Methods and compositions for chromosome-specific staining

CHROMOSOME-SPECIFIC

TO DETECT GENETIC REARRANGEMENTS

ABSTRACT OF THE DISCLOSURE

Methods and compositions for staining based upon nucleic acid sequence that employ nucleic add probes are provided. Said methods produce staining patterns that can be tailored for spedfic cytogenetic analyses. Said probes are appropriate for in situ hybridization and stain both interphase and metaphase chromosomal material with reliable signals. The nucleic add 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) 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. Further provided is a method of preparing and applying chromosome-specific single-stranded nucleic acid probes to double stranded target DNA.

METHODS AND COMPOSITIONS FOR CHROMOSOME-SPECIFIC STAINING

The United States Government has rights in this invention pursuant to Contract No. W-7405-ENG-48 between the U.S. Department of Energy and the University of California, for the operation of Lawrence Livermore National Laboratory.

HELD OF THE INVENTION

The invention relates generally to the field of cytogenetics, and more particularly, to the field of molecular cytogenetics. The invention concerns methods for identifying and classifying chromosomes. Still more particularly, this invention concerns nucleic acid probes which can be designed by the processes described herein to produce staining distributions that can extend along one or more whole chromosomes, and/or along a region or regions on one or more chromosomes, including staining patterns that extend over the whole genome. Staining patterns can be tailored for any desired cytogenetic application, including prenatal, tumor and disease related cytogenetic applications, among others. The invention provides for compositions of nucleic acid probes and for methods of staining chromosomes therewith to identify normal chromosomes and chromosomal abnormalities in metaphase spreads and in interphase nuclei. The probeproduced staining patterns of this invention facilitate the microscopic and/or flow cytometric identification of normal and abnormal chromosomes and the characterization of the genetic nature of particular abnormalities.

Although most of the examples herein concern human chromosomes and much of the language herein is directed to human concerns, the concept of using nucleic acid probes for staining or painting chromosomes is applicable to chromosomes from any source including both plants and animals.

BACKGROUND OF THE INVENTION

Chromosome abnormalities are associated with genetic disorders, degenerative diseases, and exposure to agents known to cause degenerative diseases, particularly cancer, German, Studying Human Chromosomes Today, American Scientist, Vol. 58, pp. 182-201 (1970); Yunis, The Chromosomal Basis of Human Neoplasia, Science, Vol. 221, pp. 227-236 (1983); and German, Clinical Implication of Chromosome Breakage, in Genetic Damage in Man Caused by Environmental Agents, Berg, Ed., pp. 65-86 (Academic Press, New York, 1979). Chromosomal abnormalities can be of several types, including: extra or missing individual chromosomes, extra or missing portions of a chromosome (segmental duplications or deletions), breaks, rings and chromosomal rearrangements, among others. Chromosomal or genetic rearrangements include translocations (transfer of a piece from one chromosome onto another chromosome), dicentrics

-ב(chromosomes with two centromeres), inversions (reversal in polarity of ³ chromosomal segment), insertions, amplifications, and deletions.

Detectable chromosomal abnormalities occur with a frequency of one in every 250 human births. Abnormalities that involve deletions or additions of chromosomal material alter the gene balance of an organism and generally lead to fetal death or to serious mental and physical defects. Down syndrome can be caused by having three copies of chromosome 21 instead of the normal 2. This syndrome is an example of a condition caused by abnormal chromosome number, or aneuploidy. Down syndrome can also be caused by a segmental duplication of a subregion on chromosome 21 (such as, 71 q77)_z which can be present on chromosome 21 or on another chromosome. Edward syndrome (18+), Patau syndrome (13+), Turner syndrome (XO) and Kleinfelter syndrome (XXY) are among the most common numerical aberrations. [Epstein, The Consequences of Chromosome Imbalance: Principles, Mechanisms and Models (Cambridge Univ. Press 1986); Jacobs, Am, J, Epidemiol, 105:180 (1977); and Lubs et al., Science, 169:495 (1970).]

Retinoblastoma (del 13ql4), Prader-Willi syndrome (del 15qll- ql3), Wilm's tumor (del llp!3) and Cri-du-chat syndrome (del 5p) are examples of important disease linked structural aberrations. [Nora and Fraser, Medical Genetics: Principles and Practice, (Lea and Febiger 1989).]

Measures of the frequency of structurally aberrant chromosomes, for example, dicentric chromosomes, caused by clastogenic agents, such as, ionizing radiation or chemical mutagens, are widely used as quantitative indicators of genetic damage caused by such agents, Biochemical Indicators of Radiation Injury in Man (International Atomic Energy Agency, Vienna, 1971); and Berg, Ed. Genetic Damage in Man Caused by Environmental ־4Agents (Academic Press, New York, 1979). A host of potentially carcinogenic and teratogenic chemicals are widely distributed in the environment because of industrial and agricultural activity. These chemicals include pesticides, and a range of industrial wastes and by-products, such as halogenated hydrocarbons, vinyl chloride, benzene, arsenic, and the like, Kraybill et al., Eds., Environmental Cancer (Hemisphere Publishing Corporation, New York, 1977). Sensitive measures of chromosomal breaks and other abnormalities could form the basis of improved dosimetric and risk assessment methodologies for evaluating the consequences of exposure to such occupational and environmental agents.

Current procedures for genetic screening and biological dosimetry involve the analysis of karyotypes. A karyotype is the particular chromosome complement of an individual or of a related group of individuals, as defined both by the number and morphology of the chromosomes usually in mitotic metaphase. It includes such things as total chromosome number, copy number of individual chromosome types (e.g., the number of copies of chromosome X), and chromosomal morphology, e.g., as measured by length, centromeric index, connectedness, or the like. Chromosomal abnormalities can be detected by examination of karyotypes. Karyotypes are conventionally determined by staining an organism's metaphase, or otherwise condensed (for example, by premature chromosome condensation) chromosomes. Condensed chromosomes are used because, until recently, it has not been possible to visualize interphase chromosomes due to their dispersed condition and the lack of visible boundaries between them in the cell nucleus.

A number of cytological techniques based upon chemical stains have been developed which produce longitudinal patterns on condensed chromosomes, generally referred to as bands. The banding pattern of each chromosome within an organism usually permits unambiguous identification of each chromosome type, Latt, Optical Studies of Metaphase Chromosome Organization, Annual Review of Biophysics and Bioengineering, Vol. 5, pp. 11976) 37־). Accurate detection of some important chromosomal abnormalities, such as translocations and inversions, has required such banding analysis.

Unfortunately, such conventional banding analysis requires cell culturing and preparation of high quality metaphase spreads, which is time consuming and labor intensive, and frequently difficult or impossible. For example, cells from many tumor types are difficult to culture, and it is not clear that the cultured cells are representative of the original tumor cell population. Fetal cells capable of being cultured need to be obtained by invasive means and need to be cultured for several weeks to obtain enough metaphase cells for analysis. In many cases, the banding patterns on the abnormal chromosomes do not permit unambiguous identification of the portions of the normal chromosomes that make them up. Such identification may be important to indicate the location of important genes involved in the abnormality. Further, the sensitivity and resolving power of current methods of karyotyping are limited by the fact that multiple chromosomes or chromosomal regions have highly similar staining characteristics, and that abnormalities (such as deletions) which involve only a fraction of a band are not detectable. Therefore, such methods are substantially limited for the diagnosis and detailed analysis of contiguous gene syndromes, such as partial trisomy, Prader-Willi syndrome [Emanuel, Am. T. Hum. Genet., 43:575 (1988); Schmickel. T. Pediatr., 109:231 (1986)] and retinoblastoma [Sparkes, Biochem. Biophys. Acta., 780:95 (1985)].

Thus, conventional banding analysis has several important limitations, which include the following. 1) It is labor intensive, time consuming, and requires a highly trained analyst. 2) It can be applied only to condensed chromosomes. 3) It does not allow for the detection of structural aberrations involving less than 3-15 megabases (Mb), depending upon the nature of the aberration and the resolution of the banding technique [Landegren et al. Science, 242:229 (1988)]. This invention provides for probe compositions and methods to overcome such limitations of conventional banding analysis.

The chemical staining procedures of the prior art provide patterns over a genome for reasons not well understood and which cannot be modified as required for use in different applications. Such chemical staining patterns were used to map the binding site of probes. However, only occasionally, and with great effort, was in situ hybridization used to obtain some information about the position of a lesion, for example, a breakpoint relative to a particular DNA sequence. The present invention overcomes the inflexibility of chemical staining in that it stains a genome in a pattern based upon nucleic acid sequence; therefore the pattern can be altered as required by changing the nucleic acid sequence of the probe. The probe-produced staining patterns of this invention provide reliable fundamental landmarks which are useful in cytogenetic analysis.

Automated detection of structural abnormalities of chromosomes with image analysis of chemically stained bands would require the

-Ίdevelopment of a system that can detect and interpret the banding patterns produced on metaphase chromosomes by conventional techniques. It has proven to be very difficult to identify reliably by automated means normal chromosomes that have been chemically stained; it is much more difficult to differentiate abnormal chromosomes having structural abnormalities, such as, translocations. Effective automated detection of translocations in conventionally banded chromosomes has not been accomplished after over a decade of intensive work. The probe-produced banding patterns of this invention are suitable for such automated detection and analysis.

In recent years rapid advances have taken place in the study of chromosome structure and its relation to genetic content and DNA composition. In part, the progress has come in the form of improved methods of gene mapping based on the availability of large quantities of pure DNA and RNA fragments for probes produced by genetic engineering techniques, e.g., Kao, Somatic Cell Genetics and Gene Mapping, International Review of Cytology, Vol. 85, pp. 109-146 (1983), and D'Eustachio et al., Somatic Cell Genetics in Gene Families, Science, Vol. 220, pp. 9,191983) 924־). The probes for gene mapping comprise labeled fragments of single-stranded or double-stranded DNA or RNA which are hybridized to complementary sites on chromosomal DNA. With such probes it has been crucially important to produce pure, or homogeneous, probes to minimize hybridizations at locations other than at the site of interest, Henderson, Cytological Hybridization to Mammalian Chromosomes, International Review of Cytology, Vol. 76, pp. 1-46 (1982).

The hybridization process involves unravelling, or melting, the double-stranded nucleic adds of the probe and target by heating, or other

־8־ means (unless the probe and target are single-stranded nucleic acids). This step is sometimes referred to as denaturing the nucleic add. When the mixture of probe and target nucleic adds cool, strands having complementary bases recombine, or anneal. When a probe anneals with a target nudeic add, the probe's location on the target can be detected by a label carried by the probe or by some intrinsic characteristics of the probe or probe-target duplex. When the target nucleic acid remains in its natural biological setting, e.g., DNA in chromosomes, mRNA in cytoplasm, portions of chromosomes or cell nuclei (albeit fixed or altered by preparative techniques), the hybridization process is referred to as in situ hybridization.

In situ hybridization probes were initially limited to identifying the location of genes or other well defined nucleic add sequences on chromosomes or in cells. Comparisons of the mapping of single-copy probes to normal and abnormal chromosomes were used to examine chromosomal abnormalities. Cannizzaro et al., Cytogenetics and Cell Genetics, 39:173178־ (1985). Distribution of the multiple binding sites of repetitive probes could also be determined.

Hybridization with probes which have one target site in a haploid genome, single-copy or unique sequence probes, has been used to map the locations of particular genes in the genome [Harper and Saunders, Localization of the Human Insulin Gene to the Distal End of the Short Arm of Chromosome 11, Proc. Natl. Acad. Sci., Vol. 78, pp. 44581981) 4460־); Kao et al., Assignment of the Structural Gene Coding for Albumin to Chromosome 4, Human Genetics, Vol. 62, pp. 3371982) 341־)]; but such hybridizations are not reliable when the size of the target site is small. As the amount of target sequence for low complexity single-copy probes is small, only a portion of the potential target sites in a population of cells form hybrids with the probe. Therefore, mapping the location of the specific binding site of the probe has been complicated by background signals produced by non-specific binding of the probe and also by noise in the detection system (for example, autoradiography or immunochemistry). The unreliability of signals for such prior art single-copy probes has required statistical analysis of the positions of apparent hybridization signals in multiple cells to map the specific binding site of the probe.

Different repetitive sequences may have different distributions on chromosomes. They may be spread over all chromosomes as in the just cited reference, or they may be concentrated in compact regions of the genome, such as, on the centromeres of the chromosomes, or they may have other distributions. In some cases, such a repetitive sequence is predominantly located on a single chromosome, and therefore is a chromosome-specific repetitive sequence. [Willard et al., Isolation and Characterization of a Major Tandem Repeat Family from the Human X Chromosome, Nucleic Acids Research, Vol. 11, pp. 2017-2033 (1983).]

A probe for repetitive sequences shared by all chromosomes can be used to discriminate between chromosomes of different species if the sequence is specific to one of the species. Total genomic DNA from one species which is rich in such repetitive sequences can be used in this manner. [Pinkel et al. (ΙΠ), PNAS USA, 83:2934 (1986); Manuelidis, Hum, Genet., Zl:288 (1985) and Dumam et al., Somatic Cell Molec. Genet., 11:571 (1985.]

Recently, there has been an increased availability of probes for repeated sequences (repetitive probes) that hybridize intensely and specifically to selected chromosomes. [Trask et al., Hum, Genet., 78:251 (1988) and references cited therein.] Such probes are now available for over half of the human chromosomes. In general, they bind to repeated sequences on compact regions of the target chromosome near the centromere. However, one probe has been reported that hybridizes to human chromosome lp36, and there are several probes that hybridize to human chromosome Yq. Hybridization with such probes permits rapid identification of chromosomes in metaphase spreads, determination of the number of copies of selected chromosomes in interphase nuclei [Pinkel et al. (I), PNAS USA, 83:2934 (1986); Pinkel et al. (Π), Cold Spring Harbor Symp. Quant. Biol., 51:151 (1986) and Cremer et al., Hum. Genet, 74:346 (1986)] and determination of the relative positions of chromosomes in interphase nuclei [Trask et al., supra; Pinkel et al. (I), supra; Pinkel et al. (Π), supra; Manuelidis, PNAS USA, 81:3123 (1984); Rappold et al., Hum. Genet., 67:317 (1984); Schardin et al., Hum. Genet., 71:282 (1985); and Manuelidis, Hum. Genet., 71:288 (1985)].

However, many applications are still limited by the lack of appropriate probes. For example, until the methods described herein were invented, probes with sufficient specificity for prenatal diagnosis were not available for chromosome 13 or 21. Further, repetitive probes are not very useful for detection of structural aberrations since the probability is low that the aberrations will involve the region to which the probe hybridizes. This invention overcomes the prior art limitations on the use of probes and dramatically enhances the application of in situ hybridization for cytogenetic analysis. As indicated above, prior art probes have not been useful for in-depth cytogenetic analysis. Low complexity single-copy probes do not at this stage of hybridization technology generate reliable signals. Although repetitive probes do provide reliable signals, such signals cannot be tailored for different applications because of the fixed distribution of repetitive sequences in a genome. The probes of this invention combine the hybridization reliability of repetitive probes with the flexibility of being able to tailor the binding pattern of the probe to any desired application.

The enhanced capabilities of the probes of this invention come from their increased complexity. Inaeasing the complexity of a probe increases the probability, and therefore the intensity, of hybridization to the target region, but also increases the probability of non-specific hybridizations resulting in background signals. However, within the concept of this invention, it was considered that such background signals would be distributed approximately randomly over the genome. Therefore, the net result is that the target region could be visualized with increased contrast against such background signals.

Exemplified herein are probes in an approximate complexity range of from about 50,000 bases (50 kb) to hundreds of millions of bases. Such representative probes are for compact lod and whole human chromosomes. Prior to this invention, probes employed for in situ hybridization techniques had complexities below 40 kb, and more typically on the order of a few kb.

