pGL4.26[luc2/Hygro], pGL4.27[luc2P/Hygro], pGL4.28 [luc2CP/Hygro], pGL4.29[CRE/minP/luc2P], pGL4.30 [NFAT-RE/ minP/luc2P], and pF9A CMV hRluc-neo Flexi vectors, and PMA were from Promega (Madison, WI). All other reagents were from Sigma (St. Louis, MO).

The DNA sequences of five copies of the SRE or SRF-RE were synthesized by Integrated DNA Technologies, Inc. (Coralville, Iowa). SRE- and SRF-RE- reporter were constructed by insertion of SRE or SRF-RE into pGL4.26, pGL4.27, or pGL4.28 using Hind III and Bgl II restriction enzyme sites. The resulting constructs from pGL4.27 were named as pGL4.33 [SRE/luc2P] and pGL4.34[SRF-RE/ luc2P], respectively.

The human dopamine D1 receptor (accession number NM_000794) was obtained from American Type Culture Collection (Manassas, VA). The m₃ muscarinic receptor (NM_000740.2) was acquired from Origene (Rockville, MD). The coding sequences of both receptors were subcloned into pF9A CMV hRluc-neo Flexi Vector to make receptor-Renilla fusion proteins. The constructs for m₄ muscarinic receptor were from UMR cDNA Resource Center (Rolla, MO). All sequences were verified by restriction digest and sequencing.

HEK293 cells were cultured in DMEM (Life Technologies) supplemented with 10% fetal bovine serum at 37^oC with a humidified atmosphere at 5% CO2. Transient tranfections were done using TransIT®-LT1 Transfection Reagent following manufacturer’s protocol (Mirus Bio). Cells were directly plated and transfected in 96-well plate, or plated and transfected in T75 flask, then plated into 384-well plate 4-6 hours after transfection. The cells were changed to serum-starving medium 4-6 hours after transfection and assayed 20 hours later for SRE- or SRF-RE- reporter assay, or assayed 24 hours after transfection for CRE- or NFAT-RE- reporter assay.

Stable HEK293 cell lines expressing reporter vectors (CRE-, NFAT-RE, SRE-, SRF-RE-) were generated by transient transfection followed by hygromycin selection and limiting dilution cloning. Double stable cell lines co-expressing reporter vector and receptor (D₁ dopamine receptor and m₃ muscarinic receptor) were generated by transfection of receptor construct into single stable line expressing corresponding reporter vector. Cells were plated in complete medium for CRE- or NFAT-RE- assay, or plated in DMEM plus 2% charcoal/dextran treated FBS (Thermo Scientific) for SRE- or SRF-RE- assay 24 hours before assay.

Firefly and renilla luciferase activities, as indicated by relative luminescence units (RLU) were determined using One-Glo or Dual-Glo luciferase assay kits (Promega) according to the manufacturer's instructions. For agonist, fold of induction = firefly RLU_induced / firefly RLU_uninduced. For antagonist, % of control=100 × firefly RLU_{(agonist+antagonist)} / firefly RLU_{agonist alone}, all normalized to renilla RLU. Both EC₅₀ and IC₅₀ values were generated using GraphPad Prism software. Z’ values were determined as Z’=1- [(3×SD_inudced + 3×SD_uninduced) / (average_induced – average_uninduced)">(3×SD_inudced + 3×SD_uninduced) / (average_induced – average_uninduced)].

