JP2005510691A6 - Assay - Google Patents

Abstract

The present invention rectifies inward the affinity of compounds in the “ether-a-go-go” (ERG) potassium (K ⁺ ) channel, in particular the human ERG (hERG) potassium channel. Relates to assays defined using potassium channel (IKR) labeled blockers. This assay is useful for identifying compounds that may adversely affect cardiac repolarization in humans, particularly prolonging the ECG QT interval.

Description

The present invention relates to the affinity of a compound in an “ether-a-go-go” (ERG) potassium (K ⁺ ) channel, in particular a human ERG (hERG) potassium channel, and a rapidly active delayed rectifier potassium. It relates to an assay defined with a rapid delayed rectifying potassium channel (IKR) using a labeled blocker, such as [ ³ H] -dofetilide or [ ³ H] -MK-499. This assay is useful for identifying compounds that can cause detrimental effects on cardiac repolarization in humans, particularly prolong the electrocardiogram QT interval and thus cause torsades de pointes.

  In recent years, the development of some compounds proposed for therapeutic use has been abandoned at the late drug development stage due to the detection of harmful effects on cardiac repolarization in humans. The effects of these drugs have been evaluated on the electrocardiogram (ECG) QT interval. The QT interval is a part of the ECG that represents the time from the start of ventricular depolarization to the end of ventricular repolarization. Since the QT interval can be affected by increasing the heart rate and shortening it as the heart rate increases, the QT is often “corrected” by the heart rate to make the QTc interval. In rare cases, administration of drug molecules may lead to an extension of the human ECG QT interval. The ECG of these patients is very similar to that of a person with a genetic disorder known as long QT syndrome. In these cases, drug-induced ventricular fibrillation can result in sudden death (Morganroth J et al. (1993) Am J Cardiol. 72, 26B-31B; De Ponti F. et al. (2000) Eur J. Clin. Pharmacol. 56. 1-18). Numerous drug molecules, including E-4031, cisapride, and terfenadine are well known to prolong the QT interval of the human electrocardiogram (Fuliki A, et al. (1994), Cardiovascular Pharmacol. 23: 374- 378; Van Haarst AD et al., (1998) Clin Pharmacol. Ther. 64: 542-546 ,; Honig PK et al. (1993) JAMA 269; 1513-1518).

  Launching new drugs without detection of possible cardiotoxic side effects can have dangerous consequences and can trigger fatal heart rhythm abnormalities in patients. Delayed detection of QT prolongation induced by a compound of pharmacological interest can hinder drug discovery and program development, and as a result can have a severe impact on the outcome of the program. It is therefore desirable to test for potential cardiotoxic side effects of compounds early in drug development.

According to the present invention there is provided an assay comprising or consisting of the following steps:
a) Assay buffer for ERG-expressing cells, or membranes derived from ERG-expressing cells, or membranes derived from ERG-expressing tissue, in the presence or absence of various amounts of test compound or mixture of test compounds, together with a labeled IKR blocking agent Incubating in;
b) determination of specifically bound labeled IKR blocker;
c) Calculation of inhibition of binding of labeled IKR blocker by test compound or mixture of test compounds.

This assay is useful as a predictive indicator at the preclinical stage to identify compounds that may prolong the QT interval in humans. This assay is a competitive measure that measures the ability of a test compound or mixture of compounds to displace and replace a labeled IKR blocker from the ERG K ⁺ channel (ether-a-go-go K ⁺ channel, referred to herein as ERG). Binding assay. The assay can be performed on a high throughput test system. Combining structure-activity relationships (SAR) to facilitate ligand binding assays using labeled IKR blockers to design new drugs with or without reduced affinity for ERG, particularly human ERG (hERG) Can be used for

The assay buffer used is particularly important because it optimizes the binding of the IKR blocker or test compound (s) to the ERG. It has been discovered that an optimal assay can be achieved using a Tris-based buffer containing potassium (K ⁺ ) ions (at room temperature, pH 7.2 to 7.6, preferably pH 7.4). Potassium ions in the assay buffer can be provided, for example, as potassium chloride (KCl). The concentration of potassium ions in the assay buffer determines the predicted value of the assay. In order for an assay performed in an assay buffer containing 7.5 to 12.5 mM KCl, preferably 8.5 to 11.5 mM KCl, most preferably 10 mM KCl to provide an IC ₂₀ value that predicts the onset of QT prolongation It is particularly useful.

The assay buffer of the present invention is preferably Tris. Contains or consists of Cl and KCl. If desired, MgCl ₂ may be included in the assay buffer.
Assay buffer Tris. The concentration of Cl is preferably 30 mM to 100 mM Tris. Cl, more preferably 30 mM to 70 mM Tris. Cl, even more preferably 40 mM to 60 mM Tris. Cl, more preferably 45 mM to 55 mM Tris. Cl, most preferably 50 mM Tris. Cl.

  The concentration of KCl in the assay buffer is preferably 5 to 20 mM KCl, more preferably 6 to 15 mM KCl, even more preferably 7.5 to 12.5 mM KCl, even more preferably 8.5 to 11.5 mM KCl, most preferably Is 10 mM KCl.

  In a particularly preferred embodiment, the assay buffer is 30 to 100 mM Tris. Cl and 5 to 20 mM KCl, preferably 30 to 70 mM or 30 to 100 mM Tris. Cl and 6 to 15 mM KCl, even more preferably 40 to 60 mM Tris. Cl and 7.5 to 12.5 mM KCl, more preferably 45 to 55 mM Tris. Contains or consists of Cl and 8.5 to 11.5 mM KCl.

