WO2005080992A2

WO2005080992A2 - P-glycoprotein assay

Info

Publication number: WO2005080992A2
Application number: PCT/GB2005/000614
Authority: WO
Inventors: Jenny Berry
Original assignee: GE Healthcare UK Ltd; Amersham Biosciences UK Ltd
Current assignee: GE Healthcare UK Ltd
Priority date: 2004-02-21
Filing date: 2005-02-21
Publication date: 2005-09-01
Anticipated expiration: 2006-08-21
Also published as: GB0403913D0; WO2005080992A3

Abstract

The invention relates to a radio-ligand method for measuring the binding of a test agent to a P-glycoprotein preparation. The invention is a homogeneous scintillation proximity assay for quantifying the binding of a ligand and a test agent to a P-glycoprotein preparation. The assay is suitable for high-throughput screening to identify novel substrates/ inhibitors of P-glycoprotein.

Description

P-Glycoprotein Assay

Field of Invention

The present invention relates to a homogeneous assay for measuring binding of a substrate to P-glycoprotein. More specifically, the invention provides a scintillation proximity assay for identifying substrates and inhibitors of P-glycoprotein.

Background to the Invention

P-giycoprotein (P-gp), the most extensively studied ATP-binding cassette (ABC) transporter, functions as a biological barrier by extruding toxins and xenobiotics from cells. The protein is the product of the multidrug resistance gene (MDR) and effects the absorption, distribution and excretion of a number of clinically important drugs (Fromm, 2000, Int J Clin Pharmacol Ther 38, 69-74). Both it? vitro and Hi viv^'ύ SiϋύiθS fιav^'6 G6iTiϋπ5ϊfcu6ϋ iflαi r-yιyCϋprGϊ6iπ piSy^'S S oiQTmiCcin. TGiθ in drug absorption and disposition. For example, P-gp limits the intestinal absorption of digoxin, talinolol and cyclosporine after oral dosing, limits the central nervous system penetration of human immunodeficiency virus protease inhibitors, and excretes paclitaxel (taxol) into the intestine (Lown et al., 1997, Clin Pharmacol Ther 62, 248-260; Sparreboom et al., 1997 Proc NatlAcad Sci USA 94, 2031-2035; Kim et al., 1998 J Clin Invest 101, 289-294; Schwarz et al. 2000 Int J Clin Pharmacol Ther 38, 161 -167). Apical expression of P-gp in tissues such as liver, kidney and intestine results in reduced drug absorption from the gastrointestinal tract and enhanced drug elimination into bile and urine. Moreover, expression of this glycoprotein in the endothelial cells of the blood-brain barrier prevents entry of certain drugs into the central nervous system.

A number of groups have developed techniques to monitor functional P-gp expression in cells in order to characterise multidrug resistance, for example in cancer patients undergoing chemotherapy. Thus WO 98/21325, describes the production and use of antibodies that recognise the conformational change that occurs when P-gp binds to its substrates and can be used to determine functional P-gp expression in cells, particularly multidrug resistant cells.

Modification of P-gp function is an important underlying mechanism of drug interactions in humans, both inhibition and induction of the protein having been reported as the cause of drug-drug interactions. Compounds which act as P-gp substrates potentially have an increased risk of pharmacokinetic problems in man. There is therefore considerable interest in the pharmaceutical field in determining, at an early stage, whether new drug candidates are potential P-gp substrates as this may significantly reduce their biological efficacy. Due to its importance in pharmacokinetics, P-gp transport screening has now become an integral part of the drug discovery process. However, existing technology for quantifying P-gp transport is generally low through-put, labour intensive and expensive, characteristics which are far from optimal for meeting the demands of high-throughput screening of the pharmaceutical industry.

The human colon carcinoma cell line (Caco-2) transport assay (Anderle et al., 1998, J Pharm Sci 87, 757-762; Gao et al., 2001 , Pharm Res 18, 171-176) which is the original screen for P-gp, is still widely used in industry because it provides a reliable and functional measure of how P-gp affects the permeability of a drug. However, the assay is limited by cost and throughput, culturing on filter supports and the need for LC-MS analysis of the transported compounds being required which is both time consuming and expensive.

