dithiothreitol, 10 mM ATP, 1 mg/ml bovine serum albumin) and 1 unit of T4 ligase (Bethesda Research Laboratory, Inc.) were added. The ligation reaction was incubated at 10° C. for 16-20 hours and 3 microliter aliquots were packaged into phage particles using in vitro extracts prepared from E. coli strains BHB 2688 and BHB 2690, described by Hohn in Methods in Enzymology, Vol. 68, pgs. 299-309 (1979) Molecular Cloning: A Laboratory Manual, (Cold Spring Harbor Laboratory, New York, 1982). Briefly, both extracts were prepared by sonication and combined at the time of in vivo packaging. These extracts packaged wild-type lambda DNA at an efficiency of 1-5×10⁸ plaque forming units (pfu) per microgram. The resultant phage were amplified on E. coli LE392 at a density of approximately 10⁴ pfu/150 mm dish for 8 hours to prevent plaques from growing together and to minimize differences in growth rates of different recombinants. The phage were eluted from the agar in 10 ml SM buffer (50 mM Tris HCl pH 7.5, 10 mm MgSO₄, 100 mM NaCl, 0.01% gelatin) per plate by gentle shaking at 4° C. for 12 hours. The plates were then rinsed with an additional 4 ml of SM. After pelleting cellular debris, the phage suspension was stored over chloroform at 4° C.

V.A.2. Construction and Use of Chromosome 21-Specific Stain for Staining Chromosome 21 of Human Lymphocytes

Clones having unique sequence inserts are isolated by the method of Benton and Davis, Science, Vol. 196, pgs. 180-182, (1977). Briefly, about 1000 recombinant phage are isolated at random from the chromosome 21-specific library. These are transferred to nitrocellulose and probed with nick translated total genomic human DNA.

Of the clones which do not show strong hybridization, approximately 300 are picked which contain apparent unique sequence DNA. After the selected clones are amplified, the chromosome 21 insert in each done is ³²P labeled and hybridized to Southern blots of human genomic DNA digested with the same enzyme used to construct the chromosome 21 library, i.e., Hind III. Unique sequence containing clones are recognized as those that produce a single band during Southern analysis. Roughly, 100 such clones are selected for the heterogeneous mixture. The unique sequence clones are amplified, the inserts are removed by Hind III digestions, and the inserts are separated from the phage arms for gel electrophoresis. The probe DNA fragments (i.e., the unique sequence inserts) are removed from the gel and biotinylated by nick translation (e.g., by a kit available from Bethesda Research Laboratories). Labeled DNA fragments are separated from the nick translation reaction using small spin columns made in 0.5 ml Eppendorph tubes filled with Sephadex G-50 (medium) swollen in 50 mM Tris, 1 mM EDTA, 0.1% SDS, at pH 7.5. Human lymphocyte chromosomes are prepared following Harper et al, Proc. Natl. Acad. Sci. Vol. 78, pgs. 4458-4460 (1981). Metaphase and interphase cells were washed 3 times in phosphate buffered saline, fixed in methanol-acetic acid (3:1) and dropped onto cleaned microscope slides. Slides are stored in a nitrogen atmosphere at −20° C.

Slides carrying interphase cells and/or metaphase spreads are removed from the nitrogen, heated to 65° C. for 4 hours in air, treated with RNase (100 micrograms/ml for 1 hour at 37° C.), and dehydrated in an ethanol series. They are then treated with proteinase K (60 ng/ml at 37° C. for 7.5 minutes) and dehydrated. The proteinase K concentration is adjusted depending on the cell type and enzyme lot so that almost no phase microscopic image of the chro- mosomes remains on the dry slide. The hybridization mix consists of (final concentrations) 50 percent formamide, 2×SSC, 10 percent dextran sulfate, 500 micrograms/ml carrier DNA (sonicated herring sperm DNA), and 2.0 microgram/ml biotin-labeled chromosome 21-specific DNA. This mixture is applied to the slides at a density of 3 microliters/cm² under a glass coverslip and sealed with rubber cement. After overnight incubation at 37° C., the slides are washed at 45° C. (50% formamide-2×SSC pH 7, 3 times 3 minutes; followed by 2×SSC pH 7, 5 times 2 minutes) and immersed in BN buffer (0.1 M Na bicarbonate, 0.05 percent NP-40, pH 8). The slides are never allowed to dry after this point.

The slides are removed from the BN buffer and blocked for 5 minutes at room temperature with BN buffer containing 5% non-fat dry milk (Carnation) and 0.02% Na Azide (5 microliter/cm² under plastic coverslips). The coverslips are removed, and excess liquid briefly drained and fluorescein-avidin DCS (3 microgram/ml in BN buffer with 5% milk and 0.02% NaAzide) is applied (5 microliter/cm²). The same coverslips are replaced and the slides incubated 20 minutes at 37° C. The slides are then washed 3 times for 2 minutes each in BN buffer at 45° C. The intensity of biotin-linked fluorescence is amplified by adding a layer of biotinylated goat anti-avidin antibody (5 microgram/ml in BN buffer with 5% goat serum and 0.02% Na Azide), followed, after washing as above, by another layer of fluorescein-avidin DCS. Fluorescein-avidin DCS, goat antiavidin and goat serum are all available commercially, e.g., Vector Laboratories (Burlingame, Calif.). After washing in BN, a fluorescence antifade solution, p-phenylenediamine (1.5 microliter/cm² of coverslip) is added before observation. It is important to keep this layer thin for optimum microscopic imaging. This antifade significantly reduced fluorescein fading and allows continuous microscopic observation for up to 5 minutes. The DNA counterstains (DAPI or propidium iodide) are included in the antifade at 0.25-0.5 microgram/ml.

The red-fluorescing DNA-specific dye propidium iodide (PI) is used to allow simultaneous observation of hybridized probe and total DNA. The fluorescein and PI are excited at 450-490 nm (Zeiss filter combination 487709). Increasing the excitation wavelength to 546 nm (Zeiss filter combination 487715) allows observation of the PI only. DAPI, a blue fluorescent DNA-specific stain excited in the ultraviolet (Zeiss filter combination 487701), is used as the counterstain when biotin-labeled and total DNA are observed separately. Metaphase chromosome 21s are detected by randomly located spots of yellow distributed over the body of the chromosome.

V.B. Improved Method for Efficiently Selecting Chromosome 21 Single-Copy Sequences

Fuscoe et al., Genomics, 5:100-109 (1989) provides more efficient procedures than the method described immediately above (V.A.2) for selecting large numbers of single-copy sequence or very low copy number repeat sequence clones from recombinant phage libraries and demonstrates their use to stain chromosome 21. Said article is hereby incorporated by reference. Briefly, clones were selected from the Charon 21A library LL21NS02 (made from DNA from human chromosome 21) using two basic procedures. In the first, the phage library was screened in two stages using methods designed to be more sensitive to the presence of repetitive sequences in the clones than the method of Section V.A.2. The selected clones were then subcloned into plasmids. The 450 inserts thus selected form the library pBS-U21. The second was in a multistep process in which: 1) Inserts from LL21NS02 were subcloned into Bluescribe plasmids, 2) plasmids were grown at high density in bacterial colonies on nitrocellulose filters and 3) radioactive human genomic DNA was hybridized to the plasmid DNA on nitrocellulose filters at low stringency in two steps and 4) plasmids having inserts that failed to hybridize were selected as potentially carrying single-copy sequences. Fifteen hundred and thirty colonies were picked in this manner to form the library pBS-U21/1530.

Southern analysis indicated that the second procedure was more effective at recognizing repetitive sequence than the first. Fluorescence in situ hybridization with DNA from pBS-U21/1530 allowed specific, intense staining of the number 21 chromosomes in metaphase spreads made from human lymphocytes. Hybridization with pBS-U21 gives less specific staining of chromosome 21. Details concerning the Fuscoe et al. method of selecting single-copy sequence or very low repeat sequence probes from recombinant libraries can be found in found in Fuscoe et al., id.

V.C. Hybridization with a Collection of Chromosome 4 Single-Copy Sequences

Chromosome 4 Single-Copy Sequences. One hundred and twenty clones carrying chromosome 4-specific single-copy sequence inserts selected from the Charon 21A library LL04NS01 (ATCC accession number 57700; Van Dilla et al., supra; see Table 1) were supplied by C. Gilliam (Harvard University) [Gilliam et al., Nucleic Acids Res. 15:1445 (1987)]. The human inserts were all about 3 kilobases (kb) in length, so the ratio of insert to vector DNA was <0.1. Total phage DNA was produced from each clone individually using DEAE-cellulose columns (Whatman DE-52) [Helms et al., DNA 4:39 (1985)]. DNA pooled from the 120 clones was biotinylated by nick-translation with biotin-11-dUTP (Bethesda Research Laboratories) and recovered at a concentration of about 20 nanograms per microliter (ng/ul) using Sephadex G-50 spin columns.

Cells. Metaphase spreads from human lymphocytes were prepared from methotrexate-synchronized cultures by using the procedure of Harper et al., supra. The cells were fixed in methanol/acetic acid, 3:1. Slides were stored at −20° C. in plastic bags filled with nitrogen gas.

In Situ Hybridization: Single-Copy Hybridization. Hybridization was accomplished by using a modification of the procedure described by Pinkel et al., PNAS USA, 83: 2934 (1986). The slide mounted cells were treated with RNase [100 micrograms per milliliter (ug/ml) in 0.3 molar (M) sodium chloride (NaCl)/30 millimolar (mM) sodium citrate at 37° C. for 1 hr), dehydrated in a 70%/85%/100% ethanol series, treated with proteinase K (0.3-0.6 ug/ml in 20 mM Tris/2 mM CaCl₂, pH 7.5, for 7.5 min at 37° C.), and fixed [4% paraformaldehyde in phosphate-buffered saline (PBS; in g/liter, KCl, 0.2; KH₂PO₄, 0.2; NaCl, 8; Na₂HPO₄.7H₂O, 2.16) plus 50′ mM MgCl₂ for 10 min at room temperature]. The DNA in the target cells was denatured by immersion in 70% formamide/2×SSC (0.3 M NaCl/30 mM sodium citrate) at pH 7, for 2 min at 70° C. The hybridization mixture [10 ul total volume consisting of 50% formamide, 0.3 M NaCl/30 mM sodium citrate (final concentration), 10% dextran sulfate, 50 ug of sonicated herring DNA per ml, and 3-6 ng of biotinylated chromosome 4 unique sequences (40-80 ng of total phage DNA)] was then denatured (70° C. for 5 min) and applied. Hybridization was at 37° C. overnight (16 hr). Slides were washed in three changes of 50% formamide/0.3 M NaCl/30 mM sodium citrate (final concentration), pH 7, at 45° C. for 5 min each and once in PN buffer (a mixture of 0.1 M NaH₂PO₄ and 0.1 M Na₂HPO4 to give pH 8/0.1% Nonidet P-40). The slides were then treated with alternating layers of fluoresceinated avidin and biotinylated goat antiavidin, both at 5 ug/ml in PNM buffer (PN buffer/5% non-fat dry milk/0.02% sodium azide, centrifuged to remove solids), for 20 min each at room temperature until three layers of avidin were applied. The avidin and goat anti-avidin treatments were separated by three washes of 3 min each in PN buffer [avidin (DCS grade) and anti-avidin from Vector Laboratories (Burlingame, Calif.)]. After the final avidin treatment, a fluorescence antifade solution [Johnson and Noqueria, J. Immunol. Methods, 43:349 (1981)] containing 1 ug of 4′,6-amidino-2-phenylindole or propidium iodide per ml was applied as a counterstain (1.5 ul/cm² under a no. 1 coverslip).