Staining chromosomal material with the probes of this invention is significantly different from the chemical staining of the prior art. The specificity of the probe produced staining of this invention arises from an entirely new source - the nucleic add sequences in a genome. Thus, staining patterns of this invention can be designed to highlight fundamental genetic information important to particular applications.

The procedures of this invention to construct probes of any desired specificity provide significant advances in a broad spectrum of cytogenetic studies. The analysis can be carried out on metaphase chromosomes and interphase nuclei. The techniques of this invention can be especially advantageous for applications where high-quality banding by conventional methods is difficult or suspected of yielding biased information, e.g., in tumor cytogenetics. Reagents targeted to sites of lesions known to be diagnostically or prognostically important, such as tumor type-specific translocations and deletions, among other tumor specific genetic arrangements, permit rapid recognition of such abnormalities. Where speed of analysis is the predominant concern, e.g., detection of low-frequency chromosomal aberrations induced by toxic environmental agents, the compositions of this invention permit a dramatic increase in detection efficiency in comparison to previous techniques based on conventional chromosome banding.

Further, prenatal screening for disease-linked chromosome aberrations (e.g., trisomy 21) is enhanced by the rapid detection of such aberrations by the methods and compositions of this invention. Interphase aneuploidy analysis according to this invention is particularly significant for prenatal diagnosis in that it yields more rapid results than are available by cell culture methods. Further, fetal cells separated from maternal blood, which cannot be cultured by routine procedures and therefore cannot be analysed by conventional karyotyping techniques, can be examined by the methods and compositions of this invention. In addition, the intensity, contrast and color combinations of the staining patterns, coupled with the ability to tailor the patterns for particular applications, enhance the opportunities for automated cytogenetic analysis, for example, by flow cytometry or computerized microscopy and image analysis.

This application specifically claims chromosome specific reagents for the detection of genetic rearrangements and methods of using such reagents to detect such rearrangements. Representative genetic rearrangements so detected are those that produce a fusion gene - BCR-ABL - that is diagnostic for chronic myelogenous leukemia (CML).

Chronic myelogenous leukemia (CML) is a neoplastic proliferation of bone marrow cells genetically characterized by the fusion of the BCR and ABL genes on chromosomes 9 and 22. That fusion usually involves a reciprocal translocation t(9;22)(q34;qll), which produces the cytogenetically distinctive Philadelphia chromosome (Ph¹). However, more complex rearrangements may cause BCR-ABL fusion. At the molecular level, fusion can be detected by Southern analysis or by in vitro amplification of the mRNA from the fusion gene using the polymerase chain reaction (PCR). Those techniques are sensitive but cannot be applied to single cells.

Clearly, a sensitive method for detecting chromosomal abnormalities and, more specifically, genetic rearrangements, such as, for example, the tumor specific arrangements associated with CML, would be a highly useful tool for genetic screening. This invention provides such tools.

The following references are indicated in the ensuing text by numbers as indicated:

1. A. de Klein et al., Nature 300,765 (1982).

2. J. Groffen et al., Cell 36,93 (1984).

3. N. Heisterkamp et al., Nature 306,239 (1983).

4. E. Shtivelman et al., Blood 69,971 (1987).

5. J. B. Konopka, S. M. Watanabe, O.N. Witte, Cell 37,1035 (1984).

6. Y. Ben-Neriah et al., Science 233,212 (1986).

7. P. C. Nowell and D. A. Hungerford, Science 132,1497 (1960).

8. J. D. Rowley, Nature 243,290 (June 1973).

9. G.Grosveld et al., Mol Cell Biol 6,607 (1986).

10. E. Canaani et al., Lancet 1,593 (1984).

11. R. P. Gale and E. Canaani, Proc Natl Acad Sci USA 81,5648 (1984).

12. Konopka J. B. et al., Proc Natl Acad Sci USA 82:1810 (1985).

13. P. Benn et al., Cancer Genet Cytogenet 29,1 (1987).

14. S. Abe et al., Cancer Genet Cytogenet 38,61 (1989)

15. M. Shtalrid et al., Blood 72,485 (1988).

16. I. Dube et al., Genes Chromosomes and Cancer 1,106 (1989).

17. A. J. Fishleder, B. Shadrach and C. Tuttle, Leukemia 3:10,746 (1989)

18. C. R. Bartram et al., J Exp Med 164 (5):1389 (1986).

19. S. Hiroswa et al., Am L Hematol 28,133 (1988).

20. M. S. Lee et al., Blood 73 (8):2165 (1989).

21. E.S. Kawasaki et al., Proc Natl Acad Sci USA 85,5698 (1988).

22. M. S. Roth et al., Blood 74,882 (1989).

23. A. L. Hooberman et al., Blood 74,1101 (1989).

24. C. A. Westbrook et al., Blood 71 (3):697-702 (1988).

25. B. Trask, D. Pinkel, and G. van den Engh, Genomics 5,710 (1989).

26. S. J. Collins and Μ. T. Groudine, Proc Natl Acad Sci USA 80,4813 (1983).

27. D. Pinkel et al., Proc Natl Acad Sci USA 83,2934 (1986).

28. D. Pinkel, T. Straume and J. W. Gray., Proc Natl Acad Sci USA 85,9138 (1988).

29. B. Trask and J. Hamlin, Genes and Development, 3:1913 (1989).

30. J. B. Lawrence, C.A. Villnave and R.H. Singer, Cell 42,51 (1988).

31. G. D. Johnson and J. G. Nogueria J. Immunol. Methods 43, 349 (1981).

32. Hegewisch-Becker et al., J. Cell. Biochem. (Suppl.) 13E, 289 (1989).

33. Kohler et al., Expression of BCR-ABL from Transcripts Following Bone Marrow Transplant for Philadelphia Chromosome Positive Leukemias, (manuscript submitted).

34. Heisterkamp et al., Nature, 315:758 (1985).

35. Heisterkamp et al, J. Molec. Appl. Genet., 2:57 (1983).

Fusion of the proto-oncogene c־ABL from the long arm of chromosome 9 with the BCR gene of chromosome 22 is a consistent finding in CML (1-3). That genetic change leads to formation of a BCR-ABL transaipt that is translated to form a 210 kd protein present in virtually all cases of CML (4-6). In 90% of the cases, the fusion gene results from a reciprocal translocation involving chromosomes 9 and 22 producing a cytogenetically distinct small acrocentric chromosome called the Philadelphia (Phi) chromosome (7-12), Fig. 8. However, standard cytogenetics does not have the resolution to distinguish closely spaced breakpoints, such as those characteristic of CML and acute lymphocytic leukemia (ALL), and misses fusions produced by more complex rearrangements. Mapping and cloning of the breakpoint regions in both genes has lead to molecular techniques capable of demonstrating BCR-ABL fusion in CML cases where the Phi chromosome could not be detected cytogenetically (13-16). Southern analysis for BCR rearrangements has become the standard for diagnosis of CML. More recently, fusion has been detected by in vitro amplification of a cDNA transcript copied from CML mRNA using reverse transcriptase (17-23). That • -1694906/2 technique permits detection of BCR-ABL transcript from CML cells present at low frequencies. Both of those techniques utilize nucleic acid obtained from cell populations so that correlation between genotype and phenotype for individual cells is not possible.

Described herein are chromosome-specific reagents and methods to detect genetic rearrangements, such as those exemplified herein for the BCR-ABL fusion, that supply information unavailable by existing techniques.

SUMMARY OF THE INVENTION

The present invention relates to a method of staining targeted chromosomal material based upon nucleic acid sequence to detect one or more genetic rearrangements identified with chronic myelogenous leukemia (CML) comprising the steps of:

a. hybridizing in situ to a human genome high complexity nucleic acid probes which contain nucleic acid sequences which are substantially complementary to nucleic acid sequences that flank and/or extend partially or fully across breakpoint regions known to be associated with genetic rearrangements identified with CML; and

b. observing the staining pattern produced to determine if it is indicative of one or more genetic rearrangements identified with CML; and wherein said method is performed outside the human body.

Preferably in such method the genetic rearrangement is selected from the group consisting of translocations, deletions, inversions, amplifications and insertions, and said nucleic acid probes are substantially complementary to nucleic acid

CL94906/2 sequences which flank and/or extend partially or fully across the translocation breakpoint regions of chromosomal regions 9q34 and 22q11 associated with

CML

The invention furthermore relates to chromosome-specific staining reagent comprising a high complexity heterogeneous mixture of labeled nucleic acid fragments, wherein the nucleic acid sequences of the labeled nucleic acid fragments are complementary to sites on targeted chromosomal material which flank and/or extend partially or fully across breakpoint regions known to be associated with one or more genetic rearrangements identified with chronic myelogenous leukemia (CML) and are substantially free of nucleic acid sequences having hybridization capacity to sites on non-targeted chromosomal material.

The invention also relates to a chromosome-specific staining reagent that provides staining patterns indicative of one or more genetic rearrangements associated with chronic myelogenous leukemia (CML) produced by the process of:

isolating chromosome-specific DNA which flank and/or extend across chromosomal regions wherein the BCR and ABL genes are located;

amplifying pieces of the isolated chromosome-specific DNA; disabling the hybridization capacity of and/or removing shared repetitive sequences contained in the amplified pieces of the isolated DNA to form a collection of nucleic acid fragments which hybridize predominantly to targeted chromosomal DNA which

94906/2 flank and/or extend partially or fully across breakpoint regions known to be associated with one or more genetic rearrangements identified with CML,; and labeling the nucleic acid fragments of the collection to form a heterogeneous mixture of nucleic add fragments.

The passages which do not fall within the scope of the claims do not belong to the invention. The scope of the protection is as defined by the claims and according to the stipulations of the Patent Law (1968).

This invention concerns methods of staining chromosomal material based upon nucleic acid sequence that employ one or more nucleic acid probes. Said methods produce staining patterns that can be tailored for specific cytogenetic analyses. It is further an object of this invention to produce nucleic add probes that are useful for cytogenetic analysis, that stain chromosomal material with reliable signals. Such probes are appropriate for in situ hybridization. Preferred nucleic add probes for certain applications of this invention are those of sufficient complexity to stain reliably each of two or more target sites.

The invention provides methods and compositions for staining chromosomal material. The probe compositions of this invention at the current state of hybridization techniques are typically of high complexity, usually greater than about 50 kb of complexity, the complexity depending upon the application for which the probe is designed. In particular, chromosome specific staining reagents are provided which comprise heterogeneous mixtures of nucleic add fragments, each fragment having a substantial fraction of its sequences substantially complementary to a portion

־17־ of the nucleic add for which spedfic staining is desired - the target nudeic add, preferably the target chromosomal material. In general, the nudeic acid fragments are labeled by means as exemplified herein and indicated infra. However, the nudeic add fragments need not be directly labeled in order for the binding of probe fragments to the target to be detected; for example, such nudeic add binding can be detected by anti-RNA/DNA duplex antibodies and antibodies to thymidine dimers. The nucleic acid fragments of the heterogenous mixtures include double-stranded or single-stranded RNA or DNA.

This invention concerns chromosome specific reagents and methods of staining targeted chromosomal material that is in the vidnity of a suspected genetic rearrangement. Such genetic rearrangement indude but are not limited to translocations, inversions, deletions, amplifications and insertions. When such a genetic rearrangement is associated with a disease, such chromosome spedfic reagents are referred to as disease spedfic reagents or probes. When such a genetic rearrangement is associated with cancer, such reagents are referred to as tumor spedfic reagents or probes.

This invention provides for nucleic add probes that reliably stain targeted chromosomal materials in the vicinity of one or more suspected genetic rearrangements. Such nucleic acid probes useful for the detection of genetic rearrangements are typically of high complexity. Such nucleic acid probes preferably comprise nudeic acid sequences that are substantially homologous to nudeic acid sequences in chromosomal regions that flank and/or extend partially or fully across breakpoints assodated with genetic rearrangements.

This invention further provides for methods and reagents to distinguish between cytogenetically similar but genetically different chromosomal rearrangements.

Specifically herein exemplified are chromosome specific regents and methods to detect genetic rearrangements, e.g., translocations, deletions, amplifications and insertions, that produce the BCR-ABL fusion which is diagnostic for chronic myelogenous leukemia (CML). Such chromosome specific reagents for the diagnosis of CML contain nucleic acid sequences which are substantially homologous to chromosomal sequences in the vicinity of the translocation breakpoint regions of chromosomal regions 9q34 and 22qll associated with CML.

Those reagents produce a staining pattern which is distinctively altered when the BCR-ABL fusion characteristic of CML occurs. Figure 11 graphically demonstrates a variety of staining patterns which, along with other potential staining patterns, are altered in the presence of a genetic rearrangement, such as, the BCR-ABL fusion.

The presence of a genetic rearrangement can be determined by applying the reagents of this invention according to methods herein described and observing the proximity of and/or other characteristics of the signals of the staining patterns produced.

Preferably, the chromosome specific reagents used to detect CML of this invention comprise nucleic acid sequences having a complexity of from about 50 kilobases (kb) to about 1 megabase (Mb), more preferably from about 50 kb to about 750 kb, and still more preferably from about 200 kb to about 400 kb.

This invention further provides for methods of distinguishing between suspected genetic rearrangements that occur in relatively close proximity in a genome wherein the chromosome specific reagents comprise nucleic add sequences substantially homologous to nudeic add sequences in the vicinity of said suspected genetic rearrangements. An example of such a differentiation between two potential genetic rearrangements is the differential diagnosis of CML from acute lymphocytic leukemia (ALL).

This invention still further provides methods and reagents for producing staining patterns in a patient who is afflicted with a disease associated genetic rearrangement, such as those associated with the BCR-ABL fusion in CML, wherein said staining patterns are predictive and/or indicative of the response of a patient to various therapeutic regimens, such as chemotherapy, radiation, surgery, and transplantation, such as bone marrow transplantation. Such staining patterns can be useful in monitoring the status of such a patient, preferably on a cell by cell basis, and can be predictive of a disease recurrence for a patient that is in remission. Computer assisted microscopic analysis can assist in the interpretation of staining patterns of this invention, and the invention provides for methods wherein computer assisted microscopic analysis is used in testing patient cells on a call by cell basis, for e.g., to search for residual disease in a patient.

Still further, this invention provides for methods and reagents to determine the molecular basis of genetic disease, and to detect specific genetically based diseases.

Still further, this invention provides for methods and reagents for detecting contiguous gene syndromes comprising the in situ hybridization of nucleic acid probes which comprise sequences which are substantially

־20־ homologous to nucleic add sequences characteristic of one or more components of a contiguous gene syndrome. Representative of such a contiguous gene syndrome is Down syndrome.

Also provided are methods of simultaneously detecting genetic rearrangements of multiple loci in a genome comprising in situ hybridization of high complexity nudeic add probes comprising nucleic add sequences that are substantially homologous to nucleic acid sequences in multiple lod in a genome.

Still further provided are methods of searching for genetic rearrangements in a genome. For example, conventional banding analysis may indicate an abnormality in a chromosomal region of a genome under examination. Methods of this invention may include the application of nucleic acid probes, produced from the vidnity of that chromosomal region of a normal genome, by in situ hybridization to cells containing the abnormality to detail the exact location and kind of genetic rearrangement of said abnormality by observation of the staining patterns so produced.

The invention still further provides for high complexity nucleic acid probes which have been optimized for rapid, effident and automated detection of genetic rearrangements.

One way to produce a probe of high complexity is to pool several or many clones, for example, phage, plasmid, cosmid, and/or YAC clones, among others, wherein each clone contains an insert that is capable of hybridizing to some part of the target in a genome. Another way to produce such a probe is to use the polymerase chain reaction (PCR).

Heterogeneous in reference to the mixture of labeled nudeic add fragments means that the staining reagents comprise many copies each of fragments having different sequences and/or sizes (e.g., from the different DNA clones pooled to make the probe). In preparation for use, these fragments may be cut, randomly or specifically, to adjust the size distribution of the pieces of nucleic add partidpating in the hybridization reaction.

