DETAILED DESCRIPTION OF THE INVENTION

[0043] The present invention is based upon the initial discovery of an alternative HER-2 mRNA of 4.8 kb with a 274 bp insert identified as intron 8. The retained intron is in-frame and encodes 79 amino acids [SEQ ID NO. 1] followed by a stop codon at nucleotide 236. The alternative mRNA predicts a truncated HER-2 protein that lacks the transmembrane and intracellular domains and contains 419 amino acids [SEQ ID NO. 2]; 340 residues that are identical to the N-terminus of p185HER-2 and 79 unique residues at the C-terminus [SEQ ID NO. 1]. Using specific antibodies against either the novel 79 amino acid residue C-terminal sequence [SEQ ID NO. 1] or the N-terminus of p185HER-2, a 68 kDa protein product was identified [SEQ ID NO.2]. This 68 kDa protein is the product of an alternative HER-2 transcript, and is found in cell extracts and in extracellular media from several cell lines. Expression of the alternative transcript was highest in a nontransfected human embryonic kidney cell line.

[0044] The results presented here show expression of alternative HER-2 mRNA, which contains an additional 274 nucleotides, probably intron 8. Consistent with this finding, an alternative transcript of ˜4.8 kb was detected in human fetal kidney tissue and in the human embryonic kidney cell line, HEK 293. Moreover, a transcript of 2.6 kb, which is the size expected if the sequence is retained in the 2.3 kb truncated HER-2 mRNA (Yamamoto et al., Nature 319:230-234, 1986; and Scott et al., Mol. Cell. Biol. 13:2247-2257, 1993), was detected in human fetal liver tissue by Northern blot analysis using a probe specific for the inserted sequence or for the HER-2 ECD coding sequence ( FIG. 2 ). The inserted sequence introduces a termination codon and predicts a novel 79 amino acid extension designated ECDIIIa at residue 340 of the p185HER-2 protein. The predicted protein therefore lacks the transmembrane and intracellular domains, but contains subdomains I and II of the extracellular domain of p185HER-2. As predicted, a secreted protein that contains N-terminal sequence of p185HER-2 and the C-terminal extension provided by the inclusion of the novel sequence was detected ( FIGS. 3 and 5 ). The ECDIIIa protein was found to be 68 kDa which is the approximate size expected of the protein encoded by the alternative transcript if the five N-linked glycosylation sites found in subdomains I and II of p185HER-2 are glycosylated (Stern et al., Mol. Cell. Biol. 6:1729-1740, 1986).

[0045] The data presented herein demonstrate that p68HER-2 specifically binds to p185HER-2. The association with p185HER-2 may be conferred by the novel proline rich ECDIIIa domain rather than the N-terminal subdomains I and II of p68HER-2. While the HER-2 ECD, generated by in vitro deletion mutagenesis, also contains subdomains I and II, it does not associate with the extracellular domain of p185HER-2 unless engineered to enhance their proximity (Tzahar et al., EMBO J. 16:4938-4950, 1997; O'Rourke et al., Proc. Natl. Acad. Sci. USA 94:3250-3255, 1997; and Fitzpatrick et al., FEBS Letters 431:102-106, 1998). However, the unique ECDIIIa peptide binds with high affinity (nM concentrations) to p185HER-2 and to transfected 17-3-1 cells that overexpress p185HER-2 ( FIG. 5 ). Preferential binding of the ECDIIIa domain peptide to 17-3-1 cells indicates that secreted p68HER-2 interacts with the extracellular region of p185HER-2 at the cell surface. Therefore, p68HER-2 and fragments thereof appear to be a naturally occurring HER-2 binding protein, encoded by the HER-2 gene. In contrast to EGFR family ligands (Groenen et al., Growth Factors 11:235-257, 1994), p68HER-2 lacks an EGF homology domain and contains the first 340 amino acids of the receptor itself, p185HER.

[0046] Previously described putative HER-2 ligands were found to associate indirectly with p185HER-2 only in a heterodimer with an EGFR family member (Heldin and Ostman, Cytokine Growth Factor Rev. 7:33-40, 1996). Although it is possible that ECDIIIa binds indirectly to p185HER-2 through a coreceptor, this seems unlikely since detergent solubilized p185HER-2 was specifically and efficiently “pulled down” by immobilized ECDIIIa peptide ( FIG. 5B ).

[0047] For all naturally occurring or engineered ligands for mammalian EGFR family members, binding is tightly coupled to stimulation of receptor dimerization and tyrosine phosphorylation (Hynes and Stern, Biochim. et Biophys. Acta 1198:165-184, 1994; Dougall et al., Oncogene 9:2109-2123, 1994; and Groenen et al., Growth Factors 11:235-257, 1994). Although they bind, neither p68HER-2 nor the ECDIIIa peptide was found to activate p185HER-2. Activation was assessed in two different cell lines that differ in the extent of p185HER-2 tyrosine phosphorylation, transfected 17-3-1 cells as well as SKOV-3 ovarian carcinoma cells. Furthermore in vitro self-phosphorylation activity, which is enhanced in dimeric forms of p185HER-2 (Dougall et al., Oncogene 9:2109-2123, 1994; and Lin et al., J. Cell. Biochem. 49, 290-295, 1992), was not stimulated by p68HER-2 or ECDIIIa. Similarly, the Argos protein, which is an extracellular inhibitor of the Drosophila EGF receptor and the only known antagonist of class I RTKs, did not simulate tyrosine phosphorylation of the receptor (Schweitzer et al., Nature 376:699-702, 1995). Likewise, Angiopoietin-2, a natural antagonist for the Tie 2 RTK, bound the endothelial receptor but failed to activate it (Maisonpierre et al., Science 277:55-60, 1997).

[0048] Without being bound by theory, since p68HER-2 occupies but does not activate, it could block dimerization of p185HR-2. By analogy, HER-2 ECD, when engineered to enhance its binding to RTKs, prevented the formation of productive dimers required for transphosphorylation and receptor activation thereby having a dominant negative effect (O'Rourke et al., Proc. Natl. Acad. Sci. USA 94:3250-3255, 1997). In contrast to the HER-2 ECD, soluble p68HER-2 exhibited strong binding to p185HER-2, yet also contains subdomain I and II of the ECD. Since subdomain I may be the low affinity, promiscuous ligand binding site required for recruitment of p185HER-2 into heteromeric complexes (Tzahar et al., EMBO J. 16:4938-4950, 1997), p68HER-2 could block this site and thereby obstruct recruitment of p185HER-2 into dimers. Alternatively, p68HER-2 could compete with an uncharacterized ligand for binding to p185HER-2. The tissue-specific expression of p68HER-2 in human fetal liver and kidney may function to modulate the extent to which p185HER-2 is occupied during development of these organs. Moreover, the overexpression of p185HER-2, relative to p68HER-2 in tumor cells with HER-2 gene amplification ( FIG. 3 ), could occur though a selective pressure based on overcoming the effects of a binding protein such as p68HER-2. Therefore, p68HER-2 is the first example of a naturally occurring p185HER-2 binding protein that may prevent activation of p185HER-2.