Results. As shown in FIG. 4H, individual hybridization sites could be located to within a fraction of the width of a chromatid after overnight hybridization (16 hr) and application of three layers of avidin. Analysis of three spreads from the hybridization with the 120 unique sequence probes at a total probe concentration of 1.5 pg/ul per kilobase of human insert, showed 222 fluorescent spots out of the 1440 possible on the number 4 chromosomes (120 target sites per chromatid×4 chromatids per metaphase×3 metaphases). Thus, the hybridization efficiency was 15%. There were 814 total spots on all of the chromosomes giving a hybridization specificity of 27%. The experiment demonstrates that substantial hybridization can occur with single copy probes at low probe concentrations in overnight hybridizations. The contrast ratio of chromosome 4 relative to the rest of the $\text{chromosomes was thus }\frac{\text{spots/length of chromosome 4}}{\text{spots/length of all chromosomes}}=\frac{\text{222/.06}}{\text{814/1.0}}=\text{approximately 4.}$
(Chromosome 4 comprises about 6% of the genome.)
VI. Incapacitating Shared Repetitive Sequences
VI.A. Chromosome 21-Specific Staining Using Blocking DNA

High concentration of unlabeled human genomic DNA and lambda phage DNA were used to inhibit the binding of repetitive and vector DNA sequences to the target chromosomes. Heavy proteinase digestion and subsequent fixation of the target improved access of probes to target DNA.

Human metaphase spreads were prepared on microscope slides with standard techniques and stored immediately in a nitrogen atmosphere at −20° C.

Slides were removed from the freezer and allowed to warm to room temperature in a nitrogen atmosphere before beginning the staining procedure. The warmed slides were first treated with 0.6 microgram/ml proteinase K in P buffer (20 mM Tris, 2 mM CaCl₂ at pH 7.5) for 7.5 minutes, and washed once in P buffer. The amount of proteinase K used needs to be adjusted for different batches of slides. After denaturing the slides were stored in 2×SSC. A hybridization mix was prepared which consisted of 50% formamide, 10% dextran sulfate, 1% Tween 20, 2×SSC, 0.5 mg/ml human genomic DNA, 0.03 mg/ml lambda DNA, and 3 microgram/ml biotin labeled probe DNA. The probe DNA consisted of the highest density fraction of phage from the chromosome 21 Hind III fragment library (ATCC accession number 57713), as determined by a cesium chloride gradient. (Both insert and phage DNA of the probe were labeled by nick translation.) The average insert size (amount of chromosome 21 DNA), as determined by gel electrophoresis was about 5 kilobases. No attempt was made to remove repetitive sequences from the inserts or to isolate the inserts from the lambda phage vector. The hybridization mix was denatured by heating to 70° C. for 5 minutes followed by incubation at 37° C. for 1 hour. The incubation allows the human genomic DNA and unlabeled lambda DNA in the hybridization mix to block the human repetitive sequences and vector sequences in the probe.

The slide containing the human metaphase spread was removed from the 2×SSC and blotted dry with lens paper. The hybridization mix was immediately applied to the slide, a glass cover slip was placed on the slide with rubber cement, and the slide was incubated overnight at 37° C. Afterwards preparation of the slides proceeded as described in Section V.B. (wherein chromosome 21 DNA was stained with fluorescein and total chromosomal DNA counterstained with DAPI). FIGS. 1A-C illustrate the results. FIG. 1A is a DAPI image of the human metaphase spread obtained with a computerized image analysis system. It is a binary image showing everything above threshold as a white, and the rest as black. The primary data was recorded as a gray level image with 256 intensity levels. (Small arrows indicate the locations of the chromosome 21s.) FIG. 1B is a fluorescein image of the same spread as in FIG. 1A, again in binary form. (Again, small arrows indicate the locations of the chromosome 21s.) FIG. 1C illustrates the positions of the chromosome 21s after other less densely stained objects were removed by standard image processing techniques.

VI.B. Detection of Trisomy 21 and Translocations of Chromosome 4 Using Bluescribe Plasmid Libraries

As illustrated in Section VI.A., a human chromosome-specific library, including its shared repetitive sequences, can be used to stain that chromosome if the hybridization capacity of the shared repetitive sequences is reduced by incubation with unlabeled human genomic DNA. In Section VI.A., the nucleic acid sequences of the heterogeneous mixture were cloned in the phage vector Charon 21A, in which the ratio of insert of vector DNA is about 0.1 (4 kb average insert to 40 kb of vector). In this section, we demonstrate that transferring the same inserts to a smaller cloning vector, the about 3 kb Bluescribe plasmid, which increases the ratio of insert to vector DNA to 0.5, improved the specificity and intensity of the staining.

As previously discussed, incubation of the probe can be carried out with the probe alone, with the probe mixed with unlabeled genomic DNA, and with the probe mixed with unlabeled DNA enriched in all or some shared repetitive sequences. If unlabeled genomic DNA is added, then it is important to add enough to incapacitate sufficiently the shared repetitive sequences in the probe. However, the genomic DNA also contains unlabeled copies of the sequences, the hybridization of which is desired. As explained above, Q is herein defined as the ratio of unlabeled to labeled copies of the chromosome- specific sequences in the hybridization mixture.

Cells. Metaphase spreads from human lymphocytes were prepared from methotrexate-synchronized cultures by using the procedure of Harper et al. supra. These and all other cells used in this example were fixed in methanol/acetic add, 3:1. Other human lymphocyte cultures were irradiated with ⁶⁰Co gamma rays and stimulated with phytohemagglutnin. Colcemid was added 48 hr after stimulation and metaphase spreads were prepared 4 hr later. Metaphase spreads and interphase cells from lymphoblastoid cells (GM03716A; Human Mutant Cell Repository, Camden, N.J.) carrying trisomy 21 were prepared after a 4-hr colcemid block. Interphase cells from the cell line RS4;11 carrying t(4;11) and isochromosome 7q were harvested, fixed in methanol/acetic add, and dropped onto slides [Strong et al., Blood, 65:21 (1985)]. Slides were stored at −20° C. in plastic bags filled with nitrogen gas.

pBS-4. The entire chromosome 4 library LL04NS02 (ATCC accession number 57745; Van Dilla et al., supra) was subcloned into Bluescribe plasmids (Stratagene La Jolla, Calif.) to form the library pBS4. The average insert to vector DNA ratio in pBS-4 is about 1. The plasmid library was amplified in bulk and the DNA was extracted using DEAE-cellulose columns (Whatman DE-52) [Helms et al., DNA, 4:39 (1985)]. The DNA was then biotinylated by nick translation with biotin-11-dUTP (Bethesda Research Laboratories) and recovered at a concentration of about 20 ng/ul using Sephadex G-50 spin columns. In some experiments, the biotinylated DNA was concentrated by ethanol precipitation to achieve higher probe concentrations.

pBS-21. The entire chromosome 21 library LL21NS02 (ATCC accession number 57713; Van Dilla et al., supra) was subcloned into Bluescribe plasmids to form the library pBS-21. This library was amplified and biotinylated as described above for pBS4.

Human genomic DNA. Placental DNA (Sigma) was treated with proteinase K, extracted with phenol, and sonicated to a size range of 200-600 base pairs (bp).

Whole Library Hybridization. Hybridization was as above in section V.C except that RNase, proteinase K, and paraformaldehyde were not used. The amount of probe and genomic DNA in the hybridization mixture and the length of the hybridization varied as described in Results. All probe concentrations refer to the human insert DNA unless otherwise noted. DNA concentrations were determined by fluorometric analysis (Hoeffer Scientific Instruments, San Francisco). Incubation of the hybridization mixture prior to hybridization followed two different protocols as indicated immediately below.

Protocol I. The hybridization mixture (10 ul) contained 10-150 ng of biotinylated human DNA (20-300 ng of total plasmid DNA) and 0-10 ug of unlabeled genomic DNA. The mixture was heated to denature the DNA and incubated at 37° C. for a time t before it was added to the slide. Hybridization times ranged from 2 to 110 hr.

Protocol II. Protocol II was identical to Protocol I except that an additional aliquot of freshly denatured genomic DNA was added to the hybridization mixture after an incubation time t. The mixture was then incubated an additional time t prior to starting the hybridization. The volume of the hybridization mixture was increased <20% by the additional genomic DNA.

Microscopy. Quantitative fluorescence measurements were performed using a video camera on the microscope and a digital image processing system, [Trask et al., Human Genet., 78:251 (1988)] Results. FIG. 4A shows hybridization of pBS-4 to a human metaphase spread with a probe concentration of 1 ng/ul. No genomic DNA was used and the hybridization mixture was applied immediately after denaturation. All of the chromosomes are stained, except near many centromeres, with two copies of chromosome 4 being stained most heavily. All of the chromosomes are stained along most of their lengths due to sequences in the probe which are shared with other chromosomes. Unstained regions, noted by arrows, show locations for which homologous sequences are not present in pBS-4. The unstained regions are mostly centromeric and along the long arm of the Y chromosome. Blocks of repetitive DNA specific to those sites are known to exist.

The visible contrast on chromosome 4 is the result of the interaction of several factors. (i) All of the DNA in chromosome 4 is potential target for sequences in the probe, whereas only those sequences on the other chromosomes that are shared with chromosome 4 can bind probe. (ii) The hybridization time and probe concentration were high enough to allow significant binding of the specific sequences in the probe. (iii) The ratio of probe to target sequences is higher for the specific sequences than for the shared sequences [Ten nanograms of chromosome 4 DNA was hybridized to about 200 ng of human DNA target (4×10⁴ cells), 13 ng of which is chromosome 4. Thus, the ratio of probe to target for the specific sequences was about 1, whereas for the shared sequences it was about 0.05.]

The contrast can be increased by allowing the denatured probe DNA to partially reassociate prior to adding it to the slide, preferentially depleting the single-stranded high-copy (predominantly the shared) sequences in the probe [Cantor & Schimmel, Biophysical Chemistry: The Behavior of Biological Macromolecules, (part III, p. 1228) (Freeman 1980)]. A significant increase in staining specificity resulting from probe reassociation was observed experimentally for chromosome 4 using a hybridization mixture with 1 ng of probe per microliter (ul) and a 24 hr incubation at 37° C. prior to in situ hybridization (not shown). Likewise, hybridization after a 24 hr incubation of 4 ng of chromosome 21 probe per ul resulted in a substantial contrast ratio. That result indicates that at such concentrations the chromosome-specific sequences remain substantially single stranded for times on the order of days in the hybridization mixture. It also demonstrates that other mechanisms that might inactivate the probe are not significant during the incubation.

FIGS. 4B and 4C show the result of a protocol I hybridization [0.8 ng of probe per ul and 24 ng of genomic DNA per ul (Q≦2); 1-hr probe incubation and 110-hr hybridization]. Quantitative image analysis shows that the intensity per unit length of the FITC fluorescein on chromosome 4 is approximately 20 times that of the other chromosomes, that is the contrast ratio is 20:1. Two layers of avidin-fluorescein isothiocyanate have been used here to make the non-target chromosomes sufficiently bright to be measured accurately. However, the number 4 chromosomes can be recognized easily after a single layer.

FIG. 4D demonstrates detection of a radiation-induced translocation involving chromosome 4 in human lymphocytes [protocol I, 1′ ng of probe per ul and 76 ng of genomic DNA per ul (Q=5); 1-hr probe incubation and 16-hr hybridization]. The contrast ratio was about 5. The hybridization intensity and specificity shown in FIG. 4D are such that even small portions of the involved chromosome can be detected.

The ease with which translocations can be recognized offers the opportunity for translocation detection by automated means, such as, computerized microscopy or flow cytometry. [See Section VIII infra for elaboration concerning automated detection means.]

