- Email: [email protected]
A High-Throughput Glow-Type Aequorin Assay for Measuring Receptor-Mediated Changes in Intracellular Calcium Levels
Analytical Biochemistry 286, 231–237 (2000) doi:10.1006/abio.2000.4821, available online at http://www.idealibrary.com on
A High-Throughput Glow-Type Aequorin Assay for Measuring Receptor-Mediated Changes in Intracellular Calcium Levels S. E. George,* M. T. Schaeffer,† D. Cully,† M. S. Beer,‡ and G. McAllister‡ ,1 ‡Merck, Sharp & Dohme, Terlings Park, Eastwick Road, Harlow, Essex, United Kingdom CM20 2QR; †Merck & Company, Lincoln Avenue, Rahway, New Jersey 07065; and *Pfizer Central Research, Ramsgate Road, Sandwich, Kent, United Kingdom CT13 9NJ
Received February 18, 2000
A glow-type aequorin luminescence assay for measuring receptor-mediated stimulation of intracellular calcium levels is described and characterized. The human 5-hydroxytryptamine 2A receptor stably coexpressed in human embryonic kidney cells with apoaequorin was used to characterize the system and showed that following the flash reaction, a stable luminescence signal could be measured using a microplate scintillation counter for between 3 and 7 h after the addition of receptor agonist. Furthermore, this luminescence was dependent on the concentration of agonist used and gave potency values that were stable over this time period. Testing a range of 5-hydroxytryptamine 2A receptor agonists gave the expected rank order of potency for this receptor. The glow luminescence could also be inhibited by 5-hydroxytryptamine 2A receptor antagonists, generating affinity values that directly correlated with those determined for inhibition of the flash reaction carried out under the same buffer conditions. The assay therefore gave pharmacologically relevant data and allows a significant improvement of throughput over the traditional flash-type measurements made using an injecting luminometer. © 2000 Academic Press Key Words: intracellular calcium; high-throughput screening.
Cloning of the sequence for the h5-HT 2A 2 receptor (1– 4) allowed its heterologous expression in mamma1
To whom correspondence should be addressed. Fax: (44) 1279 440712. E-mail: [email protected]. 2 Abbreviations used: h5-HT 2A, human 5-hydroxytryptamine 2A; GPCR, G-protein-coupled receptor; HEK-AEQ, human embryonic kidney aequorin; hEP1, human prostaglandin E1; DMEM, Dul0003-2697/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
lian cell lines, and hence the opportunity to characterize both its ligand binding profile (1, 3, 5) and its functional coupling (6). Given its implicated role in schizophrenia and depression (7–9), such studies are invaluable not only in characterizing a pharmacologically important receptor in the absence of highly related family members, but also as part of drug discovery programs. A member of the GPCR superfamily, the 5-HT 2A receptor has been shown to signal via the Gq G-protein ␣ subunit to alter intracellular levels of inositol phosphate, diacylglycerol, and ultimately calcium (10, 11). Direct measurement of inositol phosphate is rather slow and laborious, so the methods described in this paper relate to the indirect measurement of calcium, detected through its activation of the photoprotein aequorin. Apoaequorin was originally isolated from the coelenterate Aequorea victoria and has been biochemically exploited due to its luminescent properties in response to calcium. It has been successfully expressed in mammalian cells and reconstituted to form active aequorin by incubation of intact cells with the membrane-permeable chromophore cofactor coelenterazine (12–14). In this reconstituted state, the binding of calcium to the aequorin complex results in the oxidation of the
becco’s modified Eagle medium; FCS, fetal calf serum; DOI, ␣-methyl5-hydroxytryptamine maleate; EC 50, molar concentration required to give 50% of its own maximal stimulation; IC 50, molar concentration required to inhibit its own maximal response by 50%; pEC 50, negative logarithm 10 of the molar concentration which produced 50% of its own maximal stimulation; pIC 50, negative logarithm 10 of the molar concentration which inhibited its own maximal response by 50%; pK i, negative logarithm 10 K i, where the K i value is calculated from the IC 50 value using the Cheng–Prusoff equation; nH, Hill slope. 231
232
GEORGE ET AL.
coelenterazine cofactor to coelenteramide, with the concomitant release of CO 2 and the emission of light at 469 nm. Light emission is dependent on the calcium concentration (14), and hence the activation of coexpressed receptors which alter intracellular levels of calcium can be measured in terms of light emission using a luminometer (13–16). Light emission is considered to be transient, reaching a peak within 1–5 s of applying stimulus and then rapidly decaying to a low level. However, in the presence of an excess concentration of coelenterazine, the spent apoaequorin can become reconstituted to form active aequorin, allowing the cycle to repeat and giving rise to a level of regenerated luminescence which is dependent on the concentration of calcium (17, 18). The aim of this paper was to investigate whether this phenomenon could be exploited to give a “glow”-type luminescent readout, thus obviating the need to read samples immediately following the addition of stimulus, and significantly increasing the throughput of the assay. Such a modification would make the aequorin assay much more amenable to automation and thus more useful for high-throughput screening. To verify the pharmacological relevance of data generated by the assay, it was used to carry out a thorough pharmacological characterization of the heterologously expressed human 5-HT 2A receptor and compared to a more traditional luminometer “flash” method of measuring aequorin activation. MATERIALS AND METHODS
HEK-AEQ-17 cells stably expressing the cDNA for apoaequorin (15) were stably transfected with the human 5-HT 2A receptor (at approximately 1 pmol/mg protein). HEK-AEQ-17 cells stably expressing the human prostaglandin EP1 receptor (HEK-AEQ-EP1 cells) were kindly donated by Dr. M. Abramovitz and Dr. K. Metters (Merck Frosst, Montreal, Canada). Coelenterazine was purchased from Molecular Probes or Europa Bioproducts Limited; glutathione, dithiothreitol, carbamyl choline chloride (carbachol), dicyclomine hydrochloride, atropine, pirenzepine dihydrochloride, and 5-hydroxytryptamine creatine sulfate complex (5-HT) were from Sigma; DMEM, penicillin/ streptomycin, and FCS were purchased from Gibco BRL. RU24969 hemisuccinate and metergoline phenylmethyl ester were purchased from Tocris Cookson, R(⫺)-DOI, ␣-methyl-5-hydroxytryptamine maleate, 5-carboxamidotryptamine (5-CT), mianserin hydrochloride, ketanserin tartrate, S(⫺)-propranolol hydrochloride, S(⫺)-pindolol, and methiothepin mesylate were purchased from Research Biochemicals International. HEK-AEQ-h5HT 2A and HEK-AEQ-hEP1 cells were cultured in DMEM supplemented with 10% FCS, 100 units/ml penicillin, 100 g/ml streptomycin, and 2 mM
glutamine. G-418 (1 mg/ml) was applied to stock cells once every 3 weeks to maintain selection. Cells were seeded into white (opaque) 96-well plates at a density of 7.5 ⫻ 10 4 cells/well in 100 l volume of media, the day prior to assay. The following day, cells were approximately 80 –90% confluent and were washed once with 150 l/well of wash buffer (DMEM supplemented with 0.1% FCS and 30 M reduced glutathione). Cells were then incubated with 45 or 50 l/well of charge buffer (wash buffer supplemented with 10 M coelenterazine) for 4 h at 37°C/5% CO 2. Test compounds diluted in assay buffer (0.1 M Hepes, 0.5 mM DTT, pH 7.8) were then added as described below. For flash-type assays, agonist dilutions (50 l) were added to coelenterazine charged cells (50 l) using an MLX injecting microplate luminometer (Dynex) and luminescence measured over 15 s following the injection. The area under the peak of luminescence (in integrated relative light units) measured during the 15 s was recorded. For glow-type assays, agonist dilutions (50 l) were added to coelenterazine-charged cells (50 l) and luminescence was measured on a 1450 MicroBeta Trilux (Wallac) in luminescence mode, typically following a 130-min incubation at room temperature in the dark. Luminescence was recorded for 1 s per well using a photomultiplier tube located above the plate. Luminescence was recorded as cps. For experiments stimulating the endogenous muscarinic receptor, luminescence was quantified using a TopCount Microscintillation counter (Packard) with the same parameters. To measure antagonism, antagonists dilutions were prepared in assay buffer, and 45 l was added to coelenterazine-charged cells (45 l) by hand for both flash and glow methods. Then the mixture was incubated for 30 min at 37°C/5% CO 2. Ten microliters of agonist prepared in 10 times the final assay concentration in assay buffer was then added and luminescence was read, as described above for flash and glow reactions. Data were expressed as a percentage of the luminescence seen with that concentration of agonist in the absence of antagonist on the same plate, following the subtraction of basal luminescence in the absence of agonist. Concentration response data were analyzed by nonlinear regression using GraphPad Prism, to generate an EC 50/IC 50 value and a Hill slope. The equation used was for a sigmoid curve with variable slope: Y ⫽ Bottom ⫹ [(Top ⫺ Bottom)/1 ⫹ 10 (LogEC50 ⫺X) 䡠 Hill slope)]. For data from the endogenous muscarinic receptor, curves were fitted using the sigmoid curve equation with Hill slope ⫽ 1, since the slopes were found to be not significantly different to 1 using the Student t test. Correlation between data sets was analyzed by linear regression using GraphPad Prism.
