US20120071635A1

US20120071635A1 - Methods and compositions for modulating tumor cell activity

Info

Publication number: US20120071635A1
Application number: US13/268,020
Authority: US
Inventors: Maureen D. O'Connor-McCourt; Christiane Cantin; Anne E.G. Lenferink
Original assignee: National Research Council of Canada
Current assignee: National Research Council of Canada
Priority date: 2005-09-13
Filing date: 2011-10-07
Publication date: 2012-03-22
Also published as: CA2621363C; AU2006291990B2; CA2621363A1; US20150111250A1; EP1937815A1; EP1937815A4; US20110033471A1; WO2007030930A1; JP2013155185A; US8044179B2; US8426562B2; US20090104215A1; JP2009507476A; JP5759503B2; DK1937815T3; JP5376948B2; AU2006291990A1; EP1937815B1; ES2543341T3

Abstract

Antibodies which target clusterin, a protein involved in the epithelial-to-mesenchymal transition of carcinoma cells, are identified and characterized. The antibodies may be used to modulate tumour cell activity through binding the clusterin.

Description

PRIORITY CLAIM

This patent application is a divisional of U.S. Ser. No. 11/991,459 filed on Mar. 5, 2008, now U.S. Pat. No. 8,044,179 issued on Oct. 25, 2011 which is a national stage filing under 35 U.S.C. §371 of international application No. PCT/CA2006/001505 filed on Sep. 13, 2006 which claimed priority to U.S. provisional application No. 60/716,086 filed Sep. 13, 2005. The entire contents of each of these priority applications are incorporated herein by reference.

FIELD OF THE INVENTION

The invention relates to antibodies, peptides and small molecules which bind clusterin, and their use in modulating tumor cell activity.

BACKGROUND OF THE INVENTION

Carcinomas, the most common human malignancy, arise from epithelial cells. Progression of epithelial cancers begins with the disruption of cell-cell contacts as well as the acquisition of a migratory (mesenchymal-like) phenotype. This phenomenon, which is called an epithelial-to-mesenchymal transition (EMT), is considered to be a crucial event in late stage tumor progression and metastasis.
The secreted protein TGF-β suppresses tumor growth initially largely due to its growth inhibitory action on tumor cells of epithelial origin, then at later stages promotes tumor cell progression and metastasis. One mechanism by which TGF-β can promote tumor progression is through the induction of an EMT.
Due to the dual role that TGF-β plays in carcinogenesis, direct inhibitors of TGF-β may be risky since, while they could benefit late stage tumors, they could also accelerate preneoplastic lesions. A better therapeutic may be one that inhibits the pro-oncogenic EMT-promoting action of TGF-β, while leaving the tumor suppressor growth-inhibitory action of TGF-β unaffected. To develop such an inhibitor it would be necessary to identify the point at which there is a bifurcation of the TGF-β signaling pathway such that the mediators in one branch of the pathway participate in the EMT response, but not the growth inhibitory response to TGF-β. Therapeutics that inhibit mediators that lie exclusively in the EMT-promoting branch of the TGF-β signaling pathway will reduce metastasis while having little or no effect on the acceleration of preneoplastic lesions.
No TGF-β signal pathway specific components have been generally identified that promote or mediate the EMT-promoting action of TGF-β, yet are not involved in the growth inhibitory action of TGF-β.
In contrast, an endogenous protein (the YY1 nuclear factor) has been identified that is able to interfere with (as opposed to promote) the protumorigenic EMT action of TGF-β, while leaving the tumor-suppressing action (growth inhibition) intact (Kurisaki et al., 2004).
Inhibitors that target TGF-β ligands, receptors and the Smad signaling proteins are known. Specifically, soluble receptor ectodomains, antibodies and other binding proteins are able to act as antagonists by interacting with TGF-β ligands and sequestering them away from cell surface receptors. Small molecules are available that inhibit the kinase activity of the Type I TGF-β receptor and endogenous inhibitors of the Smad signaling proteins are also known. Since all of these signaling pathway components are involved in both the pro- and anti-carcinogenic actions of TGF-β, these inhibitors that target them may benefit late stage tumors, however, they could also accelerate preneoplastic lesions.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1: TGF-β induces an epithelial to mesenchymal transition (EMT) in JM01 cells.

(A) This transition is characterized by an elongated morphology, the relocalization of the markers E-cadherin (E-cad), β-catenin (β-Cat) and F-actin and the down-regulation of the marker Zona Occludens-1 (ZO-1). (B) This morphology change is accompanied by an increase in cell motility as shown in a wound healing assay in which the cells' ability to migrate in to a ‘scratch’ area is monitored in the absence or presence of TGF-β. (C) A complementary black ink motility assay was also used to visualize and quantify the motility of individual JM01 cells in the absence or presence of TGF-β. The black ink which is coated on the plastic sticks to the migrating cells, thereby generating the white tracks. Both assays show that the presence of TGF-β increases the motility of the JM01 cells.

FIG. 2: Analysis of TGF-β-induced gene expression changes using microarray technology. (A) Extensive analysis of microarray data obtained from 7 time-points (0.5, 1, 2, 4, 6, 12, and 24 hrs) during the TGF-β induction of the JM01 cell EMT allowed for the identification of 328 genes that are modulated during the early (0.5, 1 hr), middle (2, 4, 6 hr) or late (12, 24 hr) stages of the transition. (B) Only 5 of these genes are affected over the entire time-course. (C) By comparing our gene list with data on the basal gene expression profiles of the NCI-60 cell line panel (some of these cell lines exhibit a mesenchymal phenotype), and with expression profiling data from clinical samples, we identified 15 genes from our list that are associated with a mesenchymal tumor cell phenotype and with clinical tumor progression.

