WO1994008615A1 - Antibody to naturally-occurring aminomalonic acid and uses therefor - Google Patents

Abstract

The present invention includes a monoclonal antibody produced against an epitope on a native aminomalonic acid-containing moiety. The antibody binds to BSA-aminomalonic acid, but not to BSA. Also included is an isolated aminomalonic acid-containing polypeptide that can be found in a human cancer cell. Diagnostic and therapeutic methods for solid cell tumors are also provided.

Description

ANTIBODY TO NATURALLY-OCCURRING AMINOMALONIC ACID

AND USES THEREFOR

Background of the Invention

Adhesion molecules have been implicated in tumorigenesis, metastasis, cell-to- cell interactions (leukocyte integrins for immune recognition and leukocyte migration), and atherogenic processes.

Fibronectin, a fiber-forming glycoprotein, first attracted attention when it was found in greatly reduced numbers on tumorigenic fibroblasts as compared to normal fibroblasts. In general, there seems to be a correlation between a decrease in cell- surface fibronectin quantities and the ability of the cells in vitro to cause tumors, invade tissue, and metastasize.

Transformed cells behave differently than normal cells in vitro. They adhere poorly to substrates and fail to flatten out and develop organized intracellular actin filament bundles known as stress fibers. Transformed cells also grow to much greater densities than normal cells. If additional fibronectin is added to transformed cells that are producing only small fibronectin quantities in vitro, the cells will rapidly adhere, flatten out, and generate organized intracellular actin filament bundles.

Purified fibronectin has been shown to promote adhesion in a variety of cell types.

In tumor metastasis, adhesion molecules are necessary for binding to and migration through the endothelium in addition to cell-cell adhesion. Research has shown that tumor cells have increased amounts of integrins on their cell surface when compared to normal cells.

Leukocyte Adhesion Deficiency (LAD), a disease associated with a lack of adhesion molecules on the cell surface of leukocytes, rarely involves tumorigenic growth even though these patients develop a wide spectrum of other chrome diseases.

Moreover, a highly metastatic T-cell line, TAM202, has been mutated yielding three adhesion molecule deficient clones that metastasize very poorly compared to the original strain. Both LAD and this cell line provide additional evidence for the role of adhesion molecules in tumorigenic growth and metastasis. In an effort to identify critical epitopes of adhesion molecules, many investigators have determined the amino acid sequences from a number of these proteins. However, no specific structural features of these molecules have been identified to date.

Another way of identifying structural features of a protein is to look for modifications of the amino acid residues. These modifications can take the form of di-sulfide bonds, sugar moieties, or carboxylation. One specific modified base is

Amino malonic acid (A a), a percarboxylated glycine.

Ama was not proven to be a naturally occurring constituent of proteins until 1983. Since the percarboxyl group of Ama is extremely acid labile, traditional acid hydrolysis gradation of proteins destroyed this molecule and may have yielded false results for proteins containing Ama. In fact, amino acid analysis of Escherichia coli alkaline hydrolysates revealed high levels of Ama and Asa (B-carboxy aspartic acid). In one experiment, fractions containing Ama were pooled, then divided, with one portion thereafter treated with acid. The acid-treated sample contained Gly and no Ama whereas the untreated portion contained only Ama. The question of how the cell makes Ama-containing proteins has proven perplexing. Ama was shown to not be an artifact of acid hydrolysis by Gas chromatography/mass spectrometry. This experiment unambiguously confirmed the presence of Ama in E. coli and atherosclerotic plaque hydrolysates.

One theoretical mechanism of Ama production involves first chemically modifying a glycine residue, and then incorporating (during translation) the modified amino acid into proteins. However, this pathway is unlikely since it would not only require incorporation of tRNAs containing Ama without a proofreading mechanism, but also sufficient concentrations of free Ama in the cell.

The most plausible method of Ama incorporation is post-translational modification of the protein's amino acid residues. This could theoretically occur by one of three different amino acid modifications. Glycine could be carboxylated to form Ama, serine could be oxidized, or cysteine could be oxidized. In fact, Koch and colleagues at the University of Colorado originally concluded that Ama synthesis was a non-enzymatic, free radical-mediated, post-translational modification of peptide linked glycine. However, later work pointed away from glycine carboxylation.

Currently, it is believed that serine is the most likely Ama precursor. One piece of evidence that supports this is that normal Escherichia coli contains about .38 Ama molecules per 1000 amino acids while another strain (JC-158), that has a deficiency in the serine synthetic pathway, contains .04 Ama molecules per 1000 amino acids. It is unknown whether the particular post-translational modification yielding Ama is physiological or accidental. If it were an intentional, physiological effect in the cell, one would expect Ama containing proteins to be localized to an area where they could carry out their function. If it were accidental, one would expect these proteins to be found randomly in cells, associated with a triggering event.

Although adhesion molecules have been recognized for a long time, no one has yet developed a convenient way for identifying and inhibiting many of these known glycoproteins. It would be advantageous to have a product that specifically bound to and inhibited the function of such adhesion molecules. Such a product could be used for inhibiting tumorigenic growth or adhesion to cell surfaces, for example. The calcium binding properties of Ama have been previously remarked, and the association of Ama with calcium-laden atherosclerotic plaque has been reported in Proc. Natl Acad. Sci. USA 81: 722-725 (1984). Ama-specific monoclonal and polyclonal antibodies have been generated in mice using a chemically synthesized Ama peptide as antigen. These antibodies have been used experimentally to aid in the detection of atherosclerosis. However, these antibodies to synthetic Ama have lacked a high degree of sensitivity and specificity. Thus, the utility of these antibodies as an effective immuno-diagnostic has been limited. It would be helpful to obtain antibodies with a high specificity and without problems of cross-reactivity to improve the effectiveness of immunodiagnostics. Prostatic cancer strikes a significant number of men over the age of 40.

Prognosis generally depends on how early the cancer is detected. However, presently, the only biochemical test for detection requires repeated visits over several months or years. This test has yet to be proven. Moreover, the cancer may have developed over the time period between visits to a point where the prognosis is not as good as if positively diagnosed at an earlier stage. Thus, there remains a need for a biochemical test for prostatic cancer that does not require a significant delay. Summary of the Invention

One embodiment of the present invention is a monoclonal antibody produced against an epitope on a native aminomalonic acid-containing moiety, wherein the antibody binds to BSA-aminomalonic acid, but not to BSA. Preferably, the antibody has substantially the same binding activity as the antibody produced by hybridoma strain ATCC No. HB 11146. Most preferably, the antibody is produced by hybridoma strain ATCC No. HB 11146. In a further preferred embodiment the aminomalonic acid-containing moiety can be found in a human cancer cell. Most preferably the cancer cell is from a prostate carcinoma cell line, with the most advantageous cell line being DU-145. In yet a further preferred embodiment, the monoclonal antibody described above binds to a cell surface antigen present in malignant cells that is vital for divalent cation binding.