FIG. 4E shows that the normal and two derivative chromosomes resulting from the translocation between chromosomes 4 and 11 [t(4;11)] in cell line RS4;11 can be detected in interphase nuclei as three distinct domains [protocol I, 13.5 ng of probe per ul and 800 ng of genomic DNA per ul (Q=5); 1-hr probe incubation and 16-hr hybridization]. The increased probe concentration resulted in brighter signals relative to FIG. 4D. Approximately half of the cells clearly show the presence of three nuclear domains, presumably produced by the two portions of the involved chromosome 4 and the intact normal chromosome. The domains in the other nuclei may have been obscured by the nuclear orientation in these two-dimensional views, by nuclear distortion that occurred during slide preparation, or because the domains were too close to each other to be distinguished. Hybridization using procedures that preserve three-dimensional morphology may resolve these issues and also permit general studies of chromosomal domains in interphase nuclei [Trask et al., Hum. Genet., 78:251 (1988)].

Hybridization of pBS21 to a metaphase spread from a cell line with trisomy 21 is shown in FIG. 4F [protocol II, 4 ng of probe per ul and 250 ng of genomic DNA per ul; 3-hr incubation, additional 250 ng of genomic DNA per ul (Q=1+1); 3-hr probe incubating and 16-hr hybridization]. A small amount of hybridization is visible near the centromeres of the other acrocentric chromosomes.

FIG. 4G shows two interphase nuclei from the same hybridization which clearly show the three chromosome 21 domains. Hybridization with probe prepared according to protocol I resulted in higher relative intensity of the shared signals on the D- and G-group chromosomes, and consequently it was more difficult to determine the number of number 21 chromosomes in interphase (not shown). Increasing stringency by using a hybridization mixture with 55% formamide and 0.15 M NaCl/15 mM sodium citrate, which lowers the melting temperature about 8° C., did not reduce the unwanted hybridization. Addition of unlabeled pA ribosomal DNA [Erikson et al., Gene, 16:1 (1981)] also was ineffective at increasing specificity.

The centromeric region of the D- and G-group chromosomes contain ribosomal [Erikson et al., id] and alpha satellite sequences and perhaps others [Choo et al., Nucleic Acids Res., 16: 1273 (1988)]. These are relatively low copy sequences shared with only a few chromosomes, so Protocol I is not very effective at suppressing them relative to the chromosome 21-specific sequences. In addition, these sequences are clustered on the chromosomes, so that even much reduced hybridization is clearly visible. This is especially distracting in analysis of interphase nuclei. Calculations indicate that addition of several aliquots of freshly denatured genomic DNA periodically during the incubation (protocol II) should increase the staining specificity. FIG. 4F shows a protocol II hybridization, using two aliquots of genomic DNA, to a metaphase spread from a trisomy 21 cell line. Intense hybridization to the three number 21 chromosomes is clearly visible and hybridization to the other D- and G-group chromosomes has been reduced to an acceptable level. FIG. 4G shows that hybridization to chromosomes other than chromosome 21 is sufficiently low that the three chromosome 21 domains are clearly visible in interphase nuclei. In practice, the most convenient procedure for suppressing the shared acrocentric hybridization might be inclusion of unlabeled DNA from one of the other D- or G-group chromosome libraries (or unlabeled cloned DNA from just these sequences, if available) as additional competitor. The use of libraries from non-target chromosomes as blocker for a probe may in general improve contrast. The specific sequences in the probe will not be blocked (Q=0) no matter how much competitor for the shared sequences is added.

VI.C. Hybridization of Yeast Artificial Chromosomes (YACS) to Human Metaphase Spread

YACS. Seven yeast clones HY1, HY19, HY29, HYA1.A2, HYA3.A2, HYA3.A9, and HYA9E6 were obtained from D. Burke (Washington University, St. Louis, Mo.). The lengths of the human DNA in the clones ranged from about 100 kb to about 600 kb. Gel electrophoresis was performed to verify the size of these inserts. Each of these clones was grown up and total DNA was isolated. The isolated DNA was biotinylated by nick translation so that 10-30% of the thymidine was replaced by biotin-11-dUTP. The concentration of the total labeled DNA after nick translations is in the range of 10-20 ng/ul.

Blocking DNA. Human placental DNA (Sigma) was treated with proteinase K and extracted with phenol and sonicated to a size range of 200-600 bp. Total DNA isolated from yeast not containing an artificial chromosome was sonicated to a similar size range. Both of these DNA's were maintained at a concentration of 1-10 ug/ul.

Fluorescence in situ hybridization (FISH). Hybridization followed the procedures of Pinkel et al. (1988), supra (as exemplified in Sections V and VI, supra) with slight modifications. Metaphase spreads were prepared from methotrexate synchronized cultures according to the procedures of Harper et al. PNAS (USA) 78: 4458-4460, (1981). Cells were fixed in methanol/acetic acid, fixed (3:1), dropped onto slides, air dried, and stored at −20° C. under nitrogen gas until used. The slides were then immersed two minutes in 70% formamide/2×SSC to denature the target DNA sequences, dehydrated in a 70-85-100% ethanol series, and air dried. (SSC is 0.15 M NaCl/0.015 M′Na Citrate, pH 7). Ten—100 ng of biotinylated yeast DNA, and approximately 1 ug each of unlabeled yeast and human genomic DNA were then added to the hybridization mix (final volume 10 ul, final composition 50% formamide/2×SSC/10% dextran sulfate), heated to 70° C. for 5 min., and then incubated at 37° C. for 1 hr to allow the complementary strands of the more highly repeated sequences to reassociate.

The hybridization mixture was then applied to the slide (approximately 4 cm² area) and sealed with rubber cement under a glass cover slip. After overnight incubation at 37° C. the coverslip was removed and the slide washed 3 times 3 min each in 50% formamide/2×SSC at 42-45° C., and once in PN buffer [mixture of 0.1 M NaH₂PO₄ and 0.1 M Na₂HPO₄ to give pH 8; 0.1% Nonidet P-40 (Sigma)]. The bound probe was then detected with alternating 20 min incubations (room temperature in avidin-FITC and goat-anti-avidin antibody, both at 5 ug/ml in PNM buffer (PN buffer plus 5% nonfat dry milk, centrifuged to remove solids; 0.02% Na azide). Avidin and anti-avidin incubation were separated by 3 washes of 3 min each in PN buffer. Two or three layers of avidin were applied (Avidin, DCS grade, and biotinylated goat-anti-avidin are obtained from Vector Laboratories Inc., Burlingame, Calif.).

FIG. 5 shows the hybridization of HYA3.A2 (580 kb of human DNA) to 12q21.1. The location of the hybridization was established by using a conventional fluorescent banding technique employing the DAPI/actinomycin D procedure: Schweizer, “Reverse fluorescent chromosome banding with chromomycin and Dapi,” Chromosoma, 58:307-324 (1976). The hybridization signal forms a band across the width of each of the chromosome 12s, indicating the morphology of the packing of DNA in that region of the chromosome.

The YAC clone positions are attributed as shown in Table 2 below.

TABLE 2
YAC Competition Hybridization
	YAC Clone	Insert Size	Localization
	HY1	120	Xq23
	HY19	450	8q23.3
			21q21.1
	HY27	500	14q12
	HYA1.A2	250	6q16
	HYA3.A2	580	12q21.1
	HYA3.A9	600	14q21
	HYA9.E6	280	1p36.2
			3q22

VI.D. Hybridization With Human/Hamster Hybrid Cell

Essentially the same hybridization and staining conditions were used in this example as for those detailed in the procedure of Pinkel et al. (1988), supra and exemplified in Sections V.C. and VI.B., supra. in this example, 400 ng of biotin labeled DNA from a hamster-human hybrid cell that contains one copy of human chromosome 19 was mixed with 1.9 ug of unlabeled human genomic DNA in 10 ul of hybridization mix. Hybridization was for approximately 60 hours at 37° C. Fluorescent staining of the bound probe and counterstaining of the chromosomes was as in the other examples above. FIG. 6 shows the results of the hybridization.

VII. Specific Applications

The present invention allows microscopic and in some cases flow cytometric detection of genetic abnormalities on a cell by cell basis. The microscopy can be performed entirely by human observers, or include various degrees of additional instrumentation and computational assistance, up to full automation. The use of instrumentation and automation for such analyses offers many advantages. Among them are the use of fluorescent dyes that are invisible to human observers (for example, infared dyes), and the opportunity to interpret results obtained with multiple labeling methods which might not be simultaneously visible (for example, combinations of fluorescent and absorbing stains, autoradiography, etc.) Quantitative measurements can be used to detect differences in staining that are not detectable by human observers. As is described below, automated analysis can also increase the speed with which cells and chromosomes can be analysed.

The types of cytogenetic abnormalities that can be detected with the probes of this invention include: Duplication of all or part of a chromosome type can be detected as an increase in the number or size of distinct hybridization domains in metaphase spreads or interphase nuclei following hybridization with a probe for that chromosome type or region, or by an increase in the amount of bound probe. If the probe is detected by fluorescence, the amount of bound probe can be determined either flow cytometrically or by quantitative fluorescence microscopy. Deletion of a whole chromosome or chromosome region can be detected as a decrease in the number or size of distinct hybridization domains in metaphase spreads or interphase nuclei following hybridization with a probe for that chromosome type or region, or by a decrease in the amount of bound probe. If the probe is detected by fluorescence, the amount bound can be determined either flow cytometrically or by quantitative fluorescence microscopy. Translocations, dicentrics and inversion can be detected in metaphase spreads and interphase nuclei by the abnormal juxtaposition of hybridization domains that are normally separate following hybridization with probes that flank or span the region(s) of the chromosome(s) that are at the point(s) of rearrangement. Translocations involve at least two different chromosome types and result in derivative chromosomes possessing only one centromere each. Dicentrics involve at least two different chromosome types and result in at least one chromosome fragment lacking a centromere and one having two centromeres. Inversions involve a reversal of polarity of a portion of a chromosome.

VII.A Banding Analysis

Substantial effort has been devoted during the past thirty years to development of automated systems (especially computer controlled microscopes) for automatic chromosome classification and aberration detection by analysis of metaphase spreads. In recent years, effort has been directed at automatic classification of chromosomes which have been chemically stained to produce distinct banding patterns on the various chromosome types. These efforts have only partly succeeded because of the subtle differences in banding pattern between chromosome types of approximately the same size, and because differential contraction of chromosomes in different metaphase spreads causes a change in the number and width of the bands visible on chromosomes of each type. The present invention overcomes these problems by allowing construction of reagents which produce a staining pattern whose spacing, widths and labeling differences (for example different colors) are optimized to facilitate automated chromosome classification and aberration detection. This is possible because hybridization probes can be selected as desired along the lengths of the chromosomes. The size of a band produced by such a reagent may range from a single small dot to a substantially uniform coverage of one or more whole chromosomes. Thus the present invention allows construction of a hybridization probe and use of labeling means, preferably fluorescence, such that adjacent hybridization domains can be distinguished, for example by color, so that bands too closely spaced to be resolved spatially can be detected spectrally (i.e. if red and green fluorescing bands coalesce, the presence of the two bands can be detected by the resulting yellow fluorescence).

The present invention also allows construction of banding patterns tailored to particular applications. Thus they can be significantly different in spacing and color mixture, for example, on chromosomes that are similar in general shape and size and which have similar banding patterns when conventional techniques are used. The size, shape and labeling (e.g. color) of the hybridization bands produced by the probes of the present invention can be optimized to eliminate errors in machine scoring so that accurate automated aberration detection becomes possible. This optimized banding pattern will also greatly improve visual chromosome classification and aberration detection.

The ease of recognition of specific translocation breakpoints can be improved by using a reagent closely targeted to the region of the break. For example, a high complexity probe of this invention comprising sequences that hybridize to both sides of the break on a chromosome can be used. The portion of the probe that binds to one side of the break can be detected differently than that which binds to the other, for example with different colors. In such a pattern, a normal chromosome would have the different colored hybridization regions next to each other, and such bands would appear dose together. A break would separate the probes to different chromosomes or result in chromosomal fragments, and could be visualized as much further apart on an average.