Assay buffer was 50 mM Tris. It is particularly preferred that it comprises or consists of Cl and 10 mM KCl.
If in assay buffer containing MgCl _2, the concentration of MgCl ₂ is 2.0 mM MgCl ₂ preferably from 0.6 mM, more preferably 1.6 mM MgCl ₂ from 0.6 mM, even more preferably from 0.8 mM 1. 4 mM MgCl ₂ , more preferably 0.9 mM to 1.3 mM MgCl ₂ , still more preferably 1.0 mM to 1.2 mM MgCl ₂ , most preferably 1.0 mM or 1.2 mM MgCl ₂ .

The assay buffer used in the preferred embodiment is 30 to 100 mM Tris. Cl, 5 to 20 mM KCl, and 0.6 to 2.0 mM MgCl ₂ , preferably 30 to 100 mM or 30 to 70 mM Tris. Cl, 6 from 15 mM KCl, and 1.6 mM MgCl ₂ 0.6, even more preferably 60 mM Tris 40. Cl, 7.5 to 12.5 mM KCl, and 0.8 mM to 1.4 mM MgCl ₂ , more preferably 45 to 55 mM Tris. Cl, comprising or consisting of 8.5 to 11.5 mM KCl and 0.9 to 1.3 mM MgCl ₂ or 1.0 to 1.2 mM MgCl ₂ .

Assay buffer was 50 mM Tris. Cl, 10 mM KCl and 1.0 mM MgCl ₂ ; or 50 mM Tris. Cl may comprise or consist of 10 mM KCl and 1.2 mM MgCl ₂ .

The assay buffer is preferably between pH 7.2 and 7.6 at room temperature; it is particularly preferred that the assay buffer is pH 7.4 at room temperature.
The ERG gene (cDNA) can be derived from a vertebrate or invertebrate source; for vertebrates, the ERG gene is derived from a mammalian source (eg, human, ape, cow, pig, dog, rabbit, guinea pig, rat, or Mouse) or invertebrate sources, such as insect sources (eg Drosophila). Prokaryotic homologues for mammalian ERG may be used. The ERG gene is preferably mammalian ERG, in particular human ERG (hERG) or canine ERG (cERG).

  ERG gene can be expressed in mammalian cell lines such as HEK-293 (human embryonic kidney) cells, CHO (Chinese hamster ovary) cells; CHL (Chinese hamster lung) cells, CSO (monkey) cells; or insect cell lines such as It can be expressed in SF9. A baculovirus vector system can be used for expression of ERG in a compatible insect cell line. Alternatively, ERG can be expressed in yeast cells or bacterial cells. The ERG gene is preferably hERG or cERG and is preferably expressed in any of HEK-293 cells, CHO cells, or CHL cells.

The assay can be performed using whole ERG expressing cells, or membrane preparations from ERG expressing cells, or membrane preparations from ERG expressing tissues.
Dofetilide is an IKR blocker (a selective inhibitor of the fast current component of the delayed rectifier potassium channel) that prolongs the duration of the action potential and the effective refractory period depending on the concentration. Clinical trials have shown that dofetilide is effective in treating patients with atrial arrhythmias as well as ventricular arrhythmias. Dofetilide has the following formula I:

Dofetilide is claimed in European patent EP 0245997 and its preparation is described.
MK-499 (Merck) is a methylsulfonamide antiarrhythmic agent that acts as an IKR blocker. MK-499 has the formula II shown below.

The IKR blocker used in this assay is labeled with a detectable label, such as a radiolabeled tag or a fluorescently labeled tag. In a preferred embodiment of the invention, the labeled IKR blocker used in the assay is a labeled dofetilide, preferably a radiolabeled dofetilide, most preferably tritiated dofetilide ([ ³ H] -dofetilide). In another embodiment of the invention, the labeled IKR blocker used in this assay is labeled MK-499, preferably radiolabeled MK-499, most preferably tritiated MK-499 ([ ³ H] -MK- 499).

Preferred assay configurations include filter binding techniques, by which a labeled IKR blocking agent such as labeled dofetilide or labeled MK-499; preferably radiolabeled dofetilide or radiolabeled MK-499; most preferably [ ³ H]- The dofetilide or [ ³ H] -MK-499 bound and unbound are separated by filtration. This assay is performed using a scintillation proximity assay (SPA) using a radiolabeled IKR blocker such as radiolabeled dofetilide or radiolabeled MK-499, preferably [ ³ H] -dofetilide or [ ³ H] -MK-499. ) Can be done using technology.

In the case of a filter binding technique, ERG-expressing cells, or membranes derived from ERG-expressing cells, or membranes derived from ERG-expressing tissue, are labeled with a labeled IKR blocking agent such as [ ³ H] -dofetilide or [ ³ H] in assay buffer. Incubate with MK-499 in the presence (test) or absence (control) of test compound or mixture of test compounds. Incubation is preferably performed at room temperature for 60 to 120 minutes, preferably 90 minutes. Nonspecific binding is determined in the presence of an unlabeled IKR blocking agent such as 10 μM dofetilide or 10 μM MK-499. Bound labeled IKR blocker is separated from unbound IKR blocker by filtering through a filter mat or on a multiwell filter plate. The filter mat or filter plate is washed to remove unbound labeled IKR blocker, and the bound labeled IKR blocker is, for example, a tritiated IKR blocker, such as [ ³ H] -dofetilide or [ ³ H] -MK- 499 is quantified by scintillation spectroscopy using a suitable counter for radioactivity.