HIGH-THROUGHPUT GLOW-TYPE AEQUORIN ASSAY TABLE 1
pEC 50 Values for 5-HT 2A Receptor Agonists Measured Using the Glow Assay Compound
pEC 50
nH
n
DOI 5-HT ␣-Methyl 5-HT 5-CT RU24969
7.38 ⫾ 0.09 6.64 ⫾ 0.05 6.76 ⫾ 0.08 5.09 ⫾ 0.02 5.28 ⫾ 0.10
1.5 ⫾ 0.1 1.8 ⫾ 0.1 1.7 ⫾ 0.2 1.7 ⫾ 0.3 1.3 ⫾ 0.2
3 4 4 4 4
Note. Data are the arithmetic mean ⫾ SE of the pEC 50 values determined as described under Materials and Methods, from the indicated number of independent experiments.
RESULTS
When 5-HT was added to cells using an injecting luminometer (flash assay), the stimulation of lumines-
233
cence was dependent on the concentration of 5-HT used, giving a mean pEC 50 of 6.2 ⫾ 0.3. When 5-HT was added manually and luminescence was measured using a MicroBeta Trilux (glow assay), a concentration response could also be measured, with a comparable pEC 50 of 6.6 ⫾ 0.1 (Table 1). The stability of the glow signal was investigated by rereading the plate at intervals over a 19-h period (Fig. 1a). This showed that the signal decayed at a fast rate over the first 100 min, but then stabilized up to around 480 min, before starting to decay again at a slower rate. Figure 1b shows the stability of the potency measurements for 5-HT over the first 480 min. This potency value appears to vary with time but is reasonably stable in the range 30 to 420 min. The signal to background ratio was still very good 130 min after the addition of agonist (the maximum response to 5-HT was 26,700 ⫾ 2900 cps over a
FIG. 1. The decay of 5-HT-stimulated luminescence measured using the glow-type assay, and the effect of time on the pEC 50 of the response. (a) Concentrations of 5-HT (■, 1 nM; ƒ, 10 nM; Œ, 100 nM; , 300 nM; }, 1 M; F, 3 M; 䊐, 10 M) were prepared in assay buffer and applied to coelenterazine-charged HEK-AEQ-5HT 2A cells by hand as described under Materials and Methods. Luminescence was measured using a MicroBeta Trilux at the indicated times, with the plate stored at room temperature in the dark between readings. Data are from a single experiment carried out in triplicate and are representative of three independent experiments. (b) The data in (a) were plotted against the log 10 of the concentration of 5-HT, and a sigmoid curve was fitted as described under Materials and Methods. The resulting pEC 50 determination is plotted against time.
234
GEORGE ET AL.
FIG. 2. Stimulation of luminescence by 5-HT 2A receptor agonists. A range of concentrations of (F) DOI, (䊐) ␣-methyl 5-HT, (}) 5-HT, (‚) RU24969, (■) and 5-CT were prepared in assay buffer, and applied to coelenterazine-charged HEK-AEQ-5HT 2A cells by hand as described under Materials and Methods. Luminescence was measured on a MicroBeta Trilux following a 130-min delay at room temperature. Data are the arithmetic mean ⫾ SE of triplicate determinations from a single experiment representative of at least three carried out.
basal of 4200 ⫾ 700 cps (n ⫽ 4)), hence this time point was used for subsequent experiments. A range of agonists were tested using this glow method on the Trilux, and all gave stimulation of luminescence with the following rank order of potency: DOI ⬎ ␣-methyl 5-HT ⬎ 5-HT ⬎ RU24969 ⬎ 5-CT (Fig. 2 and Table 1). The mean Hill slopes for these curves were generally steep and, when compared using a t test, were found to be significantly greater than one for DOI (P ⫽ 0.047), 5-HT (P ⫽ 0.011), and ␣-methyl 5-HT (P ⫽ 0.022). To investigate antagonism of the 5-HT response, a range of different 5-HT 2A receptor antagonists were applied to cells for 30 min prior to adding 5-HT using both the flash and glow assays, to allow direct comparison of the two methods (Fig. 3). In both cases, the antagonists gave rise to concentration-dependent reversal of the 5-HT-stimulated luminescence, and with the same rank order of potency (Table 2). The pIC 50 values for the antagonists determined using the two methods were compared by linear regression and gave an excellent correlation with an r 2 value of 0.973 and a slope of 1.04 ⫾ 0.09. In common with the agonist data, the Hill slopes of the curves for antagonists measured using the glow assay appeared to be steep in some cases, in contrast to the flash assay data where the Hill slopes were generally lower. When agonist pEC 50 values and antagonist pIC 50 values determined using the glow assay were compared with pK i values determined for the displacement of [ 3H]ketanserin binding at the human 5-HT 2A receptor stably expressed in CHO-K1 cells (19) or in LM(TK ⫺) cells (1), there was excellent correlation giving an r 2 value of 0.976 and a slope of 0.62 ⫾ 0.04 (Fig. 4).
Parallel experiments carried out measuring glow luminescence on a TopCount microscintillation counter (Packard) rather than the Trilux gave directly comparable results, indicating that it can be used interchangeably with the MicroBeta Trilux (data not shown). To investigate whether the response was unique to the 5-HT 2A receptor, experiments were also carried out to stimulate the endogenous muscarinic receptor present in HEK cell lines (20). Here, HEK-AEQ-EP1 cells were stimulated with carbachol, and luminescence was measured using either the luminometer flash assay or the glow assay with readings taken on a TopCount. These experiments showed stimulation of luminescence which could be measured using both
FIG. 3. Inhibition of 5-HT-mediated stimulation of luminescence by 5-HT 2A receptor antagonists. A range of concentrations of (E) ketanserin, (䊐) metergoline, (Œ) methiothepin, (F) mianserin, () propranolol, and ({) pindolol were prepared in assay buffer and preincubated with coelenterazine-charged HEK-AEQ-5HT 2A cells as described under Materials and Methods. (a) 10 l of 100 M 5-HT prepared in assay buffer was then added using the luminometer to give a final concentration of 10 M, and luminescence was measured immediately over 15 s. (b) 10 l of 3 M 5-HT prepared in assay buffer was added by hand to give a final concentration of 300 nM and luminescence was read on a Trilux following a 130-min delay at room temperature. In both cases, data are the arithmetic mean ⫾ SE of triplicate determinations from a single experiment representative of at least three carried out.
