Pharmacological Characterization of Membrane-Expressed hTAAR1 Bioluminescence Resonance Resonance Energy Transfer cAMP Biosensor

Pharmacological Characterization of Membrane-Expressed Human Trace Amine-Associated Receptor 1 (TAAR1) by a Bioluminescence Resonance Energy Transfer cAMP Biosensor

0026-895X/08/7403-585–594$20.00 M

OLECULAR

HARMACOLOGY

Pharmacological Characterization of Membrane-Expressed Human Trace Amine-Associated Receptor 1 (TAAR1) by a Bioluminescence Resonance Energy Transfer cAMP Biosensor

Larry S. Barak, Ali Salahpour, Xiaodong Zhang, Bernard Masri, Tatyana D. Sotnikova, Amy J. Ramsey, Jonathan D. Violin, Robert J. Lefkowitz, Marc G. Caron, and Raul R. Gainetdinov

Departments of Cell Biology (L.S.B., A.S., X.Z., B.M., T.D.S., A.J.R., M.G.C., R.R.G.) and Biochemistry (J.D.V., R.J.L.) and the Howard Hughes Medical Institute (J.D.V., R.J.L.), Duke University, Durham, North Carolina

Received May 15, 2008; accepted June 3, 2008

ABSTRACT Trace amines are neurotransmitters whose role in regulating

expression, and thus its pharmacological characterization with invertebrate physiology has been appreciated for many de-

a novel cAMP EPAC (exchange protein directly activated by cades. Recent studies indicate that trace amines may also play

cAMP) protein based bioluminescence resonance energy a role in mammalian physiology by binding to a novel family of

transfer (BRET) biosensor. We applied this novel cAMP BRET G protein-coupled receptors (GPCRs) that are found through-

biosensor to evaluate the activity of putative TAAR1 ligands. out the central nervous system. A major obstacle impeding the

This study represents the first comprehensive investigation of careful pharmacological characterization of trace amine asso-

the membrane-expressed human TAAR1 pharmacology. Our ciated receptors (TAARs) is their extremely poor membrane

strategy to express TAARs and to identify their ligands using a expression in model cell systems, and a molecular basis for this

cAMP BRET assay could provide a foundation for characteriz- phenomenon has not been determined. In the present study,

ing the functional role of trace amines in vivo and suggests a we show that the addition of an asparagine-linked glycosylation

strategy to apply to groups of poorly expressing GPCRs that site to the N terminus of the human trace amine associated

have remained difficult to investigate in model systems. receptor 1 (TAAR1) is sufficient to enable its plasma membrane

Trace amines (TAs) are biogenic amines that include tyra-

cluding schizophrenia and Parkinson’s disease. As such, the mine, tryptamine, octopamine, and ß-phenylethylamine (ß-

TAs offer tempting targets for ameliorating the symptoms of PEA). TAs are found in the mammalian central nervous

central nervous system diseases when used in combination system at substantially lower levels than such typical trans-

with standard therapies or for reducing doses and hence the mitters as norepinephrine, serotonin, and dopamine (Boul-

side effects that accompany the long-term use of some cur- ton, 1980; Premont et al., 2001; Branchek and Blackburn, 2003; Berry, 2004; Grandy, 2007). TA dysregulation has been linked to various psychiatric and neurological disorders, in-

rently prescribed medications. This is particularly relevant for Parkinson’s disease, where long-term

-DOPA therapy often produces drug-related dyskinesias (Hornykiewicz, 2002; Fahn, 2003; Mercuri and Bernardi, 2005).

This work was supported by National Institutes of Health grants NS19576 and 1U01-DA022950, and by the Michael J. Fox Foundation for Parkinson’s

585 Trace amines are structurally related to several psycho- tropic molecules that naturally occur in plants, including Research. A.S. was supported by a fellowship from Canadian Institutes of Health Research. B.M. is supported by a European Marie-Curie Outgoing International Fellowship (FP6-2005-Mobility-6).

amphetamine-like compounds. Amphetamines are recog- nized inhibitors of inward monoamine transport (Jones et al., L.S.B. and A.S. contributed equally to this work Article, publication date, and citation information http://molpharm.aspetjournals.org.

doi:10.1124/mol.108.048884.

can be found at

1998), and it was generally proposed that TAs also affect the monoamine system indirectly via interaction with plasma membrane transporters, such as the dopamine transporter

ABBREVIATIONS: TA, trace amine; ß-PEA, ß-phenylethylamine; GPCR, G-protein coupled receptor; TAAR, trace amine-associated receptor; ßTAAR1, human TAAR1 with the first nine amino acids of the human ß2-AR inserted at the N terminus; h, human; BRET, bioluminescence resonance energy transfer; EPAC, exchange protein directly activated by cAMP; HA, hemagglutinin; PCR, polymerase chain reaction; HEK, human embryonic kidney; PBS, phosphate-buffered saline; MEM, minimal essential medium; GFP, green fluorescent protein; ß2-AR, ß2-adrenergic receptor; MDMA, methylenedioxymethamphetamine; DA, dopamine; IBMX, 3-isobutyl-1-methylxanthine; MT, methoxytyramine.