FIG. 3: Validation of the TGF-β modulation of selected gene expression and protein levels. (A) Semi-quantitative PCR confirmed the TGF-β-induced clusterin up-regulation and caveolin-1 down-regulation thereby validating the microarray analysis (microarray data shown below PCR results). (B) Western blot analysis of whole cell lysates of JM01 cells treated for 24 hrs with TGF-β demonstrated that these transcriptional changes result in increased clusterin (p-clu=pre-clusterin; s-clu=secreted mature clusterin) and decreased caveolin-1 (Cav-1) protein levels. (C) Immunofluorescent microscopy of JM01 cells treated for 24 hrs with TGF-β further confirmed these changes in clusterin and caveolin-1 protein levels through the visualization of these proteins in the intact cell. Nuclei are stained blue, caveolin-1 and clusterin are stained green and the F-actin fibers are stained red.

FIG. 4: Identification of secreted clusterin as a mediator of the TGF-β induced EMT. (A) Immunofluorescent microscopy indicated that clusterin is localized to the secretory pathway in JM01 cells and Western blot analysis of conditioned media (CM) indicated that clusterin is secreted (s-clu). (B, C) JM01 cells were treated for 24 hr with TGF-β, or CM taken from TGF-β treated JM01 cells, in the absence or presence of a antibody raised against the C-terminus of the clusterin β chain (anti-clu). Using immunofluorescent microscopy of ZO-1 as a marker of the EMT it was shown that the clusterin antibody blocks the induction of the EMT by both TGF-β and the CM indicating that secreted clusterin is a necessary mediator in the TGF-β EMT pathway. Purified clusterin alone was also shown to promote the EMT indicating that clusterin is not only necessary, but is sufficient for EMT induction. (D) The induction of the EMT by clusterin alone was further confirmed by using FACS analysis of the epithelial marker E-cadherin to monitor the EMT.

FIG. 5: Clusterin acts as an EMT mediator in cell lines other than the JM01 cells. 4T1 tumor cells (breast) and DU145 tumor cells (prostate) were observed to secrete clusterin and exhibit a motile phenotype in the absence of TGF-β stimulation. Using the wound healing assay to monitor the motility of the 4T1 and DU145 cells, it was observed that a clusterin antibody (anti-clu) inhibits the motility of these cells indicating that clusterin is important for the maintenance of the TGF-β independent mesenchymal phenotype in these cells.

FIG. 6: Clusterin is a pivotal mediator in the pathway leading to TGF-β induction of EMT but not in the pathway leading to TGF-β growth inhibition. (A) Using the black ink motility assay to monitor the EMT of the JM01 cells, it was confirmed that a clusterin antibody blocks the TGF-β induced EMT and that clusterin alone promotes the EMT. (B) This result was further confirmed by quantifying the motility change as area cleared in the ink per cell. (C) In contrast, as monitored by the incorporation of tritiated thymidine, it was shown that the clusterin antibody does not block TGF-β induced growth inhibition and that clusterin alone does not promote growth inhibition, indicating that clusterin is not a mediator in TGF-β growth inhibitory pathways.

FIG. 7: Clusterin is an essential mediator in a TGF-β tumor promoting pathway but not in its tumor suppressing pathway. TGF-β induces secretion of clusterin and antibodies raised against the C-terminus of the clusterin β chain block the TGF-β1 induced EMT, but not the growth inhibitory response of the cells to TGF-β. These results indicate that clusterin is a necessary mediator in the TGF-β EMT pathway but do not address whether other TGF-β-induced mediators act in concert with clusterin to induce the EMT; that is, do not, address the question of whether clusterin alone mediates an EMT. The fact that purified clusterin in the absence of TGF-β also promotes an EMT indicates that clusterin is sufficient to induce this transition.

FIG. 8: Analysis of the neutralizing activity of anti-clusterin polyclonal antibodies produced at BR1. Sera collected from two rabbits (#9 and #10) immunized with a clusterin peptide (a.a. 421-437) were confirmed to contain antibodies that interact with the peptide using surface plasmon resonance (data not shown), and were tested for their ability to inhibit cell motility in a wound healing assay (1/25 dilution of rabbit serum). The mouse mammary epithelial cell line, 4T1 (top), secretes clusterin and is motile in the absence of TGF-β, whereas the JM01 cell line (bottom) requires stimulation with TGF-β to induce clusterin production and cell motility. The sera of both rabbit #9 and #10 inhibit motility, with #10 serum being more potent. As expected, the pre-immune sera of both rabbits does not affect motility. A commercially available clusterin antibody is shown as a positive control (anti-clu, Santa Cruz).

FIG. 9: Analysis of the activity of the anti-clusterin monoclonal antibodies produced at BR1. (A) Immunoprecipitations of recombinant human clusterin (500 ng) using either 50 or 100 ng of each of 12 BR1-produced monoclonal antibodies (commercial polyclonal (C18) and monoclonal (B5) antibodies were used as positive controls). Samples were analyzed on a 12% reducing SDS-PAGE. All antibodies were observed to interact with recombinant clusterin by immunoprecipitation. (B) Assessment of the ability of the 12 BR1-produced monoclonal antibodies to inhibit the TGF-b induced motility of JM01 cells using the black ink motility assay (commercial polyclonal (C18) and monoclonal (B5) antibodies were used as positive controls). The bar graph shows the relative values of the motility of the TGF-b treated BR1-JM01 cells in the presence of the various antibodies. Five BR1-produced monoclonal antibodies (21B12, 20E11, 16C11, 16B5 and 11E2) inhibit the TGF-b induced motility of the BR1-JM01 cells. Values are expressed as the clearance/cell/24 hr relative to that of the TGF-b treated (control) cells. The * illustrates the cut-off value that was used when assessing neutralizing ability. When using this cut-off value in the black ink motility assay, there was a good agreement with the evaluation of the neutralizing ability of the monoclonal antibodies when using the wound healing motility assay (data not shown).

