WO2000045852A1 - Method of treating cancer by restoration of pp32 function - Google Patents

Abstract

Methods to treat existing cancer or prevent cancer formation by restoration of pp32 function in cells lacking normal pp32 activity or expression are provided. The methods provide for restoration of pp32 either through transfection of cells with DNA expressing normal pp32, administration of an inducer of pp32 activity and/or inhibition of a nuclear phosphatase that is inhibited by pp32.

Description

Method of Treating Cancer by Restoration of pp32 Function

The work leading to this invention was supported in part by Grant No. RO1 CA 54404 from the National Institutes of Health. The U.S. Government retains certain rights in this invention. BACKGROUND File of the Invention

This invention is directed to various members of a gene family with transformation modulating activity, and to diagnostic and gene therapy techniques based on the variants. Review of Related Art

Prostatic adenocarcinoma is the most frequent malignancy in adult men with approximately 317,000 new cases diagnosed each year (Parker, et al., CA, 46:8-27, 1996). In spite of the capabilities for early diagnosis and treatment (Potosky, et al., JAMA, 273:548-552, 1995), it represents the second leading cause of cancer death in men following lung cancer.

To date, the study of alterations in specific genes has not been particularly rewarding in primary prostate cancer. Most alterations in the widely studied oncogenes and tumor suppressor genes occur in only 20 - 30% of primary prostate carcinomas, except for the myc gene, where overexpression has been observed in as many as 50 - 60% of such cases (Fleming, et al., Cancer Res., 46: 1535-1538, 1986). Up to 40% of primary prostate cancers studied by comparative genomic hybridization display chromosomal aberrations (Visakorpi, et al., Cancer Res., 55:342-347, 1995), although such alterations occur more frequently as tumors recur and become refractory to hormonal therapy. Characterization of candidate proto-oncogenes or tumor suppressor genes at such altered loci may eventually shed light on tumor progression in the prostate. pp32 (GenBank HSU73477) is a highly conserved nuclear phosphoprotein. Increased expression of pp32 or closely related species is a frequent feature of clinical cancers. For example, in human prostate cancer, high-level expression of RNA hybridizing with pp32 probes occurs in nearly 90% of clinically significant prostate cancers, in contrast to the substantially lower frequencies of alterations of other oncogenes and tumor suppressors (See U.S. Patent No. 5,726,018, incorporated herein by reference). Molecular Features and Activities ofpp32. pp32 is a nuclear phosphoprotein that is differentiation-regulated during differentiation of adult prostatic epithelium (Walensky, et al., Cancer Res. 53:4720-4726, 1993). The human pp32 cDNA sequence (Gen-Bank U73477) is 1052 bp in length and encodes a protein of 249 amino acids. The protein is composed of two domains: an amino terminal amphipathic α-helical region containing a leucine zipper, and a highly acidic carboxyl terminal region. The murine and human forms of pp32 are highly conserved with over 90% nucleic acid homology and over 95% protein-level homology.

Human pp32 has been isolated independently by a number of groups. Vaesen et al. ("Purification and characterization of two putative HLA class II associated proteins: PH API and PHAPII." Biol. Chem. Hoppe-Seyler., 375: 1 13-126, 1994) cloned an essentially equivalent molecule, termed PHAPI, from an EBV-transformed human B-lymphoblastoid cell line; PHAPII, cloned by the same strategy, is unrelated to pp32. This study identified PHAPI through its association in solution with human HLA class II protein, noting membrane and cytoplasmic localization as well as nuclear; the gene has putatively been localized to chromosome 15q22.3-q23 by fluorescent in situ hybridization (Fink, et al., "Localization of the gene encoding the putative human HLA class II-associated protein (PHAPI) to chromosome 15q22.3-q23 by fluorescence in situ hybridization." Genomics, 29:309-310, 1995). More recently, a group studying inhibitors of protein phosphatases identified pp32 as IlPP2a, an inhibitor of protein phosphatase 2a (Li, et al., "Molecular Identification of II PP2 A, a novel potent heat-stable inhibitor protein of protein phosphatase 2A." Biochemistry 35:6998-7002, 1996); another phosphatase inhibitor, I2PP2a, is unrelated to pp32. Interestingly, another recent report (Ulitzur, et al., "Biochemical characterization of mapmodulin, a protein that binds microtubule-associated proteins." Journal of Biological Chemistry 272:30577-30582, 1997) identified pp32 as a cytoskeletally-associated cytosolic protein in CHO cells. It is not clear whether this finding stems from a difference in system, or whether pp32 can localize to the cytoplasm under certain circumstances. pp32 has also been identified as LANP, a leucine rich nuclear protein in the central nervous system (Matsuoka, et al., "A nuclear factor containing the leucine-rich repeats expressed in murine cerebellar neurons. Proc Nail Acad Sci USA 91 :9670-9674, 1994). There are also a number of reports of gene products bearing lesser degrees of homology to pp32. The Vaesen group has identified a series of unpublished sequences, termed PHAPI2a (EMBL Locus HSPHAPI2A) and PHAPI2b (EMBL Locus HSPHAPI2B), also cloned from an EBV-transformed human B-lymphoblastoid cell line. These variant pp32 sequences are distinct from the sequences reported herein, representing the April protein instead. April, cloned from human pancreas, is shorter than PHAPI2a by two N-terminal amino acids (Mencinger, et al., "Expression analysis and chromosomal mapping of a novel human gene, APRIL, encoding an acidic protein rich in leucines." Biochimica el Biophysica Acta. 1395: 176-180, 1998, see EMBL Locus HSAPRIL); PHAPI2b is identical to a subset of APRIL. Silver-stainable protein SSP29 (unpublished GenBank Locus HSU70439) was cloned from HeLa cells and is identical to PHAPI2a.

The nuclear phosphoprotein pp32 has been linked to proliferation. Malek and associates reported that various neoplastic cell lines showed markedly elevated expression levels and that bacterial polysaccharide induced expression of pp32 epitopes by B lymphocytes upon polyclonal expansion (Malek, et al., J. Biol. Chem., 265: 13400-13409, 1990). Walensky and associates reported that levels of pp32 expression, measured by in situ hybridization, increased in direct relation to increasing Gleason grade of human prostatic cancers. pp32 cDNA probes hybridize strongly with prostatic adenocarcinoma, whereas the hybridization signal in normal prostate is confined to basal cells. Polyclonal anti-pp32 antibodies react strongly with sections of human prostatic adenocarcinoma. The antibodies and riboprobes used by the investigators in previous studies are consistent with cross- reactivities of the reagents with all reported members of the pp32 nuclear phosphoprotein family, therefore, while previous descriptions focused upon pp32, it cannot be excluded that homologous proteins were detected. SUMMARY OF THE INVENTION

This invention is based upon the fact that the tumor suppressor function of pp32 is altered in cancer by decreased or absent expression through regulatory means not involving structural changes to the genome such as mutation or chromosomal loss. The object of the invention is a method of treating cancer by restoration of pp32 function to treat existing cancer or prevent cancer formation by: i) Induction of the endogenous pp32 gene by pharmacologic or physiologic means in cancer cells lacking expression of normal pp32 as a form of endogenous gene therapy; or ii) Restoration of pp32 function by pharmacologic or physiologic means; or, more specifically iii) Restoration of pp32 function by inhibition of nuclear phosphatases inhibitable by pp32 by pharmacologic or physiologic means. This invention provides a method of treating certain malignant cells which exhibit subnormal expression of normal, tumor suppressive pp32 and/or overexpression of a tumorigenic pp32 variant, the method comprising restoring tumor suppressive pp32 function to the cells. The methods of this invention may be used in treating neuroendocrine, neural, mesenchymal, lymphoid, epithelial, or germ cell derived tumors, and particularly breast and prostate carcinomas.

In a particular embodiment, the method of this invention for treating malignant cells comprises inducing increased expression of normal pp32 and/or decreased expression of tumorigenic pp32 variant(s) in the cells. The increased expression of normal pp32 may be induced by transfection of the cells with DNA expressing normal pp32, or the increased expression of normal pp32 may be induced by administering an inducer of normal pp32 expression. Typically, the inducer of normal pp32 expression will be a small molecule (as opposed to a biological macromolecule) and will affect the regulatory mechanism for regulating expression of pp32 and/or its variants. The invention also provides a method of testing a compound to determine whether it is an inducer of pp32 expression comprising measuring expression of pp32 and/or one or more of its variants by cells cultured in the presence and absence of said compound, where increased expression of normal pp32, or reduced expression of tumorigenic pp32 variant(s), in the presence of the compound indicates that the compound is an inducer of pp32 expression.

In another embodiment, this invention provides a method of treating malignant cells exhibiting subnormal expression of normal pp32 and/or overexpression of a pp32 variant, where the method comprises inhibiting a protein phosphatase in the cells, in particular, a protein phosphatase that is inhibited by the presence of pp32. The protein phosphatase may be a nuclear protein phosphatase, such as a member of the protein phosphatase 2A group. The invention also provides a method of testing a compound to determine whether it is an inducer of pp32 function comprising measuring protein phosphatase activity in cells cultured in the presence and absence of the compound, using cells where expression of pp32 in the cells inhibits protein phosphatase activity. Inhibition of protein phosphatase activity by the compound in such a test is an indication that that compound induces pp32 function in the cells. The protein phosphatase may be a nuclear protein phosphatase, such as a member of the protein phosphatase 2 A group. pp32 is a member of a highly conserved family of differentiation-regulated nuclear proteins that is highly expressed in nearly all human prostatic adenocarcinomas of Gleason Grade > 5. This contrasts with the low percentage of prostate tumors that express molecular alterations in proto-oncogenes or demonstrate tumor suppressor mutation or loss of heterozygosity. By analysis of specimens of human prostatic adenocarcinoma and paired adjacent normal prostate from three individual patients, the inventors have shown that normal prostate continues to express normal pp32, whereas three of three sets of RT-PCR- amplified transcripts from prostatic adenocarcinomas display multiple cancer-associated coding sequence changes. The cancer-associated sequence changes appear to be functionally significant. Normal pp32 exerts antineoplastic effects through suppression of transformation. In contrast, cancer-associated pp32 variants augment, rather than inhibit, transformation.

The PC-3 human prostatic adenocarcinoma cell line does not express normal pp32. Transfection of normal pp32 into the PC-3 cell line results in a cell-cycle blockade. The LNCaP human prostatic adenocarcimona cell line expresses normal pp32 and appears to have a lesion elsewhere in the pathway. Transfection of normal pp32 into the LNCaP cell line does not block progression through the cell cycle. Since pp32 expression is apparently not lost through a process of mutation and loss of heterozygosity, but is instead downregulated by other means with concomitant induction of expression of tumorigenic members of the pp32 gene family, this experiment demonstrates the feasibility of restoration of pp32 function by any means (see above) to treat or prevent cancer. Since pp32 may act wholly or in part through inhibition of a certain class of phosphatases, restoration of the phosphatase inhibition by any means may be sufficient to restore pp32 function. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 A shows detection of pp32-related mRNA in benign prostate and prostate cancer tissue sections by in situ hybridization.

Figure IB shows immunohistochemical stain of prostate cancer sections with anti- pp32 antibodies.

Figure 2 shows the genomic sequence of variant pp32rl isolated from human placenta.

Figure 3 provides a base-by-base comparison of the sequence of pp32rl (top) with normal human pp32 (bottom). The numbering system for pp32rl corresponds to Figure 1, and the numbering system for normal pp32 is taken from Chen, et al. Nucleotide base differences are underlined in the pp32rl sequence. Sequences within the normal pp32 sequence missing in pp32rl are represented by dashes. The open reading frame for pp32rl is indicated by overlining.

Figure 4 shows the alignment of the pp32rl amino acid sequence (top) with normal human pp32 (bottom). Residue changes are underlined in the pp32rl sequence. Amino acids missing in the pp32rl sequence compared to normal pp32 are represented by dashes. Figure 5 shows the genomic sequence of variant pp32r2.

Figure 6A shows RT-PCR amplification of pp32 and pp32 variants from human prostate cancer and prostate cancer cell line.

Figure 6B shows cleavase fragment length polymorphism analysis of pp32 detects variant pp32 transcripts in human prostate cancer. Figure 7 shows the alignment of nucleic acid (A) and amino acid (B) sequences from human prostatic adenocarcinoma and prostatic adenocarcinoma cell lines with pp32.

