AU2274595A - Culture and isolation of fetal cells from maternal peripheral blood - Google Patents

Description

CULTURE AND ISOLATION OF FETAL CELLS FROM MATERNAL PERIPHERAL BLOOD

Background of the Invention

All hematopoietic cells are derived from pluripotent stem cells which, under appropriate conditions, differentiate along one of several possible hematopoietic lineages. One of these lineages, the erythroid lineage, leads to the production of red blood cells. The development of mature, non-nucleated red blood cells, or erythrocytes, results from the commitment of pluripotent stem cells to the erythroid lineage, where they become erythroid progenitor cells, and further differentiation of erythroid progenitor cells into mature cells in response to signals received by the developing cells. An early erythroid progenitor cell is the blast forming unit-erythroid (hereafter BFU-E), which develops into a more mature erythroid progenitor cell, the colony forming unit-erythroid (hereafter CFU-E). When stimulated by a particular growth factor, erythropoietin, CFU-Es grow and differentiate into nucleated red blood cells (hereafter nRBCs), a precurser of mature, non-nucleated red blood cells. In adults, erythroid progenitor cells are located predominantly in the bone marrow, where hematopoiesis occurs in adults, and the erythroid progenitor cells do not circulate in peripheral blood. For a review of erythroid development see Beck, W.S. in Hematology. 4th edition Cambridge, MA: MIT Press (5th edition, 1991) and Stamatoyannopoulos, G., et al. in The Molecular Basis of Blood Diseases. Philadelphia, PA: W.B. Saunders (2nd edition, 1994), pp. 66-105.

The protein erythropoietin is a growth factor necessary for erythroid development in vivo which is produced primarily by the kidneys. Erythropoietin was originally isolated and purified from urine of anemic patients. Miyake, T., et al. J. Biol Chem. 252, 5558-5564 (1977). More recently, erythropoietin has been molecularly cloned and characterized. Jacobs, K., et al., Nature 313, 806-809 (1985); Lin, P., et al., Proc. Natl. Acad. Sci. USA 82, 7580-7584 (1985). Erythropoietin has been used to support the growth of erythroid progenitor cells in culture in vitro. For example, human adult erythroid progenitors from bone marrow and fetal erythroid progenitor cells from umbilical cord blood or fetal liver explants have been grown in vitro by culturing cells in media containing erythropoietin. Sutherland, H.J., et al., Blood 74, 1563-1570 (1989); Bodger, M.P., Exp. Hematol. 15, 869- 876 (1987); Peschle. A.R., et al., Blood 58, 565-572 (1981). During pregnancy, cells of fetal origin appear in the maternal peripheral circulation. For a review see Bianchi, D.W. and Klinger, K.W. in Genetic Disorders and the Fetus: Prevention and Treatment. 3rd edition. Baltimore/London: The John Hopkins University Press (1992). These cells are a potential source of information about the gender and genetic makeup of the developing fetus. Detection, isolation and manipulation of these fetal cells in a sample of maternal peripheral blood offers a non-invasive means for prenatal diagnosis and gender identification. The feasibility of such an approach, however, is hindered by a number of factors. Most importantly, fetal cells are present in maternal blood in very limited numbers, which requires that they be enriched within a mixture of fetal and maternal cells or that the fetal cells be separated in some way from maternal cells. One approach that has been used to achieve enrichment or separation of fetal cells utilizes antibodies specific for a particular fetal cell type. For example, fetal-specific antibodies can be used in order to facilitate separation of fetal cells from maternal components by flow cytometry. Herzenberg, L.A., et al., Proc. Natl. Acad. Sci. USA 16, 1453-1455 (1979); Iverson, G.M., et al., Prenatal Diagnosis 1, 61-73 (1981); Bianchi, D.W., et al., Prenatal Diagnosis 11, 523-528 (1991). However, this approach requires very specialized reagents and can be costly and labor intensive. Another limitation in using fetal cells from maternal peripheral blood for diagnostic purposes is that while cells in metaphase are preferred for analysis not all fetal cells in the maternal peripheral circulation are dividing.

Summary of the Invention

The present invention provides non-invasive methods for culture and isolation of fetal cells and for detecting nucleic acid sequences of interest in isolated fetal cells. The invention is based, at least in part, on the discovery that fetal cell, e.g., fetal progenitor cells, are present in the peripheral blood of a pregnant woman and can be selectively grown in vitro by culturing the cells in the presence of a cell growth factor.

In an embodiment, the invention provides non-invasive methods for culture and isolation of fetal erythroid cells and for detecting nucleic acid sequences of interest in isolated fetal erythroid cells. The invention is based, at least in part, on the discovery that fetal erythroid progenitor cells are present in the peripheral blood of a pregnant woman and can be selectively grown in vitro by culturing the cells in the presence of an erythroid growth factor, e.g., erythropoietin. The methods of the invention offer a non-invasive means by which to obtain fetal cells in numbers great enough to be of use diagnostically while not requiring specialized reagents such as monoclonal antibodies or expensive equipment such as a flow cytometer. Culture of fetal erythroid progenitor cells in the presence of an erythroid growth factor, e.g., erythropoietin, causes the cells to proliferate and differentiate into fetal erythroid cells. Fetal erythroid cells can then be isolated and used in further analyses. The invention provides methods which allow for enrichment of fetal cells relative to maternal cells in a peripheral blood sample, expansion of the total number of fetal cells present compared to the total numbers present in the original sample, and isolation of mitotically active fetal cells. The ability to culture fetal erythroid progenitor cells from maternal peripheral blood is of great value since this approach provides a simple, non-invasive means of obtaining fetal cells for further analysis. The advantages of this approach are that it allows for isolation of greater numbers of fetal cells for analysis than can easily or economically be separated from maternal blood by other methods and, since these cells are growing, also allows for isolation of fetal cells in metaphase for use in diagnostic tests.

Accordingly, this invention pertains to the culture and isolation of fetal erythroid cells from a maternal peripheral blood sample as a means of obtaining fetal cells for analysis and diagnosis. The method of the invention involves obtaining a sample of maternal peripheral blood, culturing cells in the sample in a culture medium containing an erythroid growth factor, e.g., erythropoietin, which causes erythroid progenitor cells to proliferate and differentiate into erythroid cells, and isolating fetal erythroid cells. This method is based at least upon: 1) the presence of fetal erythroid progenitor cells in the maternal peripheral circulation; 2) the scarcity of maternal erythroid progenitor cells in the maternal peripheral circulation; and 3) the higher sensitivity of fetal erythroid progenitors to erythropoietin compared to the sensitivity of maternal erythroid progenitor cells to erythropoietin.

Prior to culturing, the peripheral blood sample can be treated in order to remove certain cell types and/or to enrich for certain cell types so that fewer total cell numbers are utilized in the culturing step. Non-nucleated red blood cells can be removed from the sample by, for example, selective lysis or density gradient centrifugation prior to culturing. Erythroid progenitor cells can be enriched in the sample by, for example, performing dual density gradient centrifugations, wherein gradients of different densities are utilized, and isolating a cell layer which contains erythroid progenitor cells prior to culturing. After culturing, erythroid cells can be isolated by picking one or more discrete colonies of cells from the culture media, for example by using a pipette to manipulate the cells. The cells of the colony can then be transferred to fresh culture medium, placed on a microscope slide for further analysis, used as a source of DNA, or analyzed further in any suitable manner.

Another aspect of the invention relates to a method for identifying fetal erythroid cells after culturing cells of a maternal peripheral blood sample in an erythroid growth factor, e.g., erythropoietin. This method involves detecting a fetal cell marker on erythroid cells as a means of identifying the cell as being of fetal origin. A further aspect of the invention involves detection of a nucleic acid sequence of interest in fetal nucleic acid of fetal erythroid cells after culturing. Detection of Y chromosomal DNA, a gene associated with a disease-causing mutation or chromosomal abnormalities in DNA from cultured erythroid cells are all within the scope of the invention.

The invention also provides a method for isolating fetal cells in metaphase from a maternal blood sample. In this method, after culture of cells in eythropoietin, cells are exposed to an agent which inhibits progression of dividing cells through the cell cycle or an agent which synchronizes growth of cells. Cells in metaphase can be detected microscopically and isolated. Metaphase cells can then be used for further analysis and diagnostic tests. The invention still further provides a method for preferentially isolating fetal cells from a current pregnancy in a woman who has had multiple pregnancies by culture of fetal erythroid progenitor cells from a peripheral blood sample obtained from the pregnant woman. This method is based upon the short life span, about 3 months, of erythroid progenitor cells such that any fetal erythroid cells that are isolated after culturing must be derived from the current fetus rather than any previous fetus.

Brief Description of the Drawings

Figure 1 is a schematic diagram illustrating the relative distribution of erythroid progenitor cells and nucleated red blood cells along a density gradient.

Figure 2 is a photograph of a CFU-E colony growing in methylcellulose medium containing erythropoietin. Figure 3 is a photograph of an early BFU-E colony growing in methylcellulose medium containing erythropoietin.

Detailed Description of the Invention

The present invention pertains to methods of culturing fetal cells from a sample of maternal peripheral blood. The methods include obtaining a sample of peripheral blood from a pregnant woman and culturing cells within the sample in a culture medium containing a cell growth factor. Fetal cells are isolated from the culture medium and a nucleic acid sequence of interest is detected in the fetal DNA. In an embodiment, the invention pertains to methods of culturing fetal erythroid cells from a sample of maternal peripheral blood. The methods involve culturing fetal erythroid progenitor cells present in the blood sample in vitro in a medium containing an erythroid cell growth factor, e.g., erythropoietin, which causes the fetal erythroid progenitor cells to proliferate and differentiate into fetal erythroid cells. Following culture, fetal erythroid cells can be isolated. In one embodiment, the method of the invention comprises obtaining a sample of peripheral blood from a pregnant woman, culturing cells within the sample of peripheral blood in a culture medium containing an erythroid growth factor, e.g., erythropoietin, and isolating fetal erythroid cells from the culture medium.

The language "a sample of peripheral blood from a pregnant woman", also referred to herein as a maternal blood sample, is intended to include a blood sample drawn (e.g. with a needle) from a peripheral blood source (e.g. an arm vein) from a woman pregnant with a fetus. Before use in methods of the invention, the sample of peripheral blood can be treated with an agent which inhibits blood coagulation, such as heparin. Preferably, the sample of peripheral blood contains approximately 5-20 mis of blood. More preferably, the sample contains approximately 10 mis of blood. The sample may be obtained as early as 6 weeks of gestation or as late as mid-second trimester, but preferably is obtained between about 8 and about 16 weeks of gestation. Most preferably, the sample is obtained at approximately 12 weeks of gestation. It is known that at 12 weeks of gestation there are about 50,000 nRBCs per milliliter circulating in fetal blood whereas by 20 weeks of gestation there are only about 1000 nRBCs per milliliter of fetal blood. See Holzgreve,W., et al. The Journal of Reproductive Medicine 37, 410-418 (1992).

