US20090280508A1

US20090280508A1 - Methods and Reagents for Preparing and Using Immunological Agents Specific for P-Glycoprotein

Info

Publication number: US20090280508A1
Application number: US11/566,989
Authority: US
Inventors: Eugene Mechetner; John Fruehauf
Original assignee: Oncotech Inc
Current assignee: Oncotech Inc
Priority date: 1996-11-15
Filing date: 2006-12-05
Publication date: 2009-11-12
Also published as: US6365357B1; US6630327B1; WO1998021325A1; US7144704B2; US20060211051A1; US20040166110A1; US5994088A

Abstract

This invention relates to immunological reagents and methods specific for a mammalian, transmembrane protein termed Pgp, having a non-specific efflux pump activity established in the art as being a component of clinically-important multidrug resistance in cancer patients undergoing chemotherapy. The invention provides methods for developing and using immunological reagents specific for certain mutant forms of Pgp and for wild-type Pgp in a conformation associated with substrate binding or in the presence of ATP depleting agents. The invention also provides improved methods for identifying and characterizing anticancer compounds.

Description

This application is a continuation of U.S. patent application Ser. No. 10/114,847, filed Apr. 2, 2002, now U.S. Pat. No. 7,144,704, granted Dec. 5, 2006, which is a divisional of U.S. patent application Ser. No. 09/316,167, filed May 21, 1999, now U.S. Pat. No. 6,365,357, granted Apr. 2, 2002, which is a continuation-in-part of U.S. patent application Ser. No. 08/752,447, filed Nov. 15, 1996, now U.S. Pat. No. 5,994,088, granted Nov. 30, 1999, the disclosure of each of which is incorporated by reference in its entirety herein.