Figure 8 is a bar graph showing ras + myc induced transformed focus formation. Co-transfection with a pp32 expression vector reduces transformation, while co- transfection with a pp32rl expression vector stimulates transformation. Figure 9 is a bar graph showing pp32rl stimulation of ras + myc induced transformed focus formation. Co-transfection with a pp32 expression vector reduces transformation, while co-transfection with expression vectors for pp32rl sequences from prostate cancer cell lines stimulate transformation. Figure 10 is a graph of transformation assay results for cells transfected with variant pp32 species, which are shown to stimulate transformation with variable potency. DETAILED DESCRIPTION OF THE INVENTION

The inventors have discovered that phenotypic changes in pp32 are a common feature of human prostate cancer. Previous data show that 87% of prostate cancers of Gleason Score 5 and above express pp32 or closely-related transcripts (U.S. Patent No. 5,734,022, incorporated herein by reference). This is striking in comparison to the frequency of molecular alterations in other widely studied oncogenes and tumor suppressor genes in primary prostatic adenocarcinoma, which occur in a substantially smaller proportion of cases. For example, myc overexpression (Fleming, et al.) occurs in around 60% of cases, and p53 is abnormal in only around 25% of primary tumors (Isaacs, et al., in "Genetic Alterations in Prostate Cancer." Cold Spring Harbor Symposia on Quantitative Biology. 59:653-659, 1994).

Several lines of evidence suggest that pp32 may act as a tumor suppressor. Functionally, pp32 inhibits transformation in vitro by oncogene pairs such as ras with myc, mutant p53, Ela, or jun, or human papilloma virus E6 and E7 (Chen, et al., "Structure of pp32, an acidic nuclear protein which inhibits oncogene-induced formation of transformed foci." Molecular Biology of the Cell. 7:2045-2056, 1996). pp32 also inhibits growth of transformed cells in soft agar (Chen, et al.). In another system, ras-transfected NIH3T3 cells previously transfected to overexpress normal human pp32 do not form foci in vitro or, preliminarily, do not form tumors in nude mice, unlike control cells. In contrast, knockout of endogenous pp32 in the same system by an antisense pp32 expression construct markedly augments tumorigenesis (Example 12 below).

In clinical prostate cancer, the situation at first appears counterintuitive. Most human prostate cancers seem to express high levels of pp32 by in situ hybridization (see Example 1 below) and stain intensely with anti-pp32 antibodies. Because pp32 inhibits oncogene-mediated transformation (Chen, et al.), its paradoxical expression in cancer was investigated at the sequence level. The paradoxical question of why prostate cancers seem to express high-levels of an anti-oncogenic protein was addressed by comparing the sequence and function of pp32 species from paired normal prostate and adjacent prostatic carcinoma from three patients as well as from four prostate cancer cell lines. pp32 is expressed in benign prostate but pp32rl and pp32r2, closely-related genes located on different chromosomes, are expressed in prostate cancer. pp32 is a tumor suppressor whereas pp32rl and pp32r2 are tumorigenic. Alternate usage of the pp32, pp32rl, and pp32r2 genes thus modulates oncogenic potential in human prostate cancer. Prostate cancers express little or no pp32 but do express other members of the pp32 family encoded by genes on separate chromosomes. The alternate pp32 genes expressed in prostate cancer are tumorigenic, in marked contrast to pp32. It is demonstrated herein that pp32 is a member of a closely-related gene family, and that alternate expression of these closely- related genes located on different chromosomes modulates oncogenic potential in human prostate cancer. The variant pp32 species expressed in prostate cancer are closely related to pp32.

The present data indicate that prostate cancers express variant pp32 transcripts, whereas adjacent normal prostate expresses normal pp32. Two instances clearly show that expression of alternate genes on different chromosomes can lead to the phenotypic switch, rather than mutation or alternate splicing. This switch in molecular phenotype is accompanied by a switch in functional pp32 phenotype. Normal pp32 is anti-oncogenic in character, in contrast to the pro-oncogenic variant transcripts that foster oncogene- mediated transformation. The high frequency of this abnormality suggests that expression of variant pp32 species may play an etiologic role in the development of human prostate cancer. In addition, these findings have significant diagnostic and prognostic implications. Definitions

In describing the present invention, the following terminology is used in accordance with the definitions set out below. Nucleic Acids In discussing the structure of particular double-stranded DNA molecules, sequences may be described herein according to the normal convention of giving only the sequence in the 5' to 3' direction along the nontranscribed stand of DNA (i.e., the strand having a sequence homologous to the mRNA).

A DNA sequence "corresponds" to an amino acid sequence if translation of the DNA sequence in accordance with the genetic code yields the amino acid sequence (i.e., the DNA sequence "encodes" the amino acid sequence); one DNA sequence "corresponds" to another DNA sequence if the two sequences encode the same amino acid sequence.

Two DNA sequences are "substantially similar" when at least about 90% (preferably at least about 94%, and most preferably at least about 96%) of the nucleotides match over the defined length of the DNA sequences. Sequences that are substantially similar can be identified by the assay procedures described below or by isolating and sequencing the DNA molecules. See e.g., Maniatis et al., infra, DNA Cloning, vols. 1 and II infra; Nucleic Acid Hybridization, infra.

A "heterologous" region or domain of a DNA construct is an identifiable segment of DNA within a larger DNA molecule that is not found in association with the larger molecule in nature. Thus, when the heterologous region encodes a mammalian gene, the gene will usually be flanked by DNA that does not flank the mammalian genomic DNA in the genome of the source organism. Another example of a heterologous region is a construct where the coding sequence itself is not found in nature (e.g., a cDNA where the genomic coding sequence contains introns, or synthetic sequences having codons different than the native gene). Allelic variations or naturally-occurring mutational events do not give rise to a heterologous region of DNA as defined herein.

A "coding sequence" or "open reading frame" is an in-frame sequence of codons that (in view of the genetic code) correspond to or encode a protein or peptide sequence. Two coding sequences correspond to each other if the sequences or their complementary sequences encode the same amino acid sequences. A coding sequence in association with appropriate regulatory sequences may be transcribed and translated into a polypeptide in vivo. A polyadenylation signal and transcription termination sequence will usually be located 3' to the coding sequence. A "promoter sequence" is a DNA regulatory region capable of binding RNA polymerase in a cell and initiating transcription of a downstream (3' direction) coding sequence. Promoter sequences typically contain additional sites for binding of regulatory molecules (e.g., transcription factors) which affect the transcription of the coding sequence. A coding sequence is "under the control" of the promoter sequence or "operatively linked" to the promoter when RNA polymerase binds the promoter sequence in a cell and transcribes the coding sequence into mRNA, which is then in turn translated into the protein encoded by the coding sequence.

Vectors are used to introduce a foreign substance, such as DNA, RNA or protein, into an organism. Typical vectors include recombinant viruses (for DNA) and liposomes (for protein). A "DNA vector" is a replicon, such as plasmid, phage or cosmid, to which another DNA segment may be attached so as to bring about the replication of the attached segment. An "expression vector" is a DNA vector which contains regulatory sequences which will direct protein synthesis by an appropriate host cell. This usually means a promoter to bind RNA polymerase and initiate transcription of mRNA, as well as ribosome binding sites and initiation signals to direct translation of the mRNA into a polypeptide. Incorporation of a DNA sequence into an expression vector at the proper site and in correct reading frame, followed by transformation of an appropriate host cell by the vector, enables the production of a protein encoded by said DNA sequence.

An expression vector may alternatively contain an antisense sequence, where a small DNA fragment, corresponding to all or part of an mRNA sequence, is inserted in opposite orientation into the vector after a promoter. As a result, the inserted DNA will be transcribed to produce an RNA which is complementary to and capable of binding or hybridizing with the mRNA. Upon binding to the mRNA, translation of the mRNA is prevented, and consequently the protein coded for by the mRNA is not produced. Production and use of antisense expression vectors is described in more detail in U.S. Patent 5, 107,065 (describing and exemplifying antisense regulation of genes in plants) and U.S. Patent 5, 190,931 (describing antisense regulation of genes in both prokaryotes and eukaryotes and exemplifying prokaryotes), both of which are incorporated herein by reference.

"Amplification" of nucleic acid sequences is the in vitro production of multiple copies of a particular nucleic acid sequence. The amplified sequence is usually in the form of DNA. A variety of techniques for carrying out such amplification are described in a review article by Van Brunt (1990, Bio/Technol., 8(4):291 -294). Polymerase chain reaction or PCR is a prototype of nucleic acid amplification, and use of PCR herein should be considered exemplary of other suitable amplification techniques. Polypeptides

For the purposes of defining the present invention, two proteins are homologous if 80%) of the amino acids in their respective amino acid sequences are the same; for proteins of differing length, the sequences will be at least 80% identical over the sequence which is in common (i.e., the length of the shorter protein). Two amino acid sequences are "substantially similar" when at least about 87% of the amino acids match over the defined length of the amino acid sequences, preferably a match of at least about 89%, more preferably a match of at least about 95%. Typically, two amino acid sequences which are similar will differ by only conservative substitutions.

"Conservative amino acid substitutions" are the substitution of one amino acid residue in a sequence by another residue of similar properties, such that the secondary and tertiary structure of the resultant peptides are substantially the same. Conservative amino acid substitutions occur when an amino acid has substantially the same charge or hydrophobicity as the amino acid for which it is substituted and the substitution has no significant effect on the local conformation of the protein. Amino acid pairs which may be conservatively substituted for one another are well-known to those of ordinary skill in the art.

The polypeptides of this invention encompass pp32rl and pp32rl analogs, pp32r2 and pp32r2 analogs, along with other variants of pp32 and their analogs. pp32rl and pp32r2 are naturally occurring, mature proteins, and further encompass all precursors and allelic variations of pp32rl and pp32r2, as well as including forms of heterogeneous molecular weight that may result from inconsistent processing in vivo. An example of the pp32rl sequence is shown in Figure 3, top line. "pp32rl analogs" are a class of peptides which includes:

1 ) "Allelic variations of pp32rl ," which are polypeptides which are substantially similar to pp32rl . Preferably the amino acid sequence of the allelic variation is encoded by a nucleic acid sequence that differs from the sequence of pp32rl by one nucleotide in 300;

2) "Truncated pp32rl peptides," which include fragments of either pp32 or allelic variations of pp32rl that preferably retain either (i) an amino acid sequence unique to pp32rl, (ii) an epitope unique to pp32rl or (iii) pp32rl activity; 3) "pp32rl fusion proteins," which include heterologous polypeptides which are made up of one of the above polypeptides (pp32rl, allelic variations of pp32rl or truncated pp32rl peptides) fused to any heterologous amino acid sequence.

"Unique" sequences of the pp32rl variant according to this invention, either amino acid sequences or nucleic acid sequences which encode them, are sequences which are identical to a sequence of a pp32rl polypeptide, but which differ in at least one amino acid or nucleotide residue from the sequences of human pp32 (Genbank Locus HSU73477), murine pp32 (Genbank Locus MMU73478), human cerebellar leucine rich acidic nuclear protein (LANP) (Genbank Locus AF025684), murine LANP (Genbank Locus AF022957), IlPP2a or human potent heat-stable protein phospatase 2a inhibitor (Genbank Locus HSU60823), SSP29 (Genbank Locus HSU70439),HLA-DR associated protein I (Genbank Locus HSPPHAP1, Accession No. X75090), PHAPI2a (EMBL Locus HSPHAPI2A, Genbank Accession No. Y07569), PHAPI2b (EMBL Locus HSPHAPI2B, Genbank Accession No. Y07570), and April (EMBL Locus HSAPRIL), and preferably, are not found elsewhere in the human genome. (A list of these sequences is provided in Table 3 A.) Similarly, an epitope is "unique" to pp32rl polypeptides if it is found on pp32rl polypeptides but not found on any members of the set of proteins listed above. Analogs of pp32r2 and unique pp32r2 sequences are defined similarly. Of course, unique sequences of pp32rl are not found in pp32r2 and vice versa.

"Variants of pp32" are homologous proteins which differ from pp32 by at least 2 amino acids. In particular, sequence comparison between pp32 and a variant will demonstrate at least one segment of 10 amino acids in which the sequence differs by at least two (2) amino acids. More typically a variant will exhibit at least two such 10 amino acid segments. Preferably, variants of pp32 in accordance with this invention will exhibit differences in functional activity from pp32. In particular, pp32rl and pp32r2 are variants of pp32 whose activity includes stimulation of transformation in the rat fibroblast transformation assay described herein.

