Aberrantly methylated genes as markers of breast malignancy

FIELD OF THE INVENTION

[0002] The present invention relates generally to a method of diagnosing a cell proliferative disorder of breast tissue by determining the DNA methylation status of nucleic acids obtained a subject.

BACKGROUND

[0003] Methylation has been shown by several lines of evidence to play a role in gene activity, cell differentiation, tumorigenesis, X-chromosome inactivation, genomic imprinting and other major biological processes (Razin, A., H., and Riggs, R. D. eds. in DNA Methylation Biochemistry and Biological Significance, Springer-Verlag, New York, 1984). In eukaryotic cells, methylation of cytosine residues that are immediately 5′ to a guanosine, occurs predominantly in cytosine-guanine (CG) poor regions (Bird, Nature, 321:209, 1986). In contrast, CpG islands remain unmethylated in normal cells, except during X-chromosome inactivation (Migeon, et al., supra) and parental specific imprinting (Li, et al., Nature, 366:362, 1993) where methylation of 5′ regulatory regions can lead to transcriptional repression. De novo methylation of the Rb gene has been demonstrated in a small fraction of retinoblastomas (Sakai, et al., Am. J. Hum. Genet., 48:880, 1991), and recently, a more detailed analysis of the VHL gene showed aberrant methylation in a subset of sporadic renal cell carcinomas (Herman, et al., Proc. Natl. Acad. Sci., U.S.A., 91:9700, 1994). Expression of a tumor suppressor gene can also be abolished by de novo DNA methylation of a normally unmethylated CpG island (Issa, et al., Nature Genet., 7:536, 1994; Herman, et al., supra; Merlo, et al., Nature Med., 1:686, 1995; Herman, et al., Cancer Res., 56:722, 1996; Graff, et al., Cancer Res., 55:5195, 1995; Herman, et al., Cancer Res., 55:4525, 1995).

[0004] Human cancer cells typically contain somatically altered nucleic acid, characterized by mutation, amplification, or deletion of critical genes. In addition, the nucleic acid from human cancer cells often displays somatic changes in DNA methylation (Fearon, et al., Cell, 61:759, 1990; Jones, et al., Cancer Res., 46:461, 1986; Holliday, Science, 238:163, 1987; De Bustros, et al., Proc. Natl. Acad. Sci., USA, 85:5693, 1988); Jones, et al., Adv. Cancer Res., 54:1, 1990; Baylin, et al., Cancer Cells, 3:383, 1991; Makos, et al., Proc. Natl. Acad. Sci., USA, 89:1929, 1992; Ohtani-Fujita, et al., Onco-gene, 8:1063, 1993). However, the precise role of abnormal DNA methylation in human tumorigenesis has not been established.

[0005] Aberrant methylation of normally umnethylated CpG islands has been described as a frequent event in immortalized and transformed cells, and has been associated with transcriptional inactivation of defined tumor suppressor genes in human cancers. This molecular defect has also been described in association with various cancers. CpG islands are short sequences rich in the CpG dinucleotide and can be found in the 5′ region of about half of all human genes. Methylation of cytosine within 5′ CGIs is associated with loss of gene expression and has been seen in physiological conditions such as X chromosome inactivation and genomic imprinting. Aberrant methylation of CpG islands has been detected in genetic diseases such as the fragile-X syndrome, in aging cells and in neoplasia. About half of the tumor suppressor genes which have been shown to be mutated in the germline of patients with familial cancer syndromes have also been shown to be aberrantly methylated in some proportion of sporadic cancers, including Rb, VHL, p16, hMLH1, and BRCA1 (reviewed in Baylin, et al, Adv. Cancer Res. 72:141-196 1998). Methylation of tumor suppressor genes in cancer is usually associated with (1) lack of gene transcription and (2) absence of coding region mutation. Thus CpG island methylation can serve as an alternative mechanism of gene inactivation in cancer.

[0006] Breast cancer is by far the most common form of cancer in women, and is the second leading cause of cancer death in humans. Despite many recent advances in diagnosing and treating breast cancer, the prevalence of this disease has been steadily rising at a rate of about 1% per year since 1940. Today, the likelihood that a woman living in North America will develop breast cancer during her lifetime is one in eight.

[0007] Breast cancer is often discovered at a stage that is advanced enough to severely limit therapeutic options and survival rates. Accordingly, more sensitive and reliable methods are needed to detect small (less than 2 cm diameter), early stage, in situ carcinomas of the breast. In addition to the problem of early detection, there remain serious problems in distinguishing between malignant and benign breast disease, in staging known breast cancers, and in differentiating between different types of breast cancers (e.g., estrogen dependent versus non-estrogen dependent tumors). Recent efforts to develop improved methods for breast cancer detection have focused on cancer markers such as proteins that are uniquely expressed (e.g., as a cell surface or secreted protein) by cancerous cells, or are expressed at measurably increased or decreased levels by cancerous cells compared to normal cells. Accordingly, the use of the methylation status of certain genes as a marker of cancer or cancerous conditions provides an additional weapon in early detection and prognosis of breast cancers.

[0008] Identification of the earliest genetic changes in cells associated with breast cancer is a major focus in molecular cancer research. Diagnostic approaches based on identification of these changes in specific genes are likely to allow implementation of early detection strategies and novel therapeutic approaches. Targeting these early changes might lead to more effective cancer treatment.

SUMMARY OF THE INVENTION

[0009] The present invention is based on the finding that several genes are newly identified as being differentially methylated in breast cancers. This seminal discovery is useful for breast cancer screening, risk-assessment, prognosis, disease identification, disease staging and identification of therapeutic targets. The identification of new genes that are methylated in breast cancer allows accurate and effective early diagnostic assays, methylation profiling using multiple genes; and identification of new targets for therapeutic intervention.

[0010] In a first embodiment, the invention provides method of diagnosing a cellular proliferative disorder of breast tissue in a subject comprising determining the state of methylation of one or more nucleic acids isolated from the subject. The state of methylation of one or more nucleic acids compared with the state of methylation of one or more nucleic acids from a subject not having the cellular proliferative disorder of breast tissue is indicative of a cell proliferative disorder in the subject. In one aspect of this embodiment, the state of methylation is hypermethylation. The invention provides a method of diagnosing a cellular proliferative disorder of breast tissue in a subject by detecting the state of methylation of one or more of the following nucleic acids: Twist, cyclin D2, WT1, NES-1 and HOXA5. Also methylated are RARβ2, 14.3.3 sigma and estrogen receptor. In one aspect of the invention, nucleic acids are methylated in regulatory regions.

[0011] Invention methods can be used to diagnose disorders of the breast including breast cancers. In one aspect of the invention, disorders of the breast include ductal carcinoma in situ, lobular carcinoma, colloid carcinoma, tubular carcinoma, medullary carcinoma, metaplastic carcinoma, intraductal carcinoma in situ, lobular carcinoma in situ and papillary carcinoma in situ.

[0012] Another embodiment of the invention provides a method of determining a predisposition to a cellular proliferative disorder of breast tissue in a subject. The method includes determining the state of methylation of one or more nucleic acids isolated from the subject, wherein the state of methylation of one or more nucleic acids compared with the state of methylation of the nucleic acid from a subject not having a predisposition to the cellular proliferative disorder of breast tissue is indicative of a cell proliferative disorder of breast tissue in the subject. The nucleic acids can be nucleic acids encoding Twist, cyclin D2, RARβ2, HOXA5, WT1, 14.3.3 sigma, estrogen receptor, NES-1, and combinations thereof;

[0013] Still another embodiment of the invention provides a method for detecting a cellular proliferative disorder of breast tissue in a subject. The method includes contacting a specimen containing at least one nucleic acid from the subject with an agent that provides a determination of the methylation state of at least one nucleic acid. The method further includes identifying the methylation states of at least one region of at least one nucleic acid, wherein the methylation state of the nucleic acid is different from the methylation state of the same region of nucleic acid in a subject not having the cellular proliferative disorder of breast tissue.

[0014] Yet a further embodiment of the invention provides a kit useful for the detection of a cellular proliferative disorder in a subject comprising carrier means compartmentalized to receive a sample therein; and one or more containers comprising a first container containing a reagent that modifies unmethylated cytosine and a second container containing primers for amplification of a CpG-containing nucleic acid. The primers hybridize with target polynucleotide sequence having the sequence of certain nucleic acids described herein.

SUMMARY OF THE FIGURES

[0015]
FIG. 1A shows the nucleotide sequence of the cyclin D2 promoter. Regions highlighted indicate primer sequences. CG nucleotide pairs are shown capitalized and bolded. A highlighted box shows the location of an atg codon. FIG. 1B shows nucleotide sequences for forward (F) and reverse (R) primer pairs used to detect methylated (M) and unmethylated (U) nucleic acids. The base pair (BP) length of the primer pair product is also indicated.

[0016]
FIGS. 2A and 2B show the nucleotide sequence of the TWIST promoter. Regions highlighted indicate primer sequences. CG nucleotide pairs are shown capitalized and bolded. A highlighted box shows the location of an atg codon. FIG. 2C shows nucleotide sequences for forward (F) and reverse (R) primer pairs used to detect methylated (M) and unmethylated (U) nucleic acids. The base pair (BP) length of the primer pair product is also indicated.

[0017]
FIG. 3A shows the nucleotide sequence of the Retinoic Acid Receptor Beta (RARβ) promoter. Regions highlighted indicate primer sequences. CG nucleotide pairs are shown capitalized and bolded. A highlighted box shows the location of an atg codon. FIG. 3B shows nucleotide sequences for forward (F) and reverse (R) primer pairs used to detect methylated (M) and unmethylated (U) nucleic acids. The base pair (BP) length of the primer pair product is also indicated.

[0018]
FIG. 4A shows the nucleotide sequence of Homo sapiens serine protease-like protease (NES-1) mRNA. FIG. 4B shows the nucleotide sequence of the NES-1 region (exon 3) analyzed. Regions highlighted indicate primer sequences. CG nucleotide pairs are shown capitalized and bolded. FIG. 4C shows nucleotide sequences for forward (F) and reverse (R) primer pairs used to detect methylated (M) and unmethylated (U) nucleic acids. The base pair (BP) length of the primer pair product is also indicated.

[0019]
FIG. 5A shows the nucleotide sequence of HOXA5 promoter (3′ to 5′). CG nucleotide pairs are shown capitalized and bolded. A highlighted box shows the location of a cat codon. FIG. 5B shows the nucleotide sequence of the complementary region (5′ to 3″) analyzed (nucleotides -97 to -303). Regions highlighted indicate primer sequences. CG nucleotide pairs are shown capitalized and bolded. Highlighted box shows an atg codon. FIG. 5C shows nucleotide sequences for forward (F) and reverse (R) primer pairs used to detect methylated (M) and unmethylated (U) nucleic acids. The base pair (BP) length of the primer pair product is also indicated. FIG. 5D shows forward and reverse (sense and antisense) primers used for sequencing and expression of HOXA5.

[0020]
FIGS. 6A to 6F show the nucleotide sequence of Homo sapiens 14.3.3 sigma protein promoter and gene, complete cds.

[0021]
FIGS. 7A and 7B show the nucleotide sequence of Homo sapiens Wilm's tumor (WT1) gene promoter.

[0022]
FIGS. 8A and 8B show the nucleotide sequence of homo sapiens estrogen receptor beta gene, promoter region and partial cds. CG nucleotide pairs are shown capitalized and bolded. FIG. 8C shows nucleotide sequences of forward (F) and reverse (R) primer pairs used to detect methylated (M) and unmethylated (U) nucleic acids. The base pair (BP) length of the primer pair producet is also indicated.

DETAILED DESCRIPTION OF THE INVENTION

[0023] The invention is based upon the discovery that the hypermethylation of certain genes can serve as markers for cellular proliferative disorders of breast tissue. This is the first time that promoter hypermethylation of certain genes such as Twist, cyclin D2, RARβ2, WT1, NES-1 and HOXA5 have been associated with breast cancer.

[0024] It has been determined that the methylation state of nucleic acids of certain genes, particularly regulatory sequences, is diagnostic for the presence or potential development of a cellular proliferative disorder of breast tissue in subjects. More particularly, the hypermethylation of certain nucleotides localized in CpG islands has been shown to affect the expression of genes associated with the CpG islands; typically such hypermethylated genes have reduced or abolished expression, primarily due to down-regulated transcription. Hypermethylation of, for example, Twist, cyclin D2, retinoic acid receptor β (RARβ), WT1, HOXA5, 14.3.3 sigma, estrogen receptor (ER) and NES-1, allows one to diagnose a cellular proliferative disorder of breast tissue. Using a recently developed PCR-based technique called methylated specific PCR (MSP), aberrantly methylated nucleic acids in breast cancer primary tumors and biological samples from individuals with breast cancer can be identified.

[0025] In a first embodiment, the invention provides a method of diagnosing a cellular proliferative disorder of breast tissue in a subject comprising determining the state of methylation of one or more nucleic acids isolated from the subject, wherein the state of methylation of one or more nucleic acids as compared with the state of methylation of one or more nucleic acids from a subject not having the cellular proliferative disorder of breast tissue is indicative of a cellular proliferative disorder of breast tissue in the subject.

[0026] A cell proliferative disorder as described herein may be a neoplasm. Such neoplasms are either benign or malignant. The term “neoplasm” refers to a new, abnormal growth of cells or a growth of abnormal cells that reproduce faster than normal. A neoplasm creates an unstructured mass (a tumor) which can be either benign or malignant. The term “benign” refers to a tumor that is noncancerous, e.g. its cells do not invade surrounding tissues or metastasize to distant sites. The term “malignant” refers to a tumor that is metastastic, invades contiguous tissue or no longer under normal cellular growth control.

[0027] One cellular proliferative disorder is a cell proliferative disorder of breast tissue. Disorders of breast tissue or breast cancers can relate to numerous cells and tissues resulting in various disorders of the breast including ductal carcinoma in situ, lobular carcinoma, colloid carcinoma, tubular carcinoma, medullary carcinoma, metaplastic carcinoma, intraductal carcinoma in situ, lobular carcinoma in situ, and papillary carcinoma in situ.

[0028] The invention method includes determining the state of methylation of one or more nucleic acids isolated from the subject. The phrases “nucleic acid” or “nucleic acid sequence” as used herein refer to an oligonucleotide, nucleotide, polynucleotide, or to a fragment of any of these, to DNA or RNA of genomic or synthetic origin which may be single-stranded or double-stranded and may represent a sense or antisense strand, peptide nucleic acid (PNA), or to any DNA-like or RNA-like material, natural or synthetic in origin. As will be understood by those of skill in the art, when the nucleic acid is RNA, the deoxynucleotides A, G, C, and T are replaced by ribonucleotides A, G, C, and U, respectively.

[0029] The nucleic acid of interest can be any nucleic acid where it is desirable to detect the presence of a differentially methylated CpG island. The CpG island is a CpG rich region of a nucleic acid sequence. The nucleic acids includes, for example, a sequence encoding the following genes (GenBank Accession Numbers are shown, followed by the nucleotides corresponding to the region(s) examined for the presence or absence of methylation (numbers are relative to the first ATG codon unless otherwise indicated)): Twist (Accession No. AC003986; -51145 to 151750 (complement)), cyclin D2 (Accession No. U47284; -1616 to -1394), RARβ2 (Accession No. AF; 157484; -196 to -357), WT1 (Accession No. AB034940), HOXA5 (Accession No. AC004080), 14.3.3 sigma (Accession No. AF029081), estrogen receptor (ER; Accession No. X62462), and NES-1 (Accession No. AF024605) (each Accession Number nucleotide sequence incorporated by reference herein).

[0030] WT1 encodes a transcriptional regulatory protein that binds DNA via four Cys2-His2 zinc fingers. WT1 mRNA undergoes two independent splicing events leading to the expression of at least four predominant isoforms. These splices result in the inclusion or omission of exon 5 (51 base pairs) and the presence or absence of a nine base pair insert (encoding three amino acids, KTS) between the third and fourth zinc finger domains. Lack of expression has been observed in some Wilms' tumors, leading to classification as a tumor suppressor gene. However, WT1 is overexpressed in 75% of cases of acute leukemia and is upregulated as chronic myeloid leukemia progresses into blast crisis. Thus, WT1 can apparently be either a tumor suppressor or an oncogene.

[0031] The cyclin D1, D2 and D3 proteins are involved in regulation of the cell cycle through phosphorylation and inactivation of the retinoblastoma protein and activation of cyclin E, leading to transition of the cells from G1 to DNA synthesis. In addition to their role in cell cycle regulation, the D-type cyclins have been implicated in differentiation and neoplastic transformation. Overexpression of cyclin D2 has been reported in gastric cancer, and was shown to correlate with disease progression and poor prognosis. Overexpression of cyclin D2 is also noted in ovarian granulosa cell tumors and testicular germ cell tumor cell lines.

[0032] 14.3.3 σ is a member of a superfamily that is responsible for instituting the G2 cell cycle checkpoint in response to DNA damage in human cells (Hermeking, et al. (1997) Mol. Cell 1, 3-11; Chan, et al.. (1999) Nature 401, 616-20). In addition to any growth advantage resulting from a loosening of this checkpoint control mechanism, loss of σ function is predicted to cause an increase in DNA damage in response to γ-irradiation. Loss of 14.3.3. σ a in primary eptithelial cells leads to immortalization (Dellambra et al. (2000) J. Cell Biol., 149:1117-1130), one of the earliest steps towards cancer.