As discussed more fully below, preferably the heterogeneous probe mixtures are substantially free from nucleic acid sequences with hybridization capacity to non-target nudeic acid. Most of such sequences bind to repetitive sequences which are shared by the target and non-target nucleic adds, that is, shared repetitive sequences.

Methods to remove undesirable nucleic add sequences and/or to disable the hybridization capacity of such sequences are discussed more fully below. [See Section H]. Such methods include but are not limited to the selective removal or screening of shared repetitive sequences from the probe; careful selection of nucleic acid sequences for inclusion in the probe; blocking shared repetitive sequences by the addition of unlabeled genomic DNA, or, more carefully selecting nucleic add sequences for inclusion in the blocking mixture; incubating the probe mixture for sufficient time for reassociation of high copy repetitive sequences, or the like.

Preferably, the staining reagents of the invention are applied to interphase or metaphase chromosomal DNA by in situ hybridization, and the chromosomes are identified or dassified, i.e., karyotyped, by detecting the presence of the label, such as biotin or on the nucleic acid fragments comprising the staining reagent.

The invention includes chromosome staining reagents for the total genomic complement of chromosomes, staining reagents specific to single chromosomes, staining reagents specific to subsets of chromosomes, and staining reagents specific to subregions within single or multiple chromosomes. The term chromosome-specific;׳ is understood to encompass all of these embodiments of the staining reagents of the invention. The term is also understood to encompass staining reagents made from and directed against both normal and abnormal chromosome types.

A preferred method of making the chromosome-specific staining reagents of the invention includes: 1) isolating chromosomal DNA from a particular chromosome type or target region or regions in the genome, 2) amplifying the isolated DNA to form a heterogeneous mixture of nucleic acid fragments, 3) disabling the hybridization capacity of or removing shared repeated sequences in the nucleic add fragments, and 4) labeling the nucleic acid fragments to form a heterogeneous mixture of labeled nucleic add fragments. As described more fully below, the ordering of the steps for particular embodiments varies according to the particular means adopted for carrying out the steps.

The present invention addresses problems associated with karyotyping chromosomes, espedally for diagnostic and dosimetric applications. In particular, the invention overcomes problems which arise because of the lack of stains that are suffidently chromosome-specific by providing reagents comprising heterogeneous mixtures of nudeic acid fragments that can be hybridized to the target DNA and/or RNA, e.g., the target chromosomes, target subsets of chromosomes, or target regions of specific chromosomes. The staining technique of the invention opens up the possibility of rapid and highly sensitive detection of chromosomal abnormalities, particularly genetic rearrangements, in both metaphase and interphase cells using standard clinical and laboratory equipment and improved analysis using automated techniques. It has direct application in genetic screening, cancer diagnosis, and biological dosimetry.

This invention further specifically provides for methods and nucleic acid probes for staining fetal chromosomal material, whether condensed, as in metaphase, or dispersed as in interphase. Still further, the invention provides for a non-embryo-invasive method of karyotyping the chromosomal material of fetal cells, wherein the fetal cells have been separated from maternal blood. Such fetal cells are preferably leukocytes and/or cytotrophoblasts. Exemplary nucleic acid probes are high complexity probes chromosome-specific for chromosome types 13,18 and/or 21. Representative probes comprise chromosome-specific Bluescribe plasmid libraries from which a sufficient number of shared repetitive sequences have been removed or the hybridization capacity thereof has been disabled prior to and/or during hybridization with the target fetal chromosomes.

This invention still further provides for test kits comprising appropriate nucleic acid probes for use in tumor cytogenetics, in the detection of disease related lod, in the analysis of structural abnormalities, for example translocations, among other genetic rearrangements, and for biological dosimetry.

This invention further provides for prenatal screening kits comprising appropriate nucleic acid probes of this invention. This invention also provides for test kits comprising high complexity probes for the detection of genetic rearrangements, and specifically for those producing the BCR-ABL fusion characteristic of CML.

The methods and compositions of this invention permit staining of chromosomal material with patterns appropriate for a desired application.

The pattern may extend over some regions of one or more chromosomes, or over some or all the chromosomes of a genome and may comprise multiple distinguishable sections, distinguishable, for example, by multiple colors. Alternatively, the pattern may be focused on a particular portion or portions of a genome, such as a portion or portions potentially containing a deletion or breakpoint that is diagnostically or prognostically important for one or more tumors, or on those portions of chromosomes having significance for prenatal diagnosis.

The staining patterns may be adjusted for the analysis method employed, for example, either a human observer or automated equipment, such as, flow cytometers or computer assisted microscopy. The patterns may be chosen to be appropriate for analysis of condensed chromosomes or dispersed chromosomal material.

The invention further provides for automated means of detecting and analyzing chromosomal abnormalities, particularly genetic rearrangements, as indicated by the staining patterns produced according to this invention.

Another object of the present invention is to provide an alternative method to currently available techniques for preparing and applying nonselfcomplementary single stranded DNA hybridization probes.

Another object of the invention is to improve the signal-to-noise ratios attainable in in situ hybridization by reducing nonspecific and mismatched binding of probe.

Another object of the invention is to provide a method of denaturing double stranded target DNA for application of hybridization probe which minimizes single stranded regions available for hybridization that are noncomplementary to probe sequences.

DNA fragments from which probes can be constructed by treating with a restriction endonuclease which generates a collection of restriction fragments having sticky ends, or staggered cuts, characteristic of the endonuclease used. That is, the two fragment ends introduced by a cut each consist of a protruding strand and a recessed strand. The restriction fragments are inserted into vectors which have been engineered to accept that type of restriction fragment; and the vectors are transfected into host organisms which are grown to increase the number of restriction fragments. Next the vectors are separated from the host organisms, and the restriction fragments are excised and separated from the vectors. On each end of the restriction fragments the recessed strands are digested by an appropriate exonuclease. Digestion is not allowed to go to completion. The exonuclease treated restriction fragments are then used as template/primers for DNA polymerase which replaces the digested strand in the presence of a labeled precursor. Examples of enzymes suitable for this process are exonuclease ΠΙ followed by treatment with the large fragment of DNA polymerase I; or T4 DNA polymerase, which can perform both functions by changes in reaction conditions. After synthesis is completed, the restriction fragments are broken into smaller fragments such that the labeled portions of the original restriction fragment remain substantially intact. The smaller fragments are denatured, and the labeled strands are separated from the unlabeled strands to form the hybridization probes.

Under this method of using single-stranded probes, before application of the hybridization probe to the target DNA, the target DNA is first treated with the same restriction endonuclease used to excise the probe DNA from the cloning vector. This treatment breaks the target DNA into a collection of restriction fragments having tails at each end characteristic of the restriction endonuclease. Next the target DNA is treated with an exonuclease which removes the recessed strand, thereby exposing single stranded DNA in the vicinity of the cut introduced by the restriction endonuclease. Finally, the hybridization probe is applied to the target DNA, e.g., using standard in situ hybridization protocols, as desaibed more fully below.

An important feature of the single stranded probe method is treating the cloned probe DNA and the target DNA with the same restriction endonuclease. This ensures that the single stranded DNA of the target is complementary to the labeled strand of the probe. Of course many segments of the target in addition to the correct binding sites will be made single stranded because there are many restriction cuts, but there will be much less total single stranded target than would be made by indiscriminant denaturation. In addition, target DNA rendered single stranded in this manner cannot reanneal with itself and thus block access to the probe.

Another important (but not critical) feature of such a method is the selection of a label which permits labeled strands to be separated from unlabeled strands. Preferably precursors are labeled by biotinylation, and the labeled strands are separated from unlabeled strands by affinity chromatography.

-ηBRIEF DESCRIPTION OF THE DRAWINGS

Figures ΙΑ, B and C and Figures 2A and 2B illustrate the hybridization of a chromosome-specific 21 library to human metaphase spread wherein the inserts were cloned in Lambda phage Charon 21 A. The hybridization capacity of the high copy repetitive sequences in the library was reduced by the addition of unlabeled genomic DNA to the hybridization mixture. The probe was labeled with biotin, which was detected with green FITC-avidin (fluorescein isothiocyanate avidin). All of the DNA in the chromosomes was stained with the blue fluorescent dye DAPI (4,6-diamidino2־phenylindole).

Figure 1A is a binary image of the DAPI stain in the human metaphase spread obtained by using a TV camera attached to a fluorescence microscope. Filters appropriate for DAPI visualization were used. Computer processing of the image shows all portions above a chosen threshold intensity as white, and the rest as black.

Figure IB is a binary image of the FITC staining of the same human metaphase spread as in Figure 1 A. The image was processed as in Figure 1A but the filter was changed in the miaoscope such that the FITC attached to the probe is visible rather than the DAPI.

Figure 1C is a binary image of the chromosome 21s alone, nonspecifically stained objects (which are smaller) having been removed by standard image processing techniques on the binary image of Figure IB.

Figure 2A is a color photograph of the DAPI stain in a human metaphase spread which was prepared and hybridized contemporaneously with the spread shown in the computer generated binary images of Figures ΙΑ, B and C.

Figure 2B is a color photograph of the fluorescein attached to the DNA probe in the same human metaphase spread as shown in Figure 2A. It was obtained by changing the filters in the fluorescence microscope to excite fluorescein rather than DAPI. The photograph is comparable to the binary image of Figure IB.

Figure 3 is a photograph of a human metaphase spread prepared and hybridized contemporaneously with the spreads shown in Figures ΙΑ, B and C and 2A and B. The procedures used were the same except that PI (propidium iodide) instead of DAPI, was used to stain all the chromosomes. Both PI and fluorescein stains can be viewed with the same microscope filters. Color film was used such that the propidium iodide counterstain appears red and the fluorescein of the probe appears yellow on the color film.

Figure 4A shows the hybridization of the chromosome 4־specific library in Bluescribe plasmids (the library pBS-4) to a human metaphase spread wherein no unlabeled human genomic DNA was used, and wherein the hybridization mixture was applied immediately after denaturation. Both copies of chromosome 4 are seen as slightly brighter than the other chromosomes. The small arrows indicate regions that are unstained with the probe. As in Figure 3 and as in the rest of the Figures below, PI is the counterstain and fluorescein is used to label the probe.

Figure 4B shows the hybridization of pBS4־ to a human metaphase spread wherein unlabeled human genomic DNA was used during the hybridization (Q = 2 of genomic DNA; the meaning of Q is explained infra). Quantitative image analysis shows that the intensity per unit length of the chomosome 4s is about 20X that of the other chromosomes. The chromosome 4s are yellow; the other chromosomes are red due to the

־29־ propidium iodide counterstain. Two layers of avidin-fluorescein isothiocyanate have been used to make the target chromosomes sufficiently bright to be measured accurately. However, the number 4 chromosomes can be recognized easily after a single layer is applied.

Figure 4C shows the same spread as in Figure 4B but through a filter that passes only the fluorescein isothiocyanate fluorescence.

Figure 4D shows the detection of a radiation-induced translocation (arrows) involving chromosome 4s in a human metaphase spread wherein pBS-4 specific libraries are used. The contrast ratio is about 5X.

Figure 4E shows that normal and two derivative chromosomes resulting from a translocation between chromosome 4 and 11 (in cell line RS4;11) can be detected by the compositions and methods of this invention in interphase nuclei. They appear as three distinct domains.

Figure 4F shows the hybridization of the chromosome 21-specific library in Bluescribe plasmids (the library pBS21־) to a metaphase spread of a trisomy 21 cell line. A small amount of hybridization is visible near the centromeres of the other acrocentric chromosomes.

Figure 4G shows the same hybridization as in Figure 4F but with interphase nuclei. Clearly shown are the three chromosome 21 domains.

Figure 4H shows the hybridization with a pool of 120 single copy probes from chromosome 4 to a human metaphase spread. The number 4 chromosomes are indicated by arrows.

Figure 5 shows the hybridization of a yeast artificial chromosome (YAC) clone containing a 580 kb insert of human DNA to a human metaphase spread. A yellow fluorescein band on each of the chromosome 12s (at 12q21.1) is visible against the propidium iodide counterstain.

Figure 6 shows the hybridization of DNA from a human/hamster hybrid cell containing one copy of human chromosome 19 to a human metaphase spread. A little to the right of the photograph's center are the two chromosome 19s which are brighter than the other chromosomes in the spread.

Figure 7 illustrates a representative method of using the polymerase chain reaction (PCR) to produce probes of this invention which are reduced in repetitive sequences.

Figure 8 illustrates the locations of probes to the CML breakpoint and corresponding pattern of staining in both normal and CML metaphase and interphase nuclei.

The left side shows schematic representations of the BCR gene on chromosome 22, the ABL gene of chromosome 9, and the BCR-ABL fusion gene on the Philadelphia chromosome. Also shown are the locations of CML breakpoints and their relation to the probes (32). The right shows hybridization patterns expected for the c-hu-ABL and PEM12 probes to normal and CML metaphase spreads and interphase nuclei.

Figure 9 shows fluorescence in-situ hybridization (FISH) in metaphase spreads and interphase nuclei. Panels A and B show ABL and BCR hybridization to normal metaphase spreads. The ABL signal (A) is localized to the telomeric portion of 9q and the BCR signal (B) is localized near the centromere of 22q. Panel C shows that abl staining is localized to the telomeric region of Philadelphia chromosome in a case of CML with 46XY, t (9:22) (q34;qll). Panel D shows that abl staining is interstitial on the derivative 22 chromosome arising from an insertional event in a case of CML with 46XY ins (22:9)(qll;q34). Panel E illustrates that the K562 cell line

־31־ presents multiple signals localized to a region of the interphase nucleus. Identical staining pattern was seen with BCR probe indicating BCR -ABL fusion gene amplification. Panel F presents a metaphase spread from the K562 cell line showing fusion gene amplification localized to a single chromosome.

Figure 10 illustrates fluorescence in־situ hybridization in CML interphase nuclei with ABL (red) and BCR (green) probes visualized simultaneously through a double band pass filter. Cells from a CML patient show the red-green (yellow) signals resulting from the hybridization to the BCR-ABL fusion gene and single red and green hybridization signals to the normal BCR and ABL genes on chromosomes 22 and 9.

Figure 11 illustrates some exemplary probe strategies for detection of structural aberrations. The design of the binding pattern, colors etc., of the probe can be optimized for detection of genetic abnormalities in metaphase and/or interphase cells. Different patterns may have advantages for particular applications. The drawings in Figure 11 illustrate some of the patterns useful for detection of some abnormalities. The examples are representative and not meant to be exhaustive; different patterns can be combined to allow for the detection of multiple abnormalities in the same cell.

In the drawings of Figure 11, the metaphase chromosomes are shown with probe bound to both chromatids. The interphase nuclei are pictured to be in a stage of the cell cycle prior to replication of the portion of the chromosome to which the probe binds; thus there is only one chromatid for each interphase chromosome. When the probe binding is restricted to only a portion of a chromosome, the signal is indicated as either a black or white circle. Such a representation is employed to indicate different colors or otherwise distinguishable characteristics of the staining. Patterns containing more than two distinguishable characteristics (three colors, different ratios of colors etc.) permit more complex staining patterns than those illustrated. Chromosomal locations of the breakpoints in the DNA are indicated with horizontal lines next to the abnormal chromosomes.

a. Section a) represents the use of a probe which stains a whole chromosome. Such a probe can be used to detect a translocation that occurs anywhere along the chromosome. The color photograph of Figure 12 shows use of such a stain for chromosome 22 to detect a translocation, in this case that which occurs with CML. Such an approach to staining is not very useful in interphase nuclei since the region of the nucleus that is stained is relatively large; overlaps in the stained regions can make interpretation difficult in many nuclei.

b. Section b) represents the reduction of the stained region of the chromosome shown in a) to that in the vicinity of a breakpoint, providing information focused on events in that region. The staining pattern can be continuous or discontinuous aCT0ss the breakpoint, just so that some binding is on both sides of the breakpoint. Such a staining pattern requires only one color, but gives no information about which other genomic region may be involved in the exchange.

c. Section c) represents the use of a probe which binds to sequences which come together as a result of the rearrangement and allows for the detection in metaphase and interphase cells. In this case the different sequences are stained with different colors. Such a staining pattern is that used in the examples of Section VIII of the this application.

