The present invention relates to newly identified proteins as markers for the detection of tumors, or as targets for their treatment, particularly of tumors affecting lung, colon, breast and ovary. Also provided are affinity ligands capable of selectively interacting with the newly identified markers, as well as methods for tumor diagnosis and therapy using such ligands.

Background of the invention

Tumor markers (or biomarkers)

Tumor markers are substances that can be produced by tumor cells or by other cells of the body in response to cancer. In particular, a protein biomarker is either a single protein or a panel of different proteins that could be used to unambiguously distinguish a disease state. Ideally, a biomarker would have both a high specificity and sensitivity, being represented in a significant percentage of the cases of given disease and not in healthy state.

Biomarkers can be identified in different biological samples, like tissue biopsies or preferably biological fluids (saliva, urine, blood-derivatives and other body fluids), whose collection does not necessitate invasive treatments.

Tumor marker levels may be categorized in three major classes on the basis of their clinical use. Diagnostic markers can be used in the detection and diagnosis of cancer. Prognostics markers are indicative of specific outcomes of the disease and can be used to define predictive models that allow the clinicians to predict the likely prognosis of the disease at time of diagnosis. Moreover, prognosis markers are helpful to monitor the patient response to a drug therapy and facilitate a more personalized patient management. A decrease or return to a normal level may indicate that the cancer is responding to therapy, whereas an increase may indicate that the cancer is not responding. After treatment has ended, tumor marker levels may be used to check for recurrence of the tumor. Finally, therapeutic markers can be used to develop tumor-specific drugs or affinity ligand (i.e. antibodies) for a tumor treatment.

Currently, although an abnormal tumor marker level may suggest cancer, this alone is usually not enough to accurately diagnose cancer and their measurement in body fluids is frequently combined with other tests, such as a biopsy and radioscopic examination. Frequently, tumor marker levels are not altered in all of people with a certain cancer disease, especially if the cancer is at early stage. Some tumor marker levels can also be altered in patients with noncancerous conditions. Most biomarkers commonly used in clinical practice do not reach a sufficiently high level of specificity and sensitivity to unambiguously distinguish a tumor from a normal state.

To date the number of markers that are expressed abnormally is limited to certain types/subtypes of cancer, some of which are also found in other diseases, (http://www.cancer.gov/cancertopics/factsheet). For example, prostate-specific antigen (PSA) levels are often used to screen men for prostate cancer, but this is controversial since elevated PSA levels can be caused by both prostate cancer or benign conditions, and most men with elevated PSA levels turn out not to have prostate cancer.

Another tumor marker, Cancer Antigen 125, (CA 125), is sometimes used to screen women who have an increased risk for ovarian cancer. Scientists are studying whether measurement of CA 125, along with other tests and exams, is useful to find ovarian cancer before symptoms develop. So far, CA 125 measurement is not sensitive or specific enough to be used to screen all women for ovarian cancer. Mostly, CA 125 is used to monitor response to treatment and check for recurrence in women with ovarian cancer. Finally, human epidermal growth factor receptor (HER2) is a marker protein overproduced in about 20% of breast cancers, whose expression is typically associated with a more aggressive and recurrent tumors of this class. Routine screening test for tumor diagnosis

Screening tests are a way of detecting cancer early, before there are any symptoms. For a screening test to be helpful, it should have high sensitivity and specificity. Sensitivity refers to the test's ability to identify people who have the disease. Specificity refers to the test's ability to identify people who do not have the disease. Different molecular biology approaches such as analysis of DNA sequencing, small nucleotide polymorphyms, in situ hybridization and whole transcriptional profile analysis have done remarkable progresses to discriminate a tumor state from a normal state and are accelerating the knowledge process in the tumor field. However so far different reasons are delaying their use in the common clinical practice, including the higher analysis complexity and their expensiveness. Other diagnosis tools whose application is increasing in clinics include in situ hybridization and gene sequencing.

Currently, Immuno-HistoChemistry (IHC), a technique that allows the detection of proteins expressed in tissues and cells using specific antibodies, is the most commonly used method for the clinical diagnosis of tumor samples.

This technique enables the analysis of cell morphology and the classification of tissue samples on the basis of their immunoreactivity. However, at present, IHC can be used in clinical practice to detect cancerous cells of tumor types for which protein markers and specific antibodies are available. In this context, the identification of a large panel of markers for the most frequent cancer classes would have a great impact in the clinical diagnosis of the disease.

Anti-cancer therapies

In the last decades, an overwhelming number of studies remarkably contributed to the comprehension of the molecular mechanisms leading to cancer. However, this scientific progress in the molecular oncology field has not been paralleled by a comparable progress in cancer diagnosis and therapy.

Surgery and/or radiotherapy are the still the main modality of local treatment of cancer in the majority of patients. However, these treatments are effective only at initial phases of the disease and in particular for solid tumors of epithelial origin, as is the case of colon, lung, breast, ovary, prostate and others, while they are not effective for distant recurrence of the disease. In some tumor classes, chemotherapeutic treatments have been developed, which generally relies on drugs, hormones and antibodies, targeting specific biological processes used by cancers to grow and spread. However, so far many cancer therapies had limited efficacy due to severity of side effects and overall toxicity. Indeed, a major effort in cancer therapy is the development of treatments able to target specifically tumor cells causing limited damages to surrounding normal cells thereby decreasing adverse side effects. Recent developments in cancer therapy in this direction are encouraging, indicating that in some cases a cancer specific therapy is feasible. In particular, the development and commercialization of humanized monoclonal antibodies that recognize specifically tumor- associated markers and promote the elimination of cancer is one of the most promising solution that appears to be an extremely favorable market opportunity for pharmaceutical companies. However, at present the number of therapeutic antibodies available on the market or under clinical studies is very limited and restricted to specific cancer classes. So far licensed monoclonal antibodies currently used in clinics for the therapy of specific tumor classes show only a partial efficacy and are frequently associated with chemotherapies to increase their therapeutic effect. Administration of Trastuzumab (Herceptin), a commercial monoclonal antibody targeting HER2in conjunction with Taxol adjuvant chemotherapy induces tumor remission in about 42% of the cases (1). Bevacizumab (Avastin) and Cetuximab (Erbitux) are two monoclonal antibodies recently licensed for use in humans, targeting the endothelial and epithelial growth factors respectively that, combined with adjuvant chemotherapy, proved to be effective against different tumor diseases. Bevacizumab proved to be effective in prolonging the life of patients with metastatic colorectal, breast and lung cancers. Cetuximab demostrated efficacy in patients with tumor types refractory to standard chemotherapeutic treatments

(I)- In summary, available screening tests for tumor diagnosis are uncomfortable or invasive and this sometimes limits their applications. Moreover tumor markers available today have a limited utility in clinics due to either their incapability to detect all tumor subtypes of the defined cancers types and/or to distinguish unambiguously tumor vs. normal tissues. Similarly, licensed monoclonal antibodies combined with standard chemotherapies are not effective against the majority of cases. Therefore, there is a great demand for new tools to advance the diagnosis and treatment of cancer.

Experimental approaches commonly used to identify tumor markers

Most popular approaches used to discover new tumor markers are based on genome-wide transcription profile or total protein content analyses of tumor.

These studies usually lead to the identification of groups of mRNAs and proteins which are differentially expressed in tumors. Validation experiments then follow to eventually single out, among the hundreds of RNAs/proteins identified, the very few that have the potential to become useful markers. Although often successful, these approaches have several limitations and often, do not provide firm indications on the association of protein markers with tumor. A first limitation is that, since frequently mRNA levels not always correlate with corresponding protein abundance (approx. 50% correlation), studies based on transcription profile do not provide solid information regarding the expression of protein markers in tumor (2, 3, 4, 5).

A second limitation is that neither transcription profiles nor analysis of total protein content discriminate post-translation modifications, which often occur during oncogenesis. These modifications, including phosphorylations, acetylations, and glycosylations, or protein cleavages influence significantly protein stability, localization, interactions, and functions (6).