[0049] Pharmaceutical Composition

[0050] The present invention further provides a pharmaceutical composition for treating solid tumors that overexpress HER-2, comprising an agent selected from the group consisting of (a) an isolated polypeptide having from about 50 to 79 amino acids taken from the sequence of SEQ ID NO. 1, wherein the polypeptide binds to the extracellular domain ECD of HER-2 at an affinity of at least 10 8 , (b) an isolated and glycosylated polypeptide having from about 300 to 419 amino acids taken from the sequence of SEQ ID NO.2, wherein the C terminal 79 amino acids are present, and wherein at least three N-linked glycosylation sites are present, (c) a monoclonal antibody that binds to the ECD of HER-2, and (d) combinations thereof, with the proviso that the agent cannot be the monoclonal antibody alone, and pharmaceutically acceptable carrier. Preferably, the agent is the isolated polypeptide having from about 50 to 79 amino acids taken from the sequence of SEQ ID NO. 1. Most preferably, the agent is a combination of the isolated polypeptide having from about 50 to 79 amino acids taken from the sequence of SEQ ID NO. 1 and the monoclonal antibody that binds to the ECD of HER-2.

[0051] The inventive pharmaceutical composition, comprising either or both of the inventive polypeptides and/or monoclonal antibody, can be administered to a patient either by itself (complex or combination) or in pharmaceutical compositions where it is mixed with suitable carriers and excipients. Inventive polypeptide can be administered parenterally, such as by intravenous injection or infusion, intraperitoneal injection, subcutaneous injection, or intramuscular injection. Inventive polypeptide can be administered orally or rectally through appropriate formulation with carriers and excipients to form tablets, pills, capsules, liquids, gels, syrups, slurries, suspensions and the like. Inventive polypeptide can be administered topically, such as by skin patch, to achieve consistent systemic levels of active agent. Inventive polypeptide is formulated into topical creams, skin or mucosal patch, liquids or gels suitable to topical application to skin or mucosal membrane surfaces. Inventive polypeptide can be administered by inhaler to the respiratory tract for local or systemic treatment of cancers characterized by overexpressing HER-2.

[0052] The dosage of inventive polypeptide suitable for use with the present invention can be determined by those skilled in the art from this disclosure. Inventive polypeptide will contain an effective dosage (depending upon the route of administration and pharmacokinetics of the active agent) of inventive polypeptide and suitable pharmaceutical carriers and excipients, which are suitable for the particular route of administration of the formulation (i.e., oral, parenteral, topical or by inhalation). The active inventive polypeptide is mixed into the pharmaceutical formulation by means of mixing, dissolving, granulating, dragee-making, emulsifying, encapsulating, entrapping or lyophilizing processes. The pharmaceutical formulations for parenteral administration include aqueous solutions of the inventive polypeptide in water-soluble form. Additionally, suspensions of the inventive polypeptide may be prepared as oily injection suspensions. Suitable lipophilic solvents or vehicles include fatty oils such as sesame oil, or synthetic fatty acid esters, such as ethyl oleate or triglycerides, or liposomes. Aqueous injection suspensions may contain substances which increase the viscosity of the suspension, such as sodium carboxymethyl cellulose, sorbitol, or dextran. The suspension may optionally contain stabilizers or agents to increase the solubility of the complex or combination to allow for more concentrated solutions.

[0053] Pharmaceutical formulations for oral administration can be obtained by combining the active compound with solid excipients, such as sugars (e.g., lactose, sucrose, mannitol or sorbitol), cellulose preparations (e.g., starch, methyl cellulose, hydroxypropylmethyl cellulose, and sodium carboxymethyl cellulose), gelaten, gums, or polyvinylpyrrolidone. In addition, a desintegrating agent may be added, and a stabilizer may be added.

[0054] Processes for Synthesizing p68 and 79 aa C Terminal Region

[0055] Polypeptide synthesis is done by a group of standard procedures for polypeptide synthesis by sequential amino acids building through peptide synthesis equipment, following manufacturer's instructions for synthesizing peptides. Preferably, shorter polypeptides, of less than 100 amino acids, are best suited for the method of synthesis through sequential amino acid building of polypeptides. In addition, heterologous polypeptides can be expressed by transformed cells using standard recombinant DNA techniques to transform either prokaryotic or eukaryotic cells, provide appropriate growth media for their expression, and then purify the inventive polypeptide either from the media or from intracellular contents depending upon the type of cell used and its expression characteristics.

[0056] Methods for Treating Cancer with p68, 79 aa C Terminal Region, and Combinations

[0057] The present invention provides a method for treating a solid tumor characterized by overexpression of HER-2, or HER-2 variants (see Example 8) comprising administering an agent that binds to the extracellular domain (ECD) of HER-2, wherein the agent is selected from the group consisting of (a) an isolated polypeptide having from about 50 to 79 amino acids taken from the sequence of SEQ ID NO. 1, wherein the polypeptide binds to the extracellular domain ECD of HER-2 at an affinity of at least 10 8 , (b) an isolated and glycosylated polypeptide having from about 300 to 419 amino acids taken from the sequence of SEQ ID NO. 2, wherein the C terminal 79 amino acids are present, and wherein at least three N-linked glycosylation sites are present, (c) a monoclonal antibody that binds to the ECD of HER-2, and (d) combinations thereof, with the proviso that the agent cannot be the monoclonal antibody alone. Preferably, the solid tumor that overexpresses HER-2 is selected from the group consisting of breast cancer, small cell lung carcinoma, ovarian cancer, prostate cancer, gastric carcinoma, cervical cancer, esophageal carcinoma, and colon cancer. Preferably, the agent is the isolated polypeptide having from about 50 to 79 amino acids taken from the sequence of SEQ ID NO. 1. Most preferably, the agent is a combination of the isolated polypeptide having from about 50 to 79 amino acids taken from the sequence of SEQ ID NO. 1 and the monoclonal antibody that binds to the ECD of HER-2.

[0058] The p68HER-2 polypeptide described herein was found to bind to HER-2 and prevent signal transduction through the kinase domain. Without being bound by theory, the unique ECDIIIa domain mediates specific binding to p185HER-2 and the resulting interaction with p68ECDIIIa prevents p185HER-2 dimerization and subsequent signal transduction. Therefore, p68HER-2 functions as a HER-2 antagonist to prevent signal transduction by preventing dimerization as a necessary prerequisite for signal transduction. Thus, the mechanism of p68HER-2 as a HER-2 antagonist is different from the mechanism of binding agents, such as the 79 amino acid polypeptide described herein or a monoclonal antibody that binds to the EDC of HER-2. The inventive method provides that p68HER-2 inhibits tumor cell growth in tumors that overexpress HER-2 by providing a selective pressure for such tumor cells. Similarly, the HER-2 antagonists that are binding agents also inhibit tumor cell growth in tumors that overexpress HER-2 by providing selective pressure to such cells to prevent ligand binding to the ECD of HER-2 and prevent signal transduction even before potential dimerization.

[0059] Use of 79 aa C Terminal Region as a Targeting Molecule

[0060] The present invention further provides a method for targeting a therapeutic agent to solid tumor tissue, wherein the solid tumor tissue is characterized by overexpression of HER-2, comprising attaching the therapeutic agent to an isolated polypeptide having from about 50 to 79 amino acids taken from the sequence of SEQ ID NO. 1, wherein the polypeptide binds to the extracellular domain ECD of HER-2 at an affinity of at least 10 8 . Preferably, the isolated polypeptide is from about 69 to 79 amino acids in length. Preferably, the isolated polypeptide binds to a site on the ECD of HER-2 that is different from the site of binding of Herceptin® (a marketed humanized monoclonal antibody that is used for the treatment of cancer and that binds to the ECD or HER-2). It was discovered that the 79 amino acid polypeptide [SEQ ID NO. 1] exhibited surprising high affinity binding properties to the ECD of HER-2. Moreover, the site of such binding is different and unaffected by the site of binding of a marketed humanized monoclonal antibody (Herceptin®). Therefore, the high binding affinity enables the 79 amino acid polypeptide to function as a targeting molecule to tumor cells expressing HER-2.