BACKGROUND OF THE INVENTION

1. Field of the Invention
The present invention relates to uses of immunological reagents specific for a human transmembrane efflux pump protein (P-glycoprotein). Specifically, the invention relates to uses of such immunological reagents that specifically recognize P-glycoprotein that is in a biochemical conformation adopted in the presence of certain cytotoxic, lipophilic drugs that are substrates for P-glycoprotein, in the presence of cellular ATP depleting agents, and by certain mutant embodiments of Pgp. In particular, the invention provides methods of using such immunological reagents for anticancer drug screening and development.
2. Background of the Invention
Many human cancers express intrinsically or develop spontaneously resistance to several classes of anticancer drugs, each with a different structure and different mechanism of action. This phenomenon, which can be mimicked in cultured mammalian cells selected for resistance to certain plant alkaloids or antitumor antibiotics such as colchicine, vinblastine and doxorubicin (formerly known as Adriamycin), is generally referred to as multidrug resistance (“MDR”; see Roninson (ed)., 1991, Molecular and Cellular Biology of Multidrug Resistance in Tumor Cells, Plenum Press, N.Y., 1991; Gottesman et al., 1991, in Biochemical Bases for Multidrug Resistance in Cancer, Academic Press, N.Y., Chapter 11 for reviews). The MDR phenotype presents a major obstacle to successful cancer chemotherapy in human patients.
MDR frequently appears to result from decreased intracellular accumulation of anticancer drugs as a consequence of increased drug efflux related to alterations at the cellular plasma membrane. When mutant cell lines having the MDR phenotype are isolated, they are found to express an ATP-dependent non-specific molecular “pump” protein (generally known as P-glycoprotein) that is located in the plasma membrane and keeps the intracellular accumulation of an anti-cancer drug low enough to evoke the drug-resistance phenotype. This protein (which has been determined to be the gene product of the MDR1 gene in humans) facilitates active (i.e., energy-dependent) drug efflux from the cell, against a concentration gradient of (generally) lipophilic compounds, including many cytotoxic drugs.
The gene encoding P-glycoprotein (which is also known as gp170-180 and the multidrug transporter) has been cloned from cultured human cells by Roninson et al. (see U.S. Pat. No. 5,206,352, issued Apr. 27, 1993), and is generally referred to as MDR1. The protein product of the MDR1 gene, most generally known as P-glycoprotein (“Pgp”), is a 170-180 kilodalton (kDa) transmembrane protein having the aforementioned energy-dependent efflux pump activity.
Molecular analysis of the MDR1 gene indicates that Pgp consists of 1280 amino acids distributed between two homologous halves (having 43% sequence identity of amino acid residues), each half of the molecule comprising six hydrophobic transmembrane domains and an ATP binding site within a cytoplasmic loop. Only about 8% of the molecule is extracellular, and carbohydrate moieties (approximately 30 kDa) are bound to sites in this region (Chen et al., 1986, Cell 47: 381-387).
Expression of Pgp on the cell surface is sufficient to render cells resistant to many (but not all) cytotoxic drugs, including many anti-cancer agents. Pgp-mediated MDR appears to be an important clinical component of drug resistance in tumors of different types, and MDR1 gene expression correlates with resistance to chemotherapy in different types of cancer.
Pgp is also constitutively expressed in many normal cells and tissues (see Cordon-Cardo et al., 1990, J. Histochem. Cytochem. 38: 1277; and Thiebaut et al., 1987, Proc. Natl. Acad. Sci. USA 84: 7735 for reviews). In hematopoietic cells, Neyfakh et al. (1989, Exp. Cancer Res. 185: 496) have shown that certain subsets of human and murine lymphocytes efflux Rh123, a fluorescent dye that is a Pgp substrate, and this process can be blocked by small molecule inhibitors of Pgp. It has been demonstrated more recently that Pgp is expressed on the cell-surface membranes of pluripotent stem cells, NK cells, CD4- and CD8-positive T lymphocytes, and B lymphocytes (Chaudhary et al., 1992, Blood 80: 2735; Drach et al., 1992, Blood 80: 2729; Kimecki et al., 1994, Blood 83: 2451; Chaudhary et al., 1991, Cell 66: 85). Pgp expression on the cell surface membranes of different subsets of human lymphocytes has been extensively documented (Coon et al., 1991, Human Immunol. 32: 134; Tiirikainen et al., 1992, Ann. Hematol. 65: 124; Schluesener et al., 1992, Immunopharmacology 23: 37; Gupta et al., 1993, J. Clin. Immunol. 13: 289). Although recent studies suggest that Pgp plays a role in normal physiological functions of immune cells (Witkowski et al., 1994, J. Immunol. 153: 658; Kobayashi et al., 1994, Biochem. Pharmacol. 48: 1641; Raghu et al., 1996, Exp. Hematol. 24: 1030-1036), the physiological role of Pgp in normal immune cells has remained unclear to date.
Once the central role in MDR played by Pgp was uncovered, agents with a potential for reversing MDR phenotypes were developed that target Pgp. Several classes of drugs, including calcium channel blockers (e.g., verapamil), immunosuppresants (such as cyclosporines and steroid hormones), calmodulin inhibitors, and other compounds, were found (often fortuitously) to enhance the intracellular accumulation and cytotoxic action of Pgp-transported drugs (Ford et al., 1990, Pharm. Rev. 42: 155). Many of these agents were found to inhibit either drug binding or drug transport by Pgp (Akiyamaetal., 1988, Molec. Pharm. 33: 144; Horio et al., 1988, Proc. Natl. Acad. Sci. USA 84: 3580). Some of these agents themselves were found to bind to and be effluxed by Pgp, suggesting that their enhancing effects on the cytotoxicity of Pgp substrates are due, at least in part, to competition for drug binding sites on this protein (Cornwell et al., 1986, J. Bio. Chem. 261: 7921; Tamai, 1990, J. Biochem. Molec Biol. 265: 16509).
Many of these agents, however, also have strong, deleterious side effects at physiologically-achievable concentrations. These systemic side effects severely limit the clinical use of these agents as specific inhibitors of Pgp or for negative selection against Pgp-expressing tumor cells. Most of the known MDR-reversing drugs used in clinical trials have major side effects unrelated to inhibition of Pgp, such as calcium channel blockage (verapamil) or immunosuppression (cyclosporines and steroids). Similarly, targeting of cytotoxic drugs to Pgp-expressing cells is capable of compromising normal tissue function in normal cells (such as kidney, liver, colonic epithelium, etc.) that normally express Pgp. These drawbacks restrict the clinically-achievable dose of such agents and ultimately, their usefulness.
Immunological reagents also provide a means for specifically inhibiting drug efflux mediated by Pgp. Monoclonal antibodies specific for Pgp are known in the art.
Hamada et al., 1986, Proc. Natl. Acad. Sci. USA 83: 7785 disclose the mAbs MRK-16 and MRK-17, produced by immunizing mice with doxorubicin-resistant K-562 human leukemia cells. MRK-16 mAb was also reported to modulate vincristine and actinomycin D transport in resistant cells, and MRK-17 was shown to specifically inhibit growth of resistant cells with these drugs.
Meyers et al., 1987, Cancer Res. 49: 3209 disclose mAbs HYB-241 and HYB-612, which recognize an external epitope of Pgp.
O'Brien et al., 1989, Proc. Amer. Assoc. Cancer Res. 30:Abs 2114 disclose that mAbs HYB-241 and HYB-612 increased the accumulation of vincristine and actinomycin D in tumor cells and increased the cytotoxicity of combinations of these drugs with verapamil.
Tsuruo et al., 1989, Jpn. J. Cancer Res. 80: 627 reported that treatment of athymic mice that had been previously inoculated with drug resistant human ovarian cancer cells with the mAb MRK16 caused regression of established subcutaneous tumors.
Hamada et al., 1990, Cancer Res. 50: 3167 disclosed a recombinant chimeric antibody that combines the variable region of MRK-16 with the F_cportion of a human antibody, and showed this chimeric antibody to be more effective than MRK-16 mAb in increasing cytotoxicity in vitro.
Pearson et al., 1991, J. Natl. Cancer Inst. 88: 1386 disclosed that MRK-16 mAb increased the in vivo toxicity of vincristine to a human MDR colon cancer cell line grown as a xenograft in nude mice. The in vitro potentiation of drug cytotoxicity by MRK-16 mAb was, however, weak relative to known chemical inhibitors of Pgp action, and was apparently limited to only two Pgp substrates (vincristine and actinomycin D), having no effect on cytotoxicity by doxorubicin.
Cinciarelli et al., 1991, Int. J. Cancer 47: 533 disclosed a mouse IgG_2amAb, termed MAb657, having cross reactivity to Pgp-expressing human MDR cells. This mAb was shown to increase the susceptibility of MDR cells to human peripheral blood lymphocyte-mediated cytotoxicity, but was not shown to have an inhibitory effect on the drug efflux activity of Pgp.
Arcesi et al., 1993, Cancer Res. 53:310-317 disclosed mAb 4E3 that binds to extracellular epitopes of Pgp but does not disrupt drug efflux or potentiate MDR drug-induced cytotoxicity.
Mechetner and Roninson, U.S. Pat. No. 5,434,075, issued Jul. 18, 1995, disclose mAb UIC2, having specificity for extracellular Pgp epitopes. This antibody was also shown to effectively inhibit Pgp-mediated drug efflux in MDR cells, and to reverse the MDR phenotype in vitro thereby, for a number of structurally and functional different cytotoxic compounds and all tested chemotherapeutic drugs known to be substrates for Pgp-mediated drug efflux.
There is a need in the art to develop new Pgp inhibitors for preventing or overcoming multidrug resistance in human cancer. In developing new pharmaceuticals, it is essential to determine whether a drug candidate is a Pgp substrate and is effluxed by Pgp expressed in normal or tumor cells. This is important because, on the one hand, such drugs are expected to inhibit Pgp expression in normal cells (in the gastrointestinal tract, excretory organs like kidney, certain hematopoietic cells and the blood-brain and testicular barriers), as well as tumor cells, and to compromise normal function in such organs thereby. On the other hand, tumors derived from such Pgp-expressing tissues are frequently intrinsically multidrug resistant and therefore unaffected by chemotherapeutic intervention. Finally, in all multidrug resistant tumor cells, anti-cancer drugs transported by Pgp decrease intracellular drug concentration, reduce the drug's “therapeutic window” and ultimately reduce the effectiveness of chemotherapeutic treatment. Thus, there is a great need in the art for reagents and assays that permit the rapid, efficient and economical screening and development of effective Pgp inhibitors.
It has been shown that small molecules that are transported by Pgp can be used to competitively inhibit Pgp-mediated efflux of chemotherapeutic drugs that are Pgp substrates. Inhibition of cytotoxic drug efflux from tumor cells in the presence of small molecule Pgp inhibitors has been shown to increase intracellular concentrations of drug and thereby increase its cytotoxic effectiveness. Such small molecules are considered promising drug candidates for selective potentiation of the antitumor effects of several anticancer drugs, including doxorubicin, taxol, vinblastine and VP-16. For example, recent clinical trials of a (relatively) non-toxic cyclosporin analog (PSC833, Novartis Corp.) demonstrated the feasibility of using small molecule Pgp inhibitors for reversing multidrug resistance in patients with hematological malignancies. These results are being actively pursued by a variety of pharmaceutical and biotechnology companies and academic researchers. Thus, development of inexpensive and reliable tests for high throughput screening and identification of new Pgp substrates is important for the development of potent Pgp reversing agents.
At present there are two techniques available for identifying Pgp transport substrates. The first is a dye-efflux assay performed using flow cytometry and is based on competitive inhibition of Pgp-mediated efflux of fluorescent dyes such as rhodamine 123. The second is an in vitro cytotoxicity assay that uses the ability of Pgp substrates to competitively inhibit Pgp-mediated efflux of cytotoxic drugs in Pgp-expressing multidrug resistant cells. In this assay, competitive inhibition of Pgp in cells cultured in the presence of cytotoxic concentrations of Pgp-effluxed cytotoxic drugs results in increased intracellular concentration of such drugs and decreased cell growth. Both assays suffer from the disadvantage that they are laborious and time-consuming and are not suitable for high throughput screening or clinical laboratory testing. In addition, these assays are not specific for Pgp because fluorescent dyes and cytotoxic drugs are also transport substrates for related multidrug resistance transporters (such as MRP; Grant et al., 1994, Cancer Res. 54: 357-361)
There remains a need in the art for a rapid, reliable, efficient and inexpensive method for high throughput screening of compounds for Pgp inhibiting activity in order to develop more effective chemotherapeutic treatment of human cancer patients.