As a consequence, large scale studies generally result in long lists of differentially expressed genes that would require complex experimental paths in order to validate the potential markers. However, large scale genomic/proteomic studies reporting novel tumor markers frequently lack of confirmation data on the reported potential novel markers and thus do not provide solid demonstration on the association of the described protein markers with tumor. Approach used to identify the protein markers included in the present invention

The approach that we used to identify protein markers is based on an innovative immuno-proteomic technology. In essence, a library of recombinant human proteins has been produced from E. coli and is being used to generate polyclonal antibodies against each of the recombinant proteins.

The screening of the antibodies library on Tissue microarrays (TMAs) carrying clinical samples from different patients affected by the tumor under investigation leads to the identification of specific tumor marker proteins. Therefore, by screening TMAs with the antibody library, the tumor markers are visualized by IHC, the classical technology applied in all clinical pathology laboratories. Since TMAs also include healthy tissues, the specificity of the antibodies for the tumors can be immediately appreciated and information on the relative level of expression and cellular localization of the markers could be obtained. In our approach the markers are subjected to a validation process consisting in a molecular and cellular characterization.

Altogether, the detection the marker proteins disclosed in the present invention selectively in tumor samples and the subsequent validation experiments lead to an unambiguous confirmation of the marker identity and confirm its association with defined tumor classes. Moreover this experimental process provides an indication of the possible use of the proteins as tools for diagnostic or therapeutic intervention. For instance, proteins showing a cell surface localization could be both diagnostic and therapeutic markers, against which both chemical and antibody therapies can be developed. Differently, markers showing a cytoplasmic localization could be more likely considered for the development of tumor diagnostic tests and chemotherapy/small molecules treatments.

Summary of the invention The present invention provides new means for the detection and treatment of tumors, in particular colo-rectal, lung, ovary and breast cancers, based on the identification of protein markers specific for these tumor types, namely: a) Tectonic- 1 (TCTNl), Tectonic-2 (TCTN2) and Tectonic-3 (TCTN3) proteins, b) HIGD2A and HIGD2B proteins, c) chromosome 4 open reading frame 32 (C4orf32) protein and d) FAM62A protein. In preferred embodiments, the invention provides the use of a) TCTNl, TCTN2 and TCTN3 proteins, as marker or target for colon, lung and ovary tumors, b) Hypoxia induced proteins A and B (HIGD2A and HIGD2B) as markers or targets for colon and ovary tumors, c) C4orf32 proteins as markers or targets for breast and lung tumors and d) FAM62A as marker or target for breast, colon, lung and ovary tumors.

The invention also provides a method for the diagnosis of these cancer types, comprising a step of detecting the above-identified markers in a biological sample, e.g. in a tissue sample of a subject suspected of having or at risk of developing malignancies or susceptible to cancer recurrences. In particular, the protein markers of the invention allow to specifically detect lung, colon, breast and ovary cancers, according to their tumor-specificity, namely: a) TCTNl, TCTN2 and TCTN3 proteins for colon, lung and ovary tumors; b) Hypoxia induced proteins A and B (HIGD2A and HIGD2B) for colon and ovary tumors; c) chromosome 4 open reading frame 32 proteins for breast and lung tumors; d) FAM62A protein for breast, colon, lung and ovary tumors.

In addition, the tumor markers identify novel targets for affinity ligands which can be used for therapeutic applications, especially in the treatment of colo-rectal, lung, ovary and breast proliferative diseases. Also provided are affinity ligands, particularly antibodies, capable of selectively interacting with the newly identified protein markers.

Detailed disclosure of the invention

The present invention is based on the surprising finding of antibodies that are able to specifically bind tumor tissues from patients, while negative or very poor binding is observed in normal tissues from the same patients. These antibodies have been found to specifically bind proteins for which no previous association with tumor has been reported. Hence, in a first aspect, the invention provides a tumor marker which is selected from the group consisting of: a) Tectonic-1 (TCTNl) protein, in one of its isoforms SEQ ID NO: 1,

SEQ ID NO:2, SEQ ID NO:3, SEQ ID NO:4, SEQ ID NO:5, SEQ

ID NO:6, SEQ ID NO:7 or SEQ ID NO:8 or a different isoform having sequence identity of at least 80%, preferably at least 90%, more preferably at least 95% to SEQ ID NO: 1, SEQ ID NO:2, SEQ ID NO:3, SEQ ID NO:4, SEQ ID NO:5, SEQ ID NO:6, SEQ ID NO:7 or SEQ ID NO:8; or a nucleic acid molecule containing a sequence coding for a TCTNl protein, said encoding sequence being preferably selected from SEQ ID NO:20, SEQ ID NO:21, SEQ ID NO:22,

SEQ ID NO:23, SEQ ID NO:24, SEQ ID NO:25, SEQ ID NO:26 and SEQ ID NO:27; b) Tectonic-2 (TCTN2) protein, SEQ ID NO: 9 or an isoform thereof having sequence identity of at least 80%, preferably at least 90%, more preferably at least 95% to SEQ ID NO:9; or a nucleic acid molecule containing a sequence coding for a TCTN2 protein, said encoding sequence being preferably SEQ ID NO: 28; c) Tectonic-3 (TCTN3) protein in one of its isoforms SEQ ID NO: 10, SEQ ID NO:11, SEQ ID NO:12 or SEQ ID NO:13, or a different isoform having sequence identity of at least 80%, preferably at least

90%, more preferably at least 95% to SEQ ID NO: 10, SEQ ID NO: 11, SEQ ID NO: 12 or SEQ ID NO: 13; or a nucleic acid molecule containing a sequence coding for a TCTN3 protein, said encoding sequence being preferably selected from SEQ ID NO:29, SEQ ID NO:30, SEQ ID NO:31 and SEQ ID NO:32; d) Hypoxia induced protein A (HIGD2A), SEQ ID NO: 14 or an isoform thereof having sequence identity of at least 80%, preferably at least 90%, more preferably at least 95% to SEQ ID NO: 14; or a nucleic acid molecule containing a sequence coding for a HIGD2A protein, said encoding sequence being preferably SEQ ID NO:33; e) Hypoxia induced protein B (HIGD2B), SEQ ID NO: 15 or an isoform thereof having sequence identity of at least 80%, preferably at least 90%, more preferably at least 95% to SEQ ID NO: 15; or a nucleic acid molecule containing a sequence coding for a HIGD2B protein, said encoding sequence being preferably SEQ ID NO:34; f) C4orf32 protein, SEQ ID NO: 16, or an isoform thereof having sequence identity of at least 80%, preferably at least 90%, more preferably at least 95% to SEQ ID NO: 16; or a nucleic acid molecule containing a sequence coding for a C4orf32 protein, said encoding sequence being preferably SEQ ID NO:35; g) FAM62A protein, in one of its variant isoforms SEQ ID NO: 17, SEQ ID NO:18 or SEQ ID NO: 19, or a different isoform having sequence identity of at least 80%, preferably at least 90%, more preferably at least 95% to any of SEQ ID NO: 17, SEQ ID NO: 18 or SEQ ID NO: 19; or a nucleic acid molecule containing a sequence coding for a FAM62A protein, said encoding sequence being preferably selected from SEQ ID NO: 36, SEQ ID NO: 37 and SEQ ID NO: 38. TCTNl (Gene name: TCTNl ; Gene ID: ENSG00000204852, Protein names: Tectonic- 1, TCTl, TCTNl. Eight variants identified: Protein ID Transcript ID

ENSP00000366882 ENST00000377654

ENSP00000380771 ENST00000397650 ENSP00000380772 ENST00000397652

ENSP00000380774 ENST00000397654

ENSP00000380775 ENST00000397655

ENSP00000380776 ENST00000397656

ENSP00000380777 ENST00000397657 ENSP00000380779 ENST00000397659),

TCTN2 (Gene names: TCTN2, C12orf38, TECT2 Gene ID: ENSG00000168778; Transcript ID: ENST00000303372; Protein Names: Tectonic-2, TCTN2; Protein ID: ENSP00000304941 ;), and TCTN3 (Gene names: TCTN3, ClOorfόl, TECT3; Gene ID: ENSGOOOOO l 19977; protein names: Tectonic-3, TCT3, TCTN3. Four variants identified:

Protein ID Transcript ID

ENSP00000265993 ENST00000265993

ENSP00000345815 ENST00000343162 ENSP00000360253 ENST00000371209

ENSP00000360261 ENST00000371217) are proteins without previous known association in any cancer disease and are preferably used as markers for colon-, lung- and ovary-type tumors. Antibodies generated against the TCTN2 protein show a selective immunoreactivity in histological preparation of colo-rectal cancer tissues, lung cancer tissues, ovary cancer tissues, which indicates the presence of TCTN2 in these cancer samples. The most striking finding is the marker positivity in colo- rectal carcinomas, with 100% positive samples, which makes TCTN2 protein and its antibody particularly convenient tools for distinguishing a colo-rectal cancer from a normal state.