A composition comprising a selected component A is "substantially free" of another component B when component A makes up at least about 75% by weight of the combined weight of components A and B. Preferably, selected component A comprises at least about 90%) by weight of the combined weight, most preferably at least about 99% by weight of the combined weight. In the case of a composition comprising a selected biologically active protein, which is substantially free of contaminating proteins, it is sometimes preferred that the composition having the activity of the protein of interest contain species with only a single molecular weight (i.e., a "homogeneous" composition). As used herein, a "biological sample" refers to a sample of tissue or fluid isolated from a individual, including but not limited to, for example, plasma, serum, spinal fluid, lymph fluid, the external sections of the skin, respiratory, intestinal, and genitourinary tracts, tears, saliva, milk, blood cells, tumors, organs, and also samples of in vivo cell culture constituents (including but not limited to conditioned medium resulting from the growth of cells in cell culture medium, putatively virally infected cells, recombinant cells, and cell components).

"Human tissue" is an aggregate of human cells which may constitute a solid mass. This term also encompasses a suspension of human cells, such as blood cells, or a human cell line. The term "immunoglobulin molecule" encompasses whole antibodies made up of four immunoglobulin peptide chains, two heavy chains and two light chains, as well as immunoglobulin fragments. "Immunoglobulin fragments" are protein molecules related to antibodies, which are known to retain the epitopic binding specificity of the original antibody, such as Fab, F(ab)'₂, Fv, etc. Two polypeptides are "immunologically cross- reactive" when both polypeptides react with the same polyclonal antiserum. General Methods

The practice of the present invention employs, unless otherwise indicated, conventional molecular biology, microbiology, and recombinant DNA techniques within the skill of the art. Such techniques are well known to the skilled worker and are explained fully in the literature. See, e.g., Maniatis, Fritsch & Sambrook, "Molecular Cloning: A Laboratory Manual" (1982); "DNA Cloning: A Practical Approach," Volumes I and II (D.N. Glover, ed., 1985); "Oligonucleotide Synthesis" (M.J. Gait, ed., 1984); "Nucleic Acid Hybridization" (B.D. Hames & S.J. Higgins, eds., 1985); "Transcription and Translation" (B.D. Hames & S.J. Higgins, eds., 1984); "Animal Cell Culture" (R.I. Freshney, ed., 1986); "Immobilized Cells and Enzymes" (IRL Press, 1986); B. Perbal, "A Practical Guide to Molecular Cloning" (1984), and Sambrook, et al., "Molecular Cloning: a Laboratory Manual" (1989). pp32 Related Genomic DNA

Screening a human genomic library in bacteriophages with probes generated from human pp32 cDNA yielded a new sequence that contained an open reading frame encoding a protein homologous with pp32 (see Example 2; pp32 sequence, reported in Chen, et al., Mol. Biol. Cell, -.2045-2056, 1996). While the pp32rl and pp32r2 sequences (see Figures 2 and 5) are substantially homologous to pp32, multiple single nucleotide base changes and short deletions suggest that they are encoded by gene distinct from pp32 gene. The pp32 family also includes substantially homologous polypeptides reported by others: HLA-DR associated protein I (Vaesen, 1994), leucine-rich acidic nuclear protein (Matsuoka, 1994), and protein phosphatase 2 A inhibitor (Li, 1996).

DNA segments or oligonucleotides having specific sequences can be synthesized chemically or isolated by one of several approaches. The basic strategies for identifying, amplifying and isolating desired DNA sequences as well as assembling them into larger DNA molecules containing the desired sequence domains in the desired order, are well known to those of ordinary skill in the art. See, e.g., Sambrook, et al., (1989); B. Perbal, (1984). Preferably, DNA segments corresponding to all or a part of the cDNA or genomic sequence of pp32rl may be isolated individually using the polymerase chain reaction (M. A. Innis, et al., "PCR Protocols: A Guide To Methods and Applications," Academic Press, 1990). A complete sequence may be assembled from overlapping oligonucleotides prepared by standard methods and assembled into a complete coding sequence. See, e.g.. Edge π 981) Nature 292:756; Nambair, et al. (1984) Science 223: 1299; Jay, et al. (1984) J. Biol. Chem.. 259:6311. The assembled sequence can be cloned into any suitable vector or replicon and maintained there in a composition which is substantially free of vectors that do not contain the assembled sequence. This provides a reservoir of the assembled sequence, and segments or the entire sequence can be extracted from the reservoir by excising from DNA in the reservoir material with restriction enzymes or by PCR amplification. Numerous cloning vectors are known to those of skill in the art, and the selection of an appropriate cloning vector is a matter of choice (see, e.g., Sambrook, et al., incorporated herein by reference). The construction of vectors containing desired DNA segments linked by appropriate DNA sequences is accomplished by techniques similar to those used to construct the segments. These vectors may be constructed to contain additional DNA segments, such as bacterial origins of replication to make shuttle vectors (for shuttling between prokaryotic hosts and mammalian hosts), etc.

Procedures for construction and expression of proteins of defined sequence are well known in the art. A DNA sequence encoding pp32rl, pp32r2, or an analog of either pp31 Rl or pp32r2, can be synthesized chemically or prepared from the wild-type sequence by one of several approaches, including primer extension, linker insertion and PCR (see, e.g., Sambrook, et al.). Mutants can be prepared by these techniques having additions, deletions and substitutions in the wild-type sequence. It is preferable to test the mutants to confirm that they are the desired sequence by sequence analysis and/or the assays described below. Mutant protein for testing may be prepared by placing the coding sequence for the polypeptide in a vector under the control of a promoter, so that the DNA sequence is transcribed into RNA and translated into protein in a host cell transformed by this (expression) vector. The mutant protein may be produced by growing host cells transfected by an expression vector containing the coding sequence for the mutant under conditions whereby the polypeptide is expressed. The selection of the appropriate growth conditions is within the skill of the art. The assembled sequence can be cloned into any suitable vector or replicon and maintained there in a composition which is substantially free of vectors that do not contain the assembled sequence. This provides a reservoir of the assembled sequence, and segments or the entire sequence can be extracted from the reservoir by excising from DNA in the reservoir material with restriction enzymes or by PCR amplification. Numerous cloning vectors are known to those of skill in the art, and the selection of an appropriate cloning vector is a matter of choice (see, e.g., Sambrook, et al., incorporated herein by reference). The construction of vectors containing desired DNA segments linked by appropriate DNA sequences is accomplished by techniques similar to those used to construct the segments. These vectors may be constructed to contain additional DNA segments, such as bacterial origins of replication to make shuttle vectors (for shuttling between prokaryotic hosts and mammalian hosts), etc. Producing the Recombinant Peptide

Preferably, DNA from the selected clones should be subcloned into an expression vector, and the protein expressed by cells transformed with the vector should be tested for immunoreactivity with antibodies against the recombinant protein of this invention prepared as described below. Such subcloning is easily within the skill of the ordinary worker in the art in view of the present disclosure. The amino acid coding region of the DNA sequence of this invention may be longer or shorter than the coding region of the disclosed sequence, so long as the recombinant peptide expressed by the DNA sequence retains at least one epitope cross-reactive with antibodies which are specifically immunoreactive with pp32r 1 , pp32r2, or other pp32 variant as desired. The preparation of selected clones which contain DNA sequences corresponding to all or part of the sequence of pp32rl or pp32r2 may be accomplished by those of ordinary skill in the art using conventional molecular biology techniques along with the information provided in this specification.

It is possible to purify a pp32 variant protein, such as pp32rl, which is cross- reactive with antibodies specific for pp32, from an appropriate tissue/fluid source; however, a cross-reactive pp32 variant, or analog thereof, may also be produced by recombinant methods from a DNA sequence encoding such a protein or polypeptide. Polypeptides corresponding to the recombinant protein of this invention may be obtained by transforming cells with an expression vector containing DNA from a clone selected from an mammalian (preferably human) library as described herein. Suitable expression vector and host cell systems are well known to those of ordinary skill in the art, and are taught, for instance, in Sambrook, et al., 1989. The peptide may be obtained by growing the transformed cells in culture under conditions wherein the cloned DNA is expressed. Of course, the peptide expressed by the clone may be longer or shorter than pp32rl or pp32r2, so long as the peptides are immunologically cross-reactive. Depending on the expression vector chosen, the peptide may be expressed as a fusion protein or a mature protein which is secreted or retained intracellularly, or as an inclusion protein. The desired polypeptides can be recovered from the culture by well-known procedures, such as centrifugation, filtration, extraction, and the like, with or without cell rupture, depending on how the peptide was expressed. The crude aqueous solution or suspension may be enriched for the desired peptide by protein purification techniques well known to those skilled in the art. Preparation of the polypeptides may include biosynthesis of a protein including extraneous sequence which may be removed by post-culture processing.

Using the nucleotide sequences disclosed herein and the polypeptides expressed from them, antibodies can be obtained which have high binding affinity for pp32rl or pp32r2, but much lower affinity for pp32 and/or other variants of pp32. Such antibodies, whether monoclonal or purified polyclonal antibodies can be used to specifically detect pp32rl or pp32r2. Techniques for preparing polypeptides, antibodies and nucleic acid probes for use in diagnostic assays, as well as diagnostic procedures suitable for detection of pp32 are described in U.S. Patent Nos. 5,726,018 and 5,734,022, incorporated herein by reference, and these techniques may be applied to pp32rl or pp32r2 by substitution of the nucleic acid sequences disclosed herein. Similar substitution may be applied to other variants of pp32. pp32rl Promoter Sequence

Multiple consensus sequences for binding active steroid receptors found in genomic sequences upstream from the pp32rl coding region are consistent with hormone regulation of gene expression. The consensus sequences were associated with the both induction and repression of expression by steroid hormones. The combination of both positively and negatively acting elements suggests complex regulation of pp32rl expression.

Possible steroid hormone regulation of pp32rl expression is important in regard to prostate cancer. While about one-half of treated patients initially respond to androgen ablation, subsequent hormone refraction and continued aggressive tumor growth is common (Garnick, M.B., "Prostate Cancer," in Scientific American Medicine. Dale, D.C. and Federman, D.D. Eds., Scientific American Inc., New York, 1995). Many different steroid hormones regulate the growth of prostate cancer cells (Huggins, et al., "Studies on prostate cancer: I. The effect of castration, of estrogen, and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate," Cancer Res., 1 :293, 1941). These findings established a basis for androgen ablation therapy for the treatment of metastatic prostate cancer.

The present invention provides androgen-activated promoters based on the upstream portion of the genomic sequence in Figure 2. The promoter sequence provided by this invention is bounded at its 3' terminus by the translation start codon of a coding sequence and extends upstream (5' direction) to include at least the number of bases or elements necessary to initiate transcription at levels above background. Within the promoter sequence will be found a transcription initiation site (conveniently defined by mapping with nuclease SI), a protein binding domain (consensus sequence) within about 100 bases upstream of the transcription initiation site generally designated the TATA box (a binding site for TATA box binding proteins and RNA polymerase), and various other protein binding domains (consensus sequences) upstream of the TATA box that modulate the basic transcriptional activity of the transcription initiation site and the TATA box. The various other protein binding domains preferably contain recognition sequences shown in Table 1 for binding (1) androgen receptors, estrogen receptors, glucocorticoid receptors, and progesterone receptors; (2) transcription factors containing the leucine zipper motif including, but not limited to Fos, Jun, JunB, and Myc; and, (3) certain tissue specific transcription factors including, but not limited to GAT A- 1 and GATA-2. The various other protein binding domains upstream of the TATA box may contribute to specificity (tissue specific expression), accuracy (proper initiation), and strength (transcription frequency) of the promoter. The promoter elements may serve overlapping functions so that the promoter may function in the absence of subsets of these elements. Therapy

Inhibition of function of protransforming variants of pp32 by any means would be expected to be an avenue of therapy. Phenotypic changes in pp32 by alternate gene usage are a common feature of human prostate cancer, and the change in pp32 phenotype is accompanied by a change in oncogenic potential. Modulation of oncogenic potential by alternate gene usage has interesting implications. Preliminary comparison of structures predicted by energy minimization programs for pp32 and variant pp32 species suggests significant structural differences that might form the basis for interaction with different mechanistic pathways by pp32 and the variant pp32 species. From a temporal standpoint, it is not yet clear how early in the neoplastic process the use of alternate genes of the pp32 family happens. The common occurrence of alternate pp32 gene usage in prostate cancer suggests that it could be an early, important event in tumorigenesis.