SUMMARY OF THE INVENTION

The invention also provides methods for evaluating novel cytotoxic, chemotherapeutic drugs and Pgp inhibitors. The methods of the invention are based on the development of novel immunological reagents specific for Pgp in a biochemical conformation adopted in the presence of Pgp-mediated transport substrates or ATP depleting agents. The capacity to discriminate between compounds that induce this conformation in Pgp and those that do not provides a way to identify Pgp inhibitors that can be used in high throughput screening assays. These methods can be used as screening assays based on enhanced binding of certain immunological reagents such as UIC2 mAb (A.T.C.C. Accession No. HB 11027) or its derivatives in the presence of Pgp substrates and enable rapid, reliable and cost-effective characterization of potential new Pgp-targeted drugs.
The methods of the invention comprise the steps of contacting a mammalian cell expressing Pgp with an immunological reagent such as UIC2 mAb in the presence and absence of a putative Pgp binding substrate and comparing binding of the immunological reagent in the presence of the test compound with immunological reagent binding in the absence of the test compound. In preferred embodiments, the immunological binding agent is detectably labeled. More preferably, the immunological reagent is detectably labeled with a fluorescent label, and binding affinity is detected by fluorescence activated cell sorting (FACS), immunohistochemistry and similar staining methods. In one aspect of the methods of the invention, Pgp expression levels are determined, providing the capacity to quantitatively compare results between assays. In a second aspect, enhanced binding activity of the immunological reagents provide a way of determining Pgp binding capacity of the test compound. The assays of the invention thus advantageously provide information on both Pgp expression and function simultaneously.
An additional advantage of the methods of the invention is that the use of immunological reagents specific for Pgp reduces the possibility that the assay results contain contributions from related species involved in multidrug resistance, such as MRP.
Specific preferred embodiments of the present invention will become evident from the following more detailed description of certain preferred embodiments and the claims.

DESCRIPTION OF THE DRAWINGS

FIGS. 1A through 1H depict the predicted nucleic acid sequence of human Pgp (Seq. I.D. No. 1), wherein the initiation (ATG) and termination (TGA) codons, as well as codons encoding mutations at amino acid positions 433 and 1076, are underlined.

FIG. 2A illustrates flow cytometric analysis of K562/I-S9 leukemia cells incubated with phycoerythrin (PE)-conjugated mAb in the presence or absence of vinblastine.

FIG. 2B illustrates flow cytometric analysis of K562/I-S9 leukemia cells incubated with PE-conjugated UIC2 mAb in the presence or absence of vinblastine at 4° C.

FIGS. 3A and 3B illustrate flow cytometric analysis of K562/i-S9 leukemia cells incubated with PE-conjugated UIC2 mAb (FIG. 3A) or MRK16 mAb (FIG. 3B) in the presence or absence of different cytotoxic drugs.

FIG. 4 illustrates flow cytometric analysis of K562/i-S9 leukemia cells incubated with PE-conjugated UIC2 mAb in the presence of increasing concentrations of vinblastine (1-625 μM), taxol (0.96-600 μM), verapamil (1.8-1125 μM), colchicine (2-1250 μM), etoposide (1.36-850 μM) and puromycin (1.72-1075 μM).

FIGS. 5A through 5D illustrate flow cytometric analysis of mouse L cell transfectants expressing wildtype (KK-L) double mutant (MM) or single mutant (MK-H or KM-H) human Pgp incubated with PE-conjugated UIC2 (FIGS. 5A and 5C) or MRK16 (FIGS. 5B and 5D).

FIGS. 6A through 6D illustrate flow cytometric analysis of mouse L cell transfectants expressing wildtype (KK-L; FIGS. 6A and 6B), or double mutant (MM; FIGS. 6C and 6D) human Pgp incubated with PE-conjugated UIC2 (FIGS. 6A and 6C) or MRK16 (FIGS. 6B and 6D) in the presence of absence of taxol, vinblastine or etoposide.

FIGS. 7A through 7F illustrate flow cytometric analysis of mouse L cell transfectants expressing wildtype (KK-H) or single mutant (KM-H and MK-H) human Pgp incubated with PE-UIC2 (FIGS. 7A, 7C and 7E) or PE-MRK16 (FIGS. 7B, 7D and 7F) in the presence or absence of vinblastine, taxol or etoposide.

FIGS. 8A through 8E illustrate flow cytometric analysis of mouse L cell transfectants expressing wildtype (KK-L; FIG. 8A or KK-H; FIG. 8B), single mutant (MK-H; FIG. 8C; or KM; FIG. 8D) or double mutant (MM; FIG. 8E) human Pgp incubated with PE-conjugated UIC2 in the presence or absence of vinblastine and the ATP depletion agents oligomycin, azide and cyanide.

FIGS. 9A through 9C illustrate flow cytometric analysis of K562/i-S9 leukemia cells incubated with PE-conjugated UIC2 in the presence or absence of vinblastine and varying concentrations of the ATP depletion agents oligomycin, azide and cyanide.

FIGS. 10A and 10B illustrate flow cytometric analysis of KK-L cells incubated with PE-conjugated UIC2 (FIG. 10A) or MRK16 (FIG. 10B) in the presence or absence of vinblastine and varying concentrations of the ATP depletion agents oligomycin, azide and cyanide.

FIG. 11 illustrates flow cytometric analysis of K562/i-S9 leukemia cells incubated with PE-conjugated UIC2 in the presence or absence of cyclosporine, BSO or SN-38 as described in Example 5.

FIG. 12 illustrates flow cytometric analysis of MCF7-40F P4 breast cancer cells incubated with PE-conjugated UIC2 in the presence or absence of cyclosporine, cisplatin or SN-38 as described in Example 5.