[0033] Retinoic acid (RA) controls fundamental developmental processes, induces terminal differentiation of myeloid progenitors and suppresses cancer and cell growth. RA activity is mediated by nuclear receptors, the retinoic acid receptors, RARs, that act as RA-dependent transcriptional activators in their heterodimeric forms with retinoid X receptors, RXRs (Chambon, 1996). RARs induce local chromatin changes at level of target genes, containing responsive RA elements (RAREs) by recruiting multiprotein complexes with histone acetyltransferase (HAT) activity and histone deacetylase (HDAC) activity, that dynamically pattern chromatin modification and regulate gene expression. RARs and RXRs, when disrupted, result in severe developmental defects and neoplastic transformation. In breast cancer cells, the expression of one member of the RARs family, RAR β is found consistently down-regulated or lost. RAR β downregulation can be reversed by RA in estrogen receptor (ER)-positive, but not in ER-negative breast carcinoma cell lines, believed to represent more advanced forms of tumors. Loss of RA-induced RAR β expression is considered a crucial step in the development of RA-resistance in breast carcinogenesis. A complex regulatory region, with two promoters, regulates RAR β gene expression. Only one promoter, RAR β2, containing several RA-response elements, including a canonical and an auxiliary RA response element, βRARE, is active in human mammary epithelial cells (HMEC). The transcription of the RAR β2 promoter is mediated by multiple RARs including, RAR a and RAR β itself able to recruit coactivator and corepressor protein complexes with HAT/HDAC activities, respectively.

[0034] The Twist gene product is a transcription factor with DNA binding and helix-loop-helix domains. Twist is a member of the bHLH transcription factor family and is involved in the development of mesodermally derived tissue includeding the skeleton. In humans, mutations in the Twist gene have been identified in patients with Saethre-chotzen syndrome, a relatively common craniosynostosis disorder with autosomal dominant inheritance. (see Gripp et al., (2000) Hum. Mutat. 15:479.) Twit also influences osteogenic gene expression and may act as a master switch in initiating bone cell differentiation by regulating the osteogenic cell lineage (Lee et al., (1999) J. Cell Biochem. 75:566-577).

[0035] NES 1 (normal epithelial cell-specific 1) is a novel gene with a predicted polypeptide of about 30.14 kilodaltons and having a 50-63% similarity and 34-42% identity with several families of serine proteases, in particular the trypsin-like proteases, members of the glandular kallikrein family (including prostate-specific antigen, nerve growth factor gamma, and epidermal growth factor-binding protein) and the activators for the kringle family proteins (including the human tissue plasminogen activator and human hepatocyte growth factor activator) (Liu et al., (1996) Cancer Res. 56:14 3371-9). All of the residues known to be crucial for substrate binding, specificity, and catalysis by the serine proteases are conserved in the predicted NES1 protein, indicating that it has protease-like activity. Immunolocalization studies with an antipeptide antibody directed against a unique region of the NES1 protein (amino acids 120-137) detect a specific 30-kilodalton polypeptide almost exclusively in the supernatant of the mRNA-positive mammary epithelial cells (MECs), suggesting that NES1 is a secreted protein. The 1.4-kb NES1 mRNA is expressed in several organs (thymus, prostate, testis, ovary, small intestine, colon, heart, lung, and pancreas) with highest levels in the ovarys. Although expression of the NES1 mRNA is observed in all normal and immortalized nontumorigenic MECs, the majority of human breast cancer cell lines show a drastic reduction or a complete lack of its expression. The structural similarity of NES1 to polypeptides known to regulate growth factor activity and a negative correlation of NES1 expression with breast oncogenesis suggest a direct or indirect role for this novel protease-like gene product in the suppression of tumorigenesis. Studies using fluorescence in situ hybridization localized the NES1 gene to chromosome 19q13.3, a region that contains genes for related proteases (Goyal et al., (1998) Cancer Res., 58:21 4782-6).

[0036] The HOX genes are expressed during embryonic development and have a role in specifying antero-posterior positional information. The genes are arranged in four clusters and a colinear relation exists between a gene's position in the cluster and its anterior boundary of expression. Genes with more anterior boundaries are also expressed earlier than genes with more posterior boundaries. Hox genes encode transcription factors; therefore, a model for the coordinate regulation of the genes within the HOX clusters is that Hox gene products regulate their own expression. The production of HOXA5 from an expression vector can activate a transient and simultaneous expression of other upstream and downstream genes of the same HOX cluster and genes from other clusters.

[0037] The estrogen receptor gene has been implicated in the initiation and/or progresion of human breast cancer. Loss of expression of either gene has been assoicated with poorly differentiated tumors and poorer prognosis. Several studies have reported an asooication between estrogen receptor (ER) expression and breast tumors. A loss of ER expression has been assoicated with aberrant 5′ CpG isalnd methuylation in breast cancer cell lines and primary human breast tumors. Studies show that aberrant methylation of ER CpG islands begins before invasion of tumors into surrounding tissues and it increases with metastatic progression (Naas et al., (2000) Cancer Res., 60:4346-4348; incorporated by reference in its entirety).

[0038] Any nucleic acid sample, in purified or nonpurified form, can be utilized in accordance with the present invention, provided it contains, or is suspected of containing, a nucleic acid sequence containing a target locus (e.g., CpG-containing nucleic acid). One nucleic acid region capable of being differentially methylated is a CpG island, a sequence of nucleic acid with an increased density relative to other nucleic acid regions of the dinucleotide CpG. The CpG doublet occurs in vertebrate DNA at only about 20% of the frequency that would be expected from the proportion of G&Circlesolid;C base pairs. In certain regions, the density of CpG doublets reaches the predicted value; it is increased by ten fold relative to the rest of the genome. CpG islands have an average G&Circlesolid;C content of about 60%, compared with the 40% average in bulk DNA. The islands take the form of stretches of DNA typically about one to two kilobases long. There are about 45,000 such islands in the human genome.

[0039] In many genes, the CpG islands begin just upstream of a promoter and extend downstream into the transcribed region. Methylation of a CpG island at a promoter usually prevents expression of the gene. The islands can also surround the 5′ region of the coding region of the gene as well as the 3′ region of the coding region. Thus, CpG islands can be found in multiple regions of a nucleic acid sequence including upstream of coding sequences in a regulatory region including a promoter region, in the coding regions (e.g., exons), downstream of coding regions in, for example, enhancer regions, and in introns.

[0040] In general, the CpG-containing nucleic acid is DNA. However, invention methods may employ, for example, samples that contain DNA, or DNA and RNA, including messenger RNA, wherein DNA or RNA may be single stranded or double stranded, or a DNA-RNA hybrid may be included in the sample. A mixture of nucleic acids may also be employed. The specific nucleic acid sequence to be detected may be a fraction of a larger molecule or can be present initially as a discrete molecule, so that the specific sequence constitutes the entire nucleic acid. It is not necessary that the sequence to be studied be present initially in a pure form; the nucleic acid may be a minor fraction of a complex mixture, such as contained in whole human DNA. The nucleic acid-containing sample used for determination of the state of methylation of nucleic acids contained in the sample or detection of methylated CpG islands may be extracted by a variety of techniques such as that described by Sambrook, et al. (Molecular Cloning: A Laboratory Manual, Cold Spring Harbor, N.Y., 1989; incorporated in its entirety herein by reference).

[0041] A nucleic acid can contain a regulatory region which is a region of DNA that encodes information that directs or controls transcription of the nucleic acid. Regulatory regions include at least one promoter. A “promoter” is a minimal sequence sufficient to direct transcription, to render promoter-dependent gene expression controllable for cell-type specific, tissue-specific, or inducible by external signals or agents. Promoters may be located in the 5′ or 3′ regions of the gene. Promoter regions, in whole or in part, of a number of nucleic acids can be examined for sites of CG-island methylation.

[0042] Nucleic acids isolated from a subject are obtained in a biological specimen from the subject. The nucleic acid may be isolated from breast tissue, blood, plasma serum, lymph, duct cells, nipple aspiration fluid, ductal lavage fluid and bone marrow. Tissue, blood, lymph, lymph node, duct cells, nipple aspiration fluid, ductal lavage fluid and bone marrow are obtained by various medical procedures known to those of skill in the art. Duct cells can be obtained by nipple aspiration, ducal lavage, sentinel node biopsy, fine needle aspirate, routine operative breast endoscopy and core biopsy. Ductal lavage fluid can be obtained by using a DucWash procedure. In the procedure, a catheter is inserted into one or more of the four to eight ducts typically present in each human breast, lavage of the duct is performed, and the lavage fluid is collected. Alternatively, ductal lavage may be achieved through a microcatheter procedure known as ROBE (routine operative breast endoscopy) which allows visualization of a tumor at the same time as aspiration of fluid from the duct.

[0043] In one aspect of the invention, the state of methylation in nucleic acids of the sample obtained from a subject is hypermethylation compared with the same regions of the nucleic acid in a subject not having the cellular proliferative disorder of breast tissue. Hypermethylation, as used herein, is the presence of methylated alleles in one or more nucleic acids. Nucleic acids from a subject not having a cellular proliferative disorder of breast tissues contains no detectable methylated alleles when the same nucleic acids are examined.

[0044] A method for determining the methylation state of nucleic acids is described in U.S. Pat. No. 6,017,704 which is incorporated herein in its entirety and described briefly herein. Determining the methylation state of the nucleic acid includes amplifying the nucleic acid by means of oligonucleotide primers that distinguishes between methylated and unmethylated nucleic acids.

[0045] Two or more markers can also be multiplexed in a single amplification reaction to generate a low cost, reliable cancer screening test for breast cancers simultaneously. A combination of DNA markers for CpG-rich regions of nucleic acid may be amplified in a single amplification reaction. The markers are multiplexed in a single amplification reaction, for example, by combining primers for more than one locus. For example, DNA from a ductal lavage sample can be amplified with two or more different unlabeled or randomly labeled primer sets in the same amplification reaction. Especially useful are two or more markers selected from cyclin D2, RARβ2, Twist and NES-1. The reaction products are separated on a denaturing polyacrylamide gel, for example, and then exposed to film or stained with ethidium bromide for visualization and analysis. By analyzing a panel of markers, there is a greater probability of producing a more useful methylation profile for a subject.

[0046] If the sample is impure (e.g., plasma, serum, lymph, ductal cells, nipple aspiration fluid, ductal lavage fluid, bone marrow, blood or breast tissue embedded in paraffin), it may be treated before amplification with a reagent effective for lysing the cells contained in the fluids, tissues, or animal cell membranes of the sample, and for exposing the nucleic acid(s) contained therein. Methods for purifying or partially purifying nucleic acid from a sample are well known in the art (e.g., Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, 1989, herein incorporated by reference).

[0047] Primers hybridize with target polynucleotide sequences. Nucleic acid sequences including exemplary primers are set forth in SEQ ID NO:1 to SEQ ID NO:89. Oligonucleotide primers specifically targeted to methylated and unmethylated genes including Twist, cyclinD2, RARβ2, WT1, HOXA5, 14.3.3 sigma, estrogen receptor, and NES-1, and their associated CpG islands include SEQ ID NO:7-14, 21-24, 37-40, 49-64, 69-72, 77-80 and 85-90.

[0048] Detection of differential methylation can be accomplished by contacting a nucleic acid sample with a methylation sensitive restriction endonuclease that cleaves only unmethylated CpG sites under conditions and for a time to allow cleavage of unmethylated nucleic acid. The sample is further contacted with an isoschizomer of the methylation sensitive restriction endonuclease, that cleaves both methylated and unmethylated CpG-sites, under conditions and for a time to allow cleavage of methylated nucleic acid. Oligonucleotides are added to the nucleic acid sample under conditions and for a time to allow ligation of the oligonucleotides to nucleic acid cleaved by the restriction endonuclease, and the digested nucleic acid is amplified by conventional methods such as PCR wherein primers complementary to the oligonucleotides are employed. Following identification, the methylated CpG-containing nucleic acid can be cloned, using method well known to one of skill in the art (see Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, 1989).

[0049] As used herein, a “methylation sensitive restriction endonuclease” is a restriction endonuclease that includes CG as part of its recognition site and has altered activity when the C is methylated as compared to when the C is not methylated. Preferably, the methylation sensitive restriction endonuclease has inhibited activity when the C is methylated (e.g., Smal). Specific non-limiting examples of a methylation sensitive restriction endonucleases include Sma I, BssHII, or HpaII, MspI, BSTUI, and NotI. Such enzymes can be used alone or in combination. Other methylation sensitive restriction endonucleases will be known to those of skill in the art and include, but are not limited to SacII, and EagI, for example. An “isoschizomer” of a methylation sensitive restriction endonuclease is a restriction endonuclease which recognizes the same recognition site as a methylation sensitive restriction endonuclease but which cleaves both methylated and unmethylated CGs. One of skill in the art can readily determine appropriate conditions for a restriction endonuclease to cleave a nucleic acid (see Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, 1989).

[0050] A nucleic acid of interest is cleaved with a methylation sensitive endonuclease. Cleavage with the methylation sensitive endonuclease creates a sufficient overhang on the nucleic acid of interest. Following cleavage with the isoschizomer, the cleavage product can still have a sufficient overhang. An “overhang” refers to nucleic acid having two strands wherein the strands end in such a manner that a few bases of one strand are not base paired to the other strand. A “sufficient overhang” refers to an overhang of sufficient length to allow specific hybridization of an oligonucleotide of interest. Sufficient overhang is at least two bases in length or four or more bases in length. An overhang of a specific sequence on the nucleic acid of interest may be desired in order for an oligonucleotide of interest to hybridize. In this case, the isoschizomer can be used to create the overhang having the desired sequence on the nucleic acid of interest.

[0051] Cleavage with a methylation sensitive endonuclease results in a reaction product of the nucleic acid of interest that has a blunt end or an insufficient overhang. Blunt ends refers to a flush ending of two stands, the sense stand and the antisense strand, of a nucleic acid. Once a sufficient overhang is created on the nucleic acid of interest, an oligonucleotide is ligated to the nucleic acid cleaved of interest which has been cleaved by the methylation specific restriction endonuclease. “Ligation” is the attachment of two nucleic acid sequences by base pairing of substantially complementary sequences and/or by the formation of covalent bonds between two nucleic acid sequences.

[0052] An adaptor can be utilized to create DNA ends of desired sequence and overhang. An “adaptor” is a double-stranded nucleic acid sequence with one end that has a sufficient single-stranded overhang at one or both ends such that the adaptor can be ligated by base-pairing to a sufficient overhang on a nucleic acid of interest that has been cleaved by a methylation sensitive restriction enzyme or an isoschizomer of a methylation sensitive restriction enzyme. Adaptors can be obtained commercially, or two oligonucleotides can be utilized to form an adaptor. Thus, two oligonucleotides can be used to form an adaptor; these oligonucleotides are substantially complementary over their entire sequence except for the region(s) at the 5′ and/or 3′ ends that will form a single stranded overhang. The single stranded overhang is complementary to an overhang on the nucleic acid cleaved by a methylation sensitive restriction enzyme or an isoschizomer of a methylation sensitive restriction enzyme, such that the overhang on the nucleic acid of interest will base pair with the 3′ or 5′ single stranded end of the adaptor under appropriate conditions. The conditions will vary depending on the sequence composition (GC vs AT), the length, and the type of nucleic acid (see Sambrook et al., Molecular Cloning: A Laboratory Manual, 2nd Ed.; Cold Spring Harbor Laboratory Press, Plainview, N.Y., 1998).

[0053] Following the ligation of the oligonucleotide, the nucleic acid of interest is amplified using a primer complementary to the oligonucleotide. Specifically, the term “primer” as used herein refers to a sequence comprising two or more deoxyribonucleotides or ribonucleotides, preferably more than three, and more preferably more than eight, wherein the sequence is capable of initiating synthesis of a primer extension product, which is substantially complementary to a nucleic acid such as an adaptor or a ligated oligonucleotide. Environmental conditions conducive to synthesis include the presence of nucleoside triphosphates and an agent for polymerization, such as DNA polymerase, and a suitable temperature and pH. The primer is preferably single stranded for maximum efficiency in amplification, but may be double stranded. If double stranded, the primer is first treated to separate its strands before being used to prepare extension products. The primer can be an oligodeoxyribonucleotide. The primer must be sufficiently long to prime the synthesis of extension products in the presence of the inducing agent for polymerization. The exact length of primer will depend on many factors, including temperature, buffer, and nucleotide composition. The oligonucleotide primer typically contains 12-20 or more nucleotides, although it may contain fewer nucleotides.

[0054] Primers of the invention are designed to be “substantially” complementary to each strand of the oligonucleotide to be amplified and include the appropriate G or C nucleotides as discussed above. This means that the primers must be sufficiently complementary to hybridize with their respective strands under conditions which allow the agent for polymerization to perform. In other words, the primers should have sufficient complementarity with a 5′ and 3′ oligonucleotide to hybridize therewith and permit amplification of CpG containing nucleic acid sequence.

