WO2006044335A2

WO2006044335A2 - Therapy with clostridium perfringens enterotoxin to treat ovarian and uterine cancer

Info

Publication number: WO2006044335A2
Application number: PCT/US2005/036449
Authority: WO
Inventors: Alessandro D. Santin; Fabrizio Comper
Original assignee: University of Arkansas System; University of Arkansas at Little Rock
Current assignee: University of Arkansas System; University of Arkansas at Little Rock
Priority date: 2004-10-14
Filing date: 2005-10-12
Publication date: 2006-04-27
Anticipated expiration: 2007-04-14
Also published as: WO2006044335A3; US20060084594A1; US20150218644A1; US9702010B2; US8247371B2; US20090291047A1; US20120071421A1; US8664184B2

Abstract

The invention discloses high levels of receptors for Clostridium perfringens enterotoxin (CPE) have been found in ovarian cancer and uterine cancer tissue samples. In addition, successful in vivo treatment off a mouse model of ovarian cancer with intraperitoneal injection of CPE is disclosed. High levels of Ep-CAM protein is also disclosed in ovarian cancer tissue samples. Thus, the invention provides a method of treating ovarian cancer and uterine cancer by administering CPE. The invention also provides a method of treating cancer in a mammal involving intraperitoneal administration of CPE, where at least some cancerous cells are located in or adjacent to the peritoneal cavity of the mammal. The invention also provides a method of treating ovarian cancer involving administering an anti-Ep-CAM antibody. The invention also provides a method of treating cancers expressing claudin-3 or claudin-4 by administering an antibody against claudin-3 and/or an antibody against claudin-4. The invention also provides a method of protecting a mammal from CPE toxicity involving administering a non-toxic protective agent that binds to claudin-3 and/or claudin-4 and inhibits CPE binding to claudin-3 and/or claudin-4.

Description

THERAPY WITH CLOSTRIDIUM PERFRINGENS ENTEROTOXIN TO TREAT OVARIAN AND UTERINE CANCER.

Background of the Invention Ovarian carcinoma remains the cancer with the highest mortality rate among the gynecological malignancies, with 25,400 new cancer cases estimated in 2003 in the United States alone. Ovarian serous papillary carcinoma (OSPC) is the most common histologic type of ovarian carcinoma. Because of the insidious onset of the disease and the lack of reliable screening tests, two-thirds of patients have advanced disease when diagnosed; and although many patients with disseminated tumors respond initially to standard combinations of surgical and cytological therapy, nearly 90% will develop recurrence and succumb to their disease.

Uterine cancer is the most prevalent gynecological tumor in women, with an estimated 40,100 cases and 6,800 deaths in the United States in 2003. On the basis of clinical and histopathological variables, two subtypes of endometrial carcinoma, Type I and II tumors, have been described (1,2). Type I endometrial cancers, which account for about 80% of cases, are usually well differentiated and endometrioid in histology. They are diagnosed predominantly in younger women, and have a favorable prognosis. Type II endometrial cancers are poorly differentiated tumors, often with serous papillary or clear cell histology. Although type II tumors account for only a minority of endometrial carcinoma cases, about 50% of all relapses occur in this group.

Uterine serous papillary carcinoma (USPC) tumors represent the most aggressive variant of Type II endometrial cancer and may constitute up to 10% of endometrial tumors (3-10). The microscopic criteria for diagnosis of USPC were first outlined by Hendrickson in 1982 (10). Typically, the neoplastic epithelium is characterized by serous differentiation with psammoma bodies present and with predominantly papillary architecture (11). Pleomorphism, grade III nuclear atypia with prominent nucleoli and vesicular chromatin pattern, as well as a high mitotic activity are commonly detected in this tumor. Clinically, USPC has a propensity for intraabdominal and lymphatic spread even at presentation and is characterized by a highly aggressive biologic behavior (3-10, 12). USPC is a chemoresistant disease from onset, with responses to combined cisplatinum-based chemotherapy of about 20% and short duration (7-9). The survival rate is dismal, even when USPC is only a minor component of the histologically more common endometrioid adenocarcinoma, and widespread metastasis and high mortality may occur even in those cases in which tumor is confined to the endometrium or to an endometrial polyp (4, 6, 12). The overall 5 -year survival is about 30% for all stages and the recurrence rate after surgery is extremely high (50% to 80%).

Pancreatic cancer is the fifth leading cause of cancer death in the U.S. and has one of the highest mortality rates of any malignancy.

Chemotherapy resistance is a major problem in treatment of ovarian, uterine, and pancreatic cancers, and other cancers. Often a patient initially responds to chemotherapy, but the tumor develops resistance and recurs. New treatments for chemotherapy resistant tumors are needed.

Claudin-3 and claudin-4 are proteins found on the surface of cells of various tissue types involved in tight junctions connecting one cell to the adjacent cells (13). Claudins 3 and 4 have been found to be the receptors for Clostridium perfringens enterotoxin, a bacterial toxin that causes food poisoning (14-16).

Claudin-3 mRNA is expressed in normal prostrate, colon, small bowel, and pancreas (17). Claudin-4 is expressed at high levels in colon, and at moderate levels in prostate, placenta, lung, pancreas, and lower levels in small bowel, kidney, and uterus (17). Claudin-3 mRNA expression in prostate adenocarcinoma was found to be equal to or greater than expression in surrounding normal prostate tissue (17). The prostate cancer cells in tissue culture were sensitive to killing by Clostridium perfringens enterotoxin (CPE) (17).

In another report, claudin-4 mRNA was found to be not expressed or weakly expressed in normal pancreas, but was highly expressed in most primary pancreatic carcinoma samples (18). Claudin-4 expression was also found in colon cancer, breast cancer, and gastric cancer samples (18). CPE was reported to kill pancreatic cancer cell lines in vitro (18). Pancreatic tumor xenografts were induced in nude mice, and the tumors were directly injected with CPE over the course of 5 days. The CPE- treated tumors showed no increase in size and exhibited areas of necrosis, while the untreated tumors grew (18).

Hough et al. reported results of serial analysis of gene expression (S AGE) comparing expression of genes in ovarian surface epithelium, cystadenoma, and ovarian primary tumors (19). Many genes were found to be upregulated in ovarian tumors compared to normal ovarian epithelium. Several of these were surface proteins. Surface proteins shown to be upregulated included claudin-3 and -4, HE4, mucin-1, epithelial cellular adhesion molecule, and mesothelin (19). Claudin-3 and - 4 mRNA and protein levels were also reported to be elevated in ovarian carcinoma samples as compared to normal ovarian tissue and ovarian cystadenoma (20). New methods for treating cancer are needed. Methods for treating the most deadly cancers, including ovarian cancer, pancreatic cancer, and uterine serous papillary carcinoma, are particularly needed. Methods for treating chemotherapy resistant cancers are particularly needed. Preferably the methods would have reduced side effects and involve reduced toxicity for healthy non-target tissue.

Summary

The invention is based in part on the discovery that expression of claudin-3 and -4 is higher in metastatic ovarian cancer cells than cells from primary ovarian tumors, and higher in chemotherapy-resistant ovarian cancer cells than in chemotherapy-naive ovarian cancer cells. Thus Clostridium perfringens enterotoxin (CPE) is well suited for use in treating ovarian cancer patients whose tumors have spread or resisted chemotherapy.

It has also been discovered that uterine serous papillary carcinoma (USPC) cells express 12-times more claudin-4 and 8-times more claudin-3 than healthy uterine epithelial cells.

It is demonstrated herein that ovarian cancer can be treated successfully in vivo in a mouse model with administration of CPE. Thus, one aspect of the invention is treating uterine or ovarian cancer, particularly metastatic or chemotherapy resistant forms of uterine or ovarian cancer, in vivo in a mammal by administering CPE. One aspect of the invention is administering CPE intraperitoneally to treat tumors found in or adjacent to the peritoneal cavity. This reduces the toxicity of CPE administration because it spares tissues distant from the peritoneal cavity. That allows higher doses to be administered, increasing the efficacy against tumors in or near the peritoneal cavity. Another aspect of the invention involves administering to non-target tissues an agent that protects cells against CPE toxicity in conjunction with intraperitoneal administration of CPE. This further protects the non-target tissues, reducing the side effects of CPE and allowing the use of higher doses of CPE. One such protective agent is a carboxy terminal fragment of the CPE protein. Thus, one embodiment of the invention provides a method of treating cancer in a mammal involving: administering to the mammal a therapeutically effective amount of CPE or a pharmaceutically effective salt thereof; wherein the cancer is ovarian cancer or uterine cancer. Another embodiment of the invention provides a method of treating cancer in a mammal involving: intraperitoneally administering to the mammal a therapeutically effective amount of CPE or a pharmaceutically acceptable salt thereof. This method is suitable when at least some cancerous cells are located in or adjacent to the peritoneal cavity of the mammal, and the cells are sensitive to CPE. One embodiment of the invention provides a method of determining the sensitivity of a gynecological malignancy to CPE involving detecting the presence or absence of claudin-3 and/or claudin-4 in a tissue sample comprising a portion of the malignancy.

One embodiment of the invention provides a use of CPE to prepare a medicament effective to treat ovarian cancer or uterine cancer in a mammal.

One embodiment of the invention provides a use of CPE to prepare a medicament effective to treat cancer in a mammal, wherein the medicament is adapted for intraperitoneal administration.

It is also reported herein that the TROP-I /Ep-CAM gene and the Ep-CAM protein are overexpressed in ovarian serous papillary carcinoma (OSPC) compared to normal ovarian epithelium (NOVA). The TROP-I /Ep-CAM mRNA is expressed 39- fold higher in OSPC than in NOVA. Thus, antibodies against Ep-CAM can be an effective treatment for ovarian cancer.

Accordingly, one embodiment of the invention provides a method of treating cancer in a mammal involving administering to the mammal a therapeutically effective amount of an anti-Ep-CAM antibody.

Another embodiment of the invention is a method of determining the sensitivity of a gynecological malignancy to treatment with an anti-Ep-CAM antibody involving detecting the presence or absence of Ep-CAM in a tissue sample comprising a portion of the malignancy.

Another embodiment of the invention provides a method of protecting a mammal from CPE toxicity involving administering a non-toxic agent that binds to claudin-3 and/or claudin-4 and inhibits CPE binding to claudin-3 and/or claudin-4. Another embodiment of the invention provides a method of treating cancer in a mammal involving administering to the mammal an antibody against claudin-3 and/or claudin-4, wherein cancerous cells in the mammal contain claudin-3 and/or claudin-4. Another embodiment of the invention provides a humanized antibody against an extracellular portion of claudin-3 and/or an extracellular portion of claudin-4.

Another embodiment of the invention provides an antibody that includes an antigen-binding region comprising residues 290-319 of SEQ ID NO: 1 or a fragment thereof that binds specifically to claudin-3 and/or claudin-4. The fragment may include residues 315-319 of SEQ ID NO:1

Another embodiment of the invention provides a method of treating cancer in a mammal involving: administering to the mammal a therapeutically effective amount of a cytotoxic agent that binds specifically to claudin-3 and/or claudin-4; wherein the cancer is ovarian cancer or uterine cancer. Another embodiment of the invention provides an anticancer agent containing:

(a) a moiety that binds specifically to claudin-3 and/or claudin-4; coupled to (b)a cytotoxic moiety; wherein the agent is effective to kill tumor cells overexpressing claudin-3 and/or claudin-4 in vitro or in vivo, and wherein the agent is not CPE.

Brief Description of the Drawings

FIG. 1 is a schematic showing hierarchical clustering of 15 primary ovarian cell lines (10 OSPC and 5 NOVA) and 2 established OSPC cell lines (UCI-101 and UCI- 107). FIG. 2 shows bar graphs of quantitative real-time PCR and microarray expression analyses of TROP-I, CD24, claudin-3, and claudin-4 genes differentially expressed between OSPC and NOVA.

FIG. 3 shows representative FACS analyses of CD24 staining (left panel) and TROP-I /Ep-CAM staining (right panel) of 2 primary OSPC cell lines and 1 NOVA cell line. Data on CD24 and TROP-I /Ep-CAM are shown in solid black, while isotype control MAb profiles are shown in white.

FIG. 4 depicts an unsupervised hierarchical clustering of fifteen primary uterine cell lines (10 USPC and 5 NEC). FIG. 5 shows bar graphs of quantitative RT-PCR and microarray expression analysis of CDKN2A/pl6, CDKN2A/pl4ARF, LlCAM, claudin-3, claudin-4, GRB- 7, andc-erbB2 genes differentially expressed between USPC and NEC.

FIG. 6. Quantitative RT-PCR analysis of claudin-3 and claudin-4 expression. The Y axis represents the fold induction relative to normal ovary expression (sample 1). The X axis represents each sample tested for claudin-3 (A) and claudin-4 (B). The first 15 bars represent normal ovarian epithelium (1-3), normal endometrial epithelium (4-6), normal cervical keratinocytes (7), primary squamous cervical cancer cell lines (8-10), primary adenocarcinoma cervical cancer cell lines (11-13), Epstein- Barr transformed B lymphocytes (LCL) (14), and human fibroblasts (15). The following 16 bars represent primary ovarian cancer cell lines (16-21 serous papillary ovarian cancers, 22-26 clear cell ovarian tumors) and established serous ovarian cancer cell lines, 27-31 (i.e, UCI 101, UCI 107, CaO V3, OVACAR-3, and OVARK-

5). FIG. 7. Quantitative RT-PCR analysis of claudin-3 and claudin-4 expression in chemotherapy naive versus chemotherapy resistant/recurrent ovarian cancer. The Y axis represents the fold induction relative to normal ovary expression. The X axis represents each sample tested for claudin-3 and claudin-4. Figs. 7 A and 7B: 1 (chemotherapy naϊve ovarian cancers = 6 OSPC samples, mean + SEM); 2 (chemotherapy resistant/recurrent ovarian cancer = 6 OSPC samples, mean + SEM) (p < 0.05). Figs. 7C and 7D: 1 (chemotherapy naϊve) and 2 (chemotherapy resistant) represent claudin-3 and claudin-4 expression in autologous matched OVA-I tumors. 3 (chemotherapy naϊve) and 4 (chemotherapy resistant) represent claudin-3 and claudin-4 expression in autologous matched OVA-4 tumors. 5 (chemotherapy naϊve) and 6 (chemotherapy resistant) represent claudin-3 and claudin-4 expression in autologous matched OVA-6 tumors.

FIG. 8. Quantitative RT-PCR analysis of claudin-3 and claudin-4 expression in primary vs. metastatic ovarian cancer tumors. The Y axis represents the fold induction of claudin-3 (left panel) and claudin-4 (right panel) relative to normal ovary expression. The sample tested is indicated on the X axis as (1) OVA-I primary tumor, and (2) OVA-I metastatic tumor.

FIG. 9. Representative dose dependent CPE-mediated cytotoxicity of primary ovarian cancers compared to positive control Vero cells or negative controls (i.e., normal and neoplastic cells) after 24 hrs exposure to CPE. VERO: positive control cells. OVA-I to OVA-6: primary ovarian tumors. OVARK-5, CaOV3, OVACAR-3: established serous ovarian tumors. Nova: normal ovarian epithelium. Norm CX: normal cervix keratinocytes. Fibroblast: normal human fibroblasts. LCL: lymphoblastoid B-cells. PBL: normal peripheral blood lymphocytes. CXl-3: primary squamous cervical cancer. ADX 1-3: primary adenocarcinoma cervical cancer.

FIGS. 1OA and B. Survival of C.B-17/SCID Mice after i.p. injection of 5-7.5 x 10⁶ viable OVA-I tumor cells. Animals harboring four- week (A) and one- week (B) established OVA-I tumors were injected i.p. with doses of CPE ranging from 5 to 8.5μg as described in the material and methods. CPE was administered i.p. every 72 hrs until death or end of study. Mice were evaluated on a daily basis and sacrificed when moribund. In both studies, the log-rank test yielded p < 0.0001 for the differences in survival.

Detailed Description Definitions:

The term "Clostridium perfringens enterotoxin" (CPE) includes the native enterotoxin (SEQ ID NO: 1) or an engineered, mutant, or variant enterotoxin that is - toxic and binds specifically in vivo in a mammal to claudin-3 and/or claudin-4. The portion of native CPE responsible for toxicity is thought to be approximately residues 45-116 of full-length native CPE (SEQ ID NO:1). The portion of native CPE responsible for binding to claudin-3 and claudin-4 is residues 290-319 of SEQ ID NO: 1 or a portion thereof (23-27).

The term "chemotherapy-resistant" cancer cells refers to cells of a tumor that has recurred after chemotherapy, not responded to chemotherapy, has demonstrated survival in vitro or in vivo after exposure to doses of one or more chemotherapy agents that would otherwise be considered for use against the tumor, or is of a type of tumor (such as USPC) that, based on data from experimental animals or other patients, is considered to not respond to chemotherapy or recur after chemotherapy.

The term "metastasized" or "metastatic" refers to cancer cells that have spread to a different part of the body from the site where the primary tumor for the cancer originated. The term "therapeutic index" is defined herein as the ratio of the minimal toxic dose (LD₅₀) to the minimal dose that induces a measurable response in vivo, such as arresting tumor growth.

The term "cytotoxic agent" refers to an agent that kills cancerous cells in vitro or in vivo in a mammal. The term "cytotoxic agent" includes agents that are not toxic in the absence of immune system components but that promote an immune response that kills cancerous cells in vivo.

The term "Fc" refers to the constant portion of an immunoglobulin protein.

The terms "pancreatic cancer," "ovarian cancer," "uterine cancer," etc. refer to the organ or tissue type in which the primary tumor of the cancer originated.

Description:

CPE is an enterotoxin responsible for many cases of food poisoning. Native CPE is a 319 residue protein (SEQ ID NO: 1) that binds specifically to claudin-3 and claudin-4 on the surface of certain cells in the gastrointestinal tract. After CPE binds to its target cell receptor (claudin-3 or claudin-4) a 90 kDa small membrane complex is formed in the target cell plasma membrane containing intact CPE. Additional target cell proteins then associate with the small complex to form a large 160 kDa membrane complex that allows partial insertion of CPE into the membrane, allowing small molecule permeability changes and lysis (21, 22).

Functional domain mapping of the full-length 319 amino acid CPE protein showed that a CPE(290-319) COOH-terminus fragment is sufficient for high affinity binding to receptors and small complex formation, but does not initiate large complex formation and cytolysis. In fact, CPE(290-319) fragment inhibits cytolysis by full- length CPE (26). Deletion from full-length native CPE of the five residues at the C- terminus, residues 315-319, is enough to abolish binding to the receptors (25). Approximately residues 45-116 of CPE are essential for large complex formation and cytolysis. With NH2-terminal residues 1-44 deleted, CPE(45-319) is actually more toxic than full-length CPE (23-27). Thus, in particular embodiments of the invention, the CPE is a variant CPE comprising residues 45-319 of SEQ ID NO: 1. In other particular embodiments, the CPE comprises a sequence at least 80% identical, or at least 90% identical to residues 45-319 of SEQ ID NO: 1. In particular embodiments of the invention, the CPE comprises residues 290- 319 of SEQ ID NO: 1. In particular embodiments, the CPE comprises a fragment of residues 290-319 of SEQ ID NO:1 that binds specifically to claudin-3 and/or claudin- 4. In particular embodiments, the CPE comprises residues 315-319 of SEQ ID NO: 1. In particular embodiments, CPE comprises a variant of residues 290-319 of SEQ ID NO:1 that specifically binds claudin-3 and claudin-4. In particular embodiments, CPE comprises a variant of residues 290-319 of SEQ ID NO:1 that specifically binds claudin-3 or claudin-4.