[0061] Anti-p68 Antibody as a Diagnostic/Prognostic Agent

[0062] The p68HER-2 ECDIIIa variant 3 (see TABLE 1, below) glycosylated polypeptide was expressed and used as an antigen for antibody production. Specifically, antibody specific for p68HER-2 was prepared by injecting rabbits with purified polyhistidine-tagged ECDIIIa variant 3 peptide, which is the same as the intron encoded novel C-terminus or p68HER-2, the domain that binds with high affinity to p185HER-2. The isolated polyclonal antibody detected pM quantities of ECDIIIa peptide or of p68HER-2 with high specificity (see FIGS. 3 and 5 ). Thus, an antibody specific for p68HER-2 is useful as a diagnostic agent for detecting p68HER-2 in bodily fluids and tumor tissues using diagnostic techniques, such as ELISA, immunoprecipitations, immunohistochemistry or Western analysis.

[0063] Antibodies that specifically recognize one or more epitopes of ECDIIIa, or epitopes of p68HER-2, or peptide fragments, and thus distinguish among ECDIIIa variants (see TABLE 1, below) are also encompassed by the invention. Such antibodies include but are not limited to polyclonal antibodies, monoclonal antibodies (mAbs), humanized or chimeric antibodies, single-chain antibodies, Fab fragments, F(ab′) ₂ fragments, fragments produced by a Fab expression library, anti-idiotypic (anti-Id) antibodies, and epitope-binding fragments of any of the above. The antibodies of the invention may be used, for example, in the detection of a particular p68HER-2 ECDIIIa variant in a biological sample and may, therefore, be utilized as part of a diagnostic or prognostic technique whereby patients or tissue samples may be tested for the presence of particular variants, or for abnormal amounts particular variants.

[0064] Such antibodies may also be utilized in conjunction with, for example, compound screening schemes for the evaluation of the effect of test compounds on expression and/or activity of particular p69HER-2 variants. Additionally, such antibodies can be used in conjunction with the cancer treatment methods described herein.

[0065] For the production of antibodies, various host animals may be immunized by injection with e.g., polyhistidine-tagged ECDIIIa variant polypeptides, truncated ECDIIIa variant polypeptides, functional equivalents of the ECDIIIa variants or mutants of the ECDIIIa region. Such host animals may include but are not limited to rabbits, mice, hamsters and rats, to name but a few. Various adjuvants may be used to increase the immunological response, depending on the host species, including but not limited to Freund's (complete and incomplete), mineral gels such as aluminum hydroxide, surface active substances such as lysolecithin, pluronic polyols, polyanions, peptides, oil emulsions, keyhole limpet hemocyanin, dinitrophenol, and potentially useful human adjuvants such as BCG (bacille Calmette-Guerin) and Corynebacterium parvum.

[0066] Polyclonal antibodies are heterogeneous populations of antibody molecules derived from the sera of the immunized animals. Monoclonal antibodies, which are homogeneous populations of antibodies to a particular antigen, may be obtained by any technique that provides for the production of antibody molecules by continuous cell lines in culture. These include, but are not limited to, the hybridoma technique of Kohler and Milstein, (Nature 256:495-497, 1975; and U.S. Pat. No. 4,376,110), the human B-cell hybridoma technique (Kosbor et al., Immunology Today 4:72, 1983; Cole et al., Proc. Natl. Acad. Sci. USA 80:2026-2030, 1983), and the EBV-hybridoma technique (Cole et al., Monoclonal Antibodies And Cancer Therapy, Alan R. Liss, Inc., pp. 77-96, 1985). Such antibodies may be of any immunoglobulin class including IgG, IgM, IgE, IgA, IgD and any subclass thereof. Hybridomas producing mAb may be cultivated in vitro or in vivo. Production of high titers of mAbs in vivo makes this the presently preferred method of production.

[0067] Additionally, techniques developed for the production of “chimeric antibodies” (Morrison et al., Proc. Natl. Acad. Sci. USA, 81:6851-6855, 1984; Neuberger et al., Nature, 312:604-608, 1984; Takeda et al., Nature, 314: 452-454, 1985) by splicing the genes from a mouse antibody molecule of appropriate antigen specificity together with genes from a human antibody molecule of appropriate biological activity can be used. A chimeric antibody is a molecule in which different portions are derived from different animal species, such as those having a variable region derived from a murine mAb and a human immunoglobulin constant region (humanized).

[0068] Alternatively, techniques described for the production of single-chain antibodies (U.S. Pat. No. 4,946,778; Bird, Science 242:423-426, 1988; Huston et al., Proc. Natl. Acad. Sci. USA 85:5879-5883, 1988; and Ward et al., Nature 334:544-546, 1989) can be adapted to produce single-chain antibodies against ECDIIIa variant gene products. Single-chain antibodies are formed by linking the heavy and light chain fragments of the Fv region via an amino acid bridge, resulting in a single chain polypeptide.

[0069] Antibody fragments that recognize specific epitopes may be generated by known techniques. For example, such fragments include but are not limited to: the F(ab′) ₂ fragments which can be produced by pepsin digestion of the antibody molecule and the Fab fragments which can be generated by reducing the disulfide bridges of the F(ab′) ₂ fragments. Alternatively, Fab expression libraries may be constructed (Huse et al., Science, 246:1275-1281, 1989) to allow rapid and easy identification of monoclonal Fab fragments with the desired specificity.

[0070] Antibodies to particular ECDIIIa variants can, in turn, be utilized to generate anti-idiotype antibodies that “mimic” the ECDIIIa variant, using techniques well known to those skilled in the art. (Greenspan & Bona, FASEB J 7 (5):437-444, 1993; and Nissinoff, J. Immunol. 147:2429-2438, 1991). For example antibodies which bind to an ECDIIIa variant and competitively inhibit the binding of p68HER-2 to HER-2 receptor can be used to generate anti-idiotypes that “mimic” the ECDIIIa variant and, therefore, bind and neutralize HER-2 receptor. Such neutralizing anti-idiotypes or Fab fragments of such anti-idiotypes can be used in cancer therapeutic regimens.

[0071] Alternatively, antibodies to particular ECDIIIa variants that can act as agonists or antagonists of the ECDIIIa variant activity can be generated. Such antibodies will bind to the ECDIIIa variant and modulate the activity of p68HER-2 vis-à-vis p185HER-2 receptor-mediated signal transduction. Such antibodies may be particularly useful for treating particular cancers and/or modulating tumor differentiation. Accordingly, the present invention further provides a method for determining the prognosis of tumor treatment for a tumor that overexpresses HER-2, comprising: (a) obtaining a bodily fluid, wherein the bodily fluid is selected from the group consisting of blood, serum, urine, lymph, saliva, tumor tissue, and combinations thereof; and (b) measuring the amount of p68HER-2 expressed using an anti-p68HER-2 antibody-based assay, wherein the assay is selected from the group consisting of ELISA, immunoprecipitation, immunohistocytochemistry, and Western analysis. Preferably, the method for determining the prognosis of tumor treatment further comprises measuring the amount of p185HER-2 ECD in the bodily fluid, and determining a ratio between the amount of p68HER-2 and p185HER-2. The higher the ratio of p68HER-2:p185HER-2, the better the treatment prognosis.