(Sotnikova et al., 2004). In this view, the TAs function simply as neuromodulators of conventional mammalian brain amine transmitters, and at physiological levels, TAs generally should have only minor effects on neuronal excitability in the absence of conventional monoamines. However, binding sites for tryptamine, tyramine and ß-PEA have been reported in rat brain and a family of trace amine G-protein-coupled re- ceptors (GPCRs) has been recently identified in rodents and humans (Borowsky et al., 2001; Bunzow et al., 2001; Linde- mann and Hoener, 2005; Grandy, 2007). Thus solid support exists for a potential direct behavioral mechanism in which trace amines influence neuronal activity.

In humans, six trace amine-associated receptors (TAARs) and three pseudogenes have been identified (Lindemann and Hoener, 2005). Similar to their mouse counterparts on chro- mosome 10, human TAARs are coded for by intronless genes clustered tightly on chromosome 6. TAAR1 is the best char- acterized of them (Borowsky et al., 2001; Bunzow et al., 2001; Grandy, 2007). TAAR1 using cAMP, is activated couples to by TAs such G

s , as and ß-PEA when assessed in addition to metabolites of catecholamines, amphetamines (including MDMA), and other compounds known to affect monoaminer- gic transmission (Borowsky et al., 2001; Bunzow et al., 2001; Kim and von Zastrow, 2001; Scanlan et al., 2004; Lindemann and Hoener, 2005; Miller et al., 2005; Grandy, 2007; Reese et al., 2007; Wainscott et al., 2007; Wolinsky et al., 2007; Xie and Miller, 2007; Xie et al., 2007).

In the brain, the mRNA for TAAR1 is distributed through- out the limbic system and in regions containing cat- echolaminergic cell bodies and their projections (Borowsky et al., 2001; Bunzow et al., 2001; Lindemann and Hoener, 2005; Xie et al., 2007). Thus, TAAR1 is well positioned to modulate locomotor, emotional, and motivated behaviors that are tra- ditionally associated with monoaminergic activity. Recent evidence indicates that at least some TAARs can also serve as a new class of chemosensory receptors in mammals (Liberles and Buck, 2006). Because an imbalance in the func- tion of TAs might have important implications in the patho- genesis of several disorders, finding selective small-molecule agonists and antagonists for these receptors could provide novel approaches to disease management. However, a major obstacle in studying and screening TA receptors is their extremely poor plasma membrane expression in heterologous cell systems (Borowsky et al., 2001; Bunzow et al., 2001; Lindemann and Hoener, 2005; Miller et al., 2005; Grandy, 2007; Wainscott et al., 2007; Wolinsky et al., 2007), a sce- nario remarkably similar to that of the large GPCR subfam- ily of chemosensory odorant receptors (Saito et al., 2004; Zhuang and Matsunami, 2007).

The human variant of the TAAR1 has remained even more difficult to express in model systems than its poorly expressed rodent analog (Lindemann and Hoener, 2005; Grandy, 2007), thus impeding measurements of biochemical and pharmaco- logical properties of this receptor. To improve our ability to study hTAAR1, we investigated potential reasons for the poor hTAAR1 expression. In this report, we demonstrate that an N-glycosylated variant of the human TAAR1 expresses at the plasma membrane of HEK-293 cells well enough to be studied using antibody labeling. We have determined in this system that hTAAR1 can interact with ß-arrestin2. Moreover, we have used an N-glycosylated variant of the human TAAR1 for iden- tifying and studying TAAR1 ligands by constructing a novel

586 Barak et al.

cell-based screening assay that employs a bioluminescence res- onance energy transfer (BRET) exchange protein activated by cAMP (EPAC) cAMP biosensor.

Materials and Methods

Materials. Anti-HA antibody was from Roche Applied Sciences (Indianapolis, IN) and goat anti-mouse Alexa Fluor 568 antibody from Invitrogen (Carlsbad, CA). Plasmids containing the cDNA for the human trace amine receptor were obtained from the cDNA resource center at the University of Missouri-Rolla and the American Type Culture Collection (Manassas, VA). Sources of cell culture reagents and buffers were from Invitrogen and Sigma (St. Louis, MO), and fetal bovine serum from JRH Biosciences (Lenexa, KS). The enhanced green fluorescent protein plasmid was obtained from Clontech (Mountain View, CA). All compounds used in this study were obtained from Sigma.