FIG. 13 illustrates flow cytometric analysis of human KB-8-5 tumor cells incubated with PE-conjugated UIC2 in the presence or absence of taxol or taxotere as described in Example 6.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention provides a variety of methods related to P-glycoprotein mediated multidrug resistance in mammalian, most preferably human, cells. For the purposes of the present invention, “multidrug resistance” is defined as cross-resistance to at least the following cytotoxic drugs: vinblastine, vincristine, doxorubicin, colchicine, actinomycinD, etoposide, taxol, puromycin, and gramicidin D; it will be recognized that cross-resistance to other cytotoxic drugs also falls within the meaning of multidrug resistance as it is understood by those with skill in the art. Such drugs are generally referred to herein as MDR drugs.
The methods of the invention ire based in significant part on the discovery by the present inventors that the mAb UIC2, which is capable of inhibiting drug efflux from Pgp-expressing cells, specifically binds to Pgp in a particular biochemical conformation. For the purposes of this invention this biochemical conformation is functionally defined as the conformation adopted by human Pgp in the presence of Pgp substrates or ATP depleting agents, and results in enhanced binding of the mAb UIC2. Also within this definition are certain mutant forms of Pgp having disabling mutations in the nucleotide binding sites, wherein ATPase activity is disabled, as described below, in Loo and Clarke (1995, J. Biol. Chem. 270: 21449-21452) and in Müller et al. (1996, J. Biol. Chem 271: 1877-1883). For the purposes of this invention, exemplary Pgp transport substrates include a variety of lipophilic, cytotoxic natural product drugs used in cancer chemotherapy, including but not limited to Vinca alkaloids, epipodophyllotoxins, anthracyclines, etoposide, colchicine, colcemid and taxol, as well as the antibiotics monensin and actinomycin D and the interleukin cytokines. For the purposes of this invention, the term “ATP-depleting agent” is intended to include, but is not limited to, 2-deoxyglucose, cyanine, oligomycin, valinomycin and azide, as well as salts and derivatives thereof.
The invention provides methods for detecting functional P-glycoprotein expression in a mammalian cell, particularly a malignant mammalian cell and most particularly a multidrug resistant malignant mammalian cell. For the purposes of this invention, the term “functional Pgp expression” is intended to encompass the production of Pgp protein in a cell membrane, most preferably the plasma membrane, wherein the Pgp is capable of transporting an MDR drug across said membrane and against a concentration or solubility gradient. “Functional Pgp expression” is also intended to encompass Pgp protein molecules having an ATPase activity.
In the methods of the invention provided to detect functional Pgp expression in a mammalian cell, the immunological reagent is preferably provided wherein the extent and amount of specific binding of the reagent to Pgp expressed by the mammalian cell is increased in the presence of a Pgp substrate or ATP-depleting agent. For the purposes of this invention, it will be understood that the invention thus provides methods and reagents wherein specific binding of the immunological reagents is enhanced in the presence of a Pgp substrate or ATP-depleting agent, as compared with specific binding of the immunological reagent to the mammalian cell in the absence of a Pgp substrate or ATP-depleting agent. Such enhanced binding is detected using any method known to the skilled artisan, including but not limited to detection of binding of detectably-labeled embodiments of the immunological reagents of the invention, and detection of specific binding of the immunological reagents of the invention using a detectably-labeled immunological reagent that is specific for the immunological reagents of the invention (e.g., in a “sandwich-type” immunoassay). Alternatively, and preferably, the methods of the invention include conventional cell separation methods and techniques, including but not limited to fluorescence activated cell sorting techniques. In other embodiments, the methods of the invention are provided wherein the immunological reagents of the invention are recognized by detectably-labeled second immunological reagents which specifically recognize the immunological reagents of the invention (for example, based on isotypic, allotypic or species-specific antibodies or antisera).
For the purposes of this invention, the term “immunological reagents” is intended to encompass antisera and antibodies, particularly monoclonal antibodies, as well as fragments thereof (including F(ab), F(ab)₂, F(ab)′ and Fv fragments). Also included in the definition of immunological reagent are chimeric antibodies, humanized antibodies, and recombinantly-produced antibodies and fragments thereof. Immunological methods used in conjunction with the reagents of the invention include direct and indirect (for example, sandwich-type) labeling techniques, immunoaffinity columns, immunomagnetic beads, fluorescence activated cell sorting (FACS), enzyme-linked immunosorbent assays (ELISA), and radioimmune assay (RIA). For use in these assays, the Pgp-specific immunological reagents can be labeled, using fluorescence, antigenic, radioisotopic orbiotin labels, among others, or a labeled secondary immunological detection reagent can be used to detect binding of the Pgp-specific immunological reagents (i.e., in secondary antibody (sandwich) assays).
The UIC2 mAb is one example of the immunological reagents of the invention. This mAb is directed to an epitope in an extracellular domain of human Pgp, and was made by immunizing mice with mouse cells that have been made MDR by transfection with an isolated human MDR1-encoding cDNA (see U.S. Ser. No. 07/626,836, incorporated by reference). Briefly, immunogenic cells (preferably transfected syngeneic mouse fibroblasts) were used to immunize BALB/c mice (e.g., transfected BALB/c mouse 3T3 fibroblasts). MDR derivatives of mouse BALB/c 3T3 fibroblasts were generated with human MDR1-encoding DNA, and cells selected and grown in cytotoxic concentrations of an MDR drug. Once produced, MDR fibroblasts were selected in which the transfected MDR1 gene had been amplified, by consecutive steps of selection in progressively higher concentrations of an MDR drug. This produced highly multidrug resistant cells that expressed large amounts of Pgp inserted into the cellular plasma membrane resulting in high levels of MDR (e.g., BALB/c 3T3-1000 cells are resistant to vinblastine at a concentration of 1000 ng/mL).
Such cells were used to immunize syngeneic mice. Appropriate numbers of cells were injected subcutaneously (s.c.) or intraperitoneally (i.p.) by art-recognized immunization protocols (see U.S. Pat. No. 5,434,075, issued Jul. 18, 1995, incorporated in their entirety herein). Typically, 10⁵to 10⁸transfected cells were injected 5 or 6 times at two week intervals, and a final boosting was done with, for example, 10⁶cells subcutaneously and/or intravenously. At an appropriate time after the booster injection, typically 3 to 5 days thereafter, the spleen was harvested from a hyperimmune mouse, and hybridomas generated by standard procedures (see, e.g., Kearney et al., 1979, J. Immunol. 123: 1548) using human myeloma cells, for example, P3-X63-Ag 8.653 (A.T.C.C., Rockville, Md.).
Extracellular fluids from individual hybridoma cultures were screened for specific mAb production by conventional methods, such as by indirect immunofluorescence using non-Pgp expressing control cells (e.g. non-transfected fibroblasts) and human Pgp-expressing (e.g. BALB/c 3T3-1000) cells affixed to glass slides, and FITC-labeled goat anti-mouse polyvalent immunoglobulins (Sigma Chemical Co., St. Louis, Mo.) as the secondary, reporter antibody. The particular screening method used was not critical provided that it was capable of detecting anti-human MDR1 Pgp mAb. It is important, however, that cells are not permeabilized and fixed during screening (i.e., they are living cells), so that only antibodies reactive with extracellular protein domains are detected.
A stable hybridoma producing the UIC2 mAb was established by conventional methods, such as by consecutive rounds of subcloning by, e.g., end-point dilution, and screening the culture medium for monoclonal antibodies. The hybridoma was propagated by, for example, growth in ascites fluid in vivo in syngeneic animals, and the secreted antibody isolated and purified from ascites fluid by affinity chromatography with a Sepharose-Protein A matrix specific for an IgG isotype. It will be understood that other procedures for immunoglobulin purification well known in the art are also useful for producing hybridomas that express Pgp-specific antibodies.
Alternative methods for producing mAbs are known in the art (as described in U.S. Pat. No. 5,434,075, issued Jul. 18, 1995, incorporated in its entirety herein).
mAbs produced by the UIC2 hybridoma, as well as fragments and recombinant derivatives thereof, were characterized as to immunoglobulin isotype, reactivity with different Pgp-expressing cell lines and binding to Pgp in MDR cells using art-recognized techniques (see U.S. Pat. No. 5,434,075, issued Jul. 18, 1995, incorporated by reference). As provided herein, preferred mAbs of the invention specifically bind to Pgp in a biochemical conformation adopted in the presence of Pgp-mediated transport substrates or ATP depleting agents, or in certain Pgp mutants as described herein.
The effect of anti-Pgp mAbs, fragments or recombinant derivatives thereof on Pgp function was assessed by studying the efflux of fluorescent or radioactively labeled drugs from MDR cells in the presence of absence of mAb. The effects of antibody preparations on drug cytotoxicity were assessed by incubating suspensions of MDR and control cells with the antibody preparation, then testing for cell growth inhibition in the absence and presence of an anti-cancer drug such as one of the Vinca alkaloids. Such assays are by definition preferred, as the mAbs of the invention are intentionally provided to be specific for substrate-bound Pgp.
Fragments of the UIC2 mAb that maintain the antigenic specificity of the complete antibody are derived by enzymatic, chemical or genetic engineering techniques (for example, partial digestion with proteolytic enzymes such as papain, pepsin or trypsin; papain digestion produces two Fab fragments and one F_cfragment, while pepsin cleavage releases F(ab)₂(two antigen-binding domains bound together) fragments). mAb fragments lacking the constant (F_c) portion are advantageous over the complete antibody for in vivo applications, as such fragments are likely to possess improved tissue permeability. Furthermore, many cells and tissues in the body express receptors capable of binding to the F_cportion of antibodies, resulting in undesirable non-specific binding of the complete antibody.
The methods of the invention are not intended to be limited in scope to immunological reagents comprising the UIC2 mAb and hybridomas producing this mAb. The invention provides a variety of methods, all related to specific binding of mAbs to Pgp in a biochemical conformation adopted in the presence of Pgp-mediated transport substrates or ATP depleting agents. The UIC2 mAb is provided solely as one illustrative example of an mAb that specifically binds to Pgp and mutants thereof having such a biochemical conformation.
The Examples which follow are illustrative of specific embodiments of the invention, and various uses thereof. They set forth for explanatory purposes only, and are not to be taken as limiting the invention.