־33-

d. Section d) represents an extension of c) by including staining of both sides of both breakpoints involved in the rearrangement. Different colors are used as indicated. The additional information supplied by the more complex staining pattern may assist with interpretation of the nuclei. It might also permit recognition of an apparent insertional event as discussed herein.

e. Section e) represents the detection of an inversion in one homologue of a chromosome.

f. Section f) represents a staining pattern useful in the detection of a deletion. A deletion could also be detected with a probe that stains only the deleted region; however, lack of probe binding may be due to reasons other than deletion of the target sequence. The flanking regions stained a different color serve as controls for hybridization.

Figure 12 illustrates a staining pattern to detect a rearrangement by staining a whole chromosome, in this case a rearrangement of chromosome 22 associated with CML. The metaphase spread of this figure is from a CML cell that has been stained with a probe which binds all along chromosome 22. Probe-stained regions appear yellow. The rest of the DNA has been stained with the red-fluorescing chemical stain propidium iodide. The entirely yellow chromosome is a normal copy of chromosome 22. Just below said normal chromosome 22 is the Philadelphia chromosome, a small part yellow and part red chromosome. Below and to the right of the Philadelphia chromosome is the abnormal chromosome 9 (red) with the distal part of chromosome 22 (yellow) attached. The photograph of this figure illustrates the staining pattern represented in part a) of the previous figure.

DETAILED DESCRIPTION OF THE INVENTION

This invention concerns the use of nucleic acid probes to stain targeted chromosomal material in patterns which can extend along one or more whole chromosomes, and/or along one or more regions on one or more chromosomes, including patterns which extend over an entire genome. The staining reagents of this invention facilitate the microscopic and/or flow cytometric identification of normal and aberrant chromosomes and provide for the characterization of the genetic nature of particular abnormalities, such as, genetic rearrangements. The term chromosome-specific is herein defined to encompass the terms target specific and region specific, that is, when the staining composition is directed to one chromosome, it is chromosome-specific, but it is also chromosome-specific when it is directed, for example, to multiple regions on multiple chromosomes, or to a region of only one chromosome, or to regions across the entire genome. The term chromosome-specific originated from the use of recombinant DNA libraries made by cloning DNA from a single normal chromosome type as the source material for the initial probes of this invention. Libraries made from DNA from regions of one or more chromosomes are sources of DNA for probes for that region or those regions of the genome. The probes produced from such source material are region-specific probes but are also encompassed within the broader phrase chromosome-specific probes. The term target specific is interchangeably used herein with the term chromosome-specific.

The word specific as commonly used in the art has two somewhat different meanings. The practice is followed herein. Specific may refer to the origin of a nucleic acid sequence or to the pattern with which it will hybridize to a genome as part of a staining reagent. For example, isolation and doning of DNA from a specified chromosome results in a chromosomespecific library. [Eg., Van Dilla et al., Human Chromosome-Spedfic DNA Libraries: Construction and Availability, Biotechnology, 4:537 (1986).] However, such a library contains sequences that are shared with other chromosomes. Such shared sequences are not chromosome-specific to the chromosome from which they were derived in their hybridization properties since they will bind to more than the chromosome of origin. A sequence is chromosome-spedfic if it binds only to the desired portion of a genome. Such sequences indude single-copy sequences contained in the target or repetitive sequences, in which the copies are contained predominantly in the target.

Chromosome-spedfic in modifying staining reagent refers to the overall hybridization pattern of the nucleic acid sequences that comprise the reagent. A staining reagent is chromosome-specific if useful contrast between the target and non-target chromosomal material is achieved (that is, that the target can be adequately visualized).

A probe is herein defined to be a collection of nucleic acid fragments whose hybridization to the target can be detected. The probe is labeled as described below so that its binding to the target can be visualized. The probe is produced from some source of nucleic add sequences, for example, a collection of dones or a collection of polymerase chain reaction (PCR) products. The source nucleic add may then be processed in some way, for example, by removal of repetitive sequences or blocking them with unlabeled nucleic add with complementary sequence, so that hybridization with the resulting probe produces staining of sufficient contrast on the target. ׳Thus, the word probe may be used herein to refer not only to the detectable nucleic ־36acid, but also to the detectable nucleic add in the form in which it is applied to the target, for example, with the blocking nucleic add, etc. The blocking nucleic acid may also be mentioned separately. What probe refers to specifically should be clear from the context in which the word is used.

When two or more nucleic add probes of this invention are mixed together, they produce a new probe which when hybridized to a target according to the methods of this invention, produces a staining pattern that is a combination of the staining patterns individually produced by the component probes thereof. Thus, the terms probe and probes (that is, the singular and plural forms) can be used interchangeably within the context of a staining pattern produced. For example, if one probe of this invention produces a dot on chromosome 9, and another probe produces a band on chromosome 11, together the two probes form a probe which produces a dot/band staining pattern.

The term labeled is herein used to indicate that there is some method to visualize the bound probe, whether or not the probe directly carries some modified constituent. Section III infra describes various means of directly labeling the probe and other labeling means by which the bound probe can be detected.

The terms staining or painting are herein defined to mean hybridizing a probe of this invention to a genome or segment thereof, such that the probe reliably binds to the targeted chromosomal material therein and the bound probe is capable of being visualized. The terms staining or painting are used interchangeably. The patterns resulting from staining or painting are useful for cytogenetic analysis, more particularly, molecular cytogenetic analysis. The staining patterns facilitate the microscopic and/or flow cytometric identification of normal and abnormal chromosomes and the characterization of the genetic nature of particular abnormalities. Section ΙΠ infra desaibes methods of rendering the probe visible. Since multiple compatible methods of probe visualization are available, the binding patterns of different components of the probe can be distinguished-for example, by color. Thus, this invention is capable of producing any desired staining pattern on the chromosomes visualized with one or more colors (a multicolor staining pattern) and/or other indicator methods. The term staining as defined herein does not include the concept of staining chromosomes with chemicals as in conventional karotyping methods although such conventional stains may be used in conjunction with the probes of this invention to allow visualization of those parts of the genome where the probe does not bind. The use of DAPI and propidium iodide for such a purpose is illustrated in the figures.

The phrase high complexity is defined herein to mean that the probe, thereby modified contains on the order of 50,000 (50 kb) or greater, up to many minions or several billions, of bases of nucleic acid sequences which are not repeated in the probe. For example, representative high complexity nucleic add probes of this invention can have a complexity greater than 50 kb, greater than 100,000 bases (100 kb), greater than 200,000 (200 kb), greater than 500,000 bases (500 kb), greater than one million bases (1 Mb), greater than 2 Mb, greater than 10 Mb, greater than 100 Mb, greater than 500 Mb, greater than 1 billion bases and still further greater than several billion bases.

The term complexity is defined herein according to the standard for nucleic acid complexity as established by Britten et al., Methods of Enzymol., 29:363 (1974). See also Cantor and Schimmel, Biophysical Chemistry: Part ΠΙ:

The Behavior of Biological Macromolecules, at 1228-1230 (Freeman and Co. 1980) for further explanation and exemplification of nucleic add complexity.

The complexity preferred for a probe composition of this invention is dependent upon the application for which it is designed. In general, the larger the target area, the more complex is the probe. It is antidpated that the complexity of a probe needed to produce a desired pattern of landmarks on a chromosome will deaease as hybridization sensitivity increases, as progress is made in hybridization technology. As the sensitivity increases, the reliability of the signal from smaller target sites will increase. Therefore, whereas from about a 40 kb to about a 100 kb target sequence may be presently necessary to provide a reliable, easily detectable signal, smaller target sequences should provide reliable signals in the future. Therefore, as hybridization sensitivity increases, a probe of a certain complexity, for example, 100 kb, should enable the user to detect considerably more loci in a genome than are presently reliably detected; thus, more information will be obtained with a probe of the same complexity. The term complexity therefore refers to the complexity of the total probe no matter how many visually distinct lod are to be detected, that is, regardless of the distribution of the target sites over the genome.

As indicated above, with current hybridization techniques it is possible to obtain a reliable, easily detectable signal with a probe of about 40 kb to about 100 kb (eg. the probe insert capacity of one or a few cosmids) targeted to a compact point in the genome. Thus, for example, a complexity in the range of approximately 100 kb now permits hybridization to both sides of a tumor-specific translocation. The portion of the probe targeted to one side of the breakpoint can be labeled differently from that targeted to the other side of the breakpoint so that the two sides can be differentiated with different colors, for example. Proportionately increasing the complexity of the probe permits analysis of multiple compact regions of the genome simultaneously. The conventional banding patterns produced by chemical stains may be replaced according to this invention with a series of probe-based, color coded (for example), reference points along each chromosome or significant regions thereof.

Uniform staining of an extended contiguous region of a genome, for example, a whole chromosome, requires a probe complexity proportional to but substantially less than, the complexity of the target region. The complexity required is only that necessary to provide a reliable, substantially uniform signal on the target. Section V.B, infra, demonstrates that fluorescent staining of human chromosome 21, which contains about 50 megabases (Mb) of DNA, is sufficient with a probe complexity of about 1 Mb. Figure 4H illustrates hybridization of about 400 kb of probe to human chromosome 4, which contains about 200 Mb of DNA. In that case, gaps between the hybridization of individual elements of the probe are visible. Figures 4B and 4F demonstrate the results achieved with probes made up of entire libraries for chromosomes 4 and 21, respectively. The chromosomes are stained much more densely as shown in Figures 4B and 4F than with the lower complexity probe comprising single-copy nucleic acid sequences used to produce the pattern of Figure 4H.

Increasing the complexity beyond the minimum required for adequate staining is not detrimental as long as the total nucleic acid concentration in the probe remains below the point where hybridization is impaired. The decrease in concentration of a portion of a sequence in the probe is compensated for by the increase in the number of target sites. In fact.

when using double-stranded probes, it is preferred to maintain a relatively low concentration of each portion of sequence to inhibit reassociation before said portion of sequence can find a binding site in the target.

The staining patterns of this invention comprise one or more bands. The term band is herein defined as a reference point in a genome which comprises a target nucleic acid sequence bound to a probe component, which duplex is detectable by some indicator means, and which at its narrowest dimension provides for a reliable signal under the conditions and protocols of the hybridization and the instrumentation, among other variables, used. A band can extend from the narrow dimension of a sequence providing a reliable signal to a whole chromosome to multiple regions on a ! number of chromosomes.

The probe-produced bands of this invention are to be distinguished from bands produced by chemical staining as indicated above in the Background. The probe-produced bands of this invention are based upon nucleic add sequence whereas the bands produced by chemical staining depend on natural characteristics of the chromosomes, but not the actual nucleic add sequence. Further, the banding patterns produced by chemical staining are only interpretable in terms of metaphase chromosomes whereas the probe-produced bands of this invention are useful both for metaphase and interphase chromosomes.

One method of forming the probes of the present invention is to pool many different low complexity probes. Such a probe would then comprise a heterogeneous mixture of individual cloned sequences. The number of dones required depends on the extent of the target area and the capadty of the .

cloning vector. If the target is made up of several discrete, compact loci, that is, single spots at the limit of microscopic resolution, then about 40 kb, more preferably 100 kb, for each spot gives a reliable signal given current techniques. The portion of the probe for each spot may be made up from, for example, a single insert from a yeast artificial chromosome (YAC), from several cosmids each containing 35-40 kb or probe sequence, or from about 25 plasmids each with 4 kb of sequence.

Representative heterogeneous mixtures of clones exemplified herein include phage (Figures 1,2 and 3), and plasmids (Figure 4). Yeast artificial chromosomes (YACS) (Figure 5), and a single human chromosome in an inter-species hybrid cell (Figure 6) are examples of high complexity probes for single lod and an entire chromosome that can be propagated as a single clone.

A base sequence at any point in the genome can be classified as either single-copy or repetitive. For practical purposes the sequence needs to be long enough so that a complementary probe sequence can form a stable hybrid with the target sequence under the hybridization conditions being used. Such a length is typically in the range of several tens to hundreds of nucleotides.

A single-copy sequence is that wherein only one copy of the target nucleic acid sequence is present in the haploid genome. Single-copy sequences are also known in the art as unique sequences. A repetitive sequence is that wherein there are more than one copy of the same target nucleic acid sequence in the genome. Each copy of a repetitive sequence need not be identical to all the others. The important feature is that the sequence be sufficiently similar to the other members of the family of repetitive sequences such that under the hybridization conditions being used, the same fragment of probe nucleic acid is capable of forming stable hybrids with each copy. A shared repetitive sequence is a sequence with some copies in the target region of the genome, and some elsewhere.

When the adjectives single-copy, repetitive, shared repetitive, among other such modifiers, are used to describe sequences in the probe, they refer to the type of sequence in the target to which the probe sequence will bind. Thus, a repetitive probe is one that binds to a repetitive sequence in the target; and a single-copy probe binds to a single-copy target sequence.

Repetitive sequences occur in multiple copies in the haploid genome. The number of copies can range from two to hundreds of thousands, wherein the Alu family of repetitive DNA are exemplary of the latter numerous variety. The copies of a repeat may be clustered or interspersed throughout the genome. Repeats may be clustered in one or more locations in the genome, for example, repetitive sequences occurring near the centromeres of each chromosome, and variable number tandem repeats (VNTRs) [Nakamura et al, Science, 235:1616 (1987)]; or the repeats may be distributed over a single chromosome [for example, repeats found only on the X chromosome as described by Bardoni et al., Cytogenet. Cell Genet., 46:575 (1987)]; or the repeats may be distributed over all the chromosomes, for example, the Alu family of repetitive sequences.

Herein, the terms repetitive sequences, repeated sequences and repeats are used interchangeably.

Shared repetitive sequences can be clustered or interspersed. Clustered repetitive sequences include tandem repeats which are so named because they are contiguous on the DNA molecule which forms the backbone of a chromosome. Clustered repeats are associated with well-defined regions of one or more chromosomes, e.g., the centromeric region. If one or more clustered repeats form a sizable fraction of a chromosome, and are shared with one or more non-target regions of the genome and are consequently removed from the heterogeneous mixture of fragments employed in the invention or the hybridization capacity thereof is disabled, perfect uniformity of staining of the target region may not be possible. That situation is comprehended by the use of the term substantially uniform in reference to the binding of the heterogeneous mixture of labeled nucleic acid fragments to the target.

Chromosome-specific staining of the current invention is accomplished by using nucleic acid fragments that hybridize to sequences specific to the target. These sequences may be either single-copy or repetitive, wherein the copies of the repeat occur predominantly in the target area. Figure 4H and the results of the work detailed in section V infra indicate that probes can be made of single-copy sequences. However, in probes such as that of Figure 4B, low-copy chromosome-specific repeats [Nakamura et al., and Bardoni et al., supral may contribute to the hybridization as well.

If nucleic acid fragments complementary to non-target regions of the genome are included in the probe, for example, shared repetitive sequences or non-specific sequences, their hybridization capacity needs to be sufficiently disabled or their prevalence sufficiently reduced, so that adequate staining contrast can be obtained. Section V and Figure 4H show examples of hybridization with probes that contain pools of clones in which each clone has been individually selected so that it hybridizes to single-copy sequences or very low copy repetitive sequences. The remaining figures illustrate use of probes that contain fragments that could have hybridized to high-copy repetitive sequences, but which have had the hybridization capacity of such sequences disabled.