In particular embodiments, the CPE binds claudin-3 and/or claudin-4 with a K_D of less than 10 μM, or more preferably less than about 1 μM, most preferably less than about 0.1 μM. This can be assayed as described in references 14 and 26.

One embodiment of the invention is a method of treating ovarian cancer or uterine cancer in a mammal involving administering a therapeutically effective amount of CPE or a pharmaceutically acceptable salt thereof. In a specific embodiment, the cancer includes chemotherapy-resistant malignant cells. As shown below, chemotherapy-resistant ovarian cancer cells have higher expression of claudin-3 and -4 than chemotherapy-naive ovarian cancer cells. USPC, a chemotherapy-resistant and aggressive form of uterine cancer also is shown below to have high expression of claudin-3 and -4. Thus, CPE is particularly suited for treatment of ovarian and uterine cancer. Moreover, since CPE is a toxin, oncologists may be hesitant to use it on a tumor that is thought to be sensitive to less toxic conventional chemotherapy agents.

In another particular embodiment of the method, the cancer includes metastatic malignant cells. It is shown below in Example 3 that metastatic ovarian tumors express higher levels of claudin-3 and -4 than primary ovarian tumors. USPC, an aggressive form of uterine cancer that tends to readily metastasize also is shown below to have high expression of claudin-3 and -4.

In a particular embodiment, the cancer is uterine serous papillary carcinoma (USPC). In a particular embodiment, the cancer is ovarian serous papillary carcinoma

(OSPC).

In a particular embodiment of the method, the CPE or pharmaceutically acceptable salt thereof is administered intraperitoneally. The dosage of CPE or a pharmaceutically acceptable salt thereof for intraperitoneal administration is 0.2 to 0.8 mg/kg body mass, in a particular embodiment.

In other particular embodiments, the CPE or pharmaceutically acceptable salt thereof is administered intravenously or intratumorally.

In particular embodiments, the CPE or pharmaceutically acceptable salt thereof has a therapeutic index as administered of at least 5. The therapeutic index can depend on the route of administration. Thus, for tumors located in or adjacent to the peritoneal cavity, intraperitoneal administration is likely to have a higher therapeutic index than systemic intravenous administration. The therapeutic index will also depend on the type of cancer. The therapeutic index for treatment of a human is determined by data with experimental animals from the closest animal model of the particular type of cancer involved, or from data on treatment of humans when available. The therapeutic index is defined as the minimal toxic dose (e.g., the lethal dose for 50% of experimental animals) divided by the minimal effective dose (e.g., the minimal dose required to cause measurable tumor shrinkage in 50% of experimental animals).

In other particular embodiments, the CPE or pharmaceutically acceptable salt thereof has a therapeutic index as administered of at least 2. In other particular embodiments, the CPE or pharmaceutically acceptable salt thereof has a therapeutic index as administered of at least 4, at least 6, at least 8, at least 10, or at least 20.

In some embodiments, the CPE consists of SEQ ID NO:1. In some embodiments, the CPE is a mutant or engineered CPE. In particular embodiments, the mutant or engineered CPE comprises residues 45-116 of SEQ ID NO:1.

In some embodiments of the methods of the invention, the method involves administering a protective agent that protects cells against CPE toxicity.

A peptide consisting of residues 290-319 of CPE SEQ ID NO:1 has been shown to compete with native CPE for binding to cells and thus inhibit CPE binding (26). Any CPE fragment that includes residues 290-319 and does not include residues 45-116 is expected to be protective. Such a fragment will bind to the CPE receptors (claudin-3 and -4) but not lyse the cells, and will compete with full-length CPE for binding to the receptor (23-27). Fragments of residues 290-319 of SEQ ID NO: 1 are also envisioned. In specific embodiments, the fragment includes residues 315-319 of SEQ ID NO: 1.

Thus, one protective agent that protects cells against CPE toxicity is a protective agent that comprises residues 290-319 of native CPE, or a homologue thereof.

A protective agent that comprises residues 290-319 of native CPE, or a homologue thereof, is a proteinaceous agent that binds to claudin-3 and/or claudin-4 and does not lyse the cells it binds to. In particular embodiments, the homologue of residues 290-319 of native CPE is at least 80% or at least 90% identical to residues 290-319 of SEQ ID NO: 1.

Another embodiment of a protective agent is a protective agent that comprises a fragment of residues 290-319 of SEQ ID NO:1 that binds specifically to claudin-3 and/or claudin-4.

The protective agent can also be an antibody against claudin-3 and/or claudin- 4 that inhibits binding of CPE to claudin-3 and/or claudin-4. The term an "antibody against claudin-3 and/or claudin-4" refers to an antibody that specifically binds claudin-3 (i.e., binds claudin-3 and does not bind other proteins in a mammal), specifically binds claudin-4, or specifically binds to both claudin-3 and claudin-4.

The protective agent that is an antibody against claudin-3 and/or claudin-4 can be an antibody comprising an Fc region coupled to a peptide comprising residues 290- 319 of SEQ ID NO: 1 or a fragment thereof that specifically binds claudin-3 and/or claudin-4.

In particular embodiments, the protective agent is administered intravenously and the CPE or pharmaceutically acceptable salt thereof is administered by a non- intravenous route. For instance, the CPE or salt thereof may be administered intraperitoneally and the protective agent administered intravenously.

In particular embodiments, the protective agent is administered enterally. The CPE or salt thereof may be administered in conjunction with this, e.g., intraperitoneally or intravenously. In particular embodiments, the protective agent is a non-toxic agent that inhibits CPE binding to claudin-3 and/or claudin-4. The inhibition may be competitive or non-competitive. Preferably the inhibition is competitive.

One embodiment of the invention is a method of treating cancer in a mammal involving intraperitoneally administering to the mammal a therapeutically effective

I l amount of CPE or a pharmaceutically acceptable salt thereof, wherein at least some cancerous cells are located in or adjacent to the peritoneal cavity of the mammal, and the cells are sensitive to CPE. The cancerous cells are considered "adjacent to the peritoneal cavity" of the mammal if intraperitoneal administration of CPE results in a higher therapeutic index (the ratio of the toxic dose to the effective dose) than systemic intravenous administration of CPE. Cancerous cells may also be considered adjacent to the peritoneal cavity if intraperitoneal administration of CPE results in a higher concentration of CPE in the cancerous cells than systemic intravenous administration of CPE. In a particular embodiment, all detectable cancerous cells are in or adjacent to the peritoneal cavity.

In particular embodiments of the method involving intraperitoneal administration of CPE, the cancerous cells include chemotherapy-resistant cells.

In particular embodiments, the cancerous cells include metastatic cancerous cells.

In particular embodiments, the cancer treated with intraperitoneal administration of CPE is pancreatic cancer. In other embodiments, the cancer is ovarian cancer or uterine cancer.

In particular embodiments, the cancer is uterine serous papillary carcinoma (USPC).

In another particular embodiment, the cancer is ovarian serous papillary carcinoma (OSPC).

In particular embodiments, the cancer is liver cancer, stomach cancer, colon cancer, bladder cancer, kidney cancer, intestinal cancer, testicular cancer, or prostate cancer, wherein cells of the cancer have claudin-3 and/or claudin-4.

The method involving intraperitoneal administration of CPE may involve in addition administering intravenously or enterally a protective agent that protects cells against CPE toxicity. This will provide some protection to healthy non-target tissues.

One embodiment of the invention is a use of CPE to prepare a medicament effective to treat ovarian or uterine cancer in a mammal.

In a particular embodiment, the mammal is a human.

In a particular embodiment, the medicament is adapted for intraperitoneal administration to the mammal. Another embodiment of the invention is a use of CPE to prepare a medicament effective to treat cancer in a mammal, wherein the medicament is adapted for intraperitoneal administration.

In particular embodiments of the use, the cancer is ovarian cancer, uterine cancer, or pancreatic cancer.

The invention also provides a method of determining the sensitivity of a gynecological malignancy to CPE involving detecting the presence or absence of claudin-3 and/or claudin-4 in a tissue sample comprising a portion of the malignancy.

In particular embodiments, the detecting step involves contacting the tissue sample with a protein containing residues 290-319 of SEQ ID NO: 1. In specific embodiments, the protein is CPE. The CPE may be SEQ ID NO:1. In particular embodiments, the protein is an antibody comprising residues 290-319 of SEQ ID NO:1 or a fragment thereof that binds specifically to claudin-3 and/or claudin-4 (e.g., an antibody comprising an Fc coupled to an antigen-binding peptide comprising residues 290-319 of SEQ ID NO: 1).

Binding of the protein can be detected in various ways. If the protein is CPE, lysis of the cells can be detected. If the protein is labeled with radioactivity, a fluorescent label, or some other label, the label can be detected. The protein can also include a tag that is recognized by an antibody, where binding of the protein is detected by adding antibody after adding the protein, and detecting binding of the antibody to the tissue sample.

The detecting step can involve detecting lysis of cells. The detecting step can involve contacting the tissue sample with an antibody against claudin-3 and/or an antibody against claudin-4. In Example 1 below it is shown that the TROP- 1/Ep-CAM gene is overexpressed, and the Ep-CAM protein is found, in ovarian serous papillary carcinoma cells. An anti-Ep-CAM antibody, edrecolomab, has been shown to increase survival in patients harboring stage III colon cancer expressing the Ep-CAM protein (28). Thus, one embodiment of the invention is a method of treating cancer in a mammal involving administering to the mammal a therapeutically effective amount of an anti-Ep-CAM antibody; wherein the cancer is ovarian cancer. In particular embodiments, the ovarian cancer is OSPC. In particular embodiments, the antibody is edrecolomab. The antibody may be radioactive, in which case it can be used for radiation therapy of the cancer.

The detecting step can include contacting the tissue sample with an antibody against Ep-CAM. The antibody can be polyclonal or monoclonal.

In particular embodiments, the malignancy is ovarian cancer, or specifically OSPC.

Another embodiment of the invention is a method of protecting a mammal from CPE toxicity involving administering a non-toxic protective agent that binds to claudin-3 and/or claudin-4 and inhibits CPE binding to claudin-3 and/or claudin-4. The protective agent can be administered before CPE is administered for treatment. Or the protective agent can be administered after CPE is administered, e.g., as an antidote to an overdose of therapeutic CPE, or as an antidote or treatment for CPE food poisoning.

The protective agent could be an antibody against claudin-3 and/or claudin-4, e.g., a humanized antibody. The protective agent can also be an agent that comprises residues 290-319 of

SEQ ID NO:1, or a homologue thereof, or can be an agent that comprises a fragment of residues 290-319 of SEQ ID NO:1 that binds specifically to claudin-3 and/or claudin-4.

Another embodiment of the invention provides a method of treating cancer in a mammal involving administering to the mammal an antibody against claudin-3 and/or claudin-4, wherein cancerous cells in the mammal contain claudin-3 and/or claudin-4. Preferably, the cancerous cells contain at least 5-times more claudin-3 and/or claudin-4 than normal cells of the same tissue type.

In certain embodiments, the antibody may inhibit CPE binding to claudin-3 and/or claudin-4. But to treat cancer, the antibody does not have to. It need only bind to the claudin-3 or claudin-4 and attract immune attack against the antibody-labeled cells.

The antibody may also carry a therapeutic radioisotope, so that it kills the claudin-3- or claudin-4-containing cells with radioactivity. In particular embodiments, the cancer treated with antibodies is ovarian cancer (e.g., OSPC) or uterine cancer (e.g., USPC). In other embodiments, the cancer is pancreatic, prostate, or colon cancer. In other embodiments, the cancer is breast cancer or lung cancer. Antibodies against claudin-3 and claudin-4 and Ep-CAM can be generated by injecting rabbits, mice, goats, or other appropriate vertebrate animals with whole claudin-3 or claudin-4 or Ep-CAM, or peptides from the proteins can be synthesized and coupled to a carrier protein for injection. Preferably, if peptides are used, extracellular domains of the proteins are selected, since these are the portions that will be exposed on the exterior of the cell. The amino acid sequences of human claudin-4 (referred to as human CPE receptor) and claudin-3 (referred to as RVPl) and hydrophobicity plots of the proteins, which suggest the transmembrane domains and the extracellular domains, are disclosed in references 15 and 50. This can be used to select peptides to be used to generate antibodies. Polyclonal or monoclonal antibodies can be prepared from the immunized animal as described below. The antibodies used should recognize the exposed extracellular portions of claudin-3 and/or claudin-4 in order to bind to cells expressing the proteins. For this reason, monoclonal antibodies may be preferable, since they can be characterized as recognizing a specific epitope found on the outer surface of the cells expressing claudin-3 or claudin-4.

Antibodies obtained can be administered to experimental mammals at a dose of approximately 10 mg/kg every 3-7 days to treat cancer.

Antibodies can be labeled with radioisotopes, such as yttrium-90 or iodine- 131 for therapeutic radiation treatment. One procedure for attaching radioisotopes to antibodies is described in reference 122. Radiolabeled antibodies are administered to experimental animals at appropriate radiation doses. For instance, yttrium-90 is administered to humans in the monoclonal antibody ZEV ALIN™ for treatment of lymphoma at a dose of 5 mCi. Thus, in some embodiments of the antibodies used in the invention, the antibody carries a therapeutic radionuclide. One embodiment of the invention is an antibody comprising an antigen- binding region comprising residues 290-319 of SEQ ID NO: 1 or a fragment thereof that binds specifically to claudin-3 and/or claudin-4. This can be created by recombinant means as described in Example 4 below. Briefly, a recombinant DNA is engineered to express a fusion polypeptide comprising the CH2 and CH3 domains of the IgGl heavy chain coupled to residues 290-319 of SEQ ID NO: 1. Two heavy chains dimerize in vivo to form an Fc coupled to two peptides containing the residues of CPE that bind specifically to claudin-3 and claudin-4.

Thus, in particular embodiments, an antibody used in the invention is produced by recombinant means.

Other methods for the preparation of antibodies are described below.

Another embodiment of the invention provides a method of treating cancer in a mammal involving administering to the mammal a therapeutically effective amount of a cytotoxic agent that binds specifically to claudin-3 and/or claudin-4, wherein the cancer is uterine cancer or ovarian cancer.

The cytotoxic agent may be CPE or a pharmaceutically acceptable salt thereof.

The cytotoxic agent may also be an antibody that binds specifically to claudin- 3 and/or claudin-4.

The cytotoxic agent may include residues 290-319 of SEQ ID NO: 1 or a fragment thereof that binds specifically to claudin-3 and/or claudin-4; coupled to a cytotoxic moiety.

In particular embodiments, the cytotoxic moiety is an anti-cancer chemotherapy agent. An anti-cancer chemotherapeutic agent refers to an agent that kills or inhibits the growth of cancer cells while having less effect on non-cancerous cells, typically because it selectively kills dividing cells. Examples suitable for use in creating these cytotoxic agents include amsacrine, azacytidine, bleomycin, busulfan, capecitabine, carboplatin, carmustine, chlorambucil, cisplatin, cladribine, cyclophosphamide, cytarabine, dactinomycin, daunorubicin, decarbazine, docetaxel, doxorubicin, epirubicin, estramustine, etoposide, floxuridine, fludarabine, fluorouracil, gemcitabine, hexamethylmelamine, idarubicin, ifosfamide, irinotecan, lomustine, mechlorethamine, melphalan, mercaptopurine, methotrexate, mitomycin C, mitotane, mitoxantrone, oxaliplatin, paclitaxel, pemetrexed, pentostatin, plicamycin, procarbazine, ralitrexed, semustine, streptozocin, temozolamide, tenyposide, thioguanine, thiotepa, topotecan, trimitrexate, valrubicin, vincristine, vinblastine, vindestine, or vinorelbine.

The chemotherapeutic agents can be coupled to the CPE peptide that binds claudin-3/claudin-4 as described in published international patent application PCT/US2004/034704. In other embodiments, the cytotoxic moiety is a moiety that stimulates cell killing by immune system components, e.g., cell killing by the complement system or T-cells. In particular embodiments, the cytotoxic moiety is an antibody Fc. In other specific embodiments, the cytotoxic moiety is interleukin-2. It has been shown that interleukin-2 (IL2) can be coupled to an antibody that recognizes an antigen on cancer cells, and the coupled moiety binds to target cells through the antibody portion of the conjugate (123). The interleukin-2 portion of the conjugate is recognized by T-cells, including T-cells that do not specifically recognize the tumor cells (123). The conjugates have been shown to thereby couple target tumor cells to noncognate T-cells (123). The conjugates also induce cell killing of the target tumor cells by T-cells, including T-cells that do not specifically recognize the target tumor cells in the absence of the antibody-IL2 conjugate (123).

Thus, here IL2 can be coupled to residues 290-319 of SEQ ID NO: 1 or a fragment thereof that binds specifically to claudin-3 and/or claudin-4. The fusion protein will bind to cells carrying claudin-3 or claudin-4 receptors and induce killing of these cells by cytotoxic T lymphocytes through the T-cells' recognition of, binding to, and stimulation by the IL2 moiety.

In another embodiment of this method of treating cancer using a cytotoxic agent comprising residues 290-319 of SEQ ID NO: 1 or a fragment thereof that specifically binds claudin-3 and/or claudin-4; coupled to a cytotoxic moiety, the cytotoxic moiety is a therapeutic radionuclide.

The methods described herein using agents other than CPE are, like the methods using CPE, suitable for treatment of ovrian cancer and uterine cancer. They are also suitable for treatment of ovarian, uterine, and other cancers wherein the cancer includes chemotherapy resistant cells or metastatic malignant cells, or wherein the cancer is a recurrent cancer.

Raising Antibodies

To generate antibodies, claudin-3 or claudin-4 or Ep-CAM can be administered directly to a mammal, or the proteins or peptide fragments thereof can be coupled to a carrier protein. Suitable carrier proteins include keyhole limpet hemocyanin, bovine serum albumin, and ovalbumin. Methods of coupling to the carrier protein include single step glutaraldehyde coupling and other methods disclosed in Harlow, Ed et al., Antibodies: a laboratory manual, Cold Spring Harbor Laboratory (1988).

The immunogen is used to immunize a vertebrate animal in order to induce the vertebrate to generate antibodies. Preferably the immunogen is injected along with an adjuvant such as Freund's adjuvant, to enhance the immune response. Suitable vertebrates include rabbits, mice, rats, hamsters, goats, and chickens.