The language "culturing cells" in a "culture medium" is intended to refer generally to contacting cells with a culture medium appropriate for the survival of the cells and incubating the cells in the culture medium under conditions appropriate for the survival of the cells for a period of time to allow proliferation and possibly differentiation of the cells. The language "culture medium" is intended to include a liquid or semisolid solution or material which allows survival and growth of fetal cells, e.g., erythroid progenitor cells. Suitable culture media generally contain a nutrient source, such as carbohydrates (e.g. sugars), general growth factors, such as those found in blood serum (e.g. fetal bovine serum), and supplementary amino acid(s), such as L-glutamine. A culture medium can also contain antibiotics, such as penicillin and streptomycin, a reducing agent (e.g. 2-mercaptoethanol), additional protein (e.g. bovine serum albumin), a buffering agent (e.g. sodium bicarbonate) and/or a pH indicator (e.g. phenol red). Conditions appropriate for survival of mammalian cells in culture generally are 37 C, 4 %-5 % CO2. The length of time of culturing can vary depending upon such factors as the number of cells desired (i.e. the extent of proliferation needed to produce the number of cells desired) and/or the degree of cellular differentiation desired, but generally is at least about 48 hours. Preferably cells are cultured about 4 days. More preferably, cells are cultured about 6 days. In a preferred embodiment, the culture medium further contains a semisolid matrix material. The language "semisolid matrix material" is intended to include a substance which, when added to a liquid culture medium, can convert the liquid culture medium to a gelatinous, semisolid state. A semisolid medium is one in which cells can still grow (i.e. which is not so solid as to inhibit cell growth) but which is firm enough so that isolated cells within the medium cannot migrate from their site of growth. A semisolid matrix material can also allow growing cells to attach to the matrix material, thereby preventing their migration within the medium. Growth of a single cell in a semisolid medium results in the production of a discrete colony of cells derived, by division, from the original single cell. A cell colony can be distinguished from other cell colonies by light microscopic examination of the semisolid culture medium. Suitable semisolid matrix materials include methylcellulose, agargel and a plasma clot. See for example Stephenson, J.R., et al. Proc. Natl. Acad. Sci. USA 68, 1542-1546 (1971); Iscove, N.N., et al. J. Cell Physiol. 83, 309-320 (1974); and Pike. B.L. and Robinson, W.A. J. Cell Physiol. 16, 11 (1970). A preferred culture medium for isolation of fetal erythroid cells from a maternal blood sample is Iscove's Modified Dulbecco's Medium (IMDM; commercially available; Sigma, Whitaker) with added 1.1 % methylcellulose, 30 % fetal calf serum, 1 % BSA, 1 % penecillin-streptomycin, 1 % L-glutamine, 10~4 M 2-mercaptoethanol. Preferred culture conditions are incubation of cells at 37 C, 4 %-5 % CO2 for approximately 96 hours. Cells are typically cultured in an appropriate container for the culture medium, such as a culture dish or culture flask.

The culture medium in which the cells of the maternal blood sample are cultured also contains a cell growth factor. The language "cell growth factor" is intended to include those factors which stimulate proliferation or differentiation of fetal cells. Examples of cell growth factors include: factors which stimulate differentiation of lymphoid progenitor cells, e.g., interleukin-2 (IL-2), interleukin-4 (IL-4),and interleukin-7, and factors which stimulate hematopoietic progenitor cells, e.g., hematopoietic cell growth factors (including erythroid growth factors), e.g., interleukin -3 (IL-3), granulocyte-macrophage colony- stimulating factor (GM-CSF), monocyte-macrophage colony-stimulating factor (M-CSF), granulocyte colony-stimulating factor (G-CSF), erythropoietin (EPO), and human recombinant stem cell factor (rhSCF).

The culture medium in which the cells of the maternal blood sample are cultured also can contain an erythroid growth factor. The language "erythroid growth factor" is intended to include ^J.hose factors which support growth and proliferation of human erythroid progenitor cells. Examples include: erythropoietin (EPO), and burst promoting activity (BPA). These erythroid growth factors can be administered alone or in combination with each other.

The language "erythropoietin" is intended to include a preparation of the protein erythropoietin which supports growth and proliferation of human erythroid progenitor cells. Erythropoietin may be prepared by purification of the protein from a natural source or may be prepared by expression of the erythropoietin gene by recombinant DNA technology. A crude preparation of erythropoietin isolated from human urine can be used and is commercially available from Sigma (catalogue # E5011, 50 U/mg). Other commercially available preparations include Sigma # E9761 (human recombinant; 100,000 U/mg), # E9757 (from human urine; 80,000 U/mg), # E2639 (from human urine; 500 U/mg) and # E2514 (from human urine; 100 U/mg). Preferably, the erythropoietin is of human origin, although erythropoietin from another species which is capable of supporting the proliferation and differentiation of human erythroid progenitor cells can also be used. Erythropoietin has been shown to function across species. For example, human and mouse cells have both been stimulated by sheep or human erythropoietin (see Iscove, N.N. et al. J. Cell Physiol. 83, 309- 320 (1974)). A concentration of erythropoietin is used which is sufficient to achieve the desired result of proliferation and/or differentiation of erythroid progenitor cells. A preferred concentration range for erythropoietin in the culture medium is about 0.25 U/ml to 1 U/ml. Erythroid progenitor cells exposed to erythropoietin in culture will be stimulated to proliferate and differentiate. Erythropoietin preferentially stimulates the proliferation and differentiation of fetal erythroid progenitor cells in cells from a maternal blood sample for a number of reasons. First, it has been demonstrated that fetal erythroid progenitor cells from human fetal liver are more sensitive to the stimulatory effects of erythropoietin than are erythroid progenitor cells from human adults (see Peschle, C, et al. Blood 58, 565-572 (1981)). Second, the concentration of erythropoietin used (e.g., 0.25 U/ml to 1 U/ml) is lower than the optimal concentration for stimulation of adult erythroid progenitors, which is 2-4 U/ml (see Linch et al. Blood 59, 976-979 (1982)). Finally, erythroid progenitor cells in human adults are found almost exclusively in the bone marrow, the site of adult hematopoiesis, and therefore few, if any, maternal erythroid progenitor cells should be present in maternal peripheral blood.

Two types of fetal erythroid progenitor cells can respond to the stimulatory effects of erythropoietin: the mature, or late, BFU-E (hereafter M-BFU-E) and the CFU-E. Colonies derived from M-BFU-Es and CFU-Es can be distinguished from each other and from other cell types by their morphology, which can be determined microscopically. BFU- E-derived colonies usually contain over 100 cells, and can contain several thousand cells, and are arranged in a diffuse "burst" or scattered pattern of cells. CFU-E-derived colonies contain about 20 to 150 cells arranged in a smaller, more compact colony. CFU-E-derived colonies mature during the first week of culture, whereas BFU-E-derived colonies may take longer to develop. Additionally, CFU-E-derived colonies may undergo further maturation and differentiation to form erythroid cells such as nucleated erythroblasts, which display visible hemoglobinization, which can be identified microscopically.

After culturing cells in a culture medium containing an erythroid growth factor, e.g., erythropoietin, erythroid cells are isolated from the culture medium. The language "isolated" is intended to refer to separation of one or more cells or colonies of cells away from other cells and/or away from the culture medium. When cells are grown in a semisolid culture medium at low enough cell density such that discrete, distinguishable cell colonies form, these cell colonies are already isolated from other cells within the semisolid medium. Cells or cell colonies can be further isolated by removing the cells from the medium. For example, a discrete cell colony can be identified by examining the cell culture under a light microscope and can be removed from the culture medium by picking out the cells from the culture medium. Cells can be effectively picked out of the culture medium using an instrument, such as a drawn out Pasteur pipette or other suitable pipette, to manipulate the cells. Once cells are separated from the culture medium, for example by picking them with a pipette, they can be transferred to a holding container (e.g. a tube), to a solid support (e.g. a microscope slide), to fresh culture media or to any other location suitable for further growth, manipulation or analysis of the cells. In a preferred embodiment, a single colony of cells is picked from the culture medium and transferred to a microscope slide. The language "fetal cells" is intended to include those fetal cells present in the maternal peripheral blood circulation. The fetal cells include fetal progenitor cells, which after incubation with a cell growth factor, can differentiate into mature cells, e.g., granulocytes, monocytes or leukocytes. For example, fetal lymphoid progenitor cells can be cultured with IL-2, IL-7, or IL-4 in order to differentiate into mature T or B-lymphocytes. Other fetal progenitor cells can be cultured with factors such as G-CSF and /or GM-CSF in order to differentiate into mature granulocytes.

The language "erythroid cells" is intended to include those cells which are derived from erythroid progenitor cells upon culture with an erythroid growth factor, e.g., erythropoietin. Culture of erythroid progenitor cells with an erythroid growth factor, e.g., erythropoietin, can cause progenitor cells to differentiate into a more mature stage of erythroid lineage development. Thus, the cells present after culture may be a mixture of cell types at different stages of the erythroid developmental pathway. Cells at different stages of the erythroid developmental pathway can be morphologically different from each other. Cells or colonies of cells having a morphology of erythroid progenitor cells, such as BFU-E and CFU-E colonies, or more developed erythroid cells, including cells with visible hemoglobinization, such as nucleated erythroblasts, are detectable after culture of erythroid progenitor cells in an erythroid growth factor, e.g., erythropoietin, and can be isolated from the culture medium. Before culturing cells from a sample of peripheral blood obtained from a pregnant woman, the sample can be manipulated to enrich for one or more types of cells within the sample. A maternal blood sample contains both nucleated cells, such as lymphocytes, monocytes, granulocytes and, potentially, fetal cells, e.g., fetal erythroid progenitor cells, and non-nucleated mature red blood cells. Non-nucleated mature red blood cells constitute the vast majority of cells within a sample of peripheral blood and it is often desirable to remove these cells from the sample before culturing. A proportion of nucleated cells present in the sample of peripheral blood can be increased relative to a proportion of nucleated cells present in the sample of peripheral blood prior to enrichment forming a sample enriched in nucleated cells. The sample enriched in nucleated cells is then used for culture. The language "a proportion of nucleated cells" is intended to refer to the percentage of nucleated cells, relative to the total number of cells, i.e. nucleated and non-nucleated cells, in the sample. The language "increased relative to a proportion present ... prior to enrichment" is intended to mean that the percentage of nucleated cells present in the sample is greater after an enrichment procedure as compared to the percentage of nucleated cells present in the sample before the enrichment procedure. The percentage of nucleated cells in a sample containing both nucleated and non-nucleated cells, such as peripheral blood, can be increased relative to the starting sample by removing non-nucleated cells from the sample and/or by separating nucleated cells away from non-nucleated cells. The language "a sample enriched in nucleated cells" is intended to include a sample derived from a maternal blood sample in which the percentage of nucleated cells present in the sample has been increased relative to the percentage present in the maternal blood sample prior to an enrichment procedure.