Another embodiment of the present invention is an isolated aminomalonic acid-containing polypeptide that can be found in a human cancer cell. Preferably, the polypeptide plays a role in cell adhesion. Even more preferably, the polypeptide is a cell surface antigen present in malignant cells that is vital for divalent cation binding. Advantageously, the polypeptide can be found in a human prostate carcinoma cell line. Still more advantageously, the cell line is DU-145 and has an apparent molecular weight in SDS-PAGE of 14 to 80 kD. Still another embodiment of the present invention is a method of diagnosing a solid cell tumor in a tissue suspected of having such a tumor in a patient. This method comprises (1) administering to the tissue an antibody specific for aminomalonic acid; (2) determining the level of binding of the antibody to the tissue; and (3) comparing the determined level of binding of the antibody to a baseline level of binding of the antibody to normal tissue of the same type as the tissue. A higher level of antibody binding to the tissue than the baseline level indicates the presence of a solid cell tumor in the tissue.

Preferably, the baseline level is a substantially undetectable quantity when determined by the method used in determining the level of binding of the antibody to the tissue. Advantageously, the tissue in the above method comprises prostate tissue, and the tumor comprises a prostate carcinoma. Even more preferably, the

SUBSTITUTE SHEET antibody is labeled and is produced against an epitope on a native aminomalonic acid-containing moiety. Still yet a more preferable embodiment is wherein the antibody has substantially the same binding activity as the antibody produced by hybridoma strain ATCC No. HB 11146. The most advantageous antibody in this method however, is produced by hybridoma strain ATCC No. HB 11146.

In even yet another preferred embodiment of this method, the aminomalonic acid-containing moiety can be found in a human cancer cell. The moiety can alternatively be from a prostate carcinoma cell line, with the most advantageous cell line being DU-145. Additionally, the antibody in this method can be labeled with a radioisotope and the determining step can include a radioimaging procedure.

Further, the antibody of this method can be labeled with a paramagnetic moiety with the determining step including nuclear magnetic resonance imaging. In still yet a further preferred embodiment, the administering step of this method can include the in vivo administration to the patient of an immunoreactive amount of the antibody. Also, the antibody of the present invention method can advantageously bind to a polypeptide that plays a role in cell adhesion, while this polypeptide is a cell surface antigen present in malignant cells and vital for divalent cation binding.

Yet another embodiment of the present invention is a method of treating a solid cell tumor in a tissue in a patient, comprising administering to the patient an effective solid tumor-inhibiting amount of a therapeutic antibody specific for a cell surface antigen present in the tumor that is vital for divalent cation binding. Preferably, the cell surface antigen is recognized by the antibody produced by hybridoma strain ATCC No. HB 11146. Most preferably, the therapeutic antibody has substantially the same binding activity as the antibody produced by hybridoma strain ATCC No. HB 11146. In this method the solid cell tumor can advantageously be a prostate cell carcinoma. Most advantageously, the method includes administering the therapeutic antibody in an amount effective to reduce the likelihood of tumor metastasis. Still more preferably, the method includes the co- administration of other anti-cancer agents advantageously administered during or immediately after surgical excision of the tumor.

SUBSTITUTE SHEET Another embodiment of the present invention is a pharmaceutical composition comprising an antibody produced against an epitope on a native aminomalonic acid- containing moiety, wherein the antibody binds to BSA-aminomalonic acid, but not to BSA in an amount effective to bind to a solid cell tumor in a patient, and a pharmaceutically acceptable carrier. Preferably, the antibody is present in an amount effective to treat the tumor in the patient and advantageously prevent metastasis of the tumor in the patient. Even more preferably, the antibody of the above pharmaceutical composition is labeled and is present in an amount that specific binding thereof to tissue in the patient can be detected. Brief Description of the Figures

Figure 1 shows the results of SDS polyacrylamide gel electrophoresis (SDS- PAGE) of antigens in a prostatic carcinoma cell line (DU-145) from a immunoaffinity column using monoclonal antibodies directed against a native Ama- containing epitope (mAb-Ama_nat). Figure 2 shows the results of native PAGE of antigens in DU-145 cells from a immunoaffinity column using mAb-Ama_na,.

Figure 3 is a graph showing the effect of increasing concentration of a mAb- Ama_nat on DU-145 cells in culture.

Detailed Description We have discovered a monoclonal antibody produced against naturally- occurring aminomalonic acid (Ama). Such an antibody can delineate the presence and function of the Ama epitope in adhesion molecules. Thus, we have also discovered that the Ama epitope is an essential component of adhesion molecules and appears vital to malignant cells for growth and metastasis. Prostatic cancer cells appear to be particularly disposed to display surface-bound proteins with Ama epitopes. By using the monoclonal antibody of the present invention, cancer cells, and particularly prostatic cancer cells can be distinguished over normal cells. Moreover, we have demonstrated that these monoclonal antibodies can inhibit tumor growth. Thus, we believe these antibodies are useful in the treatment of solid cell tumors. Producing a Naturally-Occurring Ama Epitope

Because Ama is very acid labile, isolation of pure samples of this compound from naturally occurring sources is difficult. Even tissues with high concentrations of Ama, such as atherosclerotic plaques, include only about 5 nanomoles/mg of Ama. Accordingly, in order to produce the antibodies of the present invention directed against naturally-occurring Ama epitopes, we first obtained antibodies directed against a synthetic Ama antigen (Ama^). The antibody against Ama^ was used to identify and purify components of naturally-occurring samples immunoreactive therewith. Many methods of chemically synthesizing Ama have been described in the literature (Ann. Chem. 131: 291 (1964)), with a particularly preferable method being described in JACS 75:2499 (1953). However, any of a variety of methods of synthesizing a pure concentration of the Ama molecule can be performed in accordance with the present invention. In the methods employed in the discovery of the present invention, synthetic

Ama (Ama_syjj) was chemically linked to bovine serum albumin (BSA) to provide an immunostimulatory agent for injection and isolation of Ama specific antibodies. Example 1 below details one preferred method of conjugating Ama to BSA.

Amino malonic acid, synthesized as described above, was conjugated to bovine serum albumin (BSA) providing an injectable substrate with immunoreactivity. BSA is a known immunostimulatory agent, and was chosen for this conjugation so as to elicit the most aggressive anti-Ama immune response after injection.