[0055] Primers of the invention are employed in the amplification process which is an enzymatic chain reaction that produces exponential quantities of target locus relative to the number of reaction steps involved (e.g., polymerase chain reaction or PCR). Typically, one primer is complementary to the negative (−) strand of the locus (antisense primer) and the other is complementary to the positive (+) strand (sense primer). Annealing the primers to denatured nucleic acid followed by extension with an enzyme, such as the large fragment of DNA Polymerase I (Klenow) and nucleotides, results in newly synthesized + and − strands containing the target locus sequence. Because these newly synthesized sequences are also templates, repeated cycles of denaturing, primer annealing, and extension results in exponential production of the region (i.e., the target locus sequence) defined by the primer. The product of the chain reaction is a discrete nucleic acid duplex with termini corresponding to the ends of the specific primers employed.

[0056] The oligonucleotide primers used in invention methods may be prepared using any suitable method, such as conventional phosphotriester and phosphodiester methods or automated embodiments thereof. In one such automated embodiment, diethylphos-phoramidites are used as starting materials and may be synthesized as described by Beaucage, et al. (Tetrahedron Letters, 22:1859-1862, 1981). One method for synthesizing oligonucleotides on a modified solid support is described in U.S. Pat. No. 4,458,066.

[0057] Another method for detecting a methylated CpG-containing nucleic acid includes contacting a nucleic acid-containing specimen with an agent that modifies unmethylated cytosine, amplifying the CpG-containing nucleic acid in the specimen by means of CpG-specific oligonucleotide primers, wherein the oligonucleotide primers distinguish between modified methylated and non-methylated nucleic acid and detecting the methylated nucleic acid. The amplification step is optional and although desirable, is not essential. The method relies on the PCR reaction itself to distinguish between modified (e.g, chemically modified) methylated and unmethylated DNA.

[0058] The term “modifies” as used herein means the conversion of an unmethylated cytosine to another nucleotide which will facilitate methods to distinguish the unmethylated from the methylated cytosine. Preferably, the agent modifies unmethylated cytosine to uracil. Preferably, the agent used for modifying unmethylated cytosine is sodium bisulfite, however, other agents that similarly modify unmethylated cytosine, but not methylated cytosine can also be used in the method. Sodium bisulfite (NaHSO3) reacts readily with the 5,6-double bond of cytosine, but poorly with methylated cytosine. Cytosine reacts with the bisulfite ion to form a sulfonated cytosine reaction intermediate which is susceptible to deamination, giving rise to a sulfonated uracil. The sulfonate group can be removed under alkaline conditions, resulting in the formation of uracil. Uracil is recognized as a thymine by Taq polymerase and therefore upon PCR, the resultant product contains cytosine only at the position where 5-methylcytosine occurs in the starting template DNA.

[0059] The primers used in the invention for amplification of the CpG-containing nucleic acid in the specimen, after bisulfite modification, specifically distinguish between untreated or unmodified DNA, methylated, and non-methylated DNA. MSP primers for the non-methylated DNA preferably have a T in the 3′ CG pair to distinguish it from the C retained in methylated DNA, and the complement is designed for the antisense primer. MSP primers usually contain relatively few Cs or Gs in the sequence since the Cs will be absent in the sense primer and the Gs absent in the antisense primer (C becomes modified to U (uracil) which is amplified as T (thymidine) in the amplification product).

[0060] The primers of the invention embrace oligonucleotides of sufficient length and appropriate sequence so as to provide specific initiation of polymerization on a significant number of nucleic acids in the polymorphic locus. Where the nucleic acid sequence of interest contains two strands, it is necessary to separate the strands of the nucleic acid before it can be used as a template for the amplification process. Strand separation can be effected either as a separate step or simultaneously with the synthesis of the primer extension products. This strand separation can be accomplished using various suitable denaturing conditions, including physical, chemical, or enzymatic means, the word “denaturing” includes all such means. One physical method of separating nucleic acid strands involves heating the nucleic acid until it is denatured. Typical heat denaturation may involve temperatures ranging from about 80° to 105° C. for times ranging from about 1 to 10 minutes. Strand separation may also be induced by an enzyme from the class of enzymes known as helicases or by the enzyme RecA, which has helicase activity, and in the presence of riboATP, is known to denature DNA. The reaction conditions suitable for strand separation of nucleic acids with helicases are described by Kuhn Hoffmann-Berling (CSH-Quantitative Biology, 43:63, 1978) and techniques for using RecA are reviewed in C. Radding (Ann. Rev. Genetics, 16:405-437, 1982).

[0061] When complementary strands of nucleic acid or acids are separated, regardless of whether the nucleic acid was originally double or single stranded, the separated strands are ready to be used as a template for the synthesis of additional nucleic acid strands. This synthesis is performed under conditions allowing hybridization of primers to templates to occur. Generally synthesis occurs in a buffered aqueous solution, generally at a pH of about 7-9. Preferably, a molar excess (for genomic nucleic acid, usually about 108:1 primer:template) of the two oligonucleotide primers is added to the buffer containing the separated template strands. It is understood, however, that the amount of complementary strand may not be known if the process of the invention is used for diagnostic applications, so that the amount of primer relative to the amount of complementary strand cannot be determined with certainty. As a practical matter, however, the amount of primer added will generally be in molar excess over the amount of complementary strand (template) when the sequence to be amplified is contained in a mixture of complicated long-chain nucleic acid strands. a large molar excess is preferred to improve the efficiency of the process.

[0062] The deoxyribonucleoside triphosphates dATP, dCTP, dGTP, and dTTP are added to the synthesis mixture, either separately or together with the primers, in adequate amounts and the resulting solution is heated to about 90°-100° C. from about 1 to 10 minutes, preferably from 1 to 4 minutes. After this heating period, the solution is allowed to cool to approximately room temperature, which is preferable for the primer hybridization. To the cooled mixture is added an appropriate agent for effecting the primer extension reaction (called herein “agent for polymerization”), and the reaction is allowed to occur under conditions known in the art. The agent for polymerization may also be added together with the other reagents if it is heat stable. This synthesis (or amplification) reaction may occur at room temperature up to a temperature above which the agent for polymerization no longer functions. Thus, for example, if DNA polymerase is used as the agent, the temperature is generally no greater than about 40° C. Most conveniently the reaction occurs at room temperature.

[0063] The agent for polymerization may be any compound or system which will function to accomplish the synthesis of primer extension products, including enzymes. Suitable enzymes for this purpose include, for example, E. coli DNA polymerase I, Klenow fragment of E. coli DNA polymerase I, T4 DNA polymerase, other available DNA polymerases, polymerase muteins, reverse transcriptase, and other enzymes, including heat-stable enzymes (i.e., those enzymes which perform primer extension after being subjected to temperatures sufficiently elevated to cause denaturation such as Taq DNA polymerase, and the like). Suitable enzymes will facilitate combination of the nucleotides in the proper manner to form the primer extension products which are complementary to each locus nucleic acid strand. Generally, the synthesis will be initiated at the 3′ end of each primer and proceed in the 5′ direction along the template strand, until synthesis terminates, producing molecules of different lengths. There may be agents for polymerization, however, which initiate synthesis at the 5′ end and proceed in the other direction, using the same process as described above.

[0064] Preferably, the method of amplifying is by PCR, as described herein and as is commonly used by those of ordinary skill in the art. However, alternative methods of amplification have been described and can also be employed. PCR techniques and many variations of PCR are known. Basic PCR techniques are described by Saiki et al. (1988 Science 239:487-491) and by U.S. Pat. Nos. 4,683,195, 4,683,202 and 4,800,159, which are incorporated herein by reference.

[0065] The conditions generally required for PCR include temperature, salt, cation, pH and related conditions needed for efficient copying of the master-cut fragment. PCR conditions include repeated cycles of heat denaturation (i.e. heating to at least about 95.degree. C.) and incubation at a temperature permitting primer: adaptor hybridization and copying of the master-cut DNA fragment by the amplification enzyme. Heat stable amplification enzymes like the pwo, Thermus aquaticus or Thermococcus litoralis DNA polymerases which eliminate the need to add enzyme after each denaturation cycle, are commercially available. The salt, cation, pH and related factors needed for enzymatic amplification activity are available from commercial manufacturers of amplification enzymes.

[0066] As provided herein an amplification enzyme is any enzyme which can be used for in vitro nucleic acid amplification, e.g. by the above-described procedures. Such amplification enzymes include pwo, Escherichia coli DNA polymerase I, Klenow fragment of E. coli DNA polymerase I, T4 DNA polymerase, T7 DNA polymerase, Thermus aquaticus (Taq) DNA polymerase, Thermococcus litoralis DNA polymerase, SP6 RNA polymerase, T7 RNA polymerase, T3 RNA polymerase, T4 polynucleotide kinase, Avian Myeloblastosis Virus reverse transcriptase, Moloney Murine Leukemia Virus reverse transcriptase, T4 DNA ligase, E. coli DNA ligase or Q.beta. replicase. Preferred amplification enzymes are the pwo and Taq polymerases. The pwo enzyme is especially preferred because of its fidelity in replicating DNA.

[0067] Once amplified, the nucleic acid can be attached to a solid support, such as a membrane, and can be hybridized with any probe of interest, to detect any nucleic acid sequence. Several membranes are known to one of skill in the art for the adhesion of nucleic acid sequences. Specific non-limiting examples of these membranes include nitrocellulose (NITROPURE) or other membranes used in for detection of gene expression such as polyvinylchloride, diazotized paper and other commercially available membranes such as GENESCREEN, ZETAPROBE (Biorad), and NYTRAN Methods for attaching nucleic acids to these membranes are well known to one of skill in the art. Alternatively, screening can be done in a liquid phase.

[0068] In nucleic acid hybridization reactions, the conditions used to achieve a particular level of stringency will vary, depending on the nature of the nucleic acids being hybridized. For example, the length, degree of complementarity, nucleotide sequence composition (e.g., GC v. AT content), and nucleic acid type (e.g., RNA v. DNA) of the hybridizing regions of the nucleic acids can be considered in selecting hybridization conditions. An additional consideration is whether one of the nucleic acids is immobilized, for example, on a filter.

[0069] An example of progressively higher stringency conditions is as follows: 2×SSC/0.1% SDS at about room temperature (hybridization conditions); 0.2×SSC/0.1% SDS at about room temperature (low stringency conditions); 0.2×SSC/0.l% SDS at about 42° C. (moderate stringency conditions); and 0.1×SSC at about 68° C. (high stringency conditions). Washing can be carried out using only one of these conditions, e.g., high stringency conditions, or each of the conditions can be used, e.g., for 10-15 minutes each, in the order listed above, repeating any or all of the steps listed. However, as mentioned above, optimal conditions will vary, depending on the particular hybridization reaction involved, and can be determined empirically. In general, conditions of high stringency are used for the hybridization of the probe of interest.

[0070] The probe of interest can be detectably labeled, for example, with a radioisotope, a fluorescent compound, a bioluminescent compound, a chemiluminescent compound, a metal chelator, or an enzyme. Those of ordinary skill in the art will know of other suitable labels for binding to the probe, or will be able to ascertain such, using routine experimentation.

[0071] Another embodiment of the invention provides a method of determining a predisposition to a cellular proliferative disorder of breast tissue in a subject comprising determining the state of methylation of one or more nucleic acids isolated from the subject, wherein the nucleic acid is selected from the group consisting of Twist, cyclin D2, RARβ2, HOXA5, WT1, 14.3.3 sigma, estrogen receptor, NES-1 and combinations thereof; and wherein the state of methylation of one or more nucleic acids as compared with the state of methylation of said nucleic acid from a subject not having a predisposition to the cellular proliferative disorder of breast tissue is indicative of a cell proliferative disorder of breast tissue in the subject.

[0072] As used herein, “predisposition” refers to an increased likely that an individual will have a disorder. Although a subject with a predisposition does not yet have the disorder, there exists an increased propensity to the disease.

[0073] Another embodiment of the invention provides a method for diagnosing a cellular proliferative disorder of breast tissue in a subject comprising contacting a nucleic acid-containing specimen from the subject with an agent that provides a determination of the methylation state of nucleic acids in the specimen, and identifying the methylation state of at least one region of least one nucleic acid, wherein the methylation state of at least one region of at least one nucleic acid that is different from the methylation state of the same region of the same nucleic acid in a subject not having the cellular proliferative disorder is indicative of a cellular proliferative disorder of breast tissue in the subject.

[0074] Invention methods are ideally suited for the preparation of a kit. Therefore, in accordance with another embodiment of the present invention, there is provided a kit it useful for the detection of a cellular proliferative disorder in a subject. Invention kits include a carrier means compartmentalized to receive a sample therein, one or more containers comprising a first container containing a reagent which modifies unmethylated cytosine and a second container containing primers for amplification of a CpG-containing nucleic acid, wherein the primers distinguish between modified methylated and nonmethylated nucleic acid. Primers contemplated for use in accordance with the invention include those set forth in SEQ ID NOs: 1 to 37.

[0075] Carrier means are suited for containing one or more container means such as vials, tubes, and the like, each of the container means comprising one of the separate elements to be used in the method. In view of the description provided herein of invention methods, those of skill in the art can readily determine the apportionment of the necessary reagents among the container means. For example, one of the container means can comprise a container containing an oligonucleotide for ligation to nucleic acid cleaved by a methylation sensitive restriction endonuclease. One or more container means can also be included comprising a primer complementary to the oligonucleotide. In addition, one or more container means can also be included which comprise a methylation sensitive restriction endonuclease. One or more container means can also be included containing an isoschizomer of said methylation sensitive restriction enzyme.

[0076] The above disclosure generally describes the present invention. A more complete understanding can be obtained by reference to the following specific examples which are provided herein for purposes of illustration only and are not intended to limit the scope of the invention.

[0077] Although the invention has been described with reference to the presently preferred embodiment, is should be understood that various modifications can be made without departing from the spirit of the invention. Accordingly, the invention is limited only by the following claims.

EXAMPLE 1

Methylation Status of Wilms' tumor suppressor gene (WT1)

[0078] The extent of methylation of the WT1-associated CpG islands in normal mammary epithelium, in breast cancer cell lines, and in primary mammary tumors, and expression of the WT1 mRNA and protein in the same cells and tissues was examined.

[0079] Cell lines and finite life span cultures Cell lines were obtained from ATCC (Rockville, Md.) and grown according to conditions specified. Also utilized were three independent cultures of finite life span human mammary epithelial cells (HMEC): 16637 (Clonetics, Walkersville, Md.) and 1-26, 3-14 (kindly provided by Dr. Steve Ethier, Univ. Michigan, Ann Arbor, Mich.). When indicated, cell lines were treated with 0.75 μM 5-aza-2′-deoxycytidine (5-aza-dC) or with 100 ng/ml Trichostatin A (TSA) as described in Ferguson, et al. (Proc Natl Acad Sci U S A. (2000) 97:6049-54).

[0080] Tumors and Organoids Primary breast tumors were obtained from the Johns Hopkins frozen tumor bank. Mammary organoids were prepared from reduction mammoplasty specimens of women with benign (B) or no (N) abnormalities in the breast as described in Fujii, et al. (Oncogene. 16: 2159-64, 1998). Briefly, the specimens were enzymatically digested into duct-like structures (organoids), filtered, histologically confirmed to contain more than 80% epithelial cells, and frozen at −70° C. until used. Also utilized were highly purified myo- and luminal epithelial cells isolated by differential centrifugation and fluorescence-activated cell sorting of enzymatically digested normal mammoplasty specimens (Gomm, et al., (1995) Anal Biochem. 226 91-9).

[0081] RT-PCR for WT1 mRNA Methods for RNA extraction and RT-PCR are known to those of skill in the art. The sequences of the primers used are as follows: for amplifying the 555 bp region surrounding WT1 exon 5, 5′-GCGGCGCAGTTCCCCAACCA-3′ (sense, nucleotides 882-901; Q ID NO:1) and 5′-ATGGTTTCTCACCAGTGTGCTT-3′ (antisense, nucleotides 1416-1437; SEQ ID NO:2); for amplifying the 382 bp region surrounding the KTS insert, 5′-GCATCTGAAACCAGTGAGAA-3′ (sense, nucleotides 1320-1339; SEQ ID NO:3) and 5′-TTTCTCTGATGCATGTTG-3′ (antisense, nucleotides 1685-1702; SEQ ID NO:4). Amplification was performed using a hot-start protocol: samples were heated to 94° C. for 4 minutes and then cooled to 80° C. prior to the addition of Taq polymerase (RedTaq, Sigma, St. Louis, Mo.). Samples were then heated to 94° C. for 30 seconds followed by either 50° C. for 30 seconds (for the KTS primers) or 56° C. for 30 seconds (for the Exon 5 primers) and then 72° C. for 1 minute for 40 cycles. PCR products were resolved by electrophoresis, using a 2% agarose gel for the exon 5-splice variants and a 12% polyacrylamide gel to resolve the KTS insert variants. Co-amplification of the ribosomal RNA 36B4 was performed as an internal control using the following primers: 5′-GATTGGCTACCCAACTGTTGCA-3′ (sense; SEQ ID NO:5) and 5′-CAGGGGCAGCAGCCACAAAGGC-3′ (antisense; SEQ ID NO:6).