In the case of the scintillation proximity assay® (SPA) system (Amersham Biosciences), beads are used to bind to ERG expressing cells, or membranes derived from ERG expressing cells, or membranes derived from ERG expressing tissues. Various types of beads are suitable for use in the SPA assay according to the present invention, including PVT wheat germ agglutinin, yttrium oxide polylysine beads, or yttrium silicate beads (YSi) (Amersham Biosciences) such as YSi polylysine or YSi wheat. Contains germ agglutinin. The optimal bead type for use in the SPA assay of the invention depends on the cell or cell membrane used; the binding of the cell to the bead or membrane to the bead is evaluated to determine the optimal for the cell or cell membrane to be used. The type of bead may be identified. A bead combined with an ERG material (whole cell, cell membrane preparation, or tissue membrane preparation) is combined with a labeled IKR blocker, such as [ ³ H] -dofetilide or [ ³ H] -MK-499, or a test compound or test compound Incubate in assay buffer in the presence (test) or absence (control). The ability of a test compound or mixture of test compounds to displace a bound radiolabeled IKR blocker is determined by detecting luminescence using, for example, a standard counter that can be used with SPA technology.

The assay may also include one or more of the following steps: calculation of concentration (IC ₂₀ ) of test compound (s) that inhibits dofetilide binding by 20%, test compound that inhibits dofetilide binding by 50% (IC ₂₀ ). Calculation of concentration (IC ₅₀ ), calculation of compound affinity as Ki, or calculation of compound affinity as pKi.

IC ₂₀ values obtained from competitive replacement with IKR blocker binding, eg, [ ³ H] -dofetilide binding, are comparable to the concentration of free drug involved in QT prolongation in humans. Thus, this assay can be used to predict the concentration of a compound that can cause adverse cardiac side effects.

To assess whether a compound or a mixture of compounds may prolong the QT interval of a human electrocardiogram, the following steps are performed:
a) Perform the assay according to the present invention.
b) Obtain an IC ₂₀ value; this value indicates the actual or estimated concentration of free drug at which QT prolongation is likely to occur in humans.
c) IC ₂₀ values are compared in vivo to the concentration of free drug required for the desired therapeutic effect of the compound or mixture of compounds.

A compound or mixture of compounds if the concentration of the free drug required for the desired therapeutic effect of the compound or mixture of compounds is within the range of 10 to 30 times the IC ₂₀ value of the compound or mixture of compounds in the assay May show an extension of the QT interval in humans.

  The assay of the present invention is a method that more accurately predicts the effect of QT prolongation by a drug molecule in vivo than current assays such as the HERG patch clamp method.

Example
Example 1: Preparation of membranes from HEK-293 cells expressing human ERG or canine ERG Adherent cell line HEK-293 expressing human ERG (Zhou, Z et al (1998) Biophys. J. 74, 230-241 ) Was provided by Dr. Craig January, University of Wisconsin, USA; this cell line was designated the “January” cell line. An alternative adherent cell line HEK-293 (293S-HERG clone 15) designated cell line 15 was generated by the method described in Zhou, Z et al (1998). The full length cDNA of human ERG was inserted downstream of the CMV promoter of pcDNA3.1 (Invitrogen) and this vector also has an SV40 promoter that controls the expression of the neomycin resistance gene. The construct was transfected into human embryonic kidney cells 293S (HEK-293). Stable transformants were selected using G418 (Gibco). Cell line 15 has slightly lower hERG expression than cell line January, but improved its growth characteristics.

  Cell line 15 (293S-HERG (clone 15)) was deposited on ECACC (CAMR Salisbury, Wiltshire, SP4 OJG, UK) on June 26, 2002, under deposit accession number 02062678, according to the provisions of Budapest Treaty 1977. Deposited.

Adherent cells expressing human ERG HEK-293 supplemented with 10% fetal bovine serum (PAA Laboratories), 2 mM L-glutamine (Sigma), 1 mM sodium pyruvate (Sigma), 0.4 mg / ml G418 (Life technologies) In addition, 1 × non-essential amino acid (Life technologies) was added to MEM Earles medium (Life technologies) and cultured. The cells were cultured at 37 ° C. in a T225 cm ³ flask in a humid environment of 5% CO ₂ . After reaching 80% confluence, the cells were split 1: 3 to 1: 5 with cell dissociation solution (Sigma, cat no: C5914 in 2001) and then roller bottles filled with 850 cm ² of CO ₂ gas (Corning cat no: 430849 in 2001) in the absence of G418.

  To prepare the membrane, the cells were peeled off from the roller bottles and collected and resuspended in PBS (Life Technologies, cat no: 14190-094 in 2001). All cells were pelleted, washed twice with PBS, aliquoted on dry ice, and stored at -80 ° C until needed.