Previously, we have shown that the luciferase reporter vectors (pGL4 series) containing minimal promoter and destabilized luciferase gene improve the responsiveness of cAMP response element to G_s coupled receptor [8Fan F, Wood KV. Bioluminescent assays for high-throughput screening Assay Drug Dev Technol 2007; 5: 127-36.]. To evaluate if this is also applicable for other response elements which are specific for different signal pathways, we constructed SRE- and SRF-RE- reporter in various pGL4 vectors containing luciferase gene with or without protein degradation sequence hPEST (Pro-Glu-Ser-Thr) and CL1 [9Li X, Zhao X, Fang Y, et al. Generation of destabilized green fluorescent protein as a transcription reporter J Biol Chem 1998; 273: 34970-5., 10Gilon T, Chomsky O, Kulka RG. Degradation signals for ubiquitin system proteolysis in Saccharomyces cerevisiae EMBO J 1998; 17: 2759-66.] as shown in Fig. ( 2A). SRE is known to respond to ternary complex factor (TCF)-dependent ERK/MAPK pathway, while SRF-RE, a mutant form of SRE lacking TCF binding domain, is newly designed to respond to SRF-dependent and TCF-independent pathway such as RhoA activation [11Hill CS, Wynne J, Treisman R. The Rho family GTPases RhoA, Rac1, and CDC42Hs regulate transcriptional activation by SRF Cell 1995; 81: 1159-70.]. GPCRs, particularly those coupled with G_i and G_q activate ERK/MAPK pathway and induce transcriptional activation of SRE, while GPCRs coupled with G₁₂ family are known to activate Rho guanine nucleotide exchange factors (RhoGEFs) which leads to activation of RhoA and transcriptional activation of SRF-RE [12Hill CS, Treisman R. Differential activation of c-fos promoter elements by serum, lysophosphatidic acid, G proteins and polypeptide growth factors EMBO J 1995; 14: 5037-47.]. This is particularly important with increasing interests in HTS drug screening targeting G₁₂/RhoA pathway while there are no HTS-compatible methods available for G₁₂/RhoA so far.

Fig. (2) Destabilized luciferase genes increase the reporter response and reduce the assay time to reach maximum induction. luc2, firefly luciferase; luc2P, firefly luciferase with hPEST sequence; luc2CP, firefly luciferase with hCL1 and hPEST sequences. HEK293 cells transiently expressing various versions of SRE- or SRF-RE- reporters (luc2, luc2P or luc2CP) were serum starved, induced with 20% FBS plus 10ng/ml PMA for SRE (2B, 2D), or with 20% FBS for SRF-RE (2C, 2E) for 2-8 hours. Cells not induced after serum starvation were used as control (uninduced). Firefly luciferase activity (RLU) were determined using One-Glo assay system and measured on the GloMax 96 multiplate luminometer. (2B, 2C) Firefly luciferase activity (RLU) and fold of induction measured after six hours of induction. (2D, 2E) Time course of fold of induction of SRE- and SRF-RE measured every two hours for 8 hours after induction.

SRE-mediated gene transcription is known to be rapidly induced by serum, phorbol esters and growth factors in TCF and SRF-dependent manner, while SRF alone can mediate transcription activation of SRF-RE by serum mostly through the binding to cell surface receptors of the mitogenic agents such as LPA in serum [11Hill CS, Wynne J, Treisman R. The Rho family GTPases RhoA, Rac1, and CDC42Hs regulate transcriptional activation by SRF Cell 1995; 81: 1159-70., 12Hill CS, Treisman R. Differential activation of c-fos promoter elements by serum, lysophosphatidic acid, G proteins and polypeptide growth factors EMBO J 1995; 14: 5037-47.]. Thus, to establish our assay we first used serum plus PMA to stimulate SRE reporter, and used serum alone to simulate SRF-RE reporter. When various SRE- or SRF-RE reporters were transiently transfected into HEK293 cells, the luciferase reporters containing protein degradation sequence hPEST (luc2P), or both hPEST and CL1 (luc2CP) showed 5-10 times lower assay background due to decreased basal expression level of reporter proteins (Fig. 2B, 2C). The cells were further induced with FBS plus PMA (for SRE-) or FBS alone (for SRF-RE-) for different time points up to 24 hours (data not shown for 8+ hour time point). As shown in Fig. (2D and 2E), the destabilized SRE- and SRF-RE- reporters reached the maximum of response in a shorter time (2-6 hours) than traditional luciferase gene (8+ hours).

The long incubation time required by traditional reporter assays has limited the application of reporter technology in HTS drug screening due to the concerns of cytotoxicity by the compounds. Therefore, deployment of the destabilized reporter gene will remove this hurdle with improved response dynamics and reduced assay time (2-6 hours) which in turn could potentially minimize secondary effects (such as toxicity effect) arising from the prolonged incubation of cells with chemical compounds. Here we have shown that Luc2P versions of SRE and SRF-RE reached the peak response in six hours (Fig. 2D, 2E) with minimal sacrifice of basal reporter expression (Fig. 2B, 2C). Similar results were observed when we further measured several GPCR responses by comparing these SRE- or SRF-RE reporters with or without protein degradation sequences (data not shown), confirming the potentials of using destabilized luciferase reporters in GPCR drug screening. Thus, in consideration of both assay dynamics and signal output, we chose the reporter vectors with protein degradation sequence hPEST incorporated with luciferase, SRE-luc2P and SRF-RE-luc2P and six hour induction time for most of the following evaluation, unless otherwise indicated.