FIG. 10: Two SPR-biosensor (Biacore) approaches to analysing the relationship between the epitopes of antibodies. (A) In the first approach, a rabbit anti-mouse Fc antibody (RAMFc) is covalently immobilized on the sensor chip and one monoclonal (termed Ab 1) is captured on the surface. After binding clusterin to Ab1, the second monoclonal antibody (termed Ab 2) is flowed over the surface. If the epitopes of the two antibodies are overlapping, then Ab2 will not be able to bind to Ab1-bound clusterin. If the two antibodies have unrelated epitopes, then Ab2 will be able to bind to Ab1-bound clusterin. (B) In the second approach, one monoclonal (termed Ab 1) is covalently immobilized on the sensor chip surface. Clusterin is then incubated with a second antibody (monoclonal or polyclonal, termed Ab2) in solution and the complex is then flowed over Ab1. If the epitopes of the two antibodies are overlapping, then Ab2-bound clusterin will not be able to bind to Ab1.

FIG. 11: Results of the analysis of the relationship of the epitopes of the 5 EMT neutralizing BR1-produced anti-clusterin monoclonals antibodies with each other, and with the peptide epitopes of the C18, pAb#10 and B5 antibodies. This table summarizes all the epitope mapping results obtained using the two SPR-biosensor (Biacore) approaches. A blue + indicates that Ab1 competed with Ab2 for binding to clusterin in the first Biacore approach (i.e. the ratio of RUs of Ab2 to RUs of bound clusterin was 0.1 or less). A red + or +/− indicates that Ab2 competed with Ab1 for binding to clusterin in the second Biacore approach (i.e. the binding of clusterin to Ab1 was inhibited between 30-100% for +, and between 10-30% for +/−, when preincubated with Ab2). It is evident that all of the five neutralizing monolconal antibodies (21B12, 20E11, 16C11, 16B5 and 11 E2) interact with the overlapping peptide epitopes of pAb#10, pAbC18 and mAb B5 since they all compete for each other, and for pAb#10, pAbC18 and mAb B5. *It should be noted that all of the negative results from the first approach (blue −) occurred when Ab 20E11 was used (either as Ab1 or Ab2) indicating that this Ab did not behave well in that experimental set up. Therefore, for A δ 20E11, conclusions are taken primarily from the second experimental approach.

FIG. 12: Isolation of the Ig variable region cDNAs. Flow diagram indication the steps for the isolation, sequencing, sequence analysis of the monoclonal variable regions.

FIG. 13: Amino acid sequences of monoclonal antibodies

FIG. 14: CDR1 and CDR2 alignment of clusterin Ig VH

SUMMARY OF THE INVENTION

A first object of the invention is to identify a method for inhibiting EMT in tumour cells without inhibiting the tumour-suppressing activity of TGF-β.
A further object of the invention is to identify molecules or compositions which may inhibit TGF-β-induced EMT in tumour cells without inhibiting the tumour-supressing activity of TGF-β.
A first aspect of the invention provides for an agent having a binding affinity for clusterin, wherein binding of the agent to clusterin inhibits epithelial-to-mesenchymal transition in carcinoma cells. In particular, the agent may bind to the β-subunit of clusterin, and more specifically, it may bind to the C-terminal portion of the clusterin β-subunit. The agent may, for example, be an antibody, including a monoclonal or polyclonal antibody.
A second aspect of the invention provides for a method for modulating the activity of carcinoma cells, comprising the steps of exposing the cells to an agent having a binding affinity for clusterin.
A further aspect of the invention provides for the use of an amino acid sequence in the generation of agents having a binding affinity for clusterin, wherein the sequence comprises SEQ ID NO.: 4 or a portion thereof. In particular, the sequence may comprise shorter portions of SEQ ID NO.: 4, including SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, and SEQ ID NO.: 5.
A further aspect of the invention provides for a vaccine comprising clusterin or a portion thereof which is involved in epithelial-to-mesenchymal transition in carcinoma cells, and a pharmaceutically suitable carrier. The portion of clusterin may comprise SEQ ID NO.: 4 or a portion thereof.
A further aspect of the invention provides for the use of an amino acid sequence in the preparation of a vaccine, wherein the sequence comprises SEQ ID NO.: 4 or a portion thereof. In particular, the sequence may comprise shorter portions of SEQ ID NO.: 4, including SEQ ID NO.: 1, SEQ ID NO.: 2, SEQ ID NO.: 3, and SEQ ID NO.: 5.
A further aspect of the invention provides for a nucleic acid sequence that encodes at least one of SEQ ID NO.: 1 through SEQ ID NO.: 30.
A further aspect of the invention provides for the use of an agent with a binding affinity for clusterin as a diagnostic tool, wherein binding of the agent to clusterin inhibits epithelial-to-mesenchymal transition in carcinoma cells.