  The HEK-293 cell line expressing canine ERG was generated by transient transfection of HEK cells. A sequence completely encoding cERG cDNA (Zehlein et al (2001) Pflugers Archiv. European Journal of Physioligy. 442 (2): 188-191) is a pBluescript® vector (Stratagene) from Professor Zehelein University of Heidelberg, Germany. Received the offer inside. In the pBluescript construct, cERG cDNA was incorporated adjacent to multiple BamHI sites. Initial experiments showed low insertion efficiency for direct insertion of the cERG BamHI fragment into the desired vector, pcDNA3.1. In order to overcome this point, an indirect cloning method using the cloning vector pSP73 (Promega) was devised. In order to digest the cERG / pBluescript construct and the pSP73 vector with BamHI and reduce interference due to the presence of the pBluescript BamHI fragment in the ligation reaction, the cERG / pBluescript BamHI digested material was also digested with ScaI to cleave the pBluescript, It was confirmed that the cERG BamHI fragment was more effectively separated on an agarose gel. The restriction digest mixture was subjected to agarose gel electrophoresis and the band containing cERG BamHI and pSP73 BamHI fragments was visualized by ethidium bromide staining and UV irradiation. The cERG and pSP73 bands were excised and eluted from the gel using the QIAgen MinElute Gel extraction kit according to product instructions. To prevent re-ligation of multiple BamHI ends of pSP73 DNA during the ligation reaction, the plasmid DNA fragment was CIP treated according to standard protocols. The cERG BamHI fragment was ligated to the pSP73 BamHI fragment by standard ligation protocols. After the reaction, the ligation mixture was transformed into cJM109 competent E. coli cells using standard transformation protocols. Transformants were selected by plating on LB agar (Millers) containing ampicillin (50 μg / ml) and incubated overnight at 37 ° C. A small amount of cERG / pSP73 DNA was prepared using the QIAgen Miniprep kit according to the product instructions using the transformed cells cultured overnight. The obtained DNA was cleaved with a restriction enzyme and subjected to agarose gel electrophoresis to identify a positive clone.

  The cERG cDNA is excised from cERG / pSP73 as a Xhol (5 ′) EcoRI (3 ′) fragment, and this fragment is a Xhol / EcoRI fragment in the form of a reverse polylinker of pcDNA3.1, ie pcDNA3.1 (−) Xhol. Linked to / EcoRI. The reverse polylinker form was used in this case because the selected cERG / pSP73 clone contained the reverse orientation of cERG. After ligation into pcDNA3.1 (−), the 5 ′ end of cERG was located adjacent to the enhancer-promoter sequence from human cytomegalovirus (CMV). The ligation mixture is transformed into cJM109 competent E. coli cells using standard transformation protocols, and transformants are selected by plating on LB agar (Millers) containing ampicillin (50 μg / ml); Overnight incubation at 37 ° C. with necessary controls. Colonies randomly picked from the cERG / pcDNA3.1 (−) plate were inoculated into 5 ml of LB medium containing ampicillin (50 μg / ml) and incubated overnight at 37 ° C. and 200 rpm. These overnight cultures were then used to make small amounts of DNA using the QIAgen Miniprep kit. The resulting DNA was double digested with Xhol and EcoRI and analyzed on a 1% agarose gel. Combined with the fact that DNA from only a few clones was analyzed by a small preparation method and the low insertion efficiency of the ligation reaction, no positive cERG / pcDNA3.1 (−) clone was identified. Therefore, a larger number of colonies were screened for positive clones using the colony PCR method.

  The colony PCR protocol allowed rapid detection of cERG / pcDNA3.1 (−) clones. Three primers were designed and created for use in the PCR protocol.

Primer 1: 'CERG01' (SEQ ID NO: 1) that hybridizes with cERG at nucleotide positions 601-620 of the coding sequence:
5'-ACCACATCCACCAGGCACAG-3 '
Primer 2: 'NHE PCDN3' (SEQ ID NO: 2) that hybridizes with pcDNA3.1 (-) at nucleotide positions 886-910 (contained in the multiple cloning site adjacent to the Nhel cloning site):
5'-CCCAAGCTGGCTAGCGTTTAAACGG-3 '
Primer 3: 'T7SP73' used as a control and used for colonies known to produce cERG / pSP73 (SEQ ID NO: 3). This primer hybridized with pSP73 at nucleotide positions 98-121 contained in the T7 polymerase promoter sequence:
5'-TAATACGACTCACTATAGGGAGA-3 '
95 cJM109 colonies were picked from LB agar transformation plates and transferred to sterilized 96 well deep well plates containing 1 ml / well of LB broth supplemented with ampicillin (50 μg / ml). As a control, colonies known to contain the cERG / pSP73 plasmid were transferred to the last 96th well containing LB-amp broth. The plate was covered and incubated at 37 ° C. and 200 rpm overnight. A 70 μl aliquot of each mini-culture was transferred to a 96-well PCR plate (0.5 ml / well), placed in a Beckman Allegra 6R centrifuge and centrifuged at 2800 rpm for 10 minutes at room temperature. The supernatant was discarded and the plate was drained for 3 minutes. A PCR reaction mixture was prepared and added to the PCR plate containing the bacterial pellet as follows:

  The bacterial pellet was resuspended in the PCR reaction. The PCR reaction was performed as follows, as specified in the product protocol of the Taqman Gold PCR kit (Applied Biosystems, 1999 edition):

  The PCR products in each well were then separated by electrophoresis on a 1.5% agarose gel using a 100 bp DNA ladder marker with 1 × TAE buffer at 100 V for 25 minutes and visualized with UV light. Clones that were presumed to be positive were identified and samples from these PCR reaction mixtures were run on a second 1.5% agarose gel for 1 hour at 100 V to determine the size of the PCR products.

  Small cultures that received the amplified PCR product were seeded from each original 96-well deep well plate into sterile tubes containing 5 ml of LB broth containing 50 μg / ml ampicillin and incubated at 37 ° C. overnight at 200 rpm. Next, a small amount of DNA was prepared using the overnight culture using the QIAgen Miniprep kit. The resulting DNA was double digested with Xhol and EcoRI and examined for the presence of cERG / pcDNA3.1 (-). Restriction enzyme digests were run on a 1% agarose gel with 1 kb DNA ladder marker run at 100 V for 1 hour (20 μl of sample loaded with 2 μl of gel loading solution). Further, restriction enzyme digestion analysis was performed, and it was confirmed that the plasmid purified from the transformed cells was indeed cERG / pcDNA3.1 (-).