VII.B Biological Dosimetry

One approach to biological dosimetry is to measure frequencies of structurally aberrant chromosomes as an indication of the genetic damage suffered by individuals exposed to potentially toxic agents. Numerous studies have indicated the increase in structural aberration frequencies with increasing exposure to ionizing radiation and other agents, which are called clastogens. Dicentric chromosomes are most commonly scored because their distinctive nature allows them to be scored rapidly without banding analysis. Rapid analysis is important because of the low frequency of such aberrations in individuals exposed at levels found in workplaces (˜2×10⁻³/cell). Unfortunately, dicentrics are not stably retained so the measured dicentric frequency decreases with time after exposure. Thus low level exposure over long periods of time does not result in an elevated dicentric frequency because of the continued clearance of these aberrations. Translocations are better aberrations to score for such dosimetric studies because they are retained more or less indefinitely. Thus, assessment of genetic damage can be made at times long after exposure. Translocations are not routinely scored for biological dosimetry because the difficulty of recognizing them makes scoring sufficient cells for dosimetry logistically impossible.

The present invention eliminates this difficulty. Specifically, hybridization with a probe which substantially uniformly stains several chromosomes (e.g. chromosomes 1, 2, 3 and 4) allows immediate microscopic identification in metaphase spreads of structural aberrations involving these chromosomes. Normal chromosomes appear completely stained or unstained by the probe. Derivative chromosomes resulting from translocations between targeted and non-targeted chromosomes are recognized as being only partly stained, FIG. 4D. Such partially hybridized chromosomes can be immediately recognized either visually in the microscope or in an automated manner using computer assisted microscopy. Discrimination between translocations and dicentrics is facilitated by adding to the probe, sequences found at all of the chromosome centromeres. Detection of the centromeric components of the probe with a labeling means, for example color, different from that used to detect the rest of the probe elements allows ready identification of the chromosome centromeres, which in turn facilitates discrimination between dicentrics and translocations. This technology dramatically reduces the scoring effort required with previous techniques so that it becomes feasible to examine tens of thousands of metaphase spreads as required for low level biological dosimetry.

VII.C. Prenatal Diagnosis

The most common aberrations found prenatally are trisomies involving chromosomes 21 (Down syndrome), 18 (Edward syndrome) and 13 (Patau syndrome) and X0 (Turner syndrome), XXY (Kleinfelter syndrome) and XYY disease. Structural aberrations also occur. However, they are rare and their clinical significance is often uncertain. Thus, the importance of detecting these aberrations is questionable. Current techniques for obtaining fetal cells for conventional karyotyping, such as, amniocentesis and chorionic villus biopsy yield hundreds to thousands of cells for analysis. These are usually grown in culture for 2 to 5 weeks to produce sufficient mitotic cells for cytogenetic analysis. Once metaphase spreads are prepared, they are analyzed by conventional banding analysis. Such a process can only be carried out by highly skilled analysts and is time consuming so that the number of analyses that can be reliably carried out by even the largest cytogenetics laboratories is only a few thousand per year. As a result, prenatal cytogenetic analysis is usually limited to women whose children are at high risk for genetic disease (e.g. to women over the age of 35).

The present invention overcomes these difficulties by allowing simple, rapid identification of common numerical chromosome aberrations in interphase cells with no or minimal cell culture. Specifically, abnormal numbers of chromosomes 21, 18, 13, X and Y can be detected in interphase nuclei by counting numbers of hybridization domains following hybridization with probes specific for these chromosomes (or for important regions thereof such as 21q22 for Down syndrome). A hybridization domain is a compact, distinct region over which the intensity of hybridization is high. An increased frequency of cells showing three domains (specifically to greater than 10%) for chromosomes 21, 18 and 13 indicates the occurrence of Down, Edward and Patau syndromes, respectively. An increase in the number of cells showing a single X-specific domain and no Y-specific domain following hybridization with X-specific and Y-specific probes indicates the occurrence of Turner syndrome. An increase in the frequency showing two X-specific domains and one Y-specific domain indicates Kleinfelter syndrome, and increase in the frequency of cells showing one X-specific domain and two Y-specific domains indicates an XYY fetus. Domain counting in interphase nuclei can be supplemented (or in some cases replaced) by measurement of the intensity of hybridization using, for example, quantitative fluorescence microscopy or flow cytometry, since the intensity of hybridization is approximately proportional to the number of target chromosomes for which the probe is specific. Numerical aberrations involving several chromosomes can be scored simultaneously by detecting the hybridization of the different chromosomes with different labeling means, for example, different colors. These aberration detection procedures overcome the need for extensive cell culture required by procedures since all cells in the population can be scored. They eliminate the need for highly skilled analysts because of the simple, distinct nature of the hybridization signatures of numerical aberrations. Further, they are well suited to automated aberration analysis.

The fact that numerical aberrations can be detected in interphase nuclei also allows cytogenetic analysis of cells that normally cannot be stimulated into mitosis. Specifically, they allow analysis of fetal cells found in maternal peripheral blood. Such a feature is advantageous because it eliminates the need for invasive fetal cell sampling such as amniocentesis or chorionic villus biopsy.

As indicated in the Background, the reason such embryoinvasive methods are necessary is that conventional karyotyping and banding analysis requires metaphase chromosomes. At this time, there are no accepted procedures for culturing fetal cells separated from maternal blood to provide a population of cells having metaphase chromosomes. In that the staining reagents of this invention can be employed with interphase nuclei, a non-embryo-invasive method of karyotyping fetal chromosomes is provided by this invention.

The first step in such a method is to separate fetal cells that have passed through the placenta or that have been shed by the placenta into the maternal blood. The incidence of fetal cells in the maternal bloodstream is very low, on the order of 10⁻⁴ to 10⁻⁶ cells/ml and quite variable depending on the time of gestation; however, appropriately marked fetal cells may be distinguished from maternal cells and concentrated, for example, with high speed cell sorting.

The presence of cells of a male fetus may be identified by a label, for example a fluorescent tag, on a chromosome-specific staining reagent for the Y chromosome. Cells that were apparently either lymphocytes or erythrocyte precursors that were separated from maternal blood where shown to be Y-chromatin-positive. [Zillacus et al., Scan. J. Haematol, 15: 333 (1975); Parks and Herzenberg, Methods in Cell Biology, Vol. 10, pp. 277-295 (Academic Press, N.Y., 1982); and Siebers et al., Humangenetik, 28: 273 (1975)].

A preferred method of separating fetal cells from maternal blood is the use of monoclonal antibodies which preferentially have affinity for some component not present upon the maternal blood cells. Fetal cells may be detected by paternal HLA (human leukocyte antigen) markers or by an antigen on the surface of fetal cells. Preferred immunochemical procedures to distinguish between fetal and maternal leukocytes on the basis of differing HLA type use differences at the HLA-A2, -A3, and -B7 loci, and further preferred at the -A2 locus. Further, first and second trimester fetal trophoblasts may be marked with antibody against the internal cellular constituent cytokeratin which is not present in maternal leukocytes. Exemplary monoclonal antibodies are described in the following references.

Herzenberg et al., PNAS, 76: 1453 (1979), reports the isolation of fetal cells, apparently of lympoid origin, from maternal blood by fluorescence activated cell sorting (FACS) wherein the separation was based on the detection of labeled antibody probes which bind HLA-A2 negative cells in maternal blood. Male fetal cells separated in that manner were further identified by quinacrine staining of Y-chromatin.

Covone et al., Lancet, Oct. 13, 1984: 841, reported the recovery of fetal trophoblasts from maternal blood by flow cytometry using a monoclonal antibody termed H315. Said monoclonal reportedly identifies a glycoprotein expressed on the surface of the human syncytiotrophoblast as well as other trophoblast cell populations, and that is absent from peripheral blood cells.

Kawata et al., J. Exp. Med., 160: 653 (1984), discloses a method for isolating placental cell populations from suspensions of human placenta. The method uses coordinate two-color and light-scatter FACS analysis and sorting. Five different cell populations were isolated on the basis of size and quantitative differences in the coordinate expression of cell surface antigens detected by monoclonal antibodies against an HLA-A, B, C monomorphic determinant (MB40.5) and against human trophoblasts (anti-Trop-1 and anti-Trop-2).

Loke and Butterworth, J. Cell Sci., 76: 189 (1985), describe two monoclonal antibodies, 18B/A5 and 18A/C4, which are reactive with first trimester cytotrophoblasts and other fetal epithelial tissues including syncytiotrophoblasts.

A preferred monoclonal antibody to separate fetal cells from maternal blood for staining according to this invention is the anti-cytokeratin antibody Cam 5.2, which is commercially available from Becton-Dickinson (Franklin Lakes, N.J., USA).

Other preferred monoclonal antibodies for separating fetal cells from maternal blood are those disclosed in co-pending, commonly owned U.S. patent application, U.S. Ser. No. 389,224, filed Aug. 3, 1989, entitled “Method for Isolating Fetal Cytotrophoblast Cells”. [See also: in Fisher et al., J. Cell. Biol., 109 (2): 891-902 (1989)]. The monoclonal antibodies disclosed therein react specifically with antigen on first trimester human cytotrophoblast cells, which fetal cells have the highest probability of reaching the maternal circulation. Said application and article are herein specifically incorporated by reference. Briefly, the disclosed monoclonal antibodies were raised by injection of test animals with cytotrophoblast cells obtained from sections of the placental bed, that had been isolated by uterine aspiration. Antibodies raised were subjected to several cytological screens to select for those antibodies which react with the cytotrophoblast stem cell layer of first trimester chorionic villi.

Preferred monoclonal antibodies against such first trimester cytotrophoblast cells disclosed by Fisher et al. include monoclonal antibodies produced from the following hybridomas deposited at the American Tyupe Culture Collection (ATCC; Rockville, Md., USA) under the Budapest Treaty:

Hybridoma ATCC Accession #

J1D8      HB10096
P1B5      HB10097

Both hybridoma cultures were received by the ATCC on Apr. 4, 1989 and reported viable thereby on Apr. 14, 1989.

Fisher et al. state that fetal cells isolated from maternal blood by use of said monoclonal antibodies are capable of replication in vitro. Therefore, fetal cells isolated by the method of Fisher et al., that is, first trimester fetal cytotrophoblasts, may provide fetal chromosomal material that is both in metaphase and in interphase.

The fetal cells, preferably leukocytes and cytotrophoblasts, more preferably cytotrophoblasts, once marked with an appropriate antibody are then separated from the maternal cells either directly or by preferably separating and concentrating said fetal cells by cell sorting or panning. For example, FACS may be used to separate fluorescently labeled fetal cells, or flow cytometry may be used.

The fetal cells once separated from the maternal blood can then be stained according to the methods of this invention with appropriate chromosome-specific staining reagents of this invention, preferably those of particular importance for prenatal diagnosis. Preferred staining reagents are those designed to detect aneuploidy, for example, trisomy of any of several chromosomes, including chromosome types 21, 18, 13, X and Y and subregions on such chromosomes, such as, subregion 21q22 on chromosome 21.

Preferably, a fetal sample for staining analysis according to this invention comprises at least 10 cells or nuclei, and more preferably about 100 cells or nuclei.

VII.D Tumor Cytogenetics

Numerous studies in recent years have revealed the existence of structural and numerical chromosome aberrations that are diagnostic for particular disease phenotypes and that provide clues to the genetic nature of the disease itself. Prominent examples include the close association between chronic myelogeneous leukemia and a translocation involving chromosome 9 and 22, the association of a deletion of a portion of 13q14 with retinoblastoma and the association of a translocation involving chromosomes 8 and 14 with Burkitts lymphoma. Current progress in elucidating new tumor specific abnormalities is limited by the difficulty of producing representative, high quality banded metaphase spreads for cytogenetic analysis. These problems stem from the fact that many human tumors are difficult or impossible to grow in culture. Thus, obtaining mitotic cells is usually difficult. Even if the cells can be grown in culture, there is the significant risk that the cells that do grow may not be representative of the tumorigenic population. That difficulty also impedes the application of existing genetic knowledge to clinical diagnosis and prognosis.