235
HIGH-THROUGHPUT GLOW-TYPE AEQUORIN ASSAY TABLE 2
pIC 50 Values for 5-HT 2A Receptor Antagonists Measured Using the Flash- and Glow-Type Assays Flash
Glow
Compound
pIC 50
nH
n
pIC 50
nH
n
Propranolol Pindolol Mianserin Ketanserin Metergoline Methiothepin
4.46 ⫾ 0.43 4.51 ⫾ 0.42 7.65 ⫾ 0.13 8.07 ⫾ 0.09 8.27, 8.07 8.42 ⫾ 0.26
0.69 ⫾ 0.23 1.11 ⫾ 0.22 0.69 ⫾ 0.09 1.39 ⫾ 0.22 1.3, 1.14 1.98 ⫾ 0.46
3 3 3 6 2 3
4.83 ⫾ 0.32 4.52 ⫾ 0.32 7.64 ⫾ 0.07 8.01 ⫾ 0.17 8.89 ⫾ 0.20 8.91 ⫾ 0.25
1.2 ⫾ 0.3 1.0 ⫾ 0.7 1.8 ⫾ 0.2 1.3 ⫾ 0.3 2.1 ⫾ 0.2 1.7 ⫾ 0.1
3 3 3 3 4 3
Note. Data are the arithmetic mean ⫾ SE of the pIC 50 values determined as described under Materials and Methods, from the indicated number of independent experiments.
methods (Fig. 5). The pEC 50 values for the response measured using the glow assay (arithmetic mean ⫾ SE (n) ⫽ 5.58 ⫾ 0.1 (3)) was approximately threefold lower than that using the flash assay (4.82, 5.10 (n ⫽ 2)). Also, as seen previously with the h5-HT 2A receptor data, the Hill slope for the glow assay was steeper than with the flash assay. The maximum stimulation seen in response to carbachol with the glow assay was lower than had been seen with the h5-HT 2A receptor (8775 ⫾ 2581 cps over a basal of 2032 ⫾ 422 cps). The ability of three standard muscarinic receptor antagonists to inhibit this response was investigated as described under Materials and Methods (Fig. 6). The pIC 50 values generated using the two assays were directly comparable (Table 3). DISCUSSION
The expression of the photoprotein aequorin in mammalian cells offers an excellent opportunity to measure
FIG. 4. The correlation between agonist pEC 50 values and antagonist pIC 50 values from the glow assay, with published binding data for the h5-HT 2A receptor. Mean pEC 50 or pIC 50 values shown in Tables 1 and 2 were correlated with published pK i for metergoline, methiothepin, mianserin, ketanserin, and 5-HT (19), and for 5-HT and DOI (1), according to which compounds were included in each study.
receptor-mediated changes in intracellular calcium levels with a high level of sensitivity (13–16). A major drawback of the assay however is that the flash kinetics of the aequorin reaction would necessitate the use of specialized luminometers to allow the injection of a stimulating compound and immediate reading of luminescence for around 15 s. This gives a minimum read time of 25 min per 96-well plate, prohibiting its use in high-throughput screening. Additionally, with cells seeded in the final assay plate, a single syringe injector luminometer limits the testing of agonists since it requires a laborious transfer of injectors between stock bottles of dilutions, obviously impairing automation and reducing throughput. The aim of this paper was to determine whether an end-point glow-type aequorin assay could be developed to replace the kinetic flash assay making it more amenable for use in automated high-throughput screening.
FIG. 5. Stimulation of luminescence in response to carbachol measured using flash- or glow-type assays. Agonist dilutions were prepared in assay buffer and were added to coelenterazine charged HEK-AEQ-5HT 2A cells either using the luminometer with immediate measurement of luminescence for 15 s (flash assay, ■) or were added by hand, and luminescence was measured using a TopCount following a 10-min count delay (glow assay, 䊐). The values shown are the arithmetic mean ⫾ SE of triplicate determinations from a single experiment. Data were normalized to the maximum response.
236
GEORGE ET AL.
FIG. 6. Inhibition of carbachol-stimulated luminescence by muscarinic receptor antagonists. Muscarinic receptor antagonists were prepared in assay buffer and preincubated with coelenterazinecharged HEK-AEQ-5HT 2A cells as described under Materials and Methods. (a) Flash assay: 10 l of 1 mM carbachol prepared in assay buffer was then added using the luminometer to give a final concentration of 100 M, and luminescence was measured for 15 s. (b) Glow assay: 10 l of 30 M carbachol prepared in assay buffer was added by hand to give a final concentration of 3 M, and luminescence was measured using a TopCount. Values shown are the arithmetic mean ⫾ SE of triplicate determinations from a single experiment representative of four carried out showing the response to (■) atropine; (E) pirenzepine, and (Œ) dicyclomine.
Experiments using an HEK cell line stably transfected with apoaequorin and the h5-HT 2A receptor showed that addition of agonist dilutions directly to the charge buffer gave a luminescent signal which could be measured on a microplate scintillation counter in luminescence mode over a 19-h period. The stability of the EC 50 value of the 5-HT 2A receptor-mediated response over the period of 3–7 h allows a large time window for compound addition and stacking of plates on the scintillation counter. Clearly the stability of the luminescence decay and the generated EC 50 values should be established for other receptors employed in a similar assay system because this could be affected by variables such as the levels of receptor expression or receptor desensitization. Coupled with the 1-s read time per well with the glow assay and the use of two (TopCount)
or six (Trilux) detectors (which allow a plate to be read in as little as 60 s), this has obvious implications for the throughput of the assay. Additionally, the assay involves a minimal amount of washing and pipetting, making it fully compatible with automation. Although thought of as a flash reaction, an explanation for this phenomenon is that the luminescence measured could be “regenerated luminescence” permitted by the presence of an excess concentration of coelenterazine throughout the assay. In this case, aequorin spent due to the oxidation reaction triggered by calcium can become reconstituted to form active aequorin due to the continued presence of coelenterazine. This aequorin is then able to emit light in response to the calcium which is still elevated in the cytoplasm, resulting in a continuous luminescence with gradual consumption of coelenterazine (17, 18). In this case, the luminescent signal would be limited only by the amount of coelenterazine present (and its stability), the speed of regeneration of the active aequorin, and the level of intracellular calcium. Importantly, the luminescence measured using the glow assay was pharmacologically relevant, giving concentration response curves to a variety of 5-HT 2A receptor agonists with the expected rank order of potency for this receptor (10). Furthermore, the agonist-stimulated luminescence could be inhibited by a range of antagonists with affinities which were directly comparable to those measured using established flash assay protocols, and which correlated with previously published binding assays on this receptor (1, 19). The only possible concern was that the Hill slopes generated were generally high, particularly with glow assay. This is most likely due to methodological design and is enhanced in the glow assays by the sensitivity of the scintillation counters in luminescence mode, rather than cooperativity of the response, and does not detract from the main finding of the study that the two methods investigated yield comparable rank order pharmacological profiles.