HEK-293 cells that were not transfected were 50 ml of Minimum Essential Medium supplemented with 10% (v / v) fetal calf serum (FCS), 2 mM L-glutamine, 1 mM sodium pyruvate, and 1 mM non-essential amino acids ( MEM) and maintained periodically. Cells were seeded in 225 cm ² ventilated cap flasks and maintained in a humid environment with 5% CO ₂ . HEK-293 cells used in this study had passage numbers between 39-48. Cells were typically passaged from 80-90% confluency flasks at a ratio of 1: 3 every 3 days; washed twice with 10 ml PBS and dissociated from the flask with cell dissociation solution. After that, fresh medium was added.

HEK-293, in which the cERG / pcDNA3.1 (−) construct was grown to 80-95% saturation density in a 225 cm ² ventilable flask, was transfected by the following method. CERG / pcDNA3.1 (−) DNA (94 μg) and Lipofectamine 2000 (Gibco BRL) (94 μg) without endotoxin are added to 2.25 ml of OPTIMEM-I medium (Gibco BRL) in a 10 ml sterile centrifuge tube. After 5 minutes incubation at room temperature, mixed. Thereafter, a mixture of Lipofectamine 2000 / DNA / OPTIME-I was incubated at room temperature for 20 minutes, and further 10.5 ml of OPTIMEM-I was added. HEK-239 cells were washed with 10 ml of PBS, Lipofectamine2000 / DNA / OPTIME-I mixture was added, and incubated at 37 ° C. for 3.5 hours in a humid environment containing 5% CO ₂ . After incubation, 50 ml MEM (10% (v / v) FCS, 2 mM L-glutamine, 1 mM sodium pyruvate, and 1 mM non-essential amino acids supplemented) was added. HEK-293 cells were incubated at 37 ° C. for 24 hours. Transfected cells were washed 24 hours later with PBS, scraped in 10 ml PBS and collected by centrifugation at 1000 rpm for 5 minutes at room temperature. The resulting pellet of HEK-293 cells transfected with cERG / pcDNA3.1 (−) was stored at −80 ° C. until needed.

Preparation of membranes from HEK-293 cells expressing human ERG and canine ERG Cell membrane fractions were prepared from frozen aliquots of cells. All work was performed at 4 ° C unless otherwise stated. Frozen aliquots of cells are thawed at room temperature and resuspended in assay buffer (eg 50 mM Tris.Cl, 10 mM KCl, 1 to 1.2 mM MgCl ₂ , pH 7.4 or 50 mM Tris.Cl, 10 mM KCl, pH 7.4). did. The cells were then disrupted with an Omni LabTek homogenizer at 20,000 rpm for 30 seconds. The homogenate was centrifuged at 48.000 × g for 20 minutes (4 ° C., Sorvall RC5B centrifuge), and the supernatant was removed. The resulting pellet was resuspended in assay buffer and homogenized for 10 seconds as described above. The pellet was collected by centrifugation and the final pellet was resuspended in assay buffer. Protein content was determined using a Coomassie Blue based protein assay kit. Aliquots were stored at −80 ° C. until needed; when stored under these conditions, the binding activity of the cell membrane fraction proved to be stable for at least 4 months.

Example 2: Filter binding assay with [ < ³ > H] -dofetilide
[ ³ H] -Dofetilide (80-83 Ci / mmol) was synthesized by catalytic tritiation (eg custom service provided by Amersham Life Science). However, known to those skilled in the art other detectable label, such as a fluorescent label tag, other radiolabeled tags, antibodies, etc., it can also be used in place of ^{3 H.}

  On the day of the assay, test compounds were dissolved in 50% DMSO or 100% DMSO at a concentration of 1 mM and then diluted in assay buffer to the desired concentration (eg, up to 100 μM or to the limit of compound solubility). The final concentration of DMSO in the assay incubation is preferably 1.0 to 1.5% or less for optimal assay conditions.

Incubation is 50 μg / ml membrane homogenate in assay buffer (50 mM Tris.Cl, 10 mM KCl, 1.0 mM to 1.2 mM MgCl ₂ , pH 7.4), [ ³ H] -Dofetilide, unless otherwise stated (4 to 7 nM) and a test compound or a mixture of test compounds or a control carrier. The filtration assay was incubated at room temperature for 90 minutes. Nonspecific binding was determined in the presence of 10 μM dofetilide and was usually less than 15% of total binding. The bound ligand is rapidly passed through a GF / B glass fiber filter mat using, for example, a Brandel cell harvester or on a GF / B Unifilter 96 well filter plate (Packard) using a Packard Filtermate 96 harvester. The free ligand was separated from the free ligand by filtration. The filter mat and filter plate were pre-soaked in 5% PEI (w / v) for 60 minutes to collect the cells, and then washed 3 times with 1 ml of ice-cold assay buffer. Unifilter plates were air dried at a minimum of 37 ° C. for 1.5 hours and Microscint-0 (Packard) was added. Bound [ ³ H] -dofetilide can be liquidized using a suitable counter, such as a Packard TopCount Scintillation Counter (NXT Counter) or Wallac Counter (Trilux) for unifilter plates, or a Wallac Big Spot Counter for filter mats. Determined by scintillation spectroscopy.