Construction of Expression Vectors. A cDNA leader sequence containing the triple-HA epitope was constructed using the polymer- ase chain reaction for insertion into the plasmid pcDNA3. Full- length human TAAR1 cDNA without a stop codon was amplified by PCR with 5 and 3 in-frame restriction enzyme sites of EcoRI and KpnI, respectively. A cDNA encoding GFP in the absence of its initiation start codon was amplified by PCR with 5 (in-frame) and 3 (downstream of stop codon) restriction enzyme sites of KpnI and XbaI, respectively. Enzyme-digested TAAR1 and GFP PCR products were ligated into EcoRI and XbaI restriction enzyme sites of pcDNA3 vector with N-terminal triple HA tag to generate the triple HA- ß2N9-TAAR1-GFP construct. The cDNA corresponding to the first nine amino acids of the ß2-adrenergic receptor was inserted in-frame between the triple HA and TAAR1.

For the GFP-negative construct, the GFP moiety was removed from the above construct by PCR amplification using 5 T7 promoter sequence primer upstream of triple-HA sequence and 3 primer containing a stop codon followed by the restriction enzyme site of XbaI. The PCR product was then digested with HindIII (upstream of triple HA sequence) and XbaI, and subcloned into the pcDNA3 vec- tor. Two complementary sequences encoding the EcoRV/EcoRI-di- gested ß2N9 DNA fragment were allowed to anneal in buffer con- taining 10 mM Tris-EDTA, pH 8.0, and 50 mM NaCl and were ligated into EcoRV/EcoRI restriction enzyme sites. The cDNA illus- trative of these modifications is shown below; upper case characters correspond to the three triple-HA inserts, and the last segment corresponds to the ß2-adrenergic receptor. The corresponding cDNA sequence is: [tccaccatggctagctacccaTACGATGTTCCTGACTATG- CGggcTATCCCTATGACGTCCCGGACTATGCAggatccTAT- CCATATGACGTTCCAGATTACGCCagatctgatatcATGGGGCA- ACCCGGGAACGGCAGCGCCgaattcATGATGCCCTTTTGCC]. The translated sequence is: (s t m a s YPYDVPDY a g YPYD VPDY ags YPYDVPDY arsdi MGQPGNGSA e f M MPFC).

The BRET-based cyclic AMP biosensor (Ponsioen et al., 2004; Holz et al., 2006; Jiang et al., 2007) was constructed by the modification of ICUE2, an existing FRET-based intramolecular biosensor in which enhanced cyan fluorescent protein and citrine flank residues 149 to 881 of Epac1 (exchange proteins directly activated by cAMP) (DiPilato et al., 2004; Violin et al., 2008). The enhanced cyan fluo- rescent protein moiety of ICUE2 was removed by BamHI and KpnI restriction and replaced with a humanized Renilla reniformis lucif- erase gene that was PCR-amplified from phRluc-C1 (PerkinElmer Life and Analytical Sciences) and was engineered with BamHI and KpnI restriction sites that preserve the frame of translation. Primers for the PCR amplification of Rluc were 5 -GGATCCATGACCAG- CAAGGTGTACGACC and 5 -GGTACCCCCCCTGCTCGTTCTTCA- GCAC.

Cell Culture and Transfection of Cell Lines. Cells were cul- tured in MEM supplemented with 10% fetal bovine serum and 1:100

atmosphere. penicillin/streptomycin in a humidified 5% CO

Trans- fections were carried out by the calcium phosphate precipitation method. HEK-293T cells or U2OS cells were transfected with ali- quots containing 1 to 2 μg of ß-arrestin2 cDNA and/or 5 μg of receptor cDNA per milliliter. Cells were plated for immunofluores- cence studies into the wells of 35-mm glass-bottomed Matek dishes containing 2 ml of MEM/10% fetal bovine serum. Transfection solu- tion (125 μl) was added for a period lasting between 6 h and over- night, after which the transfection solution was replaced with fresh serum containing media. For BRET studies, the HEK-293T cells were transfected with 5 μg of cDNA in 1 ml of transfection solution in 100-mm plastic dishes overnight and selected to permanently express the EPAC sensor using phleomycin (Zeocin; 200–400 μg/ml). These cells were transfected as mentioned above with TAAR1 for evaluation of compounds in the screening assay.