In one embodiment, a sample enriched in nucleated cells is formed by separating non-nucleated cells from nucleated cells in a sample of peripheral blood obtained from a pregnant woman. The language "separating" is intended to mean that non-nucleated cells and nucleated cells are isolated away from each other. Non-nucleated cells can be separated from nucleated cells by density gradient centrifugation. The language "density gradient centrifugation" is intended to refer to a technique in which a mixture of cells, e.g. cells of a blood sample, are centrifuged in a container, such as a centrifuge tube, through a material of a particular density to form layers containing different cell types. Under centrifugation conditions, cells travel through the density gradient material as a function of their cell size and density. Therefore, density gradient centrifugation can be used to separate cells based upon differences in cell size and density. Non-nucleated red blood cells have a different cell size and density than nucleated cells and thus can be separated from nucleated cells by density gradient centrifugation. Several materials suitable for separating non- nucleated and nucleated cells by density gradient centrifugation are commercially available. These include Ficoll (Pharmacia, Uppsala, Sweden), Histopaque (Sigma Diagnostics, St. Louis, MO), Nycodenz (Nycomed Pharma, Oslo, Norway; available in the US from Gibco BRL) and Polymorph^rep (Nycomed Pharma, Oslo, Norway; available in the US from Gibco BRL). Appropriate centrifugation speeds and times have been determined by the manufacturers. Generally, appropriate centrifugation speeds and times are in the range of 400-500 g for 20-35 minutes at room temperature.

After centrifugation, the maternal blood sample is typically separated into at least three layers: a supernatant layer containing serum and platelets; a mononuclear cell layer; and an agglutinated pellet which contains non-nucleated red blood cells. One or more cell layers containing nucleated cells and devoid of most non-nucleated red blood cells (e.g. the mononuclear cell layer) is removed from the gradient to form a sample enriched in nucleated cells. The sample enriched in nucleated cells is then used for culture. In another embodiment, non-nucleated cells can be separated from nucleated cells by selective lysis of the non-nucleated cells. The language "selective lysis" is intended to refer to preferential destruction of one cell type, e.g. non-nucleated cells, compared to other cell types, e.g. nucleated cells. Cells in the maternal blood sample can be incubated in one of a number of hypotonic buffers known to be effective, and conventionally used, for lysing non-nucleated red blood cells, such as 0.17M NH4CI, 0.01M Tris, pH 7.3 (see e.g. Tiilikainen et al. Transplantation 17, 355 (1974) and Macera, et al. Leukemia Research 13, 729-734 (1989)). Buffers suitable for this purpose are also available commercially (e.g. "Lyse and Fix", GenTrak). After incubation in an appropriate buffer suitable for selective lysis of non-nucleated cells, the resultant cell sample is a sample enriched in nucleated cells. In a preferred embodiment, the maternal blood sample is manipulated before culturing cells in order to enrich for fetal cells, e.g., fetal erythroid progenitor cells, within the sample. A proportion of fetal cells, e.g., fetal erythroid progenitor cells, present in the sample of peripheral blood can be increased relative to a proportion of fetal cells, e.g., fetal erythroid progenitor cells, present in the sample of peripheral blood prior to enrichment forming a sample enriched in fetal cells, e.g., fetal erythroid progenitor cells. The sample enriched in fetal cells is then used for culture. The language "a proportion of fetal cells, e.g., fetal erythroid progenitor cells" is intended to refer to the percentage of fetal cells, relative to the total number of cells, i.e. nucleated and non-nucleated cells, in the sample. The language "increased relative to a proportion present ... prior to enrichment" is intended to mean that the percentage of fetal cells present in the sample is greater after an enrichment procedure as compared to the percentage of fetal cells present in the sample before the enrichment procedure. The percentage of a particular type of fetal cells, e.g., fetal erythroid progenitor cells, in a sample containing other cell types, as in peripheral blood, can be increased relative to the starting sample by removing other cell types, e.g. non-nucleated cells and other types of nucleated cells, from the sample and/or by separating the particular type of cells, e.g., erythroid progenitor cells, away from non-nucleated cells and other types of nucleated cells. The language "a sample enriched in fetal cells, e.g., erythroid progenitor cells" is intended to include a sample derived from a maternal blood sample in which the percentage of fetal cells, e.g., fetal erythroid progenitor cells present in the sample has been increased relative to the percentage present in the maternal blood sample prior to an enrichment procedure.

A particular type of fetal cells, e.g., erythroid progenitor cells, can be separated from non-nucleated cells, e.g. mature red blood cells, and other types of nucleated cells, e.g. lymphocytes and mononuclear cells, by density gradient centrifugation based upon differences in cell size and density of selected cell type and other cell types. In one embodiment, "dual density gradient centrifugation" is performed. The language "dual density gradient centrifugation" is intended to include techniques in which a mixture of cells, e.g. cells of a blood sample, are centrifuged in a container, such as a centrifuge tube, through two or more materials of different densities to form layers containing different cell types or, alternatively, a mixture of cells is centrifuged through a material of one density, one or more cell layers is removed and the removed cells are recentrifuged through a second material, preferably of a different density than the first, to form layers containing different cell types. Dual density gradient centrifugation can be performed with the gradient materials described for single density gradient centrifugation (e.g. Ficoll, Histopaque, Nycodenz, Polymorphprep).

A sample enriched in fetal cells, e.g., erythroid progenitor cells, can be produced, for example, by density gradient centrifugation using Polymorphprep (Nycomed; Gibco BRL catalogue #100-1971). Undiluted anticoagulated whole blood of a maternal blood sample can be centrifuged through the gradient material to form the following layers (from the top to the bottom of the tube): a supernatant layer containing serum and platelets; a mononuclear cell layer (the "buffy coat") containing lymphocytes and monocytes; a polymorphonuclear cell layer containing polymorphonuclear leukocytes, nucleated red blood cells and erythroid progenitor cells; and a non-nucleated red blood cell pellet. The polymorphonuclear cell layer can be removed to form a sample enriched in fetal erythroid progenitor cells.

Additionally, dual density gradient centrifugation using Histopaque can be used to produce a sample enriched in erythroid progenitor cells. A maternal blood sample, diluted 1 : 1 in phosphate buffered saline, is centrifuged through a dual density gradient composed of Histopaque 1077 (density = 1.077 gm/ml) layered on top of Histopaque 1119 (density = 1.119 gm ml) to form the following layers (from the top to the bottom of the tube): a supernatant layer; a mononuclear cell layer (the "buffy coat"); a nucleated red blood cell layer, containing erythroid progenitor cells, at the interface between the 1.077 and 1.119 gradients; and a non-nucleated red blood cell pellet. The nucleated red blood cell layer can be removed to form a sample enriched in erythroid progenitor cells. Additionally,

Histopaque 1077 and 1119 can be used sequentially, as described in Example 1, to form a sample enriched in fetal cells, e.g., fetal erythroid progenitor cells.

When single or dual density gradient centrifugation is performed, the cell layers may not always be discrete and identifiable. Therefore, several different cell layers may be removed and _^ultured separately in culture media containing a cell growth factor, e.g., an erythroid growth factor, e.g., erythropoietin, both to ensure that fetal cells, e.g.,erythroid progenitor cells have been properly selected and for comparison purposes (i.e., other cell layers can function as controls).

A sample enriched in fetal cells, e.g., fetal erythroid progenitor cells, can also be produced using other techniques to remove non-nucleated cells and/or other cell types from a maternal blood sample or to enrich for erythroid progenitor cells. For example, antibodies specific for a cell surface marker present on a particular cell type can be used to either remove or enrich for that cell type within a maternal blood sample. An antibody specific for a marker present on the selected fetal cell type, e.g., fetal erythroid progenitor cell, can be used to select for erythroid progenitor cells, for example by flow cytometry, panning or immunomagnetic separation (positive selection of cells). Alternatively, an antibody specific for a marker present on a cell type other than the selected fetal cell type can be used to remove that cell type from a maternal blood sample, for example by flow cytometry, panning or immunomagnetic separation (negative selection of cells) or by complement-mediated lysis of cells which have bound the antibody.

Although culture of a maternal blood sample, a sample enriched in nucleated cells or a sample enriched in fetal cells, e.g., erythroid progenitor cells, in a culture medium containing a cell growth factor, e.g., an erythroid growth factor, e.g., erythropoietin, can preferentially produce colonies of fetal cells (for example, due to the scarcity of maternal erythroid progenitor cells in peripheral blood and the higher sensitivity of fetal erythroid progenitor cells for erythropoietin), erythroid cells or cell colonies cannot be identified morphologically as being fetal-derived. Therefore, to identify fetal cells as being fetal- derived, the method of the invention can further comprise detecting a fetal cell marker on or in the fetal cells. The language "fetal cell marker" is intended to include a cell-surface, cytoplasmic or nuclear structure or molecule present on or in a fetal cell which can be used to distinguish a fetal cell from a maternal cell.

The language "detecting a fetal cell marker" is intended to include detection of the fetal cell marker itself or detection of another molecule which binds to the fetal cell marker. A fetal cell marker can be detected on or in a fetal cell by, for example, contacting the cell, or portion thereof, with an antibody reactive against the fetal cell marker. The antibody can be labelled with a detectable substance or can be further reacted with another molecule, e.g. a second antibody reactive against the first antibody, which is labelled with detectable substance. Suitable detectable substances include enzymes, prosthetic groups, fluorescent materials, luminescent materials and radioactive materials. Examples of suitable enzymes include horseradish peroxidase, biotin, alkaline phosphatase, β-galactosidase, or acetylcholinesterase; examples of suitable enzyme/prosthetic group complexes include streptavidin and biotin; examples of suitable fluorescent materials include umbelliferone, fluorescein, fluorescein isothiocyanate, rhodamine, dichlorotriazinylamine fluorescein, dansyl chloride or phycoerytlirin; an example of a luminescent material includes luminol; and examples of suitable radioactive material include 1 ? I, 131 I, 35 S or ^J 3H. Standard immunohistochemistry techniques can be used to detect fetal cell marker expression using an antibody against the fetal cell marker.