To determine if luciferase reporter assay is amenable for high-throughput screening for GPCR modulators linked with all major G protein signaling, we chose four individual receptors including D₁ dopamine receptor, m₄ muscarinic receptor, m₃ muscarinic receptor and endogenous EDG receptor, which are known to couple with G_s, G_i, G_q and G₁₂ subfamily respectively, and co-expressed them with the corresponding reporters. In the cases of G_s, G_q and G₁₂ signaling, luc2P version of reporters (CRE-luc2P, NFAT-RE-luc2P, SRF-RE-luc2P) were tested and shown to give maximal assay performance, while luc2CP of SRE reporter was used for detection of G_iβγ-linked MAP kinase activation in order to achieve optimized assay dynamics (Fig. 3, Table 1).

Fig. (3) GPCR assays for Gs, Gi, Gq, and G12 coupled receptors in 384-well format. HEK293 cells used are either stably expressing D1 dopamine receptor and CRE-luc2P (A), m3 muscarinic receptor with NFAT-RE-luc2P reporter (B), SRF-RE-luc2P reporter (test endogenous EDG receptor) (D), or transiently expressing m4 muscarinic receptor with SRE-luc2CP reporter (C). Cells were induced with various agonists in 384-well format. Agonists used (time of induction): 1 µM dopamine (4 hours); 1 µM muscarine chloride (8 hours); 10 µM acetylcholine (2 hours); 10 µM LPA (6 hours); Firefly luciferase activity was measured using One-Glo assay system. Fold of induction and Z’ values were determined as described in “Materials and Methodology”.

Table 1

Summary of Fold Change and Z’ Value of Four Different Receptors Known to Couple with Different G Protein Pathways, Measured by Individual Luciferase Reporter Assays

Cells expressing luciferase reporter and receptors were induced in a 384-well plate with various agonists for two to eight hours depending on the requirement of the response elements, RLUs were measured and Z’-factor values were then determined. Z’-factor is one of the major statistical parameters commonly used to evaluate assay performance in high throughput screening. According to the formula, the closer the Z’-factor is to 1, the better the assay quality is. In general, Z’ values greater than 0.5 are acceptable for HTS [13Zhang JH, Chung TD, Oldenburg KR. A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays J Biomol Screen 1999; 4: 67-73.]. As shown in Table T11, all four reporter assays showed Z’-factor values significantly higher than 0.5 in 384-well format with large response dynamic (fold of induction) range from 10 to 300 fold. Thus, luciferase reporter assays are robust, and can be used for high throughput screening of modulators for all major G protein pathways.

We next evaluate if luciferase reporter assay can be used to profile receptor/G protein coupling in HEK293 cells. It is known that the m₃ muscarinic receptor, when exogenously expressed, couples with both G_q and G₁₂ pathways, while endogenously expressed β₂-adrenergic receptors mostly activate G_s protein. When we overexpressed m3 muscarinic receptor in cells in four individual reporter assays, receptor activation by agonist carbachol induced dose-dependent expression of luciferase with distinguishable potencies (Fig. 4). As indicated in Table 2, NFAT-RE-, SRE- and SRF-RE assays showed similar EC₅₀ values by carbachol stimulation, suggesting that activation of m₃ muscarinic receptors leads to increase of intracellular calcium concentration, MAP kinase and RhoA activation with similar efficacies. In contrast, EC₅₀ values shown by CRE- reporter assay was almost three logs higher than those by other reporter assays tested, indicating that m₃ muscarinic receptors activate cAMP production in a much less efficient manner. In contrast, activation of β₂-adrenergic receptors by isoproterenol showed dose-dependent response in CRE-reporter assay, but not in NFAT-RE-, SRE- and SRF-RE- reporter assays, confirming that the major downstream signaling for β₂-adrenergic receptors is cAMP pathway via G_s protein. Thus, luciferase reporter technology is able to study G protein activation profiles by the same receptor in one single assay format.

Fig. (4) Profiling of receptor/G protein coupling. HEK293 cells transiently expressing various luciferase reporters as indicated: CRE-luc2P (A), NFAT-RE-luc2P (B), SRE-luc2P (C) or SRF-RE-luc2P (D) were induced with carbachol (for exogenously introduced m3 muscarinic receptor) or isoproterenol (for endogenous β2-adrenergic receptor) for six hours in 96-well format. Firefly luciferase activity was measured using One-Glo assay system.