TABLE 3
pIC 50 Values for Muscarinic Receptor Antagonists Measured Using the Flash- and Glow-Type Assays pIC 50 ⫾ SE (n) Compound
Flash
Glow
Atropine Dicyclomine Pirenzepine
9.72 ⫾ 0.26 (4) 6.72 ⫾ 0.10 (4) 6.68 ⫾ 0.08 (4)
9.53 ⫾ 0.21 (4) 6.96 ⫾ 0.19 (4) 6.88 ⫾ 0.22 (4)
Note. Data are the arithmetic mean ⫾ SE of the pIC 50 values as described under Materials and Methods, from the indicated number of independent experiments.
HIGH-THROUGHPUT GLOW-TYPE AEQUORIN ASSAY
The glow-type assay was also shown to be of value in determining agonist/antagonist activity at an endogenous muscarinic receptor, since antagonism of carbachol-stimulated luminescence could be measured giving the same rank order of potency for three standard muscarinic receptor antagonists, demonstrating that it is applicable to another member of the GPCR superfamily. Furthermore, although the maximum stimulation in response to carbachol in HEK-AEQ-hEP1 cells was lower than that in response to 5-HT in HEK-AEQh5-HT 2A cells, these experiments demonstrated that the glow assay could be used with physiologically relevant levels of receptor expression. Additionally, it has been shown to give pharmacologically relevant data for the prostaglandin EP1 receptor expressed at 5 pmol/mg protein (data not shown), indicating that it may be applicable to a variety of GPCRs, with a range of expression levels. In summary, this assay system offers a distinct ergonomic advantage over conventional luminometry, while generating comparable data. ACKNOWLEDGMENTS We thank Dr. C. Roelent (RTRT, Leuven, Belgium) for useful discussions and Packard Instrument B.V. for the introduction. We also thank Dr. M. Abramovitz and Dr. K. Metters (Merck Frosst, Montreal, Canada) for the use of HEK-AEQ-EP1 cells.
REFERENCES 1. Branchek, T., Adham, N., Macchi, M., Kao, H-T., and Hartig, P. (1990) [ 3H]-DOB-(4-bromo-2,5-dimethoxyphenylisopropylamine) and [ 3H]-ketanserin label two affinity states of the cloned human 5-hydroxytryptamine 2 receptor. Mol. Pharmacol. 38, 604 – 609. 2. Saltzman, A., Morse, B., Whitman, M., Ivanshchenko, Y., Jaye, M., and Felder, S. (1991) Cloning of the human serotonin 5-HT 2 and 5-HT 1C receptor subtypes. Biochem. Biophys. Res. Commun. 181, 1469 –1478. 3. Stam, N., Huizen, F., Alwbeek, C., Brands, J., Dijkema, R., Tonnaer, J., and Olijve, W. (1992) Genomic organisation, coding sequence and functional expression of human 5-HT 2 and 5-HT 1A receptor genes. Eur. J. Pharmacol. 227, 153–162. 4. Chen, K., Yang, W., Grimsby, J., and Shih, J. (1992) The human 5-HT 2 receptor is encoded by a multiple intron– exon gene. Mol. Brain Res. 14, 20 –26. 5. Szele, F., and Pritchett, D. (1993) High affinity agonist binding to cloned 5-hydroxytryptamine 2 receptors is not sensitive to GTP analogs. Mol. Pharmacol. 43, 915–920. 6. Newton, R., Phipps, S., Flanigan, T., Newberry, N., Carey, J., Kumar, C., McDonald, B., Chen, C., and Elliott, J. (1996) Characterisation of human 5-hydroxytryptamine 2A and 5-hydroxytryptamine 2C receptors expressed in the human neuroblastoma cell line SH-SY5Y: Comparative stimulation by hallucinogenic drugs. J. Neurochem. 67, 2521–2531.
237
7. Martin, P., Waters, N., Schmidt, C. J., Carlsson, A., and Carlsson, M. (1998) Rodent data and general hypothesis: Antipsychotic action exerted through 5-HT 2A receptor antagonism is dependent on increased serotonergic tone. J. Neural. Transm. 105, 65–396. 8. Peroutka, S. J. (1998) Serotonin receptor variants in disease: New therapeutic opportunities? Ann. NY Acad. Sci. 861, 16 –25. 9. Roth, B., Berr, S., Kroeze, W., Willins, D., and Kristiansen, K. (1998) Serotonin 5-HT2A receptors: Molecular biology and mechanisms of regulation. Crit. Rev. Neurobiol. 12, 319 –338. 10. Doyle, V., Creba, J., Ruegg, U., and Hoyer, D. (1986) Serotonin increases the production of inositol phosphates and mobilises calcium via the 5-HT 2 receptor in A7r5 smooth muscle cells. Naunyn-Schmiedeberg’s Arch. Pharmacol. 333, 98 –103. 11. Roth, B., Nakaki, T., Chuang, D., and Costa, E. (1984) Aortic recognition sites for serotonin (5-HT) are coupled to phospholipase C and modulate phosphatidyl inositol turnover. Neuropharmacology 23, 1223–1225. 12. Rizzuto, R., Simpson, A., Brini, M., and Pozzan, T. (1992) Rapid changes of mitochondrial Ca 2⫹ revealed by specifically targeted recombinant aequorin. Nature 358, 325–327. 13. Sheu, Y., Kricka, L., and Pritchett, D. (1993) Measurement of intracellular calcium using bioluminescent aequorin expressed in human cells. Anal. Biochem. 209, 343–347. 14. Brini, M., Marsault, R., Bastuianutto, C., Alvarez, J., Pozzan, T., and Rizzuto, R. (1995) Transfected aequorin in the measurement of cytosolic Ca 2⫹ concentration ([Ca 2⫹] c). J. Biol. Chem. 270, 9896 –9903. 15. Button, D., and Brownstein, M. (1993) Aequorin-expressing mammalian cell lines used to report Ca 2⫹ mobilisation. Cell Calcium 14, 663– 671. 16. Stables, J., Green, A., Marshall, F., Fraser, N., Knight, E., Sautel, M., Milligan, G., Lee, M., and Rees, S. (1997) A bioluminescent assay for agonist activity at potentially any G-protein coupled receptor. Anal. Biochem. 252, 115–126. 17. Shimomura, O. (1995) Cause of spectral variation in the luminescence of semisynthetic aequorins. Biochem. J. 306, 537–543. 18. Shimomura, O., Kishi, Y., and Inouyes, S. (1993) The relative rate of aequorin regeneration from apoaequorin and coelenterazine analogues. Biochem. J. 296, 549 –551. 19. Bonhaus, D., Weinhardt, K., Taylor, M., Desouza, A., McNeeley, P., Szczepanski, K., Fontana, D., Trinh, J., Rocha, C., Dawson, M., Flippin, L., and Eglen, R. (1997) RS-102221: A novel high affinity and selective 5-HT 2C receptor antagonist. Neuropharmacology 36, 621– 629. 20. Conklin, B., Chabre, O., Won, Y., Federman, A., and Bourne, H. (1992) Recombinant Gq␣. J. Biol. Chem. 267, 31–34. 21. Offermanns, S., and Simon, M. (1995) G␣ 15 and G␣ 16 couple a wide variety of receptors to phospholipase C. J. Biol. Chem. 270, 15175–15180. 22. Milligan, G., Marshall, F., and Rees, S. (1996) G 16 as a universal G protein adapter: Implications for agonist screening strategies. Trends Pharmacol. Sci. 17, 235–237. 23. Milligan, G., and Rees, S. (1999) Chimaeric G␣ proteins: Their potential use in drug discovery. Trends Pharmacol. Sci. 20, 118 –124.