Hybridomas to synthesize monoclonal antibodies can be prepared by methods known in the art. See, for instance, Wang, H., et al., Antibody Expression and Engineering, Am. Chem. Soc, Washington, DC (1995). Polyclonal and monoclonal antibodies can be isolated by methods known in the art. See, for instance, id. and Harlow et al.

Native antibodies are tetramers of two identical light (L) chains and two identical heavy (H) chains. The L and H chains each have variable domains that are responsible for antigen recognition and binding. The variability in the variable domains is concentrated in the complementarity determining regions (CDRs).

An antibody that is contemplated for use in the present invention can be in any of a variety of forms, including a whole immunoglobulin, an antibody fragment such as Fv, Fab, and similar fragments, a single chain antibody that includes the CDR, and like forms, all of which fall under the broad term "antibody" as used herein. The term "antibody fragment" refers to an antigen-binding portion of a full- length antibody. Antibody fragments can be as small as about 4 amino acids, about 10 amino acids, or about 30 amino acids or more. Some types of antibody fragments are the following:

(1) Fab is the fragment that contains a monovalent antigen-binding fragment of an antibody molecule. A Fab fragment can be produced by digestion of whole antibody with the enzyme papain to yield an intact light chain and a portion of one heavy chain. Two Fab fragments are obtained per whole antibody molecule.

(2) Fab' is the fragment of an antibody that can be obtained by treating whole antibody with pepsin, followed by reduction to yield an intact light chain and a portion of the heavy chain. Two Fab' fragments are obtained per whole antibody molecule. Fab' fragments differ from Fab fragments by the addition of a few residues at the carboxyl terminus of the heavy chain CHl domain including one or more cysteines. (3) F(ab')2 is the fragment that can be obtained by digestion of whole antibody with pepsin, without reduction. F(ab')₂ is a dimer of two Fab' fragments held together by two disulfide bonds.

(4) Fv is the minimum antibody fragment that contains a complete antigen recognition and binding site. Fv consists of a dimer of one H and one L chain variable domain in a tight, non-covalent association (V_H -V_L dimer). It is in this configuration that the three CDRs of each variable domain interact to define an antigen-binding site. Collectively, the six CDRs confer antigen binding specificity to the antibody. However, even a single vaiable domain (or half of an Fv comprising only three CDRs specific for an antigen) has the ability to bind antigen, although at a lower affinity than the complete binding site.

(5) A single chain antibody (SCA) is defined as a genetically engineered molecule containing the variable region of the light chain and the variable region of the heavy chain linked by a suitable polypeptide linker as a genetically fused single chain molecule.

The preparation of polyclonal antibodies is well known to those skilled in the art. See, for example, Coligan et al., in Current Protocols in Immunology, section 2.4.1 (1992). The preparation of monoclonal antibodies is likewise conventional. See, for example, Harlow et al., page 726. Methods of in vitro and in vivo manipulation of monoclonal antibodies are well known to those skilled in the art. For example, the monoclonal antibodies to be used in accordance with the present invention may be made by the hybridoma method first described by Kohler and Milstein, Nature 256:495 (1975), or may be made by recombinant methods, e.g., as described in U.S. Patent No. 4,816,567. The monoclonal antibodies for use with the present invention may also be isolated from phage antibody libraries using the techniques described in Clarkson et al., Nature 352:624 (1991), as well as in Marks et al., J. MoI Biol. 222:581 (1991). Another method involves humanizing a monoclonal antibody by recombinant means to generate antibodies containing human specific and recognizable sequences. See, for review, Holmes et al., J Immunol. 158:2192 (1997) and Vaswani et al., Annals Allergy, Asthma & Immunol. 81:105 (1998).

The monoclonal antibodies herein specifically include "chimeric" antibodies in which a portion of the heavy and/or light chain is identical with or homologous to corresponding sequences in antibodies derived from a particular species or belonging to a particular antibody class or subclass, while the remainder of the chain(s) are identical with or homologous to corresponding sequences in antibodies derived from another species or belonging to another antibody class or subclass, as well as fragments of such antibodies, so long as they exhibit the desired biological activity (U.S. Patent No. 4,816,567; Morrison et al., Proc. Nat l Acad. Sci. 81 :6851 (1984)).

Methods of making antibody fragments are also known in the art (see, for example, Harlow and Lane, Antibodies: a Laboratory Manual, Cold Spring Harbor Laboratory, New York (1988)). Antibody fragments of the present invention can be prepared by proteolytic hydrolysis of the antibody or by expression in E. coli of DNA encoding the fragment. Antibody fragments can be obtained by pepsin or papain digestion of whole antibodies by conventional methods. For example, antibody fragments can be produced by enzymatic cleavage of antibodies with pepsin to provide a 5 S fragment denoted F(ab')₂. This fragment can be further cleaved using a thiol reducing agent, and optionally a blocking group for the sulfhydryl groups resulting from cleavage of disulfide linkages, to produce 3.5 S Fab' monovalent fragments. Alternatively, an enzymatic cleavage using pepsin produces two monovalent Fab fragments and an Fc fragment directly. These methods are described, for example, in U.S. Patents No. 4,036,945, and 4,331,647, and references contained therein. Other methods of cleaving antibodies, such as separation of heavy chains to form monovalent light-heavy chain fragments, further cleavage of fragments, or other enzymatic, chemical, or genetic techniques may also be used, so long as the fragments bind to the antigen that is recognized by the intact antibody. For example, Fv fragments comprise an association of VH and VL chains. This association may be noncovalent or the variable chains can be linked by an intermolecular disulfide bond or cross-linked by chemicals such as glutaraldehyde. Preferably, the Fv fragments comprise VH and VL chains connected by a peptide linker. These single-chain antigen binding proteins (sFv) are prepared by constructing a structural gene comprising DNA sequences encoding the VH and VL domains connected by an oligonucleotide. The structural gene is inserted into an expression vector, which is subsequently introduced into a host cell such as E. coli. The recombinant host cells synthesize a single polypeptide chain with a linker bridging the two V domains. Methods for producing sFvs are described, for example, by Whitlow et al., Methods: a Companion to Methods in Enzymology, 2:97 (1991); Bird et al., Science 242:423 (1988); Ladner et al., U.S. Patent No. 4,946,778; and Pack et al., Bio/Technology 11 : 1271 (1993).

Another form of an antibody fragment is a peptide containing a single complementarity-determining region (CDR). CDR peptides ("minimal recognition units") are often involved in antigen recognition and binding. CDR peptides can be obtained by cloning or constructing genes encoding the CDR of an antibody of interest. Such genes are prepared, for example, by using the polymerase chain reaction to synthesize the variable region from RNA of antibody-producing cells. See, for example, Larrick et al., Methods: a Companion to Methods in Enzymology, 2:106 (1991).

The invention contemplates human and humanized forms of non-human (e.g., murine) antibodies. Such humanized antibodies are chimeric immunoglobulins, immunoglobulin chains, or fragments thereof (such as Fv, Fab, Fab', F(ab')₂ or other antigen-binding subsequences of antibodies) that contain minimal sequence derived from non-human immunoglobulin. For the most part, humanized antibodies are human immunoglobulins (recipient antibody) in which residues from a CDR of the recipient are replaced by residues from a CDR of a non-human species (donor antibody) such as mouse, rat, or rabbit having the desired specificity, affinity, and capacity. In some instances, Fv framework residues of the human immunoglobulin are replaced by corresponding non-human residues. Furthermore, humanized antibodies may comprise residues that are found neither in the recipient antibody nor in the imported CDR or framework sequences. These modifications are made to further refine and optimize antibody performance. In general, humanized antibodies will comprise substantially all of at least one, and typically two, variable domains, in which all or substantially all of the CDR regions correspond to those of a non-human immunoglobulin and all or substantially all of the FR regions are those of a human immunoglobulin consensus sequence. The humanized antibody optimally also will comprise at least a portion of an immunoglobulin constant region (Fc), typically that of a human immunoglobulin. For further details, see: Jones et al., Nature 321 :522 (1986); Reichmann et al., Nature 332:323 (1988); Presta, Curr. Opinion Struct. Biol. 2:593 (1992); Holmes et al., J. Immunol. 158:2192 (1997); and Vaswani et al., Annals Allergy, Asthma & Immunol. 81:105 (1998). Antibodies of the invention can also be mutated to optimize their affinity, selectivity, binding strength or other desirable property. One method of mutating antibodies involves affinity maturation using phage display. Affinity maturation using phage display refers to a process described in Lowman et al., Biochemistry 30:10832 (1991); see also Hawkins et al., J. MoI. Biol. 254:889 (1992).

The invention will now be illustrated by the following non-limiting examples.

Examples The following materials and methods are common to the Examples.

MATERIALS AND METHODS

RNA Purification and Microarray Hybridization and Analysis. Detailed protocols for RNA purification, cDNA synthesis, cRNA preparation, and hybridization to the Affymetrix Human U95Av2 GeneChip microarray were performed according to the manufacturer's protocols, as reported previously (29).

Data Processing. All data used in the analyses were derived from Affymetrix 5.0 software. GeneChip 5.0 output files are given as a signal that represents the difference between the intensities of the sequence-specific perfect match probe set and the mismatch probe set, or as a detection of present, marginal, or absent signals as determined by the GeneChip 5.0 algorithm. Gene arrays were scaled to an average signal of 1500 and then analyzed independently. Signal calls were transformed by the log base 2 and each sample was normalized to give a mean of 0 and variance of 1. Gene Expression Data Analysis. Statistical analyses of the data were performed with the software package SPSS 10.0 (SPSS, Chicago, IL). The first test applied was the detection. In each comparison, genes having "present" detection calls in more than half of the samples in the overexpressed gene group were retained. To compare gene expression levels, the nonparametric Wilcoxon rank sum (WRS) test (p < 0.05) was applied to the normalized signal call. By combining the detection and WRS data, differentially expressed genes were identified between OSPC and NOVA or USPC and NEC.

Quantitative Real-time PCR. q-RT-PCR was performed with an ABI Prism 7000 Sequence Analyzer using the manufacturer's recommended protocol (Applied Biosystems, Foster City, CA) to validate differential expression of selected genes in samples from all primary cell lines (10 OSPC and 5 NOVA in Example 1; 10 USPC and 5 NEC in Example 2). Each reaction was run in triplicate. The comparative threshold cycle (C_T) method was used for the calculation of amplification fold as specified by the manufacturer. Briefly, five μg of total RNA from each sample was reverse transcribed using SUPERSCRIPT™ II Rnase H Reverse Transcriptase

(Invitrogen, Carlsbad, CA). Ten μl of reverse transcribed RNA samples (from 500 μl of total volume) were amplified by using the TAQMAN™ Universal PCR Master Mix (Applied Biosystems) to produce PCR products specific for the target genes. Primers specific for 18s ribosomal RNA and empirically determined ratios of 18s competimers (Applied Biosystems) were used to control for the amounts of cDNA generated from each sample. Primers for TROP-I, claudin-3 and claudin-4 were obtained from Applied Biosystems as assay on demand products. Assays ID were Hs00158980_ml (TROP-I), Hs00265816_sl (claudin-3), and Hs00533616_sl (claudin-4). CD24 primers sequences were the following (forward, 5'-cccaggtgttactgtaattcctcaa; reverse, 5'-gaacagcaatagctcaacaatgtaaac). Primers for Ll CAM, claudin-3, and claudin-4 were obtained from Applied Biosystems as assay on demand products. Assays ID were Hs00170849_ml (LlCAM,), Hs00265816_sl (claudin-3), and Hs00533616_sl (claudin-4). GRB7 primer sequences were: forward, 5'-tctacgggatgaccactga-3'; reverse, 5'-cgaagccccttgtgtcca-3'. c-erbB2 primer sequences were: forward, 5'- gtatacattcggcgccagct-3'; reverse, 5'-gcagacgagggtgcagga-3'. CDKN2A/pl6 primer sequences were: forward, 5'-cccaaacgcaccgaatagttac-3'; reverse, 5'- attccaattcccctgcaaact-3'. CDKN2A/pl4ARF primer sequences were: forward, 5'- tgatgctactgaggagccagc-3'; reverse, 5'-agggcctttcctacctggtc-3'. Amplification was carried out by using 1 unit of polymerase in a final volume of 20 μl containing 2.5 mM MgCl₂. TAQGOLD™ was activated by incubation at 96⁰C for 12 min, and the reactions were cycled 26-30 times at 95°C for 1 min, 55°C for 1 min, and 72⁰C for 1 min, followed by a final extension at 72°C for 10 min. PCR products were visualized on 2% agarose gels stained with ethidium bromide, and images were captured by an Ultraviolet Products Image Analysis System. Differences among OSPC and NOVA or USPC and NEC in the q-RT-PCR expression data were tested using the Kruskal-Wallis nonparametric test. Pearson product-moment correlations were used to estimate the degree of association between the microarray and q-RT-PCR data. Claudin-4 Immunostaining of Formalin-fixed Tumor Tissues. Ovarian and uterine tumors were evaluated by immunohistochemical staining on formalin-fixed tumor tissue for claudin-4 surface expression. Study blocks were selected after histopathologic review by a surgical pathologist. The most representative hematoxylin and eosin-stained block sections were used for each specimen. Briefly, immunohistochemical stains were performed on 4 μm-thick sections of formalin-fixed, paraffin embedded tissue. After pretreatment with 10 mM citrate buffer at pH 6.0 using a steamer, they were incubated with mouse anti-claudin-4 antibodies (Zymed Laboratories Inc. San Francisco, CA) at 1 :2000 dilution. Antigen-bound primary antibodies were detected using standard avidin-biotin immunoperoxidase complex (Dako Corp., Carpinteria, CA). Cases with less than 10% staining in tumor cells were considered negative for claudin expression, while positive cases were classified as follows regarding the intensity of claudin-4 protein expression: (a) +, focal membrane staining; (b) ++, diffuse membrane staining; and c) +++, diffuse membrane and cytoplasmic staining.

Example !.

Gene Expression Profiles in Primary Ovarian Serous Papillary Tumors and Normal Ovarian Epithelium: Identification of Candidate Molecular Markers for Ovarian Cancer Diagnosis and Therapy

The goal of this study was to identify genes with differential patterns of expression between ovarian serous papillary carcinoma (OSPC) and normal ovarian (NOVA) epithelium and use this knowledge for the development of novel diagnostic and therapeutic markers for ovarian cancer. Gene expression in 10 primary OSPC cell lines, 2 establish OSPCT cell lines (UCI-101, UCI-107), and 5 primary NOVA epithelial cultures was analyzed by oligonucleotide microarrays with probe sets complementary to 12,533 genes.

MATERIALS AND METHODS Establishment of OSPC and NOVA Primary Cell Lines. A total of fifteen primary cell lines (i.e., 10 OSPC and 5 NOVA) were established after sterile processing of the samples from surgical biopsies as previously described for ovarian carcinoma specimens (30,19,31). UCI-101 and UCI-107, two previously characterized OSPC cell lines (32,33) were also included in the analysis. All fresh tumor samples were obtained with appropriate consent according to IRB guidelines. Tumors were staged according to the F.I.G.O. operative staging system. Radical tumor debulking, including a total abdominal hysterectomy and omentectomy, was performed in all ovarian carcinoma patients while normal ovarian tissue was obtained from consenting similar age donors undergoing surgery for benign pathology. No patient received chemotherapy before surgery. Patient characteristics are described in Table 1. Briefly, normal tissue was obtained by scraping epithelial cells from the ovarian surface and placing cells in RPMI 1640 medium (Sigma Chemical Co., St. Louis, MO) containing 10% fetal bovine serum (FBS, Invitrogen, Grand Island, NY), 200 u/ml penicillin, and 200 μg/ml streptomycin. The epithelial explants were then allowed to attach and proliferate. Once the epithelial cells reached confluency, explants were trypsinized and subcultured for 3 to 4 passages before being collected for RNA extraction. Viable tumor tissue was mechanically minced in RPMI 1640 to portions no larger than 1-3 mm³ and washed twice with RPMI 1640. The portions of minced tumor were then placed into 250 ml flasks containing 30 ml of enzyme solution [0.14% collagenase Type I (Sigma, St. Louis, MO) and 0.01% DNAse (Sigma, 2000 KU/mg)] in RPMI 1640, and incubated on a magnetic stirring apparatus overnight at 4° C. Enzymatically dissociated tumor was then filtered through 150 μm nylon mesh to generate a single- cell suspension. The resultant cell suspension was then washed twice in RPMI 1640 plus 10% FBS. Primary cell lines were maintained in RPMI 1640, supplemented with 10% FBS, 200 u/ml penicillin, 200 μg/ml streptomycin at 37⁰C, 5% CO₂. The epithelial nature and the purity of OSPC and NOVA cultures was verified by immunohistochemical staining and flow cytometric analysis with antibodies against cytokeratin as previously described (30,31 ). Only primary cultures which had at least ' 90% viability and contained >99% epithelial cells were used for total RNA extraction.

Gene Cluster/Treeview. The hierarchical clustering of average-linkage method with the centered correlation metric was used (34). The dendrogram was constructed with a subset of genes from 12,588 probe sets present on the microarray, whose expression levels varied the most among the 11 samples, and were thus most informative. For the hierarchical clustering shown in Figure 1 and 2, only genes significantly expressed and whose average change in expression level was at least five-fold were chosen. The expression value of each selected gene was re-normalized to have a mean of zero.

Flow Cytometry. To validate microarray data on primary OSPC and NOVA cell lines at the protein level, TROP-I /EP-CAM and CD24 expression were evaluated by flow cytometric analysis of a total of 13 primary cell lines (i.e., 10 OSPC and 3 NOVA). Unconjugated anti-TROP-1 /EP-CAM (IgG2a), anti-CD24 (IgG2a) and isotype control antibodies (mouse IgG2a) were all obtained from BD PharMingen (San Diego, CA). Goat anti-murine FITC-labeled mouse Ig was purchased from Becton Dickinson (San Jose, CA). Analysis was conducted with a FACScan, utilizing cell Quest software (Becton Dickinson).

TROP-I and CD24 Immunostaining of Formalin-fixed Tumor Tissues. To evaluate whether the differential TROP-I and CD24 expression detected by flow cytometry on primary OSPC cell lines were comparable to the expression of TROP-I and CD24 of uncultured OSPC from which the primary cell lines were derived, protein expression was evaluated by standard immunohistochemical staining on formalin fixed tumor tissue from all surgical specimens (i.e., 10 OSPC and 5 NOVA controls). Study blocks were selected after histopathologic review by a surgical pathologist. The most representative hematoxylin and eosin-stained block sections were used for each specimen. Briefly, immunohistochemical stains were performed on 4 μm-thick sections of formalin-fixed, paraffin embedded tissue. After pretreatment with 10 mM citrate buffer at pH 6.0 using a steamer, they were incubated with anti-Ep-CAM Ab-3 MAb and anti-CD24 (Neo Markers, Fremont, CA) at 1 : 2000 dilution. Slides were subsequently labeled with streptavidin-biotin (DAKO, Glostrup, Denmark), stained with diaminobenzidine and counterstained with hematoxylin. The intensity of staining was graded as 0 (staining not greater than negative control), 1+ (light staining), 2+ (moderate staining), or 3+ (heavy staining).