Some P-gp assays which are suitable for high throughput screening are already available, based either on the ability of a drug to activate ATPase in P-gp - over- expressing membranes (e.g. Gentest, Woburn, MA, USA) or cell-based assays that measure accumulation of a fluorescent dye (e.g. Cayman). However, although these assays are used by some pharmaceutical companies, issues relating to reliability and interpretation mean that they have not gained universal acceptance. In a recent study, Polli et al. (J Pharmacol Exp Ther, 2001 , 299, 620-628) compared three existing assays based upon monolayer efflux (Polli et al., 1999, Pharm Res 16, 1206-1212) , ATPase activity (Schmid et al., 1999, Biochem Pharmacol 58, 1447-1456) and calcein-AM fluorescence (Liminga et al., 1994, Exp Cell Res 212, 291-296) and concluded that the monolayer efflux assay gave the most reliable measure of efflux.

Doppenschmitt ef al. (Pharm Res 1998, 15, 1001-1006; J Pharmacol Exp Ther, 1999, 288, 348-35) have reported the use of a radio-ligand binding assay to characterise P-gp binding properties. Binding studies were performed on human Caco-2 cells to demonstrate that non-labelled substrates could be used to displace [³H]verapamil or [³H]vinblastine. However, while it was suggested that this assay could form the basis for a high-throughput screening tool, no studies were performed to translate these basic experiments into a robust screen.

Furthermore, the non-homogeneous experimental procedures described in these papers were time consuming and labour intensive, requiring filtration to quantify binding.

Scintillation Proximity Assay (SPA) is a homogeneous radioisotopic assay, as described in US 4568649. In a SPA, there is a solid phase (e.g., a bead or the bottom of a tissue culture/microplate well) that is or contains within it a substance capable of fluorescing when stimulated by a -particle that has been emitted by a weakly emitting ^-isotope such as °H or ^{l o}l. The fluorescent substance is known as a scintillant or phosphor. The surface of the solid phase is such that it has an affinity for the particular analyte the assay is designed to detect. This can be done by modifying the surface of the solid so that it is coated with a receptor where the analyte is a substance that has an affinity for the receptor (e.g., a ligand of the receptor).

Some phosphors have been designed to emit radiation of long wavelength (400- 900nm) which can be detected by, for example, a charge coupled device (CCD) to allow simultaneous imaging of all wells in a multiwell plate (e.g. GB2366370) and/or to overcome problems of colour quench and detection sensitivity (e.g. US6,524,786).

SPA can be designed to detect a binding event (e.g. binding of a receptor to a substrate as described in, for example, US 6,524,786) or to follow enzyme activity (e.g. to measure substrate production as described in WO 99/60155). A significant advantage of SPA over conventional binding assays, such as receptor binding assays described in US 2003/0059857, is that being homogeneous in nature it is suitable for high throughput screening.

The present invention addresses the need for a reliable, homogeneous and robust high-throughput screen for characterising P-gp substrates and inhibitors. The invention provides a homogeneous scintillation proximity assay (SPA) which can be used in high-throughput screening to detect substrates and inhibitors of P-gp binding. The assay further provides quantitative data on the affinity of substrate/inhibitor binding which is of use in predicting whether the substrate/inhibitor is likely to interact with P-gp at the therapeutic doses expected in man.

Summary of the Invention

According to the first aspect of the present invention there is provided a method for measuring the binding of a radio-labelled ligand to a P-glycoprotein preparation comprising the steps of i) mixing the P-glycoprotein preparation comprising a capture moiety with the radio-labelled ligand to produce a P- glycoprotein -labelled ligand complex; and ii) capturing the P-glycoprotein-labelled ligand complex on a solid phase comprising a phosphor or a scintillant and a capture reagent that specifically binds to the capture moiety whereby the phosphor or the scintillant generates a measurable signal

wherein the measurable signal is indicative of the amount of radio-labelled ligand bound to P-glycoprotein.