As described in detail in the Examples, TCTNl, TCTN2 and TCTN3 proteins share a large common domain showing short highly conserved internal sequences and thereby, polyclonal antibodies generated against TCTN2 protein are capable to recognize also the related proteins TCTNl and TCTN3.

HIGD2A (HIGl domain family member 2 A, Protein ID: ENSP00000274787; Gene name: HIGD2A, Gene ID:ENSGOOO00146066; Transcript ID: ENST00000274787) and its homologous protein HIGD2BP (HIGl domain family member 2B, Protein ID: ENSP00000307951 ; Gene name: HIGD2BP, GenelD: ENSGOOOOO 175202; Transcript ID: ENST00000311755) are proteins without previous known association in any cancer disease and are preferably used as markers for colon- and ovary-type tumors. Antibodies generated to a fragment of HIGD2A show a selective immunoreactivity in histological preparation of colo-rectal cancer tissues and ovary cancer tissues, which indicates the presence of this protein in these cancer samples. As described in detail in the Examples, antibodies generated against HIGD2A protein are capable to recognize also its homologous protein HIGD2B and viceversa, indicating that both proteins can be stained by the same antibody. C4orf32 (Protein ID: ENSP00000310182; Gene ID: ENSGOOOOO 174749,

Gene Name: C4orf32; Transcript ID:ENST00000309733) is a protein without previous known association in any cancer disease and is preferably used as a marker for breast- and lung-type cancers. Antibodies generated to a fragment of C4orf32 show selective immunoreactivity in histological preparation of breast cancer tissues and lung cancer tissues, which indicates the presence this protein in these cancer samples.

FAM62A (Gene names: FAM62A, ESYTl, KIAA0747, MBC2; Gene ID: ENSG00000139641, Protein names: E-Sytl, Extended-synaptotagmin-1; Membrane-bound C2 domain-containing protein, Protein FAM62A. Three variants identified:

Protein ID Transcript ID

ENSP00000386045 ENST00000402331 ENSP00000377612 ENST00000394048

ENSP00000267113 ENST00000267113) is a protein without previous known association with any tumor class, although the corresponding genomic sequence can be isolated from pancreas tumor tissue along with a large number of different sequences (WO9955858). Antibodies against a fragment of FAM62A showed a selective immunoreactivity in histological preparation of tumor tissue samples from breast, colon, lung and ovary.

A further aspect of this invention is a method of screening a tissue sample for malignancy, which comprises determining the presence in said sample of at least one of the above-mentioned tumor markers. This method includes detecting either the marker protein, e.g. by means of labeled monoclonal or polyclonal antibodies that specifically bind to the target protein, or the respective mRNA, e.g. by means of polymerase chain reaction techniques such as RT-PCR. The methods for detecting proteins in a tissue sample are known to one skilled in the art and include immunoradiometric, immunoenzymatic or immunohistochemical techniques, such as radioimmunoassays, immunofluorescent assays or enzyme-linked immunoassays. Other known protein analysis techniques, such as polyacrylamide gel electrophoresis (PAGE), Western blot or Dot blot are suitable as well. Preferably, the detection of the protein marker is carried out with the immune-hystochemistry technology, particularly by means of High Through-Put methods that allow the analyses of the antibody immune-reactivity simultaneously on different tissue samples immobilized on a microscope slide. Briefly, each Tissue Micro Array (TMA) slide includes tissue samples suspected of malignancy taken from different patients, and an equal number of normal tissue samples from the same patients as controls. The direct comparison of samples by qualitative or quantitative measurement, e.g. by enzimatic or colorimetric reactions, allows the identification of tumors.

In one embodiment, the invention provides a method of screening a sample of colon or colo-rectal tissue for malignancy, which comprises determining the presence in said sample of a tumor marker selected from TCTNl, TCTN2, TCTN3, HIGD2A, HIGD2B and FAM62A, variants or isoforms or combinations thereof as described above. In another embodiment, the invention provides a method of screening a sample of lung tissue for malignancy, which comprises determining the presence in said sample of a tumor marker selected from TCTNl, TCTN2, TCTN3, C4orf32 and FAM62A, variants or isoforms or combinations thereof as described above. In a further embodiment, the invention provides a method of screening a sample of ovarian tissue for malignancy, which comprises determining the presence in said sample of a tumor marker selected from TCTNl, TCTN2, TCTN3, HIGD2A, HIGD2B and FAM62A, variants or isoforms or combinations thereof as described above. In a yet further embodiment, the invention provides a method of screening a sample of breast tissue for malignancy, which comprises determining the presence in said sample of a tumor marker selected from C4orG2 and FAM62A, variants or isoforms or combinations thereof as described above. A further aspect of the invention is a method in vitro for determining the presence of a tumor in a subject, which comprises the steps of:

(1) providing a sample of the tissue suspected of containing tumor cells;

(2) determining the presence of a tumor marker as above defined, or a combination thereof in said tissue sample by detecting the expression of the marker protein or the presence of the respective mRNA transcript; wherein the detection of one or more tumor markers in the tissue sample is indicative of the presence of tumor in said subject. The methods and techniques for carrying out the assay are known to one skilled in the art and are prefarbly based on immunoreactions for detecting proteins and on PCR methods for the detection of mRNAs. The same methods for detecting proteins or mRNAs from a tissue sample as disclosed above can be applied. A further aspect of this invention is the use of the tumor markers herein provided as targets for the identification of candidate antitumor agents. Accordingly, the invention provides a method for screening a test compound which comprises contacting the cells expressing a tumor-associated protein selected from TCTNl ₃ TCTN2, TCTN3, HIGD2A and HIGD2B, C4orf32 and FAM62A with the test compound, and determining the binding of said compound to said cells. In addition, the ability of the test compound to modulate the activity of each target molecule can be assayed.

A further aspect of the invention is a method of suppressing the function or expression of a tumor-associated protein herein provided. This includes inhibiting or blocking the protein, e.g. by means of antibodies, or silencing the gene encoding therefor, e.g. by RNA interference or RNA antisense technologies. As shown in the experimental section, marker expression confers a malignant phenotype to cells, making them able to grow and proliferate in an anchorage-independent fashion in an in vitro assay.

In one embodiment, the invention provides an antibody or a fragment thereof which is able to specifically recognize and bind to one of the tumor- associated proteins described above. The term "antibody" as used herein refers to all types of immunoglobulins, including IgG, IgM ₅ IgA, IgD and IgE. Such antibodies may include polyclonal, monoclonal, chimeric, single chain, antibodies or fragments such as Fab or scFv. The antibodies may be of various origin, including human, mouse, rat, rabbit and horse, or chimeric antibodies. The production of antibodies is well known in the art. For the production of antibodies in experimental animals, various hosts including goats, rabbits, rats, mice, and others, may be immunized by injection with polypeptides of the present invention or any fragment or oligopeptide or derivative thereof which has immunogenic properties or forms a suitable epitope. Monoclonal antibodies may be produced following the procedures described in Kohler and Milstein, Nature 265:495 (1975) or other techniques known in the art.