A High-Throughput Glow-Type Aequorin Assay for Measuring Receptor-Mediated Changes in Intracellular Calcium Levels S. E. George,* M. T. Schaeffer,† D. Cully,† M. S. Beer,‡ and G. McAllister‡ ,1 ‡Merck, Sharp & Dohme, Terlings Park, Eastwick Road, Harlow, Essex, United Kingdom CM20 2QR; †Merck & Company, Lincoln Avenue, Rahway, New Jersey 07065; and *Pfizer Central Research, Ramsgate Road, Sandwich, Kent, United Kingdom CT13 9NJ
Received February 18, 2000
A glow-type aequorin luminescence assay for measuring receptor-mediated stimulation of intracellular calcium levels is described and characterized. The human 5-hydroxytryptamine 2A receptor stably coexpressed in human embryonic kidney cells with apoaequorin was used to characterize the system and showed that following the flash reaction, a stable luminescence signal could be measured using a microplate scintillation counter for between 3 and 7 h after the addition of receptor agonist. Furthermore, this luminescence was dependent on the concentration of agonist used and gave potency values that were stable over this time period. Testing a range of 5-hydroxytryptamine 2A receptor agonists gave the expected rank order of potency for this receptor. The glow luminescence could also be inhibited by 5-hydroxytryptamine 2A receptor antagonists, generating affinity values that directly correlated with those determined for inhibition of the flash reaction carried out under the same buffer conditions. The assay therefore gave pharmacologically relevant data and allows a significant improvement of throughput over the traditional flash-type measurements made using an injecting luminometer. © 2000 Academic Press Key Words: intracellular calcium; high-throughput screening.
Cloning of the sequence for the h5-HT 2A 2 receptor (1– 4) allowed its heterologous expression in mamma1
To whom correspondence should be addressed. Fax: (44) 1279 440712. E-mail: [email protected]. 2 Abbreviations used: h5-HT 2A, human 5-hydroxytryptamine 2A; GPCR, G-protein-coupled receptor; HEK-AEQ, human embryonic kidney aequorin; hEP1, human prostaglandin E1; DMEM, Dul0003-2697/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
lian cell lines, and hence the opportunity to characterize both its ligand binding profile (1, 3, 5) and its functional coupling (6). Given its implicated role in schizophrenia and depression (7–9), such studies are invaluable not only in characterizing a pharmacologically important receptor in the absence of highly related family members, but also as part of drug discovery programs. A member of the GPCR superfamily, the 5-HT 2A receptor has been shown to signal via the Gq G-protein ␣ subunit to alter intracellular levels of inositol phosphate, diacylglycerol, and ultimately calcium (10, 11). Direct measurement of inositol phosphate is rather slow and laborious, so the methods described in this paper relate to the indirect measurement of calcium, detected through its activation of the photoprotein aequorin. Apoaequorin was originally isolated from the coelenterate Aequorea victoria and has been biochemically exploited due to its luminescent properties in response to calcium. It has been successfully expressed in mammalian cells and reconstituted to form active aequorin by incubation of intact cells with the membrane-permeable chromophore cofactor coelenterazine (12–14). In this reconstituted state, the binding of calcium to the aequorin complex results in the oxidation of the
becco’s modified Eagle medium; FCS, fetal calf serum; DOI, ␣-methyl5-hydroxytryptamine maleate; EC 50, molar concentration required to give 50% of its own maximal stimulation; IC 50, molar concentration required to inhibit its own maximal response by 50%; pEC 50, negative logarithm 10 of the molar concentration which produced 50% of its own maximal stimulation; pIC 50, negative logarithm 10 of the molar concentration which inhibited its own maximal response by 50%; pK i, negative logarithm 10 K i, where the K i value is calculated from the IC 50 value using the Cheng–Prusoff equation; nH, Hill slope. 231
232
GEORGE ET AL.
coelenterazine cofactor to coelenteramide, with the concomitant release of CO 2 and the emission of light at 469 nm. Light emission is dependent on the calcium concentration (14), and hence the activation of coexpressed receptors which alter intracellular levels of calcium can be measured in terms of light emission using a luminometer (13–16). Light emission is considered to be transient, reaching a peak within 1–5 s of applying stimulus and then rapidly decaying to a low level. However, in the presence of an excess concentration of coelenterazine, the spent apoaequorin can become reconstituted to form active aequorin, allowing the cycle to repeat and giving rise to a level of regenerated luminescence which is dependent on the concentration of calcium (17, 18). The aim of this paper was to investigate whether this phenomenon could be exploited to give a “glow”-type luminescent readout, thus obviating the need to read samples immediately following the addition of stimulus, and significantly increasing the throughput of the assay. Such a modification would make the aequorin assay much more amenable to automation and thus more useful for high-throughput screening. To verify the pharmacological relevance of data generated by the assay, it was used to carry out a thorough pharmacological characterization of the heterologously expressed human 5-HT 2A receptor and compared to a more traditional luminometer “flash” method of measuring aequorin activation. MATERIALS AND METHODS
HEK-AEQ-17 cells stably expressing the cDNA for apoaequorin (15) were stably transfected with the human 5-HT 2A receptor (at approximately 1 pmol/mg protein). HEK-AEQ-17 cells stably expressing the human prostaglandin EP1 receptor (HEK-AEQ-EP1 cells) were kindly donated by Dr. M. Abramovitz and Dr. K. Metters (Merck Frosst, Montreal, Canada). Coelenterazine was purchased from Molecular Probes or Europa Bioproducts Limited; glutathione, dithiothreitol, carbamyl choline chloride (carbachol), dicyclomine hydrochloride, atropine, pirenzepine dihydrochloride, and 5-hydroxytryptamine creatine sulfate complex (5-HT) were from Sigma; DMEM, penicillin/ streptomycin, and FCS were purchased from Gibco BRL. RU24969 hemisuccinate and metergoline phenylmethyl ester were purchased from Tocris Cookson, R(⫺)-DOI, ␣-methyl-5-hydroxytryptamine maleate, 5-carboxamidotryptamine (5-CT), mianserin hydrochloride, ketanserin tartrate, S(⫺)-propranolol hydrochloride, S(⫺)-pindolol, and methiothepin mesylate were purchased from Research Biochemicals International. HEK-AEQ-h5HT 2A and HEK-AEQ-hEP1 cells were cultured in DMEM supplemented with 10% FCS, 100 units/ml penicillin, 100 g/ml streptomycin, and 2 mM
glutamine. G-418 (1 mg/ml) was applied to stock cells once every 3 weeks to maintain selection. Cells were seeded into white (opaque) 96-well plates at a density of 7.5 ⫻ 10 4 cells/well in 100 l volume of media, the day prior to assay. The following day, cells were approximately 80 –90% confluent and were washed once with 150 l/well of wash buffer (DMEM supplemented with 0.1% FCS and 30 M reduced glutathione). Cells were then incubated with 45 or 50 l/well of charge buffer (wash buffer supplemented with 10 M coelenterazine) for 4 h at 37°C/5% CO 2. Test compounds diluted in assay buffer (0.1 M Hepes, 0.5 mM DTT, pH 7.8) were then added as described below. For flash-type assays, agonist dilutions (50 l) were added to coelenterazine charged cells (50 l) using an MLX injecting microplate luminometer (Dynex) and luminescence measured over 15 s following the injection. The area under the peak of luminescence (in integrated relative light units) measured during the 15 s was recorded. For glow-type assays, agonist dilutions (50 l) were added to coelenterazine-charged cells (50 l) and luminescence was measured on a 1450 MicroBeta Trilux (Wallac) in luminescence mode, typically following a 130-min incubation at room temperature in the dark. Luminescence was recorded for 1 s per well using a photomultiplier tube located above the plate. Luminescence was recorded as cps. For experiments stimulating the endogenous muscarinic receptor, luminescence was quantified using a TopCount Microscintillation counter (Packard) with the same parameters. To measure antagonism, antagonists dilutions were prepared in assay buffer, and 45 l was added to coelenterazine-charged cells (45 l) by hand for both flash and glow methods. Then the mixture was incubated for 30 min at 37°C/5% CO 2. Ten microliters of agonist prepared in 10 times the final assay concentration in assay buffer was then added and luminescence was read, as described above for flash and glow reactions. Data were expressed as a percentage of the luminescence seen with that concentration of agonist in the absence of antagonist on the same plate, following the subtraction of basal luminescence in the absence of agonist. Concentration response data were analyzed by nonlinear regression using GraphPad Prism, to generate an EC 50/IC 50 value and a Hill slope. The equation used was for a sigmoid curve with variable slope: Y ⫽ Bottom ⫹ [(Top ⫺ Bottom)/1 ⫹ 10 (LogEC50 ⫺X) 䡠 Hill slope)]. For data from the endogenous muscarinic receptor, curves were fitted using the sigmoid curve equation with Hill slope ⫽ 1, since the slopes were found to be not significantly different to 1 using the Student t test. Correlation between data sets was analyzed by linear regression using GraphPad Prism.