Example 1

1. Cell Lines, Monoclonal Antibodies, and Reagents

MRK-16 mAb (IgG.sub.2a) was obtained from Dr. T. Tsuruo, University of Tokyo, Japan. UIC2 was produced from UIC2 and UIC2/A hybridomas as described in U.S. Pat. No. 5,434,075, issued Jul. 18, 1995.
All mAb samples were at least 95% pure according to SDS-PAGE. Concentrations of the mAb were determined by the quantitative mouse Ig radial immunodiffusion kit (ICN, Costa Mesa, Calif.). When necessary, mAb's were further concentrated and dialyzed against phosphate-buffered saline (PBS) or Dulbecco modified Eagle's medium (DMEM). mAbs were conjugated with R-phycoerythrin (PE) or fluorescein isothiocyanate (FITC) at 1:1 (PE) and 1:4 (FITC) mAb:label and purified using standard techniques (Maino et al., 1995, Cytometry 20: 127-133). IgG_2a-PE conjugates were purchased from Becton-Dickinson Immunocytometry Systems (BDIS, San Jose, Calif.) and used as a negative isotype control for nonspecific staining.
The K562/Inf cell line was derived by infection of human K562 leukemia cells with a recombinant retrovirus pLMDR1L6 carrying human MDR1 cDNA (Choi et al., 1991, Proc. Natl. Acad. Sci. USA 88: 7386-7390), and subsequently subcloned without cytotoxic selection (e.g., by FACS sorting based on Pgp-specific immunostaining or Pgp-mediated efflux of fluorescent dyes). Clones expressing relatively high levels of Pgp were selected by repeated selection of Pgp-positive clones by FACS after clonal expansion. Clone K562/I-S9 is one such FACS-selected clone (produced as described in Weisberg et al., 1996, J. Exp. Med. 183: 2699-2704).
LMtk⁻ cells transformed with wildtype and mutant forms of P-glycoprotein were prepared according to Morse (1996, Doctoral Dissertation, Department of Genetics, University of Illinois at Chicago, incorporated by reference herein). MDR1 cDNA-comprising constructs encoding wildtype (KK), single mutant (KM, MK) and double mutant (MM) forms of P-glycoprotein were prepared as described in Morse, wherein the mutant forms have a lysine→to→methionine mutation within either (single mutant) or both (double mutant) of the consensus ATP binding sites in the amino- and carboxyl-terminal halves of P-glycoprotein, introduced at amino acid positions 433 and 1076 by site-directed mutagenesis techniques (see Kramer et al., 1984, Nucleic Acids Res. 12: 9441-9456; Carter et al., 1985, Nucleic Acids Res. 13: 4431-4443). Each of these constructs further comprises the bacterial neomycin-resistance gene (neo), fused to the MDR1 gene via an overlapping translation termination/initiation codon (ATGA). As a consequence, MDR1 and neo are expressed in mammalian cells in a bicistronic messenger RNA. The MDR1-encoding portions of these constructs are shown in Seq. I.D. No. 1. These sequences, cloned into the mammalian expression vector expression vector pUCFVX were introduced into LMtk cells by calcium phosphate coprecipitation or electroporation (see Sambrook et al., ibid.) and transfectants selected in G418 (Grand Island Biological Co. (GIBCO), Long Island, N.Y.)-containing media. Clonal populations of Pgp wildtype or mutant-expressing cells expressing approximately equal amounts of Pgp at the cell surface were selected by FACS using fluorescently labeled mAb MRK16 and were then expanded under G418 selection.
All chemotherapeutic drugs were purchased from Sigma Chemical Co. (St. Louis, Mo.), diluted in water, DMSO or alcohol, aliquoted and stored at +4° C. for 10-14 days or at 20° C. until use.

2. Preparation of Anti-Pgp Monoclonal Antibodies

Monoclonal antibodies specific for human P-glycoprotein were prepared as disclosed in co-owned and co-pending U.S. patent application Ser. No. 07/854,881, filed Mar. 20, 1992, now U.S. Pat. No. 5,434,075, issued Jul. 18, 1995, and Ser. No. 08/032,056, filed Mar. 16, 1993, each of which are incorporated in its entirety herein.
Briefly, mouse fibroblast BALB/c 3T3 cells expressing the MDR1 gene encoding P-glycoprotein (Pgp) were derived by transfecting fibroblasts with isolated human MDR1 cDNA in a eukaryotic expression vector pUCFVXMDR1 (Choi et al., 1988, Cell 53: 519-529), isolating multidrug-resistant cells after cytotoxic selection in 20 ng/mL of vinblastine, and subsequently amplifying the transfected gene by consecutive steps of selection in 250 ng/mL, 500 ng/mL and 1000 ng/mL of vinblastine. The resultant multidrug-resistant fibroblasts were termed BALB/c 3T3-250, BALB/c 3T3-500 and BALB/c 3T3-1000, respectively.
BALB/c mice were immunized with 1−2×10⁷of BALB/c 3T3-1000 cells, injected subcutaneously (s.c.) and/or intraperitoneally (i.p.) six times at two-week intervals. The final immunogenic boost was done with 2×10⁷cells i.p., and 5×10⁶cells administered intravenously (i.v.). Four days after the last administration of fibroblasts, the spleen from one animal was removed, and hybridomas generated by art-recognized techniques using the human myeloma cell line P3-X63-Ag8.653 (A.T.C.C. Accession No. CRL-1580).
Tissue culture supernatant fluids from individual hybridomas were screened for monoclonal antibody (mAb) production by indirect immunofluorescence labeling of live BALB/c 3T3 and BALB/c 3T3-1000 cells attached to glass slides. Fluorescein isothiocyanate (FITC)-labeled goat anti-mouse polyvalent immunoglobulins (obtained from Sigma Chemical Co., St. Louis, Mo.) were used as a secondary antibody reagent at 1:100 dilution. Of 556 tested hybridomas, mAb produced by only two hybridomas reacted with BALB/c 3T3-1000 cells, and of these two only one hybridoma (termed UIC2) produced an antibody reactive with BALB/c 3T3-1000 cells, but not with control BALB/c 3T3 cells.
A stable hybridoma line secreting UIC2 mAb was established by three consecutive rounds of subcloning by end-point dilution and screening of the supernatant fluids, as described in co-owned and co-pending U.S. patent application Ser. No. 07/854,881, filed Mar. 20, 1992, now U.S. Pat. No. 5,434,075, issued Jul. 18, 1995, and Ser. No. 08/032,056, filed Mar. 16, 1993, each of which are incorporated in its entirety herein.
The UIC2 hybridoma was propagated as ascites in syngeneic BALB/c mice, and the immunoglobulin was purified from ascites fluid by Sepharose-Protein A (Bio-Rad, Richmond, Calif.) affinity chromatography. UIC2 mAb, tested by SDS-PAGE, was at least 95% pure IgG. The UIC2 hybridoma is on deposit in the American Type Culture Collection, 10801 University Boulevard, Manassas, Va. U.S.A. 20110-2209 (Accession No. HB11027) and is available to the public.
Application of Ouchterlony and immunoblotting tests using a standard set of anti-mouse Ig antibodies revealed that the UIC2 mAb belongs to the IgG_2asubclass.
UIC2 mAb was shown to induce complement-mediated cytotoxicity using Low-Tox-M rabbit complement (Cedarland Labs, Homby, Ontario) on BALB/c, BALB/c 3T3-1000, CEM/VLB₁₀₀, K562 and K562/Inf cell lines.