Example 1

Conjugation of Ama to BSA Example 1A

Ama was chemically linked to BSA, as previously described in Methods in Enzymology 70:159 (1980). This method is known to those of skill in the art However, in brief, a solution of 20 mg BSA and 3 mg pure synthetic Ama was dissolved in 2ml of phosphate buffered saline (pH 7.0). This corresponded to 0.3 μM BSA and 25 μM Ama resulting in a Ama:BSA ratio of approximately 83:1. Gradual addition of a 21 mM glutaraldehyde solution (in 0.1 mM PBS) was carried out with constant stirring, followed by incubation for 24 hours at room temperature. Example IB

N-acetyl-Ama-ethylamide was attached to amino groups of bovine serum albumin (BSA) with a 6-aminohexanoic acid spacer. Conjugation was achieved with diisothiocyanate binding the Ama peptide to 46 of the 59 lysine residues on the BSA molecule.

Both the methods of Example 1A and Example IB provided an effective Ama^ antigen. The resulting solution, comprising conjugated BSA-Ama_jy,,, can be dialyzed against 0.1 M PBS (pH 7.0) in preparation for murine injection.

After obtaining an Ama^ antigen, such as one of the BSA-Ama_jy,, antigens of Example 1A or IB, antibody against the antigen should be obtained. Thus, for example, after the conjugation of Example IB, the BSA-Ama^ we prepared was injected into BALB/c mice to stimulate production of antibodies against the synthetic Ama molecule. The first step in the production of this antibody is to generate an immune response in a suitable animal, such as a mouse. Murine injection can be intraperitoneally (IP), subcutaneously (SC) or a combination of both methods. Adjuvant is normally included with the injected material to increase the animal's immune response. Freund's adjuvant is one type of adjuvant normally used by those of skill in the art. Example 2 below provides one method for producing monoclonal antibodies to Ama.-. This method includes injecting BSA-Ama into mice to elicit an immune response and create hybridomas.

Example 2

Production of Hybridomas Producing Monoclonal Antibodies to Ama-BSA

Male BALB/C mice were immunized with a series of IP injections containing 50 μg BSA-Ama-_yn in Freund's adjuvant over a period of seven weeks. Three days after the final injection, the mice were sacrificed and spleens removed. The spleen cells were fused with mouse myeloma cells (SP 2/0), and the hybridized cells were suspended in HAT media (hypoxanthine, aminopterin, and thymidine in DMEM) with 20% serum and 10% NCTC-109 supplemented media at a concentration of 3 x 10⁵ cells/ml with normal spleen feeder cells. The cell suspension was dispensed into a 24-well plate at approximately 1 ml/well. After seven days, 1 ml of DMEM-

HT media (minus aminopterin) was added to each well and the plate was inspected for the presence of hybridomas. An ELISA was run on the hybridomas to determine which clone was producing the antibody with greatest binding to BSA-Ama_™.

Example 3 ELISA of the Eluted Protein Fraction An ELISA was run to confirm the binding, and estimate the strength, of the hybridoma clones created by the fusion of Example 2.

The following steps were undertaken to perform the ELISA:

1. Pipette 50 mL column eluate into a 96 well plate.

2. Add 50 mL mM Carbonate buffer (pH 9.6). 3. Incubate overnight at 4^βG

4. Wash plate with PBS 3 times and block with 315 mL plating buffer (0.05% tween, 0.25% BSA, 0.02% azide in PBS). Incubate for 30 min.

5. Wash plate with PBS 3 times and add 100 mL of anti-Ama^, (primary antibody), incubate for 60 min. 6. Add 100 mL second Ab (anti-IgG). Incubate for 60 minutes at 37 "C.

7. Wash plate with PBS 3 times then add 100 mL alkaline phosphatase (substrate) and place in the dark at room temperature.

8. Read at 30 and 60 minutes at 410 nM on an ELISA plate reader.

The monoclonal antibody showing the strongest binding to BSA-Ama was determined, such as through the ELISA assay of Example 3. To isolate and purify the antibody produced by this hybridoma, the media from the hybridoma was run over a protein A column, as explained below in Example 4.

Example 4

Protein A Purification of Monoclonal Antibodies A protein A column was run to isolate the monoclonal antibodies being produced by the preferred fusion hybridoma. Protein A binds antibody fragments, thereby providing a method for preferentially binding monoclonal antibodies from other media components.

In this case, the hybridoma media from the clone containing the secreted antibody, was brought to 1 M Tris/HCl (pH 8.0) and run over the column (1/20 column volume) by methods known to those of skill in the art. The protein A column was equilibrated with 2-3 column volumes of O.lM Tris/HCl (pH 8.0), and washed with 0.01M Tris/HCl (pH 8.0). Bound fractions comprising the desired monoclonal antibody were eluted with O.lM citrate/NaOH (pH 3 to pH 6).

These specific mAbs produced by the method of Example 4 were also screened for activity by ELISA assays to confirm the binding with BSA-Ama-,,, but not BSA.

To provide the desired antibodies directed against the naturally occurring

Ama epitope, we next used the BSA-Ama^ derived monoclonal antibody, described above, as a probe to selectively isolate proteins with naturally occurring Ama epitopes. These naturally occurring proteins were then injected into suitable animals to create monoclonal antibodies and hybridomas against Ama in its naturally occurring state.

Example 5

Purification of Natural Proteins

Bearing the Synthetic Ama-BSA Epitope Malignant (DU-145) and normal prostatic cells were detergent lysed and the lysates run on an affinity column containing the matrix-bound murine monoclonal Ab produced in Example 4 that had cross-reactivity to the BSA-Ama_j-,,. epitope. Antibodies with these characteristics are referred to herein as "mAb-Ama ''

Fresh human prostate carcinoma cells (DU-145 tumor cell line, American Type and Culture College, Rockville, MD) were detergent lysed and the lysate centrifuged. The supernate was applied to the mAb-Ama_™ immunoaffinity column and the unbound material washed through.

The DU-145 cells were grown until late log phase in media described in Table 1.

Cell type: Adherent

Doubling time: 24 hours

Source: ATCC

Media: DMEM with 10% FCS Table 1: DU-145 Growth Media Composition

PROCEDURE: 1. Media:

Reagents DMEM, penicillin/streptomycin, glutamine and fetal calf serum were thawed. Ingredients were filtered in a sterile environment until the glutamine was fully dissolved. DMEM (about 300 mL) was mixed with the other ingredients (5mL Pen/Strep, 5mL Glu, and 50 mL Fetal Calf Serum) and brought up to 500 mis. The mixture was stored at 4 ° C until it was used.