HIGH-THROUGHPUT GLOW-TYPE AEQUORIN ASSAY TABLE 1
pEC 50 Values for 5-HT 2A Receptor Agonists Measured Using the Glow Assay Compound
pEC 50
nH
n
DOI 5-HT ␣-Methyl 5-HT 5-CT RU24969
7.38 ⫾ 0.09 6.64 ⫾ 0.05 6.76 ⫾ 0.08 5.09 ⫾ 0.02 5.28 ⫾ 0.10
1.5 ⫾ 0.1 1.8 ⫾ 0.1 1.7 ⫾ 0.2 1.7 ⫾ 0.3 1.3 ⫾ 0.2
3 4 4 4 4
Note. Data are the arithmetic mean ⫾ SE of the pEC 50 values determined as described under Materials and Methods, from the indicated number of independent experiments.
RESULTS
When 5-HT was added to cells using an injecting luminometer (flash assay), the stimulation of lumines-
233
cence was dependent on the concentration of 5-HT used, giving a mean pEC 50 of 6.2 ⫾ 0.3. When 5-HT was added manually and luminescence was measured using a MicroBeta Trilux (glow assay), a concentration response could also be measured, with a comparable pEC 50 of 6.6 ⫾ 0.1 (Table 1). The stability of the glow signal was investigated by rereading the plate at intervals over a 19-h period (Fig. 1a). This showed that the signal decayed at a fast rate over the first 100 min, but then stabilized up to around 480 min, before starting to decay again at a slower rate. Figure 1b shows the stability of the potency measurements for 5-HT over the first 480 min. This potency value appears to vary with time but is reasonably stable in the range 30 to 420 min. The signal to background ratio was still very good 130 min after the addition of agonist (the maximum response to 5-HT was 26,700 ⫾ 2900 cps over a
FIG. 1. The decay of 5-HT-stimulated luminescence measured using the glow-type assay, and the effect of time on the pEC 50 of the response. (a) Concentrations of 5-HT (■, 1 nM; ƒ, 10 nM; Œ, 100 nM; , 300 nM; }, 1 M; F, 3 M; 䊐, 10 M) were prepared in assay buffer and applied to coelenterazine-charged HEK-AEQ-5HT 2A cells by hand as described under Materials and Methods. Luminescence was measured using a MicroBeta Trilux at the indicated times, with the plate stored at room temperature in the dark between readings. Data are from a single experiment carried out in triplicate and are representative of three independent experiments. (b) The data in (a) were plotted against the log 10 of the concentration of 5-HT, and a sigmoid curve was fitted as described under Materials and Methods. The resulting pEC 50 determination is plotted against time.
234
GEORGE ET AL.
FIG. 2. Stimulation of luminescence by 5-HT 2A receptor agonists. A range of concentrations of (F) DOI, (䊐) ␣-methyl 5-HT, (}) 5-HT, (‚) RU24969, (■) and 5-CT were prepared in assay buffer, and applied to coelenterazine-charged HEK-AEQ-5HT 2A cells by hand as described under Materials and Methods. Luminescence was measured on a MicroBeta Trilux following a 130-min delay at room temperature. Data are the arithmetic mean ⫾ SE of triplicate determinations from a single experiment representative of at least three carried out.
basal of 4200 ⫾ 700 cps (n ⫽ 4)), hence this time point was used for subsequent experiments. A range of agonists were tested using this glow method on the Trilux, and all gave stimulation of luminescence with the following rank order of potency: DOI ⬎ ␣-methyl 5-HT ⬎ 5-HT ⬎ RU24969 ⬎ 5-CT (Fig. 2 and Table 1). The mean Hill slopes for these curves were generally steep and, when compared using a t test, were found to be significantly greater than one for DOI (P ⫽ 0.047), 5-HT (P ⫽ 0.011), and ␣-methyl 5-HT (P ⫽ 0.022). To investigate antagonism of the 5-HT response, a range of different 5-HT 2A receptor antagonists were applied to cells for 30 min prior to adding 5-HT using both the flash and glow assays, to allow direct comparison of the two methods (Fig. 3). In both cases, the antagonists gave rise to concentration-dependent reversal of the 5-HT-stimulated luminescence, and with the same rank order of potency (Table 2). The pIC 50 values for the antagonists determined using the two methods were compared by linear regression and gave an excellent correlation with an r 2 value of 0.973 and a slope of 1.04 ⫾ 0.09. In common with the agonist data, the Hill slopes of the curves for antagonists measured using the glow assay appeared to be steep in some cases, in contrast to the flash assay data where the Hill slopes were generally lower. When agonist pEC 50 values and antagonist pIC 50 values determined using the glow assay were compared with pK i values determined for the displacement of [ 3H]ketanserin binding at the human 5-HT 2A receptor stably expressed in CHO-K1 cells (19) or in LM(TK ⫺) cells (1), there was excellent correlation giving an r 2 value of 0.976 and a slope of 0.62 ⫾ 0.04 (Fig. 4).
Parallel experiments carried out measuring glow luminescence on a TopCount microscintillation counter (Packard) rather than the Trilux gave directly comparable results, indicating that it can be used interchangeably with the MicroBeta Trilux (data not shown). To investigate whether the response was unique to the 5-HT 2A receptor, experiments were also carried out to stimulate the endogenous muscarinic receptor present in HEK cell lines (20). Here, HEK-AEQ-EP1 cells were stimulated with carbachol, and luminescence was measured using either the luminometer flash assay or the glow assay with readings taken on a TopCount. These experiments showed stimulation of luminescence which could be measured using both
FIG. 3. Inhibition of 5-HT-mediated stimulation of luminescence by 5-HT 2A receptor antagonists. A range of concentrations of (E) ketanserin, (䊐) metergoline, (Œ) methiothepin, (F) mianserin, () propranolol, and ({) pindolol were prepared in assay buffer and preincubated with coelenterazine-charged HEK-AEQ-5HT 2A cells as described under Materials and Methods. (a) 10 l of 100 M 5-HT prepared in assay buffer was then added using the luminometer to give a final concentration of 10 M, and luminescence was measured immediately over 15 s. (b) 10 l of 3 M 5-HT prepared in assay buffer was added by hand to give a final concentration of 300 nM and luminescence was read on a Trilux following a 130-min delay at room temperature. In both cases, data are the arithmetic mean ⫾ SE of triplicate determinations from a single experiment representative of at least three carried out.