DETAILED DESCRIPTION OF THE INVENTION

It is disclosed herein that clusterin is a therapeutic target whose inhibition blocks EMT without preventing TGF-β's anti-proliferative tumor suppressor action.
Clusterin was first identified as a protein possibly involved in EMT using transcriptome analysis, then was analyzed to identify potential binding sites within clusterin. Synthetic peptides were created accordingly, and antibody preparations directed against these peptides were produced or purchased. Additionally, twelve monoclonal antibodies were isolated using full-length recombinant clusterin as the antigen. Both the anti-peptide antibody preparations and the twelve monoclonal antibodies were confirmed to bind to recombinant clusterin. The anti-peptide polyclonal antibody preparations and five of the twelve monoclonal antibodies were shown to inhibit EMT. These five neutralizing monoclonal antibodies were shown to interact with the same peptide epitope as the anti-peptide antibodies.
Using semi-quantitative RT-PCR, Western blot and immunofluorescent microscopy analysis, it was confirmed that several of the EMT-associated transcriptional changes that were detected by microarray analysis were reflected in changes in message and protein abundance (clusterin and caveolin are shown in FIG. 3). Anti-peptide antibodies were used to demonstrate that clusterin is an essential EMT mediator that is not involved in TGF-β's growth inhibitory pathways (FIGS. 4-6). These results indicate that clusterin is an accessible therapeutic target whose inhibition blocks EMT without preventing TGF-β's anti-proliferative tumor suppressor action.
The epitope within clusterin that is important for the generation of EMT-inhibiting agents was elucidated using anti-peptide antibody preparations in neutralization assays. Two different commercial polyclonal antibody preparations raised against synthetic peptides corresponding to sections of the C-terminus of the clusterin β sub-unit were used. The first antibody (from RDI Research Diagnostics Inc.) was raised against the synthetic peptide corresponding to amino acids 421-437 of clusterin (VEVSRKNPKF METVAEK, SEQ ID NO 1) (termed RDI) and the second antibody (from Santa Cruz Biotechnology Inc.) was raised against the synthetic peptide corresponding to amino acids 432-443 of clusterin (ETVAEKALQ EYR, SEQ ID NO 2) (termed C-18). An anti-peptide monoclonal antibody against the same peptide (SEQ ID NO 2) was also purchased (termed B5). The overlap between these two epitopes is shown below. The ability of these antibody preparations to block EMT indicates the significance of the C-terminal portion of the clusterin β subunit in inducing EMT (FIG. 4-6, C-18 results shown; similar results obtained with RDI).

	375
	(SEQ ID NO.: 4)
	LTQGED QYYLRVTTVA SHTSDSDVPS GVTEVVVKLF
	DSDPITVTVP VEVSRKNPKF METVAEK ALQ EYRKKHREE
	449
	Antibody 1
	Antibody 2

Prediction of Putative Functional Subdomains in Clusterin Based on Structural Bioinformatics

Generally, clusterin is thought to be a protein that is only partially structured, containing molten globule fragments. Additionally, it has been classified as an intrinsically disordered protein. Clusterin is postulated to contain several independent classes of binding sites capable of interacting with numerous other binding partners.
The clusterin sequence was examined using bioinformatics programs, namely:

- PredictProtein (Rost, 1996).
- GenTHREADER (Jones, 1999).
- COILS (Lupas, 1996).
- PONDR (Li et al., 1999)