[0072] ECDIIIa Region Variants as Diagnostic/Prognostic Agents

[0073] Example 11 (below) shows that the human sequence of intron 8 is both proline-rich and polymorphic. Sequencing of genomic DNA from fifteen different individuals resulted in the identification of 10 variable sequence regions within Her-2 Intron 8. See SEQ ID NO: 10; FIG. 8 , and Table 1. FIG. 8 shows the most common nucleotide and corresponding polypeptide sequences of intron 8. This region contains 10 different polymorphisms (marked by the letters W (2×), Y (3×), R, N, M, and S (2×) in SEQ ID NO: 10; or marked by an “X” in FIG. 8 ) that result in nonconservative amino acid substitutions (see legend to TABLE 1). For example, the polymorphism (G→C) at nucleotide position 161 ( FIG. 8 ; TABLE 1) would result in a substitution of Arginine (R) for Proline (P) at amino acid residue #54 of SEQ ID NO: 1, or residue #394 of SEQ ID NO: 2. The N-terminal Glycine (G), designated as position 1 in FIG. 8 or SEQ ID NO: 10, corresponds to amino acid residue 341 in the “herstatin” sequence (Doherty et al., Proc. Natl. Acad. Sci. USA 96:10,869-10,874, 1999). The nucleotide sequence shown in FIG. 1 (A) (Doherty et al., Proc. Natl. Acad. Sci. USA 96:10,869-10,874, 1999), is a polymorphic form that differs at amino acid residues #6 and #73 from the most commonly detected sequence shown here in FIG. 8 .

[0074] This result demonstrates that in the human population there are several variations in the intron-8 encoded domain that could lead to altered biochemical and biological properties among ECDIIIa-containing protein variants. An individual may, inter alia, be genetically heterozygous for two variants, homozygous for a given variant, or homozygous for a double variant. Both tumor progression and optimal treatment may vary depending upon the particular variants represented in a given individual.

[0075] This variability has both prognostic and diagnostic utility. The present invention shows that ECDIIIa-containing polypeptides can bind tightly to, and thus antagonize the HER-2 receptor. Such a specific, high-affinity interaction is dependent upon particular primary, secondary and tertiary structure of the ECDIIIa-containing polypeptide. The ECDIIIa region is proline-rich, and it is well known in the art that nonconservative substitution of proline residues, or other residues within a proline-rich sequence, in a given protein can have profound effects on its secondary and tertiary structure. Thus, the polymorphisms of the present invention are likely to embody significant structural, biochemical and biological differences relative to the most common polypeptide structure (shown in FIG. 8 ). Structural differences among ECDIIIa variant proteins may include for example, differences in size, electronegativity, or antigenicity. Differences in biological properties among ECDIIIa variants might be seen e.g., in the relative degree of cellular secretion, the nature and/or extent of modulation of the HER-2 receptor, pharmacokinetics (e.g., serum half-life, elimination profile), resistance to proteolysis, N-linked glycosylation patterns, etc. These biological differences, in turn, would be expected to alter tumor progression and thus optimal treatment protocols. Thus, the knowledge that an individual contains a particular ECDIIIa variant or variants (e.g., in individuals heterozygous for a given variant, or individuals with compound variants like variant 11 of Table 1), may, in itself, be prognostic of particular cancer susceptibility.

[0076] The apparent genetic heterogeneity of ECDIIIa region means that the nature of the particular ECDIIIa variation carried by an individual may have to be ascertained using sequence identity assays prior to attempting genetic diagnosis of the patient. The analysis can be carried out on any genomic DNA derived from bodily fluids of the patient, typically a blood sample from an adult or child, but alternatively may be serum, urine, lymph, saliva, tumor tissue, placental tissue, umbilical cord tissue, amniotic fluid, and chorionic villi samples. It is expected that standard genetic diagnostic methods, such as hybridization or amplification assays, can be used. Either DNA or RNA, may, for example, be used in hybridization or amplification assays of biological samples to detect particular ECDIIIa variant sequences. Such sequence identity assays may include, but are not limited to, Southern or Northern analyses, single-stranded conformational polymorphism analysis, in situ hybridization assays, and polymerase chain reaction (“PCR”) analyses. Such analyses may reveal both quantitative and qualitative aspects of ECDIIIa variant sequence expression. Such aspects may include, for example, point mutations, and/or activation or inactivation of gene expression. Standard in situ hybridization techniques may be used to provide information regarding which cells within a given tissue express a particular ECDIIIa variant sequence.

[0077] Preferably, diagnostic methods for the detection of ECDIIIa variant nucleic acid molecules involve contacting and incubating nucleic acids, derived from cell types or tissues being analyzed, with one or more labeled nucleic acid reagents, or probes, specific for particular ECDIIIa variants. More preferably, PCR, or reverse transcription PCR, can be utilized to identify nucleotide variation within the ECDIIIa domain. PCR reaction conditions should be chosen which optimize amplified product yield and specificity, and, additionally, produce amplified products of lengths that may be resolved utilizing standard gel electrophoresis techniques. Such reaction conditions are well known to those of skill in the art, and important reaction parameters include, for example, length and nucleotide sequence of oligonucleotide primers, and annealing and elongation step temperatures and reaction times. Following the PCR reaction, the PCR products can be analyzed by methods such as heteroduplex detection, cleavage of RNA-DNA hybrids using Rnase A, single-stranded conformational polymorphisms, and denaturing gradient gel electrophoresis.

[0078] Additionally, if the particular ECDIIIA sequence variant is known to add or remove a restriction site, or to have significantly altered the size of a particular restriction fragment, a protocol based upon restriction fragment length polymorphism (“RFLP”) analysis may be appropriate.

[0079] ECDIIIa variants can also be analyzed at the expression level using sequence identity assays with bodily fluids derived from the patient, typically a blood sample from an adult or child, but may include serum, urine, lymph, saliva, tumor tissue, placental or umbilical cord cells, amniotic fluid, and chorionic villi samples. Well-known sequence identity assays for analyzing expression include, but are not limited to, mRNA-based methods, such as Northern blots and in situ hybridization (using a nucleic acid probe derived from the relevant cDNA), and quantitative PCR (as described by St-Jacques et al., Endocrinology 134:2645-2657, 1994).

[0080] Polypeptide-based methods (e.g., including but not limited to western blot analysis) including the use of antibodies specific for the ECDIIIa variant of interest, as discussed above, could also be used. These techniques permit quantitation of the amount of expression of a given ECDIIIa variant, at least relative to positive and negative controls. Preferably, a battery of monoclonal antibodies, specific for different ECDIIIa eptitopes or variants, could be used for rapidly screening cells or tissue samples to detect those expressing particular ECDIIIa variants, or for quantifying the level of ECDIIIa variant polypeptides. Preferred diagnostic methods for the quantitative or qualitative detection of ECDIIIa variant peptide molecules may involve, for example, immunoassays wherein particular ECDIIIa-containing peptides are detected by their interaction with anti-ECDIIIa variant specific antibodies. This can be accomplished for example, by immunofluorescence techniques employing a fluorescently labeled antibody coupled with light microscopic, flow cytometric, or fluorometric detection. The antibodies (or fragments thereof) useful in the present invention may, additionally, be employed histologically, as in immunofluorescence or immunoelectron microscopy, for in situ detection of ECDIIIa-containing peptides. Through the use of such procedures, it is possible to determine not only the presence of particular ECDIIIa-containing polypeptides, but also their distribution in the examined tissue.