The present invention overcomes these limitations by allowing detection of specific structural and numerical aberrations in interphase nuclei. These aberrations are detected as described supra. Hybridization with whole chromosome probes will facilitate identification of previously unknown aberrations thereby allowing rapid development of new associations between aberrations and disease phenotypes. As the genetic nature of specific malignancies becomes increasingly well known, the interphase assays can be made increasingly specific by selecting hybridization probes targeted to the genetic lesion. Translocation at specific sites on selected chromosomes can be detected by using hybridization probes. that closely flank the breakpoints. Use of these probes allows diagnosis of these specific disease phenotypes. Translocations may be detected in interphase because they bring together hybridization domains that are normally separated, or because they separate a hybridization domain into two, well separated domains. In addition, they may be used to follow the reduction and reemergence of the malignant cells during the course of therapy. Interphase analysis is particularly important in such a application because of the small number of cells that may be present and because they may be difficult or impossible to stimulate into mitosis.

Duplications and deletions, processes involved in gene amplification and loss of heterozygosity, can also be detected in metaphase spreads and interphase nuclei using the techniques of this invention. Such processes are implicated in an increasing number of different tumors.

VIII. Detection of BCR-ABL Fusion in Chronic Myelogenous Leukemia (CML)

Probes. This section details a CML assay based upon FISH with probes from chromosomes 9 and 22 that flank the fused BCR and ABL sequences in essentially all cases of CML (FIG. 8). The BCR and ABL probes used in the examples of this section were kindly provided by Carol A. Westbrook of the Department of Medicine, Section of Hematology/Oncology at the University of Chicago Medical Center in Chicago, Ill. (USA).

The ABL probe on chromosome 9, c-hu-ABL, is a 35-kb cosmid (pCV105) clone selected to be telomeric to the 200-kb region of ABL between exons IB and II in which the breaks occur (24). The BCR probe on chromosome 22, PEM12, is an 18-kb phage clone (in EMBL3) that contains part of, and extends centromeric to, the 5.8-kb breakpoint duster region of the BCR gene in which almost all CML breakpoints occur. FISH was carried out using a biotin labeled ABL probe, detected with the fluorochrome Texas red, and a digoxigenin labeled BCR probe, detected with the green fluorochrome FITC. Hybridization of both probes could be observed simultaneously using a fluorescence microscope equipped with a double band pass filter set (Omega Optical).

FIG. 8 is a schematic representation of the BCR gene on chromosome 22, the ABL gene of chromosome 9, and the BCR-ABL fusion gene on the Philadelphia chromosome, showing the location of CML breakpoints and their relation to the probes. Exons of the BCR gene are depicted as solid boxes. The Roman numeral I refers to the first exon of the BCR gene; the arabic numerals 1-5 refer to the exons within the breakpoint cluster region, here indicated by the dashed line. The approximate location of the 18 kb phage PEM12 probe (the BCR probe) is indicated by the open horizontal bar. Since the majority of breakpoints in CML occur between exons 2 and 4,15 kb or more of target for PEM12 will remain on the Philadelphia chromosome. In the classic reciprocal translocation a few kb of target for PEM12 (undetectable fluorescent signal) will be found on the derivative chromosome. The map and exon numbering (not to scale) is adapted from Heisterkamp et al. (ref. 34, supra).

Exons of the ABL gene are depicted as open vertical bars (not to scale). The Roman numerals Ia and Ib refer to the alternative first exons, and II to the second exon. Exon II is approximately 25 kb upstream of the end of the 28 kb cosmid c-hu-abl (the ABL probe). All CML breakpoints occur upstream of exon II, usually between exons Ib and Ia, within a region that is approximately 200kb in length. Thus, c-hu-abl will always be 25 to 200 kb away from the fusion junction. The map (not to scale) is adapted from Heisterkamp et al. (ref. 35, supra). The BCR-ABL fusion gene is depicted. In CML, PEM12 will always lie at the junction, and c-hu-abl will be separated from PEM12 by 25 to 225 kb.

Sample Preparation: CML-4: peripheral blood was centrifuged for 5 min. Ten drops of interface was diluted with PBS, spun down, fixed in methanol/acetic acid (3:1), and dropped on slides. CML-2, 3, 7: Five to 10 drops of marrow diluted with PBS to prevent clotting were fixed in methanol/acetic acid and dropped on slides. CML-1, 4, 5, 6: Peripheral blood and/or bone marrow was cultured in RPMI 1640 supplemented with 10% fetal calf serum, an antibiotic mixture (gentamycin 500 mg/ml), and 1% L-glutamine for 24 h. Cultures were synchronized according to J. J. Yunis and M. E. Chandler Prog. in Clin. Path., 7:267 (1977), and chromosome preparations followed Gibis and Jackson, Karyogram, 11:91 (1985).

Hybridization and Detection Protocol. Hybridization followed procedures described by D. Pinkel et al. (27), Trask et al. (25), and J. B. Lawrence et al (30), with modifications. The BCR probe was nick-translated (Bethesda Research Laboratories Nick-Translation System) with digoxigenin-11-dUTP (Boehringer Mannheim Biochemicals) with an average incorporation of 25%. The ABL probe was similarly nick-translated with biotin-11-dUTP (Enzo Diagnostics).

1. Hybridization. Denature target interphase cells and/or metaphase spreads on glass slides at 72° C. in 70% formamide/2×SSC at pH 7 for 2 min. Dehydrate in an ethanol series (70%, 85%, and 100% each for 2 min.). Air dry and place at 37° C. (2×SSC is 0.3M NaCl/30 mM sodium citrate). Heat 10 ml of hybridization mixture containing 2 ng/ml of each probe, 50% formamide/2×SSC, 10% dextran sulphate, and 1 mg/ml human genomic DNA (sonicated to 200-600 bp) to 70° C. for 5 min. to denature the DNA. Incubate for 30 min. at 37° C. Place on the warmed slides, cover with a 20 mm×20 mm coverslip, seal with rubber cement, and incubate overnight in a moist chamber at 37° C. Remove coverslips and wash three times for 20 minutes each in 50% formamide/2×SSC pH 7 at 42° C., twice for 20 minutes each in 2×SSC at 42° C., and finally rinse at room temperature in 4×SSC.

2. Detection of Bound Probes: All incubation steps are performed with approximately 100 ml of solution at room temperature under coverslips. The biotinylated ABL probe was detected first, then the digoxigenin-labeled BCR probe.

a. Biotinylated ABL Probe: Preblock with 4×SSC/1% bovine serum albumin (BSA) for 5 min. Apply Texas Red-avidin (Vector Laboratories Inc., 2 mg/ml in 4×SSC/1% BSA) for 45 min. Wash in 4×SSC once, 4×SSC/1% Triton-X 100 (Sigma) and then again in 4×SSC, 5 min. each. Preblock for 5 min. in PNM (PN containing 5% non-fat dry milk and 0.02% sodium azide and centrifuged to remove solids. PN is 0.1 M NaH₂PO₄/0.1M Na₂HPO₄, 0.05% NP40, pH 8). Apply biotinylated goat anti-avidin (Vector Laboratories Inc., 5 mg/ml in PNM) for 45 min. Wash twice in PN for 5 min. Apply a second layer of Texas Red-avidin (2 mg/ml in PNM) for 45 min. Wash twice in PN for 5 min. each.

b. Digoxigenin-Labeled BCR Probe: Preblock with PNM for 5 min. Apply sheep anti-digoxigenin antibody (obtained from D. Pepper, Boehringer Mannheim Biochemicals, Indianapolis, Ind.; 15.4 mg/ml in PNM) for 45 min. Wash twice in PN for 5 min. each. Preblock with PNM for 5 min. Apply rabbit-anti-sheep antibody conjugated with FITC (Organon Teknika-Cappel, 1:50 in PNM) for 45 min. Wash twice for 5 min. each in PN. If necessary, the signal is amplified by preblocking for 5 min. with PNM and applying sheep anti-rabbit IgG antibody conjugated to FITC (Organon Teknika-Cappel, 1:50 in PNM) for 45 min. Rinse in PN.

3. Visualization: The slides are mounted fluorescence antifade solution [G. D. Johnson and J. G. Nogueria, J. Immunol. Methods, 43:349 (1981)) (ref. 31, supra)] containing 1 mg/ml 4′,6-amidino-2-phenylindole (DAPI) as a counterstain, and examined using a FITC/Texas red double-band pass filter set (Omega Optical) on a Zeiss Axioskop.

The method used for BCR-ABL PCR tested herein was that described in Hegewisch-Becker et al. for CML-3, 4 and 7 (ref. 32, supra), and Kohler et al., for CML-5 and 6 (ref. 33, supra).

Results. ABL and BCR hybridization sites were visible on both chromatids of chromosomes in most metaphase spreads. The ABL probe bound to metaphase spreads from normal individuals (FIG. 9A) near the telomere on 9q while the BCR probe bound at 22q11 (FIG. 9B). Hybridization with the ABL or BCR probe to normal interphase nuclei typically resulted in two tiny fluorescent dots corresponding to the target sequence on both chromosome homologues. The spots were apparently randomly distributed in the two dimensional nuclear images and were usually well separated. A few cells showed two doublet hybridization signals probably a result of hybridization to both sister chromatids of both homologues in cells which had replicated this region of DNA (i.e., those in the S- or G2-phase of cell cycle). Dual color FISH of the ABL (red) and BCR (green) probes to normal G1 nuclei yielded two red (ABL) and two green (BCR) hybridization signals distributed randomly around the nucleus.

The genetic rearrangement of CML brings the DNA sequences homologous to the probes together on an abnormal chromosome, usually the Ph¹, and together in the interphase nucleus, as illustrated in FIG. 8. The genomic distance between the probe binding sites in the fusion gene varies among CML cases, ranging from 25 to 225 kb, but remains the same in all the cells of a single leukemic done. Dual color hybridization with ABL and BCR probes to interphase CML cells resulted in one red and one green hybridization signal located at random in the nucleus, and one red-green doublet signal in which the separation between the two colors was less than 1 micron (or one yellow hybridization signal for hybridization in very close proximity, see FIG. 10). The randomly located red and green signals are ascribed to hybridization to the ABL and BCR genes on the normal chromosomes, and the red-green doublet signal to hybridization to the BCR-ABL fusion gene. Interphase mapping studies suggest that DNA sequences separated by less than 250 kb should be separated in interphase nuclei by less than 1 micron (25). As a result, cells showing red and green hybridization signals separated by greater than 1 micron were scored as normal since this is consistent with the hybridization sites being on different chromosomes. However, due to statistical considerations, some normal cells will have red and green dots close enough together to be scored as abnormal. In these two dimensional nuclear analyses, 9 out of 750 normal nuclei had red and green hybridization signals less than 1 micron of each other. Thus, approximately 1% of normal cells were classified as abnormal.

Table 3 shows the hybridization results for 7 samples from 6 CML cases along with conventional karyotypes, and other diagnostic results (PCR and Southern blot data ). All six cases, including 3 that were found to be Ph¹ negative by banding analysis (CML-5,-6 and -7), showed red-green hybridization signals separated by less than 1 micron in greater than 50% of nuclei examined. In most, the fusion event was visible in almost every cell. One case (CML-7) showed fusion signals in almost every cell even though PCR analysis failed to detect the presence of a fusion gene and banding analysis did not reveal a Philadelphia chromosome.