Pre-made media was warmed to 37 ° C and the cells were counted and viability tested to determine the split ratio. After the media became warm, the bottle and culture flask were alcohol sterilized and placed under a sterile hood. 2. Cell Split: The cell media was aspirated and discarded. Five milliliters of Hank's buttered salt solution (HBSS) per T-150 flask without Ca+ + or Mg+ + was pipeted into each flask. The flask bottoms were rinsed, aspirated and the HBSS discarded. An additional 5 mLs of cell dissociation buffer (0.04 gms EDTA in 500 mL HBSS), was added to the flask and placed at 37 " C for 10-15 minutes. The flasks were removed from the incubator and tapped against a flat surface to dislodge the remaining adherent cells. The flask was resprayed with alcohol and again placed under the hood. A further rinse with 2 mLs media was followed by aspirating a predetermined amount of stock culture to inoculate either roller bottles or T-150 flasks. Fresh media was added to a total of 17-20mL for T-150 flasks and 150-180 for roller bottles.

3. Lvsis of DU-145 cells:

DU-145 cells were grown until late log phase. Media was decanted and Triton X-100 lysis buffer (Triton X-1000.5%, Phenylmethylsulfonyl fluoride (PMSF)

0.1%, Azide 0.2g/l in PBS) was added directly to the T-150 flask or roller bottle at 2 x 10⁷ cells/mL (about lOmL for T-150 flasks and 30 mL for roller bottles). The flasks were kept on ice for 30 to 45 min., with light mixing. The roller bottles were rotated at 100-120 rpm (4 ^βC) for approximately 15 min. then occasionally for next 30 min. The PMSF, a protease inhibitor, was added fresh, since it normally will only stay active a relatively short time.

The DU-145 cell lysate was centrifuged for 15 minutes at approximately 4000 x G (4500 rpm) in the cold (4 ^β C) to pellet the lysed membrane. The supernatant (containing cellular proteins to be column purified) was transferred to a 50 mL freezer vial and placed in the cold.

Further details of this procedure can be found in Current Protocols in Immunology, Volume 1, (1991). Colign, J.E., Kruisbeek, A.M., Margulies, D.H., Shevach, E.M., and Strober, W., eds., National Institutes of Health, Chapter 8.

Affinity chromatography of the DU-145 cell supernatant was performed to isolate the proteins displaying cross-reactivity with the Ama^ epitope. These procedures are described in Example 6 below.

Example 6

Affinity Chromatography of the DU-145 Lysates

A series of column steps was undertaken as shown in Table 2 below:

Table 2: Affinity Column Steps

Wash and elute until the peak on the recorder goes back to baseline - 10-20 column volumes.

1. cv = column volume

2. Reverse column for washing and elution.

Description of Column Solutions:

Diethylamine (500 mis, pH 11.5)

The column was precycled with 2 column volumes of diethylamine, then washed with PBS at the rates indicated in Table 2. Following PBS washes, the column was equilibrated with 2-3 column volumes of lysis buffer. DU-145 cell lysates were applied at 1 column volume/hour, then rinsed with 10-20 column volumes of PBS. Bound proteins were eluted with 0.1 M diethylamine, pH 11.0. Each eluted fraction was neutralized immediately with 1 M PBS, pH 6.8. The column was then re-equilibrated to a neutral pH with PBS. The eluted fraction comprising proteins with cross-reactivity to the synthetic

Ama epitope was buffer exchanged and concentrated by ultrafiltration as described in Example 7 below.

We also tested a variety of other human carcinoma cell lines using the methods of Example 6. In addition to the DU-145 prostatic carcinoma cells, we also tested A 427 (lung), BT 20 (breast) and CA C02 (colon) cells. We found some antigen-binding to all four cell lines. We also screened CEM leukemia cells and found a low level of binding. Both lipid-laden macrophages and malignant T lymphocytes contained significant levels of Ama-containing epitopes. However, the highest level of antigen-binding was found for the DU-145 cell line. Thus, we further tested antigens obtained from DU-145 cells.

Example 7

Concentration of DU-145 Eluate From Immunoaffinity Chromatography

A YM-10 membrane was soaked (shiny side down) in a beaker of pyrogen free water to elute off any azide and repeated once. The concentrator was assembled (membrane shiny side up) followed by addition of the DU-145 eluate.

The concentrator was placed on a stir plate (4 ° C) and stirred at a moderate speed

(not enough to disrupt or foam proteins but enough to keep them from clogging the membrane). Positive air pressure (100 kpascals) was applied. The membrane was refilled with eluate as needed, and the filtrate discarded as necessary.

The bound fraction was prepared for murine injection after the concentration step by being buffer exchanged (three times) with 10 mL of sterile PBS.

The binding and purity of the eluted material was then studied by an ELISA performed as described above in Example 3. The ELISA results confirmed the positive, strong binding of mAb-Ama_sy,, to specific proteins in the DU-145 column eluate. Monoclonal Antibody to Native Ama-Containing Antigen

To produce monoclonal antibodies against the native Ama epitope, the protein eluate from the previous immunoaffinity column (mAb-Ama_j™) was used as an immunostimulatory agent in mice. Hybridomas were then produced that secreted monoclonal antibodies against native antigens.

Briefly, eight mice were injected with approximately 200μL of the DU-145 immunoaffinity column purified, concentrated lysate in Ribbi's adjuvant (100 μL IP and 100 μL ID). On days 21, 42 and 63 the mice were boosted with a similar dose. On day 73 a small amount of blood from the tail veins of each mouse was titered for the presence of anti-Ama Antibodies. The two mice with the highest titers were sacrificed on day 84 and the spleen cells isolated and fused with NS-1 cells.

Specific procedures used in the production of monoclonal antibody to the naturally-occurring Ama-containing antigens are described in Examples 8-10 below.

Example 8 Injection of Antigen into Mice

Eight BALB/C (female) mice were immunized both intraperitoneally (IP) and subcutaneously (SC) with a series of injections (initial plus booster) comprising

20 μg/mouse of antigen in Freund's adjuvant. Antibodies from the injected mice were isolated by the following procedure and tested for cross-reactivity to the antigen.

Mice immunization protocol

Adjuvant - RIBIs

0.5 mg Monophosphoryl Lipid A Part of LPS--G-, Mitogenic for mouse lymphocytes

0.5 mg Synthetic Trehaose Dicorynomycolate

Cord factor from MTB, adjuvant, alt C path, increase lymphs, decrease PMNs 0.04 mg Squalene (40 mL) 0.004 mg Monocleate (Tween 80) (0.2%)

Antigen from immunoaffinity after concentration Immunization times:

Day 1 Immunization with adjuvant IP and Sub-Q

Day 22 First boost with adjuvant IP and Sub-Q Day 43 Second boost with adjuvant IP and Sub-Q

Day 64* Third boost with adjuvant IP and Sub-Q

Day 74 Bleed Tails and Titer

Day 85 Final boos without adjuvant IV

Day 89 Fusion (as described hereinbelow)

* Third boost is not necessary in some instances.