3. Fluorescence Activated Cell Sorting/Flow Cytometry Analysis

Cells were trypsinized, when necessary, and washed twice with PBS at room temperature or 4° C. and distributed in 2 mL conical plastic tubes at a concentration of 10⁶cells/tube in 1 mL of pre-warmed (37° C.) Ca⁺⁺, Mg⁺⁺-free PBS and incubated for 37° C. for 10 min. Thereafter, aliquots of 20 μL of drug stock solutions at 1 mg/mL (or at different concentrations, when necessary) were added. The cells were incubated with drugs at 37° C. for 10 min. Aliquots of 50 μL mAb stock solutions (UIC2 conjugated with R-phycoerythrin (UIC2-PE), MRK16-PE, and IgG2a-PE conjugates, or UIC2 conjugates with fluorescein isothiocyanate (UIC2-FITC) and IgG2a-FITC conjugates), prepared at 1:10 dilution, were added and the tubes mixed thoroughly. The amount of mAb added per 10⁶cells/mL was determined by preliminary titration. mAb stock solutions were used at a concentration of 0.08 mg/mL in all experiments with chemotherapeutic drugs. After incubation with mAb for 20-30 minutes, cells were washed twice with ice-cold PBS, transferred into plastic tubes containing 0.5 mL ice-cold PBS and 1 μg/mL propidium iodide, and kept on ice until FACS analysis. For indirect staining experiments, cell samples were washed twice, stirred, incubated with secondary antibody reagents in 100 μL PBS for 20 min. and prepared as above for FACS analysis. For ATP depletion experiments, washed cells were incubated with 20 μL aliquots of stock solutions of azide, oligomycin or cyanide for 15 min. at 37° C. and then immediately treated with chemotherapeutic drugs, antibodies and propidium iodide as described above.
Cells were analyzed by FACSort (BDIS) equipped with an argon laser (Cyonics) tuned to 488 nm, using 4 parameters (forward scatter, side scatter, FL1 for FITC, FL2 for PE conjugates and FL3 for propidium iodide); dead cells were excluded on the basis of forward and side scatter and PI (FL3) staining. The FACS data were analyzed by the Lysis II or CellQuest computer programs.

4. ATP Depletion Experiments

Cells were depleted of intracellular ATP by incubation with oligomycin, azide or cyanide at various concentrations under conditions described in Section 3 above. Intracellular ATP was measured using the Bioluminescent Somatic Cell Assay kit (Sigma, St. Louis, Mo.), whereby the amount of ATP in cell lysates is proportional to light emitted by firefly luciferase. Intracellular ATP was expressed relative to the amount present in cells treated with PBS instead of ATP depleting agents. After incubation of cell lysates with the components of the assay kit, 0.1 mL of the reaction solution was assayed spectrophotometrically over a wavelength range of 390-622 nm using an AutoLumat LB953 Universal Luminometer (EG&G Berthold, Vildbad, Germany). All measurements were performed at 8° C. in 12×75 mm polystyrene cuvettes (Analytical Luminescence Lab, San Diego, Calif.).

Example 2

mAb UIC2 Reactivity is Increased in the Presence of Pgp-Transported Compounds

Flow cytometry was used to analyze the reactivity of phycoerythrin (PE)-conjugated mAbs UIC2 and MRK16 with Pgp-expressing cells in the presence of different drugs. The range of optimal drug concentrations for these experiments (1-5 mg/mL) was determined by a series of preliminary titration experiments.
FIG. 2A illustrates the results obtained with K562/I-S9 leukemia cell line, which was selected to express Pgp by infecting K562 cells with a MDR1-transducing recombinant retrovirus and subsequent flow cytometric selection based on MRK16 antibody staining. Cells were treated in the presence or absence of 25 μM vinblastine and contacted with PE-conjugates mAbs UIC2, MRK16, IgG2a (a negative isotype control) and anti-CD54 (a positive control mAb against a cell surface marker of K562 cells). UIC2 reactivity of this cell line was increased in the presence of the Pgp-transported drug vinblastine, as seen by the rightwards shift in the flow cytometric profile at increasing drug concentrations. This profile shift was not seen with either the positive or negative control mAbs and was not seen with the Pgp-specific mAb MRK16. A similar pattern of mAb binding was observed with FITC-conjugates mAbs and in experiments performed with unlabeled mAbs detected using labeled secondary antibody (sandwich) techniques. In addition, increased UIC2 reactivity was only observed when cells were incubated with drugs and antibody at 37° C., but did not appear when incubations were performed at 4° C. (FIG. 2B), suggesting that enhanced UIC2 binding in the presence of certain Pgp substrates requires the cells to be metabolically active
A variety of MDR drugs and competitive inhibitors of Pgp were tested to determine whether these compounds could induce the FACS profile shift observed with UIC2 binding in the presence of vinblastine. The tested compounds included vinblastine, taxol, actinomycin D, gramicidin D, cyclosporine A, reserpine, 5-fluorouracil and methotrexate. The results of these experiments are shown in FIG. 3A for binding of PE-UIC2 mAb and in FIG. 3B for binding of MRK16 mAb. In these experiments, a rightwards shift in the flow cytometry profile of cells contacted with PE-UIC2 mAb was observed for cells treated with vinblastine, taxol, actinomycin D, gramicidin D cyclosporine A and reserpine. No FACS profile shift was observed in cells treated with 5-fluorouracil or methotrexate, supporting the conclusion that shifting was Pgp specific and was specifically induced with Pgp substrates (since neither 5-fluorouracil or methotrexate is (typically) a Pgp substrate). In contrast, and consistent with the earlier results disclosed above, no change in the flow cytometry profile of cells contacted with MRK16 mAb was observed in cells treated with any of the tested drugs. Stimulation of UIC2 reactivity by these compounds was dose-dependent, for some compounds, while for others no shift was observed at any concentration tested (as illustrated in FIG. 4).
Increased UIC2 reactivity in the presence of Pgp substrates was also observed with other Pgp-expressing cells and cell lines, including PA3 17 cells expressing Pgp via an MDR1-encoding retrovirus (Choi et al., ibid.), NIH 3T3 cells, KB-3-1, VSV1 and GSV1 cells transfected with MDR1 cDNA (Choi et al., 1988, Cell 53: 519-529), Pgp-positive leukemia/lymphoma and tumor samples and normal B- and T-lymphocyte subpopulations and hematopoietic stem cells expressing Pgp (Chaudhary et al., 1992, Blood ibid.; Chaudhary et al., 1992, Cell ibid.). The concentrations of Pgp substrates producing maximal stimulation of UIC2 reactivity differed slightly for different cell lines and appeared to correlate with the levels of Pgp expressed on the corresponding cell lines.
A summary of these results are shown in Table I.

	TABLE I

	UIC2	MRK 16

MDR Drugs
taxol	+	−
vinblastine	+	−
reserpine	+	−
verapamil	+	−
gramicidin	+	−
cyclosporine	+	−
vincristine	+	−
actinomycin D	+	−
colchicine*	−	−
etoposide*	−	−
puromycin	−	−
Non-MDR Drugs
5-fluorouracil	−	−
cisplatin	−	−
carboplatin	−	−
methotrexate	−	−
azidothymidine	−	−
cyclophosphamide	−	−

*weak Pgp substrate

Example 3

Mutations at Pgp Nucleotide-Binding Sites Alter UIC2 Reactivity

The ability of Pgp transport substrates to increase UIC2 reactivity as described in Example 2 suggested that mAb UIC2 reacts more strongly with Pgp having a conformation associated with functioning (i.e., drug-transporting) Pgp. To investigate the relationship between Pgp function and UIC2 reactivity, nucleotide-binding site mutants of Pgp were used. As described in Example 1, Pgp was mutagenized at highly conserved lysine residues (positions 433 and 1076) in the N-terminal and C-terminal nucleotide-binding sites of the human Pgp. These lysine residues were substituted with methionine residues (i.e., lysine-to-methionine (K→M) substitutions), and the resulting proteins were designated KK (wild-type Pgp), MM (double mutant), KM and MK (C-terminal and N-terminal single mutants, respectively). Analysis of immunoprecipitated Pgps showed that nucleotide binding, as measured by specific photolabeling with ³²P-8-azido-ATP, was decreased in the single mutants (KM and MK) and undetectable in MM (as disclosed in Müller et al., 1996, J. Biol. Chem. 271: 1877-1883). In addition, all three mutants (MM, KM and MK) lost detectable ATPase activity (see Müller et al., ibid.). The double mutant, MM, also lost the ability to confer drug resistance to all tested MDR drugs (including vinblastine and vincristine). KM and MK mutant expressing cells, however, showed a 2-3 fold greater resistance to vinblastine than control cells not expressing Pgp, and accumulated 3-4 times more vinblastine than wildtype (KK)-expressing cells with the same level of vinblastine resistance. Vinblastine resistance conferred by KK, KM and MK Pgps was equally sensitive to inhibition with mAb UIC2.
For UIC2 shift experiments, two sets of murine Lmtk⁻ transfectants were used, matched to express very similar levels of the wild-type or mutant human MDR1 Pgps. The first set includes cell lines designated KK-L (wild-type) and MM (double mutant) (FIGS. 5A through 5D and 6A through 6D). The second set, expressing about five times as much Pgp as the first set, includes cell lines KK-H (wild type), KM-H and MK-H (single mutants) (FIGS. 7A through 7F). The relative levels of Pgp expression were established on the basis of indirect immunofluorescence with PE-conjugated MRK16 (see Morse, 1996, ibid.).
FIGS. 5A and 5B show a comparison between flow cytometric analysis of KK-L and MM expressing cells contacted with LTIC2 (FIG. 5A) and MRK16 (FIG. 5B). Similarly, FIGS. 5C and 5D show a comparison between flow cytometric analysis of KK-H, MK-H and KM-H expressing cells contacted with UIC2 (FIG. 5C) and MRK16 (FIG. 5D). The flow cytometric pattern of all of these cells was the same when assayed using the MRK16 mAb (see FIGS. 5B and 5D). In contrast to the results obtained using mAb MRK16, UIC2 mAb showed a strikingly different pattern of reactivity with cell lines transfected with mutant Pgps. UIC2 reacted much more strongly with the MM double mutant than with the wild-type Pgp on KK-L cells (compare in FIG. 5A). Similarly, UIC2 binding in single mutant KM-H was equivalent to wildtype binding (KK-H), while the extent of UIC2 binding to the MK-H single mutant was diminished.
UIC2 mAb binding was compared to MRK16 binding in these cells in the presence or absence of different MDR drugs. These results are shown in FIGS. 6A through 6D. No rightwards shift in the flow cytometric profile was observed in any of the cell lines assayed using MRK16 mAb. In contrast, the wildtype KK-L cell line showed a rightward shift in the profile when cells treated with vinblastine or taxol were assayed, but not when cells treated with etoposide was assayed, consistent with the results disclosed above. The MM double mutant cell line showed no flow cytometric profile shift in the presence of these drugs, but the profile was shifted rightward using UIC2 compared with MRK16 (compare FIGS. 6C and 6D). Vinblastine induced levels of UIC2 mAb reactivity in KK-L cells were roughly equivalent to binding levels seen with MM cells. MM cells showed high levels of UIC2 mAb binding in either the presence or absence of drugs (FIG. 6D); MRK16 binding was unaffected and observed at a level consistent with binding to KK-L cells, confirming our earlier observations on the relative Pgp expression levels of these cell lines.
In contrast with these results, the single mutant MK-H cells showed lower UIC2 reactivity than the wild-type KK-H transfectants, while the reactivity of the other single mutant cell line KM-H, was similar to KK-H (FIGS. 7A through 7F). The KK-H, MK-H and KM-H transfectants were all observed to yield increased UIC2 reactivity by Pgp substrates, with the final levels becoming very similar for all three cell lines (compare FIGS. 7A, 7C and 7E). MRK16 binding levels were approximately the same for all three cell lines in the presence or absence of drug.
These results demonstrated that enhanced UIC2 mAb binding was related to the conformation of Pgp expressed in UIC2-reactive cell lines, and suggested that the MM mutant had adopted a conformation equivalent to the biochemically active conformation presumed to be recognized by UIC2 and which accounted for enhanced UIC2 mAb binding to Pgp in the presence of certain Pgp substrates.

Example 4

Intracellular ATP Depletion Maximized UIC2 Reactivity

The results described in Example 3 above indicated that maximal UIC2 mAb reactivity was associated with the MM mutant, which carries disabling mutations in both nucleotide-binding sites. This result suggested that the biochemical conformation of Pgp that is specifically recognized by UIC2 mAb could reflect a conformation in which Pgp had no bound ATP. This further suggested that intracellular ATP depleting agents would increase UIC2 mAb reactivity for Pgp. Three different agents that induce ATP depletion, sodium cyanide, sodium azide and oligomycin (all of which are specific for mitochondrial enzymes and mechanisms which generate ATP) were used to deplete Pgp-expressing cells of intracellular ATP. All three agents were found to increase UIC2 mAb reactivity to wild-type Pgp in KK-L (FIG. 8A) and K562/I-S9 cells (FIGS. 9A through 9C). The increase in UIC2 reactivity correlated with the extent of intracellular ATP depletion, as measured by the luciferase assay described in Example 1.
The addition of cyanide, azide or oligomycin to the series of LMtk⁻ cells transfected with different Pgp mutants had the same effect on UIC2 mAb reactivity as the addition of Pgp transport substrates (FIGS. 8A through 8E). These agents increased the reactivity of KK-L cells to the level of MM (compare FIGS. 8A and 8B), while having no effect on the MM cell reactivity, and increased the reactivity of KK-H, MK-H and KM-H cell lines to similar final levels. Similar results were obtained in K562/I-S9 cells expressing human Pgp (FIGS. 9A through 9C), and a comparison of UIC2 (FIG. 10A) and MRK16 (FIG. 10B) binding of KK-L cells expressing the wildtype human Pgp is shown in FIGS. 10A and 10B.
Thus, these results support the conclusion that ATP depleting agents have the same effect on UIC2 mAb reactivity as mutagenesis of both nucleotide-binding sites of Pgp.

Example 5

Characterization of Pgp Binding of a Novel Anticancer Compound

A novel anticancer drug was tested using the UIC2 binding assay to determine whether it bound to Pgp.
SN-38 is the active species of CPT-11 (Irinothecan), a newly-developed drug for treating colon cancer. Several clinical trials have demonstrated the efficacy of CPT-11 in colon cancer patients. However, it was unclear if SN-38 is a Pgp-transported substrate. This is of particular importance in colon cancer because all colon cancer tumors express Pgp and, therefore, are intrinsically resistant to Pgp-transported cyctotoxic drugs.
A modification of the UIC2 binding assay was used to determine Pgp substrate specificity of SN-38 in vitro. Flow cytometry was performed on the K562/S9-I cell line, which had been infected with a retrovirus containing human MDR1 cDNA and selected for Pgp expression by flow sorting. Washed K562/S9-I cells were incubated in phosphate-buffered solution (PBS) with 2% FCS for 10 min at 37° C. in the presence of SN-38, cyclosporin-A (a known Pgp substrate as a positive control) or buthionine sulfoximine (BSO, a non-Pgp drug as a negative control for UIC2 shift) at the final concentration of 5 mg/mL. The cells were then incubated with UIC2 labeled with phycoerythrin (UIC2-PE, obtained from Immunotech, Marseille, France), or IgG2a-PE (an mAb isotype control) for 20 minutes at 37° C. in the presence of the above drugs. The treated samples were washed twice with cold PBS, mixed with propodium iodide (to exclude non-viable cells) and analyzed using the BDIS FACSVantage cell sorter. A DMSO control was also used in the experiment to exclude the possibility that this diluent affects UIC2 staining. As shown in FIG. 11, cyclosporine A induced UIC2 shift (increased immunoreactivity, as compared to “no drug” DMSO control), while treatment with BSO, a non-Pgp drug, did not result in UIC2 shift in K562/S9-I cells. Incubation of K562/S9-I cells with UIC2-PE in the presence of SN-38 under physiological conditions induced UIC2 shift, demonstrating that this drug is a Pgp-transported substrate.
This result was confirmed in additional experiments performed using a Pgp-positive breast cancer cell line, MCF7-P4 Pgp-positive breast carcinoma cell line; in these assays, cisplatin was used as a negative control. These results are shown in FIG. 12, wherein binding of labeled UIC2 to these cells was enhanced in the presence of SN-38. There results established that SN-38 is a Pgp-biding drug. These results have been confirmed by conventional in vitro cytotoxicity tests.

Example 6

Characterization of Pgp Binding of a Novel Anticancer Compound

The cytotxic drugs taxol and taxotere are commonly-used chemotherapeutic agents for treating a variety of human cancers, including breast, lung and ovarian tumors, as well as Kaposi's sarcoma. Taxol was initially purified from Pacific yew tree bark, and currently its semi-synthetic form is manufactured from other renewable sources. Taxotere is a semi-synthetic compound derived from the needles of European yew tree. Both drugs are very similar with respect to their chemical structure and mechanism of cytotoxic action on microtubules in tumor cells.
It has been previously shown that both taxol and taxotere can be effluxed by the Pgp pump in human tumors. However, it had also been shown that taxotere had higher anticancer efficacy than taxol in clinical trials on breast and ovarian tumors. One explanation for the clinical data was that this difference in the drugs' activity was caused by their differences in their Pgp substrate specificity. The modification of the UIC2 binding assay described in Example 5 was used to determine relative Pgp transport specificities for taxol and taxotere. In these experiments, a human tumor cell line, KB-8-5, was incubated as described above with PE-labeled UIC2 in the presence of taxol or taxotere at four concentrations: 0.08 mM, 0.31 mM, 1.25 mM, and 5 mM, using DMSO diluent alone as a negative substrate control and mouse IgG2a as a negative mAb control. Flow cytometric analysis was performed as described, and the results are shown in FIG. 13. At low concentrations (0.08 mM and 0.31 mM), taxotere was found to induce a smaller binding affinity shift than taxol at equivalent concentrations (compare the two cytometric analysis curves on the lower left with the two on the lower right in FIG. 13). These results were not as evidence in the analyses performed at higher drug concentrations. This observation indicated that the taxotere is effuxed less efficiently from these cells than taxol, leading to increased intracellular taxotere accumulation and the higher therapeutic efficiency found clinically in cancer patients. This finding was confirmed by conventional in vitro cytotoxicity experiments demonstrating that intracellular accumulation of taxotere is higher than that of taxol
These results demonstrate that the methods of the invention can be used to characterize and compare novel anticancer drugs for Pgp binding efficiency, and that these results can be used to identify anticancer compounds having higher clinical efficacy based on lower affinity for Pgp.
It should be understood that the foregoing disclosure emphasizes certain specific embodiments of the invention and that all modifications or alternatives equivalent thereto are within the spirit and scope of the invention as set forth in the appended claims.

Claims

1. A method for characterizing binding of a test compound to P-glycoprotein expression in a mammalian cell, the method comprising the steps of:

(a) incubating a first mammalian cell expressing P-glycoprotein in the presence of a first P-glycoprotein binding compound having a first binding affinity for P-glycoprotein;

(b) incubating a second mammalian cell expressing P-glycoprotein in the presence of the test compound;

(c) reacting the first mammalian cell and the second mammalian cell with an immunological reagent specific for P-glycoprotein in a biochemical conformation adopted in the presence of a P-glycoprotein substrate; and

(d) comparing binding of the immunological reagent to the first mammalian cell with binding of the immunological reagent to the second mammalian cell.

2. The method of claim 1 wherein the immunological reagent is a monoclonal antibody specific for P-glycoprotein in a biochemical conformation adopted in the presence of an ATP-depleting agent.

3. The monoclonal antibody of claim 2 that is UIC2 (A.T.C.C. Accession No. HB11027).

4. The method of claim 1, wherein binding of the immunological reagent is increased in the presence of the P-glycoprotein substrate.

5. The method of claim 1 wherein the immunological reagent is detectably-labeled.

6. The method of claim 5 wherein the detectable label is a fluorescent label.

7. The method of claim 1 wherein binding of the immunological reagent is increased in the presence of the P-glycoprotein substrate and wherein enhanced binding of the fluorescently-labeled immunological reagent is detected by fluorescence-activated cell sorting.