[0081] Immunoassays for ECDIIIa variant polypeptides preferably comprise incubating a biological sample, such as the above-named bodily fluids, which have been incubated in the presence of a detectably labeled antibody capable of identifying ECDIIIa-containing peptides, and detecting bound antibody by any of a number of techniques well known in the art. The biological sample may be brought in contact with and immobilized onto a solid phase support or carrier such as nitrocellulose, or other solid support that is capable of immobilizing soluble proteins, cells, or cell particles. The support may then be washed with suitable buffers followed by treatment with the detectably labeled anti-ECDIIIa variant specific antibody. The solid phase support may then be washed with the buffer a second time to remove unbound antibody. The amount of bound label on the solid support may then be detected by conventional means.

[0082] Alternatively, anti-ECDIIIa variant specific antibodies can be detectably labeled by linking the same to an enzyme for use in an enzyme immunoassay or Enzyme Linked Immunosorbent Assay (“ELISA”). The enzyme which is bound to the antibody will react with an appropriate substrate, preferably, a chromogenic substrate, in such a manner as to produce a chemical moiety which can be detected, for example, by spectrophotometric, fluorimetirc or by visual means. Enzymes which can be used to detectably label the antibody include, but are not limited to, malate dehydrogenase, staphylococcal nuclease, delta-5-steroid isomerase, yeast alcohol dehydrogenase, alpha-glycerophosphate, dehydrogenase, triose phosphate isomerase, horseradish peroxidase, alkaline phosphatase, asparaginase, glucose oxidase, beta-galactosidase, ribonuclease, urease, catalase, glucose-6-phosphate dehydrogenase, glucoamylase and acetylcholinesterase.

[0083] The detection can be accomplished by calorimetric methods that employ a chromogenic substrate for the enzyme. Detection may also be accomplished visually by comparison of the extent of enzymatic reaction with appropriate standards. Detection may also be accomplished using any of a variety of other immunoassays. For example, by radioactively labeling the antibodies or antibody fragments, it is possible to detect ECDIIIa-containing peptides through the use of a radioimmunoassay (RIA). The radioactive isotope can be detected by such means as the use of a gamma counter or a scintillation counter or by autoradiography.

[0084] It is also possible to label the antibody with a fluorescent compound. When the fluorescently labeled antibody is exposed to light of the proper wavelength, its presence can be detected due to fluorescence. Among the most commonly used fluorescent labeling compounds are fluorescein isothiocyanate, rhodamine, phycoerythrin, phycocyanin, allophycocyanin, o-phthaldehyde and fluorescamine.

[0085] The antibody can also be detectably labeled using fluorescence emitting metals such as 152 Eu, or others of the lanthanide series. These metals can be attached to the antibody using such metal chelating groups as diethylenetriaminepentacetic acid (DTPA) or ethylenediaminetetraacetic acid (EDTA).

[0086] The antibody also can be detectably labeled by coupling it to a chemiluminescent compound. The presence of the chemiluminescent-tagged antibody is then determined by detecting the presence of luminescence that arises during the course of a chemical reaction. Examples of particularly useful chemiluminescent labeling compounds are luminol, isoluminol, theromatic acridinium ester, imidazole, acridinium salt and oxalate ester. Likewise, a bioluminescent compound may be used to label the antibody of the present invention. Bioluminescence is a type of chemiluminescence found in biological systems in which a catalytic protein increases the efficiency of the chemiluminescent reaction. The presence of a bioluminescent protein is determined by detecting the presence of luminescence. Important bioluminescent compounds for purposes of labeling are luciferin, luciferase and aequorin.

[0087] The binding activity of a given lot of anti-ECDIIIa-variant specific antibody may be determined according to well-known methods. Those skilled in the art will be able to determine operative and optimal assay conditions for each determination by employing routine experimentation.

[0088] Accordingly, the present invention, including the unexpected discovery of a plurality of variable sequence positions within the proline-rich ECDIIIa region, along with antibodies specific for particular ECDIIIa variants, provides for valuable prognostic and diagnostic information and assays.

[0089] Accordingly, the present invention further proivides a method for determining the prognosis of tumor treatment in a patient for a tumor that overexpresses HER-2, comprising: (a) obtaining a bodily fluid sample from the patient, wherein the bodily fluid is selected from the group consisting of blood, serum, urine, lymph, saliva, tumor tissue, placental tissue, umbilical cord tissue, amniotic fluid, chorionic villi tissue, and combinations thereof; and (b) measuring the amount of p68HER-2 expressed using an anti-p68HER-2 antibody-based assay, wherein the assay is selected from the group consisting of ELISA, immunoprecipitation, immunohistocytochemistry, and Western analysis. Preferably, the method for determining the prognosis of tumor treatment further comprises measuring the amount of p185HER-2 ECD in the bodily fluid, and determining a ratio between the amount of p68HER-2 and p185HER-2. The higher the ratio of p68HER-2:p185HER-2, the better the treatment prognosis. Preferably, the method for determining the prognosis of tumor treatment further comprises determining which particular ECDIIIa variants are present and optimizing tumor treatment in view of particular biochemical and biological properties among ECDIIIa protein variants.

[0090] The present invention further provides an assay for cancer treatment, prognosis or diagnosis in a patient comprising: (a) obtaining a bodily fluid sample from the patient, wherein the bodily fluid is selected from the group consisting of blood, serum, urine, lymph, saliva, tumor tissue, placental tissue, umbilical cord tissue, amniotic fluid, chorionic villi tissue and combinations thereof; (b) determining whether a particular ECDIIIa variant sequence is present in the bodily fluid sample with a sequence identity assay; and (c) correlating the presence of the ECDIIIa variant sequence to cancer treatment and diagnosis using an historical database. Preferably, the sequence identity assay is selected from the group consisting of DNA sequencing, PCR assays, ELISA immunologic assays, immunoassays, hybridization assays, and combinations thereof.

[0091] The present invention further provides an assay for cancer treatment, prognosis or diagnosis in a patient comprising: (a) obtaining a bodily fluid sample from the patient, wherein the bodily fluid is selected from the group consisting of blood, serum, urine, lymph, saliva, tumor tissue, placental tissue, umbilical cord tissue, amniotic fluid, chorionic villi tissue and combinations thereof; (b) determining whether an amount of an p68HER-2 ECDIIIa variant is present in the bodily fluid sample using an anti-p68HER-2 antibody-based assay, wherein the assay is selected from the group consisting of ELISA, immunoprecipitation, immunohistocytochemistry, and Western analysis; and (c) correlating the presence or amount of the p68HER-2 ECDIIla variant to cancer treatment and diagnosis using an historical database.

[0092] The present invention further provides for the above-mentioned cancer treatment, prognostic or diagnostic assays, further comprising measuring the amount of p185HER-2 ECD in the bodily fluid sample.