Concentrated antigen should have a total protein content between the ranges of 50-250 mg/mL or 250-100 mg/mL if weakly immunogenic or non-purified. 1. Prior to reconstituting the emulsion, the vial was placed in a water bath at 40-45 ° for 5-10 minutes.

2. The antigen/saline solution (2 mL) was injected directly into the vial through the rubber stopper using a syringe fitted with a 21 gauge needle (cap seal was left in place). 3. The vial was vortexed vigorously for 2-3 minutes to form an emulsion.

(This can then be stored up to 60 days (at 4 ^βC) or lyophilized.) Each vial contained 10 mice doses (200 mL each).

4. The mice were marked with picric acid (yellow ~ binds to mouse hairs). 5. The bottle was re-vortexed. Using a tuberculin syringe, 100 mL was injected perisplenic, IP, at the scruff of neck, subcutaneously, or alternatively the final boost comprising an antigen saline solution (without adjuvant) was injected directly into the tail vein. Warming the tail vein in hot water dilated the vessel, making it easier to find. Splenocytes from mice treated as in Example 8 were isolated as a source of antibody-producing cells in the generation of hybridomas for production of monoclonal antibody to the antigen. The splenocyte cells were then fused with a transformed cell line, in this case NS-1 cells. Example 9 below describes the first step of splenocyte isolation. Example 9

Isolation of Murine Splenocytes

Materials: Immunized and Ab titered mice

Sterile: forceps scissors towelettes Small pipette tips petri dish syringes needles 96 well plates alcohol Polyethylene Glycol (PEG)

Media with 20% FCS Fusion partner cells

The mice were boosted with a tail vein injection comprising 20 μg antigen in saline solution (without adjuvant) 3-5 days prior to harvesting/fusion.

Harvesting Splenocytes

Mice were anesthetized using ether, then sacrificed by cervical dislocation. Using sterile instruments, the skin was carefully pulled up using forceps and an opening snipped with the scissors. A diagonal cut over the spleen (left side) was opened up and spread. Using forceps, the spleen was removed, cutting away any connecting vessels or tissue. The spleen was immediately placed in a sterile petri dish containing warm media without serum.

Using syringes and needles, the splenocytes were teased out by injecting media repeatedly into the spleen. Rinsed media containing free splenocytes was transferred to a test tube and additional aliquots of fresh media added to the spleen. Repeated rinses and teasing isolated substantially all the splenocytes into the test tube.

Once the splenocytes were isolated and separated, they were fused with transformed cell line NS-1 to produce the desired hybridomas. This procedure is described in Example 10 below. Example 10

Fusion of Murine Splenocytes with NS-1 Cells

Week Before Fusion

1. The NS-1 fusion partner cell line was grown for one week in media without 8-azaguanine. NS-1 cells do not secrete antibody, making it a preferable choice for cloning those cells producing antibody.

2. PEG was made up five days before the fusion in 2X DMEM (from powder) and having a IX final concentration after a 1:1 dilution. The PEG/DMEM solution was placed in a 42 ^• C water bath until there was no evident paniculate matter. Incubator storage for several days prior to use made the solution slightly alkaline (preferable). Day of Fusion - Cell Washing

The splenocytes isolated in Example 9 were washed by centrifugation at 400 x G (1400 rpm) in medium without serum. After a second splenocyte wash the myeloma cells were rinsed in media (without serum) in a separate tube. The supernatants from both cell samples were discarded and the two pellets resuspended in 37 ^βC medium (without serum). The cell suspensions were combined and centrifuged at 800 x G (2400 rpm) to pellet the cell mixture. Day of Fusion • Fusion The PEG mixture produced above was removed from its container with a

Pasteur pipet and slowly added to the pellet of NS-1 and splenocyte cells. The cells were resuspended by gentle stirring with pipet tip. The PEG was added slowly over 1 minute, while stirring for an additional minute. A 10-mL pipet was filled with 10 mL of medium (without serum) prewarmed to 37 ° C. An additional 1.0 mL of medium was added to the cell suspension over the next minute while continuing to gently stir with 10-mL pipet tip. The remaining 9.0 mLs of warm media was added over the next 2 min. with constant stirring. The mixed cells were then centrifuged at 400 x G (1400 rpm) for 5 min.

The supernatant was removed and the cells resuspended in 10 mLs of medium supplemented with 20% pre-screened fetal bovine serum (prewarmed to

37 ° C) containing 1% OPI (Oxaloacetate, pyruvate, and insulin), 2% HAT (hypoxanthine, aminopterin, and thymidine), and 5% HCF (hybridoma cloning factor).

Media Composition: Am nt I r di nt

* For low density cultures (fusion and cloning).

The aminopterin blocks de novo purine and pyrimidine synthesis in the unfused myeloma cells which have been selected for a mutant (non-functional) hypoxanthine-guanine phosphoribosyl transferase (HRPT) gene whose product performs one the essential steps in the salvage pathway. The fused cells have a normal gene from the spleen cells.

Procedure:

All media ingredients listed above were warmed to 37 ^βC. When the ingredients were warm, the bottles and flask containing cell cultures were thoroughly sprayed with alcohol and placed under sterile hood. Approximately 20mL DMEM and other media composition ingredients (1 mL Pen/Strep, 1 mL Glu, 20 mL Fetal Calf Serum, 5 mL HCF, 1 mL OPI, 2 mL HAT) were mixed together and then brought up to 100 mL total with DMEM.

The media composition was placed in a 37 ° C water bath to keep warm. The cell counts of the hybridomas were done using traditional methods. Using serial dilutions of no more than 1:10, the hybridoma concentration was diluted to 5 cells/mL. The final volume was brought up to approximately 40 mL (enough for 3-5 96 well plates). The diluted hybridomas were then dispersed into 96 well plates and place in a C0₂ incubator.

After two days, the clones were checked for growth. Many wells had no cells, some had one clone and some had two. Those with one clone were screened, and the fastest growing were chosen for specific antibody production. The selected clones were transferred to 24 well plates and isotyped using an ELISA procedure using a technique such as is described in Example 3 above. Alternatively, any of a number of such procedures readily known to those of skill in the art can be used. The selected clones were then transferred to 6 well plates, T-25 flasks, and finally T-75 flasks. Some of the cells were frozen for later use, while others were grown up in T-75 flasks, then spinner flasks. The large cultures were used to purify monoclonal antibodies.