In each experiment, three assays were performed uniformly and the data averaged. Specific binding was analyzed by fitting a non-linear regression model using GraphPad Prism software (GraphPad, San Diego). IC ₅₀ values were calculated by fitting to a 4-parameter logistic model using PRISM and converted to Ki values using the Cheng and Prusoff equations; IC ₂₀ values were extrapolated from the graph.

Example 3: Scintillation Proximity Assay The scintillation proximity assay (SPA) is a 50 mM Tris. This was performed in assay buffer consisting of Cl, 10 mM KCl, 1.0 mM to 1.2 mM MgCl ₂ , pH 7.4 or using assay buffer consisting of 50 mM Tris base, 10 mM KCl, pH 7.4. The binding of the beads to the membrane was evaluated to determine the optimal bead type for the cell line used. January YSi wheat germ agglutinin beads were used for cell membranes derived from HEK-293 hERG expressing cell lines; YSi polylysine beads were used in studies using membranes derived from Cell Line 15 (HEK-293 hERG expressing cell lines). Conditions were optimized for bead and cell membrane homogenate concentrations and the pharmacological properties of ERG were determined. Incubation (total volume 200 μl per well for 96 well plates, total volume 60 μl per well for 384 well plates) contained 25 μg of cell membrane homogenate per mg of beads. The membrane homogenate was pre-coupled with a suspension of YSi wheat germ agglutinin beads or YSi polylysine beads at 4 ° C. on a roller shaker for about 2 hours. For competitive binding assays, membrane homogenate bead suspensions are 5 nM [in the absence and presence of competitors, ie test compounds or test compound mixtures, in 96-well or 384-well plates with clear white bottoms. Incubated with ³ H] -dofetilide. The plate was incubated at room temperature and shaken for about 1 hour. After allowing the beads to stand for a minimum of 30 minutes, the remaining radioactivity was counted in a TopCount NXT scintillation counter. Non-specific binding, ie background count, was determined by adding 10 μM dofetilide. Background counts were usually less than 15% of total binding. To examine saturation, specific binding of [ ³ H] -dofetilide was determined over a range of concentrations (5 to 500 nM) in the absence or presence of cold (ie unlabeled) 10 μM dofetilide.

Example 4: Assay optimization
To optimize the specific binding of dofetilide of Hepes or Tris in binding a) dofetilide into homogenate of cell membrane containing the effect ERG of buffers to the base, ^{[3 H]} - interaction with cell membrane preparation dofetilide Hepes-based buffer (25 mM Hepes, 135 mM NaCl, 5 mM KCl, 1 mM MgSO ₄ , 50 mM CaCl ₂ , pH 7.4) and Tris-based buffer (50 mM Tris.Cl, 10 mM KCl, 1 mM or 1.m. It was examined in the presence of ₂ mM MgCl ₂ ). Comparison of specific binding in these buffers revealed that the ratio of specific binding was similar in both Tris and Hepes basic buffers. However, as shown in Table 1, the specific binding count was twice as high in the presence of the Tris-based buffer as compared to that detected with the Hepes-based buffer.

Total binding and non-specific binding data are calculated from the mean ± 14 individual wells per buffer divided over two assays, performed at a protein concentration of 75 μg / ml and an average [ ³ H] -dofetilide concentration of 6.7 nM. Represents standard error. Incubation was for 60 minutes at room temperature. ccpm = corrected count per minute.

The assay buffer used in Examples 1 to 8 was a Tris-based incubation buffer (50 mM Tris.Cl, 10 mM KCl, 1 mM MgCl ₂ ) so that the maximum specific binding window could be achieved. In addition, experiments were performed to optimize cell membrane protein and bead concentrations for filter and SPA binding assays.

b) Changes over time were examined to determine the optimal incubation time for saturated binding activity. Incubation times were similar for both filter binding and SPA assays. The filter binding assay reached equilibrium in 90 minutes, and SPA required 60 minutes. [ ³ H] -Dofetilide binding to ERG is saturating in both filter binding and scintillation approximation assays, K _D 5.08 ± 1.0 nM, and 96 and 384 format scintillation approximation in filter binding assays each in _{K D} values the assay was 8.9 ± 0.6 nM and 9.1 ± 1.8 nM (Fig. 1a-c, Figure 1a shows the results of filter binding assay, Figure 1b is a 96-well format SPA results and FIG. 1c shows SPA results in 384 well format). Non-linear curve fitting of this data indicated that the bond was one site. Filter binding resulted in B _max 7.4 ± 0.7 pmol / mg protein for [ ³ H] -dofetilide (FIG. 1). Since the scintillation approximation assay cannot accurately determine the dpm (decay per minute) value, the B _max for SPA is not estimated.

c) Comparison of SPA and filter binding techniques Comparison of SPA and filter binding techniques revealed that the results were in very good agreement. Affinity values show a very good correlation between the two assay types, and the order of compound affinity strength is also shown in FIG. 2 (correlation plot comparing pKi values obtained from filter binding and SPA binding assays). Match as shown.

A series of compounds including hERG blocking agent d) to prolong the QT interval in studies human competitive binding are known, ^{[3 H]} - were examined to replace competitive with dofetilide. E4031, dofetilide, terfenadine, and cisapride were shown to completely inhibit specific binding within the range of affinity calculations summarized in Table 2.

Data were expressed as pKi values (negative logarithm of the molar concentration of competitive ligand replacing 50% of 5 nM [ ³ H] -dofetilide binding). Data are average values of at least n = 3 experiments.

