Anal. Chem. 2007, 79, 7534-7539
Detection of G Protein Coupled Receptor Mediated Adenylyl Cyclase Activity by Capillary Electrophoresis Using Fluorescently Labeled ATP Jennifer M. Cunliffe,† Roger K. Sunahara,‡ and Robert T. Kennedy*,†,‡
Department of Chemistry and Department of Pharmacology, University of Michigan, Ann Arbor, Michigan 48109
A capillary electrophoresis (CE) laser-induced fluorescence (LIF) assay was developed for the detection of G protein coupled receptor mediated adenylyl cyclase (AC) activity using BODIPY FL ATP (BATP) as substrate. In the assay, cell membranes coexpressing the stimulatory G protein fused to the β2 adrenergic receptor (β2AR) and AC were incubated with BATP, the resultant mixture injected, and BATP separated from product BODIPY FL cAMP (BcAMP) by CE. AC activity was quantified by measuring the rate of BcAMP formation. β2AR agonists isoproterenol and terbutaline increased basal AC activity with EC50s of 2.4 ( 0.2 and 60 ( 9 nM, respectively. The antagonist propranolol competed with terbutaline for β2AR binding sites and expectedly right-shifted the terbutaline dose-response curve to 8 ( 3 µM. The high sensitivity of the assay was demonstrated by detection of small changes in AC activity, with the partial agonist alprenolol increasing (22 ( 1%) and the inverse agonist ICI 118,551 decreasing (19 ( 2%) basal activity. The simplicity and automation of the CE-LIF assay offers advantages over the more traditional assay using radiochemical ATP and column chromatography. G protein coupled receptors (GPCRs) are the target for ∼50% of newly launched drugs, making them among the most important types of protein for pharmacological action1 As a result of the significance of GPCRs, interest remains in developing rapid, robust, and sensitive functional assays for use in uncovering novel drug candidates. Screening for GPCR-active compounds can be performed using assays that detect proximal binding or functional changes at the GPCR or G protein level;2-4 however, detection of downstream biochemical events of GPCR activation are often used because of the possibility of signal amplification.5-12 Furthermore, * Corresponding author. Phone: 734-615-4363. Fax: 734-615-6462. E-mail:
[email protected]. † Department of Chemistry. ‡ Department of Pharmacology. (1) Klabunde, T.; Hessler, G. Chembiochem 2002, 3, 928-944. (2) de Jong, L. A.; Uges, D. R.; Franke, J. P.; Bischoff, R. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2005, 829, 1-25. (3) Windh, R. T.; Manning, D. R. Methods Enzymol. 2002, 344, 3-14. (4) Ross, E. M. Methods Enzymol. 2002, 344, 601-617. (5) Chidiac, P.; Hebert, T. E.; Valiquette, M.; Dennis, M.; Bouvier, M. Mol. Pharmacol. 1994, 45, 490-499. (6) Clark, M. J.; Harrison, C.; Zhong, H.; Neubig, R. R.; Traynor, J. R. J. Biol. Chem. 2003, 278, 9418-9425.