Hybridization to metaphase spreads was performed in three cases (CML-1,-5 and -6). All of these showed red and green hybridization signals in clone proximity on a single acrocentric chromosome. In two cases, scored as t(9:22) (q34;q11) by banding, the red-green pair was in close proximity to the telomere of the long arm of a small acrocentric chromosome as expected for the Ph¹ (FIG. 9C). One case (CML-6) was suspected by classical cytogenetics to have an insertion of chromosomal material at 22q11. Dual color hybridization to metaphase spreads from this case showed the red-green pair to be centrally located in a small chromosome (FIG. 9D). That result is consistent with formation of the BCR-ABL fusion gene by an insertion. In one case (CML-1), two pairs of red-green doublet signals were seen in 3 out 150 (2%) interphase nuclei. That may indicate a double Ph¹ (or double fusion gene) in those cells. Such an event was not detected by standard cytogenetics, which was limited to analysis of 25 metaphase spreads. The acquisition of an additional Ph¹ is the most frequent cytogenetic event accompanying blast transformation, and its cytogenetic detection may herald disease acceleration.

Simultaneous hybridization with ABL and BCR probes to metaphase spreads of the CML derived cell line K-562 showed multiple red-green hybridization sites along both arms of a single acrocentric chromosome. Hybridization to interphase nuclei showed that the red and green signals were confined to the same region of the nucleus. That is consistent with their being localized on a single chromosome. Twelve to fifteen hybridization pairs were seen in each nucleus indicating corresponding amplification of the BCR-ABL fusion gene (see FIGS. 9E and 9F). These findings are consistent with previous Southern blot data showing amplification of the fusion gene in this cell line (26).

In summary, analysis of interphase cells for seven CML, and four normal cell samples using dual color FISH with ABL and BCR probes suggests the utility of this approach for routine diagnosis of CML and clinical monitoring of the disease. Among its very important advantages are the ability to obtain genetic information from individual interphase or metaphase cells in less than 24 hours. Thus, it can be applied to all cells of a population, not just to those that fortuitously or through culture, happen to be in metaphase. Further, the genotypic analysis can be associated with cell phenotype, as judged by morphology or other markers, thereby permitting the study of lineage specificity of cells carrying the CML genotype as well as assessment of the frequency of cells carrying the abnormality.

Random juxtaposition of red and green signals in two dimensional images of normal cells, which occurs in about 0.01 of normal cells, sets the low frequency detection limit. That detection limit may be lowered by more complete quantitative measurement of the separation and intensity of the hybridization signals in each nucleus using computerized image analysis. Such analysis will be particularly important in studying patient populations in which the cells carrying the BCR-ABL fusion at low frequency (e.g., during remission, after bone marrow transplantation, during relapse or in model systems).

This assay also should be advantageous for detection of CML cells during therapy when the number of cells available for analysis is low since only a few cells are required. Finally, simple counting of hybridization spots allows for the detection and quantitative analysis of amplification of the BCR-ABL fusion gene as illustrated for the K562 cell line (FIG. 9E). Quantitative measurement of fluorescence intensity may assist with such an analysis.

IX. Detection of the Retinoblastoma Gene in Metaphase Chromosomes and Interphase Nuclei

Probes. Fourteen lambda phage clones which form three contigs (overlapping nucleic acid sequences) and span all the exons of the Rb-1 gene were obtained from Yuen Kai Fung [Division of Hematology and Oncology, Childrens' Hospital of Los Angeles, University of Southern California, Los Angeles, Calif. 90027 (USA)]. The phage DNA was labeled either by biotin-dUTP or digoxigenin-dUTP using the Bio Nick™ Labeling System [BRL Life Technologies, Inc., Gaithersburg, Md. (USA)].

A probe specific to the 13/21 centromeric alphoid repetitive sequence was used in assisting identification of chromosome 13 in metaphase preparations. The 13/21 centromeric probe was prepared by the polymerase chain reaction (PCR) according to methods detailed in Weier et al., Hum. Genet. 87(4):489-494 (1991). Briefly, the probe was made by PCR using flow sorted human chromosome 21s as a template and two primers (30 μM) specific for the alphoid sequence. The product was labeled during the PCR reaction by including biotin-11-dUTP (100%). Oligonucleotide primers used were W21R1 (5′-GGATAGCTTAACGATTTCGTTGGAAAC-3′) and W21R2 (5′-CAAACGTGCTCAAAGTAAGGGAATG-3′). They were synthesized using phosphoramidite chemistry on a DNA synthesizer (Applied Biophysics, Foster City, Calif., model 380B). Synthesis and further purification of the oligonucleotides by C18 reverse phase chromatography and HPLC was performed according to the specifications of the manufacturer (Waters Chromatography, Milford, Mass., USA). Using the flow sorted chromosomal DNA as a template these primers generate a 135 bp product.

Cell samples. PHA-stimulated normal peripheral blood lymphocytes, cultured human skin fibroblasts, two fibroblast cell lines from retinoblastoma patients GM05877 46, XX, del(13) (pter-q14.1::q21.2-qter) and GM01142A 46, XX, del(13) (pter-q14.1::q22.1-qter) obtained from the NIGMS (National Institute of General Medical Science) Human Genetic Mutant Cell Repository (Camden, N.J.), and clinical human breast cancer samples obtained either by fine-needle aspiration or disaggregation of fresh tumor tissue were used [made available by Fred Waldman, M.D., Department of Laboratory Medicine, University of California, San Francisco, Calif. (USA)]. Cell lines were either treated with colcemid to prepare metaphases or grown to confluency to obtain G0/G1 interphase nuclei. All samples were fixed in 3:1 methanol-acetic acid and dropped on microscope slides. Before in situ hybridization the slides were treated with (1 ug/50 ml) Proteinase K [Boehringer Mannheim GmbH, Indianapolis, Ind.) (USA) for 7.5 min at 37° C.

In situ hybridization. FISH was done using modifications of previously published methods (Pinkel et al., 1986; Trask et al. 1989). The hybridization mixture consisting of 20-40 ng of labeled probe, 5-10 μg unlabeled human placental DNA in 50% formamide, 2×SSC and 10% dextran sulphate was denaturated for 5 min at 70° C. and then allowed to reanneal for 30-60 min at 37° C. In dual color hybridizations, 20 ng of Rb 3′ end probe and 20 ng of Rb 5′ end probe was used. In cohybridizations with the 13/21 centromeric probe, 2 ng of the centromeric probe was used with 20-40 ng of the Rb-1 probe. The slides were denatured in 70% formamide, 2×SSC at 70° C. for 2 min. Hybridization was done under a coverslip in a moist chamber at 37° C. for 2 days.

Staining. Briefly, the slides were washed three times in 50% formamide, 2×SSC for 10 min, twice in 2×SSC and once in 0.1×SSC at 45° C. Biotin-labeled specimens were stained with (5 ug/ml) FITC- or Texas Red-Avidin [Vector Laboratories, Inc., Burlingame, Calif. (USA)] in 4×SSC/1% BSA for 30 minutes at room temperature. Anti-avidin (Vector Laboratories, Inc. (5 ug/ml) incubation was done in PNM buffer for 20 min followed by a second layer of FITC/Texas Red-Avidin in PNM buffer. Digoxigenin-labeled probes were detected using an FITC-labeled sheep antibody against digoxigenin and a second layer of rabbit anti-sheep FITC antibody (Vector Laboratories, Inc., Burlingame, Calif.). Before each antibody or avidin treatment the slides were preblocked with either 1%BSA or PNM. Between the antibody incubations the slides were washed twice in 4×SSC and once in 4×SSC/0.1%=TRITON X-100 or three times in PN buffer. Nuclei were counter-stained with 0.2 ug/ml propidium iodide or 0.27 uM DAPI in an antifade solution.

Fluorescence microscopy and digital image analysis. A Nikon fluorescence microscope was used in most of the analyses. For interphase analysis at least 150 nuclei were scored from each sample. To map the metaphase hybridization signals accurately digital image analysis was used. The multicolor images were stored on computer magnetic disks at an approximate resolution of 19 pixels/em and analyzed using a specific software program based on the TCL-Image software package. The program defines the contour of the DAPI-stained chromosome and draws the longitudinal axis of the chromosome. The hybridization signals are then overlaid in pseudocolors on the chromosome image to calculate their relative position in terms of the distance from the p-telomere compared to the total chromosome length (=fractional length scale).

Results. Using fourteen lambda phage clones together (Rb-1 probe), a bright and specific hybridization signal on lymphocyte metaphase preparations in the mid-region of the q-area of chromosome 13 were obtained (FIGS. 13A and 13B). A more accurate localization of the Rb-1 gene was achieved by digital image analysis. The mean distance of the Rb-1 signal from the 13 pter (p-terminus) was determined to be compatible with the location of Rb-1 gene in the band q14.2 (FIG. 13C). Analysis of the Rb-1 hybridization from interphase nuclei was first attempted in normal lymphocytes and fibroblasts (FIG. 13D). Two hybridization signals representing the two gene alleles were detected in about 90% of the nuclei (FIG. 14A). The remaining 10% showed either one or three fluorescence signals. In interphase, the fluorescence signal was not always singular but could appear as 2-4 small adjacent individual spots probably because the Rb-1 probe consists of three separate contigs.

Two fibroblast cell lines from retinoblastoma patients with homozygous deletions affecting the Rb-1 region were used to test the sensitivity of FISH in detecting deletions using an absence of a hybridization signal as an indicator of deletions. In both cell lines one Rb-1 signal was detected in about 70-80% of the interphase nuclei (FIGS. 13G and 14B). Metaphase preparations from those cell lines hybridized simultaneously with the Rb-1 and the 13/21 centromeric probes showing that the normal chromosome 13 had both the centromeric and the Rb-1 signals, whereas the other slightly shortened chromosome 13 hybridized only with the centromeric probe (FIGS. 13E and 13F).

The Rb-1 probe was also used to study fine needle aspirations and touch preparations from different breast cancer patients. Although the breast cancer samples had more non-specific background fluorescence than cultured cells, it was still possible to evaluate Rb-1 gene copy numbers from individual tumor nuclei (FIG. 13H). As shown in Table 4 below, marked genetic heterogeneity both within and between breast tumors was found in the analysis of six cases. The modal Rb-1 gene copy number varied from 1-3 in the tumors. As compared to experiments with cell cultures, the clinical samples showed a higher percentage of cells without any Rb-1 signals. Table 4 shows the percentage of nuclei exhibiting a defined number of Rb-1 signals/nucleus in six clinical breast cancer specimens. The results represent the mean of 2-3 hybridization experiments. At least 150 cells were scored from each slide.

	TABLE 4
	No. signals/nuclei (%)
	Tumor	0	1	2	3	4	5	6	DI
	B156	22	23	45	5	5	0	0	1.50
	B245	28	36	31	3	2	0	0	1.35
	B249	2	10	23	36	25	3	0.5	1.82
	B252	21	14	43	14	3	3	1	1.87
	B259	36	49	12	2	1	0	0	1.64
	B263	16	3	11	33	20	10	6	2.25
	DI = DNA index

To detect subregions of the Rb-1 gene, single phage clones spanning only 8-20 kb of the 200 kb Rb-1 gene were used as hybridization probes. The hybridization signals from such probes could be seen both in metaphase chromosomes and interphase nuclei, but the hybridization efficiency was significantly less than with the pooled Rb-1 probe. In contrast, if 2-5 contiguous phage clones were pooled the hybridization was more efficient and more easily evaluated. This approach was used to visualize the 3′ and 5′ ends of the Rb-1 gene in interphase nuclei with differently labeled probes in a dual-color hybridization (FIG. 13I).

Thus, in conclusion, the Rb-1 gene was mapped to 13q14 by FISH and digital image analysis confirming the location of the gene to be in close proximity to the esterase D locus (Sparkes et al., 1980). Also shown in this section is that the methods of this invention can be used to detect deletions involving the Rb-1 locus. In order to verify the presence of a deletion from unbanded propidium iodide stained metaphase preparations, it was necessary to use a reference probe which in the representative example of this section was a 13/21 pericentromeric alpha satellite probe. Chromosome 13s with a deletion in the Rb-1 locus were thereby identified.