The antibodies to the tumor markers of the invention can be used to detect the presence of the marker in histologic preparations or to distinguish tumor cells from normal cells. To that purpose, the antibodies may be labeled with radiocative, fluorescent or enzyme labels. In addition, the antibodies can be used for treating proliferative diseases by modulating, e.g. inhibiting or abolishing the activity of a target protein according to the invention. Therefore, in a further aspect the invention provides the use of antibodies to a tumor-associated protein selected from TCTNl, TCTN2, TCTN3, HIGD2A and HIGD2B, C4orf32 and FAM62A, for the preparation of a therepautic agent for the treatment of proliferative diseases. For use in therapy, the antibodies can be formulated with suitable carriers and excipients, optionally with the addition of adjuvants to enhance their effetcs.

In a further embodiment, the invention provides a small interfering RNA (siRNAs) complementary to a sequence selected from the group consisting of SEQ ID NO:39 through SEQ ID NO:55, for use in tumor-gene silencing.

A further aspect of the invention relates to a diagnostic kit containing suitable means for detection, in particular the polypeptides or polynucleotides, antibodies or fragments or derivatives thereof described above, reagents, buffers, solutions and materials needed for setting up and carrying out the immunoassays, nucleic acid hybridization or PCR assays as described above. Parts of the kit of the invention can be packaged individually in vials or bottles or in combination in containers or multicontainer units. Description of the Figures

Figure 1. Analysis of purified recombinant Tectonic family proteins with specific antibodies. Panel A: Comassie staining of recombinant purified His-tag TCTN2 fusion proteins after separation by SDS-PAGE; Panel B: WB on the purified recombinant TCTN2, TCTNl or TCTN3 proteins stained with anti-TCTN2 antibodies. Arrows mark the protein bands of the expected size. Molecular weight markers are reported on the left of each panel.

Figure 2. Examples of immuno-histochemistry analysis of tumor (left panels) and normal tissue samples (right panels) stained with anti-TCTN2 antibodies. In the case of the ovarian cancer, the normal tissue surrounds the tumor. The antibody- stains specifically tumor cells (in dark grey); negative or poor staining is visible in normal cells.

Figure 3. TCTN2 expression in transiently transfected HeLa cells. Panel A) Western blot analysis of TCTN2 expression in total protein extracts from HeLa cells transfected with the empty vector pcDNA3 (lane 1) or with the plasmid encoding the TCTN2 gene (lane 2) stained with anti-TCTN2 antibody. Arrow marks the expected TCTN2 band. Molecular weight markers are reported on the left. Panel B) Flow cytometry analysis of TCTN2 surface localization in HeLa cells transfected with the empty vector pcDNA3 (dashed peak) or with the plasmid construct encoding the TCTN2 gene (solid peak). X axys, Fluorescence scale; Y axys, Cells (expressed as % relatively to major peaks).

Figure 4. TCTNl, TCTN2 and TCTN3 confer malignant cell phenotypes Panel A. Silencing of either TCTNl or TCTN3 impairs the clonogenic phenotype of the cell line HCT 15 grown on soft agar. The graphs show the reduction of the size (upper panels) and number (lower panels) of colonies formed by HCT 15 cells upon transfection with either TCTNl - siRNA (3) or TCTN3 - siRNAs (4) compared to cells transfected with a scrambled siRNA (2) or untreated cells (1). A picture of the colonies formed under each condition are reported on the left of the graphs. Panel B. Silencing of either TCTNl or TCTN3 reduces the invasive phenotype of the HCT 15 colon cell line. The graph reports the effect of the siRNA mediated inhibition of TCTNl or TCTN3 expression on the migration activity of the HCT- 15 colon tumor cell line, measured with the Boyden assay. As controls, untransfected cells or cells transfected with a scrambled siRNA were used. Small boxes under the columns show the visual counting of the migrated cells. Panel C. TCTN2 over- expression increases the cell clonogenic phenotype. The graph reports the effect of the TCTN2 over-expression on the number of colonies formed by the HCT 15 cell line on soft agar upon transfection with the TCTN2-encoding plasmid (3). Cells either untreated (1) or transfected with the empty plasmid pcDNA3 (2) were used as controls. Images of the cell colonies formed under each tested condition are reported below each histogram.

Figure 5. Analysis of purified HIGD2A recombinant protein expressed in E.coli. Left panel: Comassie staining of purified His-tag HIGD2A fusion protein espressed in E. coli separated by SDS-PAGE; Right panel: WB on the purified recombinant HIGD2A protein stained with anti-HIGD2A antibody. Arrow marks the protein band of the expected size. Molecular weight markers are reported on the left.

Figure 6. Examples of immuno-histochemistry analysis of tumor (left panels) and normal tissue samples (right panels) stained with anti-HID2A antibodies. In the case of the ovarian cancer, the normal tissue surrounds the tumor. The antibody-stains specifically tumor cells (in dark grey); negative or poor staining is visible in normal cells.

Figure 7. HIGD2A expression in transiently transfected HeLa cells. Western blot analysis of HIGD2A expression in total protein extracts from HeLa cells transfected with the empty vector pcDNA3 (lane 1) or with the plasmid construct encoding the HIGD2A gene (lane 2) stained with anti- HIGD2A (right panel) and anti-HIGD2B (left panel) antibodies. Arrow marks the expected HIGD2A band. Molecular weight markers are reported on both sides.

Figure 8. HIGD2A expression in tumor cell lines by Western blot. Total protein extracts from the human colon carcinoma tumor cell lines Colo205

(lane 1) and HCT- 15 (lane 2), the ovarian carcinoma OVCAR-3 (lane 3) and the breast tumor cell line MDA-MB231 (lane 4) were separated by SDS-PAGE, transferred onto nitrocellulose membranes and probed with anti-HIGD2A antibodies. Arrow marks the expected HIGD2A band. Molecular weight markers are reported on the left

Figure 9. Confirmation of HIGD2A antibody specificity upon gene specific silencing. At different time points, total protein extracts (corresponding to IxIO ⁶ cells) from the breast tumor cell line MDA-MB231 either untreated (left panel) or transfected with a HIGD2A-siRNA (right panel) were separated by SDS-PAGE, transferred onto nitrocellulose membranes and probed with anti-HIGD2A antibodies. As normalization control, membranes were also probed with an anti-actin antibody.

Figure 10. Analysis of purified C4orf32 recombinant protein expressed in E.coli. Left panel: Comassie staining of purified His-tag C4orf32 fusion protein espressed in E. coli separated by SDS-PAGE; Right panel: WB on the purified recombinant C4orf32 protein stained with anti-C4orf32 antibody. Arrow marks the protein band of the expected size. Molecular weight markers are reported on the left.

Figure 11. Examples of immuno-histochemistry analysis of tumor (left panels) and normal tissue samples (right panels) stained with anti-C4orf32 antibodies. The antibody-stains specifically tumor cells (in dark grey); negative or poor staining is visible in normal cells. Figure 12. C4orf32 expression in transiently transfected HeLa cells.

Western blot analysis of C4orf32 expression in total protein extracts from HeLa cells (corresponding to 1x10 ⁶ cells) transfected with the empty vector pcDNA3 (lane 1) or with the plasmid encoding C4orf32 either in the untagged native (lane 2) or the V5-fusion forms (lane 3) stained with anti-C4orf32 antibodies. A solid arrow marks the expected C4orf32 band in cells expressing the V5-fusion form (lane 3). A dashed arrow indicates the high molecular weight band in cells expressing native C4orf32 (lane 2), likely corresponding to an aggregated form of the protein. Molecular weight markers are reported on the left.

Figure 13. Detection of C4orf32 in breast tumor tissue homogenates. Examples of tumor (lanes 3, 4) and normal samples (lanes 1, 2) stained with anti-C4orf32 antibodies. Molecular weight markers are reported on the left.