7534 Analytical Chemistry, Vol. 79, No. 19, October 1, 2007
monitoring a downstream biochemical event may be more amenable to drug screening because, unlike GPCR binding assays, a known, labeled ligand is not required. This advantage is especially important when screening orphan receptors with no known endogenous ligands.13 Many GPCRs are coupled to GRs, a protein that potentiates adenylyl cyclase (AC) activity resulting in an increase in ATP to cAMP turnover when the GPCR is activated by an agonist. In such systems, detection of cAMP accumulation may be used in GPCR drug screening. Besides use in drug screening, detection of GPCR-mediated AC activity is also frequently used in fundamental biochemistry studies such as characterizing the signaling of GPCR mutations.14-16 In this paper, we describe a new assay for detecting GPCR-stimulated AC activity. AC activity is most commonly measured using [R32P]ATP as substrate and separating from product [R32P]cAMP by ionexchange column chromatography.17-19 This technique was invaluable in early AC studies due to its reliable performance, high sensitivity, and large dynamic range, but the radioactivity requirement and multiple washing/elution steps make it tedious and manually intensive. Since then, several AC assays have become commercially available (see recent reviews).11,20,21 Most of the newer assays use fluorescence or luminescence detection, but they utilize antibodies with limited shelf lives, filtering steps, or are (7) Ross, E. M.; Maguire, M. E.; Sturgill, T. W.; Biltonen, R. L.; Gilman, A. G. J. Biol. Chem. 1977, 252, 5761-5775. (8) Baker, J. G.; Hall, I. P.; Hill, S. J. Mol. Pharmacol. 2003, 64, 1357-1369. (9) Fitzsimons, C.; Engel, N.; Policastro, L.; Duran, H.; Molinari, B.; Rivera, E. Biochem. Pharmacol. 2002, 63, 1785-1796. (10) Seifert, R.; Wenzel-Seifert, K.; Lee, T. W.; Gether, U.; Sanders-Bush, E.; Kobilka, B. K. J. Biol. Chem. 1998, 273, 5109-5116. (11) Thomsen, W.; Frazer, J.; Unett, D. Curr. Opin. Biotechnol. 2005, 16, 655665. (12) Gonzalez, J. E.; Negulescu, P. A. Curr. Opin. Biotechnol. 1998, 9, 624631. (13) Milligan, G.; Feng, G. J.; Ward, R. J.; Sartania, N.; Ramsay, D.; McLean, A. J.; Carrillo, J. J. Curr. Pharm. Des. 2004, 10, 1989-2001. (14) Levin, M. C.; Marullo, S.; Muntaner, O.; Andersson, B.; Magnusson, Y. J. Biol. Chem. 2002, 277, 30429-30435. (15) Gosling, J.; Monteclaro, F. S.; Atchison, R. E.; Arai, H.; Tsou, C.-L.; Goldsmith, M. A.; Charo, I. F. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 5061-5066. (16) Green, S. A.; Cole, G.; Jacinto, M.; Innis, M.; Liggett, S. B. J. Biol. Chem. 1993, 268, 23116-23121. (17) Salomon, Y.; Londos, C.; Rodbell, M. Anal. Biochem. 1974, 58, 541-548. (18) Smigel, M. D. J. Biol. Chem. 1986, 261, 1976-1982. (19) Sunahara, R. K.; Dessauer, C. W.; Whisnant, R. E.; Kleuss, C.; Gilman, A. G. J. Biol. Chem. 1997, 272, 22265-22271. (20) Gabriel, D.; Vernier, M.; Pfeifer, M. J.; Dasen, B.; Tenaillon, L.; Bouhelal, R. Assay Drug Dev. Technol. 2003, 1, 291-303. (21) Williams, C. Nat. Rev. Drug Discovery 2004, 3, 125-135. 10.1021/ac071203+ CCC: $37.00
© 2007 American Chemical Society Published on Web 08/25/2007
mediated by a hormone-stimulated β2 adrenergic receptor. The ease-of-use, high sensitivity, and reproducibility of the assay suggests it has potential in GPCR drug screening.
Figure 1. Structure of BATP. Upon interaction with cell membrane, enzymatic products BcAMP, BAMP, and BADP are formed.