The terms 'ligand' means any chemical entity, such as a protein or synthetic organic compound, which binds to P-glycoprotein.

Suitably, the value of the first measurable signal is recorded. Preferably the recorded value is stored on an electronic or optical database. In a second aspect of the present invention, there is provided a method for measuring the binding of a test agent to a P-glycoprotein preparation comprising the steps of i) measuring the binding of a radio-labelled ligand to a P- glycoprotein preparation as hereinbefore described in the first aspect of the invention to produce a measurable signal; ii) mixing the solid phase with the test agent under conditions whereby the radiolabelled ligand bound to the P-glycoprotein is displaced by the test agent; and iii) detecting a second measurable signal generated by the phosphor or scintillant wherein any difference between the measurable signal of i) and the second measurable signal of iii) is indicative of the amount of test agent bound to P- glycoprotein.

In a third aspect of the present invention, there is provided a method for measuring the binding of a test agent to a P-glycoprotein preparation comprising the steps of i) mixing the P-glycoprotein preparation comprising a capture moiety with a mixture of a radio-labelled ligand and the test agent to produce a P-glycoprotein-labelled ligand complex and/or P-glycoprotein-test agent complex; ii) capturing the complex on a solid phase comprising a phosphor or a scintillant and a capture reagent that specifically binds the capture moiety whereby the phosphor or scintillant generates a measurable signal; and iii) comparing the measurable signal with the recorded value of binding of radio-labelled ligand as hereinbefore described for the first aspect of the invention wherein any difference between the measurable signal and the recorded value is indicative of the amount of test agent bound to P-glycoprotein.

The term 'test agent' should be construed as any compound which may act as a substrate or inhibitor of P-glycoprotein. A 'substrate' is bound to P-gp and is transported/excreted from the cell by it. An 'inhibitor' is bound to P-gp and inhibits binding and/or transportation/excretion of a P-gp substrate or ligand.

Suitably, the test agent is a chemical compound. For example, the test agent may be a naturally occurring compound, such as a peptide or a nucleic acid.

Typically, the test agent will be a drug candidate. In general, drug candidates are low molecular weight organic compounds which have been specifically synthesised or optimised to evaluate their biological efficacy in a drug discovery/evaluation programme.

Preferably, the P-glycoprotein preparation is a membrane preparation.

Alternatively, the P-glycoprotein preparation is a whole cell preparation.

Suitably, the test agent inhibits P-glycoprotein binding to the ligand and may bind to the protein at a different site to that of the substrate.

Preferably, the solid phase is a bead suitable for use in a scintillation proximity assay (SPA). Current SPA technology involves the use of scintillant beads made from either cerium-doped yttrium silicate (Y₂Si0₅:Ce) (hereinafter referred to simply as yttrium silicate) or polyvinyltoluene (PVT) containing an organic scintillant such as diphenyl anthracene. Assays are carried out in aqueous buffers using radioisotopes such as ³H, ¹²⁵l, ¹⁴C, ³⁵S or ³³P, that emit low-energy radiation, the energy of which is easily dissipated in an aqueous environment. For example, the electrons emitted by ³H have an average energy of only 6 keV and have a very short path length (~1 μm) in water. If a molecule labelled with one of these isotopes is bound to the bead surface, either directly or via interaction with another molecule previously coupled to the bead, the emitted radiation will activate the scintillant and produce light. The amount of light produced, which is proportional to the amount of labelled molecules bound to the beads, can be measured conveniently with a liquid scintillation (LS) counter. If the labelled molecule is not attached to the bead surface, its radiation energy is absorbed by the surrounding aqueous solvent before it reaches the bead, and no light is produced. Thus, bound ligands give a scintillation signal, but free ligands do not, and the need for a time-consuming separation step, characteristic of conventional radioligand binding assays, is eliminated. The manipulations required in the assays are reduced to a few simple pipetting steps leading to better precision and reproducibility.