RESULTS

Characteristics of the Patients. The characteristics of the patients from whom the 10 primary OSPC cell lines were derived are listed in Table 1. Table 1

Gene Expression Profiles Distinguish OSPC From NOVA and Identify Differentially Expressed Genes. Flash frozen biopsies from ovarian tumor tissue are known to contain significant numbers of contaminant stromal cells as well as a variety of host derived immune cells (e.g., monocytes, dendritic cells, lymphocytes). In addition, because ovarian epithelial cells represent a small proportion of the total cells found in the normal ovary, it is difficult to collect primary material that is free of contaminating ovarian stromal cells in sufficient quantities to conduct comparative gene expression analyses. However, ovarian epithelial cells can be isolated and expanded in culture for about 15 passages (30, 19) while the majority of primary ovarian carcinomas can be expanded in vitro for several passages (31). Thus, short term primary OSPC and NOVA cell cultures, minimizing the risk of a selection bias inherent in any long term in vitro growth, may provide an opportunity to study differential gene expression between relatively pure populations of normal and tumor- derived epithelial cells. Accordingly, comprehensive gene expression profiles of 10 primary OSPC and 5 primary NOVA cell lines were generated using high-density oligonucleotide arrays with 12,533 probe sets, which in total interrogated some 10,000 genes. In addition, gene expression profiles derived from two established and previously characterized OSPC cell lines (i.e., UCI-101 and UCI- 107) were also analyzed. By combining the detection levels of genes significantly expressed in primary and established OSPC cultures very little correlation between the two groups of OSPC was found. Indeed, as shown in Figure 1, UCI-101 and UCI- 107 established cell lines grouped together in the leftmost columns of the dendrogram while all ten primary OSPC clustered tightly together in the rightmost columns separately by the 5 NOVA controls. Because of these results the analysis was focused on the detection of differentially expressed genes between the two homogeneous groups of primary OSPC and NOVA cell lines. Using the nonparametric WRS test (p < 0.05) that readily distinguished between the two groups of primary cultures, 1,518 genes differentially expressed between OSPC and NOVA were found. The cluster analysis was based on hybridization intensity values for each cell line for 299 gene segments whose average change in expression level was at least five-fold (data not shown). All 10 OSPC were grouped together. Similarly, all 5 NOVA were found to cluster tightly together. The tight clustering of OSPC from NOVA was "driven" by two distinct profiles of gene expression. The first was represented by a group of 129 genes that were highly expressed in OSPC and underexpressed in NOVA (Table 2). Many genes shown previously to be involved in ovarian carcinogenesis are present on these lists, providing some validity to the array analysis, while others are novel in ovarian carcinogenesis. Included in the genes overexpressed in OSPC are laminin, claudin-3 (CLDN3) and claudin-4 (CLDN4), tumor-associated calcium signal transducer 1 and 2 (TROP-1/Ep-CAM; TROP-2), ladinin 1, S100A2, SERPIN2 (PAI-2), CD24, lipocalin 2, osteopontin, kallikrein 6 (protease M) and kallikrein 10, matiiptase (TADG- 15) and stratifin (Table 2). Importantly, TROP-I /Ep-CAM gene encoding for a transmembrane glycoprotein previously found to be overexpressed in various carcinoma types including colorectal and breast (35) and where antibody-directed therapy has resulted in improved survival of patients (28), was 39-fold differentially expressed in OSPC when compared to NOVA (Table 2). The second profile was represented by 170 genes that were highly expressed in NOVA and underexpressed in OSPC (data not shown). Included in this latter group of genes are transforming growth factor beta receptor III, platelet-derived growth factor receptor alpha, SEMACAP 3, ras homolog gene family, member I (ARHI), thrombospondin 2 and disabled-2/differentially expressed in ovarian carcinoma 2 (Dab2/DOC2).

Table 2

Upregulated genes expressed at least 5 fold higher in OSPC compared with NOVA lProbe Set I Gene Symbol I Score(d)(SAM)| p of WRS Ratio Ova/Nova|

35280_at LAMC2 1.68927386 0.006 46.45

35276_at CLDN4 1.734410451 0.O15 43.76

33904_at CLDN3 1.650076713 0.02 40.24

575_s_at TACSTDI 1.705816336 0.02 39.36

32154 at TFAP2A 1.667038647 0.002 33.31

39015_f_at KRT6E 1.062629117 0.047 28.02

1713_s_at CDKN2A 1.137682905 0.015 26.96

41376_i_at UGT2B7 0.939735032 0.O47 24.81

3855 l_at LlCAM 1.151935363 0.008 24.66

291_s_at TACSTD2 1.249487388 0.047 24.46

33282_at LADl 1.422481563 0.006 24.31

34213 at KIBRA 1.533570321 0.002 23.06 38489_at HBP17 1.522882814 0.004 22.54 3686sTat PAX8 1.43906836 0.004 22.20 38482_at CLDN7 1.307716566 0.027 20.01 37909_at LAMA3 1.121654521 0.027 19.24 34674_at SlOOAl 1.219106334 0.008 19.01 1620_at CDH6 0.908193479 0.036 18.69 3282 l_at LCN2 1.99990601 0.008 18.13 522_s at FOLR3 1.113781518 0.02 17.90 39660^~at DEFBl 0.837612681 0.036 17.34 201 1_s^_at BIK 1.594057668 0.006 17.23 41587_g_at FGF 18 0.965726983 0.02 17.10 36929~at LAMB3 1.115590892 0.047 16.76 35726_at S100A2 1.036576352 0.004 15.05 1887_g_at WNT7A 1.186990893 0.004 14.75 35879__at GAL 1.223278825 0.002 14.65 266_s_at CD24 1.756569076 0.004 14.45 1108_s at EPHAl 1.242309171 0.006 14.36 37483_at HDAC9 1.406744957 0.006 14.28 31887_at 1.311220827 0.011 13.68 1788_s_at DUSP4 1.22421987 0.003 13.65 ,32787_at ERBB3 0.996784565 0.02 13.21 41660 at CELSRl 1.634286803 0.004 13.11 •33483~at NMU 1.100849065 0.004 13.04 31792_ at ANXA3 0.896090153 0.011 12.90 36838~at KLKlO 1.026306829 0.02 12.71 1585_at ERBB3 1.102058608 0.011 12.51 1898_at TRIM29 1.071987353 0.002 12.44 37185_at SERPINB2 0.815945986 0.027 12.26 406_at ITGB4 1.296194559 0.006 11.66 1914_at CCNAl 0.936342778 0.011 11.21 977_s_at CDHl 0.93637461 0.036 11.19 37603_at ILlRN 1.103624942 0.015 11.14 35977_at DKKl 1.123240701 0.006 10.74 36133_at DSP 1.280269127 0.002 10.69 361 13_s_at TNNTl 1.269558595 0.002 10.19 1802_s_at ERBB2 0.787465706 0.006 9.61 2092_s_at SPPl 1.34315986 0.02 9.53 35699_at BUBlB 1.026388835 0.006 9.49 37554jit KLK6 0.895036336 0.027 9.45 38515lat BMP7 0.945367 0.027 9.32 34775_mat TSPAN-I 1.001195829 0.02 9.01 37558~at IMP-3 1.023799379 0.011 8.99 38324~at LISCH7 1.308000521 0.006 8.96 39610_at HOXB2 1.355268631 0.006 8.64 572_at TTK 1.122796615 0.006 8.53 1970_s_at FGFR2 1.022708001 0.02 8.30 160025_at TGFA 1.065272755 0.015 8.28 41812_s_at NUP210 1.39287031 0.006 8.26 34282_at NFE2L3 1.165273649 0.008 8.06 2017_s_at CCNDl 1.114984456 0.002 8.04 33323_r at SFN 1.202433185 0.008 8.01 38766_at SRCAP 1.131917941 0.008 7.99 41060_at CCNEl 1.151246634 0.006 7.97 39016_r_at KRT6E 0.973486831 0.008 7.91 31610_at MAP 17 1.0156502 0.027 7.81 2027_at S100A2 0.941919001 0.008 7.76 418 at MKI67 0.826426448 0.011 7.46 1536_at CDC6 1.08868941 0.017 7.37 634_at PRSS8 0.899891713 0.02 7.30 34342_s_at SPPl 1.318723271 0.02 7.27 182_at ITPR3 1.107167336 0.006 7.27 32382 at UPKlB 0.731294678 0.047 7.16 863j£_al SERPINB5 0.783530451 0.015 7.14

904_s_at TOP2A 0.971648429 0.02 7.12

40095_at CA2 0.798857154 0.027 7.02

41294_at KRT7 1.082553892 0.011 7.00

39951_at PLSl 0.995091449 0.006 6.94

3805 l_at MAL 0.819842532 0.036 6.82

4072όIat KIFI l 0.803689697 0.036 6.78

I 148_s_at — 0.683569558 0.047 6.72

37920_at PITXl 0.996497645 0.015 6.67

371 17_at ARHGAP8 1.129131077 0.002 6.65

38881_i_at TRIM 16 0.721698355 0.047 6.59

3425 l_at HOXB5 1.219463307 0.002 6.52

41359_at PKP3 1.047269618 0.004 6.50

40145_at TOP2A 0.961 173129 0.02 6.48

37534_at CXADR 0.888147605 0.006 6.32

40303_mat TFAP2C 0.948734146 0.004 6.30

31805~at FGFR3 0.969764101 0.01 1 6.28

33245_at MAPKl 3 0.877514586 0.01 1 6.27

885_g~at ITGA3 0.702747685 0.036 6.19

34693_at STHM 0.872525584 0.008 6.15

38555_at DUSPlO 0.880305317 0.008 6.12

38418_at CCNDl 1.071102249 0.002 5.97

33730_at RAI3 0.813298748 0.01 1 5.90

39109_at TPX2 1.040973216 0.01 1 5.87

36658jιt DHCR24 1.122129795 0.004 5.81

3528 Cat LAMC2 0.747766326 0.047 5.78

38749_at MGC29643 0.683275086 0.036 5.77

1083_s_at MUCl 0.746980491 0.027 5.75

40079_at RAI3 0.709840659 0.02 5.73

2047_s_at JUP 0.815282235 0.01 1 5.62

32275_at SLPI 0.940625784 0.02 5.61

'2020 at CCNDl 0.926408163 0.002 5.51

33324_s_at CDC2 1.026683994 0.008 5.47

36863_at HMMR 0.96343264 0.006 5.46

1657_at PTPRR 0.764510362 0.02 5.41

37985_at LMNBl 0.895475347 0.008 5.36

36497_at C14orf78 0.942921564 0.008 5.33

2021_s_at CCNEl 0.893228297 0.006 5.33

37890_at CD47 0.775908217 0.015 5.33

40799_at C16or04 0.852774782 0.008 5.30

35309_at STl 4 0.852534105 0.008 5.30

1599_at CDKN3 0.925527261 0.02 5.29

981_at MCM4 1.058558782 0.006 5.28

32715_at VAMP8 0.938171642 0.006 5.28

3863 l_at TNFAIP2 0.72369235 0.015 5.26

34715_at FOXMl 1.31035831 0.008 5.24

33448_at SPINTl 0.924028022 0.015 5.21

419_at MKI67 0.938133197 0.015 5.16

165Ϊ_at ^■UBE2C 1.436239741 0.008 5.14

35769_at GPR56 0.937347548 0.015 5.08

37310_at PLAU 0.8851 10741 0.036 5.08

3676 l_at ZNF339 0.937123503 0.01 1 5.05

37343_at ITPR3 1.001079303 0.003 5.05

40425_at EFNAl 0.813414458 0.047 5.04

1803 at CDC2 0.732852195 0.027 5.00

Validation of the Microarray Data. Quantitative RT-PCR assays were used to validate the microarray data. Four highly differentially expressed genes between OSPC and NOVA (i.e., TROP-I, CD24, Claudin-3 and Claudin-4) were selected for q-RT-PCR analysis. A comparison of the microarray and q-RT-PCR data for these genes is shown in Figure 2. Expression differences between OSPC and NOVA for TROP-I, (p = 0.002) , CD24 (p = 0.03), claudin-3 (p = 0.002) claudin-4 (p = 0.001) were readily apparent (Table 2 and Figure 2). Moreover, for all four genes tested, the q-RT-PCR data were highly correlated to the microarray data (p < 0.001 ) (r = 0.90, 0.82. 0.80 and 0.75, respectively), as estimated from the samples (i.e., 10 OSPC and 5 NOVA) included in both the q-RT-PCR and microarray experiments. The q-RT-PCR data mirror the microarray data, both qualitatively and quantitatively, and suggest that most array probe sets are likely to accurately measure the levels of the intended transcript within a complex mixture of transcripts.

TROP-I and CD24 Expression by Flow Cytometry on Primary OSPC and NOVA Cell Lines. An important issue is whether differences in gene expression result in meaningful differences in protein expression. Because TROP-I /Ep-CAM gene encodes the target for the anti-Ep-CAM antibody (17-1 A) Edrecolomab (Panorex) that has previously been shown to increase survival in patients harboring stage III colon cancer (28), expression of Ep-CAM protein by FACS analysis was analyzed on 13 primary cell lines (i.e., 10 OSPC and 3 NOVA). As positive controls, breast cancer cell lines (i.e., BT-474 and SK-BR-3, American Type Culture Collection) known to overexpress TROP-I /Ep-CAM were also studied. High TROP-I /Ep-CAM expression was found on all three primary OSPC cell lines tested (100% positive cells for all three OSPC), with mean fluorescence intensity (MFI) ranging from 116 to 280 (Figure 3). In contrast, primary NOVA cell lines were negative for TROP-I /Ep-CAM surface expression (p< 0.001) (Figure 3). Similarly, CD24 expression was found on all three primary OSPC cell lines tested (100% positive cells for all three OSPC), with mean fluorescence intensity (MFI) ranging from 26 to 55 (Figure 3). In contrast, primary NOVA cell lines were negative for CD24 surface expression (p< 0.005) (Figure 3). These results show that high expression of the TROP-I /Ep-CAM and CD24 gene products by OSPC correlate tightly with high protein expression by the tumor cells. Breast cancer positive controls were found to express high levels of TROP- 1/Ep-C AM (data not shown) .

TROP-1/Ep-CAM and CD24 Expression by Immunohistology on OSPC and NOVA Tissue Blocks. To determine whether the high (OSPC) or low (NOVA) expression of the genes and Ep-CAM and CD24 protein expression detected by microarray and flow cytometry, respectively, in primary cell lines is the result of a selection of a subpopulation of cancer cells present in the original tumor, or whether in vitro expansion conditions may have modified gene expression, immunohistochemical analysis of TROP- 1/Ep-C AM and CD24 protein expression on formalin fixed tumor tissue from all uncultured primary surgical specimens of OSPC and NOVA was performed. Heavy apical membranous staining for CD24 protein expression was noted in all OSPC specimens that also overexpressed the CD24 gene and its gene product by microarray and flow cytometry, respectively. In contrast, negative or low (i.e., score 0 or 1+) staining was found in all NOVA samples tested by immunohistochemistry. Similarly, heavy membranous staining for TROP-I /Ep-CAM protein expression (i.e., score 3+) was noted in all OSPC specimens that also overexpressed the TROP-I /Ep-CAM gene and its gene product by microarray and flow cytometry, respectively, hi contrast, negative or low (i.e., score O or 1+) staining was found in all NOVA samples tested by immunohistochemistry (data not shown).

DISCUSSION

Because of the lack of an effective ovarian cancer screening program and the common development of chemotherapy resistant disease after an initial response to cytotoxic agents (e.g., a platinum based regimen), ovarian cancer remains the most lethal among the gynecologic malignancies. Thus, identification of novel ovarian tumor markers to be used for early detection of the disease as well as the development of effective therapy against chemotherapy resistant/recurrent ovarian cancer remains a high priority.

High-throughput technologies for assaying gene expression, such as high- density oligonucleotide and cDNA microarrays, may offer the potential to identify clinically relevant genes highly differentially expressed between ovarian tumors and normal control ovarian epithelial cells (19,30,36,37,38,39,40,41). This report represents the communication of an investigation involving the genome- wide examination of differences in gene expression between primary OSPC and normal ovarian epithelial cells (NOVA), hi this study short-term primary OSPC and NOVA cultures were used (to minimize the risk of a selection bias inherent in any long term in vitro growth) to study differential gene expression in highly enriched populations of epithelial tumor cells. In this work, only the cancer cells derived from papillary serous histology tumors, which is the most common histological type of ovarian cancer, were included to limit the complexity of gene expression analysis. It was found that hierarchical clustering of the samples and gene expression levels within the samples led to the unambiguous separation of OSPC from NOVA. Of interest, the expression patterns detected in primary OSPC cells were consistently different from those seen in established serous papillary ovarian carcinoma cell lines (i.e., UCI-101 and UCI-107). These data, thus, further highlight the divergence of gene expression that occurs as the result of long-term in vitro growth. Furthermore, these data emphasize that, although established ovarian cancer cell lines provide a relatively simple model to examine gene expression, primary OSPC and NOVA cultures represent a better model system of normal and cancerous ovarian tissues in comparative gene expression analysis. OVA 2, an OSPC with mixed clear cell features (i.e., a biologically aggressive variant of ovarian cancer characterized by a poor prognosis) clustered on a sub-branch with OSPC. Because of these results, the decision was made to focus the analysis on the detection of differentially expressed genes between the two homogeneous groups of primary OSPC and NOVA. 299 genes differentially expressed between OSPC and NOVA whose average change in expression level between the two groups was at least five-fold were detected. The known function of some of these genes may provide insights in the biology of serous ovarian tumors while others may prove to be useful diagnostic and therapeutic markers against OSPC. For example, laminin gamma 2 gene was found to be the most highly differentially expressed gene in OSPC with over 46-fold up- regulation relative to NOVA. Cell migration of ovarian epithelial cells is considered essential for cell dissemination and invasion of the submesothelial extracellular matrix commonly seen in ovarian cancer. Consistent with this view, the laminin gamma 2 isoform has been previously suggested to play an important role in tumor cell adhesion, migration, and scattering of ovarian carcinoma cells (42,43,44). Thus, it is likely that the high laminin expression found in ovarian tumor cells may be a marker correlated with the invasive potential of OSPC. Consistent with this view, increased cell surface expression of laminin has been reported in highly metastatic tumors cells compared to cells of low metastatic potential (45). Importantly, previous work has also shown that attachment and metastases of tumor cells can be inhibited by incubation with antilaminin antibodies (46), or synthetic laminin peptides (47), underscoring a novel potential approach for the treatment of chemotherapy resistant ovarian cancer. TROP- 1/Ep-CAM (also called 17-1 A, ESA, EGP40) is a 40 kDa epithelial transmemebrane glycoprotein found overexpressed in several normal epithelia and in various carcinomas including colorectal and breast cancer (35). In most adult epithelial tissues, enhanced expression of Ep-CAM is closely associated with either benign or malignant proliferation. Among mammals, Ep-CAM is an evolutionarily highly conserved molecule (48), suggesting an important biologic function of this molecule in epithelial cells and tissue. In this regard, Ep-CAM is known to function as an intercellular adhesion molecule and could have a role in tumor metastasis (49). Because a randomized phase II trial with MAb CO 17- IA in colorectal carcinoma patients has demonstrated a significant decrease in recurrence and mortality of MAb- treated patients versus control patients (28), TROP-1/Ep-CAM antigen has attracted substantial attention as a target for immunotherapy for treating human carcinomas. Importantly, in this work TROP-I /Ep-CAM was found 39-fold overexpressed in OSPC when compared to NOVA. These data provide support for the notion that anti- Ep-CAM antibody therapy can be a novel, and effective, treatment option for OSPC patients with residual/resistant disease after surgical and cytotoxic therapy. Protein expression data obtained by flow cytometry on primary OPSC cell lines and by immunohistochemistry on uncultured OSPC blocks support this view.