expensive. A simple, nonradiochemical, sensitive, and automatable assay amenable to high-throughput screening would be a valuable addition to the suite of techniques currently used to probe GPCRmediated AC activity. Capillary electrophoresis (CE) is well-suited for detecting enzyme activity due to its low sample volume requirements, high resolution, and rapid separations.22,23 We have previously reported a CE assay with UV detection for the detection of GPCR-mediated AC activity using ATP as substrate, but the relatively high detection limit of 1 µM for cAMP limits the applicability of the method for physiological samples and prevents adoption to faster, miniaturized columns.24 Fluorescent substrates are often used in CE-enzyme assays to achieve high sensitivity.25-29 We also previously demonstrated that BODIPY FL labeled ATP (BATP, Figure 1) could be used as a substrate for purified AC cyotsolic domains in a CE with laser-induced fluorescence (LIF) detection enzyme assay.30 BATP proved to be a good AC substrate, and the assay allows detection of both direct activation and inhibition with modulators as diverse as ligands, proteins, and cations. Because only purified AC was used, GPCR binding events could not be studied in this system. In this report, BATP was explored as a substrate for full-length, membrane-bound AC for the detection of GPCR-mediated AC activity using CE-LIF. GPCRs and AC are membrane-bound proteins, and cell membranes containing overexpressed GPCR, G protein, and AC were isolated from other cellular components for use in drug screening experiments. The physical proximity of the GPCR, G protein, and AC in the fluid cell membrane enabled GPCR binding events to be detected by monitoring AC activity. We demonstrate the detection of AC activity following direct activation of AC as well as activation (22) Zhang, J.; Hoogmartens, J.; Van Schepdael, A. Electrophoresis 2006, 27, 35-43. (23) Glatz, Z. J. Chromatogr., B 2006, 841, 23-37. (24) Cunliffe, J. M.; Whorton, M. R.; Sunahara, R. K.; Kennedy, R. T. Electrophoresis 2007, 28, 1913-1920. (25) Lorieau, J.; Shoemaker, G. K.; Palcic, M. M. Anal. Chem. 2003, 75, 63516354. (26) Shoemaker, G. K.; Lorieau, J.; Lau, L. H.; Gillmor, C. S.; Palcic, M. M. Anal. Chem. 2005, 77, 3132-3137. (27) Krylov, S. N.; Zhang, Z.; Chan, N. W.; Arriaga, E.; Palcic, M. M.; Dovichi, N. J. Cytometry 1999, 37, 14-20. (28) Meredith, G. D.; Sims, C. E.; Soughayer, J. S.; Allbritton, N. L. Nat. Biotechnol. 2000, 18, 309-312. (29) Jameson, E. E.; Cunliffe, J. M.; Neubig, R. R.; Sunahara, R. K.; Kennedy, R. T. Anal. Chem. 2003, 75, 4297-4304. (30) Cunliffe, J. M.; Sunahara, R. K.; Kennedy, R. T. Anal. Chem. 2006, 78, 1731-1738.
EXPERIMENTAL SECTION Chemicals. BODIPY FL 2′ (and 3′)-O-(N-(2- aminoethyl)urethane) ATP (BATP) was from Invitrogen (Carlsbad, CA). The 10× Tris-glycine buffer was purchased from Bio-Rad Laboratories (Hercules, CA), and Tris-HCl, sodium phosphate monobasic monohydrate, and MgCl2‚6H2O were purchased from Fisher (Fair Lawn, NJ). Ro 20-1724 was purchased from A. G. Scientific, Inc. (San Diego, CA). Morphine was from the Narcotic Drug and Opioid Peptide Basic Research Center at the University of Michigan. The seven select amino acids were as follows: glutamate, aspartate, glutamine, serine, glycine, taurine, γ-aminobutyric acid. The β2AR drugs (-)-isoproterenol, ICI 118,551, terbutaline, CGP 12177, norepinephrine, salbutamol, UK 14,304, (()-propranolol, (()-alprenolol, and all other chemicals were purchased from Sigma (St. Louis, MO). All solutions were prepared in deionized water purified by E-Pure water systems (Barnstead International Co., Dubuque, IA). Preparation of Sf9 Cell Membranes. The cell membranes were abstracted from other cellular components to study the membrane-bound GPCRs and AC. Spodoptera frugiperda (fall armyworm, Sf9) cells