[0093] The present invention further provides for the above-mentioned cancer treatment, prognostic or diagnostic assays further comprising measuring the amount of p185HER-2 ECD in the bodily fluid sample, and determining a ratio between the amount of p185HER-2 ECD and a particular p68HER-2 ECDIIIa variant.

[0094] The present invention further provides for antibodies specific for ECDIIIa variants of the sequence in SEQ ID NO: 1 or SEQ ID NO: 2, below.

[0095] P68HER-2 as a Therapeutic Agent

[0096] Without being bound by theory, but it appears that p68HER-2 or ECDIIIa peptide inhibits the growth of tumor cells that overexpress HER-2 by binding to p185HER-2 at the cells surface. This hypothesis was examined by testing anchorage independent growth of cells in the presence or absence of p68HER-2 using cells that depend on p185HER-2 overexpression for their malignant growth, yet have little or no detectable p68HER-2. Anchorage independent growth of cells in soft agar was used as a predictive model for tumor cytotoxicity. This is a common and predictive procedure to examine transforming activity and reflects the tumorigenic and oncogenic potential of cells (DiFore et al., Science 237:178-182, 1987; Hudziak et al., Proc. Natl. Acad. Sci. USA 84:7159-7163, 1987; and Baasner et al., Oncogene 13:901-911, 1996).

[0097] The effects of p68HER-2 on anchorage independent growth in soft agar was determined using SKOV-3 carcinoma cells and HER-2 transfected 17-3-1 cells, which are both tumorigenic and overexpress p185HER-2. The cells were suspended in media supplemented with fetal calf serum in the presence or absence of p68HER-2 and incubated for 21 days in a humidified incubator. Anchorage independent growth was quantitated by counting the number of colonies that contained more than 50 cells. FIG. 7 shows that in the presence of p68HER-2, anchorage independent growth of both SKOV-3 cells and 17-3-1 cells was inhibited several fold. Accordingly, these data show that p68HER-2 is not just cytostatic, but cytotoxic and possibly apoptotic.

[0098] Coexpression of Herstatin Inhibits p185HER-2 Levels and Inhibits Colony Formation.

[0099] The effects of ectopic herstatin expression on p185HER-2 in transfected Cos-7 cells were examined. FIG. 9 shows that as compared to cells transfected with HER-2 alone, p185 levels were decreased about 5 fold when herstatin was coexpressed. If herstatin expression caused diminished survival of transfected cells, this may explain decreased p385HER-2 levels in the cotransfected cells. Whether expression of the marker -galactosidase plasmid was altered in the cotransfected cells was examined. FIG. 10 illustrates that the -gal activity was greatest when the -gal plasmid was cotransfected with p185HER-2, which was about 35% higher than when cotransfected with control (pcDNA empty plasmid). When cells were cotransfected with herstatin in combination with p185HER-2 and the -gal plasmid, the level of -gal activity was reduced about 3 fold. This reduction was observed in two additional cell lines, CHO and HEK-293. Moreover, a similar level of inhibition was observed when expression of an alternative marker, fluorescent green protein (FGP), was examined in the cells transfected with herstatin and HER-2 (not shown). The results show that transfection with HER-2 alone increased the level of a cotransfected marker, while coexpression with herstatin and HER-2 reduced the marker signal. One possible explanation is that survival of the transfected cells is enhanced by expression of p185HER-2 and is inhibited by coexpression of herstatin and p185HER-2.

[0100] To directly measure effects on colony formation, CHO cells were transfected, subjected to selection with G418 for 20 days, and the surviving cell colonies were stained and quantitated. CHO cells are well suited for rapid growth of stably transfected cells. As shown in FIG. 11 , transfection with p185HER-2 enhanced colony growth by about 60% relative to the control transfected cells, while herstatin expressed alone did not significantly alter growth. However, cotransfection of herstatin dramatically reduced colony formation of HER-2 transfected cells by about 7-fold ( FIG. 11 ).

[0101] Herstatin Expression Inhibits p185HER-2 Tyrosine Phosphorylation

[0102] The impact of herstatin expression on p185HER-2 was further evaluated by varying the proportion of herstatin to p185HER-2 expression plasmids. When the herstatin plasmid was introduced at five-fold lower levels relative to the p185HER-2 plasmid, there was little effect on p185. As shown in FIG. 14 , as the amount of the herstatin plasmid was increased relative to the HER-2 plasmid, p185HER-2 levels were reduced and the expression of cotransfected marker GFP plasmid was reduced as in FIG. 1 . In contrast, as shown in FIG. 13 , transfection with the control GFP expression plasmid alone, which does not affect cell survival, resulted in GFP expression that was proportional to the amount of plasmid added. Therefore, reduced levels of p185 HER-2, detected by Western analysis, appears to be related to diminished survival of the transfected cells. As shown in FIG. 15 , lane 6, while there was a three-fold reduction in p185HER-2 levels, p185 tyrosine phosphorylation signal was diminished by 25-fold, indicating an approximate 8-fold decrease in the level of tyrosine phosphorylation of p185HER-2. The extent of inhibition of p185 tyrosine phosphorylation was affected by the ratio of herstatin to HER-2 expression plasmid with the strongest effect observed at a ratio of 2 to 1. As sown in FIG. 15 , comparing lanes 1 and 6, transfection of the herstatin plasmid alone reduced the phosphotyrosine signal associated with an 185 kDa protein by about two-fold. The p185 protein comigrated with p185HER-2 expressed from the transfected plasmid and may be endogenous p185HER-2. These results show that ectopic expression of herstatin reduced the tyrosine phosphorylation of exogenous p185HER-2, and that the ratio of herstatin to HER-2 expression plasmids affects the extent of inhibition of p185 HER-2 tyrosine phosphorylation and the level of expression of p185. Further these findings show that reduced p185HER-2 tyrosine phosphorylation correlated with diminished colony formation of cells cotransfected with HER-2 and herstatin.

[0103] Herstatin expression strongly inhibits colony formation of cells that overexpress p185HER-2. In addition, herstatin suppresses the increased survival provided by overexpression of the EGF receptor. Diminished cell survival correspondes to a reduction in tyrosine phosphorylation of p185HER-2 and to interference with EGF activation of the EGF receptor. Further evidence is disclosed of the inhibitory activity against p185HER-2 and extends the negative regulatory activity of herstatin to a second member of the group I family of receptor tyrosine kinases, the EGF receptor.

[0104] Herstatin inhibits p185HER-2 tyrosine phosphorylation. Transfection of herstatin diminishes the level of constitutive tyrosine phosphorylation of ectopically expressed p185 by 8-fold. This inhibition is observed in transfected CHO cells, HEK-293 cells and Cos-7 cells. The extent of inhibition of p185 tyrosine phosphorylation by herstatin is dose dependent since it is affected by the amount of herstatin relative to HER-2 plasmid added to cells. Herstatin may inhibit p185HER-2 receptor tyrosine phosphorylation by preventing receptor dimerization, an obligatory step in receptor transphosphorylation and activation.