PCT W0 91/08489 describes a support body for use in scintillation proximity radioimmunoassay, the support body being constructed of a scintillating material, having coupled to its surface a multiplicity of ligands such a antigens, antibodies, etc. capable of selectively binding a reactant of interest. Preferably the support bodies consist of yttrium silicate activated with an inorganic cerium salt such as the oxide, carbonate, or chloride.

WO 94/26413 concerns the study of cellular and biochemical processes in living cells or in components of cells. Specifically described are devices and methods for the study of cellular and biochemical processes, using the scintillation proximity principle.

The simplicity of the scintillation proximity format allows almost complete automation of assays using robotic sample processors and microplate scintillation counters. Consequently, SPA technology is capable of high throughput, which is particularly valuable in the case of drug- or sample-screening assays.

US Patent Number 6524786 describes a method of using a charge coupled device to image wells in a scintillation proximity assay. The use of phosphors that emit radiation of longer wavelength (480 - 900 nm) than conventional phosphors overcomes problems of colour quench and sensitivity encountered with the use of conventional phosphors which emit blue light.

Preferably, the bead comprises yttrium silicate, yttrium oxide, polystyrene, or polyvinyltoluene containing an organic scintillant. Alternatively, the solid phase is a coating on the base and/or sides of a vessel. Preferably, the vessel is a microplate.

In a preferred embodiment, the capture moiety and the capture reagent are members of a specific binding pair. For example, the capture moiety/capture reagent are selected from the group consisting of N-acetyl glucosamine/Wheat Germ Agglutinin (WGA), biotin/steptavidin, biotin/avidin, biotin/neutravidin, biotin/captavidin, epitope/antibody, GST/glutathione, His tag/Nickel, FLAG/M1 , maltose binding protein/maltose, chitin binding protein/chitin, calmodulin binding protein/calmodulin, and LumioTM reagents/LumioTM recognition sequence.

Suitably, the radioactive label is selected from the group consisting of ³H, ¹⁴C, ³³P, ³⁵S and ¹²⁵l. Preferably, the ligand is selected from the group consisting of Taxol, verapamil, vinblastine and vincristine. Most preferably, the radio-labelled ligand is [³H]TaxoI.

Suitably, the preparation is obtainable from a cell over-expressing P-glycoprotein. Preferably, the cell is a mammalian cell or an insect cell.

Suitably, the insect cell is selected from the group consisting of Sf9, Sf21 and High Five™ cells.

Suitably, the mammalian cell is selected from the group consisting of human colon carcinoma cell line Caco-2, human ovarian cancer cell line, MDR1-MDCKII,

CHO and HEK cells. Preferably, the cell is a human ovarian cancer cell line.

In a fourth aspect of the present invention, there is provided a kit suitable for measuring the binding of test agent to P-glycoprotein comprising i) a solid phase suitable for use in scintillation proximity assay comprising a capture reagent; and ii) a preparation of P-glycoprotein comprising a capture moiety. Preferably, the solid phase is a bead, the P-glycoprotein preparation is a membrane preparation. Most preferably, the capture reagent is Wheat Germ Agglutinin (WGA) and the capture moiety is N-acetyl glucosamine.

Brief Description of Drawings

Figure 1 is a diagrammatic presentation illustrating the competitive displacement of [³H]Taxol by a P-gp substrate or inhibitor.

Figure 2 graphically depicts the competitive displacement of labelled taxol by unlabelled erythromycin.

Figure 3 shows the competitive displacement of labelled taxol by unlabelled taxol.

Figure 4 depicts the competitive displacement of labelled taxol by unlabelled quinidine.

Figure 5 graphically illustrates the competitive effect of a range of compounds on binding of labelled taxol.

Specific Description and Examples

Selection and Culturing of Human Ovarian Cancer Cell Lines

The selection and culturing of the human ovarian cancer cell line which over- expresses P-gp is as described in Ding et al., 2001 , Brit J Cancer 85 (8), 1175- 1184.