were infected with viruses containing the cDNAs for AC2 (AC, kindly provided by A. G. Gilman, University of Texas Southwestern Medical Center), the β2 adrenergic receptor fused to the long-splice variant of the stimulatory G protein (β2ARGRsL),31 or β2ARGRsL and AC2 (β2ARGRsL-AC), and incubated at 27 °C under rotation. Postinfection (48-72 h), cells were harvested in buffer containing 50 mM Tris-HCl pH 8.0, 150 mM NaCl, and protease inhibitors (35 mg/mL phenylmethanesulfonyl fluoride, 32 µg/mL N-tosyl-L-phenylalanine chloromethyl ketone, 32 µg/mL N-R-P-tosyl-L-lysine chloromethyl ketone, 3.2 µg/mL leupeptin, and 3.2 µg/mL soybean trypsin inhibitor). The above protease inhibitor concentrations were previously found to be suitable for AC assays in Sf9 cell membranes.31 Cells were lysed by nitrogen cavitation (Parr bomb, 600 psi, 30 min), followed by a low-speed spin (2000g, 10 min) to remove unlysed cells. Membranes were collected following ultracentrifugation (100000g, 35 min), resuspended in 50 mM Tris-HCl pH 8.0, 50 mM NaCl, and protease inhibitors (same as above), and the Bradford assay (Bio-Rad Laboratories, Hercules, CA) was used to determine protein concentrations. With the exception of the cell infection, all steps were performed at 4 °C. Aliquots were snap frozen in liquid nitrogen and stored at -80 °C. Thawed samples were homogenized (Kimble Kontes, Vineland, NJ) prior to use. Experiments were performed using at least two independent cell membrane preparations for each overexpression system. Sample Preparation for Adenylyl Cyclase Assays. Samples containing cell membrane (18-50 µg) and modulator were incubated on ice for 20-30 min in 10 mM Tris-HCl pH 8.0, 1 mM EDTA, and 5 mM MgCl2 with a final volume of 50 µL. (In cases where the modulator was a drug targeting β2ARGRsL, 50 µM GTP was also included.) Unless otherwise indicated, adenylyl cyclase activity was initiated by the addition of 50 µL of activation buffer (31) Vadakkadathmeethal, K.; Felczak, A.; Davignon, I.; Collins, J.; Sunahara, R. K. Insect Biochem. Mol. Biol. 2005, 35, 333-345.
Analytical Chemistry, Vol. 79, No. 19, October 1, 2007
7535
containing 20 mM Tris-HCl pH 8.0, 3 mM potassium phosphoenolpyruvate, 100 µM Ro 20-1724, 0.6 mM EDTA, 5 mM MgCl2, 50 µg/mL pyruvate kinase, 10 µM BATP, 400 nM fluorescein (used as internal standard, IS), and 1000 µM ATP. (In the 100-µL final volume, nucleotide concentrations were 5 µM BATP and 500 µM ATP.) Samples were incubated at 30 °C, and the reaction proceeded for 15-30 min before being quenched with 2 µL of 500 mM EDTA ([EDTA]final ) 10 mM). The autosampler was set at 4 °C except for kinetic studies where it was at 30 °C. CE-LIF. AC assays were performed using a P/ACE MDQ capillary electrophoresis unit (Beckman Coulter Inc., Fullerton, CA) equipped with a LIF detector. A third party Ar+ laser was connected to the P/ACE MDQ as previously described.32 Emission was detected after passing through a 488-nm notch filter and a 520 ( 10 nm band-pass filter. Data acquisition (16 Hz) and control were performed using P/ACE 32 Karat Software (Version 5.0, Beckman). Fused-silica capillaries (Polymicro Technologies, Phoenix, AZ) had a total length of 30 cm, length to the detector of 20 cm, inner diameter of 50 µm, and outer diameter of 360 µm. The separation temperature was maintained at 15 °C for all separations. Separation buffer was 60:40 (v/v) 25 mM Tris, 192 mM glycine pH 8.5/100 mM sodium phosphate, pH 7.1 (final pH of 7.8). At the start of the day, the capillary was rinsed with 0.1 M NaOH, water, and separation buffer for 5 min each. The capillary was rinsed with 0.1 M NaOH and separation buffer for 1 min each prior to each injection to remove membrane fragments from the capillary wall. Injections were at 0.5 psi for 3 s, and separations were at 500 V/cm for 6 min. Data