The clones were visible by light microscopy at about day 4. In 96 well plates, the clones tended to grow on the edges of the wells. Thus, it is advantageous for the screening to be done, and cells transferred, before overgrowing the plate. Freezing cell lines and hybridomas

A portion of the cells were centrifuged gently at 400 x G (1400 rpm) and the supernatant discarded. Cells were resuspended gently in sterile PBS then centrifuged at 400 x G (1400 rpm). The supernatant was again discarded.

Cells were resuspended in cold 94% fetal calf serum and 6% dimethylsulfoxide. Approximately 5 X 10⁶ cells/mL were stored in 1 mL freezer vial aliquots. The cells to be frozen were placed in the -24 ° C freezer overnight. This cooled the cells down at the even rate of 1 ^β C/minute. The next day, the freezer vials were stored in the -70 ° C (liquid N0₂) freezer.

Five hybridomas were found that produced antibodies with significant binding to BSA-ama; three of these produced IgM and two produced IgG. Antibodies to the native ama-containing epitope are referred to herein as mAb-Ama_nat. A hybridoma producing IgG_j antibody, that we call WMK-1, was selected as having the highest binding activity. The hybridoma producing this antibody has now been deposited with the ATCC as ATCC accession No. HB 11146. The WMK-1 antibody was attached to an immunoaffinity column, and the column used to select proteins expressing Ama in its native form from DU-145 cell lysates, such as those created above in Example 5.

An affinity column was made with mAb-Ama_nat to purify naturally occurring proteins containing Ama epitopes. Lysates of malignant and normal cells were pumped through this column with the bound proteins being eluted and the fractions collected and tested using ELISA blots for confirming cross-reactivity to the mAb-

SUBSTITUTE SHEET Ama_jjaf Fractions were also run on a polyacrylamide gel to estimate the size and number of proteins in each fraction. Proteins of interest were then purified and made available for further analysis, such as amino acid sequencing.

The eluate from the mAb-Ama_nat column was also studied by SDS-PAGE, in which seven protein bands were visualized by Coomassie blue staining ranging from 14 to 70 Kd with two darkly staining bands at 36 and 50 Kd as shown in Figure 1. On native-PAGE we visualized two bands at 66 and 80 Kd shown in Figure 2. Using size exclusion column chromatography we isolated seven peaks ranging from 1,000 to 12.4 Kd. A Western blot assay using the WMK-1 antibody is to be performed to more specifically identify the sizes of naturally-occurring ama- containing polypeptides in DU-145 and other cells.

The specificity of the mAb-Ama_nat has been partially determined. Neither WMK-1 nor mAB-Ama_syj, bind to normal skin. However, WMK-1 appears to bind to colon carcinoma more strongly than mAB-Ama^. We also isolated Ama- containing material from granulomatous lesions and are analyzing this material for

Ama content. This isolation is shown in Example 11.

Example 11

Isolation of Ama-like Material From Granulomatous Lesions

In these studies, normal macrophages were used as control cells against which we contrasted the studies of lipid-laden macrophages. Mass spectroscopy quantitation of Ama revealed that the granulomatous cells contained 3.7 Ama molecules/1000 amino acids versus none detectable in normal peripheral blood monocytes. This finding led us to investigate leukocytes in general.

After isolating leukocytes on Histopaque 1119 (Sigma, St. Louis, MO), they were washed, lysed and run on the Ama affinity column. Eluates were then applied to size exclusion column chromatography, from which we isolated seven separate entities ranging between 6Kd and 200Kd. Each of these were reacted in ELISA with the mAb-Ama_sy,,; reactivity was less than that of the malignant cell material but still significant. Of course, we don't know which of the leukocyte populations are contributing, but this is being studied currently. The eluate taken directly from the Ama affinity column and studied with gel electrophoresis revealed (on silver stain) two bands of material, both falling between 1.5 and 20 Kd. We have also had success in altering the course of immunoreactivity (in the mixed lymphocyte culture reaction) by the addition of mAb-Ama_j-_yn, as shown below.

Example 12 Labeling and Using m-Ah-Ama^ in Fine Structure Studies

A portion of the mAb-Ama_nat stock is conjugated to the following: (a) a fluorescein tag, (b) horseradish peroxidase, or (c) colloidal gold. This permits examination of tissues grossly and microscopically to identify and localize Ama. Colloidal gold is attached to the mAb-Aman-,. which is used to probe tumorigenic tissues in vitro. These tissues are sectioned and observed under the electron microscope revealing many cellular binding regions. The Colloidal gold monoclonal antibodies bind much more specifically to tumor cells than to normal cells. Ama Epitopes as a Marker of Malignant Cells Through immunoaffinity columns incorporating mAb-A an_at, we have isolated products from lymphocytes, macrophages, foam cells (lipid laden macrophages from atheromas) and tumor cells. During these experiments we discovered that the malignant cells contained a relatively large amount of Ama- containing material in comparison to normal cells. Thus, the present invention include characterization of the natural antigens and localization of the molecules in the malignant (versus normal) cells.

The diagnostic capabilities of the present invention can be used in vitro, as described above in connection with the discussion of Examples 11 and 12, and also includes in vivo diagnostic utilities. To use the antibodies in vivo, the antibodies are administered to a patient suspected of having a solid cell tumor. Ordinarily, the antibody is labeled with a pharmaceutically acceptable label, such as a pharmaceutically acceptable radionuclide. The amount of binding to the tissue is then determined by any of a number of techniques known to those of skill in the art, such as an immunoimaging procedure such as nuclear magnetic resonance imaging. This amount is compared to a baseline amount of binding expected for normal tissue. This baseline level can be calculated from similar procedures conducted in normal subjects. For some indications, the baseline amount will be a substantially undetectable amount. If an increased amount of binding is found within the tissue, then there is an indication of a solid cell tumor. This method is believed to be particularly advantageous in the diagnosis of prostate carcinoma due to the observed high level of binding in the prostate carcinoma cell line, DU 145. Divalent Cation Binding

We have also identified a cell surface-associated ama-containing epitope vital for divalent cation binding and have shown that it is necessary for tumor binding to the endothelium during metastasis and is found predominantly on malignant (but not normal) cell membranes. Production of a specific mAb to this epitope has provided a means of controlling tumor growth and metastasis. Such a therapeutic agent is used along with or in conjunction with other agents at various stages of cancer treatment, particularly during and immediately after surgical excision of malignant lesions.

We believe that one of the difficulties in assigning a cation binding site to the integrins lies in the misidentification of Ama (percarboxylated glycine) as Gly.