Example 5: Prediction of QT interval prolongation effect of compounds in humans Figure 3 shows a series of compounds including E-4031 (Figure 3a), dofetilide (Figure 3b), terfenadine (Figure 3c), and cisapride (Figure 3d) As shown, IC ₂₀ values obtained from competitive displacement with [ ³ H] -dofetilide binding using the assay of the present invention are comparable to the concentration of free drug involved in QT prolongation in humans. To do. For each compound, inhibition of dofetilide binding in the binding assay (filter binding technique) and inhibition of the same binding in the hERG patch clamp assay were compared to the concentration of free drug involved in QT interval extension in humans (Fuliki A, et al. al. (1994) Cardiovascular Pharmacol, 23: 374-378; Van Haarst AD et al. (1998) Clin Pharmacol. Ther. 64: 542-546; Honig PK, et al. (1993) JAMA 269: 1513-1518) Compared with.

  The ERG patch clamp assay measures the current through the ERG channel and indicates the number of ion channels present in one cell. However, the phenomenon of state-dependent blockade observed in patch clamp studies demonstrated in vivo by a number of known hERG blockers that could prolong the QT interval (Walker, BD et al (1999) British J Pharmacol 128, 444-450), the ligand binding assay provides a better prediction of the in vivo QT prolonging effect of the drug than the hERG patch clamp technique (FIG. 3d).

To assess whether a compound or a mixture of compounds may prolong the QT interval of a human electrocardiogram, the following steps are performed:
a) In accordance with the present invention, a binding assay is performed, for example as described in Example 2 or Example 3, to test the affinity of the compound or mixture of compounds for ERG, preferably hERG or cERG;
b) Obtain an IC ₂₀ as described, for example, at the end of Example 2; IC ₂₀ is the actual or estimated concentration of free drug that causes QT prolongation in humans;
c) Compare the IC ₂₀ value with the concentration of free drug required for the desired therapeutic effect of the compound in vivo.

If the concentration of free drug required for the desired therapeutic effect of the compound is within the range of 10 to 30 times the IC ₂₀ of the compound of the assay of the invention, the compound is responsible for QT interval prolongation in humans. Is likely to be.

Example 6 Comparison of Dofetilide Binding Assays Performed on HEK-293 Cells Transfected with cERG or hERG The dofetilide binding assay was performed as described in Example 2, with HEK- transfected with either human ERG or canine ERG. This was performed using 293 cells. The results are shown in FIG. 4, which shows that the IC _{50 of} dofetilide is similar in canine ERG and human ERG, 13.9 nM and 15.6 nM, respectively. The IC ₂₀ values for dofetilide were 1.92 nM and 2.15 nM for canine ERG and human ERG, respectively.

Example 7 Comparison of Terfenadine Competition Assay Using HEK-293 Cells Transfected with cERG or hERG A dofetilide binding assay was performed using terfenadine as a test compound. Transiently transfected cERG HEK-293 cell membranes (200 μg / well) or stable hERG HEK-293 cell membranes (100 μg / well) with 90 different concentrations of terfenadine and 5 nM [ ³ H] -dofetilide Incubated for minutes at room temperature. Total binding and non-specific binding were determined by incubation with 10% DMSO and 10 μM unlabeled dofetilide for a total assay volume of 200 μl. Membranes were collected by filtration through a Packard Unifilter cell harvester and radioactivity (cpm) was measured. Two saturation experiments were performed for each sample of cell membrane expressing cERG and hERG. Each experiment was performed in triplicate. FIG. 5 shows the average values of the experiments for each cell type (cERG or hERG transfection), and the IC _{50 of} terfenadine is similar for cERG and hERG, being 77.2 nM and 88.9 nM, respectively. Is shown. The IC ₂₀ values for terfenadine were 10.7 nM and 12.3 nM for canine ERG and human ERG, respectively.

Example 8 Comparison of E4031 Competition Assay in HEK-293 Cells Transfected with cERG or hERG A dofetilide binding assay was performed using E4031 as the test compound. Transiently transfected cERG HEK-293 cell membranes (200 μg / well) or stable hERG HEK-293 cell membranes (100 μg / well) with 90 different concentrations of E4031 and 5 nM [ ³ H] -dofetilide 90 Incubated for minutes at room temperature. Total binding and non-specific binding were determined by incubation with 10% DMSO and 10 μM unlabeled dofetilide for a total assay volume of 200 μl. Membranes were collected by filtration through a Packard Unifilter cell harvester and radioactivity (cpm) was measured. Two saturation experiments were performed for each sample of cell membrane expressing cERG and hERG. Each experiment was performed in triplicate. FIG. 6 shows the mean of the experiment for each cell membrane type (cERG or hERG transfection), with E4031 IC ₅₀ similar for cERG and hERG, 27.3 nM and 35.4 nM, respectively. Is shown. The IC ₂₀ of E4031 was 3.8 nM and 4.9 nM for canine ERG and human ERG, respectively.

When IC ₅₀ (or IC ₂₀ ) values were compared with the tested compounds, they were found to be very similar for cERG and hERG. This suggests that either cERG or hERG can be used in the assay of the present invention to predict the onset of QT prolongation in humans.