The nucleic acid probes of this invention need not be absolutely specific for the targeted portion of the genome. They are intended to produce staining contrast. Contrast is quantified by the ratio of the stain intensity of the target region of the genome to that of the other portions of the genome. For example, a DNA library produced by cloning a particular chromosome, such as those listed in Table I, can be used as a probe capable of staining the entire chromosome. The library contains sequences found only on that chromosome, and sequences shared with other chromosomes. In a simplified (approximately true to life) model of the human genome, about half of the chromosomal DNA falls into each class. If hybridization with the whole library were capable of saturating all of the binding sites, the target chromosome would be twice as bright (contrast ratio of 2) as the others since it would contain signal from the specific and shared sequences in the probe, whereas the other chromosome would only have signal from the shared sequences. Thus, only a modest decrease in hybridization of the shared sequences in the probe would substantially enhance the contrast. Contaminating sequences which only hybridize to non-targeted sequences, for example, impurities in a library, can be tolerated in the probe to the extent that said sequences do not reduce the staining contrast below useful levels.

In reality all of the target sites may not be saturated during the hybridization, and many other mechanisms contribute to producing staining contrast, but this model illustrates one general consideration in using probes targeted at a large portion of a genome.

־45־

The required contrast depends on the application for which the probe is designed. When visualizing chromosomes and nuclei, etc., microscopically, a contrast ratio of two or greater is often sufficient for identifying whole chromosomes. In Figures 4D-F, the contrast ratio is 35־. The smaller the individual segments of the target region, the greater the contrast needs to be to permit reliable recognition of the target relative to the fluctuations in staining of the non־targeted regions. When quantifying the amount of target region present in a cell nucleus by fluorescence intensity measurements using flow cytometry or quantitative microscopy, the required contrast ratio is on the order of 1/T or greater on average for the genome, where T is die fraction of the genome contained in the targeted region. When the contrast ratio is equal to 1/T, half of the total fluorescence intensity comes from the target region and half from the rest of the genome. For example, when using a high complexity probe for chromosome 1, which comprises about 10% of the genome, the required contrast ratio is on the order of 10, that is, for the chromosome 1 fluorescence intensity to equal that of the rest of the genome.

Background staining by the probe, that is, to the non-target region of the genome, may not be uniform. Figure 4F shows that a chromosome 21 specific probe contains probe fragments that hybridize weakly to compact regions near the centromeres of other acrocentric human chromosomes. This degree of non-spedficity does not inhibit its use in the illustrated applications. For other applications, removal of or further disabling the hybridization capacity of the probe fragments that bind to these sequences may be necessary.

*

For other applications, repetitive sequences that bind to centromeres, for example, alpha-satellite sequences, and/or telomeres can be part of the chromosome-specific staining reagents wherein the target includes some or all of the centromeres and/or telomeres in a genome along with perhaps other chromosomal regions. Exemplary of such an application would be that wherein the staining reagent is designed to detect random structural aberrations caused by clastogenic agents that result in dicentric chromosomes and other structural abnormalities, such as translocations. Addition of sequences which bind to all centromeres in a genome, for example to the probe used to create the staining pattern of Figure 4D, would allow more reliable distinguishing between dicentrics and translocations.

Application of staining reagents of this invention to a genome results in a substantially uniform distribution of probe hybridized to the targeted regions of a genome. The distribution of bound probe is deemed substantially uniform if the targeted regions of the genome can be visualized with useful contrast. For example, a target is substantially uniformly stained in the case wherein it is a series of visually separated loci if most of the loci are visible in most of the cells.

Substantial proportions in reference to the base sequences of nucleic acid fragments that are comple- mentary to chromosomal DNA means that the complementarity is extensive enough so that the fragments form stable hybrids with the chromosomal DNA under the hybridization conditions used. In particular, the term comprehends the situation where the nucleic acid fragments of the heterogeneous mixture possess some regions of sequence that are not perfectly complementary to target chromosomal

-αίmaterial. The stringency can be adjusted to control the precision of the complementarity required for hybridization.

The phrase metaphase chromosomes is herein defined to mean not only chromosomes condensed in the metaphase stage of mitosis but includes any condensed chromosomes, for example, those condensed by premature chromosome condensation.

To disable the hybridization capacity of a nucleic acid sequence is herein sometimes abbreviated as disabling the nucleic acid sequence.

The methods and reagents of this invention find a particularly appropriate application in the field of diagnostic cytogenetics, particularly in the field of diagnostic interphase cytogenetics. Detecting genetic rearrangements that are associated with a disease, such as cancer, are a specific application of the chromosome specific reagents and staining methods of this invention.

Contiguous gene syndromes are an example of the genetic rearrangements that the probes and methods of this invention can identify. Contiguous gene syndromes are characterized by the presence of several closely spaced genes which are in multiple and/or reduced copy number. Down syndrome is an example of a contiguous gene syndrome wherein an extra copy of a chromosomal region containing several genes is present.

Particularly described herein is the application of chromosome specific reagents and methods for detecting genetic rearrangements that produce the BCR-ABL fusion associated with CML. Such reagents are exemplary of disease specific, in this case tumor specific, probes which can be labeled, directly and/or indirectly, such that they are visualizable when bound to the targeted chromosomal material, which in the case of CML, is the vicinity of the translocation breakpoint regions of chromosomal regions 9q34 and 22qll known to be associated with CML. In the examples provided in Section VIII of this application, the probes are labeled such that a dual color fluorescence is produced in the staining pattern of said probes upon in situ hybridization [fluorescent in situ hybridication (FISH)]; however, staining patterns can be produced in many colors as well as other types of signals, and any visualization means to signal the probe bound to its target can be used in the methods of this invention.

Section VUI herein describes representative methods and reagents of this invention to detect genetic rearrangements. The examples of Section VIII concern genetic rearrangements that produce the BCR-ABL fusion that is characteristic of CML. The approach in such examples is based on FISH with probes from chromosomes 9 and 22 that flank the fused BCR and ABL sequences in essentially all cases of CML (Figure 8). The probes when hybridized to the chromosomal material of both normal and abnormal cells produce staining patterns that are different as illustrated in Figures 8-12. The staining patterns produced by such exemplary probes are different in normal and abnormal cells; the staining pattern present when the genetic rearrangement occurs is distinctively altered from that of the staining pattern shown by hybridizing the probes to chromosomal material that does not contain the genetic rearrangement. Further, staining patterns are distinctively different for one type of genetic rearrangement versus another. For example, the staining patterns produced upon hybridization of nucleic add probes of this invention to chromosomal material containing a genetic rearrangement associated with ALL is distinctively different from that produced upon hybridization of such probes to chromosomal material ־49containing the BCR-ABL fusion characteristic of CML. Thus, the methods and reagents of this invention provide for differential diagnosis of related diseases.

The examples of Section Vm provide for the diagnosis of CML based upon the proximity of the fluorescent signals in the staining patterns, and rely upon a 1 micron cutoff point for determination of the presence of a fusion. The proximity distance of signals is only one characteristic, among many others, of signals that can be used to detect the presence of a genetic rearrangement. Further, the proximity distance is dependent on the particular cell preparation techniques employed and the size of the nuclei therein, and for a particular cell preparation is relative depending on the distance between signals in normal and abnormal cells.

The staining patterns exemplified in the examples of Section VIII are representative of one type of probe strategy. Many other probe strategies can be employed. Figure 11 illustrates some other exemplary probe strategies for detecting genetic rearrangements, the patterns of which can be modified and optimized and otherwise varied to detect particular genetic rearrangements.

Use of other disease specific reagents of this invention would be analogous to the methods detailed in Section Vin for CML. For example, the diagnosis and study of acute lymphocytic leukemia (ALL) may be accomplished by replacing the BCR probe (PEM12) of Section VM with a probe from the 5’ end of the BCR gene. ALL is of particular interest because the Ph' chromosome is the most common cytogenetic abnormality in that disease, and the presence of such a chromosome is indicative of a very aggressive neoplasm.

־50־

The methods and reagents herein exemplified, particularly in Section VID, provide for the means to distinguish between cytogenetically similar but genetically different diseases. Cytogenetically in that particular context refers to a similarity determined by conventional banding analysis. CML and ALL are in that context cytogenetically similar in that conventional banding analysis can not distinguish them because the breakpoints associated with each are so close together in the human genome.

Further, this invention provides methods and reagents that can be used in a cytogenetic research mode for the study of the molecular bases of genetic disease. For example, if an abnormality in a person's karyotype is noted by conventional banding analysis, the probes and reagents of this invention can be used to detect any genetic rearrangements in the vicinity of said abnormality. The underlying molecular basis of the abnormality can be determined by the methods and reagents of this invention, and the resulting differences at the genetic level may be indicative of different treatment plans and prognostically important. The underlying genetic rearrangements may be found to be consistently associated with a set of phenotypic characteristics in a population.

The following sections provide examples of making and using the staining compositions of this invention and are for purposes of illustration only and not meant to limit the invention in any way. The following abbreviations are used.

Abbreviations

BN ־ bicarbonate buffer with NP-40

DAPI - 4,6-diamidino-2-phenylindole

DCS - as in fluorescein-avidin DCS (a commercially available cell sorter grade of fluorescein Avidin D)

AAF - N-acetoxy-N-2-acetyl־aminofluorene

EDTA - ethylenediaminetetraacetate

FACS ־ fluorescence-activated cell sorting

FITC - fluorescein isothiocyanate

IB - isolation buffer

NP-40 - non-ionic detergent commercially available from

Sigma as Nonidet P-40 (St. Louis, MO)

PBS - phosphate-buffered saline

PI - propidium iodide

PMSF - phenylmethylsulfonyl fluoride

PN ־ mixture of 0.1 M NaH2PO4 and 0.1 M buffer Na₂HPO4, pH 8; 0.1% NP-40

PNM - Pn buffer plus 5% nonfat dry milk (centrifuged);

buffer 0.02% Na azide

SDS ־ sodium dodecyl sulfate

SSC 0.15 ־ M NaCl /0.015 M Na citrate, pH 7

VNTR - variable number tandem repeat . Methods of Preparing Chromosome-Specific Staining Reagents

LA. Isolation of Chromosome-Specific DNA and Formation of DNA Fragment Libraries.

The first step in a preferred method of making the compositions of the invention is isolating chromosome-specific DNA (which term includes target-specific and/or region-specific DNA, as indicated above, wherein specific refers to the origin of the DNA). This step includes first isolating a sufficient quantity of the particular chromosome type or chromosomal subregion to which the staining composition is directed, then extracting the DNA from the isolated chromosome(s) or chromosomal subregion(s). Here sufficient quantity means sufficient for carrying out subsequent steps of the method. Preferably, the extracted DNA is used to create a library of DNA inserts by cloning using standard genetic engineering techniques.

Preferred cloning vectors include, but are not limited to, yeast artificial chromosomes (YACS), plasmids, bacteriophages and cosmids. Preferred plasmids are Bluesaibe plasmids; preferred bacteriophages are lambda insertion vectors, more preferably Charon 4A, Charon 21 A, Charon 35, Charon 40 and GEM11; and preferred cosmids include Lawrist 4, La wrist 5 and sCosl.

As indicated above, the DNA can be isolated from any source. Chromosome-specific staining reagents can be made from both plant and animal DNA according to the methods of this invention. Important sources of animal DNA are mammals, particularly primates or rodents wherein primate sources are more particularly human and monkey, and rodent sources are more particularly rats or mice, and more particularly mice.

־53-

1. Isolating DNA from an Entire Chromosome. A preferred means for isolating particular whole chromosomes (specific chromosome types) is by direct flow sorting [fluorescence-activated cell sorting (FACS)] of metaphase chromosomes with or without the use of interspecific hybrid cell systems. For some species, every chromosome can be isolated by currently available sorting techniques. Most, but not all, human chromosomes are currently isolatable by flow sorting from human cells, Carrano et al., Measurement and Purification of Human Chromosomes by Flow Cytometry and Sorting, Proc. Natl. Acad. Sd., Vol. 76, pp. 1382-1384 (1979). Thus, for isolation of some human chromosomes, use of the human/rodent hybrid cell system may be necessary, see Kao, Somatic Cell Genetics and Gene Mapping, International Review of Cytology., Vol. 85, pp. 109-146 (1983), for a review, and Gusella et al., Isolation and Localization of DNA Segments from Specific Human Chromosomes, Proc. Natl. Acad. Sci., Vol. 77, pp. 28291980) 2833־). Chromosome sorting can be done by commercially available fluorescenceactivated sorting machines, e.g., Becton Dickinson FACS-II, Coulter Epics V sorter, or special purpose sorters optimized for chromosome sorting or like instrument.

DNA is extracted from the isolated chromosomes by standard techniques, e.g., Marmur, A Procedure for the Isolation of Deoxyribonucleic Add from Miao-Organisms, I. Mol. Biol., Vol. 3, pp. 208-218 (1961); or Maniatis et al., Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory, 1982) pp. 280281־. These references are incorporated by reference for their descriptions of DNA isolation techniques.

Generation of insert libraries from the isolated chromosome-specific DNA is carried out using standard genetic engineering techniques, e.g.,

Davies et al., Cloning of a Representative Genomic Library of the Human X Chromosome After Sorting by Flow Cytometry, Nature, Vol. 293, pp. 374376־ (1981); Krumlauf et al., Construction and Characterization of Genomic Libraries from Specific Human Chromosomes, Proc. Natl. Acad. Sci., Vol. 79, pp. 2971-2975 (1982); Lawn et al., The Isolation and Characterization of Linked Delta-and-Beta-Globin Genes from a Cloned Library of Human DNA. Cell, Vol. 15, pp. 11571978) 1174־); and Maniatis et al., Molecular Cloning: A Laboratory (Cold Springs Harbor Laboratory, 1982), pp. 256308־; Van Dilla et al., id; Fuscoe, Gene, 52:291 (1987); and Fuscoe et al., Cytogenet. Cell Genet., 43:79 (1986). Said references are herein incorporated by reference.

Recombinant DNA libraries for each of the human chromosomes have been constructed by the National Laboratory Gene Library Project and are available from the American Type Culture Collection. [Van Dilla et al., Biotechnology, 4:537 (1986).] Small insert-containing libraries were constructed by complete digestion of flow sorted human chromosome genomic DNA with Hindi!! or EcoRI and cloning into the Lambda insertion vector Charon 21A. The vector is capable of accepting human inserts of up to 9.1'kb in size. Thus, Hindlll (or EcoRI) restriction fragments greater than 9.1'kb will not be recovered from these libraries. The observed average insert size in these libraries is approximately 4 kb. A representative list of the Hind ΙΠ chromosome-specific libraries with their ATCC accession numbers are shown in Table 1.

TABLE

HUMAN CHROMOSOME - SPECIFIC GENOMIC LIBRARIES

<td></td><td> IN CHARON 21A VECTOR</td><td></td>

<td> CHROMOSOME</td><td> ATCC#</td><td> LIBRARY</td>

<td> 1</td><td> 57753</td><td> LL01NS01</td>

<td> 1</td><td> 57754</td><td> LL01NS02</td>

<td> 2</td><td> 57744</td><td> LL02NS01</td>

<td> 3</td><td> 57751</td><td> LL03NS01</td>

<td> 4</td><td> 57700</td><td> LL04NS01</td>

<td> 4</td><td> 57745</td><td> LL04NS02</td>

<td> 5</td><td> 57746</td><td> LL05NS01</td>

<td> 6</td><td> 57701</td><td> LL06NS01</td>

<td> 7</td><td> 57755</td><td> LL07NS01</td>

<td> 8</td><td> 57702</td><td> LL08NS02</td>

<td> 9</td><td> 57703</td><td> LL09NS01</td>

<td> 10</td><td> 57736</td><td> LL10NS01</td>

<td> 11</td><td> 57704</td><td> LL11NS01</td>

<td> 12</td><td> 57756</td><td> LL12NS01</td>

<td> 13</td><td> 57705</td><td> LL13NS01</td>

<td> 13</td><td> 57757</td><td> LL13NS02</td>

<td> 14</td><td> 57706</td><td> LL14NS01</td>

<td> 14/15</td><td> 57707</td><td> LL99NS01</td>

<td> 15</td><td> 57737</td><td> LL15NS01</td>

<td> 16</td><td> 57758</td><td> LL16NS03</td>

<td> 17</td><td> 57759</td><td> LL17NS02</td>

<td> 18</td><td> 57710</td><td> LL18NS01</td>

<td> 19</td><td> 57711</td><td> LL19NS01</td>

<td> 20</td><td> 57712</td><td> LL20NS01</td>

<td> 21</td><td> 57713</td><td> LL21NS02</td>

<td> 22</td><td> 57714</td><td> LL22NS01</td>

<td> X</td><td> 57747</td><td> LL0XNS01</td>

<td> Y</td><td> 57715</td><td> LL0YNS01</td>

Alternatively, the extracted DNA from a sorted chromosome type can be amplified by the polymerase chain reaction (PCR) rather than cloning the extracted DNA in a vector or propagating it in a cell line. Appropriate tails are added to the extracted DNA in preparation for PCR. References for such PCR procedures are set out in Section I.B infra.