This may be due to possible false results when analyzing Ama-containing proteins under the conditions of (1) alkaline protein hydrolysis, used in the analysis of malonic acid bearing amino acids, as well as (2) under acid hydrolysis. Reinforcing this belief is the fact that Ama strongly binds Ca^{+ +}. All known integrins require divalent cations for binding yet this binding site has not yet been satisfactorily explained. It is likely that in the RGD (arginine, glycine, aspartate) sequence, if Ama replaces glycine, the free carboxyl groups of the Ama and Asp are utilized to bind the divalent cation. Thus, many Ama- containing epitopes are believed to function as adhesion molecules. Such adhesion molecules appear to be involved in both tumor growth and metastasis.

NMR studies of Ama-containing molecules have been performed to analyze Ca^{+ +} binding to free Ama and Ama in peptide linkages. These NMR studies, as will be readily known to those of skill in the art, include labeling the antibody with a paramagnetic moiety. We have conducted NMR studies of Wallerian degeneration, a lipid degenerative process in nerves in which foam cells appear in the course of monocyte activation and myelin breakdown. The foam cells of Wallerian degeneration do contain the Ama epitope.

We performed electronmicroscopy studies to locate Ama epitopes within cells. Some of these studies are shown as Example 13. Example 13

Studies of mAb-Ama^ with Electronmicroscopy mAb-Ama-_rn was used in immunogold-enhanced electron microscopy studies (in collaboration with Stanley Fowler, University of South Carolina) to locate Ama epitopes in or on the cell. The Ama epitope was found mainly on the cytoplasmic and in some instances, we believe, on the lysosomal membranes. In a study using the mAb- Ama,_™ affinity column, we captured an unexpectedly large amount of material from a human T-cell tumor line (CEM). ICAM-1, an RGD-associated adhesion molecule is found on T cells. In similar immunogold electron microscopy studies, we have also found that Ama epitopes are contained on cell and lysosomal membranes of DU-145 (human prostate carcinoma cell line) cells, but not free in cytoplasm. This lends credence to the physiological role for Ama, such as part of a binding protein that we have set forth herein.

The results of Example 13 together with our prior identification of the epitope at a much lower density on peripheral blood monocytes strengthens our belief that we have identified a public "marker" for tumor cells. Metastasis

Research has shown that metastasis of a mammary carcinoma in rats can be promoted by the addition of a specific complex of coagulation factors, namely the vitamin K dependent serine protease factors II, IX and X. Factor VII, another vitamin K dependent serine protease did not have any effect on metastasis. These serine proteases are activated by vitamin K in a reaction that carboxylates certain of their glutamic acid residues to form Y-carboxyglutamic acid (Gla), a congener of Ama. This carboxylation enables the factors to bind Ca^{+ +} which in turn allows for binding to phospholipids which are found on cell membranes and endothelium.

Similar events occur when tumor cells bind, degrade, and traverse the basement membrane during metastasis. Researchers in this field have shown an interaction between the coagulation system and the spread of tumors. Further research has shown that in rats, warfarin (a vitamin K antagonist anticoagulant) significantly decreases metastasis yet Arvin (an anticoagulant that cleaves fibrinogen) has no effect on metastasis. In fact, the inhibition of metastasis was reversed by replenishment of the coagulation factors (II, VII, IX and X) which warfarin suppresses.

This points away from fibrin (as was originally thought) as the metastasis enhancing culprit and toward a carboxylation reaction of some kind. It is possible that warfarin not only inhibits the Y-carboxylation of Glu, but may also block a similar reaction converting glycine to Ama. This blockage of the conversion from glycine to Ama could potentially occur in the Arg-Gly-Asp (RGD) peptide thought to be the binding site in some integrins.

High concentrations of warfarin (ImM) also decrease tumor growth in vitro, but this dose causes exsanguination in vivo in rats. As discussed above in the section entitled Divalent Cation Binding, we believe that Ama-containing epitopes are needed for adhesion. In accordance with this belief, we also believe that anti-Ama has very similar actions to warfarin in decreasing tumor growth and size. Advantageously, we expect the anti-Ama antibodies to be without the traditional anticoagulant side effects. We confirmed our belief of the therapeutic utility of anti-Ama through in vitro studies on the growth of solid tumor cells, as shown in Example 14.

Example 14

In vitro Functional Analysis of WMK-1 Antibody

DU-145 cells were split 1:2 in normal culture conditions and allowed to attach for four hours. Various concentrations of WMK-1 antibody were added and cell counts were obtained after 54 hours.

The results of the in vitro studies of Example 14 are shown in Figure 3. It can be seen that the mAb-Ama_nat. WMK-1, dramatically decreases the growth of the DU-145 prostatic cancer cells. At concentrations greater than 20 μg/mL, the effect appears to be maximized. DU-145 cells require cellular adhesion in order to grow. These results confirm our belief that Ama-containing polypeptides are involved in adhesion of malignant cells, and that mAb-Ama_nat appears to prevent the cellular adhesion required for growth of such cells. Thus, in vivo studies, such as described in Examples 15 and 16, to test the effect of mAb-Ama_nat ^t0 block tumor growth and metastasis are to be performed.

Example 15 Testing the Ability of mAb-Aman_3t to

Block Tumor Growth and Metastasis

Briefly, SCID (mutant severe combined immune deficient) mice reconstituted with human immune cells (SCID-hu) are to be inoculated with tumor cells from a human pancreatic adenocarcinoma. One group of the mice is treated with mAb- Ama_nat, ^one group with warfarin, one group with both agents, and one group is untreated (control).

Uninoculated SCID-hu mice are also to be treated with mAb-Ama_nat as a control. The mice are sacrificed after 10 weeks and their tumors weighed and their lungs and livers examined for evidence of metastasis. The relative effectiveness of the treatments is evaluated based on the rate of metastasis, size of tumor, and side effects of the regimen on the mice.

The mice treated with mAb-Aman_at show more metastasis than those of the control groups.