The C-terminal fragment of the β-subunit was identified as a putative binding region. The fragment (a.a. 375-449, SEQ ID NO.: 4), which starts after the second coiled-coil region, is likely unfolded but has some propensity for β-sheet formation.
A synthetic peptide was produced corresponding to a.a. 421-437 of clusterin in order to generate polyclonal antibody preparations at BR1 that are similar to the commercial antibody 1 preparation (RDI) (these new polyclonal preparations are termed pAb#9 and #10). Additionally, full-length human clusterin was expressed in 293 cells and purified in order to use as antigen to generate monoclonal antibodies against full-length human clusterin. Twelve monoclonal antibodies were raised against full-length clusterin and were demonstrated to interact with clusterin by ELISA. These twelve antibodies are named 6E12, 7B7, 21B12, 20G3, 20E11, 18F4, 16C11, 16B5, 11E2, 8F6, 7D6, 7C12.
The polyclonal antibody preparations raised against the a.a. 421-437 epitope (pAb#9 and #10) were confirmed to inhibit the EMT (FIG. 8).
All twelve monoclonal antibody preparations raised against full-length human clusterin were confirmed to interact with recombinant human clusterin as evidenced by their ability to immunoprecipitate clusterin (FIG. 9A). Five of the twelve monoclonals were shown to be able to neutralize the EMT promoting action of clusterin in the black ink cell motility assay (FIG. 9B) and the wound healing cell motility assay (not shown). The five monoclonal antibodies that neutralize are 11E2, 21B12, 20E11, 16C11, 16B5.
Two Surface Plasmon Resonance (SPR)-based biosensor epitope mapping assays (FIG. 10) were used to determine whether the five neutralizing monoclonal antibodies generated using full-length clusterin were interacting with the same clusterin peptide epitope as the anti-peptide antibody preparations.
The two approaches that were used are described below:
1) The monoclonal antibodies were individually captured on a CM5 sensor chip surface on which a Rabbit-anti-Mouse Fc antibody was covalently immobilized (when captured, the mAb is termed mAb1 in this experimental approach). Clusterin was then allowed to bind to mAb1. Then all five monoclonal antibodies were sequentially injected over mAb1-bound clusterin (the injected mAb is termed mAb2 in this experimental approach) in order to determine if both mAb1 and mAb2 are able to interact with clusterin simultaneously (FIG. 11). It was found that all of the five neutralizing mAbs (except 20E11 in some cases) competed with each other for binding to clusterin (when used both as mAb1 or as mAb2). Additionally, they were found to compete with the C18, pAb#10 and B5 anti-peptide antibodies, suggesting that the five neutralizing mAbs interact with the overlapping peptide epitopes of pAb#10, pAbC18 and mAb B5. It should be noted that, although Ab 20E11 appeared to have a distinct epitope in some cases (when used either as mAb1 or mAb2), this conclusion was not supported by the results of the second experimental approach.
2) The monoclonal antibodies were individually covalently immobilized on a CM5 sensor chip surface using amine coupling (when immobilized, the mAb is termed mAb1 in this experimental approach). To demonstrate competition for binding to clusterin, an Ab (termed Ab2 in this approach) was then incubated with clusterin prior to injection of the complex over the mAb1 surface (FIG. 11).
It was confirmed that all of the five neutralizing mAbs competed with each other for binding to clusterin, and with the C18, pAb#10 and B5 anti-peptide antibodies. This confirms that the five neutralizing mAbs interact with the overlapping peptide epitopes of pAb#10, pAbC18 and mAb B5.
The hypervariable complementary determining regions (CDRs) of all twelve monoclonal Abs were sequenced. Mammalian light- and heavy-chain Igs contain conserved regions adjacent to the CDRs and the use of appropriately designed oligonucleotide primer sets enabled the CDRs to be specifically amplified using PCR (FIG. 12). These products were then sequenced directly (SEQ ID NO 8-30; see FIG. 13).
By aligning the CDR sequences of four out of the five neutralizing monoclonal antibodies (11E2, 21B12, 20E11, 16C11), we were able to determine a consensus sequence for VH CDR1 and CDR2 of these anti-clusterin antibodies (see FIG. 14). The following consensus sequences were determined: CDR-1: G-Y-S/T-F-T-X-Y-X (SEQ ID NO.: 6) and CDR-2: I-N/D-P/T-Y/E-X-G-X-P/T (SEQ ID NO.: 7).
The antibodies or peptides that interact with the epitope of clusterin defined here may be applied as therapeutics, i.e. they may act as a therapeutic in their own right due to their intrinsic ability to neutralize the EMT promoting activity of clusterin. Additionally, these antibodies and peptides may be used as a therapeutic due to their ability to target toxins, suicide genes or other agents with anti-tumor activity to the vicinity of tumor cells through their interaction with secreted clusterin.
Small molecules that interact with the epitope of clusterin defined here may also act as therapeutics by blocking the EMT promoting activity of clusterin. These antibodies, peptides and small molecules that exert their therapeutic activity by interacting with this clusterin epitope may exhibit less toxicity or side-effects as compared to other agents that remove all activities of clusterin, i.e. antisense or RNAi agents, since, while the EMT activity of clusterin is neutralized when this epitope is blocked, the other activities of clusterin may remain intact.
Other applications of the antibodies and peptides that interact with the epitope of clusterin defined here may be as 1) non-imaging diagnostics, i,e, they may detect clusterin as a biomarker in accessible body fluids or in tissue/tumor samples for diagnostic and prognostic applications in cancer, and 2) imaging diagnostics, i.e. they may be used to target contrast agents to tumors for imaging in vivo due to their interaction with secreted clusterin.
Antibodies comprising the heavy and light sequences identified herein, antibodies comprising the CDRs (complementarity determining regions) identified herein (FIG. 13), and antibodies comprising the consensus sequences (FIG. 14) are expected to be useful for the above-mentioned purposes.
Clusterin itself, or the portions thereof which contain the epitope recognized by the antibodies and peptides discussed above, may be used as a vaccine. Preferably, the clusterin should be combined with a pharmaceutically suitable carrier. Clusterin or epitope-containing portions of clusterin may also be used in the generation of vaccines. Similarly, amino acid sequences having at least 90% identity with SEQ ID NO. 4 or the clusterin epitope identified herein will also be useful, since they are likely to have similar functionality to the specific sequences identified herein.
Cell culture, antibodies and reagents BR1-JM01 cells were isolated and characterized as described (Lenferink et al., Breast Cancer Res., 6, R514-30 (2004)). Cells were maintained at 37° C. in a humidified, 5% CO₂atmosphere and cultured in DF/5% FBS (1:1 mixture of Ham's F12 and Dulbecco's modified Eagles Medium (DMEM) with 5% Fetal Bovine Serum (FBS) and antibiotics/antimicotics (both Wisent Inc.)).
Human recombinant TGF-β1 and pan-TGF-β neutralizing antibody 1D11 were reconstituted according to the manufacturer's instructions (R&D Systems). Purified human serum clusterin was kindly provided by Dr MR Wilson (Wilson and Easterbrook-Smith, 1992). Purified human recombinant clusterin was produced in HEK-293 cells (general expression system described in Durocher et al, 2002). Antibodies against the following proteins were purchased and used in the indicated v/v dilutions: E-cadherin (E-cad, anti-uvomorulin clone Decma-1; Sigma), Zona Occludens-1 (ZO-1; Chemicon), polyclonal antibodies raised against the C-terminus of the human clusterin β chain (cluβ; RDI and Santa Cruz), and Caveolin-1 (cav-1; Santa Cruz). Horseradish peroxidase (HRP) conjugated antibodies were obtained from Jackson ImmunoResearch Laboratories Inc and Alexa-488 labeled antibodies and Texas-red labeled phalloidin were purchased from Molecular Probes. All experiments were carried out with 75-80% confluent monolayers of BR1-JM01 cells in DF/5%. Where indicated, cells were treated for 24 hr or 48 hr with TGF-β1 or purified clusterin at a final concentration of 100 μM or 200 nM, respectively.

RNA Isolation and Labeling

Monolayers of BR1-JM01 cells were grown in the absence or presence of TGF-β1 for 30 min, 1, 2, 4, 6, 12 or 24 hr. PolyA+ mRNA was extracted (4×150 mm dishes per time point) using the FastTrack™ 2.0 kit (Invitrogen) according to the manufacturer's instructions. RNA was isolated and labeled according to Schade et al., 2004.

Hybridization and Data Analysis

cDNA microarrays (15,264 sequence verified mouse ESTs; http://lgsun.grc.nia.nih.gov/cDNA/15k.html) were obtained from the University Health Network Microarray Center in Toronto (http://www.microarrays.ca/). Slides were hybridized with Cy3 or Cy5 labeled cDNA as described (Enjalbert et al., 2003), scanned using a ScanArray 5000 (Perkin Elmer v2.11) at a 10-micron resolution and 16-bit TIFF files were quantified using QuantArray software (Perkin Elmer, v3.0). Microarray data normalization and analysis was performed as described (Enjalbert et al., 2003).

Northern Blot and Semi-Quantitative RT-PCR (SQ-RT-PCR) Analysis

For SQ-RT-PCR, 3-5 μg of total RNA was amplified in a 20 μl first-strand RT-PCR reaction using 50 U SuperScript II (Invitrogen) according to the manufacturer's guidelines with modifications. Samples were preincubated (2 min, 42° C.) before adding SuperScript II and the RNaseOUT treatment was omitted. Samples were incubated (90 min, 42° C.) and then cooled on ice. Two μl of first-strand reaction was added to the PCR mix (2.5 U Taq polymerase (New England Biolabs), 10 μM forward/reversed primers) in a final volume of 50 μl, which was heated (2 min, 94° C.) prior to PCR amplification. Primers for the generation of the probes used for northern blot and SQ-RT-PCR are listed in Table 1.

Western Blot Analysis

BR1-JM01 cells grown in 35 mm dishes were treated with TGF-β1 (24 hr). Cells were lysed in hot 2% SDS. Fifty μg of total protein or 30 μl of conditioned medium was resolved by SDS-PAGE (10%) under reducing conditions. Proteins were transferred to nitrocellulose and membranes incubated with primary antibodies (cluβ, cav-1; 1/500) in TBS-T (20 mM Tris-HCl (pH 7.6), 137 mM NaCl, 0.1% Tween 20 (v/v)) containing 5% non-fat milk (overnight, 4° C.). Membranes were washed with TBS-T, incubated with secondary HRP-conjugated antibody (1/20,000) in TBS-T+5% milk (1 hr), and washed with TBS-T. Immunoreactive bands were visualized using Enhanced Chemiluminescence (ECL; Perkin Elmer).

Immunofluorescence Microscopy

BR1-JM01 cells were seeded in glass chamber slides (Lab-Tek) and treated with purified clusterin or TGF-β1 preincubated (30 min) with or without cluβ antibody (8 μg/ml) or 1D11 (100 nM). Conditioned medium, obtained from non-treated and TGF-β1-treated BR1-JM01 cells (24 hr), was preincubated (30 min) with these antibodies prior to incubation with non-treated BR1-JM01 cells. After 24 hr of exposure, cells were fixed with 4% para-formaldehyde (10 min), rinsed twice (PBS), permeabilized (2 min, 0.2% Triton X-100 in PBS), rinsed again, and non-specific sites were blocked with 10% FBS in PBS (40 min). Para-formaldehyde fixed cells were then incubated (1 hr) with primary antibody (E-cad, 1/200; ZO-1, 1/100; cluβ, cav-1; 1/50) in PBS/10% FBS, were rinsed (4× in PBS) and finally were incubated with fluorescently conjugated secondary antibodies (Molecular Probes). Simultaneously, F-actin filaments were labeled with Texas-red labeled phalloidin (1/100) and nuclei were counterstained with 0.4 μg/ml 4,6-diamidino-2-phenylindole (DAPI; Sigma). Slides were rinsed (PBS) and mounted using Prolong anti-fade (Molecular Probes). Fluorescent images were captured using a Princeton Instrument Coolsnap CCD digital camera mounted on Leitz Aristoplan microscope and analyzed using Eclipse (Empix Imaging Inc.) and Photoshop (Adobe) software.

Cell Proliferation Assays

BR1-JM01 cells (2.5×10⁴cells/well) were seeded in 24-well plates. The next day the medium was replenished and purified clusterin, TGF-β1, or TGF-β1 pre-incubated for 30 min with 1D11 antibody (100 nM) or cluβ antibody (8 μg/ml), was added to the cells. After 24 hr, cells were pulse-labeled with 0.5 μCi/ml [³H]thymidine (Amersham), rinsed (PBS, 4° C.), trypsinized and [³H]thymidine incorporation was evaluated by liquid scintillation counting.

Cell Motility Assays

Cells (2×10⁴cells/well) were seeded in ink-coated 12-well plates according to Al-Moustafa et al. (1999) in the absence or presence of TGF-β1, TGF-β1+cluβ antibody, or purified clusterin. Images were captured after 24 hr using a Nikon Coolpix 995 digital camera mounted on Leitz Aristoplan microscope and particle-free tracks were quantified using ImageJ freeware (http://rsb.info.nih.gov/ij/).

Black Ink Motility Assay

Wound Healing Motility Assay

Confluent cell monolayers (12-well plates) were “wounded” using a 2 μL pipet tip. The medium was then replenished, to remove cell debris, and the anti-clusterin mAbs were added (final concentration of 4 μg/mL) in the absence or presence of 100 μM TGF-β. Images of the wound were captured prior to and after 24 hr of incubation using a Nikon Coolpix 995 digital camera mounted on Leitz Aristoplan microscope.

Polyclonal Antibody Production

The peptide (a.a. 421-437 of the clusterin protein) was produced and purified at the University of Calgary (http://peplab.myweb.med.ucalgary.ca/). An extra cysteine was added to the C-terminus of the peptide to facilitate oriented coupling on the surface of the CM-5 sensor chips that were used for screening of the rabbit antisera by surface plasmon resonance (SPR, Biacore™ 2000). The peptide was coupled to Keyhole Lympet Hemocyanin (KLH, Imject Mariculture KLH; Pierce) using either glutaraldehyde (Sigma) or 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide HCL (Pierce) and dialyzed against PBS (overnight at 4° C.). The peptide preparations that were conjugated by the two methods were mixed (1:1). Pre-immune serum was drawn from two female New Zealand white rabbits (10 ml), which were then injected with the KLH-coupled peptide preparation (1.25 ug peptide per leg/0.5 ml Freund's Incomplete Adjuvant or PBS). Animals were boosted (1.25 ug peptide per leg/0.5 ml Freund's Incomplete Adjuvant or PBS) every third week and serum was drawn (6 ml/kg) every 10 days after each boost until the antibody titer did not increase, at which point the animals were euthanized and exsanguinated.
Sera were tested for antibody activity using SPR. For this, the peptide was coupled to a CM-5 sensor chip (Biacore Inc.) using the Thiol coupling method (as described by the manufacturer) and dilutions (1/50) of the pre-immune sera, the antibody-containing sera and the commercially available anti-clusterin antibody (Santa Cruz) were run over the peptide surface.