Example 9: To further optimize the assay for the specific binding of dofetilide for homogenates of cell membranes containing the study hERG the optimization of additional assays, ^{[3 H]} - interaction with dofetilide and cell membrane preparation, the Tris containing either KCl or MgCl ₂ to use a buffer which is based, was examined in the format of SPA assay. The SPA assay was performed using 50 mM Tris. Cl, 10 mM KCl, pH 7.4 or 50 mM Tris. Performed according to Example 3 in Cl, 1 mM MgCl ₂ , pH 7.4. The assay was performed using dofetilide or terodiline as the test compound. Comparison of the specific binding detected under the conditions of these buffers indicated that the assay buffer used was 50 mM Tris. In the case of Cl, 1 mM MgCl ₂ , pH 7.4, it was revealed that no specific binding was observed; Observed in assay buffer consisting of Cl, 10 mM KCl, pH 7.4. As an assay buffer, 50 mM Tris. For assays performed with Cl, 10 mM KCl, pH 7.4, IC ₅₀ and IC ₂₀ values were obtained for each test compound. The average IC ₅₀ value of dofetilide was 8.69 ± 0.45 nM, and the average IC ₅₀ value of terodiline was 1.87 ± 0.00 μM. The average IC ₂₀ value of dofetilide was 1.2 nM, and the average IC ₂₀ value of terodiline was 0.248 μM.

FIG. 1: Representative saturation curve data for [ ³ H] -dofetilide binding to HERG in (a) filter binding method, (b) SPA 96-well format, and (c) SPA 384-well format. FIG. 2: Correlation plot comparing pK _i values obtained from filter binding and SPA binding assays: (a) correlation between 96 well hERG [ ³ H] dofetilide SPA assay and radioligand binding assay, (b) hERG [ Correlation between 96 and 384 wells of the ³ H] dofetilide SPA assay. FIG. 3: Inhibition of [ ³ H] -dofetilide binding to hERG for (a) E-4031, (b) dofetilide, (c) terfenadine, and (d) cisapride, hERG patch clamp method, and QT in humans Comparison of free drug concentrations known to induce interval prolongation. FIG. 3: Inhibition of [ ³ H] -dofetilide binding to hERG for (a) E-4031, (b) dofetilide, (c) terfenadine, and (d) cisapride, hERG patch clamp method, and QT in humans Comparison of free drug concentrations known to induce interval prolongation. FIG. 4: Comparison of dofetilide IC ₅₀ in dofetilide binding assays performed on cell membranes derived from HEK-293 cells transfected with human ERG (hERG (▲)) or canine ERG (cERG (■)). FIG. 5: Comparison of terfenadine IC ₅₀ in a dofetilide binding assay on HEK-293 cell membranes transfected with cERG or hERG. FIG. 6: Comparison of IC ₅₀ of E4031 in dofetilide binding assay on HEK-293 cell membranes transfected with cERG or hERG. FIG. 7: Assay buffer 50 mM Tris. Curve showing the effect at mean concentration (n = 2) for (a) dofetilide and (b) terodiline in a tritiated dofetilide SPA assay using Cl, 10 mM KCl, pH 7.4.

[Sequence Listing]

Claims (17)

The following steps:
a) Incubating ERG-expressing cells, or membranes derived from ERG-expressing cells, or membranes derived from ERG-expressing tissue, with a labeled IKR blocker in the presence or absence of a test compound or test compound mixture in assay buffer about;
b) determination of specifically bound labeled IKR blocker;
c) calculation of inhibition of binding of the labeled IKR blocker by the test compound or mixture of test compounds,
An assay comprising or consisting of: The assay according to claim 1, wherein the assay buffer is a Tris-based buffer containing KCl. Assay buffer is 30-100 mM Tris. 3. An assay according to claim ₂ , comprising or consisting of Cl, 5 to 20 mM KCl, and optionally 0.6 to 2.0 mM MgCl2. Assay buffer is 30 to 70 mM Tris. 3. An assay according to claim ₂ , comprising or consisting of Cl, 6 to 15 mM KCl, and optionally 0.6 to 1.6 mM MgCl2. Assay buffer is 40-60 mM Tris. 3. The assay of claim ₂ , comprising or consisting of Cl, 7.5 to 12.5 mM KCl, and optionally 0.8 to 1.4 mM MgCl2. Assay buffer is 45 to 55 mM Tris. Cl, 11.5 mM from 8.5 KCl, and assays described optionally from 0.9 to 1.3 mM MgCl ₂ or 1.0 to claim 2 containing 1.2 mM MgCl _2, or if made of. The assay of claim 2, wherein the assay buffer comprises or consists of 50 mM Tris and 10 mM KCl. Assay buffer 50mM Tris, 10mM KCl, and 1.0 mM MgCl _2, or 50mM Tris,, 10mM KCl, and 1.2mM including MgCl _2, or assay of claim 2 comprising a. 9. An assay according to any one of claims 1 to 8, wherein the assay buffer is between pH 7.2 and pH 7.6 at room temperature. The assay according to claim 9, wherein the assay buffer has a pH of 7.4 at room temperature. The assay according to any one of claims 1 to 10, wherein the ERG is human ERG. The assay according to any one of claims 1 to 11, wherein the labeled IKR blocker is labeled dofetilide or labeled MK-499. 13. The assay of claim 12, wherein the labeled dofetilide or labeled MK-499 is a radioactive label. Radiolabel is tritium ⁽³ H), assay of claim 13. The following additional steps (s):
(D) Calculation of the IC ₂₀ of a test compound or mixture of test compounds and optionally (e) comparing the IC ₂₀ value of a test compound or mixture of test compounds in vivo with the concentration required for the desired therapeutic effect of the compound To do,
15. The assay according to any one of claims 1 to 14, wherein 16. An assay according to any one of claims 1 to 15, wherein the assay is performed as a filter binding assay. 16. An assay according to any one of claims 1 to 15, wherein the assay is performed as a scintillation approximation assay.

Images