Proteins which are expressed preferentially or exclusively on or in fetal cells as compared to maternal cells can be used as fetal cell markers and can be detected using an antibody against the protein. Examples of suitable proteins which can be used as fetal cell markers include fetal hemoglobin and a fetal histone protein H°l . Fetal hemoglobin is a cytoplasmic protein, whereas fetal H°l is a nuclear protein. Additionally, an antigenic determinant which is preferentially or exclusively expressed on or in fetal erythroid cells can be used as a fetal cell marker. An example of a suitable antigenic determinant which can be used as a fetal cell marker is the i antigenic determinant. The i antigenic determinant is an unbranched membrane carbohydrate present on fetal erythroid cells. Adult erythroid cells generally do not express i but rather express the I antigenic determinant, a branched carbohydrate. Other examples of fetal cell markers include nucleic acid sequences which are unique to fetal cells compared to maternal cells. Examples of unique nucleic acid sequences include Y chromosomal DNA, in the case of fetal cells from a male fetus, and DNA encompassing a paternal polymorphism or polymorphic allele, such as an HLA antigen allele. Following culture and isolation of fetal cells, the method of the invention can further comprise detecting a nucleic acid sequence of interest in fetal nucleic acid of isolated fetal cells, e.g., fetal erythroid cells. The language "a nucleic acid sequence of interest" is intended to include DNA, e.g. chromosomal DNA or a particular gene fragment within chromosomal DNA or amplified from chromosomal DNA, as well as RNA, e.g. mRNA or nuclear RNA. The language "fetal nucleic acid" is intended to include fetal DNA and RNA, for example fetal chromosomal DNA and fetal mRNA. Detection of a nucleic acid of interest in nucleic acid of fetal cells can be used to identify a fetal cell as being fetal-derived (i.e., the fetal nucleic acid can be a fetal cell marker) and also for analyses such as determination of fetal gender, detection of a genetic disease in the fetus and detection of a chromosomal abnormality in the fetus. Fetal nucleic acid in fetal cells, cultured and isolated as described herein, can be analyzed for the occurrence of a nucleic acid of interest for diagnostic or other purposes.

Cultured fetal cells can be treated such that fetal nucleic acid present in the cells is made available for detection. A nucleic acid sequence of interest can be detected in the available fetal nucleic acid by, for example, enzymatically amplifying the nucleic acid sequence of interest and/or by hybridizing a labelled nucleic acid probe to the nucleic acid sequence of interest. The labelled probe used to detect a nucleic acid sequence of interest can be, for example, a labelled DNA probe, a labelled RNA probe or labelled oligonucleotides. Fetal DNA can be made available by boiling the fetal cells to lyse them, thereby releasing fetal DNA, for instance prior to amplification of fetal DNA. Fetal cells can be attached to a solid support, e.g. a microscope slide, in such a way that fetal nucleic acid is made available, for example by fixing fetal erythroid cells or erythroid cell nuclei to a microscope slide prior to in situ hybridization. Fetal nucleic acid in fetal cells, or portion thereof (e.g. nuclei), can be detected directly, for example by in situ hybridization of a labelled nucleic acid probe complementary to a nucleic acid sequence of interest or the fetal nucleic acid can be amplified prior to detection using a known amplification technique such as the polymerase chain reaction (PCR). Primers for PCR amplification are chosen which specifically amplify a DNA sequence of interest in the fetal DNA A nucleic acid sequence which is to be detected in fetal nucleic acid of fetal erythroid cells is referred to herein as a "nucleic acid sequence of interest". In one embodiment the nucleic acid of sequence of interest is a Y chromosomal DNA sequence. The language "a Y chromosomal DNA sequence" is intended to include nucleotide sequences unique to the Y chromosome, including repetitive sequences. Detection of a Y chromosomal DNA sequence can be used both for detection of a fetal cell marker and for fetal gender identification.

In another embodiment the nucleic acid sequence of interest is a sequence of a gene associated with a disease-causing mutation. The language "a sequence of a gene associated with a disease-causing mutation" is intended to include nucleotide sequences encompassing all or part of a gene which may contain a mutation which causes or is associated with a disease, or is linked to a gene which may contain a mutation which causes or is associated with a disease. That is, the detected gene sequence or a gene sequence linked to the detected gene sequence may contain a mutation which, if the mutation is present, causes or is associated with a disease. Detection of such a sequence can be used to determine if a fetus has the disease caused by or associated with the mutation.

In yet another embodiment, the nucleic acid sequence of interest detects a chromosomal abnormality. The language "detects a chromosomal abnormality" is intended to include nucleotide sequences which can detect changes in chromosome number, chromosomal deletions, chromosomal rearrangements and/or chromosomal alterations. A nucleic acid sequence of interest which can detect a chromosomal abnormality can be, for example, a nucleic acid sequence specific for a particular chromosome, such as a repetitive sequence from a particular chromosome. Preferred chromosomes to be examined for detection of a chromosomal abnormality include chromosomes 13, 18, 21, X and Y. Chromosomal trisomies are most prevalent for these particular chromosomes. Preparation of nucleic acid probes suitable for detecting chromosomal abnormalities can be made by standard techniques. For example see Klinger, K., et al. Am. J. Hum. Genet. 51, 55-65 (1992).

In one embodiment, a nucleic acid of interest is detected in fetal nucleic acid of fetal cells by in situ hybridization. In situ hybridization can be used both to detect chromosomal DNA and to detect cytoplasmic mRNA. See for example Lichter, P., et al. Hum. Genet. 80, 224-234 (1988) and U.S. Patent No. 4,888,278 to Singer et al. If in situ hybridization is to be carried out, fetal cells are separated onto a solid support, such as a microscope slide, such that fetal nucleic acid is available for detection. In situ hybridization can be used, for example, to detect Y chromosome-specific sequences in fetal DNA in order to determine the gender of a fetus as described in greater detail in Example 3. In situ hybridization can also be used to assess chromosomal abnormalities in a fetus, including chromosomal aneuploidies, such as a trisomy, or chromosomal rearrangements or deletions, as described in greater detail in Example 5. In another embodiment, fetal DNA is enzymatically amplifed prior to detection of a nucleic acid sequence of interest. Fetal DNA can be amplified by PCR as described in detail in Example 4. If amplification is to be carried out, fetal erythroid cells can be lysed by boiling and fetal DNA can then be amplified for an appropriate number of cycles of denaturation and annealing (e.g., approximately 24-60). Control samples include a tube without added DNA to monitor for false positive amplification. With proper modification of PCR conditions, more than one separate fetal gene can be amplified simultaneously. This technique, known as "multiplex" amplification, has been used with six sets of primers in the diagnosis of Duchenne's Muscular Dystrophy (Chamberlain, J.S., et al., Prenat. Diagnosis, 9, 349-355 (1989)). When amplification is carried out, the resulting amplification product may be a mixture which contains amplified fetal DNA of the sequence of interest (i.e., the DNA whose presence is to be detected and/or quantitated) and other DNA sequences. The amplified fetal DNA sequence of interest and other DNA sequences can be separated, using known techniques, for example by gel electrophoresis. Subsequent analysis of amplified DNA can be carried out using known techniques, such as: digestion with a restriction endonuclease, ultraviolet light visualization of ethidium bromide stained agarose gels, DNA sequencing, or hybridization with a labelled DNA probe, for example, specific oligonucleotide probes (Saiki, R.K., et al, Am. J. Hum. Genet., 43 (Suppl):A35 (1988)). Amplification of and hybridization to allele-specific sequences can determine whether polymorphic differences exist between the amplified "maternal" and "fetal" samples and can be used to identify a female fetus based upon detection of paternal polymorphisms in the fetal DNA. DNA sequences from the father can be identified in the autosomal chromosomes of the fetus. Consequently, the method of the present invention can be used to separate and identify female fetal cells, as well as male fetal cells, from maternal blood. Thus, the method can be used for all nucleic acid-based diagnostic procedures currently being achieved with other methods, such as amniocentesis.

Following amplification, the amplification mixture can be separated on the basis of the size of the nucleic acid fragments and the resulting size-separated fetal nucleic acid can be contacted with an appropriate selected nucleic acid probe or probes (nucleic acid sufficiently complementary to the nucleic acid sequence of interest that it hybridizes to the nucleic acid sequence of interest in fetal nucleic acid under the conditions used). Generally, the nucleic acid probes are labelled (e.g., with a radioactive material, a fluorophore or other detectable material). After the size-separated fetal nucleic acid and the selected nucleic acid probes have been maintained for sufficient time under appropriate conditions for hybridization of complementary nucleic acid sequences to occur, resulting in production of fetal nucleic acid/nucleic acid probe complexes, detection of the complexes is carried out using known methods. For example, if the probe is labelled, a fetal nucleic acid/labelled nucleic acid probe complex can be detected and/or quantitated (e.g., by autoradiography, detection of the fluorescent label). The quantity of labelled complex (and, thus, of fetal nucleic acid) can be determined by comparison with a standard curve (i.e., a predetermined relationship between the quantity of label detected and a given amount of nucleic acid present).