Monoclonal Antibody Production

Four BALB/c mice were injected subcutaneously (s.c.) and intra-peritoneally (i.p.) with 35 μg of purified human clusterin emulsified in TiterMax adjuvant (Pierce). Animals were re-injected i.p. three weeks later and the serum titer was assessed 10 days later. Ten weeks later, responsive mice was boosted by i.p. injections (50 μg purified clusterin) and sacrificed three days later. Spleen cells harvested, fused with NSO myeloma cells and immediately plated (5×10⁴cells/well in 96-well microplates; Costar) in Iscove's medium supplemented with 20% FBS, 100 μM hypoxanthine, 0.4 μM aminopterin and 16 μM thymidine (HAT medium), murine IL-6 (1 ng/ml), penicillin (50 U/ml) and streptomycin (50 μg/ml). Supernatants (10-20 days post-fusion) were tested for anti-clusterin activity on immobilized purified clusterin by Enzyme-Linked Immunosorbent Assay (ELISA). Antibody producing cells were cloned and retested twice for anti-clusterin activity. Thirteen anti-clusterin antibody producing clones were generated of which frozen stocks were prepared and a large-scale antibody production was initiated.

SPR-Based Biosensor (Biacore) Epitope Mapping

Approach 1:

- Running buffer:
  - HBS (20 mM Hepes (pH7.4), 150 mM NaCl, 3.4 mM EDTA, 0.005% Tween 20)
  - All experiments were run at 5 μL/min
- Standard amine coupling of the anti mouse Fc immunoglobulin:
  - Inject 35 μL of a mixture of 0.05M NHS and 0.2M EDC
  - Inject antibodies diluted in 10 mM NaAc pH5.0 at concentration of 30 μg/mL until an appropriate amount in captured
  - Inject 35 μL 1M ethanolamine-HCL pH8.5
- Epitope mapping:
  - Inject 25 μL of mAb1 at a concentration of 25 or 50 μg/mL.
  - Inject 25 μL of a mixture of IgG1, IgG2a, IgG2b and IgG3 each one at a concentration of 25 μg/mL.
  - Inject 25 μL of human recombinant clusterin at a concentration of 30 μg/mL.
  - Inject 25 μL of mAb2 at a concentration of 25 or 50 μg/mL.
- Control:
  - For each pair of antibodies, the non-specific binding of mab2 was determined by repeating all injections described in the epitope mapping section but injecting running buffer instead of clusterin.
  - The response (RU) obtained 20 sec after the end of the mab2 injection in the control was subtracted from the response obtained in the presence of clusterin.
- Regeneration of the surface:
  - At the end of each cycle, inject 10 μL of 20 mM glycine pH1.7 followed with 10 μL of 100 mM HCl.

Approach 2:

- Running buffer:
  - HBS (20 mM Hepes (pH7.4), 150 mM NaCl, 3.4 mM EDTA, 0.005% Tween 20)
- Standard amine coupling of the antibodies:
  - Inject 35 μL of a mixture of 0.05M NHS and 0.2M EDC
  - Inject antibodies diluted in 10 mM NaAc (pH4.5 or 5.0) at concentration raging from 20 to 80 μg/mL until a appropriate amount in captured
  - Inject 35 μL 1M ethanolamine-HCl pH8.5
- Preparation of control surface
  - Inject 35 μL of a mixture of 0.05M NHS and 0.2M EDC
  - Inject 35 μL 1M ethanolamine-HCl pH8.5
- Competition
  - Mix human recombinant clusterin at 50 nM with 250 nM or 500 nM antibodies in PBS (without Mg++ and Ca++)
  - Prepare a tube with antibody alone
  - Inject at a flow of 5 μL/min, 25 μL of clusterin alone, antibody alone or clusterin preincubated with antibodies over the antibody and the control surfaces.
  - Subtract the response obtained for the antibody alone solution from the response obtained for clusterin preincubated with the same antibody.
  - Calculate the % binding inhibition by dividing the response obtained for the clusterin preincubated with antibody by the response obtained for clusterin alone.
- Regeneration solution
  - At the end of each cycle, inject 10 μL of 10 mM HCl at a flow rate of 20 μL/min

Immunoprecipitation

50 or 100 ng of the various monoclonal antibodies or the polyclonal antibody preparation (C18) was incubated with 204 of protein G slush (1:1 in PBS) overnight at 4° C. Then 500 ng of human recombinant clusterin was added and the mixture was incubated for another 2 hr at 4° C. Immunocomplexes were washed 3 times with 1 mL of buffer (150 nM NaCl, 50 mM Tris pH 8.0, 0.55% NP-40, 50 mM Na fluoride) and 20 μL of reducing sample buffer was added. Samples were boiled for 5 min prior to loading on a 12% SDS-PAGE. Separated proteins were then transferred to nitrocellulose and membranes were probed with anti-clusterin antibodies as described.

Sequencing of the Monoclonal Antibody Variable Region

Total RNA was isolated from the 12 hybridomas and first strand cDNA was prepared with reverse transcriptase and the Ig-3 constant region primer followed by amplification with the appropriate Ig-5′ primer. These primer sets used in conjunction with KOD Hot Start DNA Polymerase specifically amplify the variable regions of light- and heavy-chain cDNAs. PCR products can be directly cloned with Novagen's pSTBlue-1 Perfectly Blunt™ Cloning Kit or treated with the Single dA™ Tailing Kit and cloned into the pSTBlue-1 AccepTor™ Vector. For details see FIG. 13.

TABLE 1

Primer sets used for the validation of some of the 328
TGF-β modulated genes in the BRI-JM01 cells.

	Gene
Gene	Bank #	Reverse	Forward	size(bp)

Eefla1	AW556381	CTGGCTTCACTGCTCAGGT	TGGCCAATTGAGACAAACAG	457	Clusterin

	AU041878	TGGTGAAAGCTGTTTGACTCTG	AAGGCGGCTTTTATTGGATT	355	Integrin α6

	AW556992	ATGTGCCATTGTTGCTTTGA	CAAGCGATGAGCACTTTTGT	517	Caveolin-1

	AU016590	GTGCAGGAAGGAGAGAATGG	GCACACCAAGGAGATTGACC	247

Ptpn13	AW548343	CCTGCAATGGTTCTTGGTTT	GGGAAAATCGATGTTGGAGA	300	14-3-3σ

	AA410123	GGGCTGTTGGCTATCTCGTA	AGAGACCGAGCTCAGAGGTG	297

Inclusion of a reference is neither an admission nor a suggestion that it is relevant to the patentability of anything disclosed herein

Bailey et al., Biochemistry. 2001; 40:11828-40
Dunker et al., J Mol Graph Model. 2001; 19 (1):26-59
Li et al., Genome Inform. Ser. Workshop Genome Inform. 1999; 10: 30-
Jones, J. Mol. Biol. 1999; 287: 797-815
Lupas, Meth. in Enzym. 1996; 266: 513-525
Rost, Meth. in Enzym. 1996; 266: 525-539
Singh et al., Curr Opin Drug Discov Devel. 2004: 437-445
Al-Moustafa et al., Biotechniques. 1999: 60-62
Durocher et at Nucleic Acids Res 2002: E9
Enjalbert et al., Mol Biol Cell. 2003: 1460-1467
Schade et al., Mol Blot Cell 2004: 5492-5502
Wilson and Easterbrook-Smith, Biochim Biophys Acta 1992: 319-326

Claims

1-30. (canceled)

31. A monoclonal antibody or antigen binding fragment capable of specific binding to a β-subunit of human clusterin, wherein said monoclonal antibody binds to an epitope including amino acids 421-443 of the β-subunit of human clusterin more efficiently than a second monoclonal antibody raised against a synthetic peptide including or comprised within said epitope.

32. The monoclonal antibody or antigen binding fragment of claim 31, wherein said monoclonal antibody does not bind significantly to a synthetic peptide including or comprised within said epitope.

33. The monoclonal antibody or antigen binding fragment of claim 31, wherein the monoclonal antibody inhibits migration of cells in an ink clearance assay better than the second antibody.

34. The monoclonal antibody or antigen binding fragment of claim 31, wherein the monoclonal antibody or antigen binding fragment is capable of inhibiting epithelial-to-mesenchymal transitions in carcinoma cells.

35. A monoclonal antibody or an antigen binding fragment capable of specific binding to a β-subunit of human clusterin, wherein said monoclonal antibody binds to an epitope including amino acids 421-443 of the β-subunit of human clusterin and comprises a) a light chain variable region; and b) a heavy chain variable region comprising a CDR1 as set forth in SEQ ID NO.:6 and a CDR2 as set forth in SEQ ID NO.:7.

36. The monoclonal antibody or antigen binding fragment of claim 35, wherein the antibody or antigen binding fragment comprises a light chain variable region having complementary determining regions identical to those of SEQ ID NO.:8 or SEQ ID NO.:9 and a heavy chain variable having a CDR1 as set forth in SEQ ID NO.:6, a CDR2 as set forth in SEQ ID NO.:7 and a CDR3 as set forth in SEQ ID NO.165.

37. The monoclonal antibody or antigen binding fragment of claim 36, wherein the CDR1 is as set forth in SEQ ID NO.: 85.

38. The monoclonal antibody or antigen binding fragment of claim 36, wherein the CDR2 is as set forth in SEQ ID NO.:119.

39. The monoclonal antibody or antigen binding fragment of claim 35, wherein the antibody or antigen binding fragment comprises a light chain variable region having complementary determining regions identical to those of SEQ ID NO.:10 and a heavy chain variable having a CDR1 as set forth in SEQ ID NO.:6, a CDR2 as set forth in SEQ ID NO.:7 and a CDR3 as set forth in SEQ ID NO.166.

40. The monoclonal antibody or antigen binding fragment of claim 39, wherein the CDR1 is as set forth in SEQ ID NO.:86.

41. The monoclonal antibody or antigen binding fragment of claim 39, wherein the CDR2 is as set forth in SEQ ID NO.:120.

42. The monoclonal antibody or antigen binding fragment of claim 35, wherein the antibody or antigen binding fragment comprises a light chain variable region having complementary determining regions identical to those of SEQ ID NO.:11 and a heavy chain variable having a CDR1 as set forth in SEQ ID NO.:6, a CDR2 as set forth in SEQ ID NO.:7 and a CDR3 as set forth in SEQ ID NO.167.

43. The monoclonal antibody or antigen binding fragment of claim 42, wherein the CDR1 is as set forth in SEQ ID NO.:87.

44. The monoclonal antibody or antigen binding fragment of claim 42, wherein the CDR2 is as set forth in SEQ NO.:121.

45. A conjugate comprising a) an antibody or antigen binding fragment capable of specific binding to a β-subunit of human clusterin, wherein said monoclonal antibody binds to an epitope including amino acids 421-443 of the β-subunit of human clusterin and b) an agent having anti-tumor activity or a contrast agent.

46. The conjugate of claim 45, wherein the agent having anti-tumor activity is a toxin.

47. A pharmaceutical composition comprising the antibody or antigen binding fragment of claim 31.

48. A pharmaceutical composition comprising the antibody or antigen binding fragment of claim 35.

49. A pharmaceutical composition comprising the conjugate of claim 45.

50. The pharmaceutical composition, wherein the agent having anti-tumor activity is a toxin.