Alternate gene usage is potentially reversible, which has additional important clinical and mechanistic implications. The malignant potential of tumor cells may be modified by manipulating the pattern of expression of pp32 family members in a form of endogenous gene therapy. Several possibilities exist. One possibility is that mutation and loss of heterozygosity (LOH) of the pp32 locus at 15q22.3-q23 somehow leads to increased expression of pp32rl and pp32r2. However, LOH is unlikely, since chromosomal loss of 15q22.3-q23 is not a common feature of prostate cancer. More intriguing from both pathogenic and therapeutic standpoints are the possibilities that involve regulatory aberrations. Epigenetic regulation could lead to inactivation of the pp32 gene and concomitant activation of pp32rl and pp32r2 genes by means such as methylation and demethylation. Another mechanism could involve regulation by one or more transcription factors that act differentially to repress pp32 while inducing pp32rl and pp32r2. Finally, post-transcriptional mechanisms could also be invoked, whereby differential expression would be regulated by changes in mRNA or protein stability. From the standpoint of carcinogenesis, all of the latter mechanisms would contribute to tumorigenesis via a plastic, potentially reversible regulatory change rather than an irreversible structural change in the genome. It is therefore possible to envision the eventual use of the pp32 family as targets for pharmacologic chemopreventive and therapeutic strategies in prostate cancer.

U.S. Patent No. 5,726,018, incorporated herein by reference, describes various therapeutic avenues which may be applied by the skilled worker based on the nucleotides and protein sequences disclosed herein. In a particular embodiment, all or a portion of the sequence of pp32rl or pp32r2 may be supplied in the antisense orientation to block expression of the variants found in carcinomas, particularly prostate cancer. Suitable methods for preparation of antisense expression vectors and administration of antisense therapy may be found in U.S. Patent No. 5,756,676, incorporated herein by reference. Prescreening of the patient population using the diagnostic methods described herein to identify patients having tumors expressing the particular pp32 variant is preferred.

Screening for compounds having therapeutic effects in prostate cancer may also be facilitated by the present invention. Studies which may be used to screen candidate compounds are described in U.S. Patent No. 5,756,676, incorporated herein by reference, modified by the use of cell lines which express particular variants of pp32 (see, e.g., Examples below). Compounds which affect steroid dependent protein expression may also be detected according to this invention by similar screening studies using an androgen- activated promoter as provided herein operatively coupled to a DNA sequence whose expression may be detected. (Marker sequences are well known in the art, see, e.g., Sambrook, et al., and selection of an appropriate detectable expression marker is a routine matter for the skilled worker.) Screening by testing the effect of candidate compounds on recombinant cells containing an expression vector having an androgen-activated promoter operatively coupled to an expression marker, with appropriate controls, is within the skill of the art, in view of the promoter sequences provided herein. In one aspect, this invention provides a method for screening candidate compounds for pharmacological activity by (1) culturing a cell transfected with the DNA molecule containing an androgen-activated transcriptional promoter which is operatively linked to an open reading frame comprising at least one exon of a protein coding sequence, and (2) determining expression of the open reading frame in the presence and absence of the compound. In a preferred mode the androgen activated promoter may be the portion of the sequence in Figure 2 which is up- stream of the translation initiation site, or alternatively the androgen activated promoter may be the 2700 bp upstream from the translation initiation site. Similar experiments for screening potential therapeutic compounds may be performed using cell lines which express pp32 variants, such as pp32rl or pp32r2, according to the screening methods described herein and/or in International Patent Publication WO 99/29906, June 17, 1999, incorporated herein by reference.

Diagnostic Methods Based on the pp32 Gene Family

In one aspect, this invention provides methods for detecting and distinguishing among members of the pp32 gene family. As explained herein, the presence of one or more members of the gene family may be detected using assays based on common structures among the members resulting from common or similar sequences. For example, polyclonal antibodies elicited by pp32 will cross-react with pp32rl and pp32r2, including various alleles of these pp32 variants. Similarly, the full coding region of the pp32 cDNA will hybridize under suitable conditions with nucleic acid encoding any of the variants, as shown by the in situ detection of the variants in tumor sections which were subsequently shown to contain either pp32rl or pp32r2 allelic forms (Example 1). Selection of conditions that promote the immune cross-reactivity or cross-hybridization necessary for such detection is within the skill of the art, in view of the examples provided herein. For example, by using large nucleotide probes in hybridization experiments, the effects of one or a few differences in sequence may be overcome, i.e., larger probes will bind to more dissimilar target sequences, in contrast to shorter probes for which each nucleotide makes a larger percentage contribution to the affinity, and a single nucleotide alteration will cause a greater relative reduction in hybridization efficiency. Typically probes of 50 or more nucleotides are used to find homologues to a given sequence, and the studies reported in Example 1 used the entire sequence of pp32 as a probe to find cells expressing homologous members of the gene family other than pp32. Likewise, polyclonal antisera elicited to an antigen having multiple epitopes is more likely to cross-react with a second antigen that has a few of the same epitopes along with many different epitopes, while a monoclonal antibody or even a purified polyclonal antiserum might not bind to the second antigen.

In addition to determining the presence of one or more members of the pp32 gene family, this invention also provides methods for distinguishing among members. Determining which pp32 variant may be useful, for instance, to determine whether a transfomation promoting or suppressing variant is present in a tissue sample. Suitable methods for distinguishing include both immunoassay and nucleic acid binding assays. Preferred are methods which can detect a 10-fold difference in the affinity of the detecting ligand (e.g., antibody or oligonucleotide) for the target analyte. Such methods are well documented for other systems, and may be adopted to distnguish between pp32 variants by routine modification of such methods in view of the guidance provided herein.

Protein level assays may rely on monoclonal or purified polyclonal antibodies of relatively greater affinity for one variant compared to another (see, e.g., Smith, et al. ("Kinetics in interactions between antibodies and haptens," Biochemistry, 14(7): 1496- 1502, 1975, which shows that the major kinetic variable governing antibody-hapten interactions is the rate of dissociation of the complex, and that the strength of antibody-hapten association is determined principally by the activation energy for dissociation), and Pontarotti, et al. ("Monoclonal antibodies to antitumor Vinca alkaloids: thermodynamics and kinetics," Molecular Immunology, 22(3): 277-84, 1985, which describes a set of monoclonal antibodies that bind various dimeric alkaloids and can distinguish among the alkaloid haptens due to different relative affinities of the various monoclonal antibodies for particular dimeric alkaloids), each of which is incorporated herein by reference). Suitable modifications of the conditions for immunoassays to emphasize the relative affinity of monoclonal antibodies with different affinity are also discussed in U.S. Patent No. 5,759,791, incorporated herein by reference.

A number of methods are available which are capable of distinguishing between nucleic acid sequences which differ in sequence by as little as one nucleotide. For example, the ligase chain reaction has been used to detect point mutations in various genes (see, e.g., Abravaya, et al., "Detection of point mutations with a modified ligase chain reaction (Gap-LCR)," Nucleic Acids Research, 23(4):675-82, 1995, or Pfef er, et al., "A ligase chain reaction targeting two adjacent nucleotides allows the differentiation of cowpox virus from other Orthopoxvirus species," Journal of Virological Methods, 49(3):353-60, 1994, each of which is incorporated herein by reference). Amplification of a sequence by PCR also may be used to distinguish sequences by selection of suitable primers, for example, short primers, preferably 10-15 matching nucleotides, where at least one of the primers has on the 3' end a unique base that matches one variant but not other variants, and using annealing conditions under which the primer having the unique base has at least a ten-fold difference in dissociation rate between the fully matching variants and variants which do not fully match. Similar differentiation may be achieved in other methods dependent on hybridization by using short probes (typically under 50bp, preferably 25bp or less more preferably less than 20bp or even 10-12 bp) by adjusting conditions in hybridization reactions to achieve at least a ten-fold difference in dissociation rate for the probes between the fully matching variants and variants which do not fully match. Cleavase fragment length polymorphism may also be used, and a specific example below provides guidance from which the skilled worker will be able to design similar studies by routine selection of other cleavase enzymes in view of the sequences provided herein.

The diagnostic methods of this invention may be used for prognostic purposes and patient differentiation as described herein. In particular, the methods of this invention allow differentiation between products expressed from the various sequences disclosed in Figure 7. Preferred methods are those that detect and/or differentiate between pp32, pp32rl, and/or pp32r2. Situations in which differentiation between pp32 variants will be of benefit will be readily apparent to the skilled clinician, in view of the present disclosure. Selection among the diagnostic methods provided by this invention of a suitable technique to achieve the desired benefit is a routine matter for the skilled clinician.

EXAMPLES In order to facilitate a more complete understanding of the invention, a number of Examples are provided below. However, the scope of the invention is not limited to specific embodiments disclosed in these Examples, which are for purposes of illustration only. Example 1. Cellular Location of pp32 Expression pp32 mRNA can be detected by in situ hybridization with a pp32 probe under stringent conditions.

In situ hybridization. Bases 1 -298 of the pp32 cDNA sequence (GenBank HSU73477) were subcloned into the Bluescript vector by standard techniques. Digoxigenin labeled anti- sense and sense RNA probes were generated using a commercially available kit (Boehringer Mannheim). Vector DNA linearized with BamHl and Xhol served as template for antisense and sense probe generation respectively. In vitro transcription was performed for 2 hours at 37° in a final volume of 20 μl which contained 1 μg of template DNA, 2 U/μl of either T3 or T7 RNA polymerase, 1 U/μl ribonuclease inhibitor, 1 mM each of ATP, CTP, GTP, 0.65 mM UTP, 0.35 mM digoxigenin-1 1-UTP, 40 mM Tris-HCl pH 8.0, 10 mM NaCI, 10 mM DTT, 6 mM MgCl₂ and 2 mM spermidine. The reaction was stopped by adding 2 μl of 0.2M EDTA, pH 8. 0 and the synthesized transcripts were precipitated for 30 min at - 70° with 2.2 μl of 4 M LiCl and 75 μl of pre-chilled ethanol. RNA was pelleted by centrifugation, washed with 80% ethanol, mildly dried and dissolved in 100 μl of DEPC treated water. Yields of labeled probe were determined by an enzyme linked irrimunoassay using a commercially available kit (Boehringer Mannheim). Nonradioactive in situ hybridization was performed with anti-sense and sense pp32 RNA probes generated by in vitro transcription (see U.S. Patent No. 5,726,018, incorporated herein by reference). Figure 1A shows that normal prostatic basal cells are positive, whereas the clear, differentiated glandular cells are negative. In contrast, prostatic adenocarcinoma, shown at left, is strikingly positive. Note that the signal is cytoplasmic since it is mRNA and not the protein that is detected in this assay. pp32 displays a distinctive pattern of expression in vivo (Chen, et al.; Malek, et al., "Identification and preliminary characterization of two related proliferation-associated nuclear phosphoproteins." Journal of Biological Chemistry, 265:13400-13409, 1990; Walensky, et al., "A novel M(r) 32,000 nuclear phosphoprotein is selectively expressed in cells competent for self-renewal," Cancer Research 53:4720-4726, 1993). In normal peripheral tissues, expression is restricted to stem-like cell populations such as crypt epithelial cells in the gut and basal epithelium in the skin; in the adult central nervous system, cerebral cortical neurons and Purkinje cells also express pp32. In normal prostate, basal cells express pp32, whereas pp32 mRNA is not detectable by in situ hybridization in differentiated glandular cells (Figure 1 A). In contrast, strong in situ hybridization to pp32 probes is found in nearly all clinically significant human prostatic adenocarcinomas. 87% of human prostatic adenocarcinomas of Gleason Score 5 and above express mRNA that hybridizes strongly with probes to pp32 in contrast to only 11% of prostate cancers of Gleason Score 4 and below in a study of 55 patients.

Immunohistochemistry. Formalin-fixed, paraffin-embedded tissue was sectioned at 4 μM, deparaffinized, hydrated, processed for heat-induced antigen retrieval at 95 ° in 0.01 M citrate buffer, pH 6.0, for 20 min (Cattoretti, et al., "Antigen unmasking on formalin- fixed, paraffin-embedded tissue sections," Journal of Pathology 171 :83-98, 1993), then incubated overnight at room temperature with a 1/20 dilution of anti-pp32 antibody. Following washing, the slide was sequentially developed with biotinylated swine-anti-rabbit IgG at 1/100 (Dako), strepavidin peroxidase (Dako), and diaminobenzidine. Figure IB shows a representative high-grade human prostate cancer stained with affinity-purified rabbit polyclonal anti-pp32 antibody (Gusev, et al., "pp32 overexpression induces nuclear pleomorphism in rat prostatic carcinoma cells," Cell Proliferation 29:643-653, 1996). The left-hand panel shows a representative field at 250x; the rectangle indicates the area shown in computer generated detail in the right-hand panel. Strongly hybridizing tumors show intense immunopositivity with antibodies to pp32, indicating that they express pp32 or immunologically related proteins (Figure 1 A and IB).