Fetal DNA/labelled probe complexes are subsequently detected, using a known technique, such as autoradiography. Simple presence or absence of hybridization of the nucleic acid probe complementary to a nucleic acid of interest and the fetal DNA can be determined or the quantity of fetal DNA containing the nucleic acid sequence of interest can be determined. The result is a qualitative or quantitative assessment of fetal DNA obtained from a maternal blood sample. For many genes which may carry a disease-causing mutation, probes are available which detect a restriction-site polymorphism which is indicative of the presence of a disease-causing mutation within the gene. Detection of such a restriction site polymorphism in fetal DNA using a nucleic acid probe specific for a gene associated with a disease of interest is indicative that the fetal DNA carries a mutation in the gene and therefore that the fetus has the disease. The presence of fetal nucleic acid associated with diseases or conditions can be detected and/or quantitated by the present method. In each case, an appropriate probe is used to detect the nucleic acid sequence of interest. For example, sequences from probes Stl4 (Oberle, I., et al., New Engl. J. Med., 312, 682-686 (1985)), 49a (Guerin, P., et al., Nucleic Acids Res., 16, 7759 (1988)), KM-19 (Gasparini, P., et al., Prenat. Diagnosis, 9, 349- 355 (1989)), or the deletion-prone exons for the Duchenne muscular dystrophy (DMD) gene (Chamberlain, J.S., et al., Nucleic Acids Res., 16, 11141-11156 (1988)) are used as probes. Stl4 is a highly polymorphic sequence isolated from the long arm of the X chromosome that can be used to distinguish female fetal DNA from maternal DNA. It maps near the gene for Factor VIII :C and thus can also be utilized for prenatal diagnosis of Hemophilia A. Primers corresponding to sequences flanking the six most commonly deleted exons in the DMD gene, which have been successfully used to diagnose DMD by PCR, can also be used (Chamberlain, J.S. et al., Nucleic Acids Res., 16, 11141-11156 (1988)). Other conditions which can be diagnosed by the present method include cystic fibrosis (Newton, C.R., et al. Lancet 2, 1481-1483 (1989)); β-thalassemia (Cai, S-P., et.al., Blood, 73:372-374 (1989)); Cai, S-P., et al., Am. J Hum. Genet., 45:112-114 (1989)); Saiki, R.K., et al.,

New Engl. J. Med., 319, 537-541 (1988)), sickle cell anemia (Saiki, R.K., et al., New Engl. J. Med, 319, 537-541 (1988)), phenylketonuria (DiLella, A.G., et al., Lancet, 1, 497-499 (1988)) and Gaucher disease (Theophilus, B., et al., Am. J. Hum. Genet., 45, 212- 215 (1989)). An appropriate probe (or probes) is available for use in the present method for assessing each condition (see for example PCT Application WO 91/07660).

In another embodiment, the invention provides a method for isolating fetal erythroid cells in metaphase from a sample of peripheral blood from a pregnant woman. The method involves obtaining a sample of peripheral blood from a pregnant woman, culturing cells within the sample of peripheral blood in a culture media containing an erythroid growth factor, e.g., erythropoietin, exposing cultured cells to an agent which inhibits progression of dividing cells through the cell cycle or synchronizes growth of cells and isolating fetal erythroid cells in metaphase. Agent which inhibit progression of dividing cells through the cell cycle are known in the art and include colcemid, colchicine and vinblastine salt. Agents which synchronize growth of cells are also known in the art and include bromodeoxyuridine, fluorodeoxyuridine and ethidium bromide. Following culture of cells, metaphase spreads can be prepared by standard methods and used for further analysis (see Example 2).

In another embodiment, the invention provides a method for preferentially isolating fetal cells from a current pregnancy in a peripheral blood sample from a multiparous pregnant woman. The method involves obtaining a sample of peripheral blood from a multiparous pregnant woman, culturing cells within the sample of peripheral blood in a culture medium containing an erythroid growth factor, e.g., erythropoietin, and isolating fetal erythroid cells to preferentially isolate fetal cells from a current pregnancy. The language "multiparous pregnant woman" is intended to refer to a pregnant woman who has had one or more other pregnancies in addition to her current pregnancy. This method is based upon the short life span of erythroid progenitor cells (see Beck, W.S. Hematology, 5th edition, MIT Press: Cambridge, MA (1991)). Since the lifespan of erythroid progenitor cells is about three months, fetal erythroid cells cultured from a peripheral blood sample of a multiparous pregnant woman will be derived from erythroid progenitor cells of a current fetus rather than from a former fetus. This is an advantage of isolating fetal erythroid cells rather than another, longer lived, fetal cell type that may be present in a maternal blood sample. For example, although fetal lymphocytes are also present in maternal peripheral blood, lymphocytes may live as long as several years in peripheral blood (Schroder et al. Transplantation 17, 346 (1974); Ciaranfi et al. Schweizerische Medizinische Wochenschrift 107, 134-138 (1977)). Thus, a fetal lymphocyte detected in a maternal blood sample of a multiparous pregnant woman could either be from the current fetus or from a former fetus, making gender identification and/or diagnosis of the current fetus problematic.

This invention is further illustrated by the following examples which should not be construed as liiiiiting. The contents of all references and published patents and patent applications cited throughout this application are hereby incorporated by reference.

EXAMPLE 1: Culture of Erythroid Progenitor Cells from a Sample of Peripheral Blood from a Pregnant Woman

In this example, erythroid progenitor cells were cultured from samples of peripheral blood from pregnant women. The peripheral blood samples were first subjected to density gradient centrifugations (single or dual) in order to eliminate most of the abundant non-nucleated red blood cells present in peripheral blood and to remove many of the maternally-derived nucleated cell types in the sample while also enriching for erythroid progenitor cells prior to the culturing step. This allowed for culturing of much fewer total cell numbers than were present in the original peripheral blood sample.

Maternal peripheral blood samples were obtained by venipuncture with informed consent from pregnant women planning to undergo pregnancy sonogram

(ultrasound) and genetic amniocentesis. In each case, the maternal blood sample was obtained before the amniocentesis was performed. Approximately 18 ml of whole peripheral blood, divided between two Vacutainer tubes containing Acid Citrate Dextrose solution A (ACD-A) as an anticoagulant (Becton Dickson), was obtained from each pregnant woman. The maternal blood samples were received via overnight courier the day after they were obtained. After being inverted a few times to ensure mixing of the samples, 5 ml of undiluted blood was taken from each of three samples and layered atop a Polymorphprep gradient as explained below. The remaining blood was diluted with an equal volume of phosphate buffered saline and 5 ml aliquots of the diluted blood were layered atop a Ficoll-Paque gradient, a Histopaque dual gradient, and a Nycodenz dual gradient. A schematic diagram illustrating the typical relative distribution of erythroid progenitor cells and nucleated red blood cells along a density gradient is shown in Figure 1.

Ficoll-Paque Gradient

A 3.5 ml aliquot of Ficoll-Paque (1.077 g/ml; Pharmacia, Uppsala, Sweden) in a 15 ml conical centrifuge tube was overlayered with 5 ml of diluted maternal blood. The tubes were spun for 30 minutes at 400 x g in a tabletop centrifuge at room temperature. The buffy coat cell layer at the interface of the plasma and the Ficoll-Paque (termed Fl) was withdrawn, rinsed, and cultured as described below.

Histopaque Dual Gradient

A 2 ml aliquot of Histopaque 1.119 g/ml (Sigma Chemical Co.) in a 15 ml tube was carefully overlayered with 1.5 ml of Histopaque 1.077 g/ml (Sigma Chemical Co.). A 5 ml aliquot of diluted blood was then layered atop Histopaque 1.077 g/ml. The tubes were spun for 30 minutes at 400 x g in a tabletop centrifuge at room temperature. A Pasteur pipette was used to harvest the buffy coat at the interface of the plasma and the 1.077 g/ml gradient (termed HI). To harvest the region thought to be the most likely to contain nucleated red blood cells, the entire 1.119 layer from the 1.077/1.119 interface down to the red blood cell pellet (termed H2) was aspirated with a Pasteur pipette. These two cell fractions were placed into clean tubes, rinsed, and cultured as described below.

Polymorphprep Gradient

A 3.5 ml aliquot of Polymorphprep 1.113 glml (NycomedlGibco BRL) in a 15 ml tube was gently overlayered with 5 ml of undiluted whole blood. The tubes were spun for 30 minutes at 500 x g in a tabletop centrifuge at room temperature. As the red blood cells in the whole blood sample contacted the Polymorphprep medium, they were agglutinated by the Dextran 500 and sedimented through the gradient under centrifugal force. This gradient medium is designed to separate two cell populations by exploiting the tendency of red blood cells to lose water upon entering a hyperosmotic environment. As expected, the maternal blood samples demonstrated two distinct bands on the Polymorphprep gradient. The buffy coat at the plasma/Polymorphprep interface from each of the blood samples (termed PI) was withdrawn with a Pasteur pipette and transferred to a clean tube. Then the polymorphonuclear cell layer (termed P2), which theoretically contained the nucleated red blood cells and erythroid progenitors, was transferred into another clean tube. An equal volume of sterile 0.45 % NaCl in distilled H2O was added to restore the cells to the proper osmolarity. The cells were diluted up to 12 ml in phosphate buffered saline, and cultured as described below.

Nycodenz Dual Gradient

Nycodenz was obtained in powdered form (Gibco/BRL), and 27.6 g of Nycodenz powder was reconstituted up to a total volume of 100 ml in buffered medium consisting of 5 mM Tris-HCl, pH 7.5, 3 mM KC1, and 0.3 mM EDTA in distilled H2O. This 27.6 % solution, at a density of 1.150 g/ml, was sterilized by autoclaving. Iso-osmotic NaCl diluent (density 1.003 g/ml) was made by mixing 0.75 g NaCl into 100 ml of the buffered medium described above. The NaCl diluent solution was sterilized through a 0.22 μm filter. The 27.6 % Nycodenz solution and NaCl diluent were mixed in the ratios of 3:2 and 1 :1 to yield two solutions having a density of 1.090 g/ml and 1.075 g/ml respectively. A 2 ml aliquot of Nycodenz 1.090 g/ml in a 15 ml tube was carefully overlayered with 1.5 ml of Nycodenz 1.075 g/ml. A 5 ml aliquot of blood diluted 1 : 1 in phosphate buffered saline was then layered atop Nycodenz 1.075 g/ml. The tubes were spun for 30 minutes at 1500 x g in a tabletop centrifuge at room temperature. It was noted that this gradient combination resulted in cell bands which were much more diffuse than on other gradients. The cell layers were contaminated with a large number of nonnucleated red blood cells. A Pasteur pipette was used to harvest the cell layer at the interface of the plasma and the 1.075 g/ml gradient (termed Nl). Most of the 1.090 layer was harvested with a Pasteur pipette (termed N2); because there were so many cells present in this layer, not all of them were used. These cell fractions were placed into clean tubes.

Preparation of Culture Medium

The basic culture medium used consisted of 1.1 % methylcellulose (4000 centipoises; Sigma Chemical Co.), 30 % fetal calf serum (Sigma Chemical Co.), 0.1 mM 2-mercaptoethanol (Sigma Chemical Co.), and 1 % bovine serum albumin (BSA, Fraction V; Sigma Chemical Co.) in Iscove's Modified Dulbecco's Medium (IMDM; Sigma Chemical Co.). Cultures were supplemented with 2 units/ml of erythropoietin (Sigma Chemical Co.) to promote erythroid cell growth and maturation.