235
HIGH-THROUGHPUT GLOW-TYPE AEQUORIN ASSAY TABLE 2
pIC 50 Values for 5-HT 2A Receptor Antagonists Measured Using the Flash- and Glow-Type Assays Flash
Glow
Compound
pIC 50
nH
n
pIC 50
nH
n
Propranolol Pindolol Mianserin Ketanserin Metergoline Methiothepin
4.46 ⫾ 0.43 4.51 ⫾ 0.42 7.65 ⫾ 0.13 8.07 ⫾ 0.09 8.27, 8.07 8.42 ⫾ 0.26
0.69 ⫾ 0.23 1.11 ⫾ 0.22 0.69 ⫾ 0.09 1.39 ⫾ 0.22 1.3, 1.14 1.98 ⫾ 0.46
3 3 3 6 2 3
4.83 ⫾ 0.32 4.52 ⫾ 0.32 7.64 ⫾ 0.07 8.01 ⫾ 0.17 8.89 ⫾ 0.20 8.91 ⫾ 0.25
1.2 ⫾ 0.3 1.0 ⫾ 0.7 1.8 ⫾ 0.2 1.3 ⫾ 0.3 2.1 ⫾ 0.2 1.7 ⫾ 0.1
3 3 3 3 4 3
Note. Data are the arithmetic mean ⫾ SE of the pIC 50 values determined as described under Materials and Methods, from the indicated number of independent experiments.
methods (Fig. 5). The pEC 50 values for the response measured using the glow assay (arithmetic mean ⫾ SE (n) ⫽ 5.58 ⫾ 0.1 (3)) was approximately threefold lower than that using the flash assay (4.82, 5.10 (n ⫽ 2)). Also, as seen previously with the h5-HT 2A receptor data, the Hill slope for the glow assay was steeper than with the flash assay. The maximum stimulation seen in response to carbachol with the glow assay was lower than had been seen with the h5-HT 2A receptor (8775 ⫾ 2581 cps over a basal of 2032 ⫾ 422 cps). The ability of three standard muscarinic receptor antagonists to inhibit this response was investigated as described under Materials and Methods (Fig. 6). The pIC 50 values generated using the two assays were directly comparable (Table 3). DISCUSSION
The expression of the photoprotein aequorin in mammalian cells offers an excellent opportunity to measure
FIG. 4. The correlation between agonist pEC 50 values and antagonist pIC 50 values from the glow assay, with published binding data for the h5-HT 2A receptor. Mean pEC 50 or pIC 50 values shown in Tables 1 and 2 were correlated with published pK i for metergoline, methiothepin, mianserin, ketanserin, and 5-HT (19), and for 5-HT and DOI (1), according to which compounds were included in each study.
receptor-mediated changes in intracellular calcium levels with a high level of sensitivity (13–16). A major drawback of the assay however is that the flash kinetics of the aequorin reaction would necessitate the use of specialized luminometers to allow the injection of a stimulating compound and immediate reading of luminescence for around 15 s. This gives a minimum read time of 25 min per 96-well plate, prohibiting its use in high-throughput screening. Additionally, with cells seeded in the final assay plate, a single syringe injector luminometer limits the testing of agonists since it requires a laborious transfer of injectors between stock bottles of dilutions, obviously impairing automation and reducing throughput. The aim of this paper was to determine whether an end-point glow-type aequorin assay could be developed to replace the kinetic flash assay making it more amenable for use in automated high-throughput screening.
FIG. 5. Stimulation of luminescence in response to carbachol measured using flash- or glow-type assays. Agonist dilutions were prepared in assay buffer and were added to coelenterazine charged HEK-AEQ-5HT 2A cells either using the luminometer with immediate measurement of luminescence for 15 s (flash assay, ■) or were added by hand, and luminescence was measured using a TopCount following a 10-min count delay (glow assay, 䊐). The values shown are the arithmetic mean ⫾ SE of triplicate determinations from a single experiment. Data were normalized to the maximum response.
236
GEORGE ET AL.
FIG. 6. Inhibition of carbachol-stimulated luminescence by muscarinic receptor antagonists. Muscarinic receptor antagonists were prepared in assay buffer and preincubated with coelenterazinecharged HEK-AEQ-5HT 2A cells as described under Materials and Methods. (a) Flash assay: 10 l of 1 mM carbachol prepared in assay buffer was then added using the luminometer to give a final concentration of 100 M, and luminescence was measured for 15 s. (b) Glow assay: 10 l of 30 M carbachol prepared in assay buffer was added by hand to give a final concentration of 3 M, and luminescence was measured using a TopCount. Values shown are the arithmetic mean ⫾ SE of triplicate determinations from a single experiment representative of four carried out showing the response to (■) atropine; (E) pirenzepine, and (Œ) dicyclomine.
Experiments using an HEK cell line stably transfected with apoaequorin and the h5-HT 2A receptor showed that addition of agonist dilutions directly to the charge buffer gave a luminescent signal which could be measured on a microplate scintillation counter in luminescence mode over a 19-h period. The stability of the EC 50 value of the 5-HT 2A receptor-mediated response over the period of 3–7 h allows a large time window for compound addition and stacking of plates on the scintillation counter. Clearly the stability of the luminescence decay and the generated EC 50 values should be established for other receptors employed in a similar assay system because this could be affected by variables such as the levels of receptor expression or receptor desensitization. Coupled with the 1-s read time per well with the glow assay and the use of two (TopCount)
or six (Trilux) detectors (which allow a plate to be read in as little as 60 s), this has obvious implications for the throughput of the assay. Additionally, the assay involves a minimal amount of washing and pipetting, making it fully compatible with automation. Although thought of as a flash reaction, an explanation for this phenomenon is that the luminescence measured could be “regenerated luminescence” permitted by the presence of an excess concentration of coelenterazine throughout the assay. In this case, aequorin spent due to the oxidation reaction triggered by calcium can become reconstituted to form active aequorin due to the continued presence of coelenterazine. This aequorin is then able to emit light in response to the calcium which is still elevated in the cytoplasm, resulting in a continuous luminescence with gradual consumption of coelenterazine (17, 18). In this case, the luminescent signal would be limited only by the amount of coelenterazine present (and its stability), the speed of regeneration of the active aequorin, and the level of intracellular calcium. Importantly, the luminescence measured using the glow assay was pharmacologically relevant, giving concentration response curves to a variety of 5-HT 2A receptor agonists with the expected rank order of potency for this receptor (10). Furthermore, the agonist-stimulated luminescence could be inhibited by a range of antagonists with affinities which were directly comparable to those measured using established flash assay protocols, and which correlated with previously published binding assays on this receptor (1, 19). The only possible concern was that the Hill slopes generated were generally high, particularly with glow assay. This is most likely due to methodological design and is enhanced in the glow assays by the sensitivity of the scintillation counters in luminescence mode, rather than cooperativity of the response, and does not detract from the main finding of the study that the two methods investigated yield comparable rank order pharmacological profiles.