At face value, the localization data shown in Figures 1 A and 1 B would suggest that pp32 is expressed in both normal and neoplastic prostatic epithelial cells, despite its ability to inhibit neoplastic functions such as transformation. The explanation for this apparent discordant expression is that prostate tumors do not generally express pp32, but rather express variant pp32 species that promote transformation, instead of inhibiting it. Example 2. ESTs corresponding to pp32

Several potential variant pp32 species have been identified in the prostate cancer expressed sequence tag libraries of the NCI's Cancer Genome Anatomy Project. Clone 588488 encodes a protein that is 96% identical to APRIL, although absent retrieval and sequencing of the full clone, it is impossible to tell whether the entire EST clone encodes a pp32 related sequence; neither is it possible to assess the biologic function of this molecule at this time. Nevertheless, it is apparent that the sequenced portion encodes a protein bearing great similarity to pp32. This EST does not appear in the database for normal prostate. As with the variant pp32 species recovered from prostate cancer, generation of this molecule by mutation would require a complex mechanism. pp32-related genes are present in other organisms. The existence of a pp32 gene family in rodent would be consistent with the existence of a comparably sized family in human. A murine pp32 (GenBank U73478) has 89% amino acid identity to pp32, but less identity to pp32rl and APRIL. (The murine cerebellar leucine rich acidic nuclear protein has a single amino acid substitution relative to murine pp32.) We additionally identified a murine EST, GenBank AA066733, with closest identity to APRIL protein at 85% identity over 148 amino acids of a predicted open reading frame. Several other murine EST's, AA212094 and W82526, are closely related to the pp32 family but are not significantly more related to either pp32, pp32rl, or APRIL. A human homologue of such a gene would be expected to encode a fourth member of this gene family. We identified EST's predicted to encode pp32-related proteins in C. elegans, schistosomes, zebrafish, and Drosophila (data not shown). However, these sequences may not represent the complete extent of the pp32 gene family in these organisms, and thus are not informative for the likely size of the mammalian pp32 gene family. Example 3. The Structure of a Gene Encoding a Relative of the pp32 Family

Screening a human genomic library in bacteriophages with probes generated from human pp32 cDNA yielded a new sequence that contained an open reading frame encoding a protein homologous with pp32. Screening a Human Genomic Library in Bacteriophages for pp32 cDNA. A genomic library from human placenta in the Lambda Fix II vector was expressed in K. coli strain XL-1 Blue MRA (Stratagene #946206). Screening for bacteriophage clones containing DNA inserts homologous with pp32 cDNA followed routine procedures (Sambrook, et al.). Briefly, nitrocellulose filters that had overlain bacteriophage plaques were hybridized with P-32 labeled probes for pp32 cDNA. The probes were prepared by the random primer method (Stratagene #300385) using pp32 cDNA as a template (Chen, et al., Molec. Biol. Cell, 7:2045-2056,1996.). Reactive bacteriophage plaques were plugged and the bacteriophages were eluted, reexpressed, and rescreened with pp32 cDNA probes until pure. Bacteriophage DNA was prepared by the plate lysate method (Sambrook, et al.).

Identifying Restriction Fragments within Bacteriophage DNA Containing Sequences Homologous with pp 32 cDNA.

DNA from a bacteriophage clone containing pp32 cDNA sequences was digested with Hindlll. Using routine methods, the restriction fragments were separated by agarose gel electrophoresis, transferred in alkaline buffer to positively charged nylon filters, and hybridized with probes that were selective for the 5' and 3' ends of the pp32 cDNA (Sambrook, et al.). The 5' and 3 ' probes were prepared as described above except that the products of polymerase chain reactions (PCR) were used as templates for the labeling reactions (Saiki, et al., Science, 239:487-491, 1988.). One PCR product was a 249 base pair segment of pp32 cDNA containing nucleotides 32 through 279. It was the result of a reaction using a pp32cDNA template and the primers

5 '-TATGCTAGCGGGTTCGGGGTTTATTG-3 ' and 5'-GATTCTAGATGGTAAGTTTGCGATTGAGG-3' (primer set A). The other product was a 263 base pair segment of pp32 cDNA including nucleotides 677 through 938. It was the result of a reaction using a pp32 cDNA template and the primers 5'-GAATCTAGAAGGAGGAGGAAGGTGAAGAG-3 ' and 5'-CTATCTAGATTCAGGGGGCAGGATTAGAG-3' (primer set B). The PCR reactions included 35 cycles of one minute denaturations at 95°C, one minute primer annealings at 50°C, and one minute extensions at 72°C (cycling program A). A 4.7 kb Hindlll restriction fragment that hybridized with the 5 ' probe, but not with the 3 ' probe and a 0.9 kb Hindlll fragment that hybridized with the 3 ' probe, but not with the 5' probe were subcloned into pBluescript (Gibco) by routine methods (Sambrook, et al.). The nucleotide sequences of both strands of purified plasmid DNA containing the inserts were determined by automated procedures (DNA Analysis Facility, Johns Hopkins University School of Medicine).

Completion of Sequencing by Direct Sequencing of PCR Products. Alignment of the sequences of the 4.7 and 0.9 kb Hindlll restriction fragments with pp32 cDNA showed about 90% homologies between the 3 ' end of the 4.7 kb fragment and the 5' region of pp32 cDNA and the 5' end of the 0.9 kb fragment and the 3' region of the pp32 cDNA. There was an unaligned 199 base pair gap of pp32 cDNA sequence between the ends of the restriction fragments. Primers were designed to specifically anneal to relative pp32 sequences on both sides of the sequence gap. The primer sequences were 5 '-GAGGTTTATTGATTGAATTCGGCT-3 ' and 5'-CCCCAGTACACTTTTCCCGTCTCA-3' (primer set C). Polymerase chain reactions followed cycling program A with primer set C and pure bacteriophage DNA as a template. The 943 base pair products were shown by ethidium bromide staining agarose gels, extracted from excised fragments of low melt agarose (NuSieve) electrophoresis gels, and sequenced by automated procedures as described above. A sequence of 5,785 bases was obtained from the human placental genomic library bacteriophage clone containing segments homologous with pp32 cDNA (Figure 2). This sequence was deposited in Genbank under Accession No. U71084, Locus HSU71084. The sequence has an open reading frame extending from nucleotides 4,453 to 5,154. Analysis of the nucleotide sequence upstream of the open reading frame revealed consensus sequences for active steroid hormone receptors at over twenty positions (Table 1).

Sequence analysis of the open reading frame showed 94% sequence homology to pp32 (Figure 3). Alignment of the open reading frame sequence to pp32 cDNA revealed 33 scattered, solitary base differences and clustered differences of two and seven bases. There were two internal deletions of three and nine bases. The open reading frame encoded a polypeptide containing 234 amino acid residues with 88% protein-level homology to pp32 (Figure 4). Alignment of the translated sequence to the pp32 amino acid sequence revealed 18 scattered, solitary amino acid residue differences, three differences in clusters of two residues, and one difference in a clusters of four residues. There were two internal deletions of one and three residues and a terminal deletion of eleven residues. The translated sequence contained 69 acidic residues, 26 fewer than pp32. Example 4. Chromosome Mapping of pp32rl

The pp32rl gene maps to chromosome 4 as determined by PCR of the NIGMS monochromosomal panel 2 (Drwinga, et al., "NIGMS human/rodent somatic cell hybrid mapping panels 1 and 2," Genomics 16:311314, 1993) followed by sequencing of the PCR product. Interestingly, the full sequence of pp32rl including 4364 nucleotides of sequence 5' to the start ATG contained over 400 matches in a blastn search of the non-redundant GenBank database. These matches were to two short regions of about 278 and 252 base pairs (nucleotides 674-952 and 2542-2794) that represent repeats in opposite orientations. The repeats are significantly related to elements on many chromosomes. The human pp32 gene has been mapped to chromosome 15q22.3-q23 by fluorescence in situ hybridization (Fink, et al.). A Unigene entry for pp32 (Hs. 76689; HLA-DR associated protein 1) lists 93 EST sequences corresponding to this gene, 12 of which contain a mapped sequence-tagged site (STS). These STS sites are all reported to map to chromosome 15, as are many of the pp32 EST's analyzed by electronic PCR (http://www.ncbi.nlm.nih.gov). APRIL protein was also mapped to chromosome 15q25 (Mencinger, et al.; GenBank Y07969). Example 5. Sequence Analysis of pp32r2

A pp32-related sequence (designated pp32r2) has been identified on chromosome 12 by methods analogous to those described in Example 2 for isolation of the unique intronless pp32-related gene pp32rl, found on chromosome 4. It was initially thought that the chromosome 12 sequence, encoding a truncated protein, might represent a pseudogene; however that interpretation has been reassessed in view of the present findings. The sequence has been designated pp32r2, and is recorded in Genbank as locus AF008216; the sequence of pp32r2 is shown in Figure 5. By BESTFIT analysis (Genetics Computer Group, Inc., Wisconsin Package, version 9.1, Madison, WI, 1997), pp32r2 is 99.5% identical to FT 1.1 1, FT2.4 and Tl, showing four nucleotide differences over the 875 nucleotide overlap of the sequences; this level of variation is consistent with a polymorphism. Similarly, BESTFIT analysis shows that PP32R1 is 99.6 % identical to FT3.3 and 99.4% identical to FT2.2, displaying four and five nucleotide differences, respectively (see Figure 7 below).

Example 6. Sequence Comparison of Multiple Clones

Screening of a human placental genomic library in Lambda Fix II vector (Stratagene #946206) with P-32 labeled probes for pp32 cDNA yielded a clone of approximately 23 kb. 4.7 kb and 0.9 kb Hindlll restriction fragments of this clone hybridized with probes for pp32 cDNA. The 4.7 kb clone aligned with the 5' portion of the pp32 cDNA sequence, and the 0.9 kb fragment aligned with the 3' end. A small discontinuity of 0.2 kb was sequenced from a bridging PCR product. No introns were identified.

Cultured cells including the whole human embryonic line FSH173WE and the prostatic cancer cell lines PC-3 and LNCaP (American Type Culture Collection) were grown under recommended tissue culture conditions. Poly A+ RNA was prepared by oligo dT adsorption (MicroFasTrack, Invitrogen) and used as a template for the generation of cDN A through reactions with reverse transcriptase and random hexamers (GeneAmp RNA PCR Kit, Perkin Elmer). The cDNA sequences encoding the open reading frame were amplified by nested PCR using primers specifically selective for the genomic sequence over pp32 sequences. The final 298 base pair products were seen by ethidium bromide staining agarose electrophoretic gels.

Using procedures similar to those described in Example 3, except without the need for nested primers in most cases, transcripts from DU- 145 cells and from numerous patients were sequenced for comparison to the transcripts from the above samples. The results are shown in Table 2. A summary of the degree of identity between various transcripts is provided in Table 3. Example 7. Sequence Variation for Individual Isolates of Different Cell Lines and Tumor Tissue

The explanation for the apparent discordant expression of pp32 in cancer is that prostate tumors do not generally express pp32, but rather express variant pp32 species that promote transformation, instead of inhibiting it.

RT-PCR and CFLP. Sequences were reverse-transcribed and amplified using bases 32 to 52 of HSU73477 as a forward primer and 9 19 to 938 of the same sequence as a reverse primer in conjuriction with the Titan One-Tube RT-PCR kit (Boehringer). Reverse transcription was carried out at 50° for 45 min followed by incubation at 94° for 2 min; the subsequent PCR utilized 45 cycles of 92° for 45, 55° for 45 sec, and 68° for 1 min with a final extension at 68° for 10 min in a PTC 100 thermocycler (MJ Research). Template RNA was isolated from cell lines or frozen tumor samples using RNAzol B (Tel-Test) according to the manufacturer's instructions, then digested with RNAse-free DNAse I (Boehringer). pCMV32 was used as a positive control without reverse transcription. The cleavage assay was performed according to the manufacturer's specifications (Life Technologies) with digestion at 55° for 10 min at 0.2 mM MnCl₂ and electrophoresed on a 6% denaturing polyacrylamide sequencing gel.