To prepare a solution of 3.2 % methyl-cellulose in IMDM, 6.4 g of methylcellulose was weighed into a one liter autoclavable bottle containing 190 ml sterile IMDM. The solution was mixed on a stirring plate while the methylcellulose went slowly into solution. The solution was autoclaved with the stir bar still in the bottle. After autoclaving, the methylcellulose formed a solid pink plug in the bottle. The bottle was shaken vigorously until the methylcellulose broke up into a slurry. As it cooled to room temperature, the methylcellulose went back into solution. BSA at a final concentration of 10 % and pH of 7.6 was prepared by weighing out 10 g BSA and adding it slowly to about 60 ml distilled H2O in a small beaker on a stir plate. Meanwhile, 10 % NaHCθ3 was prepared by dissolving 10 g NaHCO3 in distilled H2O to a final volume of 100 ml in a volumetric flask and filter sterilizing through a 0.22 μm filter. The BSA solution was titrated to pH 7.6 with 10 % NaHCO3 while being continuously monitored with a pH meter. This solution was poured into a volumetric flask, brought up to 100 ml with distilled H2O, and filter sterilized first through a 0.45 μm filter and then through a 0.22 μm filter.

Under a fume hood, 10 μl of 2-mercaptoethanol was added to 1 ml of IMDM. In a sterile tissue culture hood, 0.75 ml of this solution was added to 49.5 ml of IMDM, mixed, and filter sterilized through a 0.22 μm filter.

To the autoclaved bottle of 3.2 % methylcellulose/IMDM was added 5 ml of penicillan-streptomycin antibiotic (5000 units/ml, and 5000 μg/ml; Gibco/BRL), 5 ml of L-glutamine (200 mM; Gibco/BRL), 180 ml of sterile heat-inactivated fetal calf serum (Sigma Chemical Co.), 60 ml of 10 % BSA pH 7.6, and 40 ml filtered 2-mercaρto- ethanol/IMDM. The contents of the bottle were mixed thoroughly and the medium was divided into 2.3 ml aliquots in 15 ml tubes. The tubes were stored frozen at -20 °C until needed.

Plating and Culture of Cells

After the whole blood sample had been separated by density gradient centrifugation on a dual Histopaque gradient (1.077 g/ml over 1.119 g/ml) or on a Polymorphprep gradient, the desired cell fraction was rinsed twice in phosphate buffered saline, rinsed once in IMDM, and resuspended up to a total volume of 0.5 ml in IMDM. This 0.5 ml of cell suspension was added to a 2.3 ml aliquot of the methylcellulose medium in a 15 ml conical centrifuge tube along with 0.2 ml of 30 units/ml erythropoietin, at a final concentration of 2 units/ml.

Tubes were vigorously vortexed to ensure complete mixing of contents, as well as to break up any cell clusters which could mimic colony formation. The final product was of the consistency of thick syrup. 1.5 ml of the medium containing cells was placed in each of two covered 35 mm Falcon plastic tissue culture dishes (Becton Dickson, Lincoln Park, NJ) within a 100 mm Falcon plastic dish (Becton Dickson). A third, uncovered, 35 mm Falcon dish was placed within the 100 mm Falcon dish and partially filled with two ml of sterile distilled water to provide humidity. The cover was placed on the 100 mm Falcon dish and the entire dish was transferred to a 37 °C incubator with a 5 % CO2 atmosphere.

The more mature CFU-E progenitors formed recognizable colonies of up to 150 cells within four days, and by eight days in culture virtually all of these maturing cells contained visible hemoglobin. Because the erythropoietin receptor is not expressed until the BFU-Es have attained a higher level of differentiation, the BFU-Es did not immediately begin to proliferate in culture. By days eight to ten of culture, large scattered BFU-E colonies were visible and by day 14 these colonies were strongly hemoglobinized.

Representative examples of CFU-E and BFU-E colonies are shown in Figures 2 and 3, respectively (the colonies depicted in the figures were isolated from fetal livers). Table 1 summarizes the number of both BFU-E and CFU-E colonies obtained from culturing each cell fraction of a blood sample from a pregnant woman. The sample was fractionated using Ficoll-Paque (F), dual Histopaque (H), Polymorphprep (P), and dual Nycodenz (N) density gradients. Colonies from fraction N2 could not be seen or counted due to the presence of a large number of red blood eells which obscured the culture.

Table 1 : Relative numbers of progenitor cells in fractionated whole blood

Density Total Number Number

Fraction Colonies BFU-E ( % CFU-E ( % )

Fl (1.077) 71 64 (90.2) 7 (9.8)

HI (1.077) 68 63 (92.6) 5 (7.4)

H2 (1.119) 1 0 (0.0) 1 (100)

PI (1.077) 121 106 (87.6) 15 (12.4)

P2 (1.113) 16 4 (25.0) 12 (75.0)

Nl (1.075) 72 71 (98.6) 1 (1.4)

N2 (1.090) not analyzable

Totals: 349 308 (88.3) 41 (11.7)

EXAMPLE 2: Isolation of Cultured Erythroid Colonies and Preparation of Metaphase Spreads

In this example, erythroid cell colonies cultured as described in Example 1 were isolated and used to prepare metaphase spreads of erythroid cells on a microscope slide for further analysis.

Slide Preparation

The erythroid colonies were harvested directly onto Teflon-backed glass microscope slides (Cel-Line Assoc. Inc., Newfield, Nl) according to the basic method of Rajendra et al. (Human Genetics (1980) 55:363-366) and Dube et al. (Cancer Genetics and Cytogenetics (1981) 4:157-168). Because these erythroid cells are normally nonadherent, the slides were first coated with 0.1 % poly-L-lysine to induce the cells to attach to the slides so that they would not be lost during the harvesting procedure. For poly-L-lysine treatment, 100 mg poly-L-lysine (Sigma Chemical Co.) was dissolved in 10 ml of distilled H2O to make a 1 % stock solution. Aliquots of one ml each were stored frozen until used. To make a 0.1 % working solution, 100 μl of 1 % stock solution was diluted with 900 μl of distilled H2O. A drop of the working solution was placed on each 5 mm site on each slide. The slides were laid flat and nonoverlapping in a humid chamber (a sealed plastic box containing moist paper towels to provide humidity and a rack to keep the slides from contacting the paper towels) and incubated for two hours at 37 °C. After the incubation the slides were rinsed individually in a beaker of distilled water and air dried at room temperature. The coated slides were stored frozen at -20 °C until ready to use.

Harvesting of Colonies

Cells from CFU-E colonies were best harvested from culture at two to six days because their growth peaked quickly. After this time the nuclei had either been extruded or were dense and pyknotic, rendering further analysis difficult. Cells derived from BFU-Es could be harvested at ten days in a proliferative state before extensive hemoglobinization had occurred, or later if larger numbers of nondividing nucleated red blood cells were desired . Colonies were harvested onto 5 mm sites on poly-L-lysine coated slides, with four to six sites per slide. Only one colony was placed on each site except in cases where a large number of colonies arose in culture, when three or more small colonies were sometimes harvested per site. Because of differences in cell growth between samples, the total number of colonies harvested per sample ianged from three to 314.

Colonies were identified under an inverted microscope and plucked from the medium with a Gilson Pipetman (Rainin Instrument Co., Inc., Woburn, MA) set to 4 μl, using a 10 μl tip. Each colony was dispersed by pipetting up and down 10 to 20 times into a 15 μl drop of IMDM containing 0.1 μg/ml colcemid (Gibco/BRL) on a poly-L-lysine coated slide. Slides were placed flat in a humid chamber at 37 °C for 15 minutes, then 20 μl of 0.075 M KC1 hypotonic solution was gently infused into each drop. After a further 15 minutes of incubation at 37 °C in a humid box, 40 μl of a solution of 30 % 3:1 methanol: glacial acetic acid fixative/70 % 0.075M KC1 was added gently to each site. The slides were incubated for five to ten minutes at room temperature, then carefully tipped sideways to drain excess fluid without disturbing the cells which had adhered to the slide.

Before the slides dried, fresh 3:1 methanol: glacial acetic acid fixative was dropped onto each slide from a pasteur pipette and the wet slides were blown on to facilitate spreading out and flattening of the cells thereon. After the slides had air dried they were soaked for five minutes in Optistain II-A (Gam Rad Corp.,Capistrano, CA) followed by ten minutes in 3:1 methanol:glacial acetic acid fixative. Slides were either dehydrated through an ethanol series and either used immediately for further analysis (e.g., as described below for fluorescence in situ hybridization), or were stored frozen at -20 °C until they could be studied.

Erythroid cells on microscope slides prepared in this manner can be used for hybridization of the cells to a nucleic acid probe (e.g., fluorescent in situ hybridization), Wright's staining, banding or storage. Erythroid cells not arrested at the metaphase stage were prepared on microscope slides in the same manner by omitting colcemid from the cell- containing solution.

EXAMPLE 3: Detection of Fetal DNA in Cultured Erythroid Cells by Fluorescent In Situ Hybridization

In this example, flourescent in situ hybridization (FISH) was used to detect fetal DNA within the cultured, isolated erythroid progenitor cells. Hybridization of erythroid cell-derived DNA was performed simultaneously with both a Y chromosome-specific probe and an X-chromosome specific probe. The two probes were labeled with different fluorescent markers, allowing them to be distinguished by different colors. Since Y-specific sequences in erythroid cell-derived DNA must be of fetal origin, hybridization of this probe to the erythroid cell-derived DNA is used as an indicator of the fetal origin of the cells.

Fluorescent In Situ Hybridization

Erythroid progenitor cells were cultured as described in Example 1 and colonies were isolated and placed onto microscope slides as described in Example 2. FISH was performed using an X-chromosome specific probe derived from a cosmid containing a pericentromeric repeat sequence of the X chromosome and a Y specific probe derived from a cosmid pDP97 containing repetitive sequences. The probes are described in further detail in Klinger et al. Am. J. Hum. Genet. 51, 55-65 (1992). The X-specific probe was labeled with biotin whereas the Y-specific probe was labeled with digoxigenin. The biotinylated X chromosome probe was used at a concentration of 2.5 ng/μl and the digoxygeninlabelled Y chromosome probe was used at a concentration of 5 ng/μl.