Other possible methods of isolating the desired sequences from hybrid cells include those of Schmeckpeper et al., Partial Purification and Characterization of DNA from Human X Chromosome, Proc. Natl. Acad. Sd., Vol. 76, pp. 65251979) 6528־); or Olsen et al., supra (in Background). Accordingly, these references are incorporated by reference.

2. Isolating DNA from a Portion of a Chromosome. Among the methods that can be used for isolating region-specific chromosomal DNA indude the selection of an appropriate chromosomal region from DNA that has previously been mapped, for example, from a library of mapped cosmids; the sorting of derivative chromosomes, for example, by FACS; the microdissection of selected chromosomal material; subtractive hybridization; identification of an appropriate hybrid cell containing a desired chromosomal fragment, extracting and amplifying the DNA, and selecting the desired amplified DNA; and the selection of appropriate chromosomal material from radiation hybrids. The standard genetic engineering techniques outlined above in subsection I.A.l are used in such procedures well-known to those in the art Amplification of the region-specific DNA can be performed by cloning in an appropriate vector, propagating in an appropriate cell line, and/or by the use of PCR (see I.B infra).

A preferred method of isolating chromosomal region-specific DNA is to use mapped short DNA sequences to probe a library of longer DNA sequences, wherein the latter library has usually been cloned in a different vector. For example, a probe cloned in a plasmid can be used to probe a cosmid or yeast artificial chromosome (YAC) library. By using an initial seed probe, overlapping clones in the larger insert library can be found (a process called walking), and a higher complexity probe can be produced for reliable staining of the chromosomal region surrounding the seed probe. Ultimately, when an entire genome for a spedes has been mapped (for example, by the Human Genome Project for the human species), ordered clones for the entire genome of the species will be available. One can then easily select the appropriate clones to form a probe of the desired specificity.

Another method of isolating DNA from a chromosomal region or regions (or also a whole chromosome) is to propagate such a chromosomal region or regions in an appropriate cell line (for example, a hybrid cell line such as a human/hamster hybrid cell), extract the DNA from the cell line and clone it in an appropriate vector and select clones containing human DNA to form a library. When a hybrid cell is used, the chromosomes in the hybrid cell containing the human chromosomal material may be separated by flow sorting (FACS) prior to cloning to increase the frequency of human clones in the library. Still further, total DNA from the hybrid cell can be isolated and labeled without further cloning and used as a probe, as exemplified in Figure 6.

3. Single-Stranded Probes. In some cases, it is preferable that the nucleic add fragments of the heterogeneous mixture consist of singlestranded RNA or DNA. Under some conditions, the binding efficiency of single-stranded nucleic add probes has been found to be higher during in situ hybridization, e.g., Cox et al., Detection of mRNAs in Sea Urchin Embryos by

In Situ Hybridization Using Asymmetric RNA Probes, Developmental

Biology, Vol. 101, pp. 4851984) 502־).

Standard methods are used to generate RNA fragments from isolated DNA fragments. For example, a method developed by Green et al., described in Cell, Vol. 32, pp. 681-694 (1983), is commercialy available from Promega Biotec (Madison, WI) under the tradename Riboprobe. Other transcription kits suitable for use with the present invention are available from United States Biochemical Corporation (Cleveland, OH) under the tradename Genescribe. Single-stranded DNA probes can be produced with the singlestranded bacteriophage M13, also available in kit form, e.g. Bethesda Research Labora- tories (Gaithersburg, MD). The hybridizations illustrated in Figure 4 were performed with the libraries of Table 1 subcloned into the Bluescribe plasmid vector (Stratagene, La Jolla, CA). The Bluesaibe plasmid contains RNA promoters which permit production of single-stranded probes.

Section IX, infra provides methods for preparing and applying nonself-complementary single-stranded nucleic acid probes that improve signalto-noise ratios attainable in in situ hybridization by reducing non-specific and mismatched binding of the probe. That section further provides for methods of denaturing double-stranded target nucleic acid which minimizes singlestranded regions available for hybridization that are non-complementary to probe sequences. Briefly, probe is constructed by treating DNA with a restriction enzyme and an exonuclease to form template/primers for a DNA polymerase. The digested strand is resynthesized in the presence of labeled nucleoside triphosphate precursor, and the labeled single-stranded fragments are separated from the resynthesized fragments to form the probe. The target ־59nucleic acid is treated with the same restriction enzyme used to construct the probe, and is treated with an exonuclease before application of the probe.

I.B. PCR

Another method of producing probes of this invention includes the use of the polymerase chain reaction [PCR]. [For an explanation of the mechanics of PCR, see Saiki et al., Science, 230:1350 (1985) and U.S. Patent Nos. 4,683,195,4,683,202 (both issued July 28,1987) and 4,800,159 (issued January 24, 1989).] Target-specific nucleic acid sequences, isolated as indicated above, can be amplified by PCR to produce target-specific sequences which are reduced in or free of repetitive sequences. The PCR primers used for such a procedure are for the ends of the repetitive sequences, resulting in amplification of sequences flanked by the repeats.

Figure 7 illustrates such a method of using PCR wherein the representative repetitive sequence is Alu. If only short segments are amplified, it is probable that such sequences are free of other repeats, thus providing DNA reduced in repetitive sequences.

One can further suppress production of repetitive sequences in such a PCR procedure by first hybridizing complementary sequences to said repetitive sequence wherein said complementary sequences have extended non-complementary flanking ends or are terminated in nucleotides which do not permit extension by the polymerase. The non-complementary ends of the blocking sequences prevent the blocking sequences from acting as a PCR primer during the PCR process.

Π. Removal of Repetitive Sequences and/or Disabling the Hybridization Capacity of Repetitive Sequences

Typically a probe of the current invention is produced in a number of steps including: obtaining source nucleic acid sequences that are complementary to the target region of the genome, labeling and otherwise processing them so that they will hybridize efficiently to the target and can be detected after they bind, and treating them to either disable the hybridization capacity or remove a sufficient proportion of shared repetitive sequences, or both disable and remove such sequences. The order of these steps depends on the specific procedures employed.

The following methods can be used to remove shared repetitive sequences and/or disable the hybridization capacity of such shared repetitive sequences. Such methods are representative and are expressed schematically in terms of procedures well known to those of ordinary skill the art, and which can be modified and extended according to parameters and procedures well known to those in the art.

1. Single-copy probes. A single-copy probe consists of nucleic add fragments that are complementary to single-copy sequences contained in the target region of the genome. One method of constructing such a probe is to start with a DNA library produced by cloning the target region. Some of the clones in the library will contain DNA whose entire sequence is single-copy; others will contain repetitive sequences; and still others will have portions of single-copy and repetitive sequences. Selection, on a done by done basis, and pooling of those dones containing only single-copy sequences will result in a probe that will hybridize spedfically to the target region. The single-copy nature of a clone can ultimately be established by Southern hybridization using standard techniques. Figure 4H shows hybridization with 120 clones selected in this way from a chromosome 4 library.

Southern analysis is very time consuming and labor intensive. Therefore, less perfect but more efficient screening methods for obtaining candidate single-copy clones are useful. In Section V.B, examples of improved methods are provided for screening individual phage and plasmid clones for the presence of repetitive DNA using hybridization with genomic DNA. The screening of plasmid clones is more efficient, and approximately 80% of selected clones contain only single-copy sequences; the remainder contain low-copy repeats. However, probes produced in this way can produce adequate staining contrast, indicating that the low-copy repetitive sequences can be tolerated in the probe (see subsection 3 of this section).

A disadvantage of clone by clone procedures is that a clone is discarded even if only a portion of the sequence it contains is repetitive. The larger the length of the cloned nucleic acid, the greater the chance that it will contain a repetitive sequence. Therefore, when nucleic acid is propagated in a vector that contains large inserts such as a cosmid, YAC, or in a cell line, such as hybrid cells, it may be advantageous to subclone it in smaller pieces before the single-copy selection is performed. The selection procedures just outlined above do not discriminate between shared and specific repetitive sequences; clones with detectable repetitive sequences of either type are not used in the probe.

2. Individual testing of hybridization properties. The hybridization specificity of a piece of nucleic acid, for example, a clone, can be tested by in situ hybridization. If under appropriate hybridization conditions it binds to single-copy or repetitive sequences specific for the desired target region, it can be included in the probe. Many sequences with specific hybridization characteristics are already known, such as chromosome-specific repetitive sequences [Trask et al., supra, (1988) and references therein], VNTRs, numerous mapped single copy sequences. More are continuously being mapped. Such sequences can be included in a probe of this invention.

3. Bulk Procedures. In many genomes, such as the human genome, a major portion of shared repetitive DNA is contained in a few families of highly repeated sequences such as Alu. A probe that is substantially free of such high-copy repetitive sequences will produce useful staining contrast in many applications. Such a probe can be produced from some source of nucleic acid sequences, for example, the libraries of Table I, with relatively simple bulk procedures. Therefore, such bulk procedures are the preferred methods for such applications.

These methods primarily exploit the fact that the hybridization rate of complementary nucleic acid strands increases as their concentration increases. Thus, if a heterogeneous mixture of nucleic acid fragments is denatured and incubated under conditions that permit hybridization, the sequences present at high concentration will become double-stranded more rapidly than the others. The double-stranded nucleic acid can then be removed and the remainder used as a probe. Alternatively, the partially hybridized mixture can be used as the probe, the double-stranded sequences being unable to bind to the target. The following are methods representative of bulk procedures that are useful for producing the target-specific staining of this invention.

3a. Self-reassociation of the probe. Double-stranded probe nucleic acid in the hybridization mixture is denatured and then incubated under hybridization conditions for a time sufficient for the high-copy sequences in the probe to become substantially double-stranded. The hybridization mixture is then applied to the sample. The remaining labeled single-stranded copies of the highly repeated sequences bind throughout the sample producing a weak, widely distributed signal. The binding of the multiplicity of low-copy sequences specific for the target region of the genome produce an easily distinguishable specific signal.

Such a method is exemplified in Section VLB (infra) with chromosome-specific libraries for chromosomes 4 and 21 (pBS4 and pBS21) as probes for those chromosomes. [Pinkel et al., PNAs (USA), 85:9138-9142 (December 1988)]. The hybridization mix, containing a probe concentration in the range of 1-10 ng/ul was heated to denature the probe and incubated at 37°C for 24 hours prior to application to the sample.

3b. Use of blocking nucleic acid. Unlabeled nucleic acid sequences which are complementary to those sequences in the probe whose hybridization capacity it is desired to inhibit are added to the hybridization mixture. The probe and blocking nucleic acid are denatured, if necessary, and incubated under appropriate hybridization conditions. The sequences to be blocked become double-stranded more rapidly than the others, and therefore are unable to bind to the target when the hybridization mixture is applied to the target. In some cases, the blocking reaction occurs so quickly that the incubation period can be very short, and adequate results can be obtained if the hybridization mix is applied to the target immediately after denaturation. A blocking method is generally described by Sealy et al., Removal of Repeat Sequences form Hybridization Probes, Nucleic Acid Research, 13:1905 (1985), which reference is incorporated by reference. Examples of blocking nucleic acids include genomic DNA, a high-copy fraction of genomic DNA and particular sequences as outlined below (i-iii).

3b.i. Genomic DNA. Genomic DNA contains all of the nucleic acid sequences of the organism in proportion to their copy-number in the genome. Thus, adding genomic DNA to the hybridization mixture increases the concentration of the high-copy repeat sequences more than low-copy sequences, and therefore is more effective at blocking the former. However, the genomic DNA does contain copies of the sequences that are specific to the target and so will also reduce the desired chromosome-specific binding if too much is added. Guidelines to determine how much genomic DNA to add (see 3.e. Concept of 0, infra) and examples of using genomic blocking DNA are provided below. The blocking effectiveness of genomic DNA can be enhanced under some conditions by adjusting the timing of its addition to the hybridization mix; examples of such timing adjustments are provided with Protocol I and Protocol Π hybridizations illustrated in Figures 4B through E (Protocol I) and Figure 4F (Protocol Π) and detailed in Section VI, infra.

3b.ii. High-copy fraction of genomic DNA. The difficulty with use of genomic DNA is that it also blocks the hybridization of the low-copy sequences, which are predominantly the sequences that give the desired target staining. Thus, fractionating the genomic DNA to obtain only the high-copy sequences and using them for blocking overcomes this difficulty. Such fractionation can be done, for example, with hydroxyapatite as desaibed below (3c.i).

3b.iii. Specified sequences. The blocking of a particular sequence in the probe can be accomplished by adding many unlabeled copies of that sequence. For example, Alu sequences in the probe can be blocked by adding cloned Alu DNA. Blocking DNA made from a mixture of a few clones containing the highest copy sequences in the human genome can be used effectively with chromosome-specific libraries for example, those of Table I. Alternatively, unlabeled nucleic acid sequences from one or more chromosome-specific libraries could be used to block a probe containing labeled sequences from one or more other chromosome-specific libraries. The shared sequences would be blocked whereas sequences occurring only on the target chromosome would be unaffected. Figure 4F shows that genomic DNA was not effective in completely blocking the hybridization of a sequence or sequences shared by human chromosome 21 and the centromeric regions of the other human acrocentric chromosomes. When a clone or clones containing such a sequence or sequences is or are eventually isolated, unlabeled DNA produced therefrom could be added to the genomic blocking DNA to improve the specificity of the staining.

3c. Removal of Sequences.

3c.i. Hydroxyapatite. Single- and double-stranded nucleic acids have different binding characteristics to hydroxyapatite. Such characteristics provide a basis commonly used for fractionating nucleic acids. Hydroxyapatite is commerically available (eg. Bio-Rad Laboratories, Richmond, CA). The fraction of genomic DNA containing sequences with a particular degree of repetition, from the highest copy-number to single-copy, can be obtained by denaturing genomic DNA, allowing it to reassociate under appropriate conditions to a particular value of C_ot, followed by separation using hydroxyapatite. The single- and double-stranded nucleic add can also be discriminated by use of SI nuclease. Such techniques and the concept of ־66C_ot are explained in Britten et al., Analysis of Repeating DNA Sequences by Reassociation, in Methods in Enzymology, Vol. 29, pp. 363-418 (1974), which article is herein incorporated by reference.

The single-stranded nucleic acid fraction produced in 3a. or 3b. above can be separated by hydroxyapatite and used as a probe. Thus, the sequences that have been blocked (that become double-stranded) are physically removed. The probe can then be stored until needed. The probe can then be used without additional blocking nucleic acid, or its staining contrast can perhaps be improved by additonal blocking.