Figure 14. Analysis of purified recombinant protein FAM62A expressed in E.coli. Left panel: Comassie staining of the purified - FAM62A protein fused to GST espressed in E. coli and separated by SDS-PAGE; Right panel: WB on the purified recombinant FAM62A protein stained with anti- FAM62A antibody. Arrow marks the protein band of the expected size. Low molecular weight bands visible on the gel correspond to degradation products of the FAM62A fusion band, as determined by mass spectrometry analysis. Molecular weight markers are reported on the left.

Figure 15. Examples of immuno-histochemistry analysis of tumor (left panels) and normal tissue samples (right panels) stained with anti-FAM62A antibodies. The antibody-stains specifically tumor cells (in dark grey); negative or poor staining is visible in normal cells. In the case of the ovarian cancer, the normal tissue surrounds the tumor.

Figure 16. FAM62A expression in transiently transfected HeLa cells. Western blot analysis of FAM62A expression in total protein extracts from HeLa cells transfected with the empty vector pcDNA3 (lane 1) or with the plasmid construct encoding the FAM62A gene (lane 2) stained with anti- FAM62A antibody. Arrow marks the expected FAM62A band. Different protein species are visible on the transfected cell extract, likely corresponding to FAM62A degradation products. Molecular weight markers are reported on the left. Figure 17. FAM62A expression in tumor cell lines.

Panel A) Western blot analysis. Total protein extracts from the human lung tumor cell line H-226 (lane 1), the ovarian carcinoma OVCAR-3, (Iane2), the breast tumor cell lines T47D and MCF7 (lanes 3, 4) and the colon carcinoma cell line HCT- 15 (lane 5), were separated by SDS-PAGE, transferred onto nitrocellulose membranes and probed with anti-FAM62A antibodies. Arrow marks the expected FAM62A band. Molecular weight markers are reported on the left. Panel B) Localization analysis. Confocal microscopy analysis of the HCTl 5 cell line stained with anti-FAM62A antibodies and DAPI to visualize the nuclei. The FAM62A specific staining accumulates at the plasma membrane.

Figure 18. Confirmation of FAM62A antibody specificity upon gene specific silencing. At different time point, total protein extracts (corresponding to 1x10 ⁶ cells) from the breast tumor cell line MCF-7 untreated (left panel), or transfected with a FAM62A-specific siRNA (right panel), were separated by SDS-PAGE, transferred onto nitrocellulose membranes and probed with anti- actin (normalization control) or anti-FAM62A antibodies.

Figure 19. Detection of FAM62A in breast tumor tissue homogenates. Examples of tumor (lanes 3, 4) and normal samples (lanes 1, 2) stained with anti-FAM62A antibodies. Molecular weight markers are reported on the left. Arrow marks the FAM62 A band of expected size.

Figure 20. FAM62A confers an invasive phenotype to breast tumor cell lines. The graphs report the effect of the siRNA-mediated inhibition of FAM62A expression on the migration activity of the MCF7 (upper graph) and the MDA-MB231 (lower graph) breast tumor cell lines, measured with the Boyden assay. As controls, cells either untreated or transfected with a scrambled siRNA were used. Small boxes under the columns show the visual counting of the migrated cells. The following examples further illustrate the invention.

EXAMPLES

Example 1. Generation of recombinant human protein antigens and antibodies to identify tumor markers

Methods The entire coding region or suitable fragments of the genes encoding the target proteins, were designed for cloning and expression using bioinformatic tools with the human genome sequence as template (Lindskog M et al (2005). Where present, the leader sequence for secretion was replaced with the ATG codon to drive the expression of the recombinant proteins in the cytoplasm of E. coli. For cloning, genes were PCR-amplified from cDNA derived from Mammalian Gene Collection (http ://mgc.nci.nih. gov/) clones using specific primers so as to fuse a 6 histidine tag sequence at the 3' end, annealed to in house developed vectors, derivatives of vector pSP73 (Promega) or pGEXβPI (GE Healthcare) adapted for the T4 ligation independent cloning method (Nucleic Acids Res. 1990 October 25; 18(20): 6069-6074) and used to transform E. coli NovaBlue cells recipient strain. E. coli tranformants were plated onto selective LB plates containing 100 μg/ml ampicillin (LB Amp) and positive E. coli clones were identified by restriction enzyme analysis of purified plasmid followed by DNA sequence analysis. For expression, plasmids were used to transform BL21-(DE3) E. coli cells and BL21-(DE3) E. coli cells harbouring the plasmid were inoculated in ZYP-5052 growth medium (Studier, 2005) and grown at 37°C for 24 hours. Afterwards, bacteria were collected by centrifugation, lysed into B-Per Reagent containing 1 mM MgC12, 100 units DNAse I (Sigma), and 1 mg/ml lysozime (Sigma). After 30 min at room temperature under gentle shaking, the lysate was clarified by centrifugation at 30.000 g for 40 min at 4°C. With the exception of Fam62, all proteins were purified from the inclusion bodies by resuspending the pellet coming from lysate centrifugation in 40 mM TRIS-HCl, 1 mM TCEP (Tris(2-carboxyethyl)- phosphine hydrochloride, Pierce} and 6M guanidine hydrochloride, pH 8 and performing an IMAC in denaturing conditions. Briefly, the resuspended material was clarified by centrifugation at 30.000 g for 30 min and the supernatant was loaded on 0.5 ml columns of Ni-activated Chelating Sepharose Fast Flow (Pharmacia). The column was washed with 50 mM TRIS-HCl buffer, 1 mM TCEP, 6M urea, 60 mM imidazole, 0.5M NaCl, pH 8. Recombinant proteins were eluted with the same buffer containing 500 mM imidazole. Fam62A protein was purified as soluble GST-fusion by subjecting the B-PER soluble lysate to glutathione affinity purification using 0.5 ml mini-columns of Glutathione-Sepharose 4B resin (GE-Healthcare) equilibrated with 10 ml PBS, pH 7.4. After column washing with equilibrium buffer the proteins were eluted with 50 mM TRIS buffer, 10 mM reduced glutathione, pH δ.O.Proteins were analysed by SD S -Page and their concentration was determined by Bradford assay using the BIORAD reagent (BIORAD) with a bovine serum albumin standard according to the manufacturer's recommendations.

The identity of recombinant affinity purified proteins was further confirmed by tandem mass spectrometry (MS/MS), using standard procedures. This analysis also confirmed that lower mass protein species sometimes visible on the gels corresponded to truncated forms of the proteins.

To generate antisera, the purified proteins were used to immunize CDl mice (6 week-old females, Charles River laboratories, 5 mice per group) intraperitoneally, with 3 protein doses of 20 micrograms each, at 2 week- interval. Freund's complete adjuvant was used for the first immunization, while Freund's incomplete adjuvant was used for the two booster doses. Two weeks after the last immunization animals were bleeded and sera collected from each animal was pooled. Results Gene fragments of the expected size were successfully isolated by PCR from specific clones of the Mammalian Gene Collection using primers specific for each gene. In particular, for the TCTN2 gene, a fragment corresponding to nucleotides 637-1458 of the transcript (SEQ ID 28) of and encoding an amino acid region from 171 to 444 (SEQ ID 9) was obtained. For the TCTNl gene, a fragment corresponding to nucleotides 134 to 1789 of the transcript (SEQ ID NO 23, corresponding to ENST00000397655) of and encoding an amino acid region from 22 to 573 (SEQ ID NO 5, corresponding to ENSP00000380775) was obtained. For the TCTN3 gene, a fragment corresponding to nucleotides 311to 2065 of the transcript (SEQ ID NO 32, corresponding to ENST00000371217) of and encoding an amino acid region from 23 to 607 (SEQ ID NO 13, corresponding to ENSP00000360261) was obtained.