[0105] Ectopic expression of herstatin also interferred with EGF-activation of the EGF receptor, a group I receptor with homology to p185HER-2. Herstatin dampened EGF-induced receptor tyrosine phosphorylation, depressed EGF-stimulated tyrosine phosphorylation of two additional cytoplasmic proteins, and interferred with receptor down regulation. Receptor tyrosine phosphorylation and down regulation are halmarks of EGF activaton of the EGF receptor. The inhibition of EGF receptor by herstatin occurred at saturating concentrations of EGF. The intron 8-encoded domain at the C-terminus of herstatin, which appears to confer high affinity binding to p185HER-2, binds to the EGF receptor but does not compete with EGF binding (Doherty et al., submitted). These results suggest that herstatin blocks activation of the EGF receptor by a mechanism that does not involve competition with growth factor binding. Other studies show that herstatin associates with the EGF receptor in “pull-down” assays (Doherty, submitted). Since herstatin appears to form a stable complex with the EGF receptor, it may inhibit EGF-induced homodimers. It is also possible that herstatin inhibits the EGF receptor indirectly by preventing EGF-induced heterodimer formation or by an alternative mechanism that does not involve complex formation with the receptors.

[0106] Overexpression of p185HER-2 or the EGF receptor enhances cell growth and/or survival. Transfection of cells with p185HER-2 enhanced the number of colonies surviving G418 selection by about 60% and transfection with the EGF receptor increased the number of cell colonies by about 40% compared to the control transfected cells. Herstatin, transfected alone, did not significantly alter the number of colonies, although it did appear to decrease the tyrosine phosphorylation associated with an endogenous protein that comigrated with p185HER-2 in CHO cells. This may have been p185HER-2, since the CHO cells have little or no group I receptors other than p185HER-2. Herstatin co-expressed with the EGF receptor suppressed colony formation relative to cells transfected with the EGF receptor alone. Since herstatin also inhibited EGF activation of the EGF receptor, and has been found to form stable complexes with the EGF receptor (Doherty, submitted), the growth inhibition may have been due to herstatin forming an inactive kinase complex with the EGF receptor or through inhibition of transactivation of the EGF receptor by forming a complex with endogenous p185HER-2. It is also possible that the growth inhibition may have been caused by a separate unknown activity of herstatin that does not directly involve interaction with the EGF receptor or p185HER-2. The most potent growth inhibitory activity was observed when herstatin was co-expressed with p185HER-2. With this combination, there was a 7-fold reduction in colony formation relative to cells transfected with p185HER-2 alone, and a 5-fold inhibition compared to cells transfected with herstatin alone or compared to control transfected cells. Overexpressed p185HER-2 is constitutively active in cells, and blocking its activity may inhibit cell survival as well as growth.

[0107] In comparison to other group I RTK inhibitors that have been described, herstatin has several similarities with the p185neu ectodomain mutant, which functions as a dominant-negative inhibitor. Like herstatin, the membrane-anchored dominant-negative neu mutant contains sequences from the receptor extracellular domain and forms complexes with RTKs (Doherty et al., and Greene). Also, both p185 neu ectodomain and ectopically expressed herstatin are capable of inhibition of p185HER-2 as well as the EGF receptor. In contrast to the previously described p185neu ectodomain, which consists of subdomains I through IV and a membrane anchor, herstatin contains only subdomains I and II from the p185 ectodomain, and is a naturally occuring product expressed in fetal kidney and fetal liver cells (Doherty et al.,1999). Herstatin is secreted and does not require a membrane anchor to form complexes nor to exert its inhibitory activity.

EXAMPLE 1

[0108] This example provides the results from an experiment to investigate HER-2 mRNA diversity within the extracellular domain (ECD) coding sequence using polymerase chain reaction (PCR). A cDNA library from SKOV-3 cells (American Type Culture Collection (Rockville, Md.) maintained in DMEM, supplemented with 10% fetal bovine serum and 0.05% gentamycin), an ovarian carcinoma cell line in which the HER-2 gene is amplified eight times (Tyson et al., Am. J. Obstet. Gynecol. 165:640-646, 1991) was examined using a forward primer specific for exon 1 (Tal et al., Mol. Cell. Biol. 7, 2597-2601, 1987) identical to nucleotides 142-161 and a reverse primer complementary to nucleotides 1265-1286 in exon 9 (Scott et al., Mol. Cell. Biol. 13:2247-2257, 1993). Briefly, The SKOV-3 cDNA library was provided by Origene Technologies, Inc. (Rockville, Md.), and was prepared from RNA extracted from SKOV-3 cells. RNA was extracted from SKOV-3 cells grown to 80% confluence on 15 cm plates with TriReagent (Molecular Research Center, Inc., Cincinnati, Ohio), according to the manufacturer's protocol, to obtain total RNA. RNA was resuspended in 10 mM Tris-EDTA, pH 8.0, for reverse transcription and cDNA library construction, or in RNA hybridization buffer (80% formamide, 40 mM PIPES, 4 mM NaCl, 1 mM EDTA, pH 7.5) for ribonuclease protection assay (RPA). RNA concentrations were determined spectrophotometrically at OD 260 . Poly A + mRNA was selected from total RNA using a mRNA extraction kit (Oligotex, Qiagen).

[0109] A product of ˜1420 bp, determined to be HER-2-specific by Southern blotting, was approximately 270 bp larger than the expected size of 1144 bp from the previously described cDNA sequence (Coussens et al., Science 230:1132-1139, 1985). Briefly, the Southern blotting procedure transferred nucleic acids from agarose gels under vacuum (Bio-Rad Model 785 Vacuum Blotter) in 0.4 M NaOH to Gene Screen Plus Hybridization Transfer Membrane (NEN Research Products, Boston, Mass.). Nucleic acids were fixed to membranes by UV crosslinking in a UV-Stratalinker (Stratagene, Inc., La Jolla, Calif.), and the membranes were blocked in hybridization buffer (50% formamide, 5×SSC, 1% SDS, 10 mg/ml herring sperm DNA) at 42° C. for 2 h. The membranes were hybridized at 42° C. for 16 h in hybridization buffer with 10 7 cpm of a 220 bp Kpn-HincII fragment from ECDIIIa cDNA labelled with (α- 32 P)dCTP (NEN Life Sciences) using a Random Prime DNA Labelling Kit (Boehringer Mannheim).

[0110] Templates were amplified in a Perkin Elmer GeneAmp PCR System 2400 (Perkin Elmer Cetus, Emeryville, Calif.) using the Expand High Fidelity PCR System (Boerhinger Mannheim) with 1× High Fidelity PCR buffer containing 2.5 mM MgCl 2 , 5 μM of each primer, and 200 μM dNTPs. All primers were obtained from GIBCO BRL (Life Technologies). Numbering of nucleotide and amino acid residues is according to the HER-2 cDNA sequence reported by Coussens et al. (Coussens et al., Science 230:1132-1139, 1985). The HER-2 extracellular domain was targeted for amplification from an SKOV-3 cDNA library (Origene Technologies, Inc.) using a forward primer (A) identical to nucleotides (nt) 142-161 of HER-2 cDNA (5′-TGAGCACC ATG GAGCTGGC-3′ [SEQ ID NO 3]), which spans the initiation codon (underlined) and a reverse primer (B) (5′-TCCGGCAGAAATGCCAGGCTCC-3′ [SEQ ID NO 4]), which is complementary to HER-2 exon sequence at nt 1265-1286. Cycling parameters were: 94° C., 30 sec; 58° C., 45 sec; 68° C., 3 min, for 30 cycles. The region spanning the alternative sequence (denoted ECDIIIa) from genomic DNA, was amplified using a forward primer (C) (5′-AACACAGCGGTGTGAGAAGTGC-3′ [SEQ ID NO 5]) identical to HER-2 exon-specific sequence at nt 1131-1152 and the reverse primer (B) [SEQ ID NO. 4] on DNA prepared as described (Bond et al., FEBS Letters 367:61-66, 1995) with cycling parameters: 94° C., 30 sec; 62° C., 30 sec; 72° C., 60 sec, for 25 cycles.