Analysis. Peak areas were calculated using software written in-house.33 EC50 and IC50 values were determined by fitting data using GraphPad Prism (Version 3.0, San Diego, CA) to y ) bottom + ((top - bottom)/(1 + 10L°gEC50-x)) where x is the logarithmic concentration of modulator (in M) and y is AC activity (in nmol mg-1 min-1). BODIPY FL cAMP (BcAMP) was quantified using the calibration curve for BATP; this was considered acceptable because fluorescence was constant over the course of the reaction (i.e., the sum of products and substrate was constant throughout the reaction).30 AC activity was calculated by converting the concentration of BcAMP formed at a fixed time point into a rate (nmol min-1 mg-1, depending on incubation time and amount of protein used), and multiplying by the ratio of background ATP/2′-BATP. Significance was examined using the Student t-test or the Z′ factor.34 Error is reported as ( standard error of the mean. RESULTS AND DISCUSSION BATP as Substrate for Membrane-Bound AC. BATP exists as two isomers with the BODIPY label on the 2′- or 3′-O-ribosyl position (Figure 1) that are referred to as 2′-BATP and 3′-BATP, respectively. We have previously demonstrated that 2′-BATP is a substrate for purified cytosolic domains of AC, while 3′-BATP is not.30 The overall goal of these experiments was to determine whether BATP could be used as a substrate for AC in a more complex, physiologically relevant cell membrane environment and whether the assay could be used to detect GPCR-mediated AC activity. (32) Cunliffe, J. M.; Liu, Z.; Pawliszyn, J.; Kennedy, R. T. Electrophoresis 2004, 25, 2319-2325. (33) Shackman, J. G.; Watson, C. J.; Kennedy, R. T. J. Chromatogr., A 2004, 1040, 273-282. (34) Zhang, J. H.; Chung, T. D.; Oldenburg, K. R. J. Biomol. Screening 1999, 4, 67-73.
7536 Analytical Chemistry, Vol. 79, No. 19, October 1, 2007
Figure 2. Conversion of BATP to products. Samples contained 5 µM BATP, 36 µg of membrane, and 200 nM fluorescein (used as internal standard, IS). (A) Samples were analyzed after 4 min in the absence (trace 1) or presence (trace 2) of 36 µg of membrane. The inset highlights the formation of products BcAMP and BAMP. Trace 2 is offset for clarity. (B) Samples were serially injected, and BcAMP formation was monitored with time in the absence or presence of 500 µM ATP. Sample preparation and CE conditions are detailed in the Experimental Section.
Figure 2A illustrates that when BATP (migration time, tmig, ) 3.7 min) was incubated with cell membranes overexpressing AC (see Experimental Section), three products were formed corresponding to BcAMP (tmig ) 2.4 min), BODIPY FL AMP (BAMP; tmig ) 2.9 min), and BODIPY FL ADP (BADP; tmig ) 3.4 min). The peak identities of BcAMP, BAMP, and BADP were confirmed using enzymes known to produce cAMP (purified AC), AMP (apyrase), and ADP (hexokinase) from ATP. The formation of BAMP and BADP was the result of endogenous ATPases present in the membrane preparation and represents a significant complication over using purified enzymes described previously. Products BAMP and BADP each formed two peaks corresponding to the 2′- and 3′-isoforms, but only a single 2′-BcAMP peak was observed (Figure 2A, inset). This result agrees well with experiments using purified AC, suggesting that only one BATP isoform is a substrate for membrane-bound AC.30 Importantly for this assay, samples were not filtered prior to injection onto the capillary, drastically reducing the time and labor required for sample preparation compared to other AC assays requiring extensive washing steps.35 Furthermore, direct injection of membrane fragments did not adversely affect the separation in terms of extra peaks or capillary clogging, with migration time relative standard deviations (RSDs) being as low as 1% (n ) 6). The product BcAMP had a limit of detection of