The representative examples of this section demonstrate that chromosome-specific staining can be used to detect Rb-1 gene deletions from interphase nuclei of cultured fibroblasts from retinoblastoma patients known to have a constitutive deletion in 13q. The usefulness of chromosome-specific staining is determined by the hybridization efficiency obtained, which in turn experientially has been found to be dependent on probe target size. Previous studies on interphase FISH have mainly been done using probes to pericentromeric repetitive sequences with a target size of a few megabases with hybridization efficiencies around 90-95% (Pinkel et al. 1986). In experiments using the 150 kb Rb-1 probe with cultured cells, the hybridization efficiency obtained in interphase was about 80-90%, whereas in clinical samples the efficiency was apparently less since a number of cells exhibited no hybridization signals. Poor hybridization efficiency might therefore lead to misinterpretation of a deletion. Further, in solid tumors having very complex karyotypic abnormalities, the distinction between numerical chromosome aberrations and structural abnormalities may be difficult to evaluate. Therefore, it is preferred in analyzing large numbers of solid tumors to co-hybridize with other reference probes for the same chromosome to control for the hybridization efficiency as well as for the presence of numerical chromosomal abnormalities. The centromeric 13/21 alpha satellite probe used successfully in metaphase preparations cannot be applied to interphase analysis because the signals for chromosome 13 and 21 cannot be distinguished. Therefore, for interphase analysis, it is preferred that a reference probe specific for chromosome 13 be used.

The studies on clinical breast cancer material described in this section demonstrate the genetic heterogeneity of breast cancer. The evaluation of this heterogeneity coupled with the possibility of studying gene copy numbers from morphologically defined individual tumor cells are major advantages of the chromosome-specific staining methods of this invention.

X. Detection of Chromosome 3 and 17 Aberrations Associated with Cancer

Probes. Two centromeric-specific alpha satellite probes are used in the representative examples of this section; one is specific to chromosome 17, and the other to chromosome 3. The centromeric-specific probes were prepared similarly as the 13/21 specific centromeric probes were, as indicated above in Section IX. Specifically, those probes were prepared using a polymerase chain reaction (PCR) process employing a thermostable enzyme [Saiki et al., Science, 239:487-491 (1988)] as follows.

Probe specific for alpha satellite centromeric repeats on human chromosome 17. Approximately 50 ng (nanograms) of DNA isolated from the Bluescribe plasmid library for chromosome 17 (pBS17) were used as the DNA template. Pinkel et al., PNAS USA, 85:9138-9142 (December 1988) describes the preparation of such Bluescribe libraries as subcloning an entire chromosome 17 library, which is publicly available as LL17NS01 or LA17NS03 [Van Dilla et al., Bio/Technology, 4:537-552 (June 1986)] into Bluescribe plasmids [Stratagene, La Jolla, Calif. (USA)].

The reaction buffer consisted of 5 units of Thermus aquaticus (Taq) DNA polymerase [Bethesda Research Laboratories, Gaithersburg, Md. (USA)]; mixed with 100 μl amplification/biotinylation buffer [10 mM Tris-HCl, pH 8.4 at 20° C.; 1.5 mM MgCl₂; 5 mM KCl; and 0.2 mM each of 2′-deoxyadenosine 5′-triphosphate (dATP), 2′-deoxyguanosine 5′-triphosphate (dGTP), and biuotin-II-dUTP [all the deoxynucleotide triphosphates were from Sigma, St. Louis, Mo. (USA)]; and 1.2 μM each of the two primers WA1 and WA2 [WA1 5′-GAAGCTTA(A/T(C/G)T(C/A)ACAGAGTT (G/T)AA-3′ and WA2 5′-GCTGCAGATC(A/C)C(A/C)AAG(A/T/C)AGTTTC-3′]. Mineral oil (100 μl) [Squibb, Princeton, N.J. (USA)] was layered on top of the reaction mixture to prevent evaporation during the PCR.

DNA amplification and simultaneous biotinylation was performed during 45 cycles using an automated thermal cycling system [Weier and Gray, DNA 7:441-447 (1988)]. Each cycle began with a thermal denaturation step of 90 seconds at 94° C. (120 seconds for the initial denaturation). Primer annealing during the second step of each cycle was performed at 53° C. for 90 seconds. The temperature was then increased slowly (7° C./minute) to 72° C. The cycle was completed by holding that temperature for 120 seconds for primer extension. Amplification of alpha satellite DNA was confirmed visually by electrophoresis of 5 μl aliquots of the PCR reaction mixture on 4% agarose gels (BRL) in 40 mM Tris-acetate, 1 mM EDTA, pH 8.0 containing 0.5 μ/ml ethidium bromide [Maniatis et al., (1986), supra]. The concentration of double stranded DNA in the reaction was determined to be 229 μg/ml by Hoeschst 33258 fluorescence using a TK 100 fluorometer [Hoefer Scientific, San Francisco, Calif. (USA)].

Probe specific for alpha satellite centromeric repeats on human chromosome 3. In vitro DNA amplification was performed using approximately 80 ng of CsCl gradient isolated DNA from the Bluescribe plasmid library for chromosome 3 (pBS3) (400 μg/l) as amplification template per 200 μl reaction mixture. The reaction buffer was the same as that used to prepare the chromosome 17 centromeric-specific probe above except that dTTP is used instead of biotin-11-dUTP.

PCR was performed for 30 cycles using an automated thermal cycler [Perkin-Elmer/Cetus, Norwalk, Conn. (USA)]. The DNA template was denatured at 94° C. for 1 minute (1 minute 30 seconds during the first cycle). Primer annealing and extension were performed at 53° C. and 72° C., respectively. Probe biotinylation and further amplification was accomplished in a second reaction by adding a 5 μl aliquot of the product to 200 μl reaction mix containing 0.25 mM Biotin-11-dUTP (Sigma St. Louis, Mo.) in the absence of dTTP and 10 units of Taq polymerase [Weier et al., J. Histochem. Cytochem, 38:421-426 (1990)]. The amplification/biotinylation reaction was performed during an additional 20 PCR cycles. Amplification of degenerate alpha satellite DNA was confirmed visually by gel electrophoresis of 10 μl aliquots of the PCR reaction in either 1.8% or 4% agarose (BRL) in 40 mM Tris-acetate, 1 mM EDTA buffer, pH 8.0 containing 0.5 μg/ml ethidium bromide. After completion of PCR, labeled probe and amplified DNA were stored at −18° C.

Chromosome 3 alpha satellite centromeric-specific repetitive probe. Another chromosome 3 centromeric-specific probe called palpha 3-5 was obtained from Huntington Willard, Ph.D. [Department of Genetics, Stanford University School of Medicine, Stanford, Calif. (USA)]. That probe was described at the Ninth International Workshop on Human Gene Mapping in Paris [Cytogenet. Cell Genet., 46 (14):424, 564 and 712 (1987), and 51 (14):111 (1989)]; and a similar probe is described in Waye and Willard, Chromosoma (Berl), 97:475-480 (1989). The palpha 3-5 probe was labeled with AAF according to conventional methodology for use in the experiments described below.

3p cosmid probe. A 3p cosmid probe called cC13-787 was obtained from Yusuke Nakamura, MD, Ph.D [Division of Biochemistry, Cancer Institute, Toshima, Tokyo, 170, Japan]. Its isolation and mapping to 3p21.2-p21.1 is described in Yamakawa et al., Genomics, 9(3):536-543 (1991). That probe was amplified and labeled with biotin according to a PCR linker/adapter method described in Johnson, Genomics, 6:243-251 (1990) and Saunders et al., Nucl. Acids. Res., 17 (22):9027-9037 (1989).

3q cosmid probe. A 3q cosmid probe named J14R1A12 was developed and provided by Wen-Lin Kuo [Biomedical Department, P.O. Box 5507 (L-452), Lawrence Livermore National laboratory, Livermore, Calif. 94550 (USA)]. It was obtained from a chromosome 21 flow sorted library prepared by conventional means. It was mapped to the location 3q26 using an extrapolation of the fractional length for the probe to a chromosome 3 ideogram. It was labeled with biotin-dUTP using the Bio Nick™ Labeling System [BRL Life Technologies, Inc. Gaithersburg, Md. (USA)].

Composite Whole Chromosome 3-Specific Probe. A probe specific for the whole chromosome 3 was a Bluescribe plasmid library for chromosome 3 prepared according to Pinkel et al., PNAS USA, 85:9138-9142 (December 1988) and named pBS3.

Cell Samples. Used in the following examples were PHA-stimulated normal peripheral blood lymphocytes; two ovarian cancer cell lines, designated as RMUG-S and RMUG-L, provided by Shiro Nozawa, Md., Ph.D. (Department of Obstetrics/Gynecology, School of Medicine, Keio University, 35 Shinano-machi, Shinjuku-ku, Tokyo, 160, Japan) and described in Sakayori et al., Human Cell, 3 (1):52-56 (1990); and a uterine cervical adenocarcinoma cell line, named TMCC-1, described in Sakamoto, J. Tokyo Med. College, 46 (5):925-936 (1988). RMUG-S is a hypodiploid cancer cell line, whereas RMUG-L is a hypertriploid cancer cell line. Both lines were cloned from the same clinical specimen.

In situ hybridization and staining. The protocols resulting in the hybridizations shown in FIGS. 15, 16 and 17 were the same as those used in Section IX except that for FIG. 15, the hybridization mixture contained 5 μg of herring sperm DNA instead of 5 μg of unlabeled placental DNA, and reannealing of the denatured probe just before in situ hybridization was not performed: and for FIG. 16, the hybridization mixture contained a reduced amount of 0,5 μg of unlabeled human placental DNA explicitly 0.5 μg.

For the results shown in FIGS. 18 and 19, the hybridization protocols differed from that detailed in Section IX in that before hybridization, the slides were pretreated with 100 μg/ml of RNase for 30 minutes at 37° C. and then treated with 1μg/ml of Proteinaise K for 7.5 minutes at 37° C. The hybridization mixture comprised 1 μl biotinylated 3p cosmid (wherein the PCR product was diluted 1:10 with double distilled, deionized water) or 30 ng-40 ng 3q cosmid; 2 ng chromosome 3 alpha satellite centromeric-specific probe (palpha 3-5) labeled with AAF; 5 μg-10 μg unlabeled human placental DNA; and 7.0 μl of master mix [which consists of 5 ml formamide to which 1 g dextran sulphate and 1 ml 20×SSC (prepared using deionized, double distilled water) was added, the pH of which was adjusted to 7.0 with 1N HCl and the final volume to 7 ml was completed with deionized, double distilled water]. The master mixture is stored at −20° C. indefinitely.

For the results shown in FIGS. 18 and 19, a dual color staining protocol was performed essentially according to Trask and Pinkel, Methods in Cell Biology, 33:383400 (1991). Briefly, the slides were washed three times in the washing solutions for 10 minutes each at 45° C., 2×SSC for 10 minutes at 45° C., 0.1×SSC for 10 minutes, and PN buffer (wherein the percentage of NP-40 is 0.05% rather than 0.1%) for 5 minutes at room temperature. The washing solutions comprise 50% formamide:2×SSC (75 ml formamide; 15 ml 20×SSC; and 60 ml deionized, double distilled water wherein the pH is adjusted to 7.0 with 1 N HCl).

The slides are preblocked with 20 μl PNM buffer under a 22×22 mm coverslip at room temperature for 5 minutes inside a dark moist chamber. The coverslip is then taken off, and the PNM buffer is drained from the slide.