For the HIGD2A gene, a fragment corresponding to nucleotides 49-366 of the transcript (SEQ ID 9) of and encoding an amino acid region from 1 to 106 (SEQ ID 2) was obtained. For HIGD2B gene, a fragment corresponding to nucleotides 525-852 of the transcript (SEQ ID 10) of and encoding an amino acid region from 1 to 106 (SEQ ID 3) was obtained.

For the C4orf32, a fragment corresponding to nucleotides 60-374 of the transcript (SEQ ID 11) of and encoding an amino acid region from 1 to 105 (SEQ ID 4) was obtained.

For the FAM62A, a fragment corresponding to nucleotides 53-257 of the transcript SEQ IDs 12) and encoding an amino acid region from 1 to 68 (SEQ IDs 5) was obtained. This fragment is identical in all available FAM62A isoforms (protein SEQ IDs 5, 6, 7; Transcript SEQ IDs: 12, 13, 14).

A clone encoding the correct amino acid sequence was identified for each gene/gene fragment and, upon expression in E. coli, a protein of the correct size was produced and subsequently purified using affinity chromatography (Figures 1, 5, 10, 14, left panel). In the case of TCTN3, different truncated forms of the protein were obtained after purification, among which a protein product of approximately 38 KDa was the major form recognized by the antibodies (Figure IB). Antibodies generated by immunization specifically recognized their target proteins in Western blot (WB) (Figures 1, 5, 10, 14 right panel). Moreover, antibodies raised against TCTN2 - also recognized TCTNl and TCTN3. Similarly, antibodies raised against HIGD2A and HIGD2B were able to recognize HIGD2A in the assay.

Example 2. Tissue profiling by immune-hystochemistry

Methods

The analysis of the antibodies capability to recognize their target proteins in tumor samples was carried out by Tissue Micro Array (TMA), a miniaturized immuno-histochemistry technology suitable for HTP analysis that allows to analyse the antibody immuno-reactivity simultaneously on different tissue samples immobilized on a microscope slide. A tissue microarray was prepared containing 100 formalin-fixed paraffin- embedded cores of human tissues from patients affected by colorectal cancer, ovarian cancer, breast cancer, lung cancer, prostate cancer and corresponding normal tissues and analyzed using the specific antibody sample. Briefly, each TMA slide included tumor tissue samples representative of different well pedigreed patients, representing the 5 cancer types, and an equal number of normal tissue samples from the same patients as controls. In total, the TMA design consisted in 10 tumor samples per each tumor class and 10 normal tissue from 5 well pedigreed patients (equal to two tumor samples and 2 normal tissues from each patient) to identify promising target molecules differentially expressed in cancer and normal tissues. The direct comparison between tumor and normal tissues of each patient allowed the identification of antibodies that stain tumor cells and provide indication of target expression in the tumor under investigation. All formalin fixed, paraffin embedded tissues used as donor blocks for

TMA production were selected from the archives at the IEO (European Institue of Oncology, Milan). Corresponding whole tissue sections were examined to confirm diagnosis and tumour classification, and to select representative areas in donor blocks. Normal tissues were defined as microscopically normal (non- neoplastic) and were generally selected from specimens collected from the vicinity of surgically removed tumors. The TMA production was performed essentially as previously described (7, 8). Briefly, a hole was made in the recipient TMA block. A cylindrical core tissue sample (1 mm in diameter) from the donor block was acquired and deposited in the recipient TMA block. This was repeated in an automated tissue arrayer "Galileo TMA CK 3500" (BioRep - Milan) until a complete TMA design was produced. TMA recipient blocks were baked at 42 <0>C for 2 h prior to sectioning. The TMA blocks were sectioned with 2-3 μm thicknes using a waterfall microtome (Leica), and placed onto poli-L-lysinated glass slides for immunohistochemical analysis. Automated immunohistochemistry was performed as previously described (Kampf C. et al. 2004 Clin. Proteomics 1 : 285-300). In brief, the glass slides were incubated for 30' min in 60 ⁰ C, de-paraffinized in xylene (2 x 15 min) using the Bio-Clear solution (Midway. Scientific, Melbourne, Australia), and re-hydrated in graded alcohols. For antigen retrieval, slides were immersed 0.01 M Na-citrate buffer, pH 6.0 at 99°C for 30 min Slides were placed in the Autostainer (R) (DakoCytomation) and endogenous peroxidase was initially blocked with 3% H2O2, for 5 min. Slides were then blocked in Dako Cytomation Wash Buffer containing 5% Bovine serum albumin (BSA) and subsequently incubated with mouse antibodies for 30' (dilution 1 :200 in Dako Real™ dilution buffer). After washing with DakoCytomation wash buffer, slides were incubated with the goat anti-mouse peroxidase conjugated Envision(R) for 30 min each at room temperature (DakoCytomation). Finally, diaminobenzidine (DakoCytomation) was used as chromogen and Harris hematoxylin (Sigma- Aldrich) was used for counterstaining. The slides were mounted with Pertex(R) (Histolab).

The staining results have been evaluated by a trained pathologist at the light microscope, and scored according to both the percentage of immunostained cells and the intensity of staining. The individual values and the combined score (from 0 to 300) were recorded in a custom-tailored database. Digital images of the immunocytochemical findings have been taken at a Leica DM LB light microscope, equipped with a Leica DFC289 color camera. Results A TMA design was obtained, representing tumor tissue samples from 5 tumor classes (lung, ovary, prostate, breast and colon) and normal tissues, derived from 5 patients for each tumor type. The results from tissue profiling showed that the antibodies specific for the recombinant proteins (see Example 2) are strongly immunoreactive on several cancer tissues, indicating the presence of the target proteins in tumors tissues, while no or poor reactivity was detected in normal tissues. Based on this finding, the detection of target proteins in tissue samples can be associated with the specific tumor/s.

The capability of target-specific antibodies to stain different tumor tissues is summarized in Table 1. Representative examples of microscopic enlargements of tissue samples stained by each antibody are reported within Figures 2, 6, 11, 15.

Example 3

Confirmation of the marker association with the tumor/s by expanded TMA analysis Method

The association of each protein with the indicated tumors was further confirmed on a larger collection of clinical samples. To this aim, a tissue microarray was prepared for each of the five tumor classes containing 100 formalin-fixed paraffin-embedded cores of human tissues from 50 patients

(equal to two tissue samples from each patient). The TMAs were stained with the marker specific antibodies, using the previously reported procedure. The staining results were evaluated, as above described, by a trained pathologist at the light microscope.

Results

Five TMA designs were obtained, for each of the five tumors, representing tissue samples from 50 patients. The results from tissue analysis showed that the antibodies specific for each of the four proteins (see Example 1) are strongly immune-reactive on a large percentage of tumor tissues, indicating that the corresponding proteins are selectively detected in the tumor/s. This finding confirms a strong association of the markers with the specific tumor/s.

The capability of marker-specific antibodies to stain different tumor tissues is summarized in Table 2, which reports the percentage of positive tumor tissue samples derived from the 50 patients

Example 4. Expression and localization of the target proteins in transfected mammalian cells Methods

The specificity of the antibodies for each target proteins was assessed by Western blot analysis on total protein extracts from eukaryotic cells transiently transfected with plasmid constructs containing the complete sequences of the genes encoding the target proteins.