Fluorescence Microscopy. Cells were plated at a density of 2 to 4 104/well in 35-mm Matek glass cover-slip–bottomed dishes. Anti-HA antibody was diluted 1:400. Alexa Fluor 568 goat anti- mouse antibody was used at a 1:1500 dilution. Fixed cells were first washed twice in PBS followed by incubation with 4% paraformalde- hyde for 15 min at room temperature, rewashed, and then incubated for 40-min intervals in PBS with 2% bovine serum albumin with either anti-HA or, subsequently, secondary antibody using an inter- vening three washes with PBS between antibody treatments. Live cells were incubated with the same antibody dilutions in MEM at room temperature or 37°C. ß-Arrestin translocation was assessed as described previously (Barak et al., 1997). Cells were examined with a Zeiss LSM-510 confocal microscope equipped with a 100 /numer- ical aperture 1.4 oil objective using excitation at 488 nm for the GFP tag and excitation at 568 nm for Alexa Fluor 568 antibody, and using the accompanying fluorescein and rhodamine emission filter sets, respectively.

Immunoblotting. HEK-293T cells were transfected with the var- ious HA-tagged receptor subtypes and lysed the following day in a buffer containing 20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, and 1% Triton X-100. Cell lysates were analyzed by SDS-polyacryl- amide gel electrophoresis and immunoblotted with anti-HA antibody at a dilution of 1:5000.

BRET Screening Assays. HEK-293T cells permanently trans- fected with the EPAC sensor and transiently with the TAAR1 were split into 96-well plates at 15 to 20 104 cells per well. On the following day, they were washed twice with PBS, and 80 μl of PBS containing calcium and magnesium was added to each well followed by addition of 10 μl of a 50 μM coelenterazine solution (final concen- tration, 5 μM). After 10-min incubation, either 10 μl of vehicle or 10 concentrated solution of drug in PBS was added, and the plate was then placed into a Mithras LB940 instrument (Berthold Tech- nologies, Bad Wildbad, Germany) that allowed the sequential inte- gration of the luminescent signals detected in the 465 to 505 nm and 505 to 555 nm windows using filters with the appropriate band pass and by using MicroWin 2000 software (Berthold Technologies). The BRET signal is determined by calculating the ratio of the light emitted at 505 to 555 nm to the light emitted at 465 to 505 nm.

Results

Expression of TAAR1 Receptors in Fixed HEK-293T Cells. Receptors of the human trace amine family have been notoriously difficult to express in model cell systems and are comparable in this regard with odorant receptors (Linde- mann and Hoener, 2005; Grandy, 2007; Zhuang and Matsu- nami, 2007). Our initial attempts at expression were also disappointing, but in the process of evaluating TAAR1 plasma membrane expression, we noted a marked absence of glycosylated bands that are common to other rhodopsin-fam- ily GPCRs. Moreover, in silico prediction suggested that the N terminus of TAAR1 should be poorly glycosylated. There-

Novel cAMP BRET Assay for GPCR Ligands 587

fore, in an attempt to augment TAAR1 glycosylation and the subsequent expression of the receptor at the plasma mem- brane (Fig. 1A), the first nine amino acids of the human ß2-adrenergic receptor (ß2-AR) were inserted between the HA tag and the N terminus of TAAR1, creating the HA- ßTAAR1 (note that an initial attempt to augment expression using the N-terminal rhodopsin signal sequence proved un- successful). This nine-residue segment contains an aspara- gine glycosylation site at position 6 (Rands et al., 1990). Moreover, for the ß2-AR, this glycosylation enhances recep- tor expression at the plasma membrane (Rands et al., 1990). Without the added insert, we observed for a TAAR1-GFP chimera a homogeneous intracellular GFP distribution con- sistent with degraded TAAR1 (Fig. 1B, top left) or receptor trapped in the endoplasmic reticulum (Fig. 1B, top right). With addition of the first nine amino acids of the human ß2-adrenergic receptor, we observed a substantial accumula- tion, at 24 h, of what appears to be plasma membrane- localized receptor at the edge of the cells, with loss of the majority of internal staining (Fig. 1B, bottom). Immunoblot- ting of cell lysates for HA antibody (Fig. 1C) demonstrates extended bands for both GFP and non-GFP receptor variants characteristic of glycosylated receptor that runs above sharp- er-well defined bands typically characteristic of the imma- ture receptor isoforms (Sadeghi et al., 1997; Barak et al., 2001). In addition, paraformaldehyde fixation without a per- meabilization step of the HA-tagged ßTAAR1 transfected cells indicates that the HA tag on the receptor is located at the external face of the plasma membrane (Fig. 1D).