The hybridization, washing, and detection protocols used were modified from Klinger et al. Am. J. Hum. Genet. 51, 55-65 (1992)and were as follows. Immediately before hybridization the slides were dehydrated through a 70%-80%-90%-100% ethanol series for one minute each and then air dried. Probe DNA for both the X and Y chromosomes was suspended in 50 % formamide/10 % dextran sulfate/6X SSC (standard sodium citrate) with 8 μg/μl sonicated salmon sperm DNA (Sigma Chemical Co.) and 2 μg/μl human Cot-1 DNA (Gibco BRL). A 2.5 μl drop of probe cocktail was added to each 5 mm site on the slides and covered with a glass coverslip. The coverslips were not sealed to the slide surfaces. The probe and the target nuclear DNA on glass slides were codenatured by placing the slides on a slidewarmer at 80 °C for five to six minutes. The slides were incubated in a humid box at 37 °C overnight. The following day the coverslips were removed and the slides were washed once in 50 % formamide/2X SSC pH 7.0 at 42 °C for ten minutes, and once in 0.1X SSC at 60 °C for ten minutes. Following a blocking step in 3 %BSA/4X SSC/0.1 % Tween 20 (polyoxyethylene-sorbitan monolaurate; Sigma Chemical Co.) at 42 °C for five minutes, the labeled probes hybridized to the nuclei on the slides were detected by incubation at 37 °C for 15 minutes in 1 % BSA/4X SSC with μg/ml anti-digoxygenin fluorescein isothiocyanate (FITC) antibody (Boehringer Mannheim Corp., Indianapolis, IN), which recognizes digoxygenin and emits green fluorescence, and 2 μg/ml Cy-3 streptavidin (Jackson

ImmunoResearch Labs, Inc., West Grove, PA), which binds tightly to biotin-labeled probe and produces red fluorescence. The slides were then washed once in 4X SSC/ 0.1 % Tween- 20 at 42 °C for ten minutes to remove excess unbound fluorochrome which might otherwise result in spurious signal background. Coverslips were placed on the slides atop a thin layer of 90 % glycerol containing 2.33 % DABCO antifade (l,4-diazabicyclo-[2.2.2]octane; Sigma Chemical Co.), 100 mM Tris pH 8.0, and 0.1 μg/ml DAPI (4,6-diamidino-2phenyl-indole; Sigma Chemical Co.). The fluorochrome DAPI emits blue fluorescence under ultraviolet light when bound to AT-rich regions of the minor groove of intact DNA, and is therefore useful as a nuclear counterstain (Schweizer, Ambrose, & Andrle, 1978). Slides were viewed with 40X and

100X oil immersion lenses on a Zeiss Axioplan epifluorescence microscope (Carl Zeiss, Inc.) with appropriate fluorescence filters (Omega Optical, Inc., Brattleboro, VT). Filters used for DAPI fluorescence were a 365 nm excitation filter, a 410 nm dichroic filter, and a 420LP nm emission filter; for FITC fluorescence, a 485/20 nm excitation filter, a 510 nm dichroic filter, and a 520-560 nm emission filter; and for Cy-3 fluorescence, a 515-560 nm excitation filter, a 580 nm dichroic filter, and a 590LP nm emission filter. Hybridized nuclei were photographed using a 100X oil immersion objective, with a 35 mm camera mounted directly on the microscope using Kodak Gold 400 film.

Maternal Blood Sample Analysis

A total of 37 peripheral blood samples were obtained from consenting pregnant women between 10 5/7 weeks and 22 weeks of gestation. Cells were fractionated from whole blood using either a dual Histopaque 1.077/1.119 g/ml density gradient (21 samples) or a Polymorphprep gradient (14 samples). Two samples (samples 11 and 21) were fractionated on dual Histopaque, Polymorphprep, Ficoll-Paque and dual Nycodenz gradients. Fractionated cells were cultured with erythropoietin as described in Example 1. In most cases, both the HI or PI and the H2 or P2 cell fractions were cultured. Colonies of CFU-E morphology were harvested from the H2 or P2 fractions, and occasionally were obtained from the HI or PI fraction as well. Colonies of BFU-E morphology which exhibited rapid growth and hemoglobinization were also harvested. Colonies were analyzed by FISH as described above. The actual karyotype of the fetuses, as determined by cytogenetic analysis, is shown in the last column (46,XY=Male, 46,XX=Female). The results are summarized in Table 2.

Of the 21 samples cultured from pregnant women whose fetuses were verified to be male by ultrasound and/or cytogenetic analysis (including sample 17, with one male and one female twin), XY colonies were identified by FISH in seven samples, yielding a male detection rate of 33.3%. The number of XY colonies per sample ranged from one to eight (2.1 % to 33.3% of the total number of colonies analyzed for those cases). From the 16 pregnant women who were actually carrying female fetuses as confirmed by cytogenetic analysis (not including sample 17, with a male and a female twin), no XY colonies were detected, yielding a false-positive rate of 0 %. The gestational ages of the 37 pregnancies analyzed varied from 10 5/7 weeks to 22 weeks. Sample 1, at 10 5/7 weeks, was the earliest pregnancy with a male fetus to be analyzed, and showed male cells in four of 120 colonies. The latest gestational age at which a male colony was cultured from a sample was 17 weeks (sample 18).

Table 3 relates the gestational age (GA) in weeks at which blood samples were obtained to the gradient fraction from which cultured colonies were harvested, the length of time (in days) that cells were cultured before harvesting, and the number and percent of the total number of colonies harvested which showed male cells by FISH.

Table 3: Harvesting Parameters of Cultured Cells from Pregnancies with a Male Fetus

Sample GA Fraction Davs # Colonies #Male % Male 1 105/7 H2 4 120 4 (3.3)

2 14 H2 7 1 0 (0) H2 10 4 0 (0)

14 P2 9 9 0 (0) PI 9 30 0 (0)

4 14 HI 8 10 0 (0) 5 143/7 H2 3 64 3 (4.7) 6 145/7 H2 4 96 2 (2.1) 7 146/7 HI 13 16 0 (0) H2 13 12 1 (8.3)

8 15 H2 5 20 0 (0) 9 15 P2 6 16 0 (0) P2 13 16 0 (0)

10 15 P2 6 16 0 (0) P2 13 8 0 (0) 11 15 Fl 8 30 0 (0) PI 8 24 0 (0) Nl 8 24 0 (0)

12 15 PI 7 24 0 (0)

13 15 H2 9 10 0 (0) HI 9 20 0 (0)

14 15 P2 9 10 2 (20.0) PI 9 35 1 (2.9)

15 15 H2 7 5 0 (0) HI 7 15 0 (0)

16 16 H2 7 10 0 (0) HI 7 10 0 (0)

17 162/7 H2 5 116 8 (6.9) H2 11 72 0 (0)

18 17 H2 4 3 1 (33.3) 19 17 P2 8 20 0 (0) 20 17 H2 7 15 0 (0) HI 7 10 0 (0)

21 22 Fl 7 28 0 (0) PI 7 31 0 (0) P2 7 12 0 (0) Nl 7 24 0 (0) HI 7 24 0 (0) H2 7 4 0 (0) O 95/26417 PCΪYUS95/03906

- 28 -

EXAMPLE 4: Use of the Polymerase Chain Reaction to Amplify Gene

Sequences in Cultured Erythroid Progenitor Cells

In this example, the polymerase chain reaction was used to amplify Y-specific DNA sequences present in fetal erythroid progenitor cells cultured from maternal peripheral blood. The polymerase chain reaction, or PCR, which has a capacity for making 10" copies of rare target gene sequences, is used to amplify gene sequences in cultured erythroid progenitor cells. For fetal sex determination, PCR primers specific for repeated sequences present on the Y chromosome are used. Repeated sequences are selected because they create a stronger amplification signal from a rare male fetal cell. Primers which hybridize to a region of the short arm of the Y chromosome, encompassed by probe Y411 (Given by Dr. Ulrich Muller, Children's Hospital, Boston, MA), are used. Y411 is identical to Y156 (Muller, U., et al., Nucleic Acids Res., 14, 1325-1329 (1986)), is repeated 10-60 fold, and is absolutely Y specific on Southern blots. Two Y411 region-specific primers, primers 411-01 and 411-03, which are designed to amplify a 222 base pair (bp) sequence detectable with the Y chromosome-specific probe Y411, are used. The primer sequences are described in detail in Bianchi, D.W., et al. Proc. Natl. Acad. Sci. USA 87, 3279-3283 (1990). Other suitable primers are described in Wachtel, S., et al. Hum. Reprod. 6, 1466-1469 (1991).

To define the minimum amount of DNA detectable in a cell sample, a series of standardization experiments is done. DNA from male and female individuals is prepared in tenfold dilutions (1 pg to 1 mg) and amplified using the standard reagents, including reaction buffer, in the GeneAmpkit (Perkin-Elmer Cetus catalogue #N801-0055) on a Perkin-Elmer DNA Thermal Cycler. Measures are taken to prevent aerosol contamination of samples with male DNA. All PCRs are performed under sterile conditions, wearing gloves, and using positive displacement pipettes. All reagents are prepared in a sterile manner and incubated overnight prior to PCR with a restriction endonuclease having a digestion site within the target sequence. These precautions result in a significant decrease and virtual absence of false positive amplification, as monitored by running control reactions with all reagents but no DNA. The number of amplification cycles is varied between 18 and 30. Each amplification cycle consists of 1 minute at 94°C, 2 minutes at 60°C and 3 minutes at 72°C, with a 10 minute extension in the last cycle. Amplified DNA samples are electrophoresed on agarose gels, transferred to nylon filters, and hybridized to 32 P-labelled Y411 probe.

Erythroid progenitor cells are cultured and isolated as described in Example 1.

Prior to amplification, the cells are lysed by boiling which makes the erythroid cell DNA available for amplification. The erythroid cell DNA is amplified for the 222 bp sequence in probe Y411 as a demonstration that the cells are derived from the fetus in male pregnancies. The conditions used, as described above, make it possible to detect a minimum of 100 pg of fetal DNA, or the equivalent of 15 fetal cells. The limit of sensitivity can be improved by extending the number of cycles used in PCR. To further decrease false positive amplification and permit detection of fetal DNA at the single cell level on agarose gels, PCR is carried out using primers derived from a single copy of sequence specific for the long arm of the Y chromosome, PY49a (Guerin, P., et al., Nucleic Acids Res. , 16, 7759 (1988)).