Claudin-3 and claudin-4, two members of claudin family of tight junction proteins, were two of the top five differentially expressed genes in OSPC. These results are consistent with a previous report on gene expression in ovarian cancer (19). Although the function of claudin proteins in ovarian cancer is still unclear, these proteins likely represents a transmembrane receptor (50). Claudin-3 and claudin-4 are low- and high-affinity receptors, respectively, for CPE and are sufficient to mediate CPE binding and trigger subsequent toxin-mediated cytolysis (51). These known functions of claudin-3 and claudin-4, combined with their extremely high level of expression in OSPC suggest the use of CPE as a novel therapeutic strategy for the treatment of chemotherapy resistant disease in ovarian cancer patients.

Plasminogen activator inhibitor-2 (PAI-2), a gene whose expression has been linked to cell invasion in several human malignancies (52,53), as well as to protection from tumor necrosis factor-α (TNF-α)-mediated apoptosis (54) was found 12-fold differentially expressed in OSPC when compared to NOVA. High PAI-2 levels are independently predictive of a poor disease-free survival (55). Interestingly, a 7-fold increase in PAI-2 content was found in the omentum of ovarian cancer patients compared to the primary disease suggesting that metastatic tumors may overexpress PAI-2 (55). Other studies, however, have identified PAI-2 production as a favorable prognostic factor in epithelial ovarian cancer (56). Indeed, high PAI-2 expression in invasive ovarian tumors was limited to a group of OSPC patients which experienced a more prolonged disease free and overall survival (56). The reasons for these differences are not clear, but, as previously suggested (57), may be related at least in part to the actions of macrophage colony stimulating factor- 1 (CSF-I), a cytokine which has been shown to stimulate the release of PAI-2 by ovarian cancer cells. CD24 is a small heavily glycosylated glycosylphosphatidylinositol-linked cell surface protein, which is expressed in hematological malignancies as well as in a large variety of solid tumors (58-62). However, it is only recently that CD24 overexpression has been reported at the RNA level in ovarian cancer (39). Consistent with this recent report, in the present study CD24 gene was found 14-fold differentially expressed in OSPC when compared to NOVA. Because CD24 is a ligand of P-selectin, an adhesion receptor on activated endothelial cells and platelets, its expression may contribute to the metastasizing capacities of CD24-expressing ovarian tumor cells (63-65). Importantly, CD24 expression has recently been reported as an independent prognostic marker for ovarian cancer patient survival (64). That data combined with the present findings further suggest that this marker may delineate aggressive ovarian cancer disease and may be targeted for therapeutic and/or diagnostic purposes.

Among the genes identified herein, lipocalin-2 has not been previously linked to ovarian cancer. Lipocalin-2 represents a particularly interesting marker because of several features. Lipocalins are extracellular carriers of lipophilic molecules such as retinoids, steroids, and fatty acid, all of which may play important roles in the regulation of epithelial cell growth (66,67). In addition, because lipocalin is a secreted protein, it may play a role in the regulation of cell proliferation and survival (66,67). Of interest, two recent publications on gene expression profiling of breast and pancreatic cancer have proposed lipocalin-2 as a novel therapeutic and diagnostic marker for prevention and treatment of these diseases (66,67). On the basis of the present findings, lipocalin-2 may be added to the known markers for ovarian cancer. Osteopontin (SPPl) is an acidic, calcium-binding glycophosphoprotein that has recently been linked to tumorigenesis in several experimental animal models and human patient studies (68,69,70). Because of its integrin-binding arginine-glycine- aspartate (RDG) domain and adhesive properties, osteopontin has been reported to play a crucial role in the metastatic process of several human tumors (68,71). However, it is only recently that the upregulated expression of osteopontin in ovarian cancer has been identified (72). Importantly, because of the secreted nature of this protein, osteopontin has been proposed as a novel biomarker for the early recognition of ovarian cancer (72). In this study, the SPPl gene was found 10-fold differentially expressed in OSPC when compared to NOVA. Taken together these data confirm a high expression of osteopontin in OSPC. The organization of kallikreins, a gene family now consisting of 15 genes which all encode for trypsin-like or chymotrypsin-like serine proteases, has been recently elucidated (73). Serine proteases have been described to have well characterized roles in diverse cellular activities, including blood coagulation, wound healing, digestion, and immune responses, as well as tumor invasion and metastasis (reviewed in 73). Importantly, because of the secreted nature of some of these enzymes, prostate-specific antigen (PSA) and kallikrein 2 have already found important clinical application as prostate cancer biomarkers (73). Of interest, kallikrein.6 (also known as zyme/protease M/neurosin), kallikrein 10 and matriptase (TADG- 15/MT-SP1), were all found to be highly differentially expressed genes in OSPC when compared to NOVA. These data confirm previous results showing high expression of several kallikrein genes and proteins in ovarian neoplasms (73,74,75,76,77). Moreover, these results obtained by high-throughput technologies for assaying gene expression further emphasize the view that some members of the kallikrein family have the potential to become novel cancer markers for early diagnosis of ovarian cancer (77) as well as targets for novel therapies against recurrent/refractory ovarian disease (78). Other highly ranked genes in OSPC included stratifin, desmoplakin, S100A2, cytokeratins 6 and 7, and MUC-I.

A large number of down-regulated (at least 5-fold) genes in OSPC versus NOVA such as transforming growth factor beta receptor III, platelet-derived growth factor receptor alpha, SEMACAP3, ras homolog gene family member I (ARHI), thrombospondin 2 and disabled-2/ differentially expressed in ovarian carcinoma 2 (Dab2/DOC2) have been identified in this analysis. Some of these genes encode for widely-held tumor suppressor genes such as SEMACAP 3, ARHI, and Dab2/DOC2 (79), others for proteins important for ovarian tissue homeostasis or that have been previously implicated in apoptosis, proliferation, adhesion or tissue maintenance.

The identification of TROP- 1/Ep-CAM and CPE epithelial receptors as some of the most highly differentially expressed genes in OSPC compared to NOVA indicate that novel therapeutic strategies targeting TROP-I /Ep-CAM by monoclonal antibodies and/or claudin-3 and 4 by local and/or systemic administration of Clostridium Perfringens enterotoxin can be an effective therapeutic modalities in patients harboring OSPC refractory to standard treatment.

CONCLUSIONS

Gene expression data identified 129 and 170 genes that exhibited > 5 -fold up- regulation and down-regulation, respectively, in primary OSPC when compared to NOVA. Genes overexpressed in established OSPC cell lines were found to have little correlation with those overexpressed in primary OSPC, highlighting the divergence of gene expression that occurs as the result of long-term in vitro growth. Hierarchical clustering of the expression data readily distinguished normal tissue from primary OSPC. Laminin, claudin-3 and claudin-4, tumor-associated calcium signal transducer 1 and 2 (TROF '-1 VEp-CAM; TROP-2), ladinin 1, S100A2, SERPIN2 (PAI-2), CD24, lipocalin 2, osteopontin, kallikrein 6 (protease M) and kallih'ein 10, matriptase (TADG- 15) and stratifin were found among the most highly overexpressed genes in OSPC when compared to NOVA. Differential expression of some of these genes including claudin-3 and claudin-4, TROP-I and CD24 was validated by quantitative RT-PCR as well as by flow cytometry on primary OSPC and NOVA. Immunohistochemical staining of formalin fixed paraffin embedded tumor specimens from which primary OSPC cultures were derived further confirmed differential expression of CD24 and TROP-I /Ep-CAM markers on OSPC vs. NOVA.

Example 2.

Gene Expression Fingerprint of Uterine Serous Papillary Carcinoma: Identification of Novel Molecular Markers for Uterine Serous Cancer Diagnosis and Therapy

The goal of this study was to identify genes with a differential pattern of expression between uterine serous papillary carcinoma (USPC) and normal endometrial cells (NEC) and to use this knowledge to develop novel diagnostic and therapeutic markers for uterine cancer. Oligonucleotide microarrays were used to interrogate the expression of 10,000 known genes to analyze gene expression profiling of 10 highly purified USPC cultures and 5 primary NEC. mRNA fingerprints readily distinguished USPC from normal endometrial epithelial cells and identified a number of genes highly differentially expressed between NEC and USPC.

MATERIALS AND METHODS

Establishment of USPC and NEC Primary Cell Lines. A total of fifteen primary cell lines (i.e., 10 USPC and 5 NEC) were established after sterile processing of samples from surgical biopsies collected between 1997 and 2003 at the University of Arkansas for Medical Sciences, as previously described for USPC specimens (80) and normal endometrial cell cultures (81,82). All fresh samples were obtained with appropriate consent according to IRB guidelines. Tumors were staged according to the F.I.G.O. operative staging system. A total abdominal hysterectomy with bilateral salpingo oophorectomy and bilateral pelvic lymphadenectomy was performed in all uterine carcinoma patients while normal endometrial tissue was obtained from consenting donors undergoing surgery for benign pathology. No patient received chemotherapy or radiation before surgery. The patient characteristics are described in Table 3. Briefly, normal tissue was obtained from healthy endometria, mechanically minced and enzymatically dissociated with 0.14% collagenase Type I (Sigma, St. Louis, MO) in RPMI 1640 as described previously by Bongso et al. with minor modifications (81). After 1-2 hrs incubation with enzyme on a magnetic stirring apparatus at 37⁰C in an atmosphere of 5% CO₂, the resulting suspension was collected by centrifugation at 100 g for 5-10 minutes and washed twice with RPMI 1640 medium (Sigma) containing 10% fetal bovine serum (FBS, Invitrogen, Grand Island, NY). The final pellet was then placed in RPMI 1640 (Sigma) containing 10% FBS, 200 u/ml penicillin, and 200 μg/ml streptomycin in tissue culture flasks or Petri dishes (Invitrogen). The epithelial explants were then allowed to attach and proliferate. Explants were trypsinized and subcultured for 1 to 2 passages before being collected for RNA extraction. Tumor tissue was mechanically minced in RPMI 1640 to portions no larger than 1-3 mm³ and washed twice with RPMI 1640. The portions of minced tumor were then placed into 250 ml flasks containing 30 ml of enzyme solution [0.14% collagenase Type I (Sigma) and 0.01% DNAse (Sigma, 2000 KU/mg)] in RPMI 1640, and either incubated on a magnetic stirring apparatus for 1-2 hrs at 37⁰C in an atmosphere of 5% CO₂ or overnight at 4⁰C. Enzymatically dissociated tumor was then filtered through 150 μm nylon mesh to generate a single cell suspension. The resultant cell suspension was then washed twice in RPMI 1640 plus 10% FBS. Primary cell lines were maintained in RPMI 1640, supplemented with 10% FBS, 200 u/ml penicillin, 200 μg/ml streptomycin at 37⁰C, 5% CO₂. Tumor cells were collected for RNA extraction at a confluence of 50% to 80% after a minimum of two to a maximum often passages in vitro. The epithelial nature and the purity of USPC and NEC cultures was verified by immunohistochemical staining and flow cytometric analysis with antibodies against cytokeratin and vimentin as previously described (80,82). Only primary cultures which had at least 90% viability and contained >99% epithelial cells were used for total RNA extraction.

Table 3

Patient Age Stage

USPC 1 65 IV B

USPC 2 75 III C

USPC 3 75 IV A

USPC 4 59 IV A

USPC 5 59 m e

USPC 6 62 IV B

USPC 7 63 IH C

USPC 8 61 m e

USPC 9 78 m e

USPC l O 64 IV A

Gene Cluster/Treeview. The hierarchical clustering of average-linkage method with the centered correlation metric was used (35). For the unsupervised hierarchical clustering shown in Figure 4, a total of 7,328 probe sets were scanned across 10 USPCs and 5 NECs. The 7,328 probe sets were derived from 12,588 by filtering out all control genes, all genes with absent detections, and genes not fulfilling the test of standard deviation greater than 0.5 (0.5 being the log base 2 of the signal). Table 4 shows only genes significantly expressed by both WRS and SAM analyses and whose average change in expression level was at least five-fold. RESULTS

Gene Expression Profiles Distinguish USPC from NEC and Identify Differentially Expressed Genes. Tumor tissue flash frozen biopsies are known to contain significant numbers of contaminant stromal cells as well as a variety of host derived immune cells (e.g., monocytes, dendritic cells, lymphocytes). In addition, USPC represent rare tumors which may present in either pure forms, or admixed with endometrioid or clear cell tumor cells (i.e., mixed USPC) (9,10). To minimize the risk of contamination of USPC RNA with that of normal cells or tumor cells with different histology (i.e., endometrioid or clear cells), as well as to reduce the complexity of gene expression data analysis, in this study RNA was extracted only from short term primary tumor cell cultures collected only from USPC with single type differentiation. Short term USPC and NEC cell cultures, minimizing the risk of a selection bias inherent in any long term in vitro growth, may provide an opportunity to study differential gene expression between highly enriched populations of normal and tumor-derived epithelial cells. Accordingly, comprehensive gene expression profiles of 10 primary USPC and 5 primary NEC cell lines were generated using high-density oligonucleotide arrays with 12,533 probe sets, which in total interrogated some 10,000 genes. Characteristics of the 10 patients from whom the primary USPC cell lines were derived are shown in Table 3. Using unsupervised hierarchical cluster analysis with 7,238 probe sets, differences in gene expression between USPC and NEC were identified that readily distinguished the two groups of primary cultures. As shown in Figure 4, all 10 USPC were found to group together in the rightmost columns of the dendrogram. Similarly, in the leftmost columns all 5 NEC were found to cluster tightly together. After filtering out most "absent' genes, the SAM and the nonparametric WRS test (p < 0.05) were performed to identify genes differentially expressed between USPC and NEC. A total of 2,829 probe sets were found differentially expressed between USPC and NEC with p < 0.05 by WRS and with a median FDR of 0.35% and a 90th percentile FDR of 0.59% by SAM. Of the 2,829 aforementioned probe sets, there were 529 probe sets showing >5-fold change. As shown in Table 4, a group of 139 probe sets were found highly expressed in USPC and underexpressed in NEC. Included in this group of genes are CDKN2A/pl6/pl4ARF, Ll cell adhesion molecule (LlCAM), claudin-3 (CLDN3) and claιιdin-4 (CLDN4) , kallikrein 6 (protease M) and kallikrein 10 (NESl), interleukin-6, interleukin-18 and plasminogen activator receptor (PLAUR) (Table 4). Importantly, c- erbB2 was 14-fold more highly expressed in USPC than in NEC (Table 4). The second profile was represented by 390 genes that were highly expressed in NEC and underexpressed in USPC (data not shown). Included in this group of genes are transforming growth factor beta receptor III, platelet-derived growth factor receptor- alpha, SEMACAP3, ras homolog gene family, member I (ARHI), and differentially downregulated in ovarian carcinoma 1 (DOCI).

Table 4

Upregulated genes expressed at least 5-fold higher in USPC compared with NEC

Validation of the Microarray Data. Quantitative-RT-PCR assays were used to validate the microarray data. Seven highly differentially expressed genes between USPC and NEC (i.e., CDKN2A/pl6, CDKN2A/pl4ARF, LlCAM, claudin-3, claudin-4, GRB- 7 and c-erbB2) were selected for q-RT-PCR analysis. A comparison of the microarray and q-RT-PCR data for these genes is shown in Figure 5. Expression differences between USPC and NEC for CDKN2A/pl6 (p = 0.002), CDKN2A/pl4ARF (p = 0.002), LlCAM, (p = 0.01), claudin-3 (p = 0.01), claudin-4 (p = 0.002) GRB-7 (p = 0.002) and c-erbB2 (p = 0.01) were readily apparent (Table 4 and Figure 4). Moreover, for all seven genes tested, the q-RT-PCR data were highly correlated to the microarray data (p < 0.001) (/^• = 0.81, 0.80, 0.75, 0.69, 0.82, 0.71 and 0.65, respectively), with all the samples (i.e., 10 USPC and 5 NEC) included in both the q- RT-PCR and microarray experiments. Thus, q-RT-PCR data suggest that most array probe sets are likely to accurately measure the levels of the intended transcript within a complex mixture of transcripts.

Claudin-4 Expression by Immunohistology on USPC and NEC Tissue Blocks. To determine whether the high or low expression of the claudin-4 gene detected by microarray and q-RT-PCR assays in primary USPC and NEC cell lines, respectively, is the result of a selection of a subpopulation of cancer cells present in the original tumor, or whether in vitro expansion conditions may have modified gene expression, immunohistochemical analysis of claudin-4 protein expression on formalin fixed tumor tissue was performed on the uncultured primary surgical specimens from which the USPC cell lines were derived. Both cytoplasmic and membranous staining for claudin-4 protein expression was noted in the majority of USPC specimens (i.e., 90% score 3+ and 2+) (Table 5, micrographs not shown). In contrast, only low levels of membranous staining for claudin-4 protein was found in the NEC tissue samples tested by immunohistochemistry (Table 5, p = 0.02 USPC vs. NEC by student t test). To confirm and validate the immunohistochemical results in an independent series of USPC, formalin fixed tumor tissue blocks from 8 further surgical specimens (i.e., USPC 11 to 18, Table 5) similarly obtained from patients harboring advanced stage disease were tested for claudin-4 expression. Again, heavy cytoplasmic and/or membranous staining for the claudin-4 receptor was found in the striking majority of the further USPC sample tested.

Table 5

Claudin-4 staining

Claudin-4

Patient positivity

NEC 1 1+

NEC 2 1 +

NEC 3 1+

NEC 4 1+

NEC 5 1+

USPC l 3+

USPC 2 3+

USPC 3 3+

USPC 4 2+

USPC 5 3+

USPC 6 2+

USPC 7 3+

USPC 8 1+

USPC 9 3+

USPC 10 3+

USPC 11 3+

USPC 12 3+

USPC 13 1+

USPC 14 2+

USPC 15 3+

USPC 16 2+

USPC 17 3+

USPC 18 2+

DISCUSSION

Hierarchical clustering of the samples and gene expression levels within the samples led to the unambiguous separation of USPC from NEC. This study identified 529 genes differentially expressed between USPC and NEC whose average change in expression level between the two groups was at least five-fold and which were found significant with both WRS test and SAM analysis. The known function of some of these genes may provide insights in the molecular pathogenesis and the highly aggressive biologic behavior of uterine serous tumors while others may prove to be useful diagnostic and therapeutic markers against this disease.