At the level of RTPCR, paired normal prostate and prostatic adenocarcinoma from three patients yielded amplification products (Figure 6A) ranging from 889 to 909 bp. The reaction employed consensus primers capable of ampliring the full-length coding sequence from pp32 and the two closely-related intronless genomic sequences pp32rl and pp32r2. The sole difference noted was a diminished amplicon yield from normal tissue as compared to neoplastic. Four human prostatic adenocarcinoma cell lines, DU-145, LNCaP, PC-3, and TSUPR-1, also yielded similar products. Figure 6A shows RT-PCR amplified DNA from human prostate and prostate cancer cell lines. Lane a is an undigested control whose band migrated substantially slower than the digestion produces; samples in all other lanes were digested with cleavage as described. The lanes show: 1 kb ladder (Life Technologies), A; pCMV32, B; DU-145, C; LNCaP, D; PC-3, E; TSUPr-1 , F; a representative sample, FT-1 , without reverse transcription, G; FN- 1 , H; FT-1 , 1; FN-2, J; FT-2, K; FN-3, L; FT-3, M; negative control with template omitted. FN indicates frozen benign prostate and the number indicates the patient; FT indicates frozen prostatic adenocarcinoma and the number indicates the patient. Numbers on the left-hand side of the figure indicate the size in kb of a reference 1 kb DNA ladder (Life Technologies). Qualitative differences between normal and neoplastic tissue began to emerge when the RT-PCR products were subcloned and analyzed by cleavage fragment length polymorphism analysis (CFLP) and sequence analysis. Figure 6B shows a cleavase fragment length polymorphism analysis of cloned cDNA amplified by RT-PCR from human prostatic adenocarcinoma, adjacent normal prostate, and human prostatic adenocarcinoma cell lines using primers derived from the normal pp32 cDNA sequence. The lanes show individual RT-PCR-derived clones from the DU-145, LNCaP, PC-3 and TSUPrl cell lines, from frozen prostate cancer (FT), and from frozen normal prostate (FN): a, undigested normal pp32 cDNA; b, normal pp32 cDNA; c, DU-145-1 ; d, DU-145-3; e, DU-145-5; f, LNCaP-3; g, PC3-3; h, PC3-8; i, TSUPrl. -I; j, TSUPrl-3; k, TSUPrl-6; 1, FT1.1 1 ; m, FT1.7; n, FT2.2; o, FT2.4; p, FT3.18; q, FT3.3; r, FN3.17; s, FN2.1. LNCaP expresses normal pp32. The band shifts correspond to sequence differences. All clones of RT-PCR product from normal prostate tissue displayed a normal CFLP pattern that corresponded precisely to that obtained from cloned pp32 cDNA template (GenBank HSU73477, Figure 6B). Prostatic adenocarcinomas yielded four distinct CFLP patterns upon similar analysis, of which three were unique and one mimicked the normal pp32 pattern. Examination of DU- 145, PC-3, and TSUPR- 1 cell lines yielded substantially similar results whereas LnCaP yielded only a normal pp32 CFLP pattern. Further analysis at the sequence level confirmed that normal prostate and LnCaP contained solely normal pp32 transcripts.

Transcripts obtained from prostatic adenocarcinomas and from most cell lines represented closely-related variant species of pp32, summarized in Table I. These transcripts varied from 92.4% to 95.9% nucleotide identity to normal pp32 cDNA (Genetics Computer Group, Inc., Wisconsin Package, version 9.1, Madison, WI, 1997). Of the sixteen variant transcripts obtained, fifteen had open reading frames encoding proteins ranging from 89.3% to 99.6% identity to normal pp32. The table summarizes data obtained for variant pp32 transcripts obtained from human prostatic adenocarcinoma and prostate cancer cell lines. Sequences falling into closely related groups are indicated by the group letters (A,B,C); U indicates unassigned sequences not clearly falling into a group. The origin of each sequence is: FT, frozen tumor followed by patient number, decimal point, and clone number; D, DU-145 followed by clone number (as are all cell line sequences); P, PC3; and T, TSUPrl. Nucleotide identity, gaps in the nucleotide sequence aligninent, and protein identity were determined from BESTFIT alignments with the normal pp32 cDNA and protein sequences. The effect on transformation is described as: stimulates, more foci obtained when transfected with ras+myc than with ras+myc+vector control; inactive, equivalent foci obtained as with ras + myc + vector control; and suppresses, fewer foci obtained as with ras + myc + vector control.

The predicted protein sequences fell into three discrete groups: [1] truncated sequences spanning the N-terminal 131 amino acids of pp32, of which one such sequence substantially equivalent to pp32r2 was obtained identically from two of three patients and from the TSUPR- 1 cell line; [2] sequences more closely homologous to a distinct pp32- related gene, pp32rl than to pp32, and [3] heterogeneous pp32-related sequences. Tumors from two of the three patients analyzed contained no detectable normal pp32 transcripts. Two of twelve cloned transcripts from the third patient tumor were normal by CFLP pattern, with sequence confirmation of normality on one clone. Two clones from cell lines were normal by CFLP screening, but were later shown to represent variant sequences. In contrast to benign prostate, which yielded only pp32 transcripts by sequence analysis, adjacent prostate cancers expressed little or no pp32. Instead, four pp32-related transcripts with distinct sequences encoding open reading frames were obtained from the adjacent prostate cancers, varying from 92.4% to 95.9% nucleotide identity to normal pp32 cDNA. Of these, two transcripts corresponding to the products of the previously identified pp32-related genes, pp32r 1 and pp32r2, were obtained multiple times from patient samples and therefore became the focus of this study.

Figures 7A and 7B show a multiple pairwise alignment of nucleotide and predicted protein sequences for all transcripts (Smith, et al., "Identification of common molecular subsequences," J. Mol. Biol, 147: 195-197 1981). The figures were compiled with the GCG Pileup and Pretty programs (Smith, et al.). Differences from the consensus sequences are shown as indicated; agreement with the consensus sequence is shown as a blank. Normal human pp32 is designated hpp32. Sequences from the TSUPrl, PC3, and DU-145 cell lines are as indicated. The designation FT indicates sequence derived from a frozen human prostatic adenocarcinoma. Only the normal pp32 sequence, hpp32, was obtained from normal prostate adjacent to turnor tissue. Figure 8A shows alignment of the amplicon nucleotide sequences with pp32 and the predicted amplicon from pp32rl; Figure 8B shows alignment of the predicted protein sequences. One sequence (FT 1.1 1), independently obtained three times from two separate patients and the TSUPR- 1 cell line, is shown only once in the diagram. The pileup and pairwise alignments illustrate several important points: [ 1 ] there is a high degree of sequence conservation at both the nucleotide and predicted amino acid levels; [2] the sequence differences are distributed throughout the length of the sequence without obvious hotspots; [3] there is no obvious clustering or segmentation of sequence differences; and [4] the variant sequences fall into the previously described groups. These points are detailed in Figures 8 A and 8B. The identical pp32r 1 sequences obtained from two patients differed by 4 nucleotides from the predicted product of the pp32rl genomic sequence. The pp32r2 sequences obtained from two patients were also identical and differed, by three nucleotides from the pp32r2 genomic sequence. Both sets of differences are considered consistent with polymorphic variation. pρ32, pp32r 1 , and pp32r2 show sequence changes along their entire length, as shown in the multiple pairwise alignment of the predicted protein sequences in Figure 7. No straightforward process of somatic mutation or alternate splicing could explain these results. Instead, given the correspondence of the variant sequences with previously identified genes on chromosomes 4 and 12, the data are consistent with alternate gene expression. Example 8. Diagnostic method to distinguish among family members

The three members of the pp32 family which are expressed in human prostate cancer are pp32, pp32rl and pp32r2. Whereas pp32 suppresses in vitro transformation and in vivo tumorigenesis in model systems, pp32rl and pp32r2 are pro-transforming and are tumorigenic in the same systems. It is possible to determine which of the three members is expressed in a tissue sample by using a protocol similar to that described in Example 7. Analysis from freshly frozen human tissue and cell lines.

Total RNA is extracted from freshly frozen human tissues or human cancer cell lines and subjected to reverse transcription and polymerase chain reaction amplification with single set of primers capable of amplifying the entire coding region of the cDNA of all the three genes. A suitable set of primers is:

Upper: 5^*GGGTTCGGGGTTTATTG3' - This corresponds to bp32 to bp48 of the pp32 cDNA sequence (Genbank U73477) Lower: 5'CTCTAATCCTGCCCCCTGAA3' - This corresponds to bp919 to bp938 of the pp32 cDNA sequence (Genbank U73477) The observed amplicon sizes with this primer set are pp32 - 907bp, pp32rl - 889bp and pp32r2 - 900bp. The three cDNAs are distinguished from each other by restriction enzyme digestion with the following enzymes - EcoR I, Hind III and Xho I. The resultant digest is run on a 2.5% agarose gel to positively identify the three different cDNAs. The table below lists the sizes of the bands observed The bolded numbers indicate the band sizes useful for identification of the three cDNAs.

Table 4A Expected band sizes upon restriction digestion of the RT-PCR product from fresh tissue and cell lines

Analysis from formalin fixed and paraffin embedded tissue. A similar approach is followed for identification of pp32, pp32rl and pp32r2 transcripts from formalin fixed and paraffin embedded tissues. Total RNA is extracted and subjected to reverse transcription and PCR amplification with a single set of primers capable of amplifying a stretch of 200bp from all the three cDNAs. A suitable set of primers is:

Upper primer - from bp394 to bp414 of the pp32 cDNA sequence (Genbank U73477)

Lower primer - from bp609 to bp629 of the pp32 cDNA sequence (Genbank U73477)

The three cDNAs are distinguished from each other by restriction enzyme digestion with the following enzymes - Hind III, Xho 1 and BseR I. The resultant digest is run on a 3% agarose gel to positively identify the three different cDNAs. The table below lists the sizes of the bands observed. The bolded numbers indicate the band sizes useful for identification of the three cDNAs.

Table 5A Expected band sizes upon restriction digestion of the RT-PCR product from formalin fixed and paraffin embedded tissues

Example 9. pp32rl Augments Oncogene-Mediated Transformation of Rat Embryo Fibroblasts. pp32rl was subcloned into a eukaryotic expression vector under the CMV promoter and analyzed for its effect on ras + myc-mediated formation of transformed foci in rat embryo fibroblasts. Genomic sequences including the entire coding region for pp32rl were amplified by PCR and subcloned into the eukaryotic TA cloning and expression vector pCR3.1 vector (Invitrogen) which contains a CMV promoter. The assay was performed as described (Chen et al. Mol Biol Cell. 7:2045-56, 1996) with each T75 flask receiving 5 micrograms of pEJ-ras, and/or 10 micrograms of pMLV-c-myc, pCMV32, pp32rl in PCR3.1, or PCR 3.1 alone. After 14 days, transformed colonies were enumerated. Figure 8 shows the results. The data represent the average of seven replicates from two separate experiments in duplicate and one in triplicate. The error bars indicate standard error of the mean. In contrast to pp32, which consistently suppresses focus formation induced by ras + myc and other oncogene pairs, pp32rl caused a statistically significant stimulation of focus formation with p=.004 by an unpaired t-test. Example 10. Effect of Transcripts from Various Cell Lines on Rat Fibroblast

Transformation Assays

Expression constructs prepared as described above from PC-3 and DU-145 cells were tested in the rat embryo fibroblast transformation assay described by Chen, et al., Mol

Biol Cell, 7:2045-56, 1996, incorporated herein by reference. The results are shown in Figure 9. Transcripts from the two cell lines stimulated ras+myc induction of transformed rat embryo fibroblast foci, in contrast to normal pp32, which suppressed transformation. The figure shows the mean +/- the standard deviation, except for DU- 145, which represents a single determination. Example 11. Transformation Activity of Various Isolates from Patient Tumors The variant transcripts isolated from prostate cancer patients differ significantly from pp32 in sequence. The isolated transcripts were found to stimulate transformation. Transformation assay. Rat embryo fibroblasts were transfected with the indicated constructs as described (Chen, et al.) and transformed foci enumerated. For each experiment, approximately 1 x 10⁶ cells were plated per T75 flask and incubated for 2 to 3 d prior to transfection to achieve approximately 40% confluency. For each flask of primary rat embryo fibroblasts, the plasmids indicated in each experiment were added in the following amounts: pEJ-ras, 5 μg; and pMLV-c-myc, pCMV32, pCMVneo, or variant pp32 constructs in pCR3.1 (Invitrogen), 10 μg. Plasmids were prepared in two volumes Lipofectin (2 μ l lipofectin per μg DNA) then gently mixed by inversion in 1.5 ml OPTIMEM in sterile 15 ml polystyrene tubes and allowed to incubate at room temperature for > 15 min. For experiments with more than one flask, mixtures of all reagents were increased in proportion to the numbers of flasks required for each transfection. Cells were washed once with OPTIMEM (Gibco-BRL), and then fed with 6 ml of OPTIMEM and 1.5 ml of the DNA/Lipofectin mix. After overnight incubation, the cells were grown in standard media and refed with fresh media twice weekly. Foci were counted fourteen days post-transfection. Figure 10 summarizes four separate experiments. Each data point represents the results from an individual flask expressed as the percent foci obtained in the contemporaneous control of ras+myc+vector.