Example 16 Blocking Tumor Growth and Metastasis in a Mammal

SCID-hu mice bearing human malignant tumors are injected with mAb- Ama_na, by methods disclosed above. Mice injected with the antibody of the present invention show slower tumor growth, and a lower level of metastasis, in comparison to the controls. The Ama-containing components are of various molecular sizes, indicating that the Ama epitope is found in more than one protein. However, we have discovered that Ama-like material is found with many fold more residues in the diseased than normal cell. Thus, we believe that mAb-Aman_at is an effective anti- tumor growth and anti-metastasis agent. Since Ama is found in adhesion molecules associated with a large number of solid cell tumors, we believe that the mAb-

Ama_nat therapy of the present invention has wide-spread utility across a large number of types of tumors. In Vivo Administration and Pharmaceutical Compositions

The present invention includes the potential for using mAb-Ama_nat as a therapeutic and diagnostic agent for solid cell tumors. Thus, also included within the scope of the present invention are pharmaceutical compositions including the mAb-Ama_nat antibodies. These compositions include the antibody itself in addition to pharmaceutically acceptable carriers for antibodies that are well known to those having ordinary skill in the art. These carriers can include, for example, phosphate buffered saline, and other suitable diluents for intravenous, intramuscular, intraperitoneal or other types of injection. The amount of antibody provided in the therapeutic and/or diagnostic compositions of the present invention can be determined through in vitro testing. Thus, for example, for therapeutic dosage determination the antibody can be determined by culturing the tumor cells to be treated as in Example 14 and determining the minimum dose at which maximal effect on cellular growth in established. Further in vivo testing can be conducted to determine the maximum safe and effective dose in accordance with methods well known in the art. Diagnostic dosages can be determined through any of a variety of in vitro tests, such as is described in Example 12.

In some embodiments, the antibody of the compositions can be labeled. The labeled antibodies can be administered in vivo to a patient suspected of having a solid cell tumor and detected using a radio-imaging procedure, such as scintillography, or another diagnostic procedure known to those of skill in the art. Any of a number of well-known labels can be used, such as a radionuclide.

All references cited herein are explicitly incorporated herein in their entirety by this reference thereto. The present invention has been described in connection with a variety of specific embodiments thereof. However, those of skill in the art will recognize that many alterations in the described techniques can be made without departing from the true scope of the present invention. Thus, the true scope of the present invention is to be measured upon reference to the appended claims.

Claims

WHAT IS CLAIMED IS:

I. A monoclonal antibody produced against an epitope on a native aminomalonic acid-containing moiety, wherein said antibody binds to BSA- aminomalonic acid, but not to BSA.

2. The antibody of Claim 1, wherein said antibody has substantially the same binding activity as the antibody produced by hybridoma strain ATCC No. HB 11146.

3. The antibody of Claim 2, wherein said antibody is produced by hybridoma strain ATCC No. HB 11146.

4. The antibody of Claim 1, wherein said aminomalonic acid-containing moiety can be found in a human cancer cell.

5. The antibody of Claim 4, wherein said cell is from a prostate carcinoma cell line.

6. The antibody of Claim 5, wherein said cell line is DU-145.

7. The antibody of Claim 1, wherein said antibody binds to a cell surface antigen present in malignant cells that is vital for divalent cation binding.

8. An isolated aminomalonic acid-containing polypeptide that can be found in a human cancer cell.

9. A polypeptide according to Claim 8, wherein said polypeptide plays a role in cell adhesion.

10. A polypeptide according to Claim 9, wherein said polypeptide is a cell surface antigen present in malignant cells that is vital for divalent cation binding.

II. A polypeptide according to Claim 7, wherein said polypeptide can be found in a human prostate carcinoma cell line.

12. A polypeptide according to Claim 11, wherein said cell line is DU-

145.

13. A polypeptide according to Claim 11, wherein said polypeptide has an apparent molecular weight in SDS-PAGE of 14 to 80 kD.

SUBSTITUTE SHEET

14., A method of diagnosing a solid cell tumor in a tissue suspected of having such a tumor in a patient, comprising: administering to the tissue an antibody specific for aminomalonic acid; determining the level of binding of said antibody to said tissue; and comparing the determined level of binding of said antibody to a baseline level of binding of said antibody to normal tissue of the same type as said tissue, wherein a higher level of antibody binding to said tissue than said baseline level indicates the presence of a solid cell tumor in said tissue.

15. The method of Claim 14, wherein the baseline level is a substantially undetectable quantity when determined by the method used in the determining step.

16. The method of Claim 14, wherein said tissue comprises prostate tissue, and said tumor comprises a prostate carcinoma.

17. The method of Claim 14, wherein said antibody is labeled.

18. The method of Claim 14, wherein said antibody is produced against an epitope on a native aminomalonic acid-containing moiety.

19. The method of Claim 18, wherein said antibody has substantially the same binding activity as the antibody produced by hybridoma strain ATCC No. HB

11146.

20. The method of Claim 19, wherein said antibody is produced by hybridoma strain ATCC No. HB 11146.

21. The method of Claim 18, wherein said aminomalonic acid-containing moiety can be found in a human cancer cell.

22. The method of Claim 21, wherein said cell is from a prostate carcinoma cell line.

23. The method of Claim 22, wherein said cell line is DU-145.

24. The method of Claim 14, wherein said antibody is labeled with a radioisotope and the determining step comprises a radioimaging procedure.

25. The method of Claim 14, wherein the antibody is labeled with a parimagnetic moiety and the determining step comprises nuclear magnetic resonance imaging.

SUBSTITUTE SHEET

26. The method of Claim 14, wherein the administering step comprises the in vivo administration to said patient of an immunoreactive amount of said antibody.

27. The method of Claim 14, wherein said antibody binds to a polypeptide that plays a role in cell adhesion.

28. The method of Claim 27, wherein said polypeptide is a cell surface antigen present in malignant cells and is vital for divalent cation binding.

29. A method of treating a solid cell tumor in a tissue in a patient, comprising administering to the patient an effective solid tumor-inhibiting amount of a therapeutic antibody specific for a cell surface antigen present in said tumor that is vital for divalent cation binding.

30. The method of Claim 29, wherein said cell surface antigen is recognized by the antibody produced by hybridoma strain ATCC No. HB 11146.

31. The method of Claim 29, wherein said therapeutic antibody has substantially the same binding activity as the antibody produced by hybridoma strain

ATCC No. HB 11146.

32. The method of Claim 29, wherein said solid cell tumor is a prostate cell carcinoma.

33. The method of Claim 29, wherein said therapeutic antibody is administered in an amount effective to reduce the likelihood of metastasis of said tumor.

34. The method of Claim 29, additionally comprising the co- administration of other anti-cancer agents.

35. The method of Claim 29, wherein the therapeutic antibody is administered during or immediately after surgical excision of said tumor.

36. A pharmaceutical composition comprising an antibody according to Claim 1, in an amount effective to bind to a solid cell tumor in a patient, and a pharmaceutically acceptable carrier.

37. The composition of Claim 36, wherein said antibody is present in an amount effective to treat said tumor in said patient.

SUBSTITUTE SHEET

38. The composition of Claim 37, wherein said antibody is present in an amount effective to prevent metastasis of said tumor in said patient.

39. The composition of Claim 36, wherein said antibody is labeled and is present in an amount that specific binding thereof to tissue in said patient can be detected.