3c.ii. Reaction with immobilized nucleic acid. Removal of particular sequences can also be accomplished by attaching single-stranded absorbing nucleic add sequences to a solid support. Single-stranded source nudeic add is hybridized to the immobilized nucleic add. After the hybridization, the unbound sequences are collected and used as the probe. For example, human genomic DNA can be used to absorb repetitive sequences from human probes. One such method is described by Brison et al., General Method for Cloning Amplified DNA by Differential Screening with Genomic Probes, Molecular and Cellular Biology, Vol. 2, pp. 578-587 (1982). Accordingly, that reference is incorporated by reference. Briefly, minimally sheared human genomic DNA is bound to diazonium cellulose or a like support. The source DNA, appropriately cut into fragments, is hybridized against the immobilized DNA to C_ot values in the range of about 1 to 100. The preferred stringency of the hybridization conditions may vary depending on the base composition of the DNA. Such a procedure could remove repetitive sequences from chromosome-specific libraries, for example, those of Table I, to produce a probe capable of staining a whole human chromosome.

d. Blocking non-targeted sequences in the targeted genome. Blocking of non-targeted binding sites in the targeted genome by hybridization with unlabeled complementary sequences will prevent binding of labeled sequences in the probe that have the potential to bind to those sites. For example, hybridization with unlabeled genomic DNA will render the high-copy repetitive sequences in the target genome double-stranded. Labeled copies of such sequences in the probe will not be able to bind when the probe is subsequently applied.

In practice, several mechanisms combine to produce the staining contrast. For example, when blocking DNA is added to the probe as in 3b above, that which remains single-stranded when the probe is applied to the target can bind to and block the target sequences. If the incubation of the probe with the blocking DNA is minimal, then the genomic DNA simultaneously blocks the probe and competes with the probe for binding sites in the target.

3e. Concept of Q. As mentioned in section 3b.i above, it is necessary to add the correct amount of genomic DNA to achieve the best compromise between inhibiting the hybridization capacity of high-copy repeats in the probe and reducing the desired signal intensity by inhibition of the binding of the target-specific sequences. The following discussion pertains to use of genomic blocking DNA with probes produced by cloning or otherwise replicating stretches of DNA from the target region of the genome. Thus, the probe contains a representative sampling of the single-copy, chromosomespecific repetitive sequences, and shared repetitive sequences found in the target. Such a probe might range in complexity from 100 kb of sequence derived from a small region of the genome, for example several closely spaced cosmid clones; to many millions of bases, for example a combination of multiple libraries from Table I. The discussion below is illustrative and can be extended to other situations where different blocking nucleic adds are used. The following discussion of Q is designed only to give general guidelines as to how to proceed.

The addition of unlabeled genomic DNA to a hybridization mix containing labeled probe sequences increases the concentration of all of the sequences, but increases the concentration of the shared sequences by a larger factor than the concentration of the target-spedfic sequences because the shared sequences are found elsewhere in the genome, whereas the targetspecific sequences are not. Thus, the reassodation of the shared sequences is preferentially enhanced so that the hybridization of the labeled copies of the shared sequences to the target is preferentially inhibited.

To quantity this concept, first consider one of the sequences, repeat or single-copy, that hybridize spedfically to the ith chromosome in a hybridization mixture containing a mass mp of probe DNA from the ith chromosome library of Table 1 (for example) and m^ of unlabeled genomic DNA. The number of labeled copies of the sequence is proportional to mp. However, the number of unlabeled copies is proportional to ijm^, where fj is the fraction of genomic DNA contained on the ith chromosome. Thus, the ratio of unlabeled to labeled copies of each of the sequences specific for the target chromosome, is fjm^/mp, which is defined herein as Q. For normal human chromosomes, 0.016&lt;fj&lt;0.08 [Mendelsohn et al. Science, 179:1126 (1973)]. For representative examples described in Section VLB (infra), f4 = 0.066 and f2! = 0.016. For a probe targeted at a region comprised of L base pairs, fj=L/G where G is the number of base pairs in a genome (approximately x 10^ bases for humans and other mammals). Thus, Q = (L/G) (m^/mp).

Now consider a shared sequence that is distributed more-or-less uniformly over the genome, for example, Alu. The number of labeled copies is proportional to mp, whereas the number of unlabeled copies is proportional to m^. Thus, the ratio of unlabeled to labeled copies is m^/mp = Q/fj. This is true for all uniformly distributed sequences, regardless of copy number. Thus adding genomic DNA increases the concentration of each specific sequence by the factor 1 + Q, whereas each uniformly distributed sequence is increased by the larger factor 1 + Q/fj. Thus, the reassociation rates of the shared sequences are increased by a larger factor than those of the specific sequences by the addition of genomic DNA.

It can be shown that roughly half of the beneficial effect of genomic DNA on relative reassociation rates is achieved when Q = 1, and, by Q = 5, there is essentially no more benefit to be gained by further increases. Thus, the protocol I hybridizations of Section VLB infra keep Q&lt;5.

To illustrate the use of genomic blocking DNA, it is convenient to consider a model of a genome wherein 50% of the DNA is comprised of specific sequences (both repetitive and single-copy) and the other 50% of the DNA is comprised of shared repetitive sequences that are distributed uniformly over the genome. ׳Thus, according to the model, if the target is L bases (that is, the probe contains fragments representing L bases of the target area or areas of the genome), sequences containing L/2 bases will be specific to the target, and L/2 will be shared with the entire genome.

ΙΠ. Labeling the Nucleic Acid Fragments of the Heterogeneous Mixture.

Several techniques are available for labeling single- and doublestranded nucleic add fragments of the heterogeneous mixture. They indude incorporation of radioactive labels, e.g. Harper et al. Chromosoma, Vol 83, pp. 431-439 (1984); direct attachment of fluorochromes or enzymes, e.g. Smith et al., Nudeic Adds Research, Vol. 13, pp. 2399-2412 (1985), and Connolly et al., Nudeic Adds Research, Vol. 13, pp. 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. Sei., Vol 81, pp. 3466-3470 (1984); Richardson et al., Biotin and Fluorescent Labeling of RNA Using T4 RNA Ligase, Nucleic Acids Research, Vol. 11, pp. 6167-6184 (1983);

Langer et al., Enzymatic Synthesis of Biotin-Labeled Polynucleotides: Novel Nucleic Add Affinity Probes, Proc. Natl. Acad. Sci., Vol. 78, pp. 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, pp. 32-50 (1983); Broker et al., Electron Microscopic Visualization of tRNA Genes with Ferritin-Avidin: Biotin Labels, Nucleic Adds Research, Vol. 5, pp. 3631978) 384־); Bayer et al., The Use of the Avidin Biotin Complex as a Tool in Molecular Biology, Methods of Biochemical Analysis, Vol. 26, pp. 11980) 45־) 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-2acetylaminofluorene, 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 add 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 circum- stances. Accordingly, the above-cited references are in- corporated 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, pp. 470-480 (1981); Henderson, Cytological Hybridization to Mammalian Chromo- somes, International Review of Cytology, Vol. 76, pp. 11982) 46־); and Angerer, et al., In Situ Hybridization to Cellular RNAs, in Genetic Engineering: Principles and Methods, Setlow and Hollaender, Eds., Vol. 7, pp. 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 to 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 VLB 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 2000 bases, more preferably in the vicinity of 1 kb. When the size of the probe fragments is in about the 800 to about 1000 base range, 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 add alcohol solutions, acid acetone solutions, Petrunkewitsch's reagent, and various aldehydes such as formaldehyde, paraformaldehyde, glutaraldehyde, or the like. Preferably, ethanol-acetic add 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, pp. 1401-1402 (1985), is useful. Accordingly, Trask et al., is incorporated by reference. Briefly, K2CO3 and dimethylsuberimidate (DMS) are added (from a 5x concentrated stock solution, mixed immediately before use) to a suspension containing about 5 x 10^ 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 add per milliliter of suspension. Nuclei are washed once by centrifugation (300g, 10 minutes, 4°C in 50 mM kCl, 5 mM Hepes buffer, at pH

9.0, and 10 mM MgS04).

A preferred fixation procedure for cells or nuclei in suspension is disclosed by Trask et al., Hum. Genet., 78:2511988) 259־), 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 MgS04, 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 adds. Pronase, pepsin or proteinase K are frequently used enzymes. A representative acid treatment is 0.02-0.2 N HC1, 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 2X 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 ΠΙ [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 2X 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 consitutents or washes in appropriate solutions. The probe and target nucleic add 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 add per unit volume, that can be maintained in solution and hybridize effectively.

In the representational examples desCTibed in Section VLB (infra), the total DNA concentration in the hybridization mixture had an upper limit on

־78־ 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 targetspecific 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 nonspecific 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, pp. 6411966) 645־), with the heterogeneous mixture, incubating for several minutes, e.g., 520־, 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 2X 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.l. 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, pp. 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., Cytometery, Vol. 5, pp. 108-123 (1984), and stained with the fluorescent dyes--Hoechst 33258 and Chromomydn A3. Chromsome 21 was purified on the Lawrence Livermore National Laboratory high speed sorter, described by Peters et al., Cytometry, Vol. 6, pp. 2901985) 301־).

After sorting, chromosome concentrations were approximately 4 x 10^ /ml. Therefore, prior to DNA extraction, the chromosomes (0.2 1.0 ־ x 106) were concentrated by centrifugation at 40,000 x 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 HCI 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 phenolxhloroform: isoamyl alcohol (25:24:1) 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 Hindin (Bethesda Research Laboratories, Inc.) in 50 mM NaCI, 10 mM Tris HCI pH

7.5,10 mM MgC121 _׳ 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 minicollodion bag and then 2 micrograms of Charon 21A arms cleaved with HindHI and treated with calf alkaline phosphatase (Boehringer Mannheim) were added. This solution was concentrated under vacuum to a volume of 50100־ 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 10X ligase buffer (0.5M Tris HCI pH 7.4,0.1 M MgC12,0.1M 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 1620־ 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, pp. 299309־ (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 15־ x 10^ plaque forming units (pfu) per microgram. The resultant phage were amplified on E. coli LE392 at a density of approximately 10^ 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 HC1 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, pp. 1801977) 182־). 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 clone 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 ΠΙ. 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

־82־ separated from the phage arms by 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. Sei., Vol. 78, pp. 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, 2X SSC, 10 percent dextran sulfate, 500 micrograms/ml carrier DNA (sonicated herring sperm DNA), and 2.0 microgram/ml biotin-labeled chromsome 21specific DNA. This mixture is applied to the slides at a density of 3 microliters/cm^ under a glass coverslip and sealed with rubber cement. After overnight incubation at 37°C, the slides are washed at 45°C (50% formamide2XSSC pH 7,3 times 3 minutes; followed by 2XSSC 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 fluoresceinavidin 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 fluoresceinavidin DCS. Fluorescein-avidin DCS, goat antiavidin and goat serum are all available commercially, e.g., Vector Laboratories (Burlingame, CA). After washing in BN, a fluorescence antifade solution, p-phenylenediamine (1.5 microliter/cm^ 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 miCT0gram/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

Single-Copy Sequences

Pinkel et al., PNAS (USA), 85:9138-9142 (December 1988) describe the procedures for preparing chromosome 4 single-copy sequences and then a protocol [modification of the procedure described in Pinkel et al; PNAS (USA), 83:2934-2938 (1986)] for hybridizing said single copy probes to a human metaphase spread. Figure 4H shows the hybridization with a pool of 120 single-copy probes from chromosome 4 to a human metaphase spread.

VI. Incapacitating Shared Repetitive Sequences

VI.A. Chromosome 21-Specific Staining Using Blocking DNA

High concentrations 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 CaC12 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 2XSSC. A hybridization mix was prepared which consisted of 50% formamide, 10% dextran sulfate, 1% Tween 20, 2XSSC, 0.5 mg/ml human genomic DNA, 0.03 mg/ml lambda DNA, and 3 miCT0gram/ml biotin labeled probe DNA. The probe DNA consisted of the highest density fraction of phage from the chromosome 21 Hind ΠΙ 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 2XSSC 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). Figures 1A-C illustrate the results. Figure 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 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.) Figure IB is a fluorescein image of the same spread as in Figure IA, again in binary form. (Again, small arrows indicate the locations of the chromosome 21s.) Figure IC illustrates the positions of the chromosome 21s after other less densely stained objects were removed by standard image processing techniques.

VLB. 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.

Pinkel et al., PNAS (USA), 85:9138-9142 (December 1988) describes the use of unlabeled genomic DNA to competitively inhibit hybridization of those sequences in a chromosome-specific library that are shared with other chromosomes. That paper describes materials and methods for fluorescence in situ hybridization with human chromosome-specific libraries [chromosome 4 library LL04NS02 subcloned into Bluescribe plasmids (pBS-4); chromosome 21 library LL21NS02 subcloned into Bluescribe plasmids (pBS-21)]. The results of those hybridizations are shown in Figure 4 A-C and 4F and G.

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 HYA9.E6 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

־89־

DNA was isolated. The isolated DNA was biotinylated by nick translation so that 1030%־ of the thymidine was replaced by biotin-ll-dUTP. The concentration of the total labeled DNA after nick translations is in the range of 1020־ ng/ul.

Blocking DNA. Human placental DNA (Sigma) was treated with proteinase K and extracted with phenol and sonicated to a size range of 200600־ 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 110־ ug/ul.

Fluorescence in situ hybridization (FISH). Hybridization followed the procedures of Pinkel et al. (1988), supra with slight modifications. Metaphase spreads were prepared from methotrexate synchronized cultures according to the procedures of Harper et al. PNAS (USA) 78: 44581981) ,4460־). 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/2xSSC 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 MTSJa 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/2xSSC/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 0 slip. After overnight incubation at 37 C the covershp was removed and the slide washed 3 times 3 min each in 50% formamide/2xSSC 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 goatanti-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, CA).

Figure 5 shows the hybridization of HYA3.A2 (580 kb of human DNA) to 12q21.1. The location of the hybridization was estabilished by using a conventional fluorescent banding technique employing the DAPI/ actinomycin D procedure: Schweizer, Reverse fluorescent chromosome banding with chromomydn 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.

91־

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

<td colspan="3"> Table 2 YAC Competition Hybridization</td>

<td> YAC Clone</td><td> Insert Size</td><td> Localization</td>

<td> HY1</td><td> 120</td><td> Xq23</td>

<td> HY19</td><td> 450</td><td> 8q23.3</td>

<td></td><td></td><td> 21q21.1</td>

<td> HY29</td><td> 500</td><td> 14ql2</td>

<td> HYA1.A2</td><td> 250</td><td> 6ql6</td>

<td> HYA3.A2</td><td> 580</td><td> 12q21.1</td>

<td> HYA3.A9</td><td> 600</td><td> 14q21</td>

<td> HYA9.E6</td><td> 280</td><td> lp36.2</td>

<td></td><td></td><td> 3q22</td>

־92־

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 VLB., 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. Figure 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 addititional 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 imrease 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 m1CT0sc0py. 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 inversions 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.

VILA 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 dastogens. 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 x lO'^/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 disaimination 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 XO (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

־98־ 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-spedfic domain following hybridization with X-spedfic and Y-spedfic probes indicates the occurrence of Turner syndrome. An increase in the frequency showing two X-spedfic domains and one Y-specific domain indicates Kleinfelter syndrome, and increase in tire 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 embryo-invasive 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 nonembryo-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 1(H to 106־ cells/ml and quite variable depending on the time of gestation; however, appropriately marked fetal cells may be

־100־ 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 were 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 desaibed in the following references.

Herzenberg et al., PNAS, 76:1453 (1979), reports the isolation of fetal cells, apparently of lymphoid 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 antiTrop-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 US Patent Application, USSN 389,224, filed August 3, 1989, entitled Method for Isolating Fetal Cytotrophoblast Cells. [See also: in Fisher et al., 1. Cell. Biol., 109 (2): 891-902 (1989)]. The monoclonal antibodies disclosed therein react

־102 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 April 4, 1989 and reported viable thereby on April 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 13ql4 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. Translocations 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 (Figure 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, Illinois (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 cluster 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).