To this aim, cDNA were generated from pools of total RNA derived from human testis, human placenta, human bone marrow, human fetal brain, in reverse transcription reactions and the entire coding regions were PCR-amplified with specific primers pairs. PCR products were cloned into plasmid pcDNA3 (Invitrogen). In the case of C4Orf32, the PCR product was also cloned into plasmid pcDNA3.1D (Invitrogen) to express a tagged form of the protein fused to a V5 epitope sequence at the carboxy-terminus. HeLa cells were grown in DMEM- 10% FCS supplemented with 1 mM Glutamine were transiently transfected with preparation of the resulting plasmids and with the empty vector as negative control using the Lipofectamine-2000 transfection reagent (Invitrogen). After 48 hours, cells were collected, lysed with PBS buffer containing 1% Triton XlOO and expression of target proteins was assessed by Western blot analysis on total cell extracts (corresponding to IxIO ⁶ cells) using antibodies specific for TCTN2, C4orf32, FAM62A, HIGD2A, and its homologous HIGD2B protein. When the C4orf32 tagged construct was used for transfection, blots were analysed with a tag-specific antibody (anti-V5 antibody). Western blot was performed by separation of the protein extracts on pre-cast SDS-PAGE gradient gels (NuPage 4-12% Bis-Tris gel, Invitrogen) under reducing conditions, followed by electro-transfer to nitrocellulose membranes (Invitrogen) according to the manufacturer's recommendations. The membranes were blocked in blocking buffer composed of Ix PBS-0.1% Tween 20 (PBST) added with 10% dry milk, for 1 h at room temperature, incubated with the antibody diluted 1 :2500 in blocking buffer containing 1% dry milk and washed in PBST- 1%. The secondary HRP- conjugated antibody (goat anti- mouse immunoglobulin/HRP, Perkin Elmer) was diluted 1 :5000 in blocking buffer and chemiluminescence detection was carried out using a Chemidoc-IT UVP CCD camera (UVP) and the Western lightning ¹ ™ cheminulescence Reagent Plus (Perkin Elmer), according to the manufacturer's protocol.

Surface localization of target proteins was assessed in HeLa transfected cells by cell surface staining and Flow Cytometry (FACS) analysis. HeLa cells transfected with each construct or with the empty vector (2x10 ⁴ per well) were pelletted in 96 U-bottom microplates by centrifugation at 200 x g for 5 min at 4°C and incubated for 1 hour at 4°C with the appropriate dilutions of marker- specific antibodies. The cells were washed twice in PBS-5% FCS and incubated for 20 min with the appropriate dilution of R-Phycoerythrin (PE)-conjugated secondary antibodies (Jackson Immυno Research, PA, USA) at 4°C. After washing, bacteria were analysed by a FACS Canto II flow cytometer (Becton Dickinson). Data were analyzed with FlowJo 8.3.3 program.

Results

To confirm the antibody specificity, the complete coding sequence/s for each target protein were cloned in a eukaryotic expression vector and used for transient transfection of HeLa cells.

Expression of each protein was detected by Western blot in total protein extracts from HeLa cells transfected with the different constructs encoding for the target proteins using their specific antibodies. As far as TCTN2, HIGD2A and FAM62A are concerned, a band of the expected size was visible in HeLa cells transfected with the corresponding plasmids while the same band was either not visible or very faintly detected in HeLa cells transfected with the empty pcDNA3 plasmid. Each antibody recognized specifically its target protein, since almost a unique single protein band was detected. Results are reported in Figures 3 A, 7 and 16.

HeLa cells transfected with the HIGD2A construct were also tested with the antibody specific for the HIGD2B homolog. Results show that the anti-HIGD2B antibody was capable of detecting HIGD2A protein in transfected cells. (Figure 7). As regards C4orf32, HeLa were transfected with the C4orf32 untagged and tagged constructs and tested with the C4orf32- and the anti-V5 antibodies. A protein of expected size was detected in HeLa cells transfected with the tagged construct using both antibodies. A protein band of higher mass was detected in cells transfected with the untagged construct using the C4orf32-specific antibody. This indicates that the native C4orf32 protein forms highly stable aggregates that are detected by the antibody in immunoblot. The presence of the fusion tag at the C4orf32 carboxy-terminus appears to prevent the C4orf32 aggregation and allows the detection of the protein species with the expected mass. Results of immunoblot analysed with the C4orf32antibody is reported in Figure 12. Data obtained with the anti-V5 antibody are not shown.

Surface localization of target proteins was addressed by FACS analysis of transiently transfected cells stained with the specific antibodies. Data are reported for cells transfected with the construct encoding TCTN2. In this experiment the TCTN2- antibody was capable of binding the surface of transfected cells, while no binding was observed on cells transfected with the empty pcDNA3 vector (Figure 3B). This indicates that this target protein is localized on the cell surface and are accessible to the external environment. This finding reinforces the relevance of TCTN2 protein for future development as either diagnostic marker or therapeutic targets.

Example 5. Expression of target proteins in tumor cell lines Expression of target proteins showing positivity by TMA was also assessed by WB on total extracts from a panel of human epithelial cell lines derived from the same tumor types. In each analysis, cells were cultured in under ATCC recommended conditions, and sub-confluent cell mono-layers were detached with PBS-0.5 mM EDTA and lysed by several freeze-thaw passages in PBS-1% Triton. Total protein extracts were loaded on SDS-PAGE (2x10 ⁵ cells/lane), and subjected to WB with specific antibodies as described above. The marker cellular localization was assessed by confocal microscopy analysis. Cells were plated on glass cover slips and after 48 h were washed with PBS and fixed with 3% formaldheyde solution in PBS for 20 min at RT. Then, after extensive washing in PBS, the cells were permeabilized with 0.01% BriJ96® (Fluka), and incubated overnight at 4 ⁰ C with polyclonal antibodies (1:200). Cells were then stained with Alexafluor 488-labeled goat anti-mouse antibodies (Molecular Probes). DAPI (Molecular Probes) was used to visualize nuclei. The cells were mounted with glycerol plastine and observed under a laser-scanning confocal microscope (LeicaSP5).

Results

Expression analysis of the target proteins was confirmed in tumor cell lines. Example data are shown for HIGD2A and FAM62A. In particular, HIGD2A expression is reported for a panel of the tumor cell lines including Ovcar 3 (ovary adenocarcinoma), and HCT- 15 and Colo205 (colon tumor cell lines) and MDA-MB231 (breast adenocarcinoma). FAM62A expression analysis is reported for the tumor cell lines H226, Ovcar 3, MCF7 and T47D (breast adenocarcinoma).

HIGD2A and FAM62A protein bands of the expected sizes were detected in total protein extracts of all tested tumor cell lines by immunoblot, confirming their expression in tumor cells derived from the different tumor types.

Results are reported in Figures 8, 17A.

Confocal microscopy analysis of the HCT- 15 cell with anti-FAM62A antibody shows that the protein is localized at the plasma membrane (Figure 17B). This suggests that the protein is accessible to the external environment and reinforces its relevance for future development as either diagnostic marker or therapeutic targets.

Example 6. Confirmation of the specificity the tumor-reactive antibodies by gene silencing experiments Methods

The specificity of the polyclonal antibodies for their targets was also confirmed by transient RNA-interference experiments, measuring the loss of detection of the expected protein bands in cell lines upon silencing. For each gene, a set of small interfering RNAs (siRNAs) and controls were obtained from QIAGEN, whose target sequence is reported in Table 3. Table 3

The expression of marker genes was knocked down in a panel of epithelial tumor cell lines using marker-specific siRNAs with the HiPerfect transfection reagent (QIAGEN) following the manufacturer's protocol. As control, cells treated with irrelevant siRNA (scrambled siRNA) were analysed in parallel. At different time points (ranging from 24 to 72 hours) post transfection, we first assessed the reduction of gene transcription by quantitative RT-PCR (QRT-PCR) on total RNA, by evaluating the relative marker transcript level, using the beta-actin, GAPDH or MAPK genes as internal normalization control. Afterwards, the loss of protein expression was also confirmed by immunoblot on total protein extracts prepared from the siRNA-treated cell lines, using the same antibodies giving positive immune- staining on tumor tissues. Blots were also probed with an anti-actin antibody as internal normalization control. Results

Gene silencing experiments with marker- specific siRNA reduced the marker transcripts (approximately 30-40 fold reduction), as determined by Q- RT-PCR. Under this condition, immunoblot analysis with marker- specific antibodies revealed that expression of protein bands of expected size were clearly detected in untreated cell lines while they were significantly reduced upon si-RNA treatment. Examples of this analysis are reported for HIGD2A and FAM62A in Figures 9 and 18, respectively. Examples 7. The tumor-reactive antibodies are able to recognize proteins of expected size in tumor tissue homogenates

The presence of protein bands corresponding to the marker proteins was also investigated in tissue homogenates of tumor biopsies from patients affected by tumor (selected from the 5 tumor classes). In these assay, tumor and normal tissues from the same patients were analysed in parallel. Homogenates were prepared by mechanic tissue disruption in buffer containing 40 mM TRIS-HCl, 1 mM TCEP {Tris(2-carboxyethyl)-phosphine hydrochloride, Pierce} and 6M guanidine hydrochloride, pH 8. Western blot was performed by separation of the total protein extracts (20 μg/lane) proteins were detected by specific antibodies. Results

All tested antibodies specifically recognized protein species on tumor tissues, while the same bands were not or faintly visible on normal tissues. Example data are represented for proteins FAM62A and C4orf32 on breast tissue homogenates. As shown in Figure 19, in the case of FAM62A a band of expected size was detected in breast tumor tissues, while no bands were detected in normal tissue homogenates, confirming the presence of the marker proteins in the tumor. As far as C4orf32 is concerned, different protein species with higher mass (50-60 KDa) were detected by the antibody specifically in tumor samples (Figure 13). This is in line with previous evidences indicating the tendency of the C4orf32 protein to form stable aggregates (see Example 4).