[0082] Northern blots Total RNA was extracted as described above. After electrophoresis through a 1.5% agarose gel in MOPS buffer with 6.7% formaldehyde, RNA was transferred to nitrocellulose. Blots were probed with a PCR product corresponding to the WT1 zinc finger region, amplified using the primers described above and labeled by random priming using standard techniques.

[0083] Methylation-specific PCR Genomic DNA was isolated using standard techniques and treated with sodium bisulfite as described elsewhere (Herman, et al., Proc Natl Acad Sci U S A. 93: 9821-6, 1996). Methylation-specific PCR was performed using the following primers: to detect methylated promoter DNA, 5′-TTTGGGTTAAGTTAGGCGTCGTCG-3′ (sense; SEQ ID NO:7) and 5′-ACACTACTCCTCGTACGACTCCG-3′ (antisense; SEQ ID NO:8); to detect unmethylated promoter DNA, 5′-TTTGGGTTAAGTTAGGTGTTGTTG-3′ (sense; SEQ ID NO:9) and 5′-ACACTACTCCTCATACAACTCCA-3′ (antisense; SEQ ID NO:10); to detect methylated intron 1 DNA, 5′-CGTCGGGTGAAGGCGGGTAAT-3′(sense; SEQ ID NO:11) and 5′-CGAACCCGAACCTACGAAACC-3′ (antisense; SEQ ID NO: 12); to detect unmethylated intron 1 DNA, 5′-TGTTGGGTGAAGGTGGGTAAT-3′ (sense; SEQ ID NO: 13) and 5′-CAAACCCAAACCTACAAAACC-3′ (antisense; SEQ ID NO:14). The PCR reaction was as described above, except that the annealing temperature was 59° C., and the extension time was 45 seconds.

[0084] Western blots Total protein from cell lines was obtained from material harvested in TriReagent (Molecular Research Center, Cincinnati, Ohio) and initially used for RNA isolation. Protein purification was according to the manufacturer's protocol. After separation by SDS-PAGE and electrophoretic transfer to nitrocellulose membranes, proteins were incubated with an anti-WT1 antibody [WT (C-19); sc-192, Santa Cruz Biotechnology, Santa Cruz, Calif.] diluted 1:1000 in the blocking solution. Horseradish peroxidase-conjugated antibody against rabbit IgG (Amersham, Arlington Heights, Ill.) was used at 1:1000, and binding was revealed using enhanced chemiluminescence (Amersham, Arlington Heights, Ill.).

[0085] Expression of WT1 mRNA in mammary epithelial and breast cancer cell lines To evaluate WT1 expression in the breast, mRNA expression was analyzed by RT-PCR in a panel of normal and transformed cell lines. No WT1 RNA was detected in 3 independently derived finite lifespan mammary epithelial strains: HMEC 16637, 1-26, and 3-14. Among the three immortal breast epithelial cell lines, WT1 expression was observed in HMECs HBL-100 and MCF-10 A, but not in H16N. WT1 mRNA expression was examined in nine breast cancer cell lines, expression was easily detectable in five: HS578T, T47D, MDA-MB-468, 21MT, and 21PT, and undetectable in the remaining four: SKBR3, MDA-MB-435, MCF-7 and MDA-MB-231.

[0086] The specific expression of WT1 isoforms lacking the fifth exon and lacking the KTS insert has been reported to occur in breast cancer (Silberstein, et al., Proc Natl Acad Sci U S A. 94: 8132-7, 1997). To determine if differential expression of WT1 splice variants is seen in breast cancer cell lines, PCR primers were designed spanning the fifth exon such that mRNA encoding the isoform containing exon 5 yielded a 555 bp PCR product, while if exon 5 were missing a 504 bp PCR product was generated. PCR primers spanning the region of the KTS insert, such that a mRNA containing the insert would yield a 382 bp product, while an RNA lacking the insert would generate a 373 bp product were also used. Contrary to the findings in the published report (Silberstein et al., supra), in the five WT1 -expressing breast cancer cell lines, and in the WT1-expressing immortalized HMECs, all four splice variants—the two exon 5 isoforms and the two KTS isoforms, were present.

[0087] To confirm these results, Northern blot analysis was performed using total RNA isolated from a number of breast cancer cell lines. Similar to the results obtained by RT-PCR, WT1 mRNA expression was readily detected in HBL-100, HS-578T, T47D, and MDA-MB-468 cells but was not detected in MDA-MB-435, MDA-MB-231, SKBR3, or MCF-7 cells.

[0088] Thus, WT1 mRNA expression was undetectable in finite life span primary breast epithelial cell cultures, but was easily detectable in the neoplastic and immortalized HMECs and in seven of twelve breast cancer cell lines. Also, the striking correlation between results from Northern blots and RT-PCR experiments validated the RT-PCR protocol for the detection of WT1 mRNA expression.

[0089] Methylation of the WT1 locus in breast cancer cell lines The promoter and first intron of the WT1 gene contain dense CpG islands. These sequence elements are frequently sites of DNA methylation, and play a role in transcriptional silencing (Nan, et al., Cell. 88: 471-81, 1997; Ng, et al., Nat Genet. 23: 58-61, 1999). To determine whether methylation silences gene expression in the WT1-negative cell lines, the status of the WT1 promoter in the breast cancer cell lines was investigated . The promoter was methylated in the 4 cell lines that did not express WT1, but not in the 5 cell lines that did, consistent with the idea that methylation is a critical determinant of WT1 expression. There was one exception to this correlation. T47D cells contained methylated WT1 sequences but nevertheless expressed WT1 mRNA, suggesting that, in this case, methylation alone is insufficient to silence expression.

[0090] Promoter methylation is postulated to silence transcription, at least in part, by recruitment of histone deacetylase (HDAC) to hypermethylated loci (Nan, et al. supra and Ng, et al., supra). In order to assess the functional significance of WT1 promoter methylation, MDA-MB-231 and MCF-7 cells were treated with 5-aza-deoxyC, an inhibitor of DNA methyltransferases, or with TSA, an inhibitor of HDAC. As demonstrated before (Laux, et al., Breast Cancer Res Treat. 56: 35-43, 1999), treatment with 5-aza-deoxyC resulted in WT1 expression by MDA-MB-231 cells. Interestingly, this treatment did not cause WT1 expression in MCF-7 cells, nor did TSA restore expression in either cell line. In the same samples, these treatments restored expression of 14.3.3σ (Ferguson, et al., Proc Natl Acad Sci USA. (2000) 97:6049-54). These findings suggest that while promoter methylation correlates with gene silencing, it may not play a causal role.

[0091] Expression of WT1 in primary breast tissue These findings from cell lines were expanded to patient samples, including normal breast epithelium and primary breast tumors. Breast carcinomas arise from luminal epithelial cells in the mammary duct. Normal breast tissue also contains a layer of myoepithelial cells which overlie the luminal epithelium. To ensure that the normal samples contained luminal epithelial cells, three different types of epithelial cell preparations were used, including (1) three short term cultures of HMECs, (2) nine organoid preparations of mammary ducts, and (3) eight samples of highly purified luminal and myoepithelial cells (isolated from 4 patient samples).

[0092] WT1 expression was not detected by RT-PCR in 3 HMEC samples, in eight out of nine breast organoid preparations, nor in any of eight purified epithelial cell preparations. By western blotting, WT1 protein was not detected in three organoid samples nor in two HMECs. In contrast, WT1 expression was easily detectable in 27 out of 31 (87%) primary breast carcinomas.

[0093] The HMECs did not express WT1; however, RT-PCR using primers described above demonstrated the expression of Exon 5 (+) and Exon 5 (−) isoforms in five out of seven tumors, while the remaining two expressed only the Exon 5 (+) isoform. KTS (+) and KTS (−) isoforms were detected in all nine tumors examined. Thus, a majority of the tumors expressed both Exon 5 splice variants of WT1, and all of the tumors express both splice variants involving the KTS insert. Interestingly, the sole breast organoid sample that expressed WT1 expressed all four splice variants as well.

[0094] Methylation of WT1-associated CpG islands in normal and malignant breast tissue Since methylation of the promoter-associated CpG island correlated with a lack of WT1 expression in breast cancer cell lines, the methylation status of the promoter and first intron CpG islands was studied in this panel of breast organoids and carcinomas. Prior studies demonstrating tumor-specific methylation of the CpG islands associated with the WT1 gene have employed methylation-sensitive restriction enzymes (Huang, et al., Cancer Res. 57: 1030-4, 1997; Laux, et al., Breast Cancer Res Treat. 56: 35-43, 1999; and Huang, et al., Hum Mol Genet. 8: 459-70, 1999). This technique is a reliable way to identify individual methylated sites, but it is unable to assess large-scale methylation patterns. The density of methylation, rather than methylation of any specific CpG dinucleotide, is responsible for gene silencing (Herman, et al., Semin Cancer Biol. 9: 359-67, 1999). Therefore, methylation of the CpG islands was evaluated using methylation specific PCR (MSP). This method allows the direct evaluation of several methylation sites per PCR reaction, and choosing a variety of sequences for PCR primers allows the rapid assessment of many CpG dinucleotides (Herman, et al., Proc Natl Acad Sci U S A. 93: 9821-6, 1996).

[0095] MSP was performed using DNA extracted from 19 primary tumors and nine breast organoid preparations. The WT1 promoter CpG island was unmethylated in DNA from all nine organoid samples. In contrast, six of 19 tumors contained methylated DNA, and the remaining 13 were completely unmethylated. This rate of promoter methylation (32%) is not dissimilar to the 25% incidence reported by Laux et al. (Breast Cancer Res Treat. 56: 35-43, 1999). Thus, methylation of the WT1 promoter is a tumor-specific phenomenon. Contrary to expectation, however, each of the six tumors that contained methylated WT1 also expressed WT1 protein. WT1gene methylation, therefore, was not effective in silencing gene expression. Next, the CpG island in the first intron of the WT1 gene, a region where tumor-specific methylation has also been previously reported was examined. Methylation of WT1 was detected in all three breast organoid preparations and in nine of ten tumor samples evaluated. Thus, the first intron of WT1 is methylated in both normal and malignant breast tissue, and is unrelated to tumorigenesis.

[0096] Methylation of the CpG island associated with the WT1 promoter is associated with a gene silencing in several breast cancer cell lines. While treatment of MDA-MB-231 cells with the methyltransferase inhibitor 5-aza-deoxyC results in re-expression of the gene, this was not seen in MCF-7 cells. Additionally, treatment with the HDAC inhibitor TSA had no effect on WT1 expression, suggesting that DNA methylation and histone acetylation play only minor roles in the regulation of WT1 expression in mammary epithelium.

[0097] This study demonstrates tumor-specific methylation of the CpG islands of WT1. Surprisingly, expression of WT1 mRNA and protein in the majority of breast cancer samples evaluated was also found, including in every sample that contained methylated DNA. these findings that breast carcinomas express WT1 despite tumor-specific gene methylation emphasizes the importance of evaluating methylation and gene expression concurrently in the same tissue.

[0098] WT1 mRNA was readily detected in tumor samples using a single step PCR protocol. While it is possible to detect WT1 expression in normal epithelium using a nested PCR, this would not alter the finding that the gene is overexpressed in tumors compared with normal tissue. The use of RT-PCR may allow the detection of a relatively weakly-expressed gene, but WT1 protein was readily detected by Western blotting in tumors. Since protein is the functional species, this finding suggests that WT1 is abundant enough in tumors to play a functional role.

[0099] These data also reveal a discrepancy between gene regulation in tissue culture and in vivo. Methylation of the WT1 promoter is associated with gene silencing in breast cancer cell lines. In contrast, the promoter-associated CpG island was methylated in 32% of the tumors examined; contrary to expectation, these tumors express WT1. These data highlight the fact that there are multiple mechanisms for gene silencing, of which hypermethylation of a CpG island is only one. More importantly, these findings emphasize the idea that cell lines do not necessarily reflect the in vivo situation. They also serve to point out that hypermethylation of a CpG island may be insufficient to silence expression, demonstrating the importance of assessing gene expression as well as promoter methylation status when evaluating the role of a particular gene in a particular tumor type.

[0100] In summary, these data demonstrate that WT1 is not expressed in normal breast epithelium and is over-expressed in the majority of primary breast tumors. Tumor-specific methylation of the CpG island occurs in breast cancer, but appears to be inconsequential to gene expression.

EXAMPLE 2

Hypermethylation and loss of expression of cyclin D2

[0101] The extent of methylation of the cyclin D2-associated CpG islands in normal mammary epithelium, in breast cancer cell lines, and in primary mammary tumors, and expression of the cyclin D2 mRNA and protein in the same cells and tissues was examined.

[0102] Cell Lines and Tissues. The breast cancer cell lines MDAMB435, MCF7, T47D, SKBR3, ZR75.1, MDAMB468, HS578T, MDAMB231 and the immortal human mammary epithelial cell lines (HMEC) MCF10A and HBL100 were obtained and maintained in culture according to instructions (ATCC, Rockville, Md.). The two matched tumor cell lines, 21PT, derived from a primary tumor and 21MT, from the metastasis of the same patient, were propagated as described elsewhere. The breast cancer cell line, MW, was obtained from Dr. Renato Dulbecco. HMEC-H16N (immortalized with HPV) was kindly provided by Dr. Vimla Band. Cultured finite life span human breast epithelial cell strains 04372, 219-6, and 166372 were obtained from Clonetics (Walkersville, Md.), and HMEC strains 1-26 and 3-14 were kindly provided by Dr. Steve Ethier. Finite life span HMEC 184, the immortalized HMECs 184A1 (passage 15 and 99) and 184B5 were kindly provided by Dr. Martha Stampfer, and grown as described (http://www.lbl.gov/LBL-Programs/mrgs/review.html). Cell extracts from finite lifespan HMECs 70N and 81N were kindly provided by Dr. Khandan Keyomarsi. Mammary organoids were prepared from reduction mammoplasty specimens of women with benign or no abnormalities in the breast following collagenase digestion as described in Bergstraessar LM, (1993). Human mammary luminal and myoepithelial cells were prepared by progressive collagenase digestion of breast tissue, sedimentated to obtain organoids (ductal and lobulo-alveolar fragments), cultured short term, and finally highly enriched by using an immunomagnetic separation technique (Niranjan B, 1995).

[0103] Primary breast tumor tissues were obtained after surgical resection at the Johns Hopkins University and Duke University, and stored frozen at −80° C. Samples containing greater than 50% tumor cells were selected following microscopic examination of representative tissue sections from each tumor. Microdissection of carcinoma and ductal carcinomas in situ (DCIS) lesions from eight micron cryosections was performed by using a laser capture microscope, or by manually scraping the cells with a 25G needle under 40× magnification. Genomic DNA was extracted by incubating the microdissected cells at 55° C.×12 h in 50 μl buffer containing 10 mM Tris Cl (pH 8.0), 1 mM EDTA, 0.1% Tween 20, and 0.5 μg/μl proteinase K. The extract was heat inactivated at 95° C. for 5 min., and used directly for sodium bisulfite treatment.

[0104] RT-PCR. RNA was treated with RNAse-free DNAse (Boehringer-Mannheim) (0.5-1u/ul) for 30 min. at 37° C., followed by heat inactivation at 65° C. for 10 min. RT reactions contained 2 μg DNAse treated RNA, 0.25 μg/μl pdN6 random primers (Pharmacia), 1× first strand buffer (GibcoBRL), 1 mM dNTP (Pharmacia), and 200 U MMLV-RT (GibcoBRL), and were incubated for 1 h at 37° C. followed by heat inactivation at 75° C. for 5 min. PCR was performed using the primers 5′-CATGGAGCTGCTGTGCCACG -3′ (sense; SEQ ID NO:15) and 5′-CCGACCTACCTCCAGCATCC -3′ (antisense; SEQ ID NO:16) for cyclin D2 and primers 5′-AGCCATGGAACACCAGCTC-3′ (sense; SEQ ID NO: 17) and 5′-GCACCTCCAGCATCCAGGT-3′ (antisense; SEQ ID NO:18) for cyclin D1. Co-amplified products of 36B4, a “housekeeping” ribosomal protein gene, was used as an internal control, using primers 5′GATTGGCTACCCAACTGTTGCA-3′ (sense; SEQ ID NO:19) and 5′-CAGGGGCAGCAGCCACAAAGGC-3′ antisense; SEQ ID NO:20). The 25 μl reactions contained 1×buffer (2×Reaction Mix, cat # 10928-026, BRL) and 100 nM of each primer. The PCR conditions were: 1 cycle of 94° C. for 1 min “hot start” then addition of 1 u of Taq polymerase (RedTaq), 1 cycle of 94° C. for 2 min, 35 cycles of: 94° C. for 15 sec, 55° C. for 30 sec, 72° C. for 45 sec, and finally 72° C. for 5 min. The PCR samples were resolved by electrophoresis on a 2% agarose gel in 1× TBE buffer.