For example, the cyclin-dependent kinase inhibitor 2A (CDKN2A) gene was found to be the most highly differentially expressed gene in USPC with over 101-fold up-regulation relative to NEC. Importantly, the CDKN2A gene is a putative oncosuppressor gene encoding two unrelated proteins, both cellular growth inhibitors, in different reading frames (83). One is pi 6, which regulates retinoblastoma protein (pRb)- dependent Gl arrest, and the second is pl4ARF, which blocks MDM2-induced p53 degradation, resulting in an increase in p53 levels and consequent cell cycle arrest (83). Although loss of p53 function is considered critical for the molecular pathogenesis of USPC (84,85), it is only recently that abnormality of the Rb pathway has been suggested to define a subgroup of aggressive endometrial carcinomas with poor prognosis (86). Quantitative RT-PCR analysis of expression of both pl6 and pi 4ARF in the USPC series found extremely high levels of both transcripts, suggesting that the marked overexpression of the CDKN2A gene may be attributable to a negative feedback loop due to the loss of function of both pRb and p53 proteins (84,85,86). Consistent with this view, an inverse relationship between the expression of pi 6 and pi 4ARF proteins and the presence of normal or functional Rb and p53 in human cancer cells has been previously demonstrated (87,88). Thus, these data suggest for the first time that CDKN2 A gene overexpression may represent a consistent genetic anomaly of USPC secondary to an autoregulatory feedback loop due to disruption of both the pi 6- CDK4/cyclin Dl-pRb pathway and the pl4ARF-MDM2-p53 pathway.

Among the genes identified here, lipocalin-2 has not been previously linked to uterine cancer. Lipocalins are extracellular carriers of lipophilic molecules such as retinoids, steroids, and fatty acid, all of which may play important roles in the regulation of epithelial cell growth (70,71). In addition, because lipocalin is a secreted protein, it may play a role in the regulation of cell proliferation and survival (70,71). Of interest, two recent publications on gene expression profiling of breast and pancreatic cancer have proposed lipocalin-2 as a novel therapeutic and diagnostic marker for prevention and treatment of these diseases (89,90). On the basis of the present findings, lipocalin-2 may be added to the known markers for USPC.

Among the several potential therapeutic target gene products identified, genes encoding tight junction (TJ) proteins claudin-3 and claudin-4 were consistently found as two of the most highly up-regulated genes in USPC, with over 8 and 12-fold up- regulation, respectively, relative to NEC. Although the exact function of claudin-3 and claudin-4 in USPC is still unclear, these proteins have been shown to represent the epithelial receptors for Clostridium perfringens enterotoxin (CPE), and to be the only family members of the transmembrane tissue-specific claudin proteins capable of mediating CPE binding and cytolysis. Protein expression data obtained by immunohistochemistry with anti-claudin-4 antibody on USPC blocks further support the proposal that claudins may represent therapeutic targets.

The organization of kallikreins, a gene family now consisting of 15 genes that encode for trypsin-like or chymotrypsin-like serine proteases, has been recently elucidated (77). Serine proteases have well characterized roles in diverse cellular activities, including blood coagulation, wound healing, digestion, and immune responses, as well as tumor invasion and metastasis (reviewed in 77). Secreted serine proteases such as prostate-specific antigen (PSA) and kallikrein 2 have already found important clinical application as prostate cancer biomarkers (77). Of interest, kallikrein 6 (also known as zyme/protease M/neurosin), and kallikrein- 10 (NES 1 ), two serine proteases recently shown to be present at high levels in the circulation of a subset of ovarian cancer patients (80,81,91), were both highly differentially expressed genes in USPC when compared to NEC. Kallikrein 6 and kallikrein 10 overexpression has been shown to correlate with intrinsic resistance to adjuvant chemotherapy and with a poor prognosis in ovarian cancer patients (81,014). These data are thus consistent with the present results showing high expression of kallikreins 6 and kallikrein 10 in USPC, a variant of endometrial carcinoma characterized by an aggressive biologic behavior and an inborn resistance to chemotherapy (7,8,9,10,12). c-erbB2 gene was found to be one of the most highly differentially expressed genes in USPC with over 14-fold up-regulation compared with NEC. Furthermore, the growth factor receptor-bound protein 7 (GRB7), a gene tightly linked to c-erbB2 and previously reported coamplified and coexpressed with this gene in several cancer types (92) was also highly differentially expressed in USPC compared to NEC. These data confirm our recent discovery of a striking overexpression of the c-erbB2 gene product HER2/neu on 80% of pure USPC (93). Thus, HER2/neu overexpression may represent a distinctive molecular marker that in addition to having the potential to facilitate differentiation of USPC from the histologically indistinguishable high grade serious ovarian tumors, may also provide insights into the disproportionately poor prognosis of USPC patients (94,95,96,97). Previous studies have reported that HER2/neu overexpression in USPC patients may be associated with resistance to chemotherapeutic drugs and shorter survival (96,97). However, high overexpression of the c-erbB2 gene on USPC provides support for the notion that trastuzumab (HERCEPTIN™, Genentech, San Francisco, CA), a humanized anti-HER-2/Neu 5 antibody that is showing great promise for treatment of metastatic breast cancer patients overexpressing HER-2/Neu protein (98), may be a novel, potentially highly effective therapy against USPC. Consistent with this view, high sensitivity of USPC natural killer (NK) cell-mediated antibody-dependent cytotoxicity triggered by anti- HER-2/Neu-specifϊc antibody in vitro (93), as well as clinical responses in vivo (99) l o have recently been reported with the use of Herceptin in USPC patients.

Ll adhesion molecule (LlCAM), is a 200-220 kD type I membrane glycoprotein of the immunoglobulin family which plays an important function for development of the nervous system by regulating cell adhesion and migration (100). In addition, LlCAM has been recently reported to be expressed on a variety of human

15 tumor cell lines such as neuroblastomas, melanomas, and lung carcinomas

(101,102,103). Because overexpression of LlCAM by tumor cells may enhance cell migration on various extracellular membrane substrates, this molecule has been suggested to play a role in the adhesion and migration events crucial for tumor spreading (100,101). In the present analysis, LlCAM was found to be one of the most

20 highly differentially expressed genes in USPC, with over 25-fold up-regulation relative to NEC. These data are in agreement with a recent report showing high expression of LlCAM in a subset of ovarian and uterine cancers (104). Importantly, patients with Ll -positive carcinomas had a poorer prognosis and shorter survival time compared with patients whose tumors were Ll-negative (104), suggesting a direct

25 correlation between L 1 CAM overexpression and aggressive biologic behavior. Of interest, however, high levels of soluble LlCAM molecules resulting from cleavage of its ectodomain by the metalloproteinase ADAMlO (105) have been detected in serum samples of patients with ovarian and uterine tumors (104). These observations, combined with the results herein, support the use of LlCAM as a novel biomarker for

30 prediction of clinical outcome in USPC patients.

Several other highly ranked genes have been identified in this USPC gene expression profiling analysis including membrane-associated protein 17 (MAP 17), galanin, urokinase plasminogen activator receptor (UPAR), interleukin-6, forkhead box Ml, interleukin-18, dickkopf homolog 1 (DKKl), coagulation factor II (thrombin) receptor-like 1, transforming growth factor, alpha, interleukin 8, topoisomerase (DNA) II alpha, hyaluronan-mediated motility receptor (RHAMM) and secretory leukocyte protease inhibitor (antileukoproteinase). For most of the genes found differentially expressed in these experiments a correlation with USPC cancer development and progression has not been recognized before. DKKl, for example, has been recently reported by our group to play an important role in the development of osteolytic lesions in multiple myeloma (106), but its possible role in USPC pathogenesis and/or progression has not been elucidated. For other genes such as UPAR, a glycosyl-phosphatidylinositol-anchored glycoprotein whose role in promoting tumor cell invasion and metastases has been well established in a number of experimental studies (107,108,109), a correlation with high expression in the uterine serous papillary carcinoma phenotype has been recently reported (110). In this regard, because UPAR protein exists in two forms, as the glycosyl- phosphatidylinositol-anchored glycoprotein (50-60 kDa) present on the surface of cells, and as a soluble form of UPAR (sUPAR), produced after cleavage of UPAR by urokinase (35 kDa), measurement of UPAR levels by ELISA, in analogy to breast cancer (111), can be used as a prognostic marker to identify early recurrences in endometrial cancer patients associated with poor outcome. Finally, the recent demonstration of UPAR as a suitable cancer target for both therapeutic and diagnostic application by specific antibody directed against its ligand binding domain (112) may provide a foundation for the development of a new type-specific therapy against this highly aggressive disease.

A large number of down-regulated (at least 5 -fold) genes in USPC versus NEC such as transforming growth factor beta receptor III, platelet-derived growth factor receptor alpha, SEMACAP3, ras homolog gene family member I (ARHI), and differentially downregulated in ovarian carcinoma I (DOCl) have been identified in this analysis. Some of these genes encode for widely-held tumor suppressor genes such as SEMACAPS, ARHI, and DOCl (113), others for proteins important for tissue homeostasis or that have been previously implicated in apoptosis, proliferation, adhesion or tissue maintenance.

In conclusion, multiple USPC restricted markers have been identified in this analysis. Most of these genes have not been previously linked with this disease and thus represent novel findings. The identification of HER2/neu and CPE epithelial receptors among the others as some of the most highly differentially expressed genes in USPC when compared to NEC indicate that therapeutic strategies targeting HER2/neu by monoclonal antibodies (93,99) and claudin-3 and claudin-4 by local and/or systemic administration of Clostridium Perfringens enterotoxin can be effective modalities for the treatment of patients harboring this highly aggressive and chemotherapy resistant variant of endometrial cancer.

CONCLUSIONS

Unsupervised analysis of mRNA fingerprints readily distinguished USPC from NEC and identified 139 and 390 genes that exhibited > 5-fold up-regulation and down-regulation, respectively, in primary USPC when compared to NEC. Many of the genes up-regulated in USPC were found to represent adhesion molecules, secreted proteins and oncogenes, such as Ll cell adhesion molecule (LlCAM), claudin-3 and claudin-4, kallikrein 6 (protease M) and kallikrein 10, (NESl), interleukin-6, interleukin-18, urokinase plasminogen activator receptor (UPAR), and c-erbB2. Down-regulated genes in USPC included transforming growth factor beta receptor III, platelet-derived growth factor receptor alpha, SEMACAP3, ras homolog gene family, member I (ARHl), and differentially downregulated in ovarian carcinoma gene 1 (DOCl). Quantitative RT-PCR was used to validate differences in gene expression between USPC and NEC for some of these genes, including cyclin-dependent kinase inhibitor 2A (CDKN2A/pl 6 and CDKN2A/pl4ARF), LlCAM, claudin-3, claudin-4, GRB-7 and c-erbB2. Finally, expression of the high affinity epithelial receptor for Clostridium perfringens enterotoxin (CPE) claudin-4, was further validated through immunohistochemical analysis of formalin-fixed paraffin-embedded specimens from which the primary USPC cultures were obtained, as well as an independent set of archival specimens.

Example 3.

Treatment of Chemotherapy-Resistant Human Ovarian Cancer Xenografts in CB- 17/SCID Mice by Intraperitoneal Administration of Clostridium Perfringens Enterotoxin (CPE)

In this study, real-time PCR was used to quantify the expression levels of claudin-3 and claudin-4 receptors in several chemotherapy naive and chemotherapy resistant freshly explanted ovarian tumors, hi addition, the ability of CPE to kill chemotherapy sensitive and chemotherapy resistant ovarian cancers overexpressing claudin-3 and/or claudin-4- in vitro was tested. Finally, the in vivo efficacy of CPE therapy in SCID mouse xenografts was studied in a highly relevant clinical model of chemotherapy resistant freshly explanted human ovarian cancer (i.e., OVA-I).

MATERIALS AND METHODS

Cloning and Purification Of NH₂ Terminus His-tagged CPE. Clostridium perfringens strain 2917 obtained from American Type Culture Collection (Manassas, VA) was grown from a single colony and used to prepare bacterial DNA with the INSTAGENE™ kit, according to manufacturer's directions (Bio-Rad Lab, Hercules, CA). The bacterial DNA fragment encoding full-length CPE gene (GenBank AJ000766) was PCR amplified (primer 1, 5'-AGA TGT TAA TGG ATC CAT GCT TAG TAA CAA TTT AAA TCC-3¹; primer 2, 5'-AAA GCT TTT AAA ATT TTT GAA ATA ATA TTG AAT AAG GG-3'). The PCR products were digested with the restriction enzymes Sphl/Hindlll and cloned into a Sphl/Hindlll-digested pQE-30 expression vector (Qiagen, Valencia, CA) to generate an in-frame NH₂-terminus His-tagged CPE expression plasmid, pQE-30-6xHIS-CPE. His-tagged CPE toxin was prepared from pQE-30-6xHIS-CPE transformed Escherichia coli M15. Transformed bacteria were grown at 37°C to 0.3-0.4 optical density at 600 nm, after which CPE protein expression was induced overnight with 1 mM isopropyl β-D-thio-galactoside, and the cells harvested, resuspended in 150 mM NaH₂PO₄, 25 mM Tris-HCL, and 8 M urea pH 8.0 buffer, and lysed by centrifugation at 10.000 rpm for 30 min. The fusion protein was isolated from the supernatant on a POLY-PREP™ Chromatography column (Bio-Rad). His-tagged CPE was washed with 300 mM NaH₂PO₄, 25 mM Tris-HCl, and 10 M urea pH 6.0, and eluted from the column with 200 mM NaH₂PO₄, 25 mM Tris-HCl, and 8 M urea pH 6.0. In order to reduce the level of endotoxin from His-tagged CPE protein 10 washings with ice-cold PBS with TRITON-X-114™ (from 1% to 0.1%) and 10 washings with ice-cold PBS alone were performed. Dialysis (M_r 3,500 cutoff dialysis tubing) against PBS was performed overnight. Purified CPE protein was then sterilized by 0.2 μm filtration and frozen in aliquots at -7O⁰C.

Primary and Established Cell Lines. Fresh human ovarian cancer cell lines (i.e., eleven chemotherapy naϊve tumors generated from samples obtained at the time of primary surgery and six chemotherapy resistant tumors obtained from samples collected at the time of tumor recurrence), and 5 established ovarian cancer cell lines (UCI 101, UCI 107, CaOV3, OVACAR-3, OVARK-5) were evaluated for claudin-3 and claudin- 4 expression by real time-PCR. Patient characteristics from which primary specimens were obtained are depicted in Table 6. Three of the six ovarian tumor specimens found resistant to chemotherapy in vivo, including 0,VA-I , a fresh ovarian serous papillary carcinoma (OSPC) used to establish ovarian xenografts in SCID mice, were confirmed to be highly resistant to multiple chemotherapeutic agents when measured as percentage cell inhibition (PCI) by in vitro Extreme Drug Resistance (EDR) assay (Oncotech Inc. Irvine, CA) (114) (data not shown). UCI-101 and UCI-107, two previously characterized and established human serous ovarian cancer cell lines were kindly provided by Dr. Alberto Manetta, University of California, Irvine, while CaO V3, OVACAR-3, were purchased from ATCC (Manassas, VA), and OVARK-5 was established from a stage FV ovarian cancer patient, as previously described (115). Other control cell lines evaluated in the CPE assays included Vero cells, normal ovarian epithelium (NOVA), normal endometrial epithelium, normal cervical keratinocytes, primary squamous and adenocarcinoma cervical cancer cell lines, Epstein-Barr transformed B lymphocytes (LCL), and human fibroblasts. With the exception of normal cervical keratinocytes and cervical cancer cell lines which were cultured in serum-free keratinocyte medium, supplemented with 5 ng/ml epidermal growth factor and 35 to 50 μg/ml bovine pituitary extract (Invitrogen, Grand Island,

N. Y.) at 37⁰C, 5% CO2, all other fresh specimens were cultured in RPMI 1640 medium (Invitrogen) containing 10% fetal bovine serum (FBS; Gemini Bio-products, Calabasas, CA), 200 u/ml penicillin, and 200 μg/ml streptomycin, as previously described (115,19,116). All samples were obtained with appropriate consent according to IRB guidelines. Tumors were staged according to the F.I.G.O. operative staging system. Radical tumor debulking, including a total abdominal hysterectomy and omentectomy, was performed in all ovarian carcinoma patients while normal tissues was obtained from consenting similar age donors undergoing surgery for benign pathology. Tumors were established after sterile processing of the samples from surgical biopsies as previously described for ovarian and cervical carcinoma specimens (115,19,116), while normal ovarian tissue was obtained by scraping epithelial cells from the ovarian surface. Briefly, viable tumor tissue was mechanically minced in RPMI 1640 to portions no larger than 1-3 mm³ and washed twice with RPMI 1640. The portions of minced tumor were then placed into 250 ml flasks containing 30 ml of enzyme solution [0.14% collagenase Type I and 0.01% DNAse 2000 KU/mg; (Sigma)] in RPMI 1640, and incubated on a magnetic stirring apparatus overnight at 4° C. Enzymatically dissociated tumor was then filtered through 150 μm nylon mesh to generate a single cell suspension. The resultant cell suspension was then washed twice in RPMI 1640 plus 10% FBS. The epithelial nature and the purity of epithelial tumor cultures was verified by immunohistochemical staining and flow cytometric analysis with antibodies against cytokeratin as previously described (115,19,116). RNA extraction was performed at a tumor cell confluence of 50% to 80% after a minimum of two to a maximum often passages in vitro. Only primary cultures which had at least 90% viability and contained >99% epithelial cells were used for sensitivity to CPE in vitro.

Table 6

*: Patients from which matched chemotherapy naive and chemotherapy resistant/recurrent disease were both available. OVA-R: Patients with chemotherapy resistant/recurrent disease. OSPC: ovarian serous papillary carcinoma; CC: clear cell ovarian carcinoma.

CPE Treatment of Cell Lines and Trypan Blue Exclusion Test. Tumor samples and normal control cells were seeded at a concentration of 1 x 10⁵ cells/well into 6-well culture plates (Costar, Cambridge, MA) with the appropriate medium. Adherent tumor samples, fibroblasts and normal epithelial control cell lines were grown to 80% confluence. After washing and renewal of the medium, CPE was added to final concentrations ranging from 0.03 to 3.3 μg/mL. After incubation for 60 minutes to 24 hrs at 37⁰C, 5% CO₂, floating cells were removed and stored, and attached cells were trypsinized and pooled with the floating cells. After staining with trypan blue, viability was determined by counting the number of trypan blue-positive cells and the total cell number.