[0111] Reverse transcriptase-polymerase chain reaction (RT-PCR) was used to investigate the structure of mRNA containing the ECDIIIa sequence. First strand cDNA was reverse transcribed (Bond et al., FEBS Letters 367:61-66, 1995) using 5 μg RNA primed with 0.5 μg oligo-dT. To amplify the ECDIIIa insert and adjacent 5′ HER-2 exon sequence, a forward primer (A) described above and a reverse primer (D) (5′- ATACCGGGACAGGTCAACAGC-3′ [SEQ ID NO 6]) which is complementary to the 3′ECDIIIa-specific sequence were used. Cycling parameters were: 94° C., 30 sec; 60° C., 40 sec; 68° C., 2 min, for 30 cycles.

[0112] Amplification of the ECDIIIa insert and adjacent 3′ HER-2 exon-specific sequence was with a forward primer (E) (5′-TCTGGGTACCCACTCACTGC-3′ [SEQ ID NO 7]) which is identical to the 5′ECDIIIa-specific sequence and contains a Kpn1 restriction site and a reverse primer (F) (5′-T TCA CACTGGCACGTCCAGACC-3′ [SEQ ID NO 8]) which is complementary to HER-2 exon sequence at nt 3898-3919 and spans the termination codon (underlined). Cycling parameters were: 94° C., 30 sec; 60° C., 40 sec; 68° C., 5 min, for 30 cycles.

[0113] The PCR product was subcloned and the nucleotide sequence was determined.

[0114] The results showed that the normal HER-2 coding sequence was present beginning with the 5′ primer sequence and continued uninterrupted through nucleotide 1171. At this position, a 274 nucleotide insertion was found, followed by the expected coding sequence, including the 3′ primer sequence. Analysis of the predicted protein product showed that the 274 nucleotide insertion encodes an extension of the known HER-2 protein, beginning at residue 340 (Coussens et al., Science 230:1132-1139, 1985), and introduces an in-frame stop codon 79 amino acids later ( FIG. 1 ). Comparison of the inserted nucleotides and their predicted amino acid sequence with sequences in Genbank showed no homologies. Examination of the 5′ and 3′ junctions of the divergent sequence revealed consensus splice donor and acceptor sites (Sharp, and Burge, Cell 91:875-879, 1997) and include a pyrimidine tract and potential branchpoint adenine residues near the 3′end of the insert sequence ( FIG. 1 ). Thus, the inserted sequence is likely to be an intron.

[0115] Inspection of the predicted amino acid sequence of the novel 79 amino acids [SEQ ID NO. 1] encoded by the inserted sequence shows a consensus N-linked glycosylation site and a high proline content of 19% ( FIG. 1 ). The inserted sequence was designated ECDIIIa since it is located at the boundary between subdomains II and III in the extracellular domain of the p185HER-2 sequence (Lax et al., Mol. Cell. Biol. 8:1831-1834, 1988). The insert sequence is in-frame with the adjacent 5′ HER-2 exon sequence for 236 nt where there is a termination codon.

EXAMPLE 2

[0116] This example provides the results from experiments characterizing ECDIIIa as contiguous with HER-2 exons in the genome. To investigate the HER-2 gene structure in the region of the ECDIIIa sequence, a forward primer, identical to nucleotides 763-785, and a reverse primer, complementary to nucleotides 1265-1286 of the HER-2 cDNA, were used in the PCR on human genomic DNA. The amplification product was anticipated to span exon 5 (Tal et al., Mol. Cell. Biol. 7:2597-2601, 1987) to an exon which is immediately 3′ of the ECDIIIa sequence. Intron number and sizes were estimated based on PCR product sizes, restriction digest analysis, and partial sequence analysis of amplification products.

[0117] Next, human genomic DNA was examined using HER-2 exon-specific primers that directly flank the insert to determine the sequences immediately flanking the ECDIIIa sequence. A ˜430 bp product was amplified from normal human genomic DNA and from genomic DNA extracted from carcinoma cell lines SKOV-3, SKBR-3 and BT474, all of which have HER-2 gene amplification (Kraus et al., EMBO J. 6:605-610, 1987) and were found to express ECDIIIa in their cDNA. The identities of the PCR products as HER-2 were verified by Southern blot analysis using the procedure described in Example 1. Nucleotide sequence analysis showed that the PCR product from human genomic DNA contained the ECDIIIa insert, flanked immediately on both sides by known HER-2 coding sequence; no mutations or rearrangements were seen. These data show that the ECDIIIa sequence represents a wholly retained intron, likely intron 8 based on the size of products amplified following intron 4 and on the location of intron 8 in the homologous EGFR gene and HER-3 gene (Lee and Maihle, Oncogene 16:3243-3252, 1998).

EXAMPLE 3

[0118] This example shows that ECDIIIa is the only retained intron within the coding sequence of HER-2 mRNA. To determine whether additional introns were retained in the mRNA containing the ECDIIIa insert sequence, the reverse transcriptase-polymerase chain reaction (RT-PCR) was employed. First, a forward primer identical to 5′ HER-2 cDNA sequence at 142-161 which spans the initiation codon, and a reverse primer complementary to the 3′ ECDIIIa sequence were employed with SKBR-3 and SKOV-3 cDNA. A product of 1.3 kb was amplified, which is the size expected if the product contained no introns other than intron 8. Amplification of the 3′HER-2 coding sequence was then performed using a forward primer identical to 5′ ECDIIIa sequence and a reverse primer complementary to 3′HER-2 cDNA sequence at nucleotides 3898-3919, which spans the p185HER-2 termination codon. A product of 2.9 kb was amplified, which is the size expected from the HER-2 cDNA if no additional introns were retained.

[0119] Further characterizations of both the 5′(1.3 kb) and 3′(2.9 kb) amplification products by restriction digest analysis and nucleotide sequencing confirmed the absence of additional retained introns. To determine the size of the products amplified when intron sequences are included, genomic DNA was used as a template for the PCR reactions, which resulted in products of approximately 10 kb for the 5′ coding sequence and 5 kb for the 3′ coding sequence. These results indicate that the alternative HER-2 transcript, resulting from retention of an intron of 274 bp, was expected to be about 4.8 kb in size, assuming that the 5′untranslated (5′UTR) and 3′untranslated (3′UTR) regions are identical in size to the previously described ˜4.5 kb HER-2 cDNA (Coussens et al., Science 230:1132-1139, 1985).

EXAMPLE 4

[0120] This example illustrates the expression of a protein containing an ECDIIIa sequence. To assess whether the alternative sequence is translated into a protein product, the ECDIIIa sequence was expressed as a polyhistidine-tagged peptide in bacteria, purified the peptide by nickel-affinity chromatography, and raised antisera against the purified peptide. Briefly, the bacterial expression vector was prepared by amplifying the ECDIIIa sequence from the SKOV-3 cDNA library using primer E and a reverse primer complementary to the 3′ end of the ECDIIIa insert sequence. The reverse primer contained a BamH1 restriction site sequence, and was identical to that used for template construction in the RPA (described in examples 1 and 2). The PCR amplification product of ˜280 bp was di