[0105] Methylation-specific PCR (MSP). One μg genomic DNA or the 50 ul extract of microdissected cells was treated with sodium bisulfite as described in Herman J G, (1996), and was analyzed by MSP using primer sets located within the CpG-rich island in the cyclin D2 promoter. Primers specific for unmethylated DNA were 5′-GTTATGTTATGTTTGTTGTATG-3′ (sense; SEQ ID NO:21) and 5′-GTTATGTTATGTTTGTTGTATG-3′ (antisense; SEQ ID NO:22) and yielded a 223 base-pairs PCR product. Primers specific for methylated DNA were 5′-TACGTGTTAGGGTCGATCG-3′ (sense; SEQ ID NO:23) and 5′-CGAAATATCTACGCTAAACG-3′ (antisense; SEQ ID NO:24) and yielded a 276 base-pair PCR product. The PCR conditions were as follows: 1 cycle of 95° C. for 5 min; 35 cycles of 95° C. for 30s, 55° C. for 30s and 72° C. for 45s; and 1 cycle of 72° C. for 5 min. The PCR products were resolved by electrophoresis in a 2% agarose gel in 1×TBE buffer.

[0106] Treatment of Cells with 5′-aza-2′ -deoxycytidine (5-aza-dC) and Trichostatin A (TSA). Cells were seeded at a density of 1×106 cells per 100-mm plate. 24 h later cells were treated with 0.75 μM 5-aza-dC (Sigma) or with 100 ng/ml of TSA (Sigma). Total cellular DNA and RNA were isolated at 0, 3 and 5 days after addition of 5-aza-dC and at 0, 24 and 48 hours after addition of TSA, as described above.

[0107] Western Blot Analysis: Proteins were extracted from cell pellets and from 8 micron sections of primary breast tumors in buffer containing 20 mM Tris pH 7.5, 150 nM NaCl and PMSF, and sonicated. Twenty μg of protein were fractionated on 12.5% SDS-PAGE and transferred by electrophoresis to a nylon membrane. The blot was incubated with anti-cyclin D2 antibody (Ab-4, “cocktail” mouse monoclonal antibodies, Neomarkers, San Diego, Calif.) diluted 1:200 in 5% skim milk, for 2 h at room temperature. Horseradish peroxidase-conjugated antibody anti-mouse IgG (Amersham) was used at 1:1000, and binding was revealed using enhanced chemiluminescence (Amersham).

[0108] Cyclin D2 mRNA expression in breast cancer Serial analysis of gene expression (SAGE) and subsequent microarray analysis preveously revealed that, compared with finite lifespan HMECs, cyclin D2 expression was significantly lower in a small panel of primary breast tumors [Nacht M, et al., Cancer Research 59:5464-5470 (1999). To confirm the validity of these findings, we investigated expression of cyclin D2 by RT-PCR in three finite life span and 6 immortal HMECs, 11 breast cancer cell lines and 24 primary breast carcinomas. A ribosomal protein RNA, 36B4, was co-amplified as an internal control. Abundant expression of cyclin D2 mRNA was noted in all three finite life span HMECs and in 4 of 6 immortalized HMECs. The two immortalized HMEC lines lacking cyclin D2 expression were HBL100 and MCF10A. In contrast, 10 of 11 breast cancer cell lines showed no detectable expression of cyclin D2. Only one breast cancer cell line, HS578T, expressed a low but detectable level of cyclin D2. Likewise, the results with primary tumors reflected the findings in cultured cells. Eighteen of 24 primary breast carcinomas expressed significantly lower levels of cyclin D2 mRNA as compared with finite lifespan HMEC 184 and five other HMECs. As an additional control for cyclin D2 expression, the expression of cyclin D1 was analyzed in the same panels of cell lines and tumors. Consistent with previous observations Cyclin D1 mRNA was detectable in all the cell lines and primary breast tumors tested. Thus, in both breast cancer cell lines and primary tumors specific loss of cyclin D2, but not cyclin D1, mRNA expression was observed.

[0109] Cyclin D2 mRNA expression in luminal and myoepithelial cells of the breast It has been reported that cyclin D2 is expressed in myoepithelial but not in luminal epithelial cells of the breast (Lukas J, (1995)). Therefore, lack of expression of cyclin D2 in breast cancers would be expected, since the vast majority of these tumors originate from luminal rather than myoepithelial cells. This conclusion was based, however, on the results from a single HMEC preparation. The present study used a larger panel of tissues. Luminal and myoepithelial cells isolated from four normal mammoplasty specimens from women aged 18 to 33 were used. Paired luminal and myoepithelial cells were obtained from the same breast of two women. Each cell type was purified using immunomagnetic beads. The human luminal and myoepithelial cells were separated by virtue of their exclusive expression of epithelial membrane antigen (EMA) and common acute lymphoblastic leukemia antigen (CALLA) respectively. The purity of the populations was checked by immunocytochemistry using cytokeratins 18 and 19 as markers for luminal cells and cytokeratin 14 as a marker for myoepithelial cells. These tests showed that the final population was 95-99% pure in each case. Cyclin D2 expression was assessed in the purified cell preparations by RT-PCR. Cyclin D2 expression was observed in four of four purified luminal epithelial cells, as well as four of four myoepithelial cells. However, one luminal epithelial cell sample had a significantly lower expression of cyclin D2. Four HMECs of the 184 series, which stain for luminal cell markers-cytokeratins 8 and 18 and mucin, but not for myoepithelial cell marker-cytokeratin 14, also expressed cyclin D2 mRNA. Thus, cyclin D2 mRNA was expressed in all eight of eight luminal and four of four myoepithelial cell preparations from the normal breast.

[0110] Western analysis reveals loss of cyclin D2 protein in primary tumors For Western blot analysis specific anti-cyclin D2 antibodies that did not cross-react with cyclin D1 were used. While cyclin D2 protein was clearly detected in all seven HMECs tested (11-24, 1-26, 70N, 166372, 81N, 9F1403 and 184A1), it was undetectable in the majority (10/13) of primary breast tumors. Thus, HMECs that were derived from normal breast tissue and expressed high levels of cyclin D2 mRNA show clearly detectable levels of cyclin D2 protein as well. In contrast, primary breast tumors that exhibited low or absent cyclin D2 mRNA showed a corresponding loss of the cyclin D2 protein.

[0111] The cyclin D2 promoter is hypermethylated in breast cancer cell lines and primary tumors In somatic cells, about 80% of the CGs are methylated. Exceptions to this are the CpG islands in the promoter region of many genes. CpG islands are GC-rich regions of DNA, approximately 1 kb in length, present in the promoters of more than 60% of human genes. Normally CpG islands are unmethylated and the chromatin in those sites is enriched in hyperacetylated histone and deficient in histone H1, characteristic of active chromatin. Both unmethylated and methylated DNA are assembled into nucleosomes.

[0112] The cyclin D2 promoter contains a CpG-rich region at 1000 to 1600 base-pairs 5′ to the translation start site. To test whether aberrant methylation is associated with loss of cyclin D2 expression, primers for a Methylation Specific PCR (MSP) assay were designed to rapidly screen for cyclin D2 promoter methylation. Hypermethylation of the CpG rich region was detected in 11 of 11 breast cancer cell lines which also lacked expression of cyclin D2 protein. Aberrant methylation was also noted in 49 of 106 (46%) primary breast carcinomas.

[0113] Next, to determine whether cyclin D2 promoter-methylation is a tumor-specific phenomenon, DNA from histopathologically normal breast tissue adjacent to the surgically resected cancer was tested. All 11 samples of normal breast epithelial tissue adjacent to carcinoma were unmethylated at the CpG sites tested by MSP.

[0114] To further support the observation that cyclin D2 hypermethylation does not occur in normal HMECs and is associated with malignancy, normal breast epithelial cells prepared by a variety of techniques was examined. By MSP analysis, cyclin D2 promoter was found to be unmethylated in seven mammary organoid preparations from reduction mammoplasties, and in five finite life span HMECs cultured from non-malignant breasts. The only exception to this finding was in immortalized HMECs HBL100 and MCF10A, which contained hypermethylated cyclin D2. As expected, these HMECs were the only two that did not express cyclin D2 mRNA.

[0115] To rule out the contribution of inflammatory blood cells present in breast cancer specimens as the source of methylated cyclin D2, ten samples of peripheral blood cells (PBLs) from non-cancer patients were tested. All ten PBLs contained unmethylated cyclin D2 alleles.

[0116] Expression of cyclin D2 protein was undetectable in 10 of the 1 3 primary breast cancers tested. However, methylation of the cyclin D2 promoter was noted only in six of these ten primary tumors. This finding suggests that while methylation may cause silencing of cyclin D2 expression in many breast cancers, alternative pathways account for the loss of the protein in a proportion of these tumors.

[0117] Cyclin D2 promoter hypermethylation in preneoplasia Ductal carcinoma in situ (DCIS) is a preneoplastic lesion with a potential for progression to invasive cancer. To determine if hypermethylation of the cyclin D2 promoter occurs early in the evolution of breast cancer, MSP analysis was performed on DNA from carefully microdissected samples of DCIS. Hypermethylation was noted in 44% of DCIS samples. In the cases where adjacent invasive cancer was present as well, the methylation status of both lesions was concordant. This finding suggests that alteration of cyclin D2 expression may be an early event, and may precede transformation to the fully malignant stage of invasive carcinoma.

[0118] Re-expression of cyclin D2 mRNA in breast cancer cell lines Breast cancer cell lines MdAMB231 and MCF7 do not express cyclin D2 mRNA or protein. If silencing of expression is mediated by promoter methylation and/or altered chromatin conformation, then demethylation of the gene by exposure to 5-aza 2′-deoxycytidine (5aza-dC), or treatment with the histone deacetylase inhibitor, trichostatin A (TSA), should result in removal of the repressive mechanism and re-expression of the gene. Indeed, when MDAMB231 and MCF7 cells were exposed to 5-aza-dC in culture, the cyclin D2 promoter was partially demethylated (as analyzed by MSP), and cyclin D2 mRNA expression was restored (as analyzed by RT-PCR). Further, exposure to TSA also led to re-expression of the cyclin D2 mRNA. These results suggest that methylation at the promoter region plays a functional role in suppressing the expression of cyclin D2 in breast cancer.

[0119] Using RT-PCR, cyclin D2 expression was detected in four normal luminal epithelial cultures of the 184 series, in four of four purified luminal epithelial cell extracts, and in four of four myoepithelial cell extracts Using MSP, promoter hypermethylation was detected in 49/106 (46%) of the tumors. Hypermethylation of the gene correlated with lack of cyclin D2 mRNA and/or protein expression. Thus, in about 50% of breast cancers, cyclin D2 silencing may be attributed to tumor-specific methylation.

EXAMPLE 3

Hypermethylation and loss of expression of 14-3-3 sigma

[0120] The extent of methylation of the cyclin 14-3-3 sigma-associated CpG islands in normal mammary epithelium, in breast cancer cell lines, and in primary mammary tumors, and expression of the 14.3.3 sigma mRNA and protein in the same cells and tissues was examined.

[0121] Cell Lines and Tissues. The breast cancer cell lines Hs578t, MdA-MB-231, MdA-MB-435 and MCF-7 and the human mammary epithelial cell lines, MCF-10A and HBL-100 were obtained and maintained according to instructions (ATCC). The two matched tumor cell lines, 21PT and 21MT were propagated as described (Band, et al. (1990) Cancer Res. 50:7351-7357). Cultured normal human breast epithelial cell (HMEC) strains, 161, 184, 172, and 48, and the conditionally and fully immortal cell lines, 184A1(passage 15 and 99), and 185B5 were grown as described (http://www.lbl.gov/LBL-Programs/mrgs/review.html). Three additional short term cultures of HMECs,(#04372 and #16637) were grown according to specifications (Clonetics). Primary breast tumor tissues were obtained immediately after surgical resection at the Johns Hopkins University or Duke University, and stored frozen at −80° C. Microscopic examination of representative tissue sections from each tumor revealed that these samples contained greater than 50% tumor cells. Microdissection of primary tumor cryosections was performed by using a laser capture microscope (Schutze, et al. (1998) Nat Biotechnol 16:737-42) or by manually scraping the cells with a 20G needle under 40× magnification (Umbricht, et al. (1999) Oncogene 18:3407-14.).

[0122] Northern Blot Analysis. Total RNA was isolated from primary tumor tissues using Trizol Reagent (Life Technologies). Five micrograms were resolved on 1.5% agarose/formaldehyde gels, and transferred to a nylon filter using standard methods (Gene Screen, DuPont). A 375 bp σ-specific probe was generated using MCF-10A cDNA as a template and the primers 5′-ACAGGGGAACTTTATTGAGAGG-3′ (SEQ ID NO:25) and 5′-AAGGGCTCCGTGGAGAGGG-3′ (SEQ ID NO:26). Hybridizations were done in Quikhyb (Stratagene) according to the manufacturer's instructions. Filters were exposed to autoradiographic film for up to 5 days. To test for uniform loading of the samples, blots were stripped and reprobed with a 1.5 kb DNA fragment specific for 18S rRNA (ATCC, Clone #HHCSA65).

[0123] Loss of Heterozygosity (LOH) Studies. A TG repeat sequence in the 3′UTR of σ was amplified using: 5′-GAGGAGTGTCCCGCCTTGTGG-3′ (sense; SEQ ID NO:27) and 5′- GTCTCGGTCTTGCACTGGC-3′ (antisense; SEQ ID NO:28) primers, which yields a product of 117 bp. The 25 μl reactions contained 50 ng of template DNA (10), 17 mM NH4SO4, 67 mM TrisCl (pH 8.8), 6.7 mM MgCl2, 1% DMSO, 1.5 mM dNTP, 20 ng of each primer, 2 ng of γ-32P-labeled sense primer, and 0.5 μl Taq polymerase. PCR conditions were as follows: 1 cycle of 94° C. for 90s; 35 cycles of 94° C. for 1 min, 57° C. for 30s, 72° C. for 30s; and 1 cycle of 72° C. for 5 min. PCR products were fractionated on a sequencing gel, which was exposed to autoradiographic film overnight (Evron, et al. (1997) Cancer Res, 57:2888-9).

[0124] Mutation Analysis. A 1.2 kb PCR product, encompassing the entire σ coding sequence, was generated using two primers, 5′-GTGTGTCCCCAGAGCCATGG-3′ (sense; SEQ ID NO:29) and 5′- GTCTCGGTCTTGCACTGGCG-3′ (antisense; SEQ ID NO:30). The PCR reaction contained 50 ng of DNA, 6.4% DMSO, 1.5 mM dNTPs, 100 ng of each primer and 0.5 μl Taq polymerase in a 50 μl reaction volume. α-33P cycle sequencing was performed using the Amplicycle sequencing kit (Perkin Elmer). Four different α-33P-labeled primers were used to sequence the entire σ coding sequence: 5′-CACCTTCTCCCGGTACTCACG-3′ (antisense; SEQ ID NO:31), 5′-GAGCTCTCCTGCGAAGAG-3′ (sense; SEQ ID NO:32), 5′-GAGGAGGCCATCCTCTCTGGC-3′ (sense; SEQ ID NO:33) and 5′-TCCACAGTGTCAGGTTGTCTCG-3′ (antisense; SEQ ID NO:34).

[0125] Transfection of Human Breast Cancer Cell Lines. 1.5×105 of MCF-7, Md.A-MB-231, and Hs578t, or 2.5×105 cells of MdA-MB-435 breast cancer cells were seeded in six-well plates. The following day, transfections were performed using Trans IT-LT1 (Mirus Corp.) as per manufacturer's instructions. Plasmids used in the transient transfections include: KKH luciferase, containing 4 kb of the σ-promotor linked to the luciferase gene in the pGL3-Basic vector (Promega); pCMV-β-gal (Clontech), which was used to correct for the efficiency of transfection; and pGL3-Basic (Promega), which was used as a negative vector control against which KKH luciferase activities were compared. Two μg of luciferase reporter plasmid or the pGL3-Basic vector control and 0.5 μg of CMV-β-gal reporter plasmid were used for each transfection.

[0126] Luciferase and β-galactosidase Assays. Cell lysates were made approximately 48 hr post-transfection as per manufacturers instructions (Promega, Luciferase Assay System). Luciferase and β-galactosidase activities were quantitated using the luciferase assay system (Promega) and the Aurora GAL-XE reporter gene assay (ICN Pharmaceuticals, Inc), respectively. Experiments were done in triplicate. Luciferase activity was first normalized for efficiency of transfection by using the ratio of luciferase to β-galactosidase activity. For each transfected cell line, the results were compared with the mean of pGL3 vector control levels and expressed as fold elevated expression above pGL3. The means and standard deviations of the results of all experiments were calculated.

[0127] Sodium Bisulfite DNA Sequencing. Genomic DNA was subjected to sodium bisulfite modification as described in Herman et al. ((1996) Proc. Natl. Acad. Sci. USA, 93:9821-9826). Bisulfite-converted DNA was amplified, as described above, using primers that encompass the first exon of the a gene: 5′-GAGAGAGTTAGTTTGATTTAGAAG-3′ (sense primer with start at nt 8641; SEQ ID NO:35) and 5′-CTT ACTAATATCCATAACCTCC-3′ (antisense primer with start at nt 9114; SEQ ID NO:36) which generated a 474 bp PCR product. Conditions for PCR were as follows: 1 cycle at 95° C. for 5 min; 35 cycles at 95° C. for 45s, 55° C. for 45s and 72° C. for 60s; and 1 cycle at 72° C. for 4 min. The product was purified using a Qiagen PCR purification kit (Qiagen Corp) and sequenced using the sense primer with an ABI automated fluorescent sequencer according to the manufacturer's instructions.

[0128] Methylation-specific PCR (MSP). One μg genomic DNA was treated with sodium bisulfite as described in (Herman, et al. (1996) Proc. Natl. Acad. Sci. USA 93, 9821-9826), and was analyzed by MSP using a primer set that covered CG dinucleotide numbers 3, 4, 8 and 9. Primers specific for methylated DNA: 5′-TGGTAGTTTTTATGAAAGGCGTC-3′ (sense; SEQ ID NO:37) and 5′-CCTCTAACCGCCCACCACG-3′ (antisense; SEQ ID NO:38), and primers specific for unmethylated DNA: 5′-ATGGTAGTTTTTATGAAAGGTGTT-3′ (sense; SEQ ID NO:39) and 5′-CCCTCTAACCACCCACCACA-3′(antisense; SEQ ID NO:40) yielded a 105-107 bp PCR product. The PCR conditions were as follows: 1 cycle of 95° C. for 5 min; 31 cycles of 95° C. for 45s, 56° C. for 30s and 72° C. for 30s; and 1 cycle of 72° C. for 4 min.

[0129] Treatment of Cells with 5′-aza-2′-deoxycytidine (5-aza-dC). Cells were seeded at a density of 2×106 cells per 100-mm plate. 24 h later cells were treated with 0.75 μM 5-aza-dC (Sigma) (Ferguson, et al. (1995) Cancer Res 55:2279-2283). Total cellular RNA and genomic DNA were isolated from the cells at 0 and 3 days after addition of 5-aza-dC as described herein.

[0130] RT-PCR. RNA was treated with RNase-free DNAse (Boehringer-Mannheim) (1 μg/μl) for 2 h at 37° C., followed by heat inactivation at 65° C. for 10 min. RT reactions contained 1 μg DNAse treated RNA, 0.25 μg/μl pdN6 random primers (Pharmacia), 1× first strand buffer (GibcoBRL), 0.5 mM dNTP (Pharmacia), and 200 U MMLV-RT (GibcoBRL), and were incubated for 1 h at 37° C. PCR was performed using the σ-specific primers 5′-GTGTGTCCCCAGAGCCATGG-3′ (SEQ ID NO:41) and 5′-ACCTTCTCCCGGTACTCACG-3′ (SEQ ID NO:42) using buffer conditions described herein. The PCR conditions were: 1 cycle of 95° C. for 5 min; 30 cycles of 60° C. for 45s, 72° C. for 45s and 95° C. for 45s. PCR samples were resolved by electrophoresis in a 2% agarose gel.

[0131] Assay for G1 and G2 checkpoint and chromosomal aberrations. The G1 cell cycle checkpoint and chromosomal aberrations in mitosis were assessed as described previously (Pandita, et al. (1996) Oncogene 13:1423-1430). Specifically, cells in plateau phase were irradiated with 3 Gy, sub-cultured after 24 h, and metaphases were collected. G1 type aberrations were examined at metaphase. All categories of asymmetric chromosome aberrations were scored: dicentrics, centric rings, interstitial deletions/acentric rings, and terminal deletions.

[0132] The efficiency of G2 checkpoint control was evaluated by measuring the proportion of cells in metaphase after irradiation. Chromosomal aberrations at mitosis were assessed by counting chromatid breaks and gaps per metaphase as described elsewhere (Morgan, et al. (1997) Mol Cell Biol 17:2020-2029). Specifically, cells in exponential growth phase were irradiated with 1 Gy. Metaphases were harvested 45 and 90 minutes following irradiation and examined for chromatid type breaks and gaps. Fifty metaphases each were scored for G1 and G2 types of chromosomal aberrations.

[0133] Introduction of σ into the σ-negative breast cancer cell line MDA-MB-435 by adenoviral infection. Cells were seeded and grown to 50% confluency. Adenovirus encoding either or β-galactosidase (Hermeking, et al. (1997) Mol. Cell 1:3-11) was added to the culture at a multiplicity of infection of 5000:1 and infection was allowed to take place overnight. The cells were harvested, fixed and stained with Hoechst dye and subjected to FACS analysis.

[0134] σ Expression in Normal, Immortalized and Tumorigenic Breast Epithelial Cells. By SAGE analysis, σ was found to be expressed at an average of 7-fold lower levels in three human breast cancer cell lines, 21PT, 21MT and MDA-MB-468 than in two populations of normal human mammary epithelial cells (HMEC). Northern blot analysis was performed to confirm this finding in other breast cancer cell lines and in primary breast tumors. No expression of σ was detected in 45 of 48 (94%) primary tumors. In contrast, σ was expressed at easily detectable levels in all 6 cultured HMEC populations and 5 immortalized but nontumorigenic cell lines. These results indicate that loss of a gene expression is a frequent event in human breast cancer.

[0135] Genetic alterations within the σ gene. Possible causes for loss of σ gene expression in breast tumors include deletion of the chromosomal region containing the gene or intragenic mutations that lead to decreased mRNA stability. σ localizes to chromosome 1p35, an arm that has been extensively studied for LOH in breast cancer (Hermeking, et al. (1997) Mol. Cell 1,3-Bieche, et al. (1995) Genes Chrom. Cancer 14:227-251). LOH has been observed for the 1p32-36 region at a frequency of 15-25%. However, it is not known whether the region lost in these tumors includes σ (Genuardi, et al.. (1989) Am. J Hum. Genet. 45:73-82; Trent, et al. (1993) Genes Chrom Cancer 7:194-203; Nagai, et al. (1995) Cancer Res. 55:1752-1757; Tsukamoto, et al. (1998) Cancer 82:317-322). Therefore, the loss of σ by utilizing a TG repeat sequence within the 3′ UTR of the σ gene itself was examined. Using primers that span the TG repeats, the locus in 45 sets of normal and tumor DNA pairs was studied. Twenty of 45 (44%) of the patients were found to be heterozygous with respect to the length of the PCR-product. Only one of the 20 tumor specimens exhibited LOH (Table 1). Eleven of these 20 samples were tested by Northern blot analysis, and no σ transcripts were detectable. These results prompted an examination whether there were smaller genetic changes within the coding region of σ. The entire 1190 bp coding region from σ-nonexpressing (σ-negative) breast cancer cell lines, MdA-MB-435 and Hs578t and 7 primary tumor tissues was amplified with PCR and sequenced. No mutations were found. In addition, 25 primary tumor DNA samples were analyzed by single stranded conformation polymorphism, and no abnormalities were detected. These results suggest that genetic alterations within σ are not a primary mechanism for loss of gene expression.

1TABLE 1Incidence of σ alterations in breast cancerNo. withσ expression,No. with methylatedLOH/total,No. withNorthern blot[σ/totalTG repeatmutation/totalSampleanalysisSequencingMSPPCRSequencingSSCPNormal breastMortal HMEC strains6/6010/3Immortal HMEC lines5/5010/5Reduction0/601mammoplasty,microdissectedepitheliumBreast cancerCell LinesMCF-7+XXMDA-MB-231+XXMDA-MB-435X++XHs578tX++XPrimary tumors 2/4510/1043/501/200/70/25Microdissected Carinoma32/32

[0136] Epigenetic alterations of the σ gene. Next tested was whether the lack of σ mRNA was due to deficiencies in factors required for σ transcription. The two breast cancer cell lines, MDA-MB-435 and Hs578t, served as model systems for σ-negative primary tumors that harbored wild type σ alleles, while the two breast cancer cell lines, MCF-7 and MDA-MB-231, served as σ-positive controls since they both express detectable levels of σ. The plasmid KKH-luciferase contains 4 kb of sequence upstream of the transcriptional start site of σ linked to the luciferase reporter gene; this upstream region contains the sequences necessary for p53 and γ-irradiation-inducible transcription of σ (5). Following transient transfection of the four cell lines with the reporter plasmid, high levels of expression was observed (70- to 300-fold above the promoterless parental vector) in both σ-negative and σ-positive breast cancer cell lines. These results indicate that the σ-negative breast cancer cells, like the σ-positive cells, are able to support transcription from the σ promoter equally well, and contained factors required for transcription.

[0137] σ has a CpG rich region (CpG island) within its first and only exon that begins near the transcription initiation site and ends approximately 800 bp downstream. To explore a role of hypermethylation in silencing a gene expression, the nucleotide sequence of this region was determined after treating the DNA with sodium bisulfite (Frommer, et al. (1992) Proc. Natl. Acad. Sci. USA 89:1827-1831). PCR primers were designed to amplify a region spanning 27 CpG dinucleotides within the CpG island. No significant methylation was observed using DNAs from four σ-positive cell lines including 2 HMECs (184, MCF-10A) and two tumorigenic breast cancer cell lines (MCF-7 and MDA-MB-231). In contrast, DNAs from two σ-negative breast cancer cell lines, HS578t and MDA-MB-435, were fully methylated at all of the CpG sites. Since there was a strong correlation between σ-methylation status and mRNA expression in all the cell lines examined, 10 σ-negative primary breast tumors were also examined. All of the tumor DNAs exhibited partial or complete methylation of the 27 CpG dinucleotides.

[0138] Next, an MSP assay was utilized to detect methylation of the CpG island, using primers spanning the region between CpG dinucleotides 3 and 9 within the σ gene. Primers were designed that take advantage of the nucleotide sequence differences between methylated and unmethylated DNA as a result of bisulfite modification. By this method, 5/5 σ-positive HMEC strains were completely unmethylated. In addition, DNAs from the σ-positive immortalized breast epithelial cells (MCF-10A, HBL-100) and breast cancer cell lines (MCF-7 and MDA-MB-231), were also unmethylated at the sites examined. In contrast, DNAs from the σ-negative breast cancer cell lines, Hs578t and MDA-MB-435, were fully methylated. Similarly, 43 of 50 samples from primary breast tumors were partially or completely methylated. Of these 43 tumors, 26 were examined by Northern blot analysis, and all 26 lacked detectable σ gene expression. Three of the seven unmethylated breast tumor samples also lacked σ transcripts; the expression pattern for the remainder was not tested. These results demonstrate that aberrant methylation of σ is a frequent event in breast cancer, but that other mechanisms are responsible for silencing the gene in a small fraction of breast tumors.

[0139] Previous reports indicate that σ gene expression is restricted to differentiated epithelial cells. In order to clearly ascertain the cellular origin of methylated DNA, normal and tumor tissues were microdissected and analyzed for a methylation by MSP. All six DNA samples of microdissected mammary epithelial cells obtained from reduction mammoplasty specimens were unmethylated. In contrast, all 32 samples of DNA from microdissected breast carcinomas were methylated within the σ CpG island. These results indicate that hypermethylation of the σ gene is associated with loss of gene expression in the majority of primary breast tumors. The data from gene expression, genetic and epigenetic studies are summarized in Table 1.

[0140] In order to determine the effect of methylation on σ gene expression, two fully methylated, σ-negative cell lines, Hs578t and MDA-MB-435, were treated with the DNA methyltransferase inhibitor, 5-aza-dC. Treatment of cells with 0.75 μM 5-aza-dC for 3 days led to demethylation of the CpG rich region encompassed by the MSP primers. Moreover, 5-aza-dC treatment resulted in reactivation of gene expression, as demonstrated by RT-PCR. These results demonstrate that methylation is at least partially responsible for loss of a transcription in breast cancer cells.

[0141] Functional consequences of loss of σ in breast cancer cells. The function of human σ has been analyzed in human colon carcinoma cells. These studies demonstrated that following ionizing irradiation, σ sequesters cdc2-cyclin B1 complexes in the cytoplasm, thus arresting the cells in G2. These actions prevent the cell from initiating mitosis before repair of its damaged DNA. Colon carcinoma cells lacking σ can still initiate, but do not maintain, G2 arrest, leading to mitotic catastrophe and cell death.

[0142] In an attempt to determine the effects of loss of σ gene expression on cell cycle regulation in breast cancer cells, the effects of γ-irradiation on the σ-negative breast cancer cell lines, MDA-MB-435, 21NT, and 21MT, and the σ-positive breast cancer cell line, MCF-7 were tested. First, G1 type chromosomal aberrations were examined 24 h after cells were exposed to 3 Gy of γ-irradiation. All categories of G1-type chromosomal aberrations were scored at metaphase; their frequency was identical in the two cell types. These results indicate that the examined cell lines have similar G1 cell cycle checkpoint control responses to ionizing radiation.

[0143] Next, the G2 checkpoint function in the four cell lines was evaluated. Cells in exponential growth phase were γ-irradiated with 1 Gy and metaphases were examined for chromatid type breaks and gaps. Defective G2 arrest will increase these values. The results show a striking difference in the ability of σ-negative cells and σ-positive cells to repair their damaged DNA. Forty-five minutes post irradiation, σ-negative cells exhibited up to twice as many G2 type chromosomal aberrations as MCF-7 cells. This number increases to three-fold by 90 minutes. Moreover, while repair of DNA damage was evident in the MCF-7 cells, as evidenced by a decrease in the number of G2 type aberrations between 45 and 90 minutes, no decrease was seen in σ-negative cells.

[0144] Finally, in order to further demonstrate the role of σ in G2 checkpoint function in breast cells, a cloned copy of the gene was overexpressed in the σ-negative cell line MDA-MB-435 as well as in normal breast epithelial cells using the adenovirus expression system used to express a in colon cancer cells (5). Overexpression of σ in these breast epithelial cells led to a rapid and permanent G2 arrest, whereas the control virus-infected cells showed no effect. These results indicate that although the σ-negative cell lines have a functional G1 cell cycle checkpoint, they accumulate more genetic damage following irradiation which is consistent with its failure to arrest in G2 in response to DNA damage.

[0145] In summary, these results show that in striking contrast to normal breast tissue, greater than 90% of breast cancers lack detectable expression of σ. Hypermethylation of the a gene occurs in a CpG-rich region that extends from the transcriptional initiation site to the middle of the coding region. Bisulfite genomic sequencing of this 500 bp region showed that is consistently and densely methylated in σ-negative cell lines and primary breast tumors. Several studies have clearly documented that gene activity correlates inversely with the density of gene-specific CpG island methylation, but is less dependent on the position and distance of the methylated DNA sequences from the transcriptional initiation site. With respect to σ, dense methylation just downstream of its transcriptional start site is strongly associated with gene silencing. Furthermore, in σ-negative cell lines, 5-aza-dC-induced demethylation of the CpG island leads to reactivation of gene expression, indicating that hypermethylation plays a causal role in σ gene inactivity.

EXAMPLE 4

Hypermethylation and loss of expression of RAR β2

[0146] The extent of methylation of the RAR β2-associated CpG islands in normal mammary epithelium, in breast cancer cell lines, and in primary mammary tumors, and expression of the RAR β2 mRNA and protein in the same cells and tissues was examined.

[0147] Cell cultures Human epithelial mammary cells (HEMC) from reduction mammoplasty including three mortal strains, 184, 48R and 172R, and two immortal strains, 184A1 and 184B5, were obtained and cultured according to the protocols designed by Dr Martha Stampfer (see the HMEC Homepage, http://www.lbl.gov/*mrgs/index.htlm) using Clonetics (Walkersville, Md., USA) reagents. Human breast cancer cell lines were maintained in Dulbecco's modified Eagle's medium (GIBCO) (Hs578t,MCF-7, MDA-MB-231 and T47D) or IMEM medium (Biofluids) (MDA-MB-435, MDA-MB-468, ZR751) with 5% fetal calf serum (FCS). For drug treatments, exponentially growing cells were seeded in 10 cm2 plates at a density of 36105 cells/plate or in 6-well plates at 16105 cells/well. Cells were allowed to attach overnight before the addition of the appropriate concentration of 5-Aza-2′ deoxycytidine (5-Aza-CdR) (Sigma), Trichostatin A (TSA) (Sigma) or RA(Sigma). When reduction of retinoids was required, cells were treated in either medium with 0.5% FCS or charcoal-dextran stripped FCS (Hyclone). At the indicated time points, both attached and detached cells were harvested, counted with Trypan Blue (Life Technologies) and processed for DNA or RNA extraction. 5-Aza-CdR was dissolved in 0.45% NaCl containing 10 mM sodium phosphate (pH 6.8). Trichostatin A and all-trans-retinoic acid (RA) (Sigma) were reconstituted in absolute ethanol (solvent). The growth inhibition (%) was calculated as: (1-NT/NC)6100, where NT is the number of treated cells and NC is the number of control cells.

[0148] Tissue samples: Normal and tumor tissues were collected from existing tumor banks (Instituto per lo Studio e la Cura dei Tumori, Milan; the Cancer Center, Rotterdam, the Johns Hopkins Breast Cancer Program, Baltimore, Md., USA). All tumor samples were obtained from excess clinical specimens and institutional guidelines for the acquisition and maintenance of such specimens were followed. DNA and RNA extraction: Extraction of DNA and RNA from breast cancer cell lines was performed by using DNAzol and Trizol respectively (LifeTechnologies) according to the manufacturer's instructions. Genomic DNA was further treated with 500 mg/ml proteinase K at 55° C., extracted with phenol-chloroform-isoamylic alcohol (24:24:1) (CIA) and ethanol precipitated. Extraction of DNA from paraffinated breast cancer and lymph node tissues was essentially performed as previously described (Formantici et al., 1999). One to three consecutive sections estimated to contain at least 90% tumor cells were incubated at 58° C. overnight in 200 ml of extraction buffer (50 mM KCl, 10 mM Tris-HCl (pH 7.5), 2.5 mM MgCl2, 0.1 mg/ml gelatin,0.45% NP-40, 0.45% Tween 20, and the solution was heated at 95° C. for 15 min to inactivate the proteinase K and then centrifuged at 6000 r.p.m. The DNA in the supernatant was used for analysis.

[0149] Southern blotting Genomic DNA (7 mg) was digested overnight with 15 U/mg of XbaI, HpaII and MspI enzymes, electrophoresis on a 0.8% agarose gel and transferred to Hybond-N filter. A 227 bp probe was amplified using the sense 5′-AGA GTT TGA TGG AGTTGG GTG GAG-3′ (SEQ ID NO:43) and antisense 5′-CAT TCG GTT TGGGTC AAT CCA CTG-3′ (SEQ ID NO:44) primers, gel purified and labeled with 32P-dCTP using the Megaprime DNA labeling system (Amersham). After hybridization the filters were washed and exposed to X-ray film at −80° C. for autoradiography.

[0150] Methylation specific PCR (MSP) Bisulfite modification of genomic DNA was essentially performed as described by Herman et al. (1996) and described herein. Modified DNA was used immediately or stored in aliquots at −20° C. The PCR mixture contained 1×PCR buffer (16.6 mM ammonium sulfate, 67 mM Tris (pH 8.7), 1.5 mM MgCl2),dNTPs (each at 1.25 mM), primers (300 ng each per reaction),and bisulfite-modified DNA (50 ng) or unmodified DNA(50 ng). Reactions were hot started at 95° C. before the addition of 2.5 U of Taq polymerase (Qiagen). Amplification was carried out in a Thermal Cycler 480 Perkin Elmer for 30cycles (1 min at 94° C., 1 min at the annealing temperature (at) selected for each primer pair, 1 min at 72° C.), followed by 4 min at 72° C. Twelve μl of the PCR reaction were electrophoresed onto 1.5% agarose gels, stained with ethidium bromide and visualized under UV. Two primer pairs, W3 sense 5′-CAGCCCGGGTAGGGTTCACC-3′ (SEQ ID NO:45), W3 antisense 5′-CCGGATCCTACCCCGACGG-3′ (SEQ ID NO:46), and W4sense 5′-CCGAGAACGCGAGCGATCC-3′ (SEQ ID NO:47) and W4 anti-sense 5′-GGCCAATCCAGCCGGGGCG-3′ (SEQ ID NO:48), were designed on the human RAR β sequence (Shen et al., 1991) and used to control the Na bisulfite modification. The primer pairs selected to detect the unmethylated DNA were as follows: Ulsense 5′-GTG GGT GTA GGT GGA ATA TT-3′ (SEQ ID NO:49) and U1antisense 5′-AAC AAA CAC ACA AAC CAA CA-3′ (SEQ ID NO:50) (at 55° C.); U2 sense 5′-TGT GAG TTA GGA GTA GTG TTTT-3′ (SEQ ID NO:51) and U2 antisense 5′-TTC AAT AAA CCC TAC CCA-3′ (SEQ ID NO:52) (at 49° C.); U3 sense 5′-TTA GTA GTT TGG GTA GGGTTT ATT-3′ (SEQ ID NO:53) and U3 antisense 5′-CCA AAT CCT ACC CCAACA-3′ (SEQ ID NO:54) (at 55° C.); U4 sense 5′-GAT GTT GAG AAT GTGAGT GAT TT-3′ (SEQ ID NO:55) and U4 antisense 5′-AAC CAA TCC AACCAA AAC A-3′ (SEQ ID NO:56) (at 55° C.); The sequences of the primers to detect the methylated DNA were: M1 sense 5′-AGC GGGCGT AGG CGG AAT ATC-3′ (SEQ ID NO:57) and M1 antisense 5′-CAACGA ACG CAC AAA CCG ACG-3′ (SEQ ID NO:58) (at 63° C.); M2 sense 5′-CGT GAG TTA GGA GTA GCG TTT C-3′ (SEQ ID NO:59) and M2 antisense 5′-CTT TCG ATA AAC CCT ACC CG-3′ (SEQ ID NO:60) (at 57° C.); M3 sense 5′-GGT TAG TAG TTC GGG TAG GGTTTA TC-3′ (SEQ ID NO:61) and M3 antisense 5′-CCG AAT CCT ACC CCGACG-3′ (SEQ ID NO:62) (at 64° C.); M4 sense 5′-GTC GAG AAC GCG AGCGAT TC-3′ (SEQ ID NO:63) and M4 antisense 5′-CGA CCA ATC CAA CCGAAA CG-3′ (SEQ ID NO:64) (at 64° C.).

[0151] M and U primers were designed in the same regions, with one or two nucleotide differences to meet annealing requirements. Fragment M3 (position 773±1007) contains the βRARE (792±808) and the transcription start site (position 844);fragment M4 (position 949±1096) contains an Sp1 element (position 1074±1081).

[0152] RT±PCR The exon 5 (sense primer 5′-GAC TGT ATG GAT GTTCTG TCA G-3′; SEQ ID NO:65) and exon 6 (antisense primer 5′-ATT TGTCCT GGC AGA CGA AGC A-3′; SEQ ID NO:66) were designed on the basis of published RAR β2 transcript (de The' et al., 1990;van der Leede et al., 1992) and used to amplify 50 ng of DNase treated total RNA using the Superscript One-StepRT±PCR System (Life Technologies). RT±PCR with actin primers (sense primer 5′-ACC ATG GAT GAT GAT ATCG-3′; SEQ ID NO:67 and antisense primer 5′-ACA TGG CTG GGG TGTTGA AG-3′; SEQ ID NO:68) was used as an internal RNA control.

[0153] The RAR β2 promoter is methylated in breast cancer cell lines independently of their ER status and RA-inducibility RAR transcription was first tested in a panel of breast cancer cell lines grown in the absence of exogenous RA, by reverse transcriptase-PCR (RT±PCR), using primers encompassing exons 5 and 6 (de The' et al.,1990; van der Leede et al., 1992; Toulouse et al., 1997). Under these conditions, only one cell line, Hs578t,produced a detectable 256 bp RT±PCR product. Thus, previous reports were confirmed that RAR β gene expression is down regulated/lost in breast cancer cell lines. Growing cells in the presence of RA can assess the distinction between down regulation and loss. As previously reported (Swisshelm et al., 1994; Liu et al., 1997; Shang et al., 1999), we observed induction of RAR β expression and growth inhibition inT47D, MDA-MB-435, MCF7 and ZR75-1 cell lines treated for 48 h with 1 μM RA, but not in the MDA-MB-231 and MDA-MB-468 cell lines.

[0154] To see whether the RARβ2 methylation status correlated with the ER status, the methylation status was examined at RARβ2 in a panel of ER-positive (MCF7,T47D, ZR75-1) and ER-negative (Hs578t, MDA-MB-231, MDA-MB-435, MDA-MB-468) cell lines.

[0155] By Southern blotting, the CpG island of the RAR β2 promoter within a 7.5 kb XbaI DNA fragment encompassing the TATA box, the βRARE, the transcriptional start site (TS) and the 5′ untranslated region of exon 5 was examined. In this region nine HpaII sites can be identified (Shen et al., 1991; Baust et al.,1996). The DNA methylation status was analyzed by using the methylation-sensitive enzyme, HpaII. MspI, the isoschizomer of HpaII, insensitive to methylation, was used as a positive control. The PCR probe spans the βRARE and the TATA box regions. The same 7.5 kb region was previously analyzed in a colon carcinoma cell line, and the size of all the possible fragments relative to the most 3′HpaII site were reported (Cote' and Momparler, 1997). Genomic DNA from the ER-positive, RA-inducible cell line T47D is digested to completion, indicating that it is not methylated at any of the HpaII sites. In contrast, DNA from the ER-positive, RA-inducible ZR75-1 cell line and DNA from the ER-negative, RA-resistant MDA-MB-231 cell line showed to be differentially methylated at the methylation-sensitive sites. Using methylation-specific PCR (MSP), we further analyzed a 616 bp long RAR β2 region from nucleotide 481 to nucleotide 1096 (Shen et al., 1991) in all the cell lines. MSP entails the modification of genomic DNA by sodium bisulfite that converts all unmethylated, but not methylated, cytosine to uracil (Herman et al., 1996). The genomic DNAs from four breast cancer cell lines ZR75 1,MCF7, MDA-MB-231, MDA-MB-468 showed partial to complete methylation of the promoter region. The human mammary epithelial cell (HMEC) strain48R, expressing RAR β and three breast cancer cell lines, the RAR β-positive Hs578t and the RA-inducible MDA-MB-435 and T47D, revealed only the (U) unmethylated PCR products.

[0156] These results indicate that hypermethylation of the RARβ2 promoter occurs in breast cancer cell lines irrespective of the ER status, and can be detected in both RA-inducible, and RA-resistant breast cancer cells.

[0157] RAR β2 is unmethylated in both mortal and immortalized HMEC, but is methylated in primary breast tumors The next question examined was whether hypermethylation of RAR β2 promoter in cell lines has correlates in clinical breast cancer. As a normal control, the HMEC mortal strains (48R, 172R), that are the closest representation of normal mammary epithelial cells available were examined. Also analyzed were two immortal mammary epithelial strains (184A1 and 184B5). The DNA of these strains was found to be unmethylated. Consequently, methylation of RAR β2 may be an event in the progression of breast cancer, following immortalization. Genomic DNAs from three paraffinated samples of breast tumors, two ER-positive (T1, T2)and one ER-negative (T3), estimated to contain more than 90% tumor cells, were analyzed with all MSP primer pairs, and shown to be partially methylated. Both microdissected breast stroma, and microdissected normal epithelial cells were found unmethylated at RAR β2, making it very likely that the U products in the tumor samples were amplified either from residual normal epithelial cells, or stromal cells mixed to tumor cells. DNAs from matching histologically tumor free lymph node samples (N1±N3), were similarly analyzed and produced only the unmethylated PCR products. The DNA of additional 21 tumors was performed using two sets of primer pairs (U3/M3 andU4/M4). Fifteen (7 ER-positive and 8 ER-negative) of the 24 tumors presented methylation at the RAR β2promoter. With the same primer sets hypermethylation at RAR β2 was detected in the DNA of ten out of 39 primary breast tumors collected, and analyzed independently, at the Johns Hopkins University. The overall data indicate that hypermethylation at RAR β2 promoter occurs in approximately one third of primary breast tumors, and that the RAR β2 methylation state is independent of the ER status of the tumor.

[0158] 5-Aza-CdR induces partial demethylation at the RAR β2CpG island and reactivation of RAR β gene expression. In order to determine whether DNA methylation is affecting, at least in part, RAR β gene expression, all cell lines showing methylation at the RAR β2 promoter were treated with the DNA methyl transferase inhibitor, 5-Aza-CdR. Treatment of cells with either 0.4 or 0.8 mM 5-Aza-CdR for 3 days, led to partial demethylation of the CpG rich RAR β2 region. This was evident both by Southern analysis in the MDA-MB-23I cell line, and by MSP in all cell lines. Moreover, 5-Aza-CdR treatment resulted in reactivation of gene expression both in RA-inducible MCF7 and ZR75-1, and RA-resistant MDA-MB-23 1 and MDA-MB-468 cells. Subsequent studies examined whether reactivation of RAR β expression by 5-Aza-CdRA-resistant cells could be enhanced by RA. Using non-quantitative RT±PCR, a difference could not appreciated in the level of RAR β transcription in MDA-MB-231 cells treated with 0.4 mM5-Aza-CdR alone, or in combination, with 1 μM RA. In this experiment, 5-Aza-CdR alone, or in combination with RA, produced 63 and 96% growth inhibition respectively. In the same experiment, treatment with 1 μM RA alone produced a negligible effect on growth inhibition (52%). A synergistic effect of the two drugs on cancer cells was previously reported (Cote' and Momparler,1997; Bovenzi et al., 1999).

[0159] These data indicate that DNA methylation is, at least, one factor influencing the down regulation/loss of RAR β transcription in breast cancer cell lines with a methylated RAR β2 promoter. Cells treated with 5-Aza-CdR alone, or in combination with RA, showed re-expression of RAR b, which may have contributed, along with the toxic5-Aza-CdR, to the observed growth inhibition.

[0160] The HDAC inhibitor TSA can reactivate RAR β expression in RA-resistant cells; demethylation of the RAR β2 promoter is not an absolute requirement for RAR β reactivation. The chromatin status at a given locus can be dynamically influenced by the degree of acetylation/deacetylation due to HAT/HDAC activities. Absence of RAR β regulatory factors, like RAR a, as well as DNA-methylation, can contribute to pattern chromatin modifications at RAR β promoter in RA-resistant cell lines. One of these cell lines, MDA-MB-231, lacks RA-inducible RARα activity (Shao et al., 1994) and displays a RAR β2 methylated promoter. A subsequent study was designed to probe indirectly whether the level of HDAC at RAR β2 can influence RAR β expression, by testing the effect of TSA, a HDAC inhibitor on MDA-MB-231 cells (Yoshida et al., 1995). Cells were treated for 2 days, in the presence or absence of 100 ng/ml TSA alone, or in combination, with 1 μM RA. By using RT±PCR, it was clear that, unlike cells treated with RA alone, cells treated with a combination of RA and TSA re-expressed RAR β mRNA. Under the same experimental conditions, 100 ng/ml TSA alone, or in combination with 1 mM RA, produced 77and 92% growth inhibition, respectively. Treatment with 1 μRA alone did not affect significantly growth inhibition (52%). By MSP analysis, it was assessed that RAR β expression was restored in the presence of a methylated RAR β2 promoter. This finding indirectly shows that global alterations of HDAC activity, generated by TSA in MDA-MB-23 1 cells, involved RAR β2 resulting in RA-induced RAR β expression. Further, demethylation at RAR β2 did not seem to bean absolute requirement for RAR β gene expression in MDA-MB-231 cells. Noteworthy, persistence of methylation at RAR β2 was observed also in MCF7 cells where RAR β transcription could be restored in the presence of RA. Growth inhibition was observed in cells treated with TSA alone, or in combination, with RA. Very likely, RAR β along with TSA, a drug known to induce growth inhibition(Yoshida et al., 1995), contributed to the massive growth inhibitory effect that was observed.

[0161] These results show that RAR β2 promoter is methylated in breast cancer. This study presents evidence that, in breast cancer cells, RAR β2 promoter undergoes DNA hypermethylation, an epigenetic change known to induce chromatin modifications and influence gene expression. Methylation of the RAR β2 promoter region was detected, both in breast carcinoma cell lines, and a significant proportion of primary breast tumors. RAR β2 methylation status did not correlate with the ER status of breast cancer cells and was observed both in in situ lesions and invasive tumors. It is not clear when epigenetic changes occur during breast cancer progression. However, methylation of the promoter was not detected in both mortal, and immortal human mammary epithelial cell (HMEC) strains, as well as in normal microdissected breast epithelial cells. These results suggest that aberrant methylation of the RAR β2CpG island may be a later event following immortalization. Treatment of breast cancer cells presenting with a methylated RAR β2, with the demethylating agent 5-Aza-CdR, induced partial DNA demethylation and restored RAR β gene expression. This evidence clearly indicates that DNA methylation is at least a component contributing to RAR β downregulation/loss.

EXAMPLE 5

Hypermethylation of HOXA5

[0162] The extent of methylation of the HOXA5-associated CpG islands in normal mammary epithelium, in breast cancer cell lines, and in primary mammary tumors was examined.

[0163] Tissue preparations and cells: Freshly excised primary breast carcinomas or mammoplasty specimens were minced fine with razor blades and digested with 0.15% collagenase A and 0.5% dispase II (Boehringer Mannheim) prepared in RPMI 1640 medium. The cell clumps were separated from the lighter fibroblasts by gravity separation 3 times. The cell clumps were then digested for 15′ with trypsin, washed, and immunostained with anti-cytokeratin-specific antibody (CAM 5.2, Becton-Dickinson) to assess the level of epithelial cell enrichment. The epithelial cells comprised between 70-80% of the enriched cell population.

[0164] Frozen, surgically excised breast tumor samples were cryosectioned, and representative sections were screened by a pathologist after staining with hematoxylin and eosin. Sections containing more than 70% carcinoma cells were used for RNA and protein extractions directly. Breast cancer cell lines and immortalized HMECs were obtained from ATCC (Rockville, Md.). Finite life span HMECs were obtained from Dr. Martha Stampfer, HMEC strain 9F1403 was obtained from Clonetics.

[0165] Methylation specific PCR (MSP) and sodium bisulfite DNA sequencing: One μg of genomic DNA was treated with sodium bisulfite21 and was analyzed for MSP using primer sets specific for methylated DNA: 5′TTTAGCGGTGGCGTTCG-3′ (sense; SEQ ID NO:69) and 5′-ATACGACTTCGAATCACGTA-3′ (antisense; SEQ ID NO:70), and primers specific for unmethylated DNA: 5′-TTGGTGAAGTTGGGTG-3′ (sense; SEQ ID NO:71), and 5′AATACAACTTCAAATCACATAC-3′(antisense; SEQ ID NO:72) which yielded products of 183 and 213 bp respectively. Sodium bisulfite treated DNA was used to PCR-amplify the HOXA5 promoter region -97 to -303 bp, using the primers 5′ATTTTGTTATAATGGGTTGTAAT3′ (sense; SEQ ID NO:73) and 5′AACATATACTTAATTCCCTCC-3′ (antisense; SEQ ID NO:74). The product was purified using a Qiagen PCR purification kit (Qiagen Corp.) and was sequenced using the sense primer with an ABI automated fluorescent sequencer according to the manufacturer's instructions.

[0166] Treatment of cells with 5′-aza-2′-deoxycytidine (5-aza-dC): MDA-MB-231 breast cancer cells were treated with 0.75 μM 5-aza-dC (Sigma), and collected at 0, 3 and 5 days later. RT-PCR was performed using primers: 5′ TCATTTTGCGGTCGCTATCC-3′ (sense; SEQ ID NO:75) and 5′GCCGGCTGGCTGTACCTG-3′ (antisense; SEQ ID NO:76).

[0167] Immunoblot Analysis: Proteins were visualized by Western analysis and 10% SDS-PAGE. The primary antibodies [anti-HOXA5 (HOXA5-2, BABCO), anti-p53 (AB-6, Oncogene Science), or anti-β-actin (AC-15, Sigma), anti-p21 (15091A, Pharmingen), anti-Mdm2 (65101A, Pharmingen), anti-PARP (AB-2, Oncogene Sciences), anti-dynein (Zymed) and anti- Na+, K+-ATPase (Ed Benz, Johns Hopkins) (also used as loading controls, with actin)] were used at 1:1000 dilution.

[0168] p53 inactivation by mutation is low (20%) in human breast cancer. Looking for other mechanisms that may account for loss of p53 function in these tumors, the levels of p53 mRNA in breast cancer cell lines and in primary tumors was examined. p53 mRNA levels were 5-10 fold lower in tumor cells than in normal breast epithelium. A subsequent study looked for a consensus protein binding sites in the p53 promoter (Reisman, et al. Proc. Natl. Acad. of Sci. USA 85:5146-5150 (1988)), including those of HOX proteins which are known to function as transcription factors (Deschamps, et al. Crit. Rev. Oncog. 3:117-173 (1992); Scott, Nat. Genetics 15:117-118 (1997)). Selected HOX genes are differentially expressed in neoplasms of a number of tissues, but their functional relationship to the neoplastic phenotype remains to be elucidated. Six putative HOX-core binding sequences (ATTA) were identified within the 2.4 kb human p53 promoter. Of a number of HOX genes examined in breast tumor cells and control breast epithelium, HOXA5 mRNA levels were drastically reduced in breast cancer cells. In fact, there was a tight correlation between p53 and HOXA5 mRNA levels in the ten cell lines tested for both genes, with a correlation coefficient r=0.942. No such decreased expression was observed for HOXA10, B3, B7, or C8 mRNAs.

[0169] To test for a causal relationship between the decreased expression of p53 and HOXA5 mRNAs, ZR75.1 breast cancer cells or SAOS2 osteosarcoma cells were co-transfected with the −356 bp or the −2.4 kb human p53 promoter-Luciferase reporter together with HOX expression plasmids. HOXA5 transactivated the p53 promoter-dependent reporter activity up to 25-fold in ZR75.1 cells and up to 7-fold in SAOS2 cells. This effect was not seen with other homeotic genes HOXB4, HOXB5 and HOXB7.

[0170] Positive regulation of transcription by HoxA5 was observed with the mouse p53 promoter as well. A single putative Hox-binding sequence (located at nts −204 to −201) was identified in the upstream regulatory region of the murine p53 gene. SAOS2 cells were cotransfected with a −320 bp mouse p53 promoter fused to the CAT gene, together with expression plasmids encoding full-length murine HoxA5, HoxA7, or HoxC8 proteins. Similar to human HOXA5, a 15- to 20-fold increase in CAT activity in the SAOS2 cells cotransfected with the HoxA5 expression plasmid was observed, but no significant effect of HoxA7 or HoxC8. These results suggest that expression from the mouse p53 promoter is specifically stimulated by HoxA5. To define the sequence requirements for the transactivation finction, a deletion construct of the p53-promoter CAT construct was tested in cotransfection assays with the full-length HoxA5 expression plasmid. A deletion to −153 bp in the promoter region of the p53-CAT construct eliminated stimulation of CAT activity by the effector plasmid. A truncated HoxA5 protein termed pCMVΔHoxA5, lacking the homeodomain, was completely inactive in these experiments. Finally a “TT” to “GG” mutation in the core-binding site (−320 mp53MutCAT) that abolished DNA/protein complex formation in cell extracts (see below), completely abrogated transactivation of the CAT reporter gene by HoxA5.

[0171] Direct binding of HoxA5 to the ATTA-containing site in the p53 promoter (positions -204 to -201) was demonstrated by electrophoretic mobility shift (EMSA) and supershift assays. A band was observed in cell extracts from HoxA5 transfected cells, but not in extracts from control cells. This band was competed out by an excess of unlabeled oligonucleotide but not by an oligonucleotide with an unrelated sequence. No protein/DNA complex was observed in extracts mixed with an oligonucleotide primer which carries two mutations (TT to GG) in the core binding site. Finally, HOXA5 antibodies, but not pre-immune serum, caused a supershift of the bound HOXA5 protein/oligonucleotide complex. This supershift was abrogated by pre-incubation with excess antibody (antigen depletion). Similar shift patterns were observed in extracts of RKO cells transfected with the effector plasmid. These results indicate that the ATTA-containing sequence in the mouse p53 promoter is indeed a HoxA5-binding sequence.

[0172] The above results suggest that HOXA5 may possess growth-suppressive properties through activation of p53 expression. To test this possibility, breast cancer cells, MCF-7 and ZR75.1, which harbor wildtype p53 genes, were transfected with the full length HOXA5 and the ΔHOXA5 (homeodomain-deleted) expression plasmids and tested for colony-forming ability. No surviving colonies were obtained from HOXA5-transfected cells whereas those transfected with ΔHOXA5 and the vector control generated colonies with equal efficiency. To obtain stable cultures that could express HOXA5, clones of MCF-7 cells were generated in which the HOXA5 gene was placed under the control of an ecdysone-inducible promoter. Within 3 hours after induction of HOXA5 expression by the ecdysone analog, Ponasterone A (Pon A), the levels of p53 mRNA rose by 2-fold. Western blotting showed that p53 and its downstream targets, p21 and Mdm2, as well as HOXA5 were reproducibly induced 2-5 fold following treatment with Pon A. Moreover, addition of Pon A resulted in cell shrinkage by 24 hours followed by significant cell death (80-90%) after 48 hours. Cell death occurred by apoptosis according to the following criteria: 1) cells shrank and formed contractile bodies; 2) DNA laddering was observed; 3) poly (ADP-ribose) polymerase, a substrate for caspases, underwent cleavage by 12 hours; and 4) 70% of the cells showed micronucleus formation, membrane blebbing, and ghost cell features upon staining with acridine orange. This apoptosis was not accompanied by a detectable change in the levels of Bax protein.

[0173] The results herein are consistent with the hypothesis that an increase in the level of HOXA5 in MCF-7 cells leads to an increase in p53 levels, which in turn results in apoptosis. As a further proof of this model, MCF-7 cells expressing the E6 gene of human papilloma virus, when transfected with the HOXA5 expression vector, were fully able to form colonies. Presumably, the induced p53 in these cells was sequestered by E6 protein and was unable to induce apoptosis. These results support the idea that HOXA5 induces apoptosis through a p53-dependent pathway in MCF-7 cells. This is the first demonstration of the involvement of a HOX protein in apoptosis.

[0174] The hypothesis that HOXA5-induced apoptosis is mediated by p53 was tested as follows. The p53+/+ HCT 116 line of colon carcinoma cells and its p53−/− derivative clone 379.2 were transfected with HOXA5 and p53 expression vectors. Expression of HOXA5 or p53 in the parental HCT1 16 cells reduced the ability of the cells to form colonies. In contrast, HOXA5 and p53 expression led to different phenotypes in p53 null 379.2 cells. Whereas expression of p53 in these cells abrogated colony formation, expression of HOXA5 had no detectable effect. In the HOXA5-transfected cultures, stable colonies, expressing detectable amounts of HOXA5 protein, and of a size and number comparable to the vector control were observed. Thus, HOXA5 induces cell death only in the presence of a wild-type p53 gene, adding further evidence that p53 mediates HOXA5 activity. Conversely, cells lacking HOXA5 and p53 would be unable to mount a normal response to treatments, such as DNA damage, that normally raise p53 levels by stabilizing the protein. To test this possibility, the two tumor cell lines 21PT and 21MT, which have low expression of HOXA5 and p53 were treated with γ-radiation. No detectable increase in p53 level in 21 PT and 21 MT was observed, while, as expected, p53 was induced in MCF-7 cells.

[0175] These findings in cell culture experiments have in vivo correlates. In sixty-seven percent (20/30) of primary breast tumors, HOXA5 protein was undetectable. Strikingly, concurrent loss of p53 expression was observed in the same tumors that lacked HOXA5. Among those tumors expressing HOXA5, one showed a band migrating faster than wild-type HOXA5 present in the RKO cells. HOXA5 cDNAs from eleven p53-negative breast cancer samples and two finite life span human breast epithelial cell (HMEC) strains were sequenced. All HOXA5 coding regions were wild type, except that of tumor #5 which contained a frameshift mutation (G insertion at codon 204) that created a premature stop codon. There is a coupled loss of p53 and HOXA5 expression in primary breast carcinomas, possibly due to lack of expression or mutational inactivation of HOXA5.

[0176] Seeking an explanation for the absence of the protein in the tumors,HOXA5 DNA of 20 HOXA5-negative and 5 HOXA5-positive primary tumors was sequenced. All contained the wild-type sequence except tumor #5, in which the insertion of G was again found and which contained no wild-type allele. In the absence of mutations, loss of HOXA5 may be a consequence of a loss of upstream regulatory factors or may reflect some repressive phenomenon such as methylation of the gene. Methylation specific PCR (MSP) of sodium bisulfite-treated DNA showed that {fraction (16/20)} of the tumors contained partially or completely methylated CpGs in the HOXA5 promoter region (ACCN No. AC004080). In contrast, this region was completely unmethylated in human mammary epithelial cells (HMEC) of finite life span, 184 and 9F1403, and in 4 immortalized HMECs, HBL100, MCF10A, 184B5 and 184A1. Nucleotide sequencing of the region −97 bp to −303 bp of the HOXA5 promoter, using sodium bisulfite-treated DNA from HMEC 184, and cancer cell lines, MCF-7 and MDA-MB-23 1, showed that methylation correlated with silencing of gene expression. Expression of HOXA5 mRNA could be re-initiated in MDA-MB-231 cells by treatment with the DNA methyl transferase inhibitor, 5-aza-2′-deoxycytidine (5-aza-dC). These results are strong preliminary evidence that methylation of the HOXA5 promoter region may be responsible for silencing of gene expression.

[0177] Unlike most tumor types, up to 80% of sporadic breast cancers do not contain p53 mutations. These results suggest that the reduced p53 levels in these tumors result from the absence of a positive regulator of p53 mRNA synthesis. p53 normally functions as a tetramer, so even a small reduction in the concentration of p53 monomers can greatly reduce the effective concentration of tetramers. These results show for the first time that transfected HOXA5 upregulates both p53 promoter-reporter constructs and endogenous p53 synthesis, leading to apoptosis. Finally, HOXA5 was detectable in only one-third of the primary tumors. In the majority of the remaining tumors, lack of HOXA5 expression strongly correlated with methylation of its promoter region, suggesting a causal role for methylation in the silencing of HOXA5 gene expression.

[0178] In summary, these experiments show that HOXA5 is a positive regulator of p53 transcription and function in cultured cells. The correlation observed between HOXA5 and p53 levels in clinical breast cancer demonstrates that loss of HOXA5 expression is an important step in tumorigenesis.

EXAMPLE 6

Hypermethylation of NES-1

[0179] The extent of methylation of the NES-1-associated CpG islands in normal mammary epithelium, in breast cancer cell lines, and in primary mammary tumors was examined.

[0180] Cell Lines and Tissues. The immortalized HMECs 184A1 (passage 15 and 99) were kindly provided by Dr. Martha Stampfer, and grown as decribed (http://www.lbl.gov/LBL-Programs/mrgs/review.html; incorporated by reference herein). Mammary organoids were prepared from reduction mammoplasty specimens of women with benign or no abnormalties in the breast following collagenase digestion as described (Bergstraessar and Weitzman (1993) Cancer Res., 53:2644-2654). Primary breast tumor tissues were obtained after surgical resection at the John Hopkins University, and stored frozen at −80° C. DNA was extracted by standard methods (). RNA was extracted with Triazol.

[0181] Methylation-specific PCR (MSP). One Tg genomic DNA was treated with sodium bisulfite as described in Herman et al. (supra), and was analyzed by MSP using primer sets located within the third exon of Nes 1 gene. Primers specific for unmethylated DNA were 5′-TTGTAGAGGTGGTGTTGTTT-3′ (sense; SEQ ID NO:77) and 5′-TTGTAGAGGTGGTGTTGTTT-3′ (antisense; SEQ ID NO:78) and yielded a 128 base-pairs PCR product. Primers specific for methylated DNA were 5′-TTCGAAGTTTATGGCGTTTC-3′ (sense; SEQ ID NO:79) and 5′-TTATTTCCGCAATACGCGAC-3′ (antisense; SEQ ID NO:80) and yielded a 137 base-pairs PCR product. The PCR conditions were as follows: 1 cycle of 95° C. for 5 min “hot start”, then addition of lu of Taq polymerase (RedTaq); 35 cycles of 95° C. for 30s, 55° C. for 30s and 72° C. for 45s; and 1 cycle of 72° C. for 5 min. The PCR products were resolved by electrophoresis in a 2% agarose gel 1×TBE buffer.

[0182] RT-PCR. RNA was treated with RNAse-free DNAse (Boehringer-Mannheim) (0.5 1 u/ul) for 30 min. at 37° C., followed by heat inactivation at 65° C. for 10 min. RT reactions contained 2 Tg DNAse treated RNA, 0.25 Tg/T1 pdN6 random primers (Pharmacia), 1× first strand buffer (GibcoBRL), 1 mM of each dNTP (Pharmacia), and 200 U MMLV-RT (GibcoBRL), and were incubated for 1 h at 37° C. followed by heat inactivation at 75° C. for 5 min. PCR was performed using the primers 5′-ACCAGAGTTGGGTGCTGAC-3′ (sense; SEQ ID NO:81) and 5′-ACCTGGCACTGGTCTCCG-3′ (antisense; SEQ ID NO:82) for Nes1. A “housekeeping” ribosomal protein gene 36B4 was co-amplified as an internal control, using primers 5′-GATTGGCTACCCAACTGTTGCA-3′(sense; SEQ ID NO:83) and 5′-CAGGGGCAGCAGCCACAAAGGC-3′ (antisense; SEQ ID NO:84). The 25T1 reactions contained 1×buffer (1:10 of 10×PCR buffer BRL#, 1.2 mM MgSO4, 0.2 mM of each dNTP) and 100 nM of each primer. The PCR conditions were: 1 cycle of 94° C. for 1 min “hot start” then addition of 1 u of Taq polymerase (RedTaq); 1 cycle of 94° C. for 2 min; 35 cycles of: 94° C. for 30 sec, 55° C. for 30 sec, 72° C. for 45 sec and finally 72° C. for 5 min. The PCR samples were resolved by electrophoresis on a 2% agarose gel in 1×TBE buffer.

[0183] NES-1 expression was observed in mammary organoids and HMEC's from mammoplasty specimens of normal and benign disease breast. In finite life span HMEC primary breast carcinomas analyzed by RT-PCR, NES-1 expression was observed in seven of eleven samples. MSP analysis for a CpG-rich island at NES-l third exon in the same samples showed methylated sequences in samples that showed NES-1 expression and unmethylated sequences in samples without NES-1 expression. Methylated NES-1 is absent in normal tissue.

[0184] Although the invention has been described with reference to the presently preferred embodiments, it should be understood that various modifications can be made without departing from the spirit of the invention. Accordingly, the invention is limited only by the following claims.

Aberrantly methylated genes as markers of breast malignancy

Information

Publication Number

Date Filed

Date Published

Inventors

CPC

US Classifications

International Classifications

Abstract

Description

Claims

STATEMENT AS TO FEDERALLY SPONSORED RESEARCH