Preparation of cell membranes

Confluent cell monolayers were washed with cold PBS and cells harvested by scraping into PBS using a cell scraper. The resultant cell suspension was centrifuged for 10min at 600xg and the pellet re-suspended in 10mM Tris buffer, pH 7.4 and incubated on ice for 10 minutes to swell cells. The cells were homogenised using a Dounce homogeniser, centrifuged for 10min at 600xg and the pellet discarded. The supernatant was centrifuged for 30min at 30000xg and the supernatant discarded. The pellet was re-suspended in buffer and protein content determined.

f³HlTaxol [³H]taxol was purchased from Moravek Biochemicals, USA (Moravek Biochemicals, 577 Mercury Lane, Brea, CA 92821 USA. Catalogue code MT1646).

Assay Procedure Test compound (10μl) in 100% (v/v) dimethyl sulphoxide (DMSO) was pipetted into 'test' wells/ tubes. 100μM Taxol (10μl) in 100% (v/v) DMSO was pipetted into 'non-specific binding' wells/ tubes. DMSO (10μl) was pipetted into 'total binding' wells/ tubes. Assay buffer (40μl) was pipetted into all wells/ tubes. 20nM (37kBq/ml) [³H]taxol (50μl) was pipetted into all wells/ tubes. 400μg/ml ovarian cancer cell membranes (50μl) were pipetted into all wells/ tubes. 20mg/ml YSi- WGA bead (50μl) (Amersham Biosciences, UK, Catalogue ref. RPNQ0011 ) was pipetted into all wells/ tubes. Assays were incubated at room temperature for 16 hnπr.ς nrinr tn nnuntinπ n.ςinn a MiπrπRota™ tPfirkin FlmΩr I ife Sniffnπesl scintillation counter.

Results

Figure 2 illustrates the competitive displacement of labelled taxol by unlabelled erythromycin. Erythromycin, a known P-gp substrate, was shown to competitively displace [³H]taxol from membranes expressing P-gp, an IC₅₀ value of 14μM being obtained.

Figure 3 shows the competitive displacement of labelled taxol by unlabelled taxol. Taxol, a known P-gp substrate, was shown to competitively displace [³H]taxol from membranes expressing P-gp, an IC₅₀ value of 0.2μM being obtained.

Figure 4 depicts the competitive displacement of labelled taxol by unlabelled quinidine. [³H]taxol was competitively displaced by quinidine, a known P-gp substrate, and an IC50 value of 50nM was obtained. Figure 5 shows the competitive effect of a range of compounds on binding of labelled taxol. A panel of compounds containing known P-gp substrates/inhibitors (quinidine, erythromycin, nicardipine, desipramine, ketoconazole, loperamide, digoxin and propranolol) and compounds known not to be P-gp substrates or inhibitors (mannitol, ranitidine and lidocaine) were assayed at a final concentration of 10μM. The known P-gp substrates/inhibitors (i.e. quinidine, erythromycin, nicardipine, desipramine, ketoconazole and loperamide) were shown to displace [³H]taxol when added at a final concentration of 10μM. In contrast, the non-P-gp substrates (i.e. mannitol, ranitidine and lidocaine) did not displace [³H]taxol when added at a final concentration of 10μM.

Claims

1. A method for measuring the binding of a radio-labelled ligand to a P- glycoprotein preparation comprising the steps of i) mixing said P-glycoprotein preparation comprising a capture moiety with said radio-labelled ligand to produce a P- glycoprotein -labelled ligand complex; and ii) capturing said P-glycoprotein-labelled ligand complex on a solid phase comprising a phosphor or a scintillant and a capture reagent that specifically binds said capture moiety whereby said phosphor or said scintillant generates a measurable signal wherein said measurable signal is indicative of the amount of radio- labelled ligand bound to P-glycoprotein.

2. The method according to claim 1 , wherein the value of the first measurable signal is recorded.

3. The method according to claim 2, wherein said recorded value is stored on an electronic or optical database.

4. A method for measuring the binding of a test agent to a P-glycoprotein preparation comprising the steps of i) measuring the binding of a radio-labelled ligand to a P- glycoprotein preparation according to the method of claim 1 to produce a measurable signal; ii) mixing the solid phase with said test agent under conditions whereby the radiolabelled ligand bound to the P-glycoprotein is displaced by the test agent; and iii) detecting a second measurable signal generated by the phosphor or scintillant wherein any difference between said the measurable signal of i) and the second measurable signal of iii) is indicative of the amount of test agent bound to P-glycoprotein.