Figure 10 shows that expressed variant transcripts from prostate cancer cell lines and from human prostatic adenocarcinoma generally produce increased numbers of transformed foci when co-transfected with ras and myc, as compared to the number of foci obtained when ras and myc are transfected with blank vector. Variant pp32 transcripts from DU-145 (D3), and from three prostate cancers (FT 1.7, FT 2.2, and FT3.18) yield increased numbers of transformed foci over those produced by ras and myc alone with blank vector. This stands in marked contrast to normal pp32, which consistently suppresses transformation. These activities are also summarized in Table I. Example 12. Effect of pp32 Variants on Tumorigenesis In Vivo

Experiments testing the effect of transfection of NIH3T3 cells on tumorigenesis in vivo are consistent with in vitro results in rat embryo fibroblasts. NIH3T3 cells were stably transfected by lipofection with the pp32 species indicated in Table 6A carried in the pCR3.1 -Uni CMV-driven mammalian expression vector (Invitrogen). The G418-resistant clones employed in these experiments were all shown by genomic PCR to carry the indicated pp32 species. For analysis of tumorigenesis, 5 x 10⁶ cells in 100 microliters of unsupplemented Dulbecco's modified Eagle's medium without phenol red were injected into the flanks of female athymic nude mice on an outbred background of greater than six weeks in age (Harlan). For logistical reasons, inoculations of the various groups were staggered over a seven day period. Each group of mice was euthanized precisely seven weeks after inoculation. Where a mouse had a tumor, the tumor was dissected, measured, and weighed, and Table 6A reports the average weight of tumors in mice injected with cells carrying various vectors. One tumor from each group was examined histologically. All tumors were fibrosarcomas without noteworthy inflammation present. Data obtained with N1H3T3 cells indicate that NIH3T3 cells stably transfected with the variant pp32 species P3, P8, FT1.7, FT2.2, and FT2.4 form tumors when inoculated into nude mice. In contrast, NIH3T3 cells stably transfected to express human pp32 fail to form tumors in vivo even when further transfected with ras. Lines ofNIH3T3 cells were also established that were stably transfected with expression constructs encoding pp32 or pp32-antisense. Basal expression of pp32 is essential for maintenance of contact inhibition and serum-dependent cell growth; antisense ablation of endogenous pp32 synthesis permitted cells to grow normally following serum withdrawal. Constitutive over-expression of pp32 potently suppressed ras-mediated transformation of NIH3 T3 cells in vitro and tumorigenesis in vivo. In contrast, antisense ablation of endogenous pp32 dramatically increased the number and size of ras-transformed foci; in vivo, tumors obtained from ras-transformed antisense pp32 cells were approximately 50-fold greater in mass than tumors obtained from ras-transformed control cells. A switch in the oncogenic potential of the expressed pp32 family members accompanies the switch in pp32 phenotype in prostate cancer. Example 1 1 shows an experiment demonstrating that expressed pp32rl and pp32r2 frequently fail to inhibit or, indeed, sometimes stimulate transformed focus formation when co-transfected with ras and myc, as compared to the number of foci obtained when ras and myc are transfected with blank vector. This stands in marked contrast to normal pp32, which consistently suppresses transformation. Similarly, Example 12 shows that both pp32r 1 and pp32r2 are tumorigenic when stably transfected into NIH3T3 cells, in contrast to pp32, which is non-tumorigenic.

For purposes of clarity of understanding, the foregoing invention has been described in some detail by way of illustration and example in conjunction with specific embodiments, although other aspects, advantages and modifications will be apparent to those skilled in the art to which the invention pertains. The foregoing description and examples are intended to illustrate, but not limit the scope of the invention. Modifications of the above-described modes for carrying out the invention that are apparent to persons of skill in medicine, immunology, hybridoma technology, pharmacology, pathology, and/or related fields are intended to be within the scope of the invention, which is limited only by the appended claims.

All publications and patent applications mentioned in this specification are indicative of the level of skill of those skilled in the art to which this invention pertains. All publications and patent applications are herein incorporated by reference to the same extent as if each individual publication or patent application was specifically and individually indicated to be incorporated by reference.

TABLE 1

Position Strand Consensus Sequence Factor

H C TTTCCT PEA3

21 N CAAGGTCA ELP

23 N AGGTCA PPAR

32 C CCCTAA TBF1

41 N CTTGGC NF-1 (-like proteins)

81 N TAAACAC Pit- 1

82 N AAACACA HiNF-A

1 13 C CTTCCC c-Ets-2

1 18 N CTATCA GATA-1

122 N CAGTTG c-Myc

212 C AATAAATA TFIID

213 N ATAAATA ETF

247 N TATCTA NIT2

261 C AAGGAA c-Ets-2

262 N AGGAAA PEA3

283 C l l l l lC l l l^'lTC Hb

320 C TTATAT GAL4

333 N TAAAAAA TBP

349 N TTATACATT TBP

363 C AAGGAA c-Ets-2

394 C TTTCTATA TBP

398 N TATAAA TBP

398 N TATAAA TFIID

41 1 C CTGAATT Pit-1

420 N TGTCCC GR

423 C CCCTAA TBF1

434 N TTCCTT c-Ets-2

447 C CTTCCC c-Ets-2

514 N TTATCTCT GATA-1

514 C TTATCT GATA-2

515 N TATCTC NΓΠ

537 N TATGCA EFII

553 N AAGTCA GCN4

608 N TGACTA GCN4

628 N CCTCCCAAC LyF-1

640 N TGTCCT GR

648 N TTAAAATTCA 1-Oct

648 N TTAAAATTCA 4-Oct Table 1 - Continued

649 N ^"TAAAAT F2F

649 N TAAAAT Pit- 1

661 N TAAAAAA TBP

673 N CTTGGC NF-l (-like proteins)

"25 N AGGCGG Spl

729 N GGGCGG ETF

"29 N GGGCGG Spl

"29 C GGGCGG Spl

741 N AGGTCA PPAR

793 N TATAAATA B factor

793 N TATAAA TBP

793 N TATAAATA TFIID

793 N TATAAAT TMF

794 N ATAAATA ETF

809 N TTATCT GATA-1

809 C TTATCT GATA-2

815 N GGGTGTGG TEF-2

826 C CACATG muEBP-C2

826 C CACATG TFE3-S

826 N CACATG USF

978 N ATGTAAAACA 1-Oct

978 N ATGTAAAACA 2-Oct

978 N ATGTAAAACA NF-IL-2A

1000 N ATGTCAGA CSBP-l

1006 N GATTTC H4TF-1

1034 C I l l l CAT Pit-1

1047 N AAGATAAAACC RVF

1048 C AGATAA GATA-1

1048 N AGATAA GATA-2

1049 N GATAAA TFΠD

1083 C GCCAAG NF-l (-like proteins)

1 124 N CGCCAT UCRF-L

1 163 C GACCTG TGT3

1307 N CAGTCA GCN4

1347 C TGCATA EFII

1373 C AGAACA AR

1373 N AGAACAT GR

1373 N AGAACA GR

13^*3 C AGAACA GR

13 ^"3 N AGAACA PR

1 "3 C AGAACA PR Table 1 - Continued

13^*3 N AGAACA PR A

13^*3 C AGAACA PR A

1393 C TCACTT IRF- 1

1393 C TCACTT IRF-2

1395 C ACTTCCT E1A-F

1423 N TTATCT GATA-1

1423 C TTATCT GATA-2

1424 N TATCTA NIT2

1452 N TTACTC GCN4

1471 N TGGGTCA c-Fos

1471 N TGGGTCA c-Jun

1471 N TGGGTCA ER

1496 N TCTCTTA c-Myc

151 1 N TATAAA TBP

151 1 N TATAAA TFIID

1549 C TTTGAA TFΠD

1568 C AATGTATAA TBP

1581 C TTTGAA TFIID

1590 C AGATAA GATA-1

1590 N AGATAA GATA-2

1591 C GATAATTG Dfd

165" C AGGACA GR

1670 C AΓΠTA F2F

1670 C A i l l lA Pit-1

1671 C 1 1 1 1 ATA B factor

16" ! C 1 1 1 1 ATA Drl

1671 C ITITATA En

1671 C 1 i 1 1ATA TBP

1671 C 1 1 1 1ATA TBP-1

1671 C ΓΠTATA TFIIA

1671 C i 1 i'lATA TFIIB

1671 C TTTTATA TFIID

1671 C TTTTATA TFIIE

1671 C 1 i 1 1 ATA TFIIF

1671 C I 1 1 1 ATA TRF

1672 C TTTATA TBP

1694 C AATAAATA TFIID

1695 N ATAAATA ETF

1733 N AGGAAA PEA3

P49 C TTATAT GAL4

1783 N TAACTCA AP-1 Table 1 - Continued

1 829 C TAGATA NIT:

1857 N CGCCAT UCRF-L

1875 N TTCTGGGAA IL-6 RE-BP

1895 N TGACTA GCN4

1899 N TATTTAA TBP

1942 N ATATAA GAJL4

1985 C TTTATA TBP

1985 C TTTATA TFIID

2010 C AATAAATA TFIID

201 1 N ATAAATA ETF

2058 C TGCATA EFII

2095 N CAGTCA GCN4

2146 C AAGGAA c-Ets-2

2147 N AGGAAA PEA3

2190 N AGGAAA PEA3

2220 C GGCACA GR

22^ς2 N CCAATAG gammaCAAT

2286 N TGTGCC GR

2292 N ATGGGA PTFl-beta

2314 N TATGCA EFII

2328 C GGCACA GR

2350 C ATGATAAG GATA-1

2351 N TGATAAG GATA-1

2363 N GGGAAG c-Ets-2

2367 N AGCCACT CP2

2369 C CCACTGGGGA AP-2

2404 N TAAAAT F2F

2404 N TAAAAT F2F

2404 N TAAAAT Pit-1

2409 N TTGTCATA 77+82K protein

2409 N TTGTCATA VETF

2415 N TATCTA NΓΠ

2451 C TTTATC TFIID

2452 N TTATCT GATA-l

2452 C TTATCT GATA-2

2486 N CTCTCTCTCTCTC GAG

2644 N AGGCGG Sp l

2658 N ACAGCTG GT-IIBalpha

2658 N ACAGCTG GT-IIBbeta

2709 C GGCCAGGC AP-2

2"23 N TGAACT GR Table 1 - Continued

2"3 1 C TGACCT PPAR

2731 C TGACCTCA URTF

2753 N CTTGGC NF- l (-like proteins

2818 C TGATGTCA AP-1

2818 C TGATGTCA c-Fos

2818 C TGATGTCA c-Jun

2818 C TGATGTCA CREB

2845 N GGGAAG c-Ets-2

2858 N AGATAG GATA-1

2858 C AGATAG GATA-1

2864 C AGTTCA GR

2899 N ATATAA GAL4

2900 N TATAAAA B factor

2900 N TATAAAA Drl

2900 N TATAAAA En

2900 N TATAAAA TBP

2900 N TATAAA TBP

2900 N TATAAAA TBP- 1

2900 N TATAAAA TFIIA

2900 N TATAAAA TFΠB

2900 N TATAAAA TFΠD

2900 N TATAAAA TFIIE

2900 N TATAAAA TFIIF

2900 N TATAAAA TRF

2921 C TTTGAA TFIID

2924 C GAAATC H4TF-1

2930 C CATTAG Isl-1

2948 C TGTACA GR

2948 C TGTACA PR

2948 C TGTACA PR A

2964 C ATTTGAGAA VITF

3030 N AGTGTTCT GR

3032 N TGTTCT AR

3032 N TGTTCT GR

3032 C TGTTCT GR

3032 N TGTTCT PR

3032 C TGTTCT PR

3032 N TGTTCT PR A

3032 C TGTTCT PR A

3104 C GGATTATT Til

3106 C ATTATTAA AFP1 Table 1 • Continued

3111 N TAAAAT F2F

3111 N TAAAAT Pit-1

3125 C C ATTTTA F2F

3125 C C ATTTTA Pit-1

3142 N N TGTGAT GR

3169 N GTTTTATT HOXDIO

3169 N GTTTTATT HOXD8

3169 N GTTTTATT HOXD9

3175 C TTTGAA TFIID

3185 N TTGCTCA Zta

3206 N GATTTC H4TF-1

3212 N AGGAAA PEA3

3238 C C ATTTTA F2F

3238 CC ATTTTA Pit-1

3256 CC TTTGAA TFIID

3266 N TTGCTCA Zta

3320 C ATTTTA F2F

3320 C ATTTTA Pit-1

3358 N ATGGGA PTFl-beta

3360 C GGGACA GR 3440 C CACTCA GCN4 3460 C TTTCCT PEA3

3483 N GACACA GR 3491 C TTTCCT PEA3 3495 N CTAATG Isl-1 3523 C AGAACA AR

3523 N N AGAACA GR

3523 C C AGAACACT GR

3523 C C AGAACA GR

3523 N AGAACA PR

3523 C AGAACA PR

3523 N AGAACA PR A

3523 C C AGAACA PR A

3538 CC TTTATC TFIID

3539 N TTATCT GATA-1

3539 C TTATCT GATA-2

3551 N TGAGTG GCN4

3569 C TCCCAT PTFl-beta

3594 N TTAGGG TBF1

3653 C CCTGCTGAA LyF- 1

3668 N CTCATGA 1 -Oct Table 1 - Continued

3668 N CTCATGA 2-Oct

3668 N CTCATGA Oct-2B

3668 N CTCATGA Oα-2B

3668 N CTCATGA Oct-2C

3679 C TGTGTAA Zta

3685 C AGAACT GR

3712 C TTTCCT PEA3

3713 N TTCCTT c-Ets-2 3717 N TTGCTCA Zta

3727 C AAAACATAAAT ssARS-T

3749 N TAAAAAA TBP

3784 C CACTCA GCN4

3791 C ATTTTA F2F

3791 C ATTTTA Pit-1

3815 N TATCTA N T2

3829 C TAG ATA NIT2

3859 C AGAACA AR

3859 N AGAACAG GR

3859 N AGAACA GR

3859 C AGAACA GR

3859 N AGAACA PR

3859 C AGAACA PR

3859 N AGAACA PR A

3859 C AGAACA PR A

3860 N GAACAG LVa 3877 C ATCACA GR 3886 N TGAGTCA AP-1 3886 C TGAGTCA AP-1 3886 C TGAGTCA c-Fos 3886 C TGAGTCA c-Jun 3886 C TGAGTCA Fral

3886 C TGAGTCA NF-E2

3887 C GAGTCA GCN4 3931 N AGATAG GATA-1 3931 C AGATAG GATA-1 3960 N TTGGCA NF-l/L 3965 C ATTTTA F2F 3965 C ATTTTA Pit-1 4026 N TATTTAA TBP 403" N TGTGAT GR 4040 N GATGCAT Pit-1 Table 1 - Continued

4042 C TGCATA EFII

4079 N TTCAAAG SRY

4079 N TTCAAAG TCF-1A

4079 N TTCAAA TFIID

4097 N CAGGTC TGT3

4140 N TGATTCA AP-1

4140 C TGATTCA AP-1

4140 N TGATTC GCN4

4164 N GGGAGTG p300

4205 C AGATAA GATA-1

4205 N AGATAA GATA-2

4219 C TTAGTCAC AP-1

4219 C TTAGTCA AP-1

4219 C TTAGTCAC c-Fos

4219 C TTAGTCAC ⁾ c-Jun

4219 C TTAGTCA c-Jun

4219 C TTAGTCA Jun-D

4220 C TAGTCA GCN4

4271 N TGTTCT AR

4271 N TGTTCT GR

4271 C TGTTCT GR

4271 N TGTTCT PR

4271 C TGTTCT PR

4271 N TGTTCT PR A

4271 C TGTTCT PR A

4280 C TGACCCA c-Fos

4280 C TGACCCA c-Jun

4280 C TGACCCA ER

4292 C CTTATCAG GATA-1

4292 C CTTATCA GATA-1

4361 N TTCAAAG SRY

4361 N TTCAAAG TCF-1A

4361 N TTCAAA TFIID c c o o o o c a vo vo vc o c o e

W

CO

— -I- fN

— _. <-^ ( — - N N

"i C H OT '- H t- H '-i '-i OT O H OT . J _Mi k r- U. U. . a l-. α. ^ f- a ^β. few fr- few . few at Table 2 - Continued

91 105 106 120 121 1 IS 136 ISO 1S1 165 166 1UO

1.3 TIILNLSUNKIKDLST IEPLKKLENI.KSLDI. FNCEVTNLNDYRENV FKLLPQLTYI.IXSYDR DDKEΛPDSDAEGYVE GLDDEEEDEDEEEYI) 1HO l'P»2 Tlll.NLSGNKIKOLST IEPLKKLENI.KSLDI. FNCEVTNLNDYRENV FKLI.PQLTYLDGYDR DUKEAPDSDΛEGYVE GLDDEEEDEDEEEYD lβϋ θ IIILNLSGNKIKDLST IEPLKKLENLKSLDL SNCEVTNLNDYRENV FKLI.PQLTYLDGYDR DDKEΛPDSDAEGYVE GLDDEEEDEOEEEYD 1Θ0 lθl 19S 196 210 211 225 226 240 241

TSU6

Dϊ

PG

FT 1.11

TSUI

ET1. lβ o FT2 4

FT2 2 EDAQWEDEEQBEBB BBGEEEDVSOGDEED EEGYNDGEVDGEEDE EELGEEERQQKRK-

KG EDAQWEDEEGBEEB BBGBBEDVSGGDBED EBGYNDαEVDQBEDB EELGEBERQQKRK- - -

FT 1.7 EDAQWEDEEOBEOE BBGBBEDVSOGDBED EEGYNDGEVDDEEDE EELGEEERGQKRKRE PBDEGEDDD

P) EDAQWEDBEDEDEE BEGBBEDVSGEBEED EEGYNDGEVDDEEDE EELGEEERGQKRKRE PEDEGEDDD I EDAQVVEDEEDEDEE EEGBBEDVSGEBBED EEGYNDGBVDDEEDE EELGEEERGQKRKRE PEDEGEDDD pp!2 EDAQWEDEEDEDEE EEGEEEDVSGBEEBD EEGYNDGEVDDEBDE EELGEEERGQKRKRE PEDEGEDDD β EDAQWEDEEDBDBB EEGEBEDVSGBEEBD EEGYNDGEVDDBEDB EELGEEBRGQKRKRE PEDEGEDDD

TSII6 and TSUI from TSU cell line; D3 from DU-145 cell line; P3 and Pθ from PC 3 cell line; FT1 , FT2 and F 3 f i om patient carcinoma; LE from LNCAP; KG from placenta

TABLE 3

Comparison to pp32 Sequences % Identity % Similarity

CLONE cDNA Protein Protein

D3, DU-145 cells 95 90 95 P3, PC-3 86 94 96 P8, PC-3 98 97 97

FT1.1 1 97 86 92

FT1.7 95 95 95

FT2.2 94 85 88

FT2.4 99 86 92

FT3.18 99 90 94

TABLE I A

t

l ahlc 2Λ

I'erccnl a inn acid identity of pp32 and related proteins. Sequences were aligned using the (JAP program (7). The number ol gaps in llic alignment is indicated as well as the Z score, a statistical measure of protein relaledness derived from 50 comparisons of randomi/cd protein scψicnccs.

Table 3 . pp32 Homologs

human pp32 (Genbank Locus HSU73477) murine pp32 (Genbank Locus MMU73478) human cerebellar leucine rich acidic nuclear protein (LANP) (Genbank Locus AF025684) murine LANP (Genbank Locus AF022957) murine RFC1 (Genbank Locus MUSMRFC, Accession NO. L23755)

<-Λ I l PP2a or human potent heat-stable protein phospatase 2a inhibitor (Genbank Locus HSU60823) E SSP29 (Genbank Locus HSU70439)

HLA-DR associated protein I (Genbank Locus HSPPHAPI, Accession No. X75090) PHAPI2a (EMBL Locus HSPHAPI2A, Genbank Accession No. Y07569) PHAPI2b (EMBL Locus HSPHAPI2B, Genbank Accession No. Y07570) April (EMBL Locus HSAPRIL)

Table 6A. Tumorigeniciry in Nude Mice of NIH3T3 Cells Transfected with pp32 and pp 2 Variants

pp Species Clone Tumors/ Average Tnmnr Weight

FTI .7 1 3/3 14.9 + 2.1

2' 3/3 13.3 + 3.7

FT2 2 1 3/3 10.5 + 2.8

2 3/3 3.8 + 2.1

FT2.4 1 3/3* 1.3 + 0.9

2 3/3 13.8 ± 3.3

D3 5² 0/3 6² 0/3.

P3 I I 3/3 5 5.77 ± + 00..55

14' 3/3 22..11 ±± 11.22

P8 l⁴ 3/3 6 6..44 ± + 55..33

2 3/3 111..33 ++ 33..99 4^s 3/3 100..11 ± + 44..88

L (pp32) 5⁵ 0/3 6* 0/3

Vector Control 2^J 0/3

3' 0/3

FT I J. clone 2 and Vector Control, clone 3 were tested on contralateral sides of a single group of animals.

~D3 clone 5 was tested on the contralateral sides of a group of animals simultaneously injected with N1H3T3 cells transfected with a clone of pp32rl (data not shown). D3 clone 6 was tested on the contralateral sides of a group of animals simultaneously injected with a second clone of NIH3T3 celts transfected with pp32rl (data not show).

P3, clone 14 and Vector Control, clone 2 were tested on contralateral sides of a single group of animals.

P8, clone I and pp32, clone 6 were tested on contralateral sides of a single group of animals.

⁵P8. clone 4 and pp32. clone 5 were tested on contralateral sides of a single group of animals.

One tumor in this group, weighing 0.5 gm, was detected only upon post mortem dissection.

Claims

CLAIMS:

1. A method of treating malignant cells, said cells exhibiting subnormal expression of normal pp32 or overexpression of a pp32 variant, said method comprising restoring pp32 function to said cells.

2. The method for treating malignant cells of claim 1, said cells exhibiting subnormal expression of normal pp32 or overexpression of a pp32 variant, said method comprising inducing increased expression of normal pp32 in said cells.

3. The method of claim 2, wherein increased expression of normal pp32 is induced by transfection with DNA encoding normal pp32.

4. The method of claim 2, wherein increased expression of normal pp32 is induced by administering an inducer of normal pp32 expression.

5. A method of screening to determine whether a compound is an inducer of pp32 expression comprising measuring pp32 expression by cells cultured in the presence and absence of said compound, wherein increased expression of pp32 in the presence of said compound indicates that said compound is an inducer.

6. The method of treating malignant cells according to claim 1 , said cells exhibiting subnormal expression of normal pp32 or overexpression of a pp32 variant, said method comprising inhibiting a protein phosphatase in said cells, wherein said phosphatase is inhibited by incubation with pp32.

7. The method of treating malignant cells according to claim 6, said cells exhibiting subnormal expression of normal pp32 or overexpression of a pp32 variant, wherein said phosphatase is a nuclear protein phosphatase.

8. The method of treating malignant cells according to claim 7, said cells exhibiting subnormal expression of normal pp32 or overexpression of a pp32 variant, said method comprising inhibiting a nuclear protein phosphatase in said cells, wherein the phosphatase is a member of the protein phosphatase 2A group.

9. A method of screening to determine whether a compound is an inducer of pp32 function comprising measuring protein phosphatase activity in cells cultured in the presence and absence of said compound, wherein expression of pp32 in said cells inhibits protein phosphatase activity, inhibition of protein phosphatase activity by said compound being an indication that said compound induces pp32 function in said cells.

10. The method of screening to determine whether a compound is an inducer of pp32 function according to claim 9, comprising measuring nuclear protein phosphatase activity in cells cultured in the presence and absence of said compound, wherein expression of pp32 in said cells inhibits nuclear protein phosphatase activity, inhibition of nuclear protein phosphatase activity by said compound being an indication that said compound induces pp32 function in said cells.

1 1. The method of screening to determine whether a compound is an inducer of pp32 function according to claim 10, comprising measuring nuclear protein phosphatase activity in cells cultured in the presence and absence of said compound, wherein expression of pp32 in said cells inhibits nuclear protein phosphatase activity, inhibition of nuclear protein phosphatase activity by said compound being an indication that said compound induces pp32 function in said cells, wherein said nuclear protein phosphatase is a member of the protein phosphatase 2A group.