TABLE 3
pIC 50 Values for Muscarinic Receptor Antagonists Measured Using the Flash- and Glow-Type Assays pIC 50 ⫾ SE (n) Compound
Flash
Glow
Atropine Dicyclomine Pirenzepine
9.72 ⫾ 0.26 (4) 6.72 ⫾ 0.10 (4) 6.68 ⫾ 0.08 (4)
9.53 ⫾ 0.21 (4) 6.96 ⫾ 0.19 (4) 6.88 ⫾ 0.22 (4)
Note. Data are the arithmetic mean ⫾ SE of the pIC 50 values as described under Materials and Methods, from the indicated number of independent experiments.
HIGH-THROUGHPUT GLOW-TYPE AEQUORIN ASSAY
The glow-type assay was also shown to be of value in determining agonist/antagonist activity at an endogenous muscarinic receptor, since antagonism of carbachol-stimulated luminescence could be measured giving the same rank order of potency for three standard muscarinic receptor antagonists, demonstrating that it is applicable to another member of the GPCR superfamily. Furthermore, although the maximum stimulation in response to carbachol in HEK-AEQ-hEP1 cells was lower than that in response to 5-HT in HEK-AEQh5-HT 2A cells, these experiments demonstrated that the glow assay could be used with physiologically relevant levels of receptor expression. Additionally, it has been shown to give pharmacologically relevant data for the prostaglandin EP1 receptor expressed at 5 pmol/mg protein (data not shown), indicating that it may be applicable to a variety of GPCRs, with a range of expression levels. In summary, this assay system offers a distinct ergonomic advantage over conventional luminometry, while generating comparable data. ACKNOWLEDGMENTS We thank Dr. C. Roelent (RTRT, Leuven, Belgium) for useful discussions and Packard Instrument B.V. for the introduction. We also thank Dr. M. Abramovitz and Dr. K. Metters (Merck Frosst, Montreal, Canada) for the use of HEK-AEQ-EP1 cells.
REFERENCES 1. Branchek, T., Adham, N., Macchi, M., Kao, H-T., and Hartig, P. (1990) [ 3H]-DOB-(4-bromo-2,5-dimethoxyphenylisopropylamine) and [ 3H]-ketanserin label two affinity states of the cloned human 5-hydroxytryptamine 2 receptor. Mol. Pharmacol. 38, 604 – 609. 2. Saltzman, A., Morse, B., Whitman, M., Ivanshchenko, Y., Jaye, M., and Felder, S. (1991) Cloning of the human serotonin 5-HT 2 and 5-HT 1C receptor subtypes. Biochem. Biophys. Res. Commun. 181, 1469 –1478. 3. Stam, N., Huizen, F., Alwbeek, C., Brands, J., Dijkema, R., Tonnaer, J., and Olijve, W. (1992) Genomic organisation, coding sequence and functional expression of human 5-HT 2 and 5-HT 1A receptor genes. Eur. J. Pharmacol. 227, 153–162. 4. Chen, K., Yang, W., Grimsby, J., and Shih, J. (1992) The human 5-HT 2 receptor is encoded by a multiple intron– exon gene. Mol. Brain Res. 14, 20 –26. 5. Szele, F., and Pritchett, D. (1993) High affinity agonist binding to cloned 5-hydroxytryptamine 2 receptors is not sensitive to GTP analogs. Mol. Pharmacol. 43, 915–920. 6. Newton, R., Phipps, S., Flanigan, T., Newberry, N., Carey, J., Kumar, C., McDonald, B., Chen, C., and Elliott, J. (1996) Characterisation of human 5-hydroxytryptamine 2A and 5-hydroxytryptamine 2C receptors expressed in the human neuroblastoma cell line SH-SY5Y: Comparative stimulation by hallucinogenic drugs. J. Neurochem. 67, 2521–2531.
237
7. Martin, P., Waters, N., Schmidt, C. J., Carlsson, A., and Carlsson, M. (1998) Rodent data and general hypothesis: Antipsychotic action exerted through 5-HT 2A receptor antagonism is dependent on increased serotonergic tone. J. Neural. Transm. 105, 65–396. 8. Peroutka, S. J. (1998) Serotonin receptor variants in disease: New therapeutic opportunities? Ann. NY Acad. Sci. 861, 16 –25. 9. Roth, B., Berr, S., Kroeze, W., Willins, D., and Kristiansen, K. (1998) Serotonin 5-HT2A receptors: Molecular biology and mechanisms of regulation. Crit. Rev. Neurobiol. 12, 319 –338. 10. Doyle, V., Creba, J., Ruegg, U., and Hoyer, D. (1986) Serotonin increases the production of inositol phosphates and mobilises calcium via the 5-HT 2 receptor in A7r5 smooth muscle cells. Naunyn-Schmiedeberg’s Arch. Pharmacol. 333, 98 –103. 11. Roth, B., Nakaki, T., Chuang, D., and Costa, E. (1984) Aortic recognition sites for serotonin (5-HT) are coupled to phospholipase C and modulate phosphatidyl inositol turnover. Neuropharmacology 23, 1223–1225. 12. Rizzuto, R., Simpson, A., Brini, M., and Pozzan, T. (1992) Rapid changes of mitochondrial Ca 2⫹ revealed by specifically targeted recombinant aequorin. Nature 358, 325–327. 13. Sheu, Y., Kricka, L., and Pritchett, D. (1993) Measurement of intracellular calcium using bioluminescent aequorin expressed in human cells. Anal. Biochem. 209, 343–347. 14. Brini, M., Marsault, R., Bastuianutto, C., Alvarez, J., Pozzan, T., and Rizzuto, R. (1995) Transfected aequorin in the measurement of cytosolic Ca 2⫹ concentration ([Ca 2⫹] c). J. Biol. Chem. 270, 9896 –9903. 15. Button, D., and Brownstein, M. (1993) Aequorin-expressing mammalian cell lines used to report Ca 2⫹ mobilisation. Cell Calcium 14, 663– 671. 16. Stables, J., Green, A., Marshall, F., Fraser, N., Knight, E., Sautel, M., Milligan, G., Lee, M., and Rees, S. (1997) A bioluminescent assay for agonist activity at potentially any G-protein coupled receptor. Anal. Biochem. 252, 115–126. 17. Shimomura, O. (1995) Cause of spectral variation in the luminescence of semisynthetic aequorins. Biochem. J. 306, 537–543. 18. Shimomura, O., Kishi, Y., and Inouyes, S. (1993) The relative rate of aequorin regeneration from apoaequorin and coelenterazine analogues. Biochem. J. 296, 549 –551. 19. Bonhaus, D., Weinhardt, K., Taylor, M., Desouza, A., McNeeley, P., Szczepanski, K., Fontana, D., Trinh, J., Rocha, C., Dawson, M., Flippin, L., and Eglen, R. (1997) RS-102221: A novel high affinity and selective 5-HT 2C receptor antagonist. Neuropharmacology 36, 621– 629. 20. Conklin, B., Chabre, O., Won, Y., Federman, A., and Bourne, H. (1992) Recombinant Gq␣. J. Biol. Chem. 267, 31–34. 21. Offermanns, S., and Simon, M. (1995) G␣ 15 and G␣ 16 couple a wide variety of receptors to phospholipase C. J. Biol. Chem. 270, 15175–15180. 22. Milligan, G., Marshall, F., and Rees, S. (1996) G 16 as a universal G protein adapter: Implications for agonist screening strategies. Trends Pharmacol. Sci. 17, 235–237. 23. Milligan, G., and Rees, S. (1999) Chimaeric G␣ proteins: Their potential use in drug discovery. Trends Pharmacol. Sci. 20, 118 –124.