Figure 8 is a schematic representation of the BCR gene on chromosome 22, the ABL gene of chromosome 9, and the BCR-ABL fusion

־106־ 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 15־ 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 la and lb refer to the alternative first exons, and Π to the second exon. Exon Π is approximately 25kb upstream of the end of the 28 kb cosmid c-hu־abl (the ABL probe). All CML breakpoints occur upstream of exon Π, usually between exons lb and la, 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 225kb.

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 RPMI1640 supplemented with 10% fetal calf serum, an antibiotic mixture (gentamycin 500 mg/ml), and 1% L-glutamine for 24h. 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 digoxigenin11-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/2xSSC 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 (2xSSC is 0.3M NaCl/30 mM sodium citrate). Heat 10 ml of hybridization mixture containing 2 ng/ml of each probe, 50% formamide/2xSSC, 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 x 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/2xSSC pH 7 at 42°C, twice for 20 minutes each in 2xSSC at 42°C, and finally rinse at room temperature in 4xSSC.

־108־

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 4xSSC/l% bovine serum albumin (BSA) for 5 min. Apply Texas Red-avidin (Vector Laboratories Inc., 2 mg/ml in 4xSSC/l% BSA) for 45 min. Wash in 4xSSC once, 4xSSC/l% Triton-X 100 (Sigma) and then again in 4xSSC, 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, IN; 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 (Figure 9 A) near the telomere on 9q while the BCR probe bound at 22qll (Figure 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 Ph^, and together in the interphase nucleus, as illustrated in Figure 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 clone. 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 dose proximity, see Figure 10). The randomly located red and green signals are asaibed to hybridization to the ABL and BCR genes on the normal chromosomes, and the red-green doublet signal to hybridization to the BCRABL fusion gene. Interphase mapping studies suggest that DNA sequences separated by less than 250kb 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 Ph¹negative 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.

Summary of cytogenetic, fluorescence in situ hybridization and other analyses of BCR-ABL rearrangements in 6 CML cases

Table3

<td> Sample</td><td> Cytogenetics</td><td colspan="2"> Fluorescence in situ hybridization Metaphase Inlttptat Mldti</td><td> Other infonnation</td>

<td> CML-l^a</td><td> 46XX,t(9;22)(q34;qll)</td><td> Hybridization to telomere of small acrocentric</td><td> 80% showed red-green fusion 2% showed red-green doublets 18% not interpretable</td><td></td>

<td> CML-?¹</td><td> 46XY,t«i2)(q34x|ll)</td><td> Not available</td><td> 60% showed red-green fusion Hybridization efficiency was low</td><td></td>

<td> CML-3^</td><td> 46XY,t(9;22)(q34;qll)</td><td> Not available</td><td> 75% showed red-green fusion 25% appeared normal</td><td> BCR-ABL fusion positive by PCR</td>

<td> CML-44*</td><td> WWiqll)</td><td> Not available</td><td> 1ש% showed red-green fusion</td><td> BCR-ABL fusion positive by PCR</td>

<td> CML-5‘</td><td> 47XY,+8,del(22)(qll)</td><td> 47 chromosomes. Red. green fusion at telomere of small acrocentric</td><td> 100% showed red-green fusion</td><td> BCR-ABL fusion positive by PCR</td>

<td> CML#</td><td> 46XY ins(22;9)qll;q34;q?)</td><td> Red-green fusion interstitial on small acrocentric</td><td> 100% showed red-green fusion</td><td> BCR-ABL fusion positive by PCR</td>

<td> CML-7¹¹</td><td> 46XYt(5;9)q(?;q?)</td><td> Not available</td><td> 100% showed red-green fusion</td><td> BCR-ABL fusion negative in two tests by PCR</td>

BCR rearrangment detected by Southern blot analysis

Clinical data:^a CML,chronic phase receiving no treatment; CML, chronic phase receiving hydroxyurea;^c CML,blast crisis receiving no treatment; CML,blast crisis receiving hydroxyurea; and^e CML-3 and CML-4 represent respectively bone marrow and blood samples from one patient.

Hybridization to metaphase spreads was performed in three cases (CML-1,-5 and -6). All of these showed red and green hybridization signals in close proximity on a single acrocentric chromosome. In two cases, scored as t(9:22)(q34;qll) 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 (Figure 9C). One case (CML-6) was suspected by classical cytogenetics to have an insertion of chromosomal material at 22qll. Dual color hybridization to metaphase spreads from this case showed the red-green pair to be centrally located in a small chromosome (Figure 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 Phi 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 Figures 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 (Figure 9E).

־114־

Quantitative measurement of fluorescence intensity may assist with such an analysis.

IX. Method of Preparing and Applying Single Stranded Nucleic

Add Probes to Double Standed Target DNAs

Generally the method for preparing and applying single stranded DNA hybridization probes to double stranded target DNA involves treating both target DNA and probe DNA with the same restriction endonuclease followed by digestion of single strands adjacent to the restriction cuts. Probes are constructed by resynthesizing the digested single strands with labeled nucleotides. The labeled strands are substantially complementary to the undigested single strands of the target DNAs. The double stranded DNA fragments containing the labeled single strands are broken into smaller pieces and denatured. The hybridization probes are obtained by separating the labeled single stranded fragments from the unlabeled fragments.

DNA to be used in the probes is treated with a restriction endonuclease to form restriction fragments having sticky ends. That is, it is important that the restriction endonuclease make a staggered cut through the double stranded DNA. Suitable restriction endonuclease include, but are not limited to, Hind ΙΠ, Bam Hl, Eco Rl, or the like, all of which are commercially available, e.g., Promega Biotec (Madison, WI), or Boehringer Mannheim (Indianapolis, IN). In selecting a restriction endonuclease it is preferable that the resulting restriction fragments be within a size range which allows them to be directly inserted into available cloning vectors. Suitable cloning vectors included plasmids, such as pBR322, and phages, such as lambda phage, various derivatives of both of these being commercially available, e.g., Promega Biotec (Madison, WI), and Boehringer Mannheim (Indianapolis, IN). Amplified copies of the restriction fragments are isolated using standard techniques, e.g., Maniatis et al., Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory, 1982).

Alternatively, for some applications restriction fragments can be obtained from existing libraries. For example, the American Type Culture Collection, Rockville, MD, holds collections of human chromosome-specific libraries of restriction fragments which are available to the public.

Standard procedures are followed in treating the restriction fragments with exonuclease, and in enzymatically re-synthesizing the digested strands in the presence of labeled precursors. In particular the technique disclosed by James and Leffak, Anal. Biochem., Vol. 141, pp. 331984) 37־), is followed. Accordingly this reference is incorporated by reference. Briefly, to the restriction fragments about 3 units of exonuclease III are added per miCTOgram of DNA in a solution consisting of 100 mM NaCl, 50 mM Tris-HCl (pH 8.0), 10 mM MgC12 and 1 mM dithiothreitol, at 37°C. Digestion is terminated by heating the sample to 60°C for 5-10 minutes. James and Laffak report that these conditions result in the digestion of about 80-200 nucleotides per minute. The actual digestion rate will vary depending on the source and batch of exonuclease ΙΠ as well as the source of the DNA substrate, e.g., Guo et al., Nucl. Acids Res., Vol. 10 pg. 2065. Some experimentation may be necessary to obtain labeled stands of the desired length. Exonuclease ΙΠ is available commercially, e.g., Boehringer Mannmheim (Indianapolis, IN), or Promega Biotec (Madison, WI). Also, T4 polymerase (BRL, Bethesda, MD) can be used for both the exonuclease and resynthesis steps.

־116The exonuclease treated restriction fragments serve as primer/templates for a DNA polymerase which re-synthesizes the digested strands in the presence of labeled precursors. The preferred labeled precursor is biotinylated uracil, as a substitute for thymidine. Re-synthesis is accomplished using DNA polymerase I or T4 DNA polymerase following the procedure of Langer et al., Proc. Nat’l. Acad. Sci., Vol. 78, pp. 66331981) 6637־) which, in turn, is an adaption of the basic nick translation technique disclosed by Rigby et al., T. Mol. Biol., Vol. 113, pg. 237 (1977), e.g., 1 unit E. Coli polymerase I per microgram of DNA is incubated at 37°C in a solution consisting of 100 mM Nacl, 50 mM Tris-HCl (pH 8.0), 10 mM MgC12,1 mM dithiothreitol, and 50 mM KC1. Also included in the solution are appropriate amounts of the triphosphate precursors (one or more of which are labeled),

e.g., 50100־ micromolar of each for 20-50 micrograms per milliliter of restriction fragments. Under these conditions resynthesis is completed in about 40-60 minutes.

The labeled restriction fragments are broken into smaller fragments to ensure that the labeled regions on either end of the labeled restriction fragments are separated. (Otherwise, the labeled fragment on one end would be a part of a larger piece of single stranded DNA which contained complementary regions to the labeled fragment on the other end). Such breaking into smaller fragments is accomplished by any number of standard techniques, e.g. sonication, enzymatic treatment, or the like, Maniatis, et al. Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory, 1982).

After the labeled restriction fragments are appropriately broken into smaller pieces, they are denatured and single stranded labeled fragments are separated from unlabeled fragments. The separation can be accomplished in several ways. Whenever the preferred label, biotin, is used the preferred separation means is by way of a standard avidin affinity column, e.g. Bayer and Wilchek, The Use of the Avidin-Biotin Complex as a Tool in Molecular Biology, Methods of Biochemical Analysis, Vol 26, pp. 11980) 45־); and Manning et al., Biochemistry, Vol. 16, pp. 13641977) 1370־). Avidin can be covalently coupled to a number of different substrates, such as glass, Sepharose, agarose, and the like, with standard techniques as described in the above references. Accordingly, Manning et al. and Bayer and Wilchek, pp. 9־ 16, are incorporated by reference. Avidin affinity columns are also available commercially, e.g. Zymed Laboratories, Inc. (South San Francisco, CA). The biotinylated probes are removed from the avidin column following the procedure of Chollet and Kawashima, Nucleic Acids Resources, Vol. 5, pp. 15291985) 1541־).

Alternatives to the above labeling procedure are available. For example, after the DNA to be used in the probes is treated with a restriction endonuclease, the resulting restriction fragments are separated into two portions. The first portion undergoes treatment as described above. That is, it is treated with exonuclease to form template/primers for resynthesizing a labeled strand of DNA. The resulting resynthesized restriction fragments are then broken into smaller pieces, as described above. The label in this case, need not be biotin. For example, a radioactive label can be used. The second portion is also treated with an exonuclease, preferably exonuclease ΙΠ. However, the reaction is allowed to proceed to completion so that each restriction fragment is converted into two noncomplementary single stranded pieces approximately half the length of the parent strand. These

־118־ resulting single strands are then covalently linked to DBM paper using standard techniques, e.g. Maniatis et al; Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory, 1982) pp. 335-339; and Alwine et al., Methods in Enzymology, Vol. 68, pp. 220242־ (Academic Press, New York, 1979). Accordingly, the cited pages of these references are incorporated by reference. The fragments of the first portion are denatured and combined with the DBM paper containing the covalently linked fragments of the second portion. Conditions are adjusted to permit hybridization of the labeled strands to complementary strands covalently linked to the DBM paper. The unlabeled strands from the first portion are washed from the paper (there being no complementary strands for them to hybridize to). After the washing the labeled strands are removed by heating, for example, and are ready for use.

Before application of the probe to the target DNA, the target DNA is treated with the same restriction endonuclease as was used in construction of the probe. After restriction endonuclease treatment the target DNA is treated with an exonuclease, preferably exonuclease III or T4 polymerase. Preferably, the conditions of exonuclease treatment are adjusted so that the lengths of single stranded regions created are substantially the same as the lengths of the probe DNA.

Hybridization of probe to target DNA is carried out using standard procedures, e.g. Gall and Pardue, Methods in Enzymology, Vol. 21, pp. 270480־ (1981); Henderson, International Review of Cytology, Vol. 76, pp. 11982) 46־); and Angerer et al., in Genetic Engineering: Principles and Methods, Setlow and Hollaender, Eds., Vol. 7, pp. 4365־ (Plenum Press, New York, 1985). Accordingly, these references are incorporated by reference as guides for the use of the invention in in situ hybridization. Briefly, probe prepared in accordance with the invention is combined with several other agents for reducing nonspecific binding, for maintaining the integrity of the biological structure being probed, and the like. The resulting mixture is referred to herein as the hybridization mix. Below, the method is applied in the chromosome-specific staining of human chromosome 21.

Hind ΠΙ restriction fragments of human chromosome 21 are available from the National Laboratory Gene Library Project through the American Type Culture Collection, Rockville, MD, Van Dilla et al., Human Chromosome-Specific DNA Libraries: Construction and Availability, Biotechnology, Vol. 4, pp. 537-552 (1986). Alternately, such fragments can be produced following the disclosures in Van Dilla et al., cited above, or Fuscoe et al., Contruction of Fifteen Human Chromosome-Specific DNA Libraries from Flow-Purified Chromosomes, Cytogenet Cell Genet., 43:79-86 (1986).

Clones from the library having unique sequence inserts are isolated by the method of Benton and Davis, Science, Vol. 196, pp. 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 clone 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 ΠΙ. Unique sequence containing clones are recognized as those that produce a single band during Southern analysis. Roughly, 100 such clones are selected ־120for the heterogeneous mixture of probe DNA. The unique sequence clones are amplified, the inserts are removed by Hind ΙΠ digestions, and the inserts are separated from the phage arms by gel electrophoresis. The probe DNA fragments (i.e., the unique sequence inserts) are removed from the gel and treated with exonuclease ΠΙ as described above, followed by resynthesis in the presence of biotinylated UTP precursor. The resulting double stranded fragments are sonicated so that on an average each fragment receives about

1.5-2.0 breaks. The resulting pieces are denatured and the biotinylated fragments are isolated by avidin affinity chromatography as described above.

Human lymphocyte chromosomes are prepared following Harper et al., Proc. Nat'l Acad. Sei., Vol. 78, pp. 4458-4460 (1981). Metaphase and interphase cells are 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 treated with RNase (100 micrograms/ml for 1 hour at 37°C), treated for about 116־ hours with Hind ΠΙ at 37°C (10 units M 10 mM Tris-HCl, 50 mM NaCl, 10 mM MgC12, and 14 mM dithioenythritol at pH 7.6), treated with exonuclease III as described above, 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 chromosomes remains on the dry slide. The hybridization mix consists of (final concentrations) 2X SSC (0.15 M NaCl and 0.015 M sodium nitrate) 10 percent dextran sulfate, 500 micrograms/ml carrier DNA (sonicated herring sperm DNA), and 2.0 microgram/ml biotin-labeled ־121probe DNA. This mixture is applied to the slides at a density of 3 microliters/cm^ under a glass coverslip and sealed with rubber cement. After overnight incubation at 37°C, the slides are washed at 45°C (50% formamide2XSS pH 7,3 times 3 minutes; followed by 2XSSC 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 0.92% Na Azide (5 microliter/cm² under plastic coverslips). The coverslips are removed, and excess liquid briefly drained and fluoresceinavidin DCS (3 microgram/ml in BN buffer with 5% milk and 0.02% Na Azide) 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 miCT0gram/ml in BN buffer with 5% goat serum and 0.02% NaAzide) followed, after washing as above, by another layer of fluoresceinavidin DCS. Fluorescein-avidin DCS, goat antiavidin and goat serum are all available commercially, e.g., Vector Laboratories (Burlingame, CA). 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 DNAspecific stain excited in the ultraviolet (Zeiss filter combination 487701), is used as the counterstain when biotin-labeled and total DNA are observed separately. Metaphase chromosomes 21s are detected by randomly located spots of yellow distributed over the body of the chromosome.

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.