Expression of TAAR1 in Live Cells. To estimate the expression level of the ßTAAR1 and verify its orientation and integrity in the plasma membrane, we performed live cell immunofluorescence labeling with anti-HA antibody. Figure 2A shows the relative expression of HA tagged ß2-AR (Fig. 2A, left) compared with the HA-ßTAAR1 GFP and non-GFP variants (Fig. 2A, middle and right). A more substantial expression of fluorescence associated with the ß2-AR is evi- dent, as is a smaller expression level for the non–ßTAAR1- GFP variant compared with the GFP version. The ability of GFP tagging to stabilize the expression of the ßTAAR1 was also apparent from the immunoblotting in Fig. 1B. Labeling N-terminal plasma membrane receptor in live cells with an- tibody can result in receptor aggregation, and this inhomo- geneous distribution can be used to verify the intactness of the receptor construct (Barak et al., 1997). Figure 2B shows that in nonpermeabilized live antibody-labeled cells, the C-terminal GFP distribution coincides with the HA epitope distribution, indicating that a fully formed full-length recep- tor is being expressed at the plasma membrane. Our ability to express ßTAAR1 at the plasma membrane is not simply confined to HEK cells, in that transient labeling of the cell edge by ßTAAR1 can also be demonstrated in a U2OS cell line (Fig. 2C). To demonstrate receptor functionality, we evaluated whether the ßTAAR1 could undergo desensitiza- tion to known agonists using a ß-arrestin2-GFP transloca- tion assay. However, because of the relatively smaller expres- sion levels of the ßTAAR1, we first aggregated the receptors using mouse monoclonal anti-HA and Alexa Fluor 568 goat anti-mouse antibodies. As seen in Fig. 2D, the aggregated ßTAAR1 in the presence of p-tyramine results in the trans- location of a small amount of ß-arrestin2-GFP, indicating that ßTAAR1 can recruit ß-arrestin2 similar to other GPCRs.

However, this modest translocation resembles that of the dopamine D3 receptor, which is known to desensitize rela- tively poorly in a ß-arrestin–dependent manner (Kim et al., 2001, 2005). In addition, the reduced translocation may also be indicative of the lesser expression of ßTAAR1 relative to the ß2-AR. Although the ß-arrestin translocation assay can be used to measure receptor activation, evaluate both ago- nists, and antagonists at many receptors, in this case, the weak response precludes using the direct visualization assay for this receptor variant. Therefore, we investigated a novel cAMP signaling sensor that could function well at these smaller but still greatly improved TAAR1 expression levels. BRET Sensor for Measuring TAAR1 in Real Time in Live Cells. cAMP production provides sufficient amplifica- tion to measure the functional activity of TAAR1 and so far has been the only practical way to approach the pharmaco- logical study of this very poorly expressed receptor. A signif- icant limitation of current methods to measure cAMP in cells for compound library screening is the extended activation period necessary to accumulate enough of the second mes- senger product to measure in antibody mediated assays. Here we evaluated a biosensor for real time screening of cAMP production in TAAR1-transfected cells. The cAMP

588 Barak et al.

Fig. 1. Expression of the TAAR1 in HEK-293 Cells. A, pictured are the two variants of the human trace amine receptor that were used throughout the study. Insertion of the nine-amino-acid proximal portion of the human ß2-adrenergic receptor into the N terminus of the TAAR1 was a key to stabilizing it at the plasma membrane. B, confocal micrograph of the fluorescence distribution of GFP in HEK-293 cells transfected with 5 μg of cDNA of the GFP variants of the TAAR1 without and with (ßTAAR1) the ß2-adrenergic receptor insert. C, Western Blot for the HA epitope of the triple HA-tagged, ß2-adrenergic receptor (amino acids 1–9) modified TAAR1 variants transiently expressed in HEK-293 cells. Note that the right image represents the same blot with a lower exposure period. D, immunostaining using mouse anti-HA antibody followed by goat anti-mouse Alexa Fluor 568 secondary antibody of fixed, nonpermeabilized HEK cells transiently expressing either the ßTAAR1 (top), or the ßTAAR1 conjugated to GFP (bottom).

BRET biosensor was generated by modification of the origi- nal ICUE2 cAMP FRET biosensor (Violin et al., 2008). The sensor consists of an N-terminal truncated variant of the EPAC tagged with the R. reniformis luciferase and a yellow fluorescent protein variant (citrine) attached at the N and C termini, respectively (Fig. 3A). Upon binding cAMP, the sig- nal of the biosensor decreases because of a conformational change that presumably increases the distance between the R. reniformis luciferase donor and the yellow fluorescent protein acceptor. The utility of the biosensor for measuring ßTