CZ53196A3 - Isolated and purified dna sequence encoding suppressor protein of prostate tumor and a recombinant vector and an expression vector containing thereof, system host-vector for producing polypeptide or protein, pharmaceutical preparation, dna probe, isolated and purified protein and method of its production, antibody and method of detecting absence of ptsg protein in tumorous cells in vitro, and rna probe - Google Patents

Abstract

This invention provides a novel nucleic acid molecule encoding a prostate/colon tumor suppressor gene product. The means and methods for detecting mutations and/or loss of prostate/colon tumor suppressor gene are provided. Also included within the scope of this invention are methods of suppressing the neoplastic phenotype of cancer cells having a defect in the prostate/colon tumor suppressor gene product. The invention also includes the means and methods for treating the cancer by administering the prostate/colon tumor suppressor gene.

Description

The present invention relates to an isolated and purified DNA sequence encoding a prostate tumor suppressor protein and a recombinant host and tel-vector protein system, a pharmaceutical composition, a purified protein, and a method of detecting the absence of PTSG and expression vector containing it, polypeptide or DNA probe. isolated and production, antibodies and proteins in tumor cells in vitro and RNA probes.

This invention is part of U.S. Pat. No. 08/246,604, filed May 20, 1994.

This invention has been implemented in part with government support of Grant No. CA 60358, granted by the Department of Health and Public Services, and Grant No. CA 55231, awarded by the National Cancer Institute. The government has certain rights to this invention.

Throughout this application, publications are given by first author and date of publication in parentheses. The findings of these publications are hereby incorporated by reference into the present application to more fully describe the state of the art to which this invention belongs.

BACKGROUND OF THE INVENTION

The present invention belongs to the field of tumor suppressor genes (anti-oncogenes) and relates generally to products and methods of performing broad spectrum therapy of various types of human cancer with tumor suppressor genes. In particular, the invention relates to methods of treating tumor cells by: i) administering vectors containing the

3. a nucleic acid sequence encoding novel proteins referred to herein as a prostate tumor suppressor gene (PTSG products); or 2) administering an effective amount of a protein encoded by that nucleic acid sequence. The invention also relates to the diagnosis of certain cancers, such as prostate cancer and colon cancer, using the cloned nucleic acid of the invention.

Cancer and tumors are the second leading cause of death in the United States and cause 547,000 deaths per year. One in three Americans will develop cancer and one in five will die of cancer (Scientific American Medicine, Part 12, I, 1, paragraph dated 1987). Although substantial progress has been made in identifying certain environmental and genetic causes of cancer, cancer mortality statistics indicate the need to improve cancer treatment and diseases and disorders.

what,

In connection with hereditary forms of cancer and a large number of well-studied tumor cells, a number of so-called cancer genes, i.e. genes involved in the etiology of cancer, have been identified. The study of tumor genes has helped to gain some understanding of the tumor development process. Although much of the knowledge about tumor genes has yet to be gained, the known tumor genes serve as useful models for understanding tumor development.

Tumor genes are broadly classified as oncogenes which, when activated, promote tumor development, and tumor suppression genes, which, when damaged, lose the ability to suppress tumor formation. While this classification provides a useful way of introducing the concept into the knowledge of tumor development, it is also possible that a particular gene will play different roles depending on the particular allelic form of the gene, its regulatory elements, genetic basis and the tissue environment in which it operates.

The widely used hypothesis of cancer is as follows:

(1) most human cancers are hereditary diseases; and (2) are caused by the expression or failure to express specific genes (ie mutant versions of normal regulatory cell growth genes or viral or other foreign genes in mammalian cells, causing inappropriate, poorly timed or ectopic expression of other classes of vital growth regulation genes.

A simplified view of the biological basis of neoplasia is that there are two major classes of tumor genes. The first class consists of mutated or otherwise aberrated alleles of normal cell genes that are involved in controlling cell growth and replication. These genes are cellular proto-oncogenes. When mutated, they encode new cellular functions that interfere with normal cell growth and regulation. The consequence of these changes is the generation of dominant expressed tumor phenotypes. In this model of dominantly expressed oncogenes, in a view that has prevailed since the conception of the genetic and mutational basis of neoplasia, the perception that persistence of a single wild-type allele is not sufficient to prevent neoplastic changes in the cell development or growth program. Thus, the genetic events responsible for the activation of these oncogenes can be viewed as a single intervention event. Some evidence for the involvement of such transforming oncogenes in clinical human cancer is represented by activation of tumorigenic activities of nyc oncogenes in Burkitt's lymphoma, expression of the bcr-abl chimeric gene in patients with chronic myelogenous leukemia, and activation of H-ras and K-ras oncogenes in other tumors. Access to genetically based therapy for dominant-expressed neoplastic diseases will be expected to require silencing or cessation of expression of the responsible gene.

Genes of tumor suppression

A newly discovered group of cancer-related genes are the so-called tumor suppression genes, sometimes referred to as antioncogenes, growth suppressors, or cancer suppression genes. Recent research suggests that mutations of this class, associated with loss of function, are likely to be involved in a high proportion of human cancer cases: More than a dozen good candidates for human tumor suppression genes have been identified for some types of human cancer. Tumor suppression genes involved in the pathogenesis of retinoblastoma (RB), breast and other carcinomas (p53), Wilm's tumor (wt 1, 2), and colon cancer (DCC) have been identified and cloned. Some aspects of their role in human tumorigenesis have been elucidated.

Retinoblastoma gene (RB) is a prototype tumor suppressor. The RB gene encodes a nuclear protein that is phosphorylated at both serine and threonine residues in a cell cycle-dependent manner (Lee et al., Nature 329: 642-645 (1987), Bunkovich et al.,

Cell, 58: 1097-105 (1989); Chen et al. Cell, 58: 1193-1198 (1989); DeCaprio et al., Cell, 58:

the mechanism by which RB operates in clarified. with many

1085-1095 (1989)). The molecular activity of the cell was not entirely The current model sticks to the fact that RB interacts with various proteins of the cell and that it can perform its functions through these complexes. When the function of the RB protein is to maintain the cell cycle in the G0 / G1 phase, then the RB must shut down or inactivate other proteins that are active and essential to enter the continuation of G1 (Lee et al., CSHSOB, LVI: 211-217 (1991)) . This closure hypothesis is consistent with recent observations that the important growth-promoting transcription factor E2F-1 is tightly regulated by the RB negative pathway (Helin et al., Cell, 70: 337-350 (1992), Kaelin et al., Cell, 70: 351-364 (1992), Shan et al., Mol Cell Biol., 12: 5650-5631 (1992), Helin et al., Mol Cell Biol., 13: 6501-6508 (1993). , Shan et al., Mol Cell Biol., 14: 229-309 (1994)). The currently published PTSG protein binds to the RB protein and is therefore involved in the regulation of mitosis.

The familial breast cancer gene, BRCA-1, was mapped on chromosome 17 q-21-22 by binding analysis. It is unclear whether this gene will act as a tumor suppressor or a dominant oncogene. The gene involved in human familial cancer syndrome, such as Li-Frameni syndrome, p53, appears to act as a classical suppressor; similarly, the RB gene is associated with a hereditary retinoblastoma (Knudson, 1993, supra).

Multiple steps and oncogenetic interplay

Between these two extremes of transforming oncogenes and the purely recessive genes of tumor suppression lie many other mechanisms that are apparently involved in the development of neoplastic changes characteristic of many human tumors. For many years, it has been assumed that most human tumors are likely to be due to more interactive genetic disorders, none of which alone is sufficient, but the sum of which is needed to develop the tumor. The actual roles of both cellular proto-oncogenes and growth-regulating tumor suppression genes in mammalian cell neoplasia are thought to represent a complex set of interactions between the two types of genes.

One current theory of cancerogenesis is that in some cancer pathologies, such as prostate adenocarcinoma, oncogenesis occurs through several genetic changes, each of which affects the expression or function of genes controlling cell growth or differentiation. (Nowell, P.C., Science 194: 23-28 (1976); Weinberg, R., Cancer Res. 49: 3713-3721 (1989)). Although adenocarcinoma ranks first in incidence and second in mortality among neoplasms in men (Coffey, D. S., Cancer 71:

880-886 (1993)), little is known about the molecular basis of this widespread disease. For example: Genetic changes in colon cancer have been extensively studied and a model has been proposed in which oncogen activation and loss of tumor suppression gene function correlate with the progressing clinical and histopathological changes observed during colorectal carcinogenesis (Fearon, ER and Vogelstein, Β. Cell 61, 759-767 (1990)). Indeed, it has been suggested that a similar process of progressing genetic change occurs in prostate cancer (Isaacs, W. B. and Carter, B.S., Cancer Surveys, Vol. 11, Cold Spring Harbor Laboratory, Cold.

Spring Harbor, NY, pp. 15-24 and the mechanism of change upon which (1991)), but the exact location should be based, remain unknown.

For known tumor genes, it has been shown that they are not primarily responsible for prostate cancer. For example: Mutations in a tumor gene, such as a ras oncogene or a tumor suppressor gene p53, were found only in a small fraction (<5%) of early prostate tumors (Carter et al., Proc. Natl. Acad. Sci. USA, 87: 8751-). 8755 (1990)), Gumerlock et al., Cancer Res. 51: 1632-1637 (1991); Bookstein et al., Cancer Res. 53: 3369-3373 (1993)), but p53 gene mutations were detected in 20-25% of late stage primary tumors, suggesting that the p53 gene could be involved in one of several alternative pathways for prostate tumor development (Bookstein et al. , Cancer Res 53: 3369-3373 (1993).

In order to find the genetic mechanisms responsible for prostate cancer, the determination of the karyotype or allelotype of tumor cells has also been used. Cytogenetic studies of short-term cultures of primary prostate cancer revealed some consistent chromosomal aberrations, such as deletion of lp, 7q, or 10q (Atkin, NB and Baker, MC, Hum. Genet., 70: 359-364 (1985), Cytogenet. 16: 301 304 (1985),

Cancer, 4: 16-24 (1992)), meanwhile loss of the allele indicated a slightly different set of often chromosomal regions. Carter et al., Proc. Nati.

Gibas et al. Cancer Genet. Lundgren et al. , Genes Chromium. what study lost

Acad. Sci. USA, 87: 8751-8755 (1990) first reported an accidental loss of chromosomes 10q and 16q, each in 30% of 28 tumors, and Kunny et al., Genomics 11: 530-536 (1991) showed losses in the same area as ip arms of chromosomes 8 and 10 with a frequency exceeding 50% in their set of 18 tumors.

Allelic loss of 8 p chromosomes is detected in 65% of prostate cancer cases, the highest frequency among all chromosome arms. These frequencies are comparable to the allelic loss of RB in retinoblastomas, 100% of which have an RB mutation, and suggest inactivation of the tumor suppression gene to 8β. Interestingly, a karyotypic deletion has been reported in sublines of the LNCaP cell line not responding to androgens. None of the so far cloned suppressor genes were located at 8p.

In a study by Bergerheim et al. (Bergerheim et al., Genes Chrom. Cancer, 3: 215-220 (1991), the alleles of the NEFL locus at chromosomes 8p12-p22 were lost in tumors from 7 out of eight informative patients, and alleles of the lipoprotein lipase locus (8p22 were lost in 6 out of seven patients) PLAT alleles (8p12-qll) alleles have been retained in some tumors that have lost more distal 8p loci, suggesting that the putative suppressor locus is located distally to the PLAT The most distal marker, D8S7, has been lost in 3 out of 6 tumors. the frequency of allelic loss of LPL and NEFL and the absence of observation of allelic loss starting distally from these loci further suggest that the suppressor locus may be relatively close to LPL and NEFL.

Therefore, in order to effectively diagnose the susceptibility to prostate cancer, related pathologies and other related cancers, and to treat, it is therefore necessary to identify and locate the site of the tumor suppression gene responsible for these pathologies. The present invention addresses this need and provides advantages associated with it.

SUMMARY OF THE INVENTION

The present invention is based on the discovery of a nucleic acid encoding a novel prostate / colon tumor suppression gene product (PTSG protein) that has the ability to suppress a tumor. The nucleic acid has been mapped in the p22 chromosome 8 region. Expression of PTSG product in normal prostate tissues and its loss in some cases of prostate and colon cancer supports its identification as a tumor suppressor gene. The newly discovered full-length cDNAs encode two new 348 and 347 amino acid proteins. The present invention has established for the first time that inactivation of PTSG or the PTSG product is responsible for prostate adenocarcinomas, colon cancer and related cancer pathologies as disclosed herein.

Diagnostic methods using this nucleic acid and PTSG are disclosed. In one embodiment, oligonucleotide fragments capable of hybridizing to the PTSG gene and assays utilizing them are provided. The number of nucleotides in these oligonucleotides may be as low as 5, but oligonucleotides consisting of about 20 to 30 nucleotides are preferred. These oligonucleotides may optionally be labeled with radioisotopes (such as tritium, ^Jz phosphorus, sulfur), enzymes (e.g. alkaline phosphatase and horseradish peroxidase), fluorescent compounds (e.g. fluorescein, Ethidium, terbium chelate) or chemiluminescent compounds (such as acridinium esters, isoluminol etc). Such or other labeling, such as that discussed in Non-isotopic DNA Probe Techniques, LJ Kricka, eds., Academic Press, New York, 1992 (incorporated herein by reference), may be used with oligonucleotides. They can be used in DNA probe assays in a conventional embodiment, such as Southern or Northern analysis. A description of such conventional embodiments can be found, for example, in Nucleic Acid Hybridization - A Practical Approach, BD Hames and SJ Higgins, eds., IRL Press, Washington, DC, 1985, incorporated herein by reference. Preferably, these probes are capable of hybridization under stringent conditions. Oligonucleotides can also be used as primers in polymerase chain reaction techniques, such as those described, for example, in PCR Technology,

H.A. Ehrlich, eds., Stockton Press, New York, 1989 and similar references.

According to the diagnostic method of the invention, the loss of wild-type PTSG is detected. The loss may be due to deletion or point-mutation events. PTSG alleles that are not deleted can be analyzed for the occurrence of point mutations, such as mutations that change meaning or shift the frame. Both of these types of mutations lead to non-functional PTSG products. In addition, instances of point mutations can occur in regulatory regions, such as the PTSG promoter, and result in the loss or decrease of PTSG mRNA expression.

To detect loss of wild-type PTSG in tissue, it is useful to isolate tissue from surrounding normal tissues. Means for enriching a tissue preparation by tumor cells are known in the art. For example, the tissue may be isolated from paraffin or cryostat sections. Cancer cells can be separated from normal and flow cytometry. These as well as other techniques for separating tumor cells from normal are well known in the art. When tumor tissue is highly contaminated with normal cells, detection of mutations is more difficult.

Detection of point mutations can be accomplished by molecular cloning of the PTSG allele (or alleles) present in the tumor tissue and sequencing the allele (s) using techniques well known in the art. Alternatively, a polymerase chain reaction can be used to amplify PTSG sequences directly from genomic DNA preparations from tumor tissue. The polymerase chain reaction itself is well known in the art. See, eg, Saiki et al., Science, 239: 487 (1988); U.S. Pat. U.S. Patent 4,683,203 and U.S. Pat. No. 4,683,195.

Specific PTSG deletions can also be detected. For example: Probes for restriction fragment length polymorphism (RFLP) of the PTSG gene or flanking marker genes can be used to assess the loss of the PTSG allele. Other techniques for detecting deletions as known in the art may be used.

The loss of wild-type PTSG may also be detected by loss of expression of the wild-type PTSG product. Such expression products include both the mRNA and the PTSG protein product itself. Point mutations can be detected by sequencing mRNA directly or through molecular cloning of cDNA generated from mRNA. The sequence of the cloned cDNA can be determined using DNA sequencing techniques well known in the art. The cDNA can also be sequenced by polymerase chain reaction (PCR), which will be discussed in more detail below.

Alternatively, mismatch detection can be used to detect point mutations in PTSG or its mRNA product. While these techniques are less sensitive, they are easier to perform with a large number of tumors. An example of such a splitting technique is the RNase protection method described in detail in Winter et al., Proc. Nati. Acad. Sci. U.S.A., 82: 7575 (1985) and Mayers et al., Science 230: 1242 (1985). In the practice of the invention, this method involves the use of a labeled RNA probe that is complementary to human wild-type PTSG. The ribon probe and either the mRNA or DNA isolated from the tumor tissue are hybridized together and subsequently digested with the RNase A enzyme, which is able to detect certain mismatches in the RNA duplex structure. When RNase A mismatch is detected, the enzyme cleaves at this site. Thus, when the hybridized RNA preparation is separated from the electrophoretic gel matrix and the mismatch has been detected and cleaved by RNase A, the RNA product appears to be smaller than the full-length RNA duplex for the ribosonde and PTSG mRNA or DNA. The ribon probe may not be a full-length PTSG mRNA or gene, but may only be a segment of some. When the ribon probe contains only a PTSG mRNA or gene segment, it will be desirable to use multiple such probes to analyze mismatches throughout the mRNA sequence.

Similarly, they can be used to detect mismatches of DNA probes and enzymatic or chemical cleavage. See, eg, Cotton et al. , Why. Nati. Acad. Sci. U.S. Pat. A., 85: 4397 (1988) and Shenk et al. , Why. Nati. Acad. Sci. U.S. Pat. A., 72: 989 (1975). Alternatively, mismatches can be detected by shifts in the electrophoretic mobility of mismatched duplexes compared to mismatched correctly. See, eg, Cariello Human Genetics, 42: 726 (1988). Whether with ribosomes or DNA probes, cell mRNA or DNA that may contain mutations can be amplified prior to hybridization by PCR (see below).

PTSG DNA sequences from tumor tissue that have been amplified using a polymerase chain reaction can also be evaluated using allele-specific probes. These probes are oligomeric nucleic acids each containing a DNA sequence region of the PTSG sequence containing a single known mutation. For example: A single oligonucleotide may be about 30 nucleotides in length and correspond to a portion of the PTSG DNA sequence. At the position that encodes the 175th codon, this oligomer encodes alanine instead of the wild-type valine codon. Using a set of such allele-specific probes, PCR amplification products can be evaluated and identified for the presence of previously identified PTSG mutations. Hybridization of allele-specific probes with the amplified PTSG sequence can be performed, for example, on a nylon filter. Hybridization to a particular probe indicates the presence of a mutation in the tumor tissue as in the allele-specific probe.

The kit of the invention is useful for determining the nucleotide sequence of PTSG using a polymerase chain reaction. The kit contains a set of single-stranded DNA primer pairs that can be annealed to sequences within or around the PTSG to begin (to prime) DNA synthesis of PTSG itself. The full set allows synthesis of all nucleotides of the PTSG coding sequence. The primer set may or may not allow the synthesis of both intron and exon sequences: However, it must allow the synthesis of all exon sequences.

The present invention is also directed to administering a tumor suppression gene of PTSG or a wild-type PTSG protein to suppress, eradicate or reverse the neoplastic phenotype of established cancer cells having no endogenous wild-type PTSG protein. The wild-type PTSG gene can be used to suppress or reverse the neoplastic type or properties of established human tumor cells lacking the wild-type PTSG protein. Such suppression of the neoplastic phenotype consequently suppresses or eradicates the abnormal mass of such cancer cells, i.e., the tumor, which in turn may reduce the burden on the animal by such tumors, which may in turn increase the survival of the treated animals. Neoplastic properties that have been monitored and reversed include morphology, cancer cells, tumorigenicity of the PTSG protein in the animal, and most importantly lacking normal. Thus: Reducing the burden of cancer cells is due to suppression of the neoplastic phenotype due to administration of the wild type PTSG tumor suppression gene product. It is understood that the neoplastic phenotype refers to phenotypic changes in cellular characteristics such as morphology, growth rate (i.e., doubling time), saturation density, soft agar colony formation, and tumorigenicity.

The invention therefore provides vectors that encode PTSG and PTSG proteins for use in the treatment of tumors or cancer, and methods of making PTSG proteins and vectors suitable for use in methods of treatment. The invention also provides methods for testing molecules that bind and affect PTSG.

The invention also provides methods of treating mammals, such as humans, as well as methods of treating abnormally proliferating cells, such as cancer cells, such as prostate tumor cells, colon cancer, and other tumor cells, and for suppressing the neoplastic phenotype. Broadly, the invention contemplates the treatment of abnormally proliferating cells, or mammals having a disease characterized by abnormally proliferating cells, by any suitable means known to allow entry of a vector encoding PTSG and compatible with host cells, or a PTSG protein derivative, into cells to be treated to achieve suppression of one or more characteristics of the neoplastic phenotype or suppression of proliferation.

In one embodiment, the invention includes a method of treating a disease characterized by abnormally proliferating cells in a mammal by administering an expression vector encoding PTSG to a mammal having a disease characterized by abnormally proliferating cells, introducing the expression vector into the abnormally proliferating cells and expressing PTSG in the abnormally proliferating cells in an amount effective suppressing proliferation of these cells. The expression vector is introduced into abnormally proliferating cells by viral infection or transduction, liposome-mediated transfection, polybrene-mediated transfection, CaPO _4- mediated transfection, or electroporation. If necessary, treatment is repeated.

In another embodiment, the invention encompasses a method of treating abnormally proliferating cells in a mammal by introducing an expression vector encoding PTSG into the abnormally proliferating cells and expressing PTSG therein in an amount effective to suppress proliferation of these cells. If necessary, treatment is repeated.

In another alternative embodiment, the invention provides a DNA molecule capable of suppressing the growth of abnormally proliferating cells. An example of a prostate / colon tumor suppressor protein is the PTSG protein product having an amino acid sequence substantially according to SEQ. id. No. 1, see Figure 12. In a more preferred embodiment, the DNA molecule has the DNA sequence of SEQ. id. No. 1 and is expressed by an expression vector. The expression vector may be any vector compatible with the host cells.

Preferably, the vector is selected from the group consisting of retroviral vectors, adenoviral vectors and herpesviral vectors. In another, more preferred embodiment, the DNA molecule has the DNA sequence of SEQ. id. No. 2 (see Figure 14) and is expressed by an expression vector. The expression vector may be any vector compatible with the host cells. Preferably, the vector is selected from the group consisting of retroviral vectors, adenoviral vectors and herpesviral vectors.

In another alternative embodiment, the invention provides a PTSG protein product having an amino acid sequence substantially according to SEQ ID NO: 2. id. No. 2, and biologically active fragments thereof. In yet another alternative embodiment, the invention provides a PTSG protein having an amino acid sequence substantially according to SEQ. id. No. 4 (see Figure 14), and biologically active fragments thereof.

In another alternative embodiment, the invention provides a method of producing a PTSG protein product by the steps of: introducing a compatible expression vector containing a gene encoding PTSG into a host cell and causing expression of the PTSG protein by the host cells.

In another alternative embodiment, the invention comprises a method of treating abnormally proliferating cells of a mammal ex vivo by the steps of: collecting a tissue sample in need of treatment from the mammal, wherein the tissue sample comprises abnormally proliferating cells; contacting the tissue sample in need of treatment with an effective dose of an expression vector encoding PTSG; expressing PTSG in abnormally proliferating cells in amounts effective to suppress proliferation of abnormally proliferating cells. The treatment is repeated as necessary and the treated tissue sample is returned to the original or other mammal. Preferably, the tissue treated is ex vivo tissue of blood or bone marrow.

In another alternative embodiment, the invention encompasses a method of treating a disease characterized by abnormally proliferating cells in a mammal by a method comprising the steps of administering PTSG protein to a mammal having a disease characterized by abnormally proliferating cells by introducing the PTSG protein into abnormally proliferating cells in an amount effective to suppress proliferation of these cells. In a preferred embodiment, for insertion into the cells to be treated, the PTSG protein fragments or derivatives thereof are encapsulated into liposomes. Treatment is repeated as necessary.

Brief description of the pictures

Figure 1 shows KSR2 (8p22) Southern analysis in human prostate cancer. Paired purified prostate cancer DNA (T) with non-cancerous DNA (N) from the same patients.

The 1.9-kilobase allele is lost in tumor tissue of patient 4; The 3.3-kilobase allele is lost in the tumor tissues of patients 5 and 6. Patient 7 is not informative at this locus.

Figure 2 shows the percentage of prostate cancers with loss at the loci studied on chromosome 8.

Figure 3 shows the homozygous deletion of MSR in human prostate cancer. Primary tumor 23 retained both alleles in D8S201, is non-informative in D8S163, lost the 6.3-kilobase allele in MSR, and is non-informative in D8S39. Metastasis tumor N2 lost one allele in D8S201 and D8S163, while showing complete loss of sequences in MSR. New probing of the same transfer filter (blot) with 15-65 on DCC (18q) gave a strong signal at 8 kilobases (kb) and showed the presence of high molecular weight DNA in the tumor pathway. Both D8S39 alleles are present in the N2 tumor and the intensity of the lower allele is three-fold. Image legend: bp, couples crave. For definitions of T and N, see the legend to Figure 1.

Figure 4 shows a deletion map in human prostate cancer. Only tumors that show a loss of chromosome 8p are shown. Samples 1-27 are primary tumors. Legend of the figure: N1 - N5 are metastatic prostate tumors. O, conserved alleles, loss of heterozygosity, X, homozygous deletions.

Figure 5. Artificial yeast chromosome and radiation hybrid map of loci in chromosome band 8p22, common areas of allelic loss in multiple human cancers. Genomics 24: 317-323.

Figure 6 shows the homologous integration of the conversion vector, resulting in amplification of the 1855 bp band.

«

Figure 7 shows Southern blotting of yeast DNA with a radiolabeled probe of the hygro gene, confirming the presence of the hygro ^R gene on the YAC arm.

Figure 8. Long-range artificial yeast chromosome (YAC) restriction map containing markers on the 8p22 chromosome band. DNA from YACs 946_c_9, 877_f_2, 932_e_9, 766_a_12 incorporated into the agarose beads (beads) were digested with various rarely cleaving restriction enzymes (A: Asc I, M:

Mlu I, N: Not I, Nr: Nru I, Sf: Sfi I) and separated PFGE as described in the Methods. Southern blot with selected cDNA (italic) probes and genomic restriction fragments Probes in which N2 tumor (Figures and derived minimum (740 kb, thin line) and maximum (920 kb, thick line) deletion range in this tumor is shown above) As shown, the N33 gene is located within this deletion.

Figure 9. Nucleotide sequence and selected restriction sites of the plasmid insert of pBS-N33C (7), derived by cloning the 1.3 kb EcoRI-EcoRI insert from the lambda phage clone \ N33C (SEQ ID NO: 5), obtained by library screening human placenta cDNA selected by the cDNA probe N33, into plasmid pBluescript. Selected restriction sites are shown. The sequence of the first -20 bp containing the Not I site was predicted to be introduced artificially during the construction of the cDNA library.

DNA (antiquated) was performed to identify containing any of the probes (parentheses), found to be homozygous deleted 3 and 4) or in bold.

Figure 10. Annotated double stranded sequence of N33 cDNA derived from sequencing of phage clone N33C (7) and RT-PCR of clones A4 and A5. The 65 bp segment from nt 1186 to 1250 N33C (7) and the A4 sequence are not present in clone A5, so that clones N33C (7) and A4 represent mRNA of longer form 1, while A5 represents mRNA of shorter form 2. using either of the two translation stop signal sites as indicated. The predicted translation start site is also shown and preceded by a stop codon in the same frame (*).

Figure 11. Open reading frame map, ORF, for clone N33C (7), representing mRNA of Form 1. The longest ORF is nt 158-1202.

Figure 12. Translation of the longest ORF from Form 1 mRNA (SEQ ID NO: 1). The predicted 348 amino acid polypeptide has a molecular weight of 39,674.13 daltons (SEQ ID NO: 3). The last 5 amino acids differ from the Form 2 polypeptide.

Figure 13. ORF Map for Derived Form 2 mRNA. The longest ORF is nt 158 - 1199.

Figure 14. Translation of the longest ORF from Form 2 mRNA (SEQ ID NO: 2). The predicted 347 amino acid polypeptide has a molecular weight of 39,556.18 daltons (SEQ ID NO: 4). The last 4 amino acids deviate from the Form 1 polypeptide.

Figure 15. Alignment of N33 polypeptides of Forms 1 and 2 with a hypothetical 33.7 kD protein encoded by ORF ZK686.3 from C. elegans. Three gaps were introduced into the N33 to optimize alignment. 42% of residues are identical between human and C. elegans (underlined). The protein encoded by the ORF of ZK686.3 has a molecular weight of 33.7 kD.

Figure 16. Northern transfer of mRNA from normal human tissues (Clontech), hybridized with selected cDNA probes J2, J28 and N33. N33 mRNA is about 1.5 kb in size and is expressed in most tissues, including heart, placenta, lung, liver, pancreas, prostate, testes, ovaries and colon. Expression in the spleen, thyroid, small intestine and peripheral lymphocytes was low.

Figure 17. Northern transfer of mRNA from human tumor cell lines hybridized with selected cDNA probes N33, P10, J2 and P16. Actin was used to control mRNA loading. N33 expression was not detected in 13 of 14 colorectal carcinoma cell lines (SW480, SW837, SW1417, HT-29, SW403, LS174T, DLD-1,

CACO-2, EB, SK-CO-1, RKO, HCT116 and COLO-302).

Figure 18. Northern transfer of mRNA from PPC-1, WI-38, H460, A549 tumor lines (lanes 1-4), normal colon mucosa (lane 5), and the SW837 and SW840 colon tumor lines (lanes 6 and 7). N33 mRNA is expressed in mucosa excised from the colon.

Figure 19. RT-PCR assay for N33 expression in RNA from nine prostate cancer samples (lanes 1-9). C: PCR control. N33 primers were N33GEX-f and -r. Primers for p53, Rb, and G3PD genes were used as RNA / cDNA quality controls. N33, Rb, and p53 primers lie within the boundaries of exons and do not specifically amplify genomic DNA. Significantly reduced expression of N33 was observed in cases 3, 6 and 9. In tissues expressing N33, mRNA is also visible above

1) and below (Form 2).

Figure 20. Predicted N33 polypeptide form of Form 1. The conserved 16 C-terminal amino acids (framed) were attached to KLH and used to generate a rabbit polyclonal antibody.

Figure 21. Antibody recognition of fusion protein N33-glutathione-S-transferase in E. coli. N33 RT-PCR products from placental mRNA (primers N33GEX-f and -r) were cloned into pGEX-2T (Pharmacia). Clones A4 and 6, respectively, were isolated. A5, which represent mRNA forms 1 and 2, respectively. Form 2. Protein expression was induced by IPTG and cell lysates were resolved by PAGE and transferred to membrane. In western analysis, incubation with affinity purified polyclonal antibody against peptides N33 was used, and reactive bands were visualized with alkaline phosphatase conjugated to secondary antibody and NBT / BCIP substrate. A 57 kD fusion protein band was detected in induced cells containing clone A4 but not A5 or other clones.

DETAILED DESCRIPTION OF THE INVENTION

The present invention provides a novel gene that encodes a protein referred to as a PTSG protein. PTSG refers to two proteins: one consisting of 348 amino acids and the other consisting of 347 amino acids, each having a molecular weight of approximately 40 kD.

As used herein, nucleic acid will mean single stranded genomic DNA, cDNA, mRNA, and cRNA. The term isolated when used for a nucleic acid refers to a nucleic acid deprived of at least a portion of the molecules associated with or occurring in the natural environment.

The invention also provides a recombinant expression vector or a recombinant replication vector comprising a molecule corresponding to a suppressor gene of a cell, eg, a bacterial cell, an isolated nucleic acid, a tumor, as well as a host comprising these vectors.

Treatment of human diseases by gene transfer has moved from theoretical to practical. The first trial of human gene therapy began in September of adenosine deaminase (ADA) into a lethal defect of this

1990 and involved the transfer of a lymphocyte gene to a patient who otherwise had an enzyme causing immune deficiency. The results of this initial test were very encouraging and contributed to the stimulation of further tests (Culver, K.W., Anderson, W.F., Blaese, R. R.., Hun. Gene Ther., 2: 107 (1991)).

Most human gene transfer assays so far rely on retroviral vectors for gene transduction. Retroviral vectors in this context are retroviruses from which all viral genes have been removed so that no viral proteins are produced in the cell infected with the vector. Viral replication functions are provided by the retrovirus packaging cells used, which produce all viral proteins but do not produce infectious virus. Introduction of retroviral vector DNA into packaging cells results in the production of virions that are not vector RNA, which is capable of infecting target cells, but no further spread of the virus upon infection. To distinguish this process from natural infection, where the virus continues to replicate and spread, the term transduction is used rather than infection.

For illustration purposes only: The nucleic acid delivery system is a replication incompetent retroviral vector. As used herein, the term retroviral includes, but is not limited to, a vector or delivery device that has the ability to selectively target and introduce a nucleic acid into a dividing cell. As used herein, the term replication incompetent is defined as the inability to produce viral proteins, which excludes the spread of this vector in infected host cells.

Another example of a replication incompetent retroviral vector is LNL6 (Miller, A. D. et al., BioTechniques 7: 980-990 (1989)), incorporated herein by reference. Methodology of using replication incompetent retroviruses to mediate gene transfer

retroviruses is good introduced (Correll , P. H. et al. , Why. Nati. Acad. Sci. U.S. Pat. A., 86: 8912 (1989); Borningnon, C. et al., Nati. Acad. Sci. U.S. Pat. A., 86: 8912 - 8952 (1989); Culver, K. et al. , Why. Nati Acad. Sci. U.S. Pat. AND . , 88: 3155 ; Rill, D. R. et al. , Blood 79 (9): 2694-2700 (1991) , everything included by reference.

Clinical investigations have shown that the harmful effects associated with viral vectors are few or none (Anderson, Science 256: 808-813 (1992)).

The main advantages of retroviral gene therapy vectors are the high efficiency of gene transfer into replicating cells, the precise integration of the transferred genes into cellular DNA, and the absence of further sequence propagation after gene transduction (Miller, A. D., Nature, 357:

overlapped with auxiliary

455-460 (1992)). The potential for producing a replication-competent (helper) virus in the generation of a retroviral vector remains a matter of caution, although for practical purposes this problem has been solved. To date, all FDA-approved retroviral vectors have been generated using PA317 amphotropic retrovirus packaging cells (Miller, A. D. and Buttimore, C., Molec. Cell Biol., 6: 2896-2902 (1986)). The use of vectors having little or no viral sequences in PA317 cells eliminates viral production even in stringent assays that allow the amplification of such cases (Lynch, C. M., and Miller, A. D., J. Virol. 65: 3887-3890 (1991)). Other packaging cell lines are available. For example: Cell lines designed to divide different retroviral coding regions into different plasmids should reduce the possibility of helper virus generation by recombination. Vectors produced in such packaging cells may also provide an effective system for human gene therapy (Miller, A. D., Nature, 357: 455-460 (1992)).

Non-retroviral vectors have been contemplated for use in genetic therapy. One such alternative is adenovirus (Rosenfeld, MA et al., Cell 68: 143-155 (1992); Jaffe, HA et al., Proc. Natl. Acad. Sci. USA, 89: 6482-6484 (1992)). The main advantages of adenoviral vectors are the potential to carry large stretches of DNA (36 kb genome), very high titer (10 -? ¹ ml ^-1 ), the ability to infect tissues in situ, especially in the lungs. So far, the most surprising use of this vector is the delivery of the human cystic fibrosis (CFTR) transmembrane conductivity regulator gene by intratracheal instillation on the airway epithelium of cotton rats (Rosenfeld, MA et al., Cell 68: 143-155 (1992)). Similarly, herpes viruses can become valuable for human gene therapy (Wolfe JH et al., Nature Genetics 1: 379-384 (1992)). Other suitable viral vectors can of course be used for the genetic therapy of the invention.

Another method of gene transfer that has been approved by the FDA for use in humans is to transfer plasmid DNA in liposomes directly to human cells in situ (Nabel, E. G. et al., Science 249: 1285-1288 (1990)). Plasmid DNA should be in human gene therapy since it is readily released for use, unlike retroviral vectors, it can be purified to homogeneity.

In addition, liposome-mediated delivery has been shown to be promising in human gene therapy by some other methods of physical DNA transfer, such as targeting DNA to receptors on cells by introducing plasmid DNA into protein complexes (Wu, GY et al., J. Biol. Chem 266: 14338-14342 (1991), Curiel, DT et al., Proc Natl Acad Sci USA, 88: 8850-8854 (1991)).

The PTSG of the invention can be placed by methods well known in the art into an expression vector, such as a plasmid or viral expression vector. The plasmid expression vector can be introduced into tumor cells by calcium phosphate transfection, liposome-mediated transfection (e.g., LIPOFECTIN), DEAE dextran-mediated transfection, polybrene-mediated transfection, electroporation, and any method for introducing DNA into the cell.

The viral expression vector can be introduced into the target cell in an expressible form by infection or transduction. Such viral vectors include, but are not limited to, retrovirus, adenovirus, herpes virus and avipox virus. When PTSG is expressed in any abnormally proliferating cell, the replication cycle of that cell is inhibited, resulting in cell aging and death, and ultimately reducing the mass of abnormal tissue, i.e., tumor or cancer. A vector capable of introducing a gene construct into a target cell and expressing its H-NUC in a cell in an amount that suppresses proliferation can be administered by any effective means.

For example: A physiologically acceptable solution containing an effective concentration of active vectors may be administered topically, intraocularly, parenterally, orally, intranasally, intravenously, intramuscularly, subcutaneously, or by any other effective means. In particular, the vector may be directly injected into the target cancer or tumor tissue by a needle, in an amount effective to treat the tumor cells of the target tissue.

Alternatively, a cancer or tumor present in the body cavity, such as the eyes, the gastrointestinal tract, the genitourinary tract (e.g., bladder cancer), the pulmonary or bronchial system, and the like, may receive a physiologically acceptable formulation (e.g., solution such as saline or phosphate) buffer, suspension, or emulsion, which is, except for the vector, sterile) containing an effective concentration of the active vectors by direct injection via a needle or catheter or other delivery tube placed in a hollow organ affected by cancer or tumor. Any effective imaging device, such as X-ray, sonograph or fiber optic visualization, can be used to locate the target tissue and guide the needle or catheter tube.

In another alternative, a physiologically acceptable solution containing an effective concentration of active vectors can be administered systemically into the bloodstream to treat a cancer or tumor that is directly unavailable or anatomically isolated.

In yet another alternative, tumor or cancer cells can be treated by introducing PTSG protein into the cells by any known method. For example: Liposomes are artificial membrane vesicles that are available for introduction of drugs, proteins and plasmid vectors both in vitro and in vivo (Mannino, RJ, et al., BioTechniques 6: 682-690 (1988)) into target cells (Newton, AC and Huestis, W. Η., Biochenistry 27: 4655-4659 (1988), Transwell, AK et al., Biochinica et Biophysica Acta 1044: 269-274 (1990); Ceccoll, J. et al., Journal of Investigative Dernatology 93: 190-194 (1989)). Thus, the PTSG protein can be encapsulated with high efficiency in liposome vesicles and delivered to mammalian cells in vitro or in vivo.

Liposome-sealed PTSG protein can be administered topically, intraocularly, parenterally, orally, intranasally, intratracheally, intrabronchially, with doses effective to treat abnormally proliferating cells of the target tissue. The liposomes may be administered in any physiologically acceptable formulation comprising an effective concentration of the sealed PTSG protein.

The host-vector system refers to host cells that have been transfected with vectors constructed using recombinant DNA techniques. The introduction of the vector or DNA can be accomplished by microcell transfer, retrovirus-mediated gene transfer, transfection, cell fusion, etc. The vectors and methods disclosed herein are suitable for a wide range of prokaryotic and eukaryotic organisms. In addition, the invention provides a method for transforming a cell by contacting the cell with a vector or DNA of the invention under appropriate conditions.

Reference is made to standard molecular biology textbooks that include definitions, methods, and means for practicing the basic techniques of the present invention. See, for example, Maniatis et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, New York (1989) and various references cited therein. This reference and the cited publications are expressly incorporated by reference in this specification.

In addition, recombinant DNA methods commonly used by those skilled in the art include a polymerase chain reaction which, when associated with oligonucleotide synthesis, allows for a slight increase in DNA sequences. A stretch of DNA up to approximately 6000 base pairs in length allows exponential amplification by PCR, starting from as few as a single copy of the gene. In this technique, a sample of denatured DNA is incubated with two oligonucleotide primers that drive complementary strand synthesis dependent on DNA polymerase. Each of the multiple cycles yields approximately double the amount of the target sequence. Each cycle is controlled by temperature changes that allow denaturation of DNA strands, hybridization with primers, and synthesis of new DNA strands. The use of thermostable DNA polymerase eliminates the need to add a new enzyme for each cycle and thus allows complete automation of DNA amplification. Twenty-five amplification cycles will increase the amount of target sequence approximately 10 ⁴ times. PCR technology is the subject of U.S. Pat. Nos. 4,683,195, 4,800,159, 4,754,065 and 4,683,202.

It is understood that limited modifications of the tumor suppressor gene sequence of the invention can be made without impairing its biological function and that only part of the entire primary structure is required for activity. It is further understood that minor modifications of the amino acid sequence may give rise to proteins having substantially the same or enhanced function as compared to the molecule contained in the vector pBS-N33c (7). These modifications may be deliberate, such as by directed mutagenesis, or random, such as by mutations in hosts. All these modifications are included as long as tumor suppression function is maintained. Other unique nucleic acid fragments of at least 10 nucleotides are useful as hybridization probes. These probes are useful for detecting cancer predisposition caused by malfunction of this gene.

Isolated nucleic acid fragments are also useful for generating new peptides. These peptides are in turn useful as immunogens for generating polyclonal and monoclonal antibodies useful in the diagnostic methods outlined below. Methods for preparing and using probes and immunogens are well known in the art and are briefly described below.

Also encompassed by the present invention are nucleic acid molecules that hybridize under stringent conditions to an isolated nucleic acid molecule that encodes this tumor suppression protein, such hybridizing nucleic acid molecules or probes can be prepared, for example, by priming. For a methodology for preparing such fragments, see Sambrook et al. (Sambrook et al. Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, NY., pp. 1.98-1.104 (1989)).

Purified tumor suppressor polypeptide or protein are also provided by the present invention. These polypeptides or proteins are useful for the production of antibodies, which in turn are useful for diagnosis. They can be produced by recombinant expression of an isolated nucleic acid molecule of the invention using well known molecular biology techniques.

Purified, when used to describe the state of a protein, polypeptide or antibody, refers to such a protein deprived of a portion of other proteins and molecules normally associated with it in the natural environment, or occurring with a tumor suppression polypeptide, protein, or antibody. As used herein, the term native refers to a form of a protein, polypeptide, antibody, or fragment thereof that is isolated from nature or that is free from deliberate amino acid substitutions.

As used herein, the term antibody or immunoglobulin refers to a protein that is produced in response to an immunization with an antigen and specifically reacts with the antigen. This includes polyclonal as well as monoclonal antibodies. The purpose of this definition is to include human and mammalian, for example, murine, rat, rabbit and sheep antibodies. The predominant human generated antibodies, which account for about 80% of total serum, were grown in LB medium diluted from 1:10 to 1: 150, with and purified by electrophoresis of antibodies, are of IgG isotype and have two light and two heavy chains linked by disulfide bonds.

Anti-tumor suppressor antibodies can be generated as follows. The insert DNA fragments in pBS-N33c (7) were fused with glutathione-S-transferase protein DNA. The fusion proteins were then expressed in E. coli. Transformed E. coli cells plus ampicillin. The culture mixture was preferably 1: 100, LB medium with ampicillin added. A procedure for the construction of recombinant plasmids is described in Sambrook et al. (Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, NY.).

1.98-1.104 (1989)). Fusion of fragments into vector frames at restriction enzyme sites is described in Proc. Nati. Acad. Sci. U.S.A., 83: 4685-4689 (1986).

Using the procedures described above for fusing GST to PTSG DNA fragments, an amount of fused protein by preparative SDS / polyacrylamide gel was prepared according to the procedure described in Sambrook et al.

(Sambrook et al. Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.)

1.98-1.104 (1989)) and in Harlow and Lane (Harlow and Lane, Antibodies, A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY., (1988)). The fusion protein is eluted by overnight extraction and SDS. Soluble acrylamide can be removed by dialysis. The proteins are then concentrated. The purified fusion protein is useful as an antigen to generate specific anti-PTSG antibodies.

Rabbits may receive repeated injections, preferably at 14-day intervals, 1-20 μ9, preferably 10 μg, purified fusion protein mixed with complete Freund's adjuvant (initial injection) and then receive booster injections of the same fused protein in incomplete Freund's adjuvant (repeated injections) ). The complete Freund's adjuvant generally consists of an antigen emulsion, in this case a fusion protein, in saline, and a mixture of emulsifying agents such as Arlacel A, in mineral oil with dead mycobacteria. Incomplete Freund's adjuvant is the same except that it has no mycobacteria.

does not detect a sufficiently high response with both GST and fused for two months. For enrichment

Injections are repeated until two or more affinity columns can be prepared using the method generally described in Harlow and Lane (Harlow and Lane, Antibodies, A), which are approximately antibodies that recognize only determinants of prostate tumor proteins. Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1988). At least one column is coupled to glutathione-S-transferase (GST) and at least one column is fused with protein. Both columns are appropriately pre-cycled. The antibody is loaded first onto a Sepharose fusion protein column and eluted with glycine buffer pH 2.3. The eluate is neutralized by passing through the GST column several times to remove antibodies specifically directed against GST. The purified prostate tumor suppressor antibody protein is useful for immunoprecipitation, immunochemical staining, for localizing the prostate tumor suppressor protein, and will also be useful in diagnostic identification of PTSG in mammalian and human tissue samples. Thus, these purified proteins are also within the scope of this invention. They may be labeled with detectable labels such as radioisotopes, dyes, enzymes, and biotin.

The above methods can be modified using standard procedures shown, for example, in Harlow and Lane (Harlow and Lane, Antibodies, A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY., (1988)).

Fused proteins can also be used to generate monoclonal antibodies. Thus, the invention provides a monoclonal antibody directed to an epitope on a prostate tumor suppressor protein or polypeptide. In one embodiment of the invention, the monoclonal antibody is a murine monoclonal antibody. In another embodiment of the invention, the monoclonal antibody is a human monoclonal antibody.

To isolate murine monoclonal antibodies, eight week old mice can be injected interperitoneally with 50 micrograms of purified prostate tumor suppressor polypeptide (prepared as described above) in complete Freund's adjuvant in 1: 1 volumes. Mice are then boostered with these polypeptide booster in incomplete Freund's adjuvant at monthly intervals, and allowed to bleed through the tail vein. On days 4.3 and 2 prior to fusion, mice receive an intravenous booster of 50 micrograms of polypeptides in saline. Splenocytes are fused with non-secreting myeloma cells according to the procedures described and known to those of ordinary skill in the art to which this invention belongs. A little later, after about two weeks, hybridoma supernatants are evaluated for binding activity against prostate tumor polypeptides as described herein below. Positive clones are isolated and propagated.

In addition, the invention also provides the monoclonal antibody described above conjugated to a therapeutic composition. For the purposes of this invention, suitable therapeutic agents include, but are not limited to, therapeutic agents selected from the group consisting of radioisotopes, toxins, toxoids, and chemotherapeutic agents. The invention also provides the monoclonal antibody described above conjugated to a detectable marker. Suitable detectable markers include, but are not limited to, enzymes, radioisotopes, dyes, and biotin. The present invention further provides a monoclonal antibody as described above conjugated to an imaging agent. Suitable imaging agents include, but are not limited to, radioisotopes such as P, S

Also provided by the present invention are pharmaceutical compositions comprising a purified prostate tumor suppressor polypeptide or protein described above, alone or conjugated to any of the following, a detectable marker, therapeutic agent or imaging agent as described above and a pharmaceutically acceptable carrier. Further provided are pharmaceutical compositions comprising the monoclonal antibody described above, alone or conjugated to any of the following: A detectable marker, therapeutic agent or imaging agent. As used herein, the term pharmaceutically acceptable carrier encompasses any of the standard pharmaceutical carriers such as phosphate buffered saline, water, emulsions such as oil / water emulsions, and various types of humectants.

As used herein, the antibody also covers antibody fragments. Antibody fragments retain at least some ability to bind selectively to their antigen. Also encompassed by this invention are antibody fragments that have been recombinantly or chemically synthesized and retain the ability to bind antigen corresponding to a native antibody. The ability to bind with an antigen or hapten is determined by antigen binding assays known in the art as antibody capture assays (see, e.g., Harlow and Lane (Harlow and Lane), Antibodies, A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 1988))). Antibody fragments that retain some binding affinity include, but are not limited to: Fab (a fragment that contains a monovalent, antigen-binding fragment of an antibody molecule, generated by digestion with a papain enzyme to yield an intact light chain and part of the heavy chain); Fab '(a fragment of the antibody molecule obtained by treatment with pepsin followed by reduction to obtain an intact light chain and part of the heavy chain; two Fab' fragments are obtained from one antibody molecule); (Fab ') ₂ , an antibody fragment obtained by the action of the enzyme pepsin, but without subsequent reduction; (Fab ') ₂ is a dimer of two Fab' fragments held together by two disulfide bonds; Fv and single chain antibodies (SCA). Also within the scope of this invention are grafted and chimeric CDRs that retain the ability to bind prostate tumor suppressor protein.

As used herein, the term chimeric antibody refers to an antibody in which the variable regions of antibodies derived from one (biological) species are combined with constant regions of antibodies derived from another species. Chimeric antibodies are constructed by recombinant DNA technology and are described, for example, in Shaw et al., J. Inonun. 131: 4538 (1987); Sun, L.K. et al., Proc. Nati. Acad. Sci. U.S. Pat. A., 84: 214-218 (1987); Neuberger, M.S. et al., Nature 314: 268 (1985); Boulianne, G.L. et al. Nature 312: 643-646 (1984) and Morrison, S. L. et al., Proc. Nati. Acad. Sci. U.S. Pat. A., 81: 6851-6855 (1984).

As used herein, the term CDR grafted antibody refers to an antibody having an amino acid sequence, European Patent Publication No. 3, which provides at least a portion of one or more CDRs (complementarity defining) sequences in a light or variable domain replaced by analogous portions CDR sequences from an antibody having a different binding specificity for a given hapten or antigen. Analogous CDR sequences are said to be inoculated on a substrate or recipient antibody (see 0 239 400). The donor antibody CDR sequence, and the antibody that substitutes for the sequence, is a substrate antibody.

The present invention provides a method by which the presence or absence of a protein whose absence is associated with neoplasm is detected in a sample. For protein detection, the method will include staining cells with polyclonal or monoclonal antibodies raised against the protein. For example, the method comprises the step of obtaining a suitable sample from the subject. Suitable samples include, but are not limited to, prostate tumor tissue, colon tumor tissue, lymph node tissue, and bone marrow cells. The method requires contacting a sample and an agent specifically unique to the tumor suppressor protein under conditions that favor complexing with the agent and then detecting any complex formed. The absence of the complex indicates the absence of a protein that is associated with a neoplastic condition such as prostate adenocarcinoma. Thus: This method is useful for diagnosing prostate adenocarcinoma. For the purposes of this invention, suitable labeling agents are radioisotopes such as ³² P, ³⁵ S and ¹³¹ I, but the invention includes, but is not limited to, dyes and enzymes.

The agent for use in the method may be an antibody raised against a protein, or against a unique subregion of that protein, the absence of which is associated with prostate cancer.

The present invention provides a method by which a tumor suppression gene or nucleic acid is detected in a sample, the absence of which is associated with neoplasm. The method comprises the step of obtaining a suitable sample from the subject. A method for detecting the presence of nucleic acid in cells comprises hybridizing the probe nucleic acid to the nucleic acid of the cells. Such techniques are carried out by methods well known to those of ordinary skill in the art. See, for example, Sambrook et al. (Sambrook et al. Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY., Pp. 1.98-1.104 (1989)).

Suitable samples include, but are not limited to, prostate tumor tissue, colon tumor tissue, lymph node tissue, and bone marrow cells. The method requires contacting a sample and an agent specifically unique to the normal tumor suppressor gene or the wild-type gene under conditions that favor complexing with the agent and then detecting any complex formed. The absence of the complex indicates the absence of a wild-type gene that is associated with a neoplastic condition such as prostate adenocarcinoma. Thus: This method is useful for diagnosing prostate adenocarcinoma. For the purposes of this invention, suitable labeling agents are radioisotopes such as ³² P, ³⁵ S and ¹³¹ I, but the invention includes, but is not limited to, other labeling agents such as dyes and enzymes. The agent may be a nucleic acid corresponding to a tumor suppressor protein or a unique subregion thereof.

The present invention provides a kit for the detection, diagnosis and prognosis of prostate cancer. The kit contains reagents useful for carrying out the methods described above and instructions for their use in such methods. The kit can be used for direct genetic detection of pathological changes in the prostate tumor suppressor gene, and can include, for example, oligonucleotides, primers for PCR analysis, reagents for SSCP, or sequencing. These kits, reagents and methods are also useful for prognosis. For example: A deletion may be indicative of a less favorable prognosis for recovery.

Within the scope of the invention are compositions comprising at least one of the nucleic acids, peptides and antibodies mentioned above. These compositions may also contain carriers or diluents, such as phosphate-buffered saline, emulsions, and various wetting agents.

The following embodiments are intended to explain, but not limit, the scope of the invention.

DETAILED DESCRIPTION OF THE INVENTION

ALLELA LOSS IDENTIFICATION

1. Tissue specimens have not previously been treated with primary prostate tissues with a range

Prostate cancer cells were obtained from patients undergoing radical prostatectomy between clinically localized prostate cancer between August 1988 and November 1994. None of the patients enrolled in the study, chemotherapy or hormone therapy, and seminal vesicles and PTSG were collected at -80C as described in Bova, G. S. et al. (Bova, G. S. et al. Homozygous deletion and frequent allelic loss of chromosome 8p22 loci in human prostate cancer. Cancer Res. 53: 3969-3973 (1993)). In short: Only clinically palpable tumors qualified for the study, and only palpable palpable tumors were collected. The mean Gleason value (Meilinger, G.T. et al., (1967)) for the 42 tumors included in the study was 7.4 + 1.1 (SD).

Focal or developed capsular penetration was observed in all 42 primary tumors studied, and thus all tumors studied fall into the T ₃ category used in the recent classification of prostate cancer metastasis (Schroder, FH et al., (1992)). Histological confirmation of invasion of seminal vesicles was observed in 16 of 42 (38% of primary tumors enrolled in the study. Microscopic lymph node raetastasis was observed in 11 of 42 (26%) of enrolled trials.

Collected primary tumors were incorporated, and 6-μιη sections were stained with hematoxylin and eosin. Forty-two primary prostate adenocarcinomas were selected for DNA analysis and could be dissected to provide tissues containing more than 70% of the tumor nuclei. Metastatic prostate adenocarcinoma tissue was available in ten cases from patients who had palpable enlarged abdominal lymph nodes at the time of the intended radical prostatectomy. One PTSG section taken at the time of surgery showed metastatic adenocarcinoma and radical prostatectomy was not performed. Node tissue that was not needed for PTSG histological analysis was separated, set at -80 ° C and used for this study.

Paired non-cancerous tissue (seminal vesicles, prostate, or blood lymphocytes) was obtained from each patient. Seminal vesicle or prostate tissue, which served as source material for non-cancerous DNA, was scanned every 300 μπι of PTSG sections, and all tissues containing dysplastic or cancerous epithelium were discarded.

Preoperative serum prostate-specific antigen levels were measured by a monoclonal immunoradiometric assay (Hybritech, San Diego, CA).

2. DNA preparation

Tissue containing 70% of the tissue cancer nuclei (containing lymphocytes, etc.) as well as the prostate specific antigen, with more than the prostate, was separated from the surrounding benign prostatic epithelium, stroma, perfectly as possible using the cryostat cutting technique described in Bos, JL et al. , (1987). All prostate carcinomas studied were of the common ascinar type and were <2 cm in diameter. DNA isolation and quantification were performed as described in Carter, B. S. et al. (1990) and Burton, K. et al. (1968).

3. Southern analysis

The samples were digested with restriction endonucleases (BRL and New England Biolabs) with buffers recommended by the supplier, using 10 units of enzyme / µg DNA for Msp1 and 7.5 units / µg for Taq1 digests. The samples were electrophoresed in 0.8% agarose gels and transferred to Nytran nylon membranes (Schleicher & Schuell) in 0.4 M sodium hydroxide / 0.6 M sodium chloride after depurination in 0.25 N HCl for 10 minutes. After covalent binding of the DNA to the membrane using UV irradiation (Stratagene), the membranes were prehybridized in 10 ml of 1 M NaCl / 1% sodium dodecyl sulfate /% Dextran sulfate at 65 ° C for 1 hour. KSR2, NF 5.1 and MCT 128.2 DNA probes were obtained from the American Type Culture Collection. Probes CI8-1, MSR-32 (MSR-macrophage scavenger receptor), CI8-319 and CI8-277 are cosmid probes as described in Emi, M. et al., (1993). Probes were labeled using random hexamers starting and incorporation of [α- ³² P] dCTP (Amersham) with Klenow DNA polymerase I fragment (Amersham). CI8-1, MSR-32, CI8-319 and CI8-277 probes were boiled with sheared human placental DNA (Sigma), briefly cooled on ice, and hybridized at 65 ° C overnight. After hybridization, membranes were washed in 0.1X phosphate-EDTA standard saline with 0.1% sodium dodecyl sulfate for 15 minutes and subsequently exposed to Kodak XAR-5 film at -80 ° C in amplification foil cassettes.

Allele loss was defined as the absence of one allele in prostate tumor DNA compared to non-cancer mating control DNA. In some cases, when there was a residual signal from contaminating normal tissue, it was used for densitometry analysis. The sample was ranked as having an allele loss when a 60% reduction in the disappearing allele was present compared to its normalized conserved counterpart.

Allelic multiplication using the MCT 128.2 probe was defined as an increase in intensity greater than 100% for one of the two alleles present in the tumor sample, or as intensity differences greater than 100% between tumor and normal alleles in homozygous cases when previous probing of the same blots showed equal DNA loading in tumor and normal DNA pathways.

4. Analysis of microsatellites

The sequences for lipoprotein lipase (LPL) (GZ 14) and a set of Mfd primers were previously published in Tomfohrde, J. et al., (1992). One of each primer pair (LPL GZ 14 and Mfd 199R) was terminally labeled with [t- ³² P] ATP (ICN Biochemicals) using polynucleotide kinase (Boehringer-Mannheim) and 5X kinase buffer [0.25 M Tris, (pH 9, 50 mM MgCl50 mM dithiotreitol and 0.25 mg / ml bovine serum albumin]. Six μΐ primer (10 μΜ), 2.8 μΐ 5X kinase buffer, 0.7 μΐ kinase (9 units / μΙ), 1.5 μΐ sterile deionized water and 3.0 μΐ [t- ³² P] -ATP were pooled. and incubated at 37 ° C for 1 hour. The products were purified using G-25 centrifuge columns (Boehringer-Mannheim). One μΐ of labeled primer was added to 1 μΐ of unlabeled primer (10 μΜ) of 0.5 ml of a mixture of desoxynucleotide triphosphates (equal volumes of dATP, dCTP, dGTP and dTTP, each 10 mM),

5.5 μΐ of sterile deionized water and 10X Taq DNA polymerase buffer (Perkin-Elmer), 10 μΐ of genomic DNA (2.5 ng / μΐ) was added and the mixture heated to 94 ° C. After addition of Taq DNA polymerase solution (5 units), 30 cycles of denaturation at 94 ° C for 30 seconds, annealing at 62C (LPL) or 58 ° C (Mfd 199) for 30 seconds and extension at 72 were performed. ° C for 30 seconds. This was followed by 72 ° C for 7 minutes. The products were then mixed with an equal volume of stop buffer containing 95% formamide, 0.05% xylene cyanol, 0.05% bromophenol blue and 20 mM EDTA.

The samples were thermally denatured at 94 ’and a 3 μΐ aliquot of each sample was applied to a 6% acrylamide gel containing 8.0 dried and exposed to Kodak XAR film. the use of microsatellite analysis established criteria similar to those used

M urea. In this study, allele loss was performed according to the Southern analysis described above.

5. Immunohistochemistry of MSR proteins

Sections from primary prostate cancer and adjacent non-cancerous prostate (including areas of benign prostatic hypertrophy and normal prostate) were examined in five patients. Autopsy-derived liver tissue from a single patient served as a positive control for MSR staining. A well preserved central and peripheral zone of the prostate was obtained from that particular patient and stained for MSR protein. In this patient, there was no evidence of malignancy at autopsy and the prostate tissue was normal at a cursory examination and histologically. Non-fixed, air-dried 6-μιη frozen slides on slides were warmed to room temperature and fixed in 2% formaldehyde / 10 mM Tris, pH 7.4 / 150 mM NaCl / 2 mM CaCl ₂ for 10 minutes and then incubated for 20 minutes. minutes at room temperature in 0.3% H ₂ O ₂ / absolute methanol. The slides were then washed twice with 10 mM Tris, pH

7.4 / 150 mM NaCl / 2 mM CaCl ₂ and then incubated at 37 ° C for 10 minutes in serum blocking solution (Zymed). Rabbit IgG against the human scavenger receptor synthetic peptide (kindly provided by Dr. Tatsuhiko Kodama, University of Tokyo, described in Kodama et al. (1988)) was then added (1:50) to each slide and incubated at 37 ° C for 30 minutes. minutes. Primary antibodies were detected with a biotinylated secondary antibody-streptavidin-peroxidase conjugate (Zymed).

6. Results

Fifty-two (52) prostate cancer samples were examined for allele loss using 8 polymorphic probes per short arm of chromosome 8. Overall: 32 of 51 (63%) informative tumor samples showed a loss of at least one locus on chromosome 8p. The most frequently deleted region is observed on chromosome 8p22-21.2. One allele loss is identified in 14 of 23 (61%) tumors in D8S163 (12 of 19 primary tumors and 2 of 4 lymph node metastases) (Figure 1), in 15 of 32 (47%) tumors in LPL (15 of 30 primary tumors and 0 out of 2 metastases), and 20 out of 29 (69%) MSR tumors (17 out of 26 primary tumors and 3 out of 3 to 8p22. One allele loss identifies tumors in D8S200 (12 out of 22 primary tumors and 4 of 5 metastases) to 8p21.3-21.2 (Figure 2; Table 1).

In addition to the loss of one allele at the MSR locus in most tumors, one sample of metastatic prostate cancer (N2) showed a homozygous deletion of the MSR sequence. Hybridization of the same blot with DCC probe 15-65 determined the presence of intact DNA of equivalent or larger size in the N2 pathway of the tumor (Figure 3). Repeated digestion of N2 DNA with MspI, Taql and EcoRI and probing on MSR confirmed this finding, but at least one allele is present. Thus, the interface of the homozygous deletion is delimited by D8S163 metastases), all in 16 of 27 (59%) and LPL.

and in just studied at

Unlike 8p22-21.2, the telomeric and centromeric loci to this region are mostly retained, with the loss of one or more loci in only 9 out of 48 (19%) informative cases.

The distal loci studied at 8p23 are mostly retained, with only 4 out of 38 (11%) informative cases lost in D8S140 3 out of 22 (14%) cases in D8S201 (Table 1). The 8p12.2 and 8q24 loci are also deleted at low frequency, with the loss identified in 3 out of 26 (12%) informative cases in D8S194 and in 2 out of 17 (12%) in D8S39.

Evidence for chromosome 8q multiplication was detected in 5 of 32 (16%) tumors probed in D8S39, including cases 4, 20, 21, N1 and N2 (Figure 3). The signals for one or two D8S39 (8q24) alleles were multiplied 2-3 times after correction for differences in DNA loading. All tumors with 8q multiplication had an 8p loss at least one locus.

Data from all primary and metastatic prostate cancers with evidence of chromosome 8p loss are summarized in Figure 4. Fifteen of the 42 (36%) primary tumors studied exhibited heterozygous retention or were not informative at the 8 loci studied on chromosome 8p, and these cases are not illustrated in Figure 4. All 8p-loss tumors, which are informative for MSR, lost at least one allele at this locus. Tumors 1, 18, 25 and N5 retained D8S163 but lost proximal loci, including MSR. Tumors 24 and 25 retained LPL but lost more distal loci, including MSR. These results enclose the smallest overlap region within the D8S163 and LPL intervals adjacent to the MSR locus. Based on the genetic map presented by Emi et al. (1993), this interval covers 14 cM in men.

Observing the homozygous deletion at the MSR locus prompted us to make a preliminary determination of the macrophage scavenger receptor gene as a possible tumor suppressor gene. Prostate tissue was analyzed for expression of the MSR protein using a highly specific polyclonal antibody as described by Kodama et al. (1988). The macrophage scavenger receptor protein was not detected in prostate cancer cells or non-cancerous prostate epithelium. The scattered cells contained in the stroma of each such prostate section were stained positively in accordance with the staining of the macrophages themselves.

To determine whether allele loss on chromosome 8p significantly correlated with clinical parameters, preoperative serum PSA levels were examined in each study enrolled patient, Gleason parameter and endpoint pathology were evaluated. The mean Gleason value did not differ in these two groups, with a mean of 7.3 in patients with 8p loss and a mean of 7.6 in those where no 8p loss was demonstrated. Preoperative serum PSA levels were available in 34 of the 42 patients studied in primary prostate cancer tissue. The mean PSA level for the whole group was 11.2 ng / ml (range 1.6 - 23.6). The mean PSA level for patients with 8p loss was 12.6 ng / ml and for patients without loss on chromosome 8p it was 9.3

P = 0.105). Seminal vesicle invasion (41%) of patients with 8p loss and in 5 seminal vesicle invasion (X ² , P = 0.055). Microscopic lymph node metastases were found in 9 of 27 (33%) patients with 8p loss, and in 3 of 15 (20%) patients without 8p loss ^(X2, P = 0.35). In summary: There are trends for higher pre-operative levels of serum PSA, more frequent lymph node involvement, and more severe seminal vesicle involvement in patients with 8p loss demonstrated in their prostate cancer, but these trends are not statistically significant.

ng / ml (variation analysis) was observed in 11 of 27 out of 15 (33%) patients without

B. 8p PROBE INSULATION AND MAPPING

1. Origin of somatic cell probes, primers and hybrids

The plasmid probe pABL-2, which detects D8S21, was obtained from R. White and its preparation is published in Tsui, LC et al. (1989). Its insert was partially sequenced starting from E. coli amber ^{T yr} suppressor tRNA using oligonucleotide 5'-GAATCCTTCCCCCAC-3 ', and two PCR primers were designed to generate STS (Table 2). Lambda phage CRI-R191, which detects D8S26, was obtained from the ATCC. 4.2 kb restriction fragment

37 of this phage were subcloned and partially sequenced, from which STS was designed (Table 2). Cosmid CI8-487, which detects D8S233, was obtained from the Japanese Bank of Oncology Research Resources. The 2.2 kb EcoRI restriction fragment of this cosmid was subcloned and partially sequenced to generate STS (Table 2). New STSs were generated by partial sequencing of random subclones of purified YAC DNA (see below). The terminal YAC fragments (Table 3) were obtained by the inverted PCR method of Albertsen and Thliveris (Joslyn et al., 1991). PCR products were ligated into the TA cloning vector (Invitrogen) and sequenced to make STS (Table 3). Other primer sequences were obtained from the sources shown in Tables 2 and 3. DNA from human chromosome 8 x CHO somatic cell hybrid mapping panel (Wagner et al., A hybrid mapping panel for regional localization of human chromosome 8 Genomics 10: 114- 125 (1991)) was kindly provided by M. Wagner.

2. Radiation hybrids

The hybrid human x hamsters line, GM10156b, containing human chromosome 8 as its sole human component, was obtained from a NIGMS Mutant Cell Repository (Camden, NJ). The hybrid was exposed to 5,000 rows of τ radiation and fused to the APRT- and HPRT-deficient Chinese hamster ovary cell line CH-ATS-49tg by the method of Cox et al. (Cox et al., Science 250: 245-250 (1990). A total of 97 hybrid clones were obtained after HAT selection. The presence or absence of six marker loci (D8S26, MSR, D8S233, D8S261, D8S21 and LPL) in radiation hybrid DNA was The distances and order of these markers were determined using the Radiation Hybrid Mapping Statistical Package (Cox et al.), and the somatic cell genetic method for constructing high-resolution maps of mammalian chromosome. Science 250: 245-250.) The TWOPOINT program was used to determine recombination fractions and retention frequencies, and test maps were tested to confirm the order with the FOURPOINT program.

3. Evaluation of YAC library

A copy of the YAC library (Albertsen et al., Proc. Natl. Acad. Sci. U.S.A., 87: 4256-4660 (1990)) was obtained from CEPH (Paris, France). The library was evaluated by PCR with ten loci from the list in Table 2 by a hierarchical screening method (Green,

E. D. and Olson, 1213-1217 (1990)) in 8 rows and represented 384

Μ. V. Why. Nati. Acad. Sci. U.S.A., 87: DNA pools were made from 4 plates of 12 columns; each of the 58 super-clones. Clones identified by plate / row / column address were plated on AHC-agar plates and verified by direct PCR colony or yeast DNA PCR (Ausubel et al., Current Protocols in Molecular Biology. Greene Publishing Associates / J. Wiley & Sons, Inc.). , New York, NY. (1992)).

4. Enclosure of yeast DNA into beads

The YAC clone conservation was plated on AHC-agar plates. Single pink colonies were picked and grown in 5 ml YPD medium at 30 ° C overnight, and then expanded to 100 ml for an additional hour. The yeast cells were sealed into agarose beads by Overhauser and Radic method (Focus 9 (3): 8-9, Bethesda Research Laboratories, Inc., Gaithersburg, Md., 1987) as follows: Cells were harvested by centrifugation and washed twice in 20 ml SE (75 mM NaCl, mM Na ₂ EDTA, pH 8.0) then resuspended in 4 mL SE. The cell suspensions were transferred to 125 ml Erlenmeyer flasks and heated to 45 ° C. Genomic-grade low melting point agarose (1% in SE) and mineral oil were separately equilibrated to 45 ° C, and beakers containing 100 ml ice-cold SE and stirrer were placed in an ice bath on medium speed magnetic stirrers. 5 ml of agarose was added to the cells in each flask and mixed. Twenty ml of mineral oil was then added and the flask was swirled vigorously for 30 seconds to emulsify the contents, which were then immediately poured into chilled SE beakers. The beads formed within 5 minutes. The mixture from each flask was transferred to several 50 ml centrifuge tubes and centrifuged at low speed to separate the aqueous and oil layers. Excess oil was removed and the contents recharged. Residual oil, SE and floating grains were discarded, and the other grains (5-10 ml) were taken three more times with SE. The inside of the cuvette was wiped to remove traces of oil, and the beads were put together into one cuvette. The seeded beads were resuspended in 1 volume of SE and cells were digested with 0.5 ml mercaptoethanol and 10 ml freshly dissolved yeast lytic enzyme (70,000 U / g, ICN) to 10 ml final volume at 37 ° C for 2 hours. The beads were then centrifuged as before, resuspended in 20 ml 1% (w / v) sarcosyl, 25 mM Na ₂ EDTA, pH 8.0, 50 µg / ml proteinase K and incubated overnight at 50 ° C. The supernatants were discarded, the beads were washed in 20 ml TE with 0.1 mM phenylmethylsulfonyl fluoride (PMSF), followed by washing twice more in TE.

5. Preparation of YAC DNA by PFGE

A 0.6 cm thick, 1% agarose gel in 0.5X TBE, with 2 or 3 preparative wells, was poured into a 20 cm x 14 cm long gel forming unit. The wells were loaded with low melting agarose beads containing YAC DNA and the wells were sealed with low melting agarose. Yeast chromosomes were resolved on a BioGE Rad CHEF-DR III PFGE (running at 60-120 second switching time, 120 ° angle inclination, 24 hours at 6 V / cm, with 0.5X TBE, at 14'C. The gel was stained in 1 µg / ml ethidium bromide in 0.5 X TBE for 30 minutes and the chromosomes visualized by UV light. Notches 5 - 7 mm wide were cut at the head of each YAC that was to be isolated and parallel to it, and the gel was put back on its platform. Excess buffer and gel fragments were aspirated, and notches were filled with 1 percent low melting agarose (InCert, BioRad) in 0.5X TBA, which allowed to solidify. The gel was returned to CHEF-DR III and equilibrated to 14 ° C. PFGE ran with a 180 second constant switching time for 4 hours. YAC bands were again visualized by UV illumination and cut from notches of low melting material.

The gel sections were equilibrated with 1X β-agarase buffer (New England Biolabs, NEB) with two changes, the buffer was removed, and the sections thawed at 65-70 ° C for 30 minutes. The dissected sections were placed at ° C and incubated for 1-2 hours with β-agarase I (NEB) (5 units per gram of agarose), then cooled on ice and spun to remove undigested agarose. Supernatants were placed on Centricon 100 filter units (Amicon) with excess TE buffer and centrifuged at 500 x g for 30 minutes to concentrate and purify YAC DNA. The resulting '80 μΐ preparations were further concentrated to 50 μΐ by Speed-Vac to obtain supernatants as final products.

YAC DNA (’50 μΐ) was digested with Bgl II and ligated into Bam HI-digested pBluescript (Stratagene) by standard procedures (Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY).

1.98-1.104 (1989)). The ligation mixture was re-digested with BamH I to reduce the non-recombinant background and transformed into E. coli DH10B (GIBCO-BRL) with X-gal and IPTG for blue-white selection according to the supplier's recommendations. White colony-derived plasmids were adopted for use as single-copy probes in Hind III-digested human genomic DNA (Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY., Pp. 1.98-). 1.104 (1989)), and then mapped on somatic cell hybrids and YAC clones as follows.

6. YAC DNA labeling with biotin

Isolated YAC DNA (10-20 μΐ) was labeled with biotin by extension of random primers in the presence of biotinylated dATP (BioPrime kit, BRL) in 50 μΐ volumes according to the kit instructions. Successful labeling was verified by electrophoresis of the 5 μΐ reaction product on agarose gels and either by visualization of a faint smeared band with ethidium bromide and UV irradiation, or by transfer of DNA to the nylon membrane by standard methods. The membrane was blocked as for Western analysis, and streptavidin-alkaline phosphatase conjugate without primary or secondary antibodies was added directly. Blurred bands of biotin-labeled DNA were visualized by the BCIP / NBT substrate reaction.

7. Oligonucleotides for linkers and PCR

Two oligonucleotides were synthesized,

5'-CGATCTAGACCAGCACAATGG-3 '(primer 1) and 5'-CCATTGTGCTGGTCTAGATCGCACA-3' (primer 2). Primer 2 was 5'-phosphorylated with ATP and T4 kinase (37 ° C, 30 min), heated to inactivate the enzyme, and hybridized with an equimolar amount of primer 1 to form a linker

5'-CGATCTAGACCAGCACAATGG-3 '

3'-ACACGCTAGATCTGGTCGTGTTACC-P-5 '

Xba I BstX I which is blunt and phosphorylated at one end and non-sticky to itself at the other. DNA fragments that are terminated by such splices can be amplified by PCR with primer I.

8. Create libraries of amplifiable short cDNA fragments

RNA was isolated from tissues and cells using TriReagent (Molecular Research Center, Inc) according to the manufacturer's instructions. Poly-A ⁺ RNA was selected from total RNA using biotinylated oligo-dT primers and streptavidin-conjugated paramagnetic bodies (PolyATract kit, Promega). Double-stranded cDNA was generated from poly-A ⁺ RNA and one total RNA sample according to the manufacturer's random primer instructions and M-MLV RT (RiboClone cDNA synthesis kit, Promega). The final step with T4 DNA polymerase yielded blunt-ended cDNA. cDNA generated from total DNA was set aside for later use as ribosomal DNA probes (rDNA). Excess linkers (see above) were ligated to cDNA derived from poly-A ⁺ material by T4 DNA ligase. cDNA (1 - 5 μΐ) was amplified in 100 μΐ volumes using ~ 500 ng of primer 1 and other PC components at usual concentrations. The conditions were (95 ° C, 2.5 ') - (94 ° C, 40; 60 ° C, 40; 72 ° C, 2.5') x 20 - (72 ° C, 10 '). PCR products were purified with Wizard PCR Prep Spin columns (Promega) and eluted at 50 μΐ 0.5 x TE. DNA was quantified by DNA Dipstick (Invitrogen) and typical yields were 500 ng of purified product per 100 μΐ reaction. Amplified cDNAs examined on ethidium bromide staining agarose gel electrophoresis formed a broad band with a maximum intensity at about 500 bp.

9. Blocking of repetitive sequences in cDNA

Purified amplified cDNA (1-2 µg) was mixed with an equal amount (w / w) of Cot 1 DNA (GIBCO-BRL) and the reaction volume was adjusted to 80 µg / ml in 120 mM NaPO ₄ buffer pH 7 (eg 50 µΐ). The cDNA was reduced to 21 μΐ by SpeedVac, to which 1 μ 1 Cot 1 DNA and 3 μΐ 1 M NaPO ₄ pH 7 were added, and the presence of TE was ignored.

and heated to 100C after 10 minutes to denature, then to incubated for 20 hours at 60 ° C (c _o t = 20). 10. Hybridization cDNA to YAC Morgan et al al. (1992) was adapted to smaller ones

modifications. Biotin-labeled YAC cDNAs (100 ng or 10 μΐ labeling reactions per hybridization) were thermally denatured and loaded onto Centricon 100 blocked cDNA filter units (1 µg, excluding Cot 1 DNA) and 2 ml of 1 mM NaPO ₄ pH 7, and spun at 1000 xg for 25 minutes. Washing with phosphate buffer was repeated once, and retained aliquots (60-80 μΐ) were transferred to microcentrifuge tubes. Volumes were reduced to '5 μΐ in SpeedVac, at which point hybridization mixtures were adjusted to 120 mM NaPO ₄ pH 7, 1 mM EDTA pH 8, and DNA concentration (excluding Cot 1 DNA) ~ 160 µg / ml (e.g. 1.1 μg at 7 μΐ). Reaction mixtures were overlaid with mineral oil and then incubated for 60 hours at 60 ° C (C _Q t = 120).

Capture, amplification and cloning of selected cDNAs

Streptavidin-conjugated paramagnetic bodies (Promega) were prewashed twice with TE + 1 M NaCl and then incubated with completed hybridization reactions in 200 μΐ TE + 1 M NaCl at room temperature for 15 minutes. The beads were collected magnetically and the supernatants discarded. The beads were washed 5 times for 15 minutes by incubation in 200 μΐ 0.1 X SSC + 0.1% SDS, two at room temperature and then three at 60 ° C, with magnetic collection between all washes. Bound cDNA was eluted from the beads with 100 μ Na 50 mM NaOH for 15 minutes, neutralized with 100 μΐ 1 M Tris-HCl pH 7.5, and transferred to clean tubes. The supernatants were desalted and concentrated using NaI and kieselguhr matrix (Geneclean kit, Bio 101) according to the manufacturer's instructions to 20 μΐ volumes TE. These cDNAs were re-amplified exactly as with the original libraries (see above) except that 5 μΐ templates were used and PCR was performed for 30 cycles. YAC DNA selection was also performed as above for the second time. The cDNAs selected in the second round were captured as above and amplified once more. The final PCR products were directly cloned into a T-vector (Novage) and their transformants were plated on Tet + Amp LB-X-gal and IPTG plates for blue-white selection according to the kit instructions.

12. Evaluation of recombinant clones

White colonies (75 / selection) were picked with wooden toothpicks in duplicate on two grid-lined Amp-agar plates, one of which was covered with a circular nylon membrane and the other not. A unique, ordered pattern of short lines and dotted lines was created so that duplicate colonies on these dishes could be easily identified. After growing overnight at 37 ° C, the master dish was stored, the filter was picked up from the second dish, and processed as in a hybridization evaluation of the colonies on the filter (Sambrook et al. Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring). Harbor, NY., P.

1.98-1.104 (1989)), with 10% SDS, denaturation, neutralization and 2 X SSC. The filters were baked, prewashed (p. 1.101), prehybridized in 0.05 X BLOTTO, then hybridized, either sequentially or simultaneously, with nick-translated total human DNA (Alu probe) and primed rDNA (ribosomal probe) DNA). After washing and autoradiography, colonies were removed from the master plate to further characterize the colonies that were negative in hybridization. Mini-scale DNA preparations were prepared and analyzed by BstX I or (HindIII + EcoRI) digestion and agarose gel electrophoresis. The size of the inserts was about 250-500 bp; inserts were obtained by cutting from low melting agarose and radioactively labeled with random start. The probes were hybridized to filters containing HindIII digested DNA from YAC clones and to human DNA triplets, chromosome 8 of the human-mouse cell hybrid, and mouse genomic DNA to identify single copy probes located to human chromosome 8.

13. YAC content mapping

About 30 random subclones (Table 6) YACs 932_e_9, 767_h_8, 802_f_11, 832_a_10, 821_f_7 and 885_c_8 were isolated and mapped to human chromosome 8 by Southern analysis with the DNA of a human chromosome hybrid cell line of 8 x mouse or its regional panel (Wag). Genomics 10: 114-125 (1991)). These probes were mapped with a contig of YAC hybridization to the transferred YAC DNA digested with Hind III (Note E1 = E1). The three terminal YAC clones (YE766, YE843 and YE932, respectively) were isolated by inverse PCR, as described above, from YAC 766_a_12, 843_g_3 and YAC7, respectively. 932_e_9 and mapped to chromosome 8 and the associated YAC portions as above.

The CEPH megabase YAC library (22,000 clones) was evaluated by PCR with primers for six single tandem repeat (STRP) polymorphisms and four RFLP containing loci on several independent binding maps (Fig. 5, Table 2). About 30 YAC clones were identified and confirmed by initial screening. Addresses of other clones were obtained by screening for AluPCR and fingerprinting overlap rating tables (Cohen et al., Science 250: 245-250 (1993)). These clones were included in the YAC map after testing for STS content. A set of 31 markers was used to build the map, including ten search STSs, one additional published STRP (D8S206), two expressed tags (D8S294E and D8S297E) (Adams et al., Nature 355: 632-634 (1992)) , fifteen random YAC subclones and three terminal YAC clones (Table 3). PCR methods and Southern analysis were used sequentially to minimize false positive and negative values. The YAC map is anchored to cytogenetic maps by chromosomal localization of the MSR and LPL (lipoprotein lipase) genes (8p22) (Mattei et al., Cytogenet. Cell Genet. 63: 45-46 (1993); Emi, M. et al., J. Biol Chem 268: 2120-2125 multiple probes at intervals A or

Wagner et al.

(Table 3), was created (1993)) or by location

B panel of the somatic cell hybrid hybrid described (Wagner et al., Genomics 10: 114-125 (1991)) wherein interval A is telomeric to interval B.

A single large contig of thirty-six YACs (Table 4, Figure 5). Rare Content Mapping STS dictated a unique order for all ten original search markers such as tel - D8S26 - D8S511 - D8S549 - MSR - D8S254 - D8S233 - D8S261 - D8S21 - LPL - D8S258 - prices. Cosmid CI8-245 (D8S335), which contains a centromeric interface for one or more allele loss regions (Ohata et al., Genes Chromosome. Cancer 7: 85-88 (1993), Emi, M. et al., Genes Chromosome. Cancer

7: 152-157 (1993); Fujiwara et al., Cancer Res. 53: 1172-1174 (1994)) was not available from the Cancer Bank of Japan and could not be included in our map. It is tightly bound and seemingly centromeric to LPL (Emi, M. et al., J. Biol. Chem. 268: 2120-2125 (1993a); Emi, M. et al., Genomics 15: 530-534 (1993)) . CTSB (cathepsin), another RFLP marker that was mapped in 8p22-p23.1 (Fong et al.

Hum.

Genet. 89: 10-12 (1992)), was placed in hybrid interval A (MacGrogan et al., Genes Chromosom. Cancer 10: 151-159 (1994)) and excluded from the physically mapped region D8S26 - D8S258 for its absence in this YAC file. clones (data not shown). Due to a lack of sufficient YAC all, we were unable to uniquely order arrange some accessory markers such as D8S206, D8S294E, and D8S297E. as an incidental finding the probe El, a random subclone of YAC ⁹³⁹ _θ ^"9 Hind III RFLP in human DNA , with two alleles, 12 kb and (8 kb + 4 kb).

YACs are subject to two types of remodeling artifacts, chimerism and internal deletions, which potentially can affect various aspects of physical mapping. Chimerism potentially did not affect our mapping of STS content and the derived sequence of loci, since all markers were independently mapped to chromosome 8. Insurance against internal deletion effects was provided by many discarded auxiliary probes and large attached depth (redundancy). For example: A large internal deletion, including probes E15, YE766, E1, E3, MSR, and E20, was postulated in YAC 747_h_8 to maintain uniformity of at least six other YACs. On the other hand, this apparent deletion provided two additional endpoints to resolve the order of the two marker pairs.

Of the 97 isolated radiation hybrids, 17 retained one or more of the six markers tested, with retention frequencies of individual loci ranging from 0.12 to 0.17. At least one breakpoint was detected in 10 of the 17 hybrids. The determination of the distances between the locus pairs was generated by TWOPOINT (Table 4). The order of the markers prompted by YAC mapping was tested by four-point analysis of radiation hybrid data (Fig. 1). The calculated probability against inversion was greater than 1: 1000 for all adjacent markers except D8S261 and D8S21, which were separated by only one breakpoint and a calculated aunt value of 0.05 or 5 cRay _500Q . Thus, the order of the markers was consistent for genetic maps and YAC and radiation hybrid maps. The distance between D8S26 and LPL was '9 cM on the genetic map and 90 cR ₅₀₀₀ on the radiation hybrid map, suggesting a ratio of' 10 cR ₅₀₀₀ per cM in this region.

14. Restrictive mapping of larger distances

Forty selected cDNA fragments (Table 7) were isolated and mapped to human chromosome 8 and the YAC panel as above. Selection was made with YAC 932_e_9, 802_f_11, 821_f_7, 877_f_2 and 946_c_9. A larger distance restriction map was constructed for a portion of the 8p22 region (Figure 8). The map covers at least 25 probes of Tables 6 and 7. YAC DNA was digested with various sparsely cleaving restriction enzymes, Asc I, Mlu I, Not I, Nru I, or Sfi I and divided into PFGE as described above. Southern analysis was performed to determine restriction fragments containing each of the probes. The map was compiled based on standard mapping methodologies, including partial digest analysis and two enzyme digest analysis. One notable piece of evidence is that cosmid CI8-2644, obtained from Dr. Y. Nakamury, was located telomerically to the MSR gene and not centromerically, as suggested by Fujiwara et al. (Fujiwara et al., Genes Chromos. Cancer 10: 7-14 (1994)).

15. Mapping of homozygous deletions in N2 tumor

DNA from a single prostate metastatic tumor with a homozygous deletion of MSR (Bova et al., Cancer Res., 53: 3869-3873 (1993)) was examined by Southern analysis with numerous newly isolated genomic and selected cDNA probes to map the extent of this deletion. Probes found to have a complete deletion in this tumor (bold, Figure 8) begin with MSR and continue telomerically through probes 877-15 and CCI8-2644. Ele markers resp. 877-13 are the closest preserved loci to centromeric resp. the telomeric end. Based on the positions of the lost and conserved loci within the mapped restriction fragments (Figure 8), the minimum and maximum size of the homozygous deletion in this tumor were determined to be 740 kb and 6 kb, respectively. 9 20 kb. The target tumor suppressor gene was predicted to be located within this deletion and was inactivated by this deletion. N33 is located within this region (Figure 8).

The mapping of CCI8-2644 to a position close to the telomeric interface of the deletion was significant because it suggests that a common region of allelic loss detected in colorectal cancer, liver cancer and lung cancer by Fujiwarou et al. (Fujiwara et al., Genes Chromosome. Cancer 10: 7-14 (1994)) extensively overlap with this region of homozygous deletion. Thus, any gene within this homozygous deletion may be important for these other cancers. In addition, the size of the allele loss region of this publication must be larger than claimed in this paper (600 kb) and larger than the homozygous deletion in this tumor, i.e., this homozygous deletion defines the smallest known critical region containing the putative tumor suppressor gene.

16. Sequence analysis of selected cDNA fragments identical in GENBANK by sequencing genomic P34669, GENBANK M88869,

Sequencing of selected cDNA probes from Table 7 revealed the following: 1) P3 and P26 are identical to the 5 'end of the MSR cDNA sequence, whereas P34 is derived from the 3' non-translated region of the MSR. Isolation of fragments of known genes from this region showed that the cDNA selection method was successful. 2) J28 overlaps P27, L3 and N28, and contains a partial ORF encoding a new predicted amino acid sequence without any close relationship in GENBANK or PIR. The other portions of this DNA sequence were close to the random cDNA sequencing sequences deposited 3). Sequence differences were non-conservative and we suspect that J12 represents either a closely related gene or a pseudogen. We then cloned and sequenced the J12 locus at the genomic DNA level and found that it lacked introns and that it contained a single base insertion that would abrogate the conserved ORF. We conclude that J12 is a pseudogen for the alpha subunit of human phosphatase type 2C. 4) L21, N21, N33, N36 and P14 overlap and define a partial ORF with highly significant homology to the predicted gene in C. elegans, identified by random cosmids or cDNA clones (SWISS-PROT T01933, L17337, PIR S44911). The function of this C. elegans gene is unknown.

17. Cloning and sequencing of longer N33 cDNAs

Based on preliminary expression data (see below), the N33 cDNA clone was used as a screen probe in a lambda-phage placental cDNA library (Clontech). Clone N33C was isolated and its 1.3 kb EcoRI-EcoRI insert was subcloned into pBluescript to give pBS-N33C (7). Sequencing revealed a 1342-bp insert with EcoRI sites at the edges (Fig. 9, 10) that encoded a long ORF (nt 158-1202) (Fig. 11). Based on this sequence, primers N33GEX-f and N33GEX-r were synthesized (Fig. 10) and used to amplify a region of N33 mRNA by RT-PCR of placental mRNA. Two close-lying lanes with '950 bp were detected, with a ratio of approximately 1: 2 (upper lane: lower lane). To further characterize these bands, RT-PCR products were cloned into pGEX-2T (Pharmacia) and two clones, A4 and A5, were isolated. Clone A4 was collinear with pBS-N33C (7), while A5 lacked nt 1186-1250 (65 bp) compared to other clones. This suggests that N33C (7) and A4 represent longer mRNA (Form 1), while A5 represents shorter mRNA (Form 2). The ORFs encoded by these two forms differ in the last 20 bp and utilize different termination codons (Figs. 10-14). The two ORFs are identical for residue 343 and then each encodes 4 or 5 different C-terminal amino acids.

One further sequence characteristic is that nt 1252 is C in N33C (7), but T in A4 and A5 (Fig. 10).

one coded N33C (7),

1. It is not known whether this change represents a natural polymorphism, a cloning artifact, or a mutation in one or more of these clones.

Both predicted N33 polypeptides are highly homologous (p < ^-100 ) with the predicted C. elegans cDNA ZK686.3. The comparison was optimized by introducing four gaps into N33, giving 43% identical residues in the human and C. elegans gene (Fig. 15). Three N33 sub-regions with 12 to 21 residues (eg ⁸¹ PRNYSMIVMFTALQP) retain> 90% identity with ZK686.3, suggesting highly conserved functional motifs. On the other hand, the C. elegans gene

This change does not affect because it occurs beyond

The ORF forms of the stop codon lack homologous residues to the first 35 amino acids of N33, and N33 lacks approximately 16 amino acids internally compared to ZK686.3 (Fig. 15), suggesting a significant developmental divergence of transcriptional units. N33 was not significantly related to any other sequence in GENBANK, PIR, SWISS-PROT or EMBL.

18. Expression of N33 in tissues, tumors and cultured cells

Various selected cDNA clones were used to probe Northern blots containing mRNA from several normal human tissues; examples are shown in FIG. A single mRNA of about 1.5 kb was detected with N33 probes in most tissues, including heart, placenta, lung, liver, pancreas, prostate, testes, ovaries and colon. Expression in spleen, thyroid, small intestine and peripheral lymphocytes was low. Expression detected by another clone, J2, was seen mostly in skeletal muscle and testis, whereas two mRNAs detected by clone J28 were found essentially in placenta, testis and ovaries. Tumor suppressor gene expression is expected in tissues originating in the target tumor types, so that N33 but not J2 or J28 have expression patterns consistent with the prostate cancer suppressor gene, colorectal, and possibly other, cancer.

Northern analysis of mRNA from tumor cell lines showed N33 expression in 3 of 3 prostate lines 3 of 3 lung lines, but only in 1 of 14 colorectal cancer cell lines (Fig. 17). To further determine the significance of this finding, mucosa from a colon sample (precursor tissue for colon adenocarcinoma) was excised from the intestinal wall and tested for N33 mRNA: Specific expression was observed (Figure 18, lane 5).

Finally: Small amounts of total RNA were extracted from nine fresh prostate cancer samples (7 primary tumors and 2 metastases). Cryomicrotome-driven dissection was used to reduce the number of contaminating non-neoplastic cells in the primary samples, but a certain level (typically 20%) was irremovable. Due to the limiting amounts of available RNA, RT-PCR with N33-specific primers was used to quantify N33 expression. Primers from Rb, p53 and G3PD were used to control RNA quality and cDNA synthesis. Significantly reduced expression of N33 was observed in three cases (lanes 3, 6 and 9), where RNA of lane 6 was obtained from N2 tumor to verify the functioning of this assay. RNA lanes 3 and 9 were obtained from primary tumors that are expected to contribute some amount of message to N33 from non-neoplastic cells. These results, together with the absence of expression in colorectal cell lines, supported the identification of N33 as a candidate suppressor gene for prostate and colorectal tumors.

19. Mechanism of loss of N33 expression in tumor cells and tissues

The basis for loss of N33 expression in colorectal tumor cells and prostate cancer tissues is unknown, but may be due to somatic mutations (which affect, e.g., mRNA expression or stability), methylation changes, or other epigenetic regulatory factors. While one prostate tumor is known to have a large homozygous deletion in the 8p22 lane, the genetic status of other primary and metastatic prostate and colorectal tumors was determined in different ways as follows: 1) Southern tumor DNA transfers were hybridized with N33 cDNA probes and other 8p22 markers to detect homozygous deletions or genetic rearrangements. 2) The structure of the PTSG locus was determined by cloning / sequencing at the genomic DNA level by standard techniques. For example: A PI clone containing the N33 gene was isolated and sequenced with primers from the cDNA sequence, revealing the exon / intron interface and sequences in adjacent introns. Primers are synthesized to amplify each exon. Amplification and sequencing are then performed to detect the presence of fine small deletions or point mutations. 3) The presence of LOH is determined by comparing alleles of polymorphic markers in tumor vs. tumor. normal DNA of each patient. 4) A specific DNA methylation test is performed by comparing the results of Southern analysis of tumor DNA digested with enzymes sensitive and resp. insensitive to methylation. For example:

DNA digested with MspI and Hpall is compared. The VHL gene, a renal cell carcinoma tumor suppressor gene, is known to be somatically inactivated by methylation in some cases (Herman et al., Proc. Natl. Acad. Sci. U.S. A., 91: 9700-9704 (1994)).

20. Improved tools for detecting N33 inactivation by antibody visualized

Detection of N33 expression or absence thereof should be significantly simplified by immunohistochemical assays for N33 polypeptides in tissue sections. Antibodies reactive to one form of N33 protein were prepared as follows:

The 16-amino acid peptide of the C-terminus of N33 (Fig. 20) was coupled to KLH and used to immunize rabbits. After six weeks, serum was collected and antibodies were affinity purified on a peptide column. These polyclonal antibodies were tested in Western analysis of recombinant fused N33 proteins expressed in E. coli (Fig. 21). As described above, clones A4 and 6, respectively, were obtained. A5 (partial N33 proteins fused to the glutathione-S-reductase gene, carried on the pGEX-2T expression vector), which represented Form 1 and mRNA, respectively. 2. Protein expression was induced by IPTG and cell lysates were separated by PAGE and transferred to membrane. Western blotting was incubated with affinity purified polyclonal against N33 peptide, and the reactive bands were alkaline phosphatase conjugated to the secondary antibody and NBT / BCIP substrate reaction. The 57 kD fusion protein band was detected in induced cells containing an A4 clone but not A5 or other clones.

C. TRANSFER OF YAC TO MAMMAL CELLS

1. Re-adaptation of YAC clones by resistance to hygromycin

Plasmid vector pLUSH, which contains the telomeric tail regions of the YAC4 vector, bacterial Kan ^R gene, yeast auxotrophy gene Lys2, and mammalian hygromycin ^R gene (see map), was kindly provided by D. McElligott (Sripps Research Institute). pLUSH DNA was linearized by digestion with Sal I and 5-10 µg was used to transform YAC-containing yeast cells using an alkaline cation yeast transformation kit (Bio 101, Inc) according to the manufacturer's instructions. Cells were plated on triple drop-out media (trp ^- ura ^- lys ^- ) to select clones containing both YAC and conversion factor. Colonies were picked after

3-4 days and grown overnight in 2 ml YPD medium. Yeast DNA was prepared and tested for homologous integration of pLUSH using primers 5'-CTTGAGATCGGGCGTTCGACTCGC-3 'and 5'-TGAACGGTGATCCCCACCGGAATTG-3' (Hermanson et al., Nud. Adds Res. 19: 4943-4948 (1991)). Reactions were performed in 20 μΐ volumes with 100 ng of each primer in standard buffers plus 10% DMSO. Reaction conditions were 95 ° C, 2.5 min, then 35 cycles of 95 °, 40 sec; 60 ° C, 40 sec and 72 ° C, 2 min, followed by 72 ° C for 10 min. Homologous integration of the conversion vector resulted in an amplification of the 1855 bp band (Figure 6). The presence of the hygro ^R gene was confirmed by Southern analysis of yeast DNA with a radiolabeled probe of the hygro gene (Figure 7).

2. Spheroplast fusion and transformant selection

A variety of methods are available for transferring YAC transformants into mammalian cells. The Biol protocol was used. 13: 5469-5478 (1993) on Yeast cells were left

Silverman et al., Mol. Cell. fusion of spheroplasts. Briefly, grow by standard methods and centrifuged, washed and resuspended in isotonic medium and the cell wall was digested with a yeast lytic enzyme to form spheroplasts. These were superimposed on coiled cultured mammalian cells, such as NIH 3T3 cells or human tumor cells (in a 50: 1 ratio) and incubated in the presence of polyethylene glycol 1500 (Boehringer-Mannheim) for 2 min. at room temperature to induce fusion. Cells were diluted in tissue culture medium and incubated for 48 hours, after which selection for hygromycin was started. Hygro-resistant colonies were evident after approximately 3 weeks.

The YAC of interest was with genetic markers in YAC or analysis (Table 8). To be transmitted

3. Genetic analysis of transformants

The presence of a substantial part was verified by PCR amplification known for their detection by Southern

YACs to human cells were used with polymorphic alleles so that allele sizes in YACs differ from those already present in the parent cell. Retention of only a portion of YAC transferred can also be detected by these methods, and correlation of retained YAC portions with phenotypic properties can be used to localize tumor suppression activity to sub-regions from the YAC-covered region.

4. Phenotypic analysis of transformants

The phenotype of tumor cells after transfer of tumor suppressor genes can be determined regardless of the mode of transmission, eg, microcell transfer, retrovirus-mediated gene transfer, transfection, cell fusion, etc. To determine tumor suppressing activity and thus vector or gene insertion , the in vitro growth rates ( ³ H-thymidine incorporation), growth of soft agar transformants, and tumorigenicity in nude mice can be compared for modified and parent cells.

The foregoing examples were provided merely to illustrate, not to limit, the present invention. It will be understood that various modifications and additions may be made to the disclosure without departing from the spirit of the invention. Accordingly, the invention is defined by the following claims.

D. TESTING PTSG ACTIVITY IN TUMOR SUPRESS

PTSG activity in tumor suppression is determined both in in vitro cell culture conditions and in an nude mouse animal model. Any of the 13 colon cancer cell lines N33 shown in Figure 17 can be used to determine PTSG activity in tumor suppression (SW480, SW837, SW1417, HT-29, SW4O3, LS174T, DLD-1,

CACO-2, EB, SK-CO-1, RKO, HCT116 and COLO-302).

Briefly, the effect of PTSG on cell proliferation of the above lines is determined after expression of PTSG using an adenoviral expression vector. ACN is a control adenoviral vector lacking a cDNA insert, while AC-PTSG are adenoviral vectors that express PTSG products under the control of the human CMV promoter.

and

In vitro transcription-translation of PTSG

Plasmid pBS-N33C (7) was tested for its ability to produce a 39 kD protein in the TNT Coupled Reticulocyte Lysate System (Promega, Madison, Wisconsin). The T7 promoter in the Bluescript vector (Stratagene) allows transcription and translation of the PTSG coding sequence in rabbit reticulocytes. One microgram of mini-lysate DNA was placed per TnT reticulocyte reaction and incubated for 1 hour at 30 degrees Celsius. Ten microliters of the reaction mixture was mixed with sample buffer and electrophoresed in a 10% polyacrylamide gel (Novex) for 11 hours at 165 V. The gel was dried and exposed to film overnight.

Construction of adenoviral vectors containing PTSG

To construct recombinant adenoviruses, pBS-N33C insert (7) was obtained by digestion with EcoRI and cloned into the EcoRI site of pcDNA3 (Invitrogen) to produce pcDNA3-N33 clones. The orientation of the insert was tested by Κρη I digestion, and then clones with antisense orientation to the CMV promoter in pcDNA3 were used. To construct adenoviruses with Form 1, pcDNA3-N33 was digested with Xba I

- BamH I, and the insert was oriented into the Xba I - BamH I sites of the pAdCMVb vector oriented to give pACN33-1. For the construction of adenoviruses with Form 2, pBS-N33C (7) was digested with Ava III - EcoRI and the 5 'portion (nt 1 - 616) of the N33 insert was purified. Clone A5 was also digested with Ava III - EcoRI to release the 3 'portion of the N33 Form 2 insert (nt 617)

- EcoRI). The two halves of the gene were ligated and cloned into

EcoRI sites of pcDNA3. The orientation of the reconstructed insert was again tested by Κρη I digestion and sequencing. One clone with an antisense orientation was then digested with Xba I and BamH I and the insert cloned into pAdCMVb as above, giving pACN33-2.

The above plasmids are linearized with Nru I and co-transfected with a large fragment of d1309 mutants digested with Cla I (Jones and Shenk, Cell, 17: 683-689 (1979), incorporated herein by reference), using a CaPO ₄ transfection kit ( Stratagene). Viral plaques are isolated and recombinants are identified by both restriction digest analysis and PCR using primers against the PTSG cDNA sequence. Recombinant virus is further purified by limiting dilution, and viral particles are purified and titrated by standard methods (Graham and van der Erb, Virology 52: 456-457 (1973), Graham and Prevec, Manipulation of Adenovirus Vectors. In: Methods in Molecular Biology, Vol. 7, Gene Transfer and Expression Protocols, Murray EJ (ed.) The Humana Press Inc., Clifton NJ, 7: 109-128 (1991), both incorporated herein by reference).

To verify that the above PTSG vector expresses a protein of appropriate size, colon carcinoma cell lines were infected with either control or PTSG-containing recombinant adenoviruses for 24 hours at an increased plaque multiplicity of infection (MOI) generating virus / cell units. The cells were then washed once with PBS and collected in lysis buffer (50 mM Tris-HCl pH 7.5, 250 mM NaCl, 0.1% NP40, 50 mM NaF, 5 mM EDTA, 10 µg / ml aprotinin, 10 µg / ml leupeptin and 1 mM PMSF). Protein cells were resolved on 10% SDS-PAGE and transferred to nitrocellulose. Membranes were incubated with anti-PTSG antibody, followed by sheep anti-mouse IgG, conjugated with horseradish peroxidase. True expression of PTSG protein is visualized by chemiluminescence (ECL kit, Amersham) on Kodak XAR-5 film.

In vi tro

N33-negative colon cancer cells (selected from the cell lines shown in Figure 17) are seeded with 1 x 10 ⁴ cells per 100 mm dish in Kaighn F12 / DME medium (Irvin Scientific) supplemented with 10% FBS and 0.2 IU insulin (Sigma). Incubate overnight at 37 ° C in 7% CO ₂ . The next day, cells are given 10 ml of fresh growth medium and infected with either control ACN virus lysate (MOI 10) or AC-PTSG virus lysates (MOI 10) and incubated at 37 ° C. After 3 days, the medium is removed and the cells are fixed with a 1: 5 acetic acid-methanol solution. The cells were stained with 20% methanol-0.5% crystal violet for 30 minutes and washed with tap water to remove excess dye.

Thymidine incorporation is also used to determine the effect of PTSG on cell proliferation. Briefly, approximately 3 x 10 ³ cells are plated in each well of a 96-well plate (Costar) and allowed to incubate overnight (37 ° C, 7% CO ₂ ). Serial dilutions of ACN or AC-PTSG in DME are made: F12 / 15% FBS / 1% glutamine, and cells are infected at a multiplicity of infection (MOI) of 10 to 100 (4 replicates of wells per MOI) with each adenovirus. One half of the cell medium volume is exchanged 24 hours after infection and every 48 hours until harvest. 18 hours before collection, 1 μθβ of ³ H-thymidine (Amersham) is added to each well. Cells are harvested on glass fiber filters 5 days after infection and ³ H-thymidine incorporated into the cell nucleic acid is detected using liquid scintillation (TopCount ^R , Packard Instruments). Cell proliferation (cpm / well) at each MOI is expressed as a percentage of the average proliferation of untreated control cells.

Ex vivo gene therapy

To determine the effect of PTSG on tumorigenicity, the above tumor cell lines were tested for the ability to form tumors in nude mouse models. Approximately 2 x 10 ⁷ cells are plated in T225 flasks and treated with sucrose buffer containing ACN or AC-PTSG adenoviruses with an MOI of 3 or 30. After overnight infection, cells are harvested and approximately 10 ⁷ cells are injected subcutaneously on the left and the right flank of BALB / c nude mice (4 / group) that had previously received subcutaneous 17β-estradiol pellets. One flank is injected with ACN-treated cells while the collateral flank is injected with AC-PTSG-treated cells so that each mouse serves as its own control.

Animals receiving bilateral injections of untreated cells serve as additional controls for tumor growth. Tumor dimensions (length, width, height) and body weight are measured twice a week. Tumor volumes are determined for each animal, assuming spherical geometry with a radius equal to one-half the diameter of the measured tumor dimensions.

In vivo PTSG tumor suppression

Colon cancer cells are injected subcutaneously in female BALB / c athymic nude mice. Tumors are allowed to develop for 32 days. At this point, a single injection of either ACN (control) or AC-PTSG adenoviruses is injected into the peritumoral space surrounding the tumor. Tumors are then excised at day 2 or day 7 after adenovirus injection, and poly-A ⁺ RNA is isolated from each tumor. RT-PCR with PTSG-specific primers is then used to detect PTSG RNA in treated tumors. Amplification with actin primers serves as a control for the RT-PCR reaction, while plasmids containing recombinant PTSG sequences serve as a positive control for bands specific for recombinant PTSG.

In a separate experiment, cells are injected into the subcutaneous space on the right side of the mouse and the tumors are allowed to grow for 2 weeks. Mice receive peritumoral injections of buffer or recombinant virus twice a week for a total of 8 doses. Tumor growth is monitored throughout treatment in animals receiving ACN and buffer and in animals receiving AC-PTSG. Body weight and survival time are also monitored.

Although the invention has been described with reference to embodiments considered to be preferred at present, 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.

Table 1

8P CHROMOSE DELCE MAP IN HUMAN CANCER PROSTATE Loss of allele on chromosome 8 in prostate cancer

Locus Type polymorphism Probe Enzyme Localization Number cases Allele Loss / Informative Cases (All Tumors) Loss of allele / informative Ί cases (lymph node metastases) D8S140 RFLP ^a C18-115 MspI 8p23.2-23.3 49 4/28 (11) ^b 1/6 (16) D8S201 mikrosatelit Mfcl99 - 8p23 30 3/21 (14) 1/2 (49) D8S163 RFLP KSR2 Taql 8p22-pier 50 14/24 (61) 2/4 (49) MSR RFLP MSR32 MspI 8p22 50 20/29 (70) 3/3 (99) LPL mikrosatelit GZ14.15 - 8p22 45 15/32 (46) 0/2 (0) D8S220 RFLP C18-319 Taql 8p21.2-21.3 51 16/26 (59) 4/5 (79) NEFL RFLP NF5.1 Taql 8p21 12 2/6 (34) not studied D8S194 RFLP C18-277 MspI 8pll.21-11.22 51 3/20 (11) 1/4 (25) D8S39 RFLP MCT128.2 Taql 8q24 43 2/16 (12) 0/2 (0)

^a RFLP: restriction fragment length polymorphism ^b Numbers in parentheses: percent

Table 2. Polymorphic loci on chromosomal arm 8p included in mapping warp. Chromosomal intervals (A or B) are defined as in Wagner et al., (1991). All loci were used to assess * * associated YACs. : evaluation performed in Genethon, ⁺ : STS generated within the RFLP probe as described.

O. Name locus Chromo- somální interval Type Primer sequence (5'-3 ') Size product (bp) Hybr. pulse. (•C) link 1 D8S26 AND RFLP TAGCTCCTTCGAAACCCTCA TGGCAGGAAAAGCTCTCAAT 124 60 this message ⁺ 2 D8S511 AND STRP TTGTCCCTGTTGGCAGA TGATTTTTGTGTCCTGAAACTTA 135 55 this message 3 * D8S549 AND STRP AAATGAATCTCTGATTAGCCAAC TGAGAGCCAACCTATTTCTACC '170 55 this message 4 MSR26 AND RFLP TTCATCTATTGCATTCC CAAAATTTCAGCATGACAACTG 102 50 Matsumoto et al., 1990 5 D8S254 (B) STRP TGCCGGACATACATTAGTGA TTGTAAACACCACAAGCAGG '70 55 J. Weber communication 6 D8S233 (B) RFLP TTTGAGTAGCCAGAGTCCAG CGTACCATTTCCATCTGCT 84 55 this message 7 D8S261 (B) STRP TGCCACTGTCCTGAAAATCC TATGGCCCAGCAATGTGTAT '135 55 Tississnbach et al., 1992 8 D8S21 (B) RFLP CACTGAGGAAGAGGTTtGAAG ATCCATCACCAGGTTTGG 86 55 this message ⁺ 9 LPL (B) STRP ATCTGACCAAGGATAGTGGGAT CCTGGGTAACTGAGCGAGACT '130 60 Zulian and Hobbs, 1990 10 * D8S258 (B) STRP CTGCCAGGAATCAACTGAG TTGACAGGGACCCACG '150 55 Weissenbach et al., 1992

Table 3

Sequently tagged locations used to refine the physical map. Chromosomal intervals (A or B) are defined as in Wagner et al., (1991). Earlier reports were of D8S206 (Hudson et al., 1992), D8S294E and D8S297E (Adams et al., 1992); the other STSs were generated by partial sequencing of subcloned probes as described herein. * Detects Hind III RFLP. EST: designation of the expressed sequence (tag).

C. name locus / probes chromos. type interval primer sequences size (5'-3 ') of the product hybr. temp. ("C) D8S206 AND STOP gaaaaccatggctgggtg acatgcattagcactaccatgc -130 33 2 D8S2SME (B) EST TOACCTGAAATrACAAGGTA AGCAGCTTGACAATCTTAAG 12 33 3 D8S297E (B) EST 67 33 t.Li 1 AGl. 1 IjL-ALj 1 1 Li i LUAL CATTCTQACTACTACTTTCAO 4 El · AND accidental subclone TOACACACTTGCCATTTGAT U1 33 TTCCATTAGTCCCAGTTGTC 5 E3 accidental subclone 83 35 AND GCCTGTTTCATCGAACC HIVE. 1 LiLiCA 1 1 V. iLALU i AÚA O E15 accidental subclone 86 53 AND Lil IL.1 1 UL.CA 1 Ll 1 LiA l U IU GTCGCA1CIGCTTCTGG 7 E17 AND accidental subclone CAAGGCATATCACAACTGC GATAATTCAACTGTCACCTCTG 12t 55 8 E10 (B) accidental subclone 107 55 TQAATTTGCATAGTCTGCAG ΟΑΟΟτσΓλΑΟΆΑαοσΓΐΧΤΑ 9 Eli accidental subclone 97 55 AND TCaGGlCl3 iCaT TQGGAACTTCAAGCATAGG 10 E56 (B) accidental subclone 170 55 1 I lIiI I GAtRiALLAAAACAC lGICACUATGAGGA11U1ΙΛ 11 YE766 89 55 AND yac end gactv π occaccttotaaa AIVICCAAAUL. 1 AL 1 IC 1L.L 12 YEA43 (B) yac end aocaaagtgatggtggtaac GGACTAATTACCTCAGGCCT 82 55 13 32332 (B) yac end atggaaatgcacggga 173 55

CCATTCTGTCCCAATGATC ogo oo £ og oo g £ Ogoo

col

Z Z Z Z Z Z Z Z Z. O - OJ (D

JL £ zzzzzzzz;

: for 0. 0.0. · o

Τ ' ^κ κ

ZZZZZZZZZZi £ -á SžSž zzzzzžžžzzž Ž ZQ- tL žo. <La: O_jx Z Z Z

V »** K 3 zzzz z z z z z z z z: o. o. (to. o.:z of O š‘3 iiii žžmiiiinl.z Me and O s ΐ zzzz z z z z o. o z z z z: o. o. o. o. o. z zzzz Σ izi ín u _? z z zz z o. and o z z & 0.0. o. o. z z z X U. S </> a zzzz z z z z z z z o z z z z z z z z z z .... ... . ·; ·. ·. ··. ’·· r and. <.s lJ iii and <N Ο.Π Hive i and »N u-n Hive m and K‘S and ii O G “š tt e q " tn 3 iišl iii and < 1  · - ti * » l / t UJ of z z z < • 'ia.' I L> LU of tax ^ aa ^ zzzzzzzzz < i ^UQ 1 ΰ £ zzzz z z z a. and cl. α, αζζζζζζζζζ z z z of < ΙΛ uj ';

£ 3m

Η

HiiiiiiiMiMiiiiiii 1

E ^and L zzzzzzzzxf f for -aa-g-g-g · * f ** g * | xx | l? z z < ζζζζζζζχο.ο.α.α.α.αα.α.ο.αζζζζχχζζζχ zzz of z < iiiiMiMtžMMiiiiiiiii iii i E “3 “IiiiiiMMUMiiiiii and and> IX - O iMMMiiiii 1 ^q 5 iMMMiiiiiii iiiii E á zxza.a.aia-u.Q.a.a.zzxxxzxzzzzz < "WITH iiiMMMi i rí and·· O iMMiiiiiiiiiii ii E § ^1/1 ΙΛ Utt.t ».lVZZZZZXZZZ Z X Z Z Z z z z z zz zzz < κ “κ O. Q.Q. Q..Z z z z z z z z z z z z z z z z z z Z Z Z <

OOO γμ o _-

JJJJJJJJjJJJJJjJ JJJJJJJJJLUJJJJJJJJJJ ϊ§ sssíKSSsdidísÉssžKESSíísíÍSíÍ ^ ssí S

Table 5

Pair analysis of six markers in the panel of radiation hybrids Statistics were calculated with TWOPOINT. ^cR 5ooo ^{: cent} and ^ra irradiation 5000 Rad.

Marker AND Marker (B) D8S26 MSR D8S26 D8S233 D8S26 D8S261 D8S26 D8S21 D8S26 LPL MSR D8S223 MSR D8S261 MSR D8S21 MSR LPL D8S233 D8S261 D8S233 D8S21 D8S233 LPL D8S261 D8S21 D8S261 LPL D8S21 LPL

Observed clone count * " ⁺ " Thcta cR _{50 O} o LOD total

11 Λ j 73 92 9 6 j. 74 91 9 7 1 75 92 10 6 1 75 92 10 O 3 73 92 11 3 1 77 92 9 6 2 76 93 10 3 2 76 93 10 5 4 74 93 9 3 1 79 92 10 2 1 79 92 9 3 4 76 92 ll 0 1 81 93 10 1 4 78 93 11 1 3 78 93

0.3115 37 7.17 0.3394 41 5.99 0.3463 43 6.08 0.2948 35 7.21 0.3595 45 6.08 0.1741 19 Dec 9.87 0.3563 44 6.06 0.3026 36 7.20 0.3675 46 6.10 0.1995 22nd 8.31 0.1444 16 9.89 0.3147 38 6.45 0.0496 5 12.91 0.2305 26 8.46 0.1736 20 May 9.89

TABLE 6

Genomic subclones series name probe fragments human kb (enzyme) chrome. 8 Hind III fragments detected, kb

932E9

E

E

E

E

E

E

E

E

E

E

E

E

E

E

E

E

767H8 lc le

2d

10 ’

17 '

18 '

2θ '

23 '

31 '

1.6 (EcoRI)

1.0 (EcoRI)

0.85, 1.0 (EcoRI)

0.8 (EcoRI)

0.5, 4.5 (EcoRI + NotI)

1.2 (EcoRI + NotI)

0.25, 3.0 (EcoRI + NotI)

0.9 (HindIII + SacI)

0.3, 1.2 (EcoRI + Notl)

1.0, 1.8, 4.5 (HindIII) 1.5, 1.9 (EcoRI + NotI)

1.6 (EcoRI + NotI)

1.9, 2.5, 3 (EcoRI + NotI) 0.2, 0.5, 1.0, 2.1 (EcoRI) 0.2, 0.4, 1 (EcoRI + Style) 0.8, 1, 2 (EcoRI + SacII)

Yes 8 Yes [4 + 8] Yes 7.5 Yes 5.5 + 3.8 Yes “12 Yes 3.8 Yes 7 Yes 1,2 Yes 5 Yes 1.0 Yes 7 Yes 8 Yes 3.5 Yes “8 Yes 5.6 Yes 8.5

H 23 1, 1.3, 2.1 (EcoRI + SacII) Yes H 25 0.5, 1.7 (PstI + SacII) Yes H 29 1.0, 1.5 (EcoRI + SacII) Yes H 31 0.45 (EcoRI + SacII) Yes

802F11

1.6

4.2, 2.4, 1.3 ?? 7

F 832A10 4 0χ5, 1.6 (EcoRI + SacII) Yes 1.6 AND 33 0j 35 (EcoRI + SacI) Yes 1 AND 37 ixl, 3.2 (EcoRI + SacII) Yes 12 821F7 & 885C8 G 2 0χ5, 6 (HindIII + SacII) Yes “12 G 4 0.4 (Laughing) Yes 3.2 + high background G 10 0.25 6 (EcoRI + SacII) Yes “12 G 14 0.45 (EcoRI + SacII) Yes 2 G 18 1,2 (EcoRI + SacII) Yes 10

YAC end clones

YE1-766A12 YE1-843G3 0.25 (EcoRI) Yes Yes 5 YE1-932E9 0.6 (EcoRI) Yes 1,8 PAC Ele subclone PAC A2 Yes PAC A3 Yes PAC B3 Yes

* contains no Alu, * polymorphism, fragments without Alu used as probes

TABLE 7 selected cDNA series number probe fragments human det. HindIII ID or bp (enzyme) chromium. 8 fragments, kb homology selected with 802F11

J 2 325 BstXI Yes 2.5 J 10 '350 (BstXI) yes, ⁴ * H J 12 350 BstXI Yes 10 + 6 hom. to PP2Ca J 28 400 (BstXI) Yes 4 (+3, weak) new ORF P 3 350 BstXI Yes 12 MSR (1-280 bp) P 10 400 (BstXI) Yes 2.5 P 14 450 BstXI Yes 3.5 P 16 1 belt (BstXI) Yes 2.2 P 25 300 BstXI Yes 4 P 27 Mar: 450 + 350 (BstXI) Yes 4 sequ. with J28 P 28 400 (BstXI) Yes 12 (+5.2) MSR (1-450bp) P 34 250 BstXI Yes 3.8 MSR (3'UTR) & w 17 400 (RI + HII) Yes 3.5+ background HSDHEHC01 selected with K 26 821F7 or 877F2 350 RI + HIII Yes > 12 TO 27 Mar: 500 RI + HIII Yes > 12 TO 36 250 RI + HIII Yes 2.7 selected with L 3 946C9 400 (RI + HII) Yes 4 L 5 325 RI + HIII Yes 0.5 L 12 300 RI + HIII Yes 3.8 L 14 550 (RI + HII) Yes 6 + certain background L 21 450 (RI + HII) Yes 5.4 + 4 + 3 L 30 RI + HIII Yes 4.5 + certain background L 31 RI + HIII Yes 7 N 1 RI + HIII Yes 4.8 + certain background N 7 RI + HIII Yes 3+ background N 14 600 (RI + HII) Yes 2.6 N 18 250 RI + HIII Yes > 12 N 19 Dec 600 (RI + HII) Yes > 1,6 N 21 800 RI + HIII Yes overlaps with L21 N 27 Mar: 500 RI + HIII Yes 3 + certain background N 28 550 (RI + HII) Yes 4 overlaps with J28 N 33 RI + HIII Yes 12, 11, 4, 3.2, 2 overlaps with L21 N 35 RI + HIII Yes 12 N 36 RI + HIII Yes 12, 11, 4, 3, 2 overlaps with N33 X 27 Mar: 500 RI + HIII Yes 12 X 27 Mar: 500 RI + HIII Yes 1,8 selected with Q 30 932E9 500 RI + HIII Yes 3.5. selected with 946C9 and 932E9 X 1A1 U & L stripes X 1A8 X 1C8 * polymorphic, ** 6 kb is not Yes Yes Yes on chromosome 8

JA13 IN1SV'Λ

OHjAOAS λίΛ Qyd 'avy o

Claims (31)

PATENT CLAIMS An isolated and purified DNA sequence encoding a prostate tumor suppressor protein comprising a nucleotide sequence. id. no. 1. 9 6 .XI L l ΟΊ ^ Οα An isolated and purified DNA sequence encoding a prostate tumor suppressor protein comprising the nucleotide sequence of SEQ. ID No. 2. A recombinant vector comprising an isolated and purified DNA sequence according to claims 1 or 2. The recombinant vector of claim 3, wherein the vector is a cosmid, plasmid, or virus-derived. An expression vector comprising said DNA molecule according to claims 1 or 2, capable of inserting said DNA molecule into a mammalian host cell and expressing a protein therein. The expression vector of claim 5, wherein said vector is selected from the group consisting of a plasmid and a viral vector. The expression vector of claim 6, wherein the vector is a viral vector and is selected from the group consisting of a retroviral vector and an adenoviral vector. The expression vector of claim 7, wherein said expression vector is an adenoviral vector. A host-vector system for producing polypeptides or proteins having the biological activity of a PTSG protein or biologically active derivative thereof, comprising the vector of claims 5, 6, 7, or 8 in suitable host cells. - 67 10. A system designating a prokaryotic cell. The host-vector of claim 9, wherein the host cell is 11. A system characterized by a eukaryotic cell. the host-vector of claim 9, wherein the host cell is 12. A pharmaceutical composition comprising a vector of an acceptable carrier. according to claim 5 and pharmaceutically acceptable 13. A pharmaceutical composition comprising a vector of an acceptable carrier. according to claim 6 and pharmaceutically acceptable 14. A pharmaceutical composition comprising an AC-PTSG vector and a pharmaceutically acceptable carrier. 15. A DNA probe comprising at least complementary to the DNA sequence of claim 1. 16. A DNA probe comprising at least complementary to the DNA sequence of claim 2. 15 nucleotides 15 nucleotides An isolated and purified mammalian protein comprising the amino acid sequence of SEQ. ID No. 3. 18. An isolated and purified mammalian protein comprising the amino acid sequence of SEQ. ID No. 4. 68 T3 Xle <O 1 - X> O cc 2 -i - < σ o CPX n < proteins according to < < ^m ^ rzrcků 03 '°; —4 17 or 48, 19. A method of production characterized by a. inserting into the host cell a gene encoding the protein of claim 18 b. causing expression by said cell. thereby comprising the steps of a compatible expression vector of claim 17 or a host protein 20. The method of claim 19, wherein said host cell is selected from the group consisting of prokaryotic cells and eukaryotic cells. 21. The method of claim 20, wherein said host cell is a eukaryotic cell that is a mammalian cell, and said expression vector is compatible with said mammalian host cell. An isolated and purified DNA sequence according to claim 1 or 2 for suppressing the neoplastic phenotype of a cancer cell that does not have an endogenous PTSG protein. The isolated and purified DNA sequence according to claim 1 or 2 incorporated in a recombinant vector for suppressing the neoplastic phenotype of a cancer cell that does not have an endogenous PTSG protein. The isolated and purified mammalian protein of claim 17 or 18 for suppressing a neoplastic phenotype of a cancer cell that lacks an endogenous wild-type PTSG product, 25. An antibody that binds a PTSG peptide, consisting of the protein sequence of SEQ. ID No. 3. The antibody of claim 25, which binds to a PTSG protein having the amino acid sequence of SEQ. ID C. 3. 27. An antibody that binds a PTSG peptide consisting of the protein sequence of SEQ. ID No. 4. The antibody of claim 27, which binds to a PTSG protein having the amino acid sequence of SEQ. ID No. 4. 29. A method for detecting the absence of PTSG protein in tumor cells in vitro, comprising the steps of Preparation of tumor sections from the tumor: b. contacting the antibody of claims 25-28 with said tissue sections; and c. detecting the presence or absence of said antibody that binds to said tissue sections. 30. An RNA probe comprising at least about 15 nucleotides complementary to the DNA sequence of claim 1. 31. An RNA probe comprising at least about 15 nucleotides complementary to the DNA sequence of claim 2.