SCID Mouse Tumor Xenografts and CPE Treatment. C.B-17/SCID female mice 5-7 weeks old (16-18 g in weight) were obtained from Harlan Sprague-Dawley (Indianapolis, IN) and housed in a pathogen-free environment at the University of Arkansas for Medical Sciences (UAMS). They were given commercial basal diet and water ad libitum. The experimental protocol for the use of these animals for these studies was approved by the UAMS Institutional Animal Care and Use Committee. Animals were used to generate ovarian tumor xenografts. The OVA-I cancer cell line was injected intraperitoneally (i.p.) at a dose of 5 to 7.5 x 10⁶into C.B-17/SCID mice in groups of five. hi the first set of experiments (i.e., large ovarian tumor burden challenge), four weeks after i.p. tumor injection, mice were injected i.p. with 5.0, 5.5, 6.5 μg of CPE dissolved in 1 ml of sterile saline at 72 hrs intervals. In a second set of experiments, groups of five mice received 7.5 μg or 8.5 μg of CPE i.p. at 72 hrs intervals 1 week after i.p. OVA-I tumor injection at a dose of 5 x 10⁶ tumor cells. AU animals were observed twice daily and weighed weekly, and survival was monitored. In addition, groups of mice injected i.p. at a dose of 5 to 7.5 x 10⁶ OVA-I tumor cells were killed at one, two, three and four weeks for necropsy and pathologic analysis. The remaining animals were killed and examined just before they died of intraperitoneal carcinomatosis and malignant ascites. Statistics. Statistical differences in claudin-3 and claudin-4 expression between chemotherapy naϊve and chemotherapy recurrent/resistant ovarian tumors were tested using the student t test. For the OVA-I animal model, survivals were plotted using

Kaplan-Meier methods and compared using the log-rank test. A p value less than 0.05 (p < 0.05) was used for statistical significance.

RESULTS Claudin-3 and Claudin-4 Transcript Levels in Chemotherapy Sensitive and Chemotherapy Resistant Ovarian Tumors. Quantitative RT-PCR assays were used to get highly sensitive measurements of claudin-3 and claudin-4 expression in normal tissues and fresh and established human tumors. Both claudin-3 and/or claudin-4 genes were highly expressed in all primary ovarian cancers studied when compared with normal ovarian epithelial cells as well as other normal cells or other gynecologic tumors (Figure 6). Of interest, established ovarian cancer cell lines (UCI 101, UCI 107, CaOV3₅ OVACAR-3, OVARK-5), were found to express much lower levels of claudin-3 and/or claudin-4 compared to primary ovarian tumors (Figure 6). Finally, claudin-3 and/or claudin-4 expression was extremely low in all control tissues examined including normal ovarian epithelium, normal endometrial epithelium, normal cervical keratinocytes, and normal human fibroblasts (Figure 6).

When OSPC collected at the time of primary debulking surgery (6 cases) were compared for claudin-3 and/or claudin-4 receptor expression to those collected at the time of tumor recurrence after multiple courses of chemotherapy (6 cases), chemotherapy resistant tumors were found to express significantly higher levels of claudin-3 and/or claudin-4 receptors (p < 0.05, Figure 7). Importantly, when 3 primary ovarian cancers naϊve to chemotherapy were compared to recurrent ovarian cancers recovered from the same patients following chemotherapy (i.e., matched autologous tumor samples), chemotherapy resistant tumors were again found to express higher levels of claudin-3 and claudin-4 (Figure 7).

Claudin-3 and Claudin-4 Transcript Levels in Metastatic and Primary Tumors. Claudin-3 and claudin-4 mRNA copy numbers were quantified in an OVA- 1 primary tumor and in a metastatic ovarian tumor (omentum metastasis) later taken from the same patient. The results are shown in Figure 8. Transcript levels for both claudin-3 and claudin-4 were more than twice as high in the metastatic cells as in the primary tumor.

Claudin-4 Expression by Immunohistology on OSPC and NOVA Tissue Blocks. To determine whether the high expression of the claudin-4 gene detected by q-RT-PCR assays in primary ovarian cancer cell lines is the result of a selection of a subpopulation of cancer cells present in the original tumor, or whether in vitro expansion conditions may have modified gene expression, immunohistochemical analysis of claudin-4 protein expression was performed on formalin fixed tumor tissue from the uncultured primary surgical specimens from which fresh ovarian cancers were derived. As shown in Table 7 , moderate to heavy membranous staining for claudin-4 protein expression was noted in all the cancer specimens that overexpressed the claudin- 4 transcript, hi contrast, negative or low staining was found in all the normal ovarian epithelium tested by immunohistochemistry (Table 7). Table 7

Claudin-4 staining

Patient Claudin-4 positivity

NOVA 1 1 +

NOVA 2 1+

OVA 1 3+

OVA 2 3+

OVA 3 3+

OVA 4 2+

OVA 5 3+

OVA 6 2+

OVA 7 3+

OVA 8 2+

OVA 9 3+

OVA 10 3+

OVA 11 3+

Effects of CPE on Fresh Ovarian and Cervical Cancer Cell Lines and Normal Control Cells. On the basis of the high expression of claudin-3 and/or claudin-4 on primary ovarian cancer cell lines, it was suspected that ovarian tumors expressing either claudin-3 or claudin-4 would be sensitive to CPE-mediated lysis. It was important to demonstrate this directly on fresh human ovarian carcinoma cells, particularly in a clinically relevant setting of ovarian cancer disease for which current salvage therapies are ineffective (i.e., chemotherapy resistant disease). For this reason this study examined short term in vitro primary cultures of ovarian carcinomas obtained either from chemotherapy naive patients (i.e., OVA-2, OVA-3 and OVA-5) or patients heavily treated with different combinations of chemotherapy (i.e., OVA-I, OVA-4 and OVA-6) and now with disease progression after multiple chemotherapy regimens. The sensitivity of these primary ovarian tumor cultures to CPE-mediated cytolysis was tested along with an appropriate claudin-3 and claudin-4-expressing positive control (i.e., Vero cells), established OSPC cell lines (OVARK-5, CaOV3 and OVACAR-3), and negative controls which do not express detectable levels of either claudin-3 or claudin-4. As shown in Figure 9, regardless to their sensitivity or resistance to chemotherapy all ovarian tumors tested were found sensitive to CPE- mediated cytolysis. The cytotoxic effect was dose dependent and was positively correlated to the levels of either claudin-3 or claudin-4 expression as tested by RT- PCR in tumor samples, importantly, although ovarian tumors demonstrated different sensitivities to CPE exposure, no ovarian cancer was found viable after 24 hrs exposure to CPE at the concentration of 3.3 μg/mL. In contrast, all normal controls tested including ovarian epithelium, cervical keratinocytes and mononuclear cells as well as cervical cancer cell lines lacking claudin-3 or claudin-4 were not affected by CPE (Figure 9).

Effect of CPE on Chemotherapy Resistant Ovarian Cancer Cells in Vivo

For in vivo confirmation of the in vitro data, xenograft tumors in SCID mice were developed by i.p. injection with OVA-I, a primary ovarian tumor resistant to multiple chemotherapeutic agents in vitro (by EDR assay) as well as in vivo. Primary OVA-I tumor cells grew progressively as numerous serosal nodules adherent to virtually all intraabdominal organs (peritoneum, omentum, diaphragm, bowel, liver, pancreas, spleen) and exhibited the capacity for local tissue invasion and formation of malignant ascites after 2 to 3 weeks from injection. Tumors first appeared grossly by the second week as small nodules on the omentum and continuously grew to form a confluent omental mass by the time the animals died (i.e., mean survival 38 days after i.p. injection with 7.5 x 10⁶ OVA-I cells). Necropsies revealed massive hemorrhagic ascites and numerous tumor nodules, measuring 1 to 8 mm in diameter, studding the entire peritoneal surface and implanting the serosa of virtually all intraabdominal organs. Previous toxicology studies in mice have reported 0.5 μg/g of CPE administered i.p. to be a well tolerated and safe dose in 100% of the animals (i.e., 16.5 + 1.O g male SW mice) (117,118). In contrast, some animals injected with 0.75 μg/g died after CPE injection and all animals injected with 1 μg/g of CPE died within 1 to 2 hrs (117, 118). Maximum tolerated dose in healthy female mice was determined and is consistent with these observations (data not shown). In one experiment, groups of mice harboring large ovarian tumor burden xenografts (i.e., four weeks after OVA-I tumor injection) were treated with repeated i.p. CPE injections every 72 hrs at three different doses (5.0 μg, 5.5 μg or 6.5 μg). Control mice harboring OVA-I received saline alone. CPE injections were well tolerated, and no adverse events were observed throughout the complete treatment protocol either in control mice receiving CPE alone or CPE-treated mice harboring large tumor burden. Mice harboring OVA-I treated with saline all died within 6 weeks from rumor injection with a mean survival of 38 days (Figure 10A). In contrast, animals treated with multiple CPE injections survived significantly longer than control animals did (p < 0.0001, Figure 10A). The increase in survival in the different groups of mice treated with the diverse doses of CPE was clearly dose dependent, with the highest dose injected (i.e., 6.5 μg every 72 hrs) found to provide the longer survival (Figure 10A). In another set of experiments, mice harboring OVA-I (a week after tumor injection with 5 x 10⁶ cells) were treated with i.p. CPE injections at a dose ranging from 7.5 μg to 8.5 μg every 72 hrs. While mice harboring OVA-I treated with saline all died within 9 weeks from tumor injection (Figure 10B), three out of five (60%) of mice receiving 7.5 μg doses of CPE and five out of five (100%) of mice receiving 8.5 μg doses of CPE by multiple i.p. injections remained alive and free of detectable tumor for the duration of the study period (i.e., over 120 days, p < 0.0001).

DISCUSSION:

In this study, after having confirmed at both the RNA and protein levels the expression of high levels of CPE receptors in multiple primary ovarian cancers and tested the in vitro sensitivity of tumor cells to CPE therapy, the efficacy and toxicity of i.p. injection of CPE in vivo in a clinically relevant animal model of chemotherapy resistant ovarian tumor xenografts were evaluated.

100% (seventeen out of seventeen) of the primary ovarian cancer cell lines tested for claudin-3 and claudin-4 expression by quantitative RT-PCR overexpress either the high affinity CPE receptor (claudin-4) or the low affinity CPE receptor (claudin-3). Of interest, all the established ovarian cancer cell lines tested (UCI 101, UCI 107, CaOV3, OVACAR-3, OVARK-5), were found to express much lower levels of claudin-3 and/or claudin-4 compared to primary ovarian tumors. These data suggest that prolonged in vitro culture may significantly alter claudin-3 and claudin-4 gene expression in ovarian cancer. In addition, q-RT-PCR showed a consistent downregulation of claudin-3 and claudin-4 expression levels in the more advanced in vitro passages of primary OSPC (data not shown). Thus, established ovarian cancer cell lines may represent suboptimal models to evaluate the potential of CPE-mediated therapy against ovarian cancer in vitro as well as in vivo. Importantly, all primary ovarian tumors evaluated, including those found to be resistant to chemotherapy in vivo as well as in vitro, were found highly sensitive to CPE-mediated killing in vitro, hi this regard, although ovarian tumors demonstrated different sensitivity to CPE exposure, no ovarian cancer was found viable after 24 hrs exposure to CPE at the concentration of 3.3 μg/mL, a dose well tolerated by in vivo i.p. administration of CPE in these experiments. This was in strong contrast with the lack of sensitivity of normal ovarian epithelium as well as other normal control cells to CPE-mediated cytolysis. These findings are likely explained by a limited expression of claudin-3 and claudin-4 in normal epithelia compared to ovarian tumor cells.

When the efficacy of multiple i.p. injection of sublethal doses of CPE in vivo in a clinically relevant animal model of chemotherapy resistant ovarian tumor xenografts was tested, doses of CPE ranging from 5 to 8.5 μg were well tolerated, and no adverse events were observed throughout the complete treatment protocol either in control mice receiving CPE alone or CPE-treated mice harboring small and large tumor burden. These data demonstrate that CPE doses found effective in vitro to kill ovarian tumor cells may be safely administered i.p. in mice harboring ovarian cancer disease. More importantly, survival of mice harboring a large burden of chemotherapy resistant ovarian disease was significantly prolonged in a dose dependent manner by repeated i.p. injections of CPE. Finally, when animals harboring 1 week OVA-I xenografts were treated with repeated i.p. injections of CPE, most of the mice remained alive and free of detectable tumor for the duration of the study (i.e., over 120 days). Collectively, these results provide strong evidence that CPE-based therapy can be useful in the treatment of ovarian cancer patients refractory to standard treatment modalities.

CPE probably must distribute through a tumor by passive diffusion. So intraperitoneal CPE administration would likely be most effective in patients with microscopic residual disease or small tumor burden.

CPE-mediated cytolysis requires only the single step of CPE binding to its receptors and takes place after only a few minutes of tumor cell exposure to toxic CPE concentrations. Thus, the simplicity and rapidity of CPE-mediated cytolysis may result in increased efficacy, reduced opportunity for the development of resistance, and the possibility that high local concentrations of CPE may need to be maintained for only a relatively short period of time. Furthermore, because the efficacy of CPE therapy against chemotherapy resistant ovarian cancer xenografts has been demonstrated in SCID mice (i.e., severely immunocompromised animals), it seems unlikely that the host immune system is required to play a significant role in the in vivo efficacy of CPE therapy. This point is noteworthy because several biologic response modifiers previously used for the therapy of chemotherapy resistant ovarian cancer, including cytokines (119) and humanized monoclonal antibodies (120), unlike CPE therapy, rely for most of their efficacy on the activation of an uncompromised host immune system (121), a major limitation when dealing with elderly ovarian cancer patients heavily pretreated with multiple regimens of immunosuppressive chemotherapy.

In conclusion, it is shown that primary and metastatic ovarian cancers that have acquired in vivo resistance to chemotherapeutic drugs are susceptible to killing by CPE- mediated therapy in vitro as well as in vivo.

CONCLUSIONS

100% (17 out of 17) of the primary ovarian tumors tested overexpressed one or both CPE-receptors by quantitative RT-PCR. All ovarian tumors demonstrated a dose- dependent cytotoxic effect to CPE in vitro. Importantly, chemotherapy resistant/recurrent ovarian tumors were found to express cIaudin-3 and claudin-4 genes at significantly higher levels when compared to chemotherapy naϊve ovarian cancers and all primary ovarian tumors tested, regardless of their resistance to chemotherapeutic agents, died within 24 hrs to the exposure to 3.3 μg/mL of CPE in vitro. Metastatic OVA-I tumors were also found to express more claudin-3 and claudin-4 than primary OVA-I tumors. In addition, the in vivo efficacy of intraperitoneal (i.p.) CPE therapy in SCID mouse xenografts was studied in a highly relevant clinical model of chemotherapy resistant freshly explanted human ovarian cancer (i.e., OVA-I). Repeated i.p. administration of sublethal doses of CPE every 3 days significantly inhibited tumor growth in 100% of mice harboring 1-week established OVA-I . Repeated i.p. doses of CPE also had a significant inhibitory effect on tumor progression with extended survival of animals harboring large ovarian tumor burdens (i.e., 4-weeks established OVA-I).

Example 4 Production of a chimeric antibody containing a claudin-binding peptide of CPE.

In this Example, production of a chimeric antibody having residues 290-319 of SEQ ID NO: 1 (CPE), which is a peptide of CPE that binds to claudin-3 and claudin-4, conjugated to the Fc portion of IgGl. A nucleic acid segment encoding residues 290-319 of SEQ ID NO:1 is cloned into vector _PFUSEhFc2(IL2ss) (SEQ ID NO:2) (INVIVOGEN company, www.invivogen.com).

Two single-stranded oligonucleotides 30-CPE-forward (SEQ ID NO:3) and 30 CPE-reverse (SEQ ID NO:4) are synthesized. Annealing the two strands produces a double-stranded DNA segment with a blunt end and an Ncol-compatible sticky end (a CATG-5' overhang). pFUSEhFc2 is digested with EroRV and Ncol. The hybridized insert is ligated into the double-digested pFUSEhFc2. The result is a vector that expresses a protein with SEQ ID NO:5. Residues 1 -21 of SEQ ID NO:5 are a signal peptide of interleukin-2 (ssIL2) for secretion of the expressed protein in mammalian cells. The signal peptide is cleaved during secretion. Residues 23-52 of SEQ ID NO:5 are a claudin-3- and claudin- 4-binding peptide from CPE (residues 290-319 of SEQ ID NO:1). Residues 22 and 53 result from insertion of the CPE residues. Residues 54-280 of SEQ ID NO:5 are human Fc (CH2 and CH3 domains and the hinge region of human IgGl heavy chain).

The plasmid encoding the 30aaCPE-Fc fusion protein is amplified in E. coli, and then transfected into the murine myeloma cell line Sp2/0 Ag 14 from the American Type Culture Collection for expression and secretion of the 30aaCPE-Fc fusion protein. The secreted fusion protein is purified by protein A or protein G affinity chromatography

References Cited:

1. Bohkman JV. 1983. Gynecol. Oncol. 15:10-17.

2. Rose PG. 1996. New Engl. J. Med. 335:640-649.

3. Sherman ME et al. 1992. Am. J. Surg. Pathol. 16:600-610.

4. Carcangiu ML et al. 1992. Gynecol. Oncol. 47:298-305.

5. Goff BA et al. 1994. Gynocol. Oncol. 54:264-268.

6. Carcangiu et al. 1995. Int. J. Gynecol. Pathol. 14:30-38.

7. Levenback C et al. 1992. Gynecol. Oncol. 46:317-321.

8. Nicklin JL. et al. 1996. Clin. Obstet. Gynecol. 39:686-695.

9. Trope C et al. 2001. Best Practice & Research in Clinical Obstetrics &

Gynaecology 15:433-446.