5. A method for measuring the binding of a test agent to a P-glycoprotein preparation comprising the steps of i) mixing said P-glycoprotein preparation comprising a capture moiety with a mixture of a radio-labelled ligand and said test agent to produce a P-glycoprotein-labelled ligand complex and/or P-glycoprotein-test agent complex; ii) capturing said complex on a solid phase comprising a phosphor or a scintillant and a capture reagent that specifically binds said capture moiety whereby said phosphor or scintillant generates a measurable signal; and iii) comparing said measurable signal with the recorded value of binding of radio-labelled ligand according to claim 2 or 3 wherein any difference between said measurable signal and said recorded value is indicative of the amount of test agent bound to P-glycoprotein.

6. The method according to any preceding claim wherein the P-glycoprotein preparation is a membrane preparation.

7. The method of any of claims 1 to 5, wherein the P-glycoprotein preparation is a whole cell preparation.

8. The method of any of claims 4 to 7, wherein the test agent inhibits P- glycoprotein binding to the ligand.

9. The method of any of claims 4 to 8, wherein the test agent is a chemical compound.

10. The method of any of claims 4 to 9, wherein the test agent is a drug candidate.

11. The method according to any preceding claim, wherein the solid phase is a bead suitable for use in a scintillation proximity assay.

12. The method according to claim 11 , wherein said bead comprises yttrium silicate, yttrium oxide, polystyrene or polyvinyltoluene containing an organic scintillant.

13. The method according to any of claims 1 to 11 , wherein the solid phase is a coating on the base and/or sides of a vessel.

14. The method according to claim 13, wherein said vessel is a microplate.

15. The method according to any preceding claim wherein the capture moiety and the capture reagent are members of a specific binding pair.

16. The method of claim 15, wherein the capture moiety/capture reagent are selected from the group consisting of N-acetyl glucosamine/Wheat Germ Agglutinin (WGA), biotin/steptavidin, biotin/avidin, biotin/neutravidin, biotin/captavidin, epitope/antibody, GST/glutathione, His tag/Nickel, FLAG/M1 , maltose binding protein/maltose, chitin binding protein/chitin, calmodulin binding protein/calmodulin, and LumioTM reagents/LumioTM recognition sequence.

17. The method according to any preceding claim wherein the capture moiety is N-acetyl glucosamine and the capture reagent is Wheat Germ Agglutinin (WGA).

18. The method according to any preceding claim wherein the radioactive label is selected from the group consisting of ³H, ¹⁴C, ³³P, ³⁵S and ¹²⁵l.

19. The method according to any preceding claim wherein the ligand is selected from the group consisting of Taxol, verapamil, vinblastine and vincristine.

20. The method according to any preceding claim wherein the radio-labelled ligand is [³H]Taxol.

21. The method according to any preceding claim, wherein said preparation is obtainable from a cell over-expressing P-glycoprotein.

22. The method according to 21 , wherein said cell is a mammalian cell or an insect cell.

23. The method according to claim 22, wherein said insect cell is selected from the group consisting of Sf9, Sf21 and High Five™ cells.

24. The method according to 22, wherein said mammalian cell is selected from the group consisting of human colon carcinoma cell line Caco-2, human ovarian cancer cell line, MDR1-MDCKII, CHO and HEK cells.

25. The method according to any of claims 22 or 24, wherein said cell is a human ovarian cancer cell line.

26. A kit suitable for measuring the binding of test agent to P-glycoprotein comprising i) a solid phase suitable for use in scintillation proximity assay comprising a capture reagent; and ii) a preparation of P-glycoprotein comprising a capture moiety.

27. The kit according to claim 26, wherein said solid phase is a bead and said P-glycoprotein preparation is a membrane preparation.

8. The kit according to claim 26, wherein said capture reagent is Wheat Germ Agglutinin (WGA) and said capture moiety is N-acetyl glucosamine.