Twenty μl of anti-AAF and Avidin-Texas Red solution are added to the cells+area per slide and then covered with a 22×22 mm coverslip. The slides are incubated within a dark moist chamber for 1 hour at room temperature. The anti-AAF and Avidin-Texas Red solution is prepared by adding 8 μl of 0.25 μg/μl Avidin-Texas Red (Vector) to 1 ml of the supernatant of anti-AAF producing mouse cells.

The coverslips are then removed, and the slides are washed with intermittent shaking in the PN buffer thrice for 10 minutes each in a dark place.

The cells are then preblocked as described above. Twenty μl of goat-anti-mouse-FITC antibody and biotinylated anti-Avidin antibody solution is added to the cells' area on each slide. The cells are covered with a coverslip and incubated inside a dark moist chamber for one hour at room temperature. The antibody solution per ml comprises 20 μl of goat-anti-mouse-FITC antibody [from Cal Taq (Burlingame, Calif.) i.e., the final concentration is 20 μg/ml] and 10 μl biotinylated anti-avidin antibody solution at the concentration of 0.5 mg/ml (from Vector Laboratories, i.e., the final concentration is 5 μg/ml) to 970 μl antibody dilution buffer [IX Dulbecco's PBS (Ca, Mg free), 0.05% TWEEN 20, 2% normal goat serum].

The coverslips were removed from the slides, and the slides are then washed with intermittent shaking in the PN buffer thrice for 10 minutes each in a dark place. The cells are then preblocked as indicated above.

Twenty μl of the Avidin-Texas Red solution were added to the cells' area per slide, and then a coverslip was applied. The slides are then incubated inside a dark moist chamber for one hour at room temperature. The Avidin-Texas Red Solution comprises 8 μl of a 0.25 μg/μl Avidin-Texas Red (Vector) to 1 ml of the antibody dilution buffer.

The coverslips are then removed from the slides, and the slides are washed in the PN buffer thrice for 10 minutes each in a dark place. About 8 μl of 0.8 μm DAPI (counterstain) in an anti-fade solution is prepared according to J. Immuno. Methods, 43:349 (1981) [100 mg p-phenylene-diamine dihydrochloride (Sigma P1519) in 10 ml Dulbecco's PBS; pH adjusted to 8 with 0.5 M carbonate-bicarbonate buffer; 90 ml glycerol added; filtered through 0.22 μm; stored at −20° C.] was added to the slides. The slides mounted with the anti-fade solution can be stored in a dark chamber at 4° C.

Results. FIG. 15A shows the hybridization of the chromosome 17 centromere-specific alpha satellite probe to normal lymphocytes wherein in metaphase chromosomes, two bright signals are seen, and in interphase nuclei, two bright, tight hybridization domains are visible. FIG. 15B shows the hybridization of that probe to the human ovarian mucinous cysto-adenocarcinoma (RMUG-L), wherein in both metaphase and interphase, four signals are visible.

These examples are representative of the use of chromosome-specific repeat probes for the detection of numerical chromosome aberrations on chromosome 17 which are used as a component of the high complexity staining probes of this invention.

FIGS. 16A and B show hybridization of the whole chromosome composite probe for chromosome 3 (pBS3) (A) to normal lymphocytes and (B) to the ovarian cancer cell line (RMUG-L). Two normal chromosome 3s are seen in FIG. 16A, whereas four chromosome 3s are seen in the ovarian cancer cell line (FIG. 16B), of which two are apparently shorter than the intact chromosome 3s, a pattern which is congruent with a 3p deletion in the karyotype.

Chromosome-specific recombinant lambda libraries have been constructed for all the human chromosomes by the National Laboratory Gene Library Project [Van Dilla et al. (1986), supra]. Subsequently, those libraries were subcloned into Bluescribe plasmid vectors (Stratagene), and whole chromosome composite probes were generated from the DNA extracted from those plasmids [Fuscoe et al., Genomics, 5:100-109 (1989); Collins et al., Genomics 11(4):997-1006(1991)]. Staining with such whole chromosome composite probes can be used to detect not only large deletions but also subtle translocations and to identify the origin of marker chromosomes.

FIGS. 17A and B (as well as FIGS. 9, 10 and 13, among others) provide examples of the use of locus-specific probes to count the copy number of specific genes in tumor cells and to detect changes in patterns of hybridization domains. FIGS. 17A and B provide representative examples of the use of locus-specific probes to detect translocations. As indicated above, such examples are the first step in locating exact information on genetic rearrangements within a locus. The 3q cosmid probe employed in these studies is just one of many potential probes from chromosome 3 that can be used [Yamakawa et al., Genomics, 9(3):536-543 (1991)]. Preferably, in metaphase spreads probes with different labels according to their order in a normal chromosome 3 may be used to detect any structural chromosomal aberrations differing from the standard. Further, in either metaphase spreads or interphase nuclei, probes with different labels according to their location in normal chromosomes may be used to detect structural chromosomal aberrations, for example, commonly deleted lesions found in cancer cells.

FIGS. 17A and B show the hybridization of a chromosome 3 centromeric-specific alpha satellite repeat probe (the one generated by the PCR process with the primers WA1 and WA2) and a 3q cosmid probe (J14R1A12 mapped to 3q 26) to, respectively (A) normal lymphocytes and (B) a uterine cervical adenocarcinoma cell line (TMCC-1). As indicated in the description of the figures above, two pairs of normal chromosome 3s are illustrated in (A) whereas a a pair of cosmid signals specific to chromosome region 3q was found to be translocated to another chromosome.

FIGS. 18A and 18B show dual color hybridization to normal lymphocytes, metaphase spread and interphase nucleus, respectively. FIGS. 19A and 19B show comparable hybridizations to an ovarian cancer cell line. A chromosome 3 centromeric-specific repetitive probe (AAF labeled palpha 3-5 from Huntington Willard) and 3p region-specific (3p21.2 p21.1) cosmid probe (cC13-787) (that is biotinylated and amplified and labeled by linker adapter PCR) were employed in such hybridizations. Two hundred interphase nuclei were scored for each experiment.

In FIG. 18A, the image of chromosome 3 from normal lymphocytes was digitized by the digital fluorescent microscope and shows one chromosome 3 centromeric-specific green signal and one pair of chromosome 3p region-specific red signal for each chromatid were visible.

In FIG. 18B, the picture of an interphase nucleus from normal lymphocytes, taken with a conventional fluorescent microscope, shows two greenish hybridization domains for the centromeric specific probes and two reddish domains for the 3p probe. It was commonly observed that a pair of cosmid signals on both chromatids of one chromosome 3 fuses into a single spot in interphase nuclei.

FIG. 19A shows a partial metaphase spread, one chromosome 3 shows a normal hybridization pattern whereas the other shows a 3p deletion. FIG. 19B shows in interphase nuclei, four large greenish domains for the centromeric probe and two small reddish hybridization domains for the 3p probe, indicating aneuploidy of chromosome 3, wherein two out of four of the chromosome 3s have a 3p deletion.

Eighty-six percent of the interphase nuclei of the normal lymphocytes showed normal pattern of two green centromeric signals and two red signals of the 3p cosmid probe. However 98% (94%) of interphase nuclei of the RMUG-S (RMUG-L) cells showed a lesser number of red signals (3p cosmid) than green signals (centromere) suggesting a chromosome 3p deletion in those cell lines. Among those nuclei, 53% (52%) of interphase nuclei of the RMUG-S (RMUG-L) cells showed two domines of the 3p cosmid signal and 4 domains of the chromosome 3 centromeric-specific signals.

FIGS. 20 and 21 show the results of simultaneous hybridizations of an AAF-labeled chromosome 3 centromere-specific probe (from H. Willard) and a biotinylated chromosome 3q cosmid probe (J14R1A12) wherein in FIG. 20 the target is a metaphase spread and interphase nucleus of normal lymphocytes and wherein in FIG. 21 the target is an interphase nucleus from the ovarian cancer cell line (RMUG-S). A pattern for a normal chromosomal complement is shown in FIG. 20 as two chromosome 3 centromere-specific green signals and two pairs of chromosome 3q cosmid red signals per cell. An abnormal pattern is shown in FIG. 21 as four chromosome 3 centromere-specific green signals and four chromosome 3q cosmid red signals, indicating that the cell contains four long arms of chromosome 3. The results shown in FIGS. 20 and 21 support the feasibility of detecting 3p deletions in interphase nuclei of tumor cells if combined with the findings of domain number for 3p cosmid signals. Such a methodology can be applied to detect 3p deletions in clinical tumor specimens for basic research on tumorigenesis and progression, and adjunctive diagnosis of cancers associated with 3p deletions, such as, small cell lung cancer, renal cell cancer and ovarian cancer.

The descriptions of the foregoing embodiments of the invention have been presented for purpose of illustration and description. They are not intended to be exhaustive or to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teachings. The embodiments were chosen and described in order to best explain the principles of the invention and its practical application to thereby enable others skilled in the art to best utilize the invention in various embodiments and with various modifications as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims appended hereto. All references cited herein are hereby incorporated by reference.

Claims

1. A method of staining targeted interphase chromosomal material based upon a nucleic acid segment employing a unique sequence high complexity nucleic acid probe of greater than about 50,000 bases, wherein said targeted chromosomal material is a genetic rearrangement associated with at least one chromosome in humans, said method comprising employing said chromosomal material and a unique sequence high complexity nucleic acid probe of greater than about 50,000 bases in in situ hybridization, wherein the chromosomal material is present in a morphologically identifiable cell nucleus; allowing said probe to bind to said targeted chromosomal material; and detecting said bound probe, wherein bound probe is indicative of the presence of target chromosomal material.

2. A method of staining targeted interphase chromosomal material based upon a nucleic acid segment employing a unique sequence high complexity nucleic acid probe of greater than about 40 kb, wherein said targeted chromosomal material is a genetic rearrangement associated with at least one chromosome in humans, said method comprising contacting said chromosomal material with a unique sequence high complexity nucleic acid probe of greater than about 40 kb, wherein the chromosomal material is present in a morphologically identifiable cell nucleus; allowing said probe to bind to said targeted chromosomal material; and detecting said bound probe, wherein bound probe is indicative of the presence of target chromosomal material.

3. A method of staining targeted interphase chromosomal material based upon a nucleic acid segment employing a unique sequence high complexity nucleic acid probe of greater than about 50,000 bases, wherein said targeted interphase chromosomal material is a genetic rearrangement associated with at least one chromosome in humans, said method comprising contacting said interphase chromosomal material with a unique sequence high complexity nucleic acid probe of greater than about 50,000 bases, wherein the chromosomal material is present in a morphologically identifiable cell nucleus; allowing said probe to bind to said targeted interphase chromosomal material; and detecting said bound probe, wherein bound probe is indicative of the presence of target interphase chromosomal material.

4. The method of claim 2, wherein the genetic rearrangement is a translocation or an inversion.

5. The method of claim 2, wherein the unique sequence high complexity nucleic acid probe is labeled.

6. The method of claim 5, wherein said labeled unique sequence high complexity nucleic acid probe comprises fragments complementary to a single chromosome, fragments complementary to a subregion of a single chromosome, fragments complementary to a genome or fragments complementary to a subregion of a genome.

7. The method of claim 2, wherein the interphase chromosomal material is interphase chromosomal DNA.

8. The method of claim 3, wherein the genetic rearrangement is a translocation or an inversion.

9. The method of claim 3, wherein the unique sequence high complexity nucleic acid probe is labeled.

10. The method of claim 9, wherein said labeled unique sequence high complexity nucleic acid probe comprises fragments complementary to a single chromosome, fragments complementary to a subregion of a single chromosome, fragments complementary to a genome or fragments complementary to a subregion of a genome.

11. The method of claim 3, wherein the interphase chromosomal material is interphase chromosomal DNA.

12. The method of claim 2, wherein complexity of the unique sequence high complexity nucleic acid probe is greater than about 100,000 bases.

13. The method of claim 3, wherein complexity of the unique sequence high complexity nucleic acid probe is greater than about 100,000 bases.