Example 8. The marker proteins are involved in cell malignant phenotypes To verify that the proteins included in the present invention can be exploited as targets for therapeutic applications, the effect of alteration of marker expression, either depletion or over-expression (obtained upon transfection with specific siRNAs or expression plasmids, respectively), was evaluated in in vitro studies generally used to define the role of newly discovered proteins in tumorigenesis or tumor progression. Marker-specific knock-down or transfected tumor cell lines and their respective controls were analysed for their migration property and the ability to proliferate in an anchorage-independent fashion using the Boyden in vitro invasion and the soft agar assays, respectively. A brief description of these assays is provided below. The Boyden chamber assay is based on a chamber of two medium-filled compartments separated by a microporous membrane. Cells are placed in the upper compartment and are allowed to migrate through the pores of the membrane into the lower compartment, in which chemotactic agents are present. After an appropriate incubation time, the membrane between the two compartments is fixed and stained, and the number of cells that have migrated to the lower side of the membrane is determined. Therefore, the Boyden chamber-based cell migration assay has also been called filter membrane migration assay, trans-well migration assay, or chemotaxis assay. The Soft Agar Assay for Colony Formation is an anchorage independent growth assay in soft agar, which is considered the most stringent assay for detecting malignant transformation of cells. Many primary cell lines must attach to a solid surface before they can divide and proliferate, while they fail to grow when suspended in a viscous fluid or gel (e.g. agar or agarose). However, when these cell lines are transformed, they are able to grow in a viscous fluid or gel and become anchorage-independent. The process, by which these phenotypic changes occur, is assumed to be closely related to the process of in vivo carcinogenesis. Thereby, the acquisition of an anchorage independent growth style on soft agar is viewed as indicative of in vivo carcinogenesis. For this assay, cells are cultured with appropriate controls in soft agar medium for 21-28 days. Following this incubation period, formed colonies can either be analyzed morphologically using cell stain and quantifying the number of colonies formed per well. Method When the effect of marker depletion was analysed, tumor cell lines previously shown to express the proteins were treated with any of the marker specific siRNA molecules proved to inhibit marker expression (see Table 3 in the Example 6) and then tested in the Boyden invasion and the anchorage independent soft agar assays, as compared to control cell lines treated with a scramble siRNA. When the effect of marker overexpression was assessed, tumor cell lines showing a low endogenous marker expression were transfected with corresponding pcDN A3 -derived plasmids and then tested in the Boyden and the soft agar assays.

For the Boyden in vitro invasion assay, a transwell system, equipped with 8-μm pore polyvinylpirrolidone-free polycarbonate filters, was used. The upper sides of the porous polycarbonate filters were coated with 50 μg/cm ² of reconstituted Matrigel basement membrane and placed into six-well culture dishes containing complete growth medium. Cells (1x10 cells/well) were loaded into the upper compartment in serum- free growth medium. After 16 h of incubation at 37°C, non invading cells were removed mechanically using cotton swabs, and the microporous membrane was stained with Diff-Quick solution. Chemotaxis was evaluated by counting the cells migrated to the lower surface of the polycarbonate filters (six randomly chosen fields, mean ± SD).

For the anchorage independent soft agar assay, cells were suspended in growth medium containing 10% FBS and 0.3% agar, seeded onto a solidified base of growth medium containing 0.6% agar, and overlaid with growth medium. Cell growth was monitored for 2 weeks by scoring the formation of colonies (all values were determined in triplicate). Photographs were taken with a phase-contrast microscope.

Results

Examples of the phenotypic changes induced by reducing the marker expression with specific siRNA are shown for FAM62A, TCTNl and TCTN3. Inhibition of either TCTNl or TCTN3 expression reduces the capability of the

HCTl 5 tumor cell line to grow on soft agar, as shown by the lower number and size of colonies formed by cells treated with TCTNl- and TCTN3-siRNAs

(Figure 4A). Inhibition of TCTNl and TCTN3 also impaires the invasive phenotype of the HCTl 5 cell line (Figure 4B). Similarly, inhibition of FAM62A expression significantly impairs the invasive phenotype of the MCF7 and MDA-MB231 tumor cell lines (Figure 20).

Examples of the phenotypic changes induced by an increased marker expression in cell transfected with the marker- encoding plasmids are reported for TCTN2. TCTN2 over-expression increases the capability of the HCTl 5 cell line to proliferate on soft agar, as shown by the higher number of colonies formed by TCTN2-transfected cells (Figure 4C).

Altogether, the results indicate that the proteins are involved in tumorigenesis and/or tumor progression. References

1. Adams G.P. and Weiner L.M. (2005) Monoclonal antibody therapy cancer. Nat Biotechnol 23:1147-57; 2. Anderson, L., and Seilhamer, J. (1997). A comparison of selected mRNA and protein abundances in human liver. Electrophoresis 18, 533 - 537;

3. Chen, G., Gharib, T. G., Wang, H., Huang, C. C, Kuick, R., Thomas, D. G., Shedden, K. A., Misek, D. E., Taylor, J. M., Giordano, T. J., Kardia, S. L., Iannettoni, M. D., Yee, J., Hogg, P. J., Orringer, M. B., Hanash, S. M., and Beer, D. G. (2003) Protein profiles associated with survival in lung adenocarcinoma. Proc. Natl. Acad. Sci. U. S. A 100, 13537 - 13542;

4. Ginestier, C, Charafe-Jauffret, E., Bertucci, F., Eisinger, F., Geneix, J., Bechlian, D., Conte, N., Adelaide, J., Toiron, Y., Nguyen, C, Viens, P., Mozziconacci, M. J., Houlgatte, R., Birnbaum, D., and Jacquemier, J. (2002) Distinct and complementary information provided by use of tissue and DNA microarrays in the study of breast tumor markers. Am. J. Pathol. 161, 1223 - 1233;

5. Gygi, S. P., Rochon, Y., Franza, B. R., and Aebersold, R. (1999) Correlation between protein and mRNA abundance in yeast. MoI. Cell. Biol. 19, 1720 - 1730; Nishizuka, S., Charboneau, L., Young, L., Major, S., Reinhold, W. C, Waltham, M., Kouros-Mehr, H., Bussey, K. J., Lee, J. K., Espina, V., Munson, P. J., Petricoin, E., Ill, Liotta, L. A., and Weinstein, J. N. (2003) Proteomic profiling of the NCI-60 cancer cell lines using new high-density reverse-phase lysate microarrays. Proc. Natl. Acad. Sci. U. S. A 100, 14229 - 14234; 6. Tyers, M., and Mann, M. (2003) From genomics to proteomics. Nature 422, 193 - 197;

7. Kononen J et al (1998) Nature Med. 4:844-847;

8. Kallioniemi OP et al (2001) Hum. MoI. Genet. 10:657-662.