Table 2

Comparison of EC₅₀ of Agonist-Induced Activation of m₃ Muscarinic Receptor or of Endogenous β₂-Adrenergic Receptor Measured by Four Luciferase Reporter Assays

Many GPCRs activate a variety of signaling pathways via distinct G proteins in cell type dependent manner, which is crucial for their precise regulation of tissue-specific physiological functions [14Castaño JP, Martínez-Fuentes AJ, Gutiérrez-Pascual E, et al. Intracellular signaling pathways activated by kisspeptins through GPR54: do multiple signals underlie function diversity? Peptides 2009; 30: 10-5.]. Unfortunately many of the current available screening methods have limitations and are unable to measure multiple GPCR signaling events in one assay format. Also, not all GPCR signaling such as RhoA activity can be measured in HTS format. Thus, despite significant interest in profiling for cell type dependent receptor/G protein coupling, it has been progressed rather slowly in current drug discovery practice, partly due to the burden setting up multiple assay platforms and lack of existing detection tools. The bioluminescent reporter assay using destabilized luciferase is fast, highly sensitive and easy to set up. More importantly, as we have shown here in Fig. (4), bioluminescent reporter assays can measure all major GPCR signaling via G_s, G_i, G_q and G₁₂ pathways in one single assay format with great assay dynamics and Z’ values, thereby they could be potentially used to monitor cell type-dependent signaling via G protein coupled receptors.

Since NFAT-RE- or SRE-luc2P reporters respond nicely to activation of m₃ muscarinic receptor (Fig.4), we next compared the potency ranking of a series of well-known compounds for m₃ muscarinic receptor using both reporter assays. As expected, co-transfection of either reporter vector with m₃ muscarinic receptor showed a large dynamic range of assay response upon agonist stimulation and both reporter assays showed identical orders of potency ranking for the agonists and antagonists we tested (Fig. 5). The order of potency rankings for agonists were: muscarine chloride>carbachol>pilocarpine; The IC_50s for antagonists were ranked as: scopolamine>pirenzipine, in the presence of muscarine chloride. Both potency rankings are consistent with those reported in the literature (Table 3). The potencies of same compound/receptor pairs attained from two reporter assays are all within a half-log range, suggesting none of the compounds showed functional selectivity (Fig. 5).

Fig. (5) Potency ranking of agonists and antagonists for m3 muscarinic receptor using NFAT-RE- and SRE- reporters. HEK293 cells were transiently transfected with NFAT-RE-lluc2Pl (A, B) or SRE-luc2P (C, D) along with m3 muscarinic receptor (M3R)-lRenillal fusion for twenty-four hours. Firefly luciferase activity was measured after 6 hours induction with addition of 1:3 serial dilutions of various agonists (A, C) or antagonists in the presence of 100 nM muscarine chloride (B, D) using the Dual-Glo assay system in 96-well format. Fold of induction and % of control were determined as described in “Materials and Methodology”.

Table 3

Comparison of EC₅₀ (Agonists) and IC₅₀ (Antago-nists) for m₃ Muscarinic Receptor by NFAT-RE- and SRE- Luciferase Reporter Assays

More evidence indicates that same receptor could activate various signaling pathways with distinct potencies depending on the modulators which could bind to the receptor orthosterically or allosterically. This new concept offers great potential for developing pathway-specific drugs that increase efficacy and reduce side effects, but at the same time add another dimension in assay complexity to drug discovery. In the case of m₃ muscarinic receptor, activation of m₃ muscarinic receptor increases intracellular Ca²⁺ mobilization and induces ERK/MAPK pathway at the same time, however, it is also reported that MAPK activity via m₃ muscarinic receptor is regulated in a calcium-dependent or independent manner in a variety cell types, suggesting the involvement of distinct signal mechanisms [15Slack BE. The m3 muscarinic acetylcholine receptor is coupled to mitogen-activated protein kinase via protein kinase C and epidermal growth factor receptor kinase Biochem J 2000; 348: 381-7.]. Therefore, by comparing the potencies of same compound/receptor pairs using multiple reporter vectors which measure different G protein downstream signaling, this approach has great potential in drug screening for GPCR modulators with functional selectivity in one simple assay format.