10. Hendrickson M et al. 1982. Am. J. Surg. Pathol. 6:93-108. l l. Ossowski L et al. 1991. J. Cell Biol. 115:1107-12.

12. Chan JK et al. 2003. Gynecologic Oncology 90:181-185.

13. Tsukita S et al. 2000. Ann. NY Acad. Sci. 915:129-135.

14. Katahira J et al. 1997. J. Cell Biol. 136:1239-47.

15. Katahira J et al. 1997. J. Biol. Chem. 272:26652-58.

16. McClane BA. 1996. Toxicon 34:1335-43.

17. Long H et al. 2001. Cancer Res. 61:7878-81.

18. Michl P et al. 2001. Gastroenterology 121 :678-684. 19. Hough CD 2000. Cancer Res. 60:6281-6287.

20. Rangel LBA et al. Clin. Cancer Res. 9:2567-75.

21. McClane BA et al. 1990. Infect. Immun. 58:3109-3115.

22. McClane BA. 1994. Toxicology 87:43-67.

23. Kokai-Kun JF et al. 1996. Infect. Immun. 64:1020-1025.

24. Kokai-Kun JF et al. 1997. Clin. Infect. Dis. 25 (Suppl. 2):S165-S167.

25. Kokai-Kun JF et al. 1997. Infect. Immun. 65:1014-1022.

26. Hanna PC et al. 1991. J. Biol. Chem. 266:11037-43.

27. Kokai-Kun JF et al. 1999. Infect. Immun. 67:5634-41.

28. Riethmuller G et al. 1998. J. Clin. Oncol. 16:1788-94.

29. Zhan F et al. 2002. Blood 99:1745-57.

30. Ismail RS, et al. 2000. Cancer Res. 60:6744-6749.

31. Santin AD, et al. 2000. Obstet. Gynecol. 96:422-430.

32. Fuchtner C.et al. 1993. Gynecol. Oncol. 48: 203-209.

33. Gamboa G et al. 1995. Gynecol. Oncol. 58:336-343.

34. Eisen MB et al. 1998. Proc. Natl. Acad. Sci. USA 95:14863-68.

35. Packeisen J et al. Hybridoma 18:37-40.

36. Welsh JB et al. 2001. Proc. Natl. Acad. Sci. USA 98:1176-1181.

37. Ono K et al. 2000. Cancer Res. 60: 5007-11.

38. Shridhar V et al. 2001. Cancer Res. 61 :5895-904.

39. Hough CD et al. 2001. Cancer Res. 6:3869-76.

40. Shridhar V et al. 2002. Cancer Res. 62:262-70.

41. Jazaeri AA et al. 2003. MoI. Carcinogenesis 36:53-9.

42. Maatta M et al. 2001. J. Histochem. & Cytochem. 49:711-26.

43. Casey RC et al. 2000. Clinical & Experimental Metastasis 18:67-75.

44. Yoshida Y et al. 2001. Inter. J. Oncol. 18:913-21.

45. KoshikawaN et al. 1999. Cancer Res. 59:5596-601.

46. Vollmers HP et al. 1984. FEBS Letters 172:17-20.

47. Iwamoto Y et al. 1987. Science 238:1132-4.

48. Linnenbach AJ et al. 1993. MoI. & Cell Biol. 13:1507-15

49. Litvinov SV et al. 1994. J. Cell Biol. 125:437-46

50. Morita K et al. 1999. Proc. Natl. Acad. Sci. USA 96:511-6.

51. McClane BA. 2001. Toxicon 2001, 39:1781-1791.

52. Ganesh S et al. 1995. Cancer Res. 54:4065-71.

53. Hasina R et al. 2003. Cancer Res. 2003; 63:555-9.

54. Dickinson JL et al. 1995. J. Biol. Chem. 270:27894-904.

55. Chambers SK et al. 1997. Int. J. Cancer 74: 71-575.

56. Chambers SK et al. 1997. Clin. Cane. Res. 3:999-1000.

57. Droz D et al. 1990. Am. J. Pathol. 137:895-905.

58. Raife TJ et al. 1994. Am. J. Clin. Pathol. 101:296-299.

59. Jackson D et al. 1992. Cancer Res. 52:5264-5270.

60. Huang LR et al. 1995. Cancer Res. 55:4717-4721.

61. Senner V et al. 1999. J. Neuropathol. Exp. Neurol. 58:795-802.

62. Fogel M et al. 1999. Cancer Lett. 143:87-94.

63. Aigner S et al. 1997. Blood 89:3385-3395.

64. Aigner S et al. 1998. EMBO J. 12:1241-1251.

65. Friederichs J et al. 2000. Cancer Res. 60:6714-6722.

66. Bratt T et al. 2000. Biochimica et Biophysica Acta 1482:318-26.

67. Flower DR. 1996. Biochemical Journal 318:1-14.

68. Oates AJ et al. 1997. Invasion & Metastasis 17:1-15. 69. Koeneman KS et al. 1999. Prostate 39:246-61.

70. Chambers AF et al. 1996. Lung Cancer 15:311-23.

71. Denhardt DT et al. 2003. Clinical & Experimental Metastasis 20: 77-84.

72. Kim JH et al. 2002. JAMA 287:1671-9.

73. Diamandis, EP et al. 2002. Clinical Chemistry 48:1198-205.

74. Tanimoto H et al. Cancer 86:2074-82.

75. Tanimoto H et al. 2001. Tumor Biology 22:104-14.

76. Tanimoto H et al. 2001. Tumor Biology 2001; 22:11-8.

77. Diamandis EP et al. 2003. J. Clin. Oncol. 21:1035-43.

78. Cannon MJ et al. 2002. Expert Review of Anticancer Therapy 2:97-105.

79. Liu Y et al. 2002. Reproduction 123:341-353.

80. Santin AD et al. 2002. Brit J Cancer 86(l):151-7.

81. Bongso A et al. 1988. Human Reproduction 3(6):705-13.

82. Meresman GF et al. 2003. Fertility & Sterility 80 Suppl 2:702-7.

83. Quelle DE et al. 1995. Cell 83(6):993-1000.

84. Moll UM et al. 1996. Human Pathology 27(12): 1295-300.

85. Kovalev S et al. 1998. Human Pathology 29(6):613-9, 1998.

86. Salvesen HB et al. 2000. Clinical Cancer Research 6(l):153-9..

87. Sano T et al. 2002. Path. Intern. 52(5-6):375-83.

88. Khleif SN et al. 1996. Proc. Natl. Acad. Sci. USA. 93(9):4350-4.

89. Terris B et al. 2002. Am. J. Pathol. 160:1745-54.

90. Seth P et al. 2002. Cancer Research 62:4540-4.

91. Luo LY et al. 2003. Cancer Research. 63(4):807-l 1.

92. Janes PW et al. 1997. J. Biol. Chem. 272:8490-7.

93. Santin AD et al. 2002. Clin. Cancer Res 8:1271-9.

94. Santin AD et al. 2003. Gynecol. Oncol. 88:263-5.

95. Santin AD et al. Gynecol. Oncol. 92(1):387.

96. Lukes AS et al. 1994. Cancer 73(9):2380-5.

97. Bristow, RE et al. 2004. Gynecol. Oncol. 92;480; Absl94..

98. Slamon DL et al. 2001. N. Engl. J. Med. 344:783 - 79.

99. Villella JA et al. 2003. Proc. Am. Soc. Clin. Oncol. Abs No: 1870.

100. Hortsch M. 1996. Neuron 17(4):587-93.

101. Patel K et al. 1991. Hybridoma 10(4):481-91.

102. Linnemann D et al. 1989. International Journal of Cancer 43(4):709-12.

103. Katayama M et al. 1997. Cell Structure & Function 22(5):511-6.

104. Fogel M et al. 2003. Lancet 362(9387):869-75.

105. Gutwein P et al. 2003. FASEB Journal 17(2):292-4.

106. Tian E et al. 2003. New Engl. J. Med. 349(26):2483-94.

107. Ossowski L et al. 1991. Journal of Cell Biology 115(4): 1107-12.

108. Xing RH et al. 1996. International Journal of Cancer 67(3):423-9.

109. Rabbani SA et al. 2001. Surgical Oncology Clinics of North America 10(2):393-

415.

110. Memarzadeh S et al. 2002. Proc Natl Acad Sci USA 99(16): 10647-52.

111. Riisbro R et al. 2002. Clin Cancer Research 8(5): 1132-41.

112. Rabbani SA et al. 2002. Cancer Research 62(8):2390-7.

113. Liu Y et al. 2002. Reproduction 123:341-353.

114. Holloway RW et al. 2002. Gynecol Oncol. 87:8-16.

115. Santin AD et al. 2000. Am. J. Obstet. Gynecol. 183: 601-609.

116. Santin AD et al. 1999. Journal of Virology 3: 5402-5410.

117. Niilo L et al. 1975. Infect. Immun. 12: 440-442. 118. Wallace FM et al. 1999. Current Microbiology 38:96-100.

119. Berek JS. 2000. Lancet 356:6-7.

120. Bookman MA et al. 2003. J. Clin. Oncol. 21 :283-90.

121. Clynes R et al. 1998. Proc. Natl. Acad. Sci. USA 95:652-656.

122. Casey, JL et al. 1996. Br. J. Cancer 74:1397-1405.

123. Lustgarten, J. et al. 1999. J. Immunol. 162:359-365.

AU patents, patent documents, and other references cited are hereby incorporated by reference.

Claims

CLAIMS What is claimed is:

1. A method of treating cancer in a mammal comprising: administering to the mammal a therapeutically effective amount of

Clostridium perfringens enterotoxin (CPE) or a pharmaceutically effective salt thereof; wherein the cancer is ovarian cancer or uterine cancer.

2 The method of claim 1 wherein the cancer includes chemotherapy-resistant malignant cells.

3. The method of claim 1 wherein the cancer includes metastatic malignant cells.

4. The method of claim 1, 2, or 3 wherein the cancer is a recurrent cancer.

5. The method of claim 1 wherein the cancer is uterine serous papillary carcinoma.

6. The method of claim 1 wherein the cancer is ovarian serous papillary carcinoma.

7. The method of claim 1 wherein the CPE or pharmaceutically acceptable salt thereof is administered intraperitoneally.

8. The method of claim 7 wherein the CPE or pharmaceutically acceptable salt thereof is administered at a dosage of 0.2 to 0.8 mg/kg body mass.

9. The method of claim 1 wherein the CPE or pharmaceutically acceptable salt thereof is administered intravenously.

10. The method of claim 1 wherein the CPE or pharmaceutically acceptable salt thereof is administered intratumorally.

11. The method of claim 1 wherein the CPE or pharmaceutically acceptable salt thereof has a therapeutic index as administered of at least 5.

12. The method of claim 1 wherein the CPE consists of SEQ ID NO:1.

13. The method of claim 1 wherein the CPE is a mutant or engineered CPE.

14. The method of claim 13 wherein the mutant or engineered CPE comprises residues 45-116 of SEQ ID NO:1.

15. The method of claim 1 further comprising administering a protective agent that protects cells against CPE toxicity.

16. The method of claim 15 wherein the protective agent is administered intravenously, and the CPE or pharmaceutically acceptable salt is administered by a non-intravenous route.

17. The method of claim 15 wherein the protective agent is administered enterally.

18. The method of claim 15 wherein the protective agent inhibits CPE binding to claudin-3 and/or claudin-4.

19. The method of claim 18 wherein the protective agent comprises residues 290- 319 of SEQ ID NO: 1 , or a homologue thereof.

20. The method of claim 18 wherein the protective agent comprises a fragment of residues 290-319 of SEQ ID NO:1 that binds specifically to claudin-3 and/or claudin- 4.

21. The method of claim 20 wherein the fragment of residues 290-319 of SEQ ID NO:1 comprises residues 315-319 of SEQ ID NO:1.

22. The method of claim 18 wherein the protective agent is an antibody against claudin-3 and/or claudin-4.

23. A method of treating cancer in a mammal comprising: intraperitoneally administering to the mammal a therapeutically effective amount of Clostridium perfringens enterotoxin (CPE) or a pharmaceutically acceptable salt thereof, wherein at least some cancerous cells are located in or adjacent to the peritoneal cavity of the mammal, and the cells are sensitive to CPE.

24. The method of claim 23 wherein the cancerous cells include chemotherapy- resistant cells.

25. The method of claim 23 wherein the cancerous cells include metastatic cancerous cells.

26. The method of claim 23 wherein the cancer is pancreatic cancer.

27. The method of claim 23 wherein the cancer is ovarian cancer or uterine cancer.

28. The method of claim 27 wherein the cancer is uterine serous papillary carcinoma.

29. The method of claim 27 wherein the cancer is ovarian serous papillary carcinoma.

30. The method of claim 23 wherein the cancer is liver cancer, stomach cancer, colon cancer, bladder cancer, kidney cancer, intestinal cancer, testicular cancer, or prostate cancer, wherein cells of the cancer have claudin-3 and/or claudin-4.

31. The method of claim 23 further comprising administering to the mammal intravenously or enterally a protective agent that protects cells against CPE toxicity.

32. The method of claim 31 wherein the protective agent inhibits CPE binding to claudin-3 and/or claudin-4.

33. The method of claim 32 wherein the protective agent comprises residues 290- 319 of SEQ ID NO:1, or a homologue thereof.

34. The method of claim 32 wherein the protective agent comprises a fragment of residues 290-319 of SEQ ID NO:1 that specifically binds claudin-3 and/or claudin-4.

35. A method of treating cancer in a mammal comprising: administering to the mammal a therapeutically effective amount of an anti-Ep- CAM antibody; wherein the cancer is ovarian cancer.

36. The method of claim 35 wherein the antibody is edrecolomab.

37. The method of claim 35 wherein the antibody is radioactive.

38. A method of determining the sensitivity of a gynecological malignancy to CPE comprising: detecting the presence or absence of claudin-3 and/or claudin-4 in a tissue sample comprising a portion of the malignancy.

39. The method of claim 38 wherein the detecting step comprises contacting the tissue sample with a protein comprising residues 290-319 of SEQ ID NO:1.

40. The method of claim 38 wherein the protein is CPE.

41. The method of claim 39 wherein the protein is an antibody comprising an Fc coupled to an antigen-binding peptide comprising residues 290-319 of SEQ ID NO: 1.

42. The method of claim 40 wherein the detecting step comprises detecting lysis of cells.

43. The method of claim 38 wherein the detecting step comprises contacting the tissue sample with an antibody against claudin-3 and/or an antibody against claudin-4.

44. A method of determining the sensitivity of a gynecological malignancy to treatment with an antibody against Ep-CAM comprising: detecting the presence or absence of Ep-CAM in a tissue sample comprising a portion of the malignancy.

45. The method of claim 44 wherein the detecting step comprises contacting the tissue sample with an antibody against Ep-CAM.

46. The method of claim 44 wherein the malignancy is ovarian cancer.

47. The method of claim 46 wherein the ovarian cancer is OSPC.

48. A use of Clostridium perfringens enterotoxin (CPE) to prepare a medicament effective to treat ovarian cancer or uterine cancer in a mammal.

49. The use of claim 48 wherein the mammal is a human.

50. The use of claim 48 wherein the medicament is adapted for intraperitoneal administration.

51. A use of Clostridium perfringens enterotoxin (CPE) to prepare a medicament effective to treat cancer in a mammal, wherein the medicament is adapted for intraperitoneal administration.

52. The use of claim 51 wherein the cancer is ovarian cancer, uterine cancer, or pancreatic cancer.

53. A use of an anti-Ep-CAM antibody to prepare a medicament effective to treat ovarian cancer in a mammal.

54. The use of claim 53 wherein the mammal is a human.

55. A method of protecting a mammal from Clostridium perfringens enterotoxin (CPE) toxicity comprising: administering to the mammal a protective agent that binds to claudin-3 and/or claudin-4 and inhibits CPE binding to claudin-3 and/or claudin-4.

56. The method of claim 55 wherein the protective agent is an antibody against claudin-3 and/or claudin-4.

57. The method of claim 56 wherein the antibody comprises an Fc coupled to a peptide comprising residues 290-319 of SEQ ID NO: 1 or a fragment thereof that specifically binds claudin-3 and/or claudin-4.

58. The method of claim 55 wherein the protective agent comprises residues 290- 319 of SEQ ID NO: 1 , or a homologue thereof.

59. The method of claim 55 wherein the protective agent comprises a fragment of residues 290-319 of SEQ ID NO:1 that specifically binds claudin-3 and/or claudin-4.

60. The method of claim 55 wherein the protective agent is administered after the mammal is exposed to CPE.

61. A method of treating cancer in a mammal comprising: administering to the mammal an antibody against claudin-3 and/or claudin-4; wherein cancerous cells in the mammal contain claudin-3 and/or claudin-4.

62. The method of claim 61 wherein the cancerous cells contain at least 5 -times more claudin-3 and/or claudin-4 than normal cells of the same tissue type.

63. The method of claim 61 wherein the antibody is humanized.

64. The method of claim 61 wherein the cancer is pancreatic, ovarian, uterine, prostate, or colon cancer.

65. The method of claim 61 wherein the antibody carries a therapeutic radioisotope.

66. A humanized antibody against an extracellular portion of claudin-3 and/or an extracellular portion of claudin-4.

67. The antibody of claim 66 wherein the antibody inhibits CPE binding to claudin-3 and/or claudin-4.

68. A method of treating cancer in a mammal comprising: administering to the mammal a therapeutically effective amount of a cytotoxic agent that binds specifically to claudin-3 and/or claudin-4; wherein the cancer is ovarian cancer or uterine cancer.

69. The method of claim 68 wherein the cancer includes chemotherapy-resistant malignant cells.

70. The method of claim 68 wherein the cancer includes metastatic malignant cells.

71. The method of claim 68 wherein the cancer is a recurrent cancer.

72. The method of claim 68 wherein the cytotoxic agent is Clostridium perfringens enterotoxin (CPE) or a pharmaceutically acceptable salt thereof.

73. The method of claim 68 wherein the cytotoxic agent is an antibody that binds specifically to claudin-3 and/or claudin-4.

74. The method of claim 73 wherein the antibody comprises residues 290-319 of SEQ ID NO:1 or a fragment thereof that binds specifically to claudin-3 and/or claudin-4.

75. The method of claim 68 wherein the cytotoxic agent comprises (i) residues 290-319 of SEQ ID NO: 1 or a fragment thereof that binds specifically to claudin-3 and/or claudin-4; coupled to (ii) a cytotoxic moiety.

76. The method of claim 75 wherein the cytotoxic moiety is an anti-cancer chemotherapy agent.

77. The method of claim 75 wherein the cytotoxic moiety is a moiety that stimulates cell killing by immune system components.

78. The method of claim 77 wherein the cytotoxic moiety is an antibody Fc.

79. The method of claim 77 wherein the cytotoxic moiety is interleukin-2.

80. An antibody comprising: an antigen-binding region comprising residues 290-319 of SEQ ID NO:1 or a fragment thereof that binds specifically to claudin-3 and/or claudin-4.

81. The antibody of claim 80 wherein the antibody is produced by recombinant means.

82. The antibody of claim 80 wherein the antibody comprises a therapeutic radionuclide.

83. An anticancer agent comprising: a moiety that binds specifically to claudin-3 and/or claudin-4; coupled to a cytotoxic moiety; wherein the agent is effective to kill tumor cells overexpressing claudin-3 and/or claudin-4 in vitro or in vivo, and wherein the agent is not Clostridium perfringens enterotoxin.

84. The anticancer agent of claim 83 wherein the agent is effective to treat cancer in a mammal in vivo, wherein the cancer comprises tumor cells overexpressing claudin-3 and/or claudin-4.

85. The anticancer agent of claim 83 wherein the cytotoxic moiety is an anticancer chemotherapeutic agent.

86. The anticancer agent of claim 83 wherein the cytotoxic moiety is a moiety that stimulates cell killing by immune system components.

87. The anticancer agent of claim 86 wherein the cytotoxic moiety is an antibody Fc.

88. The anticancer agent of claim 86 wherein the cytotoxic moiety is interleukin-

2.

89. The anticancer agent of claim 83 wherein the cytotoxic moiety is a therapeutic radionuclide.

90. The anticancer agent of claim 83 wherein the moiety that binds specifically to claudin-3 and/or claudin-4 comprises residues 290-319 of SEQ ID NO:1 or a fragment thereof that binds specifically to claudin-3 and/or claudin-4.