EXAMPLE 5: Detection of Chromosomal Abnormalities in Erythroid

Progenitor Cells Cultured from Maternal Blood

In this example, in situ hybridization using chromosome-specific probe sets is performed on fetal erythroid progenitor cells cultured from maternal blood in order to detect chromosomal abnormalities in fetal cells. A DNA probe set specific for a particular chromosome that provides both good signal to noise ratios and good spatial resolution of the fluorescent signals is used in in situ hybridization. Specific probe sets have been developed for five chromosomes frequently seen as liveborn aneuploidies, chromosomes 13, 18, 21, X and Y. A probe for chromosome 1 is used as a control. In constructing the probes, the general strategy was to identify a starting clone that mapped to the desired chromosomal region by multiple genetic and physical methods, and then to use that clone to identify matching cosmid "contigs" which are then used as hybridization probes. The chromosome 21 probe set is a three-cocmid contig containing 80 kb of nonoverlapping DNA. The chromosome 18 probe set is a three-cosmid contig containing 109 kb of nonoverlapping DNA. The chromosome 13 probe set is a three-cosmid contig containing approximately 97 kb of nonoverlapping DNA. The X chromosome probe is a cosmid containing a pericentromeric repeat sequence. The Y probe is pDP97, a repetitive clone (a 5.3 kb EcoRI Y fragment from cosmid Y97 subcloned into EcoRI site of pUC-13). All the probes are described in further detail in Klinger et al. Am. J. Hum. Genet. 51, 55-65 (1992).

To diagnose a chromosomal abnormality in a fetus, erythroid progenitor cells are cultured from maternal blood as described in Example 1 and chromosomal abnormalities are assessed by performing in situ hybridization using chromosome-specific probes such as those described above. In situ hybridization is performed as described in Example 3 and in Klinger et al. Am. J. Hum. Genet. 51, 55-65 (1992) under suppression conditions (Cremer et al., Hum Genet. 80, 235-246 (1988); Lichter et al., Hum Genet. 80, 224-234 (1988)). Hybridization of high copy number repeat sequences is suppressed by inclusion of total genomic human DNA, and the chromosomal specificity can be verified by hybridization to metaphase spreads. Probes are labelled with biotin-UTP, hybridized under suppression conditions, and specific hybridization detected by conjugated streptoavidin-FITC, which shows as a single "dot" in the FITC image upon microscopic analysis. Alternatively, if in situ hybridization is to be performed with multiple probes, the individual probes can be differentially labelled to allow for them to be distinguished fluorescently. For example, one probe can be labelled with biotin-UTP and the another with digoxigenin-UTP. The former probe can be detected with Cy-3 streptavidin while the latter can be detected with anti- digoxigenin-FITC. The probes give sharp, punctate fluorescent signals in interphase cells that are easily discriminated and enumerated.

Diagnosis of a chromosomal abnormality is accomplished by comparing the hybridization of a chromosome-specific probe to DNA from fetal erythroid cells to hybridization of the same probe to DNA from normal cells (a normal control). Normal cells, which may be any cells which do not contain a chromosomal abnormality in the chromosome(s) being examined, can be hybridized at the same time as the fetal erythroid cells to provide a normal control. For example, maternal cells can be used as a normal control. Alternatively, a previously established normal control can be used for comparison. A common chromosomal abnormality, a trisomy (in which three copies of a particular chromosome are present in a cell rather that the normal two), can be diagnosed by detection of three fluorescent signals for a particular chromosome, e.g. commonly chromosomes 21, 18 or 13, in a fetal erythroid cell as compared to only two fluorescent signals for the same chromosome in a normal control. A sufficient number of hybridized cells are examined to make a statistically significant determination of the number of fluorescent signals present per cell.

EQUIVALENTS

Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the following claims.

Claims (49)

1. A method for detecting a nucleic acid sequence of interest in fetal DNA of fetal cells in a sample of peripheral blood obtained from a pregnant woman, comprising:

obtaining a sample of peripheral blood from a pregnant woman;

culturing cells within the sample of peripheral blood in a culture medium containing a cell growth factor;

isolating fetal cells from the culture medium; and

detecting a nucleic acid sequence of interest in fetal DNA of fetal cells isolated from the culture medium.

2. The method of claim 1 , wherein the cell growth factor is a hemopoietic cell growth factor.

3. The method of claim 1, wherein the cell growth factor is an erythroid growth factor.

4. A method for detecting a nucleic acid sequence of interest in fetal DNA of fetal erythroid cells in a sample of peripheral blood obtained from a pregnant woman, comprising:

obtaining a sample of peripheral blood from a pregnant woman;

culturing cells within the sample of peripheral blood in a culture medium containing erythropoietin;

isolating fetal erythroid cells from the culture medium; and

detecting a nucleic acid sequence of interest in fetal DNA of fetal erythroid cells isolated from the culture medium.

5. The method of claim 4, wherein the nucleic acid sequence of interest is a Y chromosomal DNA sequence.

6. The method of claim 4, wherein the nucleic acid sequence of interest is a sequence of a gene associated with a disease-causing mutation.

7. The method of claim 4, wherein the nucleic acid sequence of interest detects a chromosomal abnormality.

8. A method for isolating fetal cells from a sample of peripheral blood obtained from a pregnant woman, comprising:

obtaining a sample of peripheral blood from a pregnant woman;

culturing cells within the sample of peripheral blood in a culture medium containing a cell growth factor; and

isolating fetal cells from the culture medium.

9. The method of claim 8, wherein the cell growth factor is a hemopoietic cell growth factor.

10. The method of claim 8, wherein the cell growth factor is an erythroid growth factor.

11. A method for isolating fetal erythroid cells from a sample of peripheral blood obtained from a pregnant woman, comprising:

obtaining a sample of peripheral blood from a pregnant woman;

culturing cells within the sample of peripheral blood in a culture medium containing erythropoietin; and

isolating fetal erythroid cells from the culture medium.

12. The method of claim 11, wherein a proportion of nucleated cells present in the sample of peripheral blood obtained from the pregnant woman is increased relative to a proportion of nucleated cells present in the sample of peripheral blood prior to enrichment forming a sample enriched in nucleated cells prior to culturing the sample enriched in nucleated cells.

13. The method of claim 12, wherein the sample enriched in nucleated cells is formed by separating non-nucleated cells from nucleated cells in the peripheral blood sample forming a sample enriched in nucleated cells.

14. The method of claim 13, wherein non-nucleated cells are separated from nucleated cells by density gradient centrifugation.

15. The method of claim 14, wherein density gradient centrifugation is performed with Ficoll.

16. The method of claim 14, wherein density gradient centrifugation is performed with Histopaque.

17. The method of claim 14, wherein density gradient centrifugation is performed with Nycodenz.

18. The method of claim 14, wherein density gradient centrifugation is performed with Polymorphprep.

19. The method of claim 13, wherein non-nucleated cells are separated from nucleated cells by selective lysis of non-nucleated cells.

20. The method of claim 11, wherein a proportion of erythroid progenitor cells present in the sample of peripheral blood obtained from the pregnant woman is increased relative to a proportion of erythroid progenitor cells present in the sample of peripheral blood prior to enrichment forming a sample enriched in erythroid progenitor cells prior to culturing the sample enriched in erv throid progenitor cells.

21. The method of claim 20, wherein the sample enriched in erythroid progenitor cells is formed by separating erythroid progenitor cells from non-nucleated cells and other nucleated cells in the sample of peripheral blood forming a sample enriched in erythroid progenitor cells.

22. The method of claim 21, wherein erythroid progenitor cells are separated from non- nucleated cells and other nucleated cells in the sample of peripheral blood by density gradient centrifugation.

23. The method of claim 22, wherein density gradient centrifugation is performed with a dual density gradient.

24. The method of claim 22, wherein density gradient centrifugation is performed with Ficoll.

25. The method of claim 22, wherein density gradient centrifugation is performed with Histopaque.

26. The method of claim 22, wherein density gradient centrifugation is performed with Nycodenz.

27. The method of claim 22, wherein density gradient centrifugation is performed with Polymorphprep.

28. The method of claim 11, wherein the culture media contains a semisolid matrix material.

29. The method of claim 28, wherein the semisolid matrix material allows cells to attach to the semisolid matrix material.

30. The method of claim 29, wherein the semisolid matrix material is methylcellulose.

31. The method of claim 29, wherein the semisolid matrix material is agargel or a plasma clot.

32. The method of claim 11 further comprising detecting a fetal cell marker on or in the fetal erythroid cells to identify them as being fetal-derived.

33. The method of claim 32, wherein the fetal cell marker is fetal hemoglobin.

34. The method of claim 32, wherein the fetal cell marker is a fetal histone protein Hl°.

35. The method of claim 32, wherein the fetal cell marker is a fetal antigenic determinant.

36. The method of claim 35, wherein the fetal antigenic determinant is an antigenic determinant i.

37. The method of claim 11, wherein fetal erythroid cells are isolated from the culture medium by picking one or more colonies of fetal erythroid cells from the culture media.

38. The method of claim 37, wherein the fetal erythroid cells are further isolated by transferring one or more colonies of erythroid cells from the culture medium to a microscope slide.

39. The method of claim 11 further comprising detecting a nucleic acid sequence of interest in fetal nucleic acid of isolated fetal erythroid cells.

40. The method of claim 39, wherein the nucleic acid sequence of interest is a Y chromosomal DNA sequence.

41. The method of claim 39, wherein the nucleic acid sequence of interest is a sequence of a gene associated with a disease-causing mutation.

42. The method of claim 39, wherein the nucleic acid sequence of interest detects a chromosomal abnormality.

43. A method for isolating fetal erythroid cells in metaphase from a sample of peripheral blood from a pregnant woman, comprising:

obtaining a sample of peripheral blood from a pregnant woman;

culturing cells within the sample of peripheral blood in a culture media containing erythropoietin;

exposing cultured cells to an agent which inhibits progression of dividing cells through the cell cycle or synchronizes growth of cells; and

isolating fetal erythroid cells in metaphase.

44. The method of claim 43, wherein the agent which inhibits progression of dividing cells through the cell cycle is colcemid.

45. The method of claim 43, wherein the agent which inhibits progression of dividing cells through the cell cycle is colchicine.

46. The method of claim 43, wherein the agent which inhibits progression of dividing cells through the cell cycle is a vinblastine salt.

47. The method of claim 43, wherein the agent which synchronizes growth of cells is bromodeoxyuridine or fluorodeoxyuridine.

48. The method of claim 43, wherein the agent which synchronizes growth of cells is ethidium bromide.

49. A method for preferentially isolating fetal cells from a current pregnancy in a peripheral blood sample from a multiparous pregnant woman, comprising: obtaining a sample of peripheral blood from a multiparous pregnant woman;

culturing cells within the sample of peripheral blood in a culture media containing erythropoietin; and

isolating fetal erythroid cells to preferentially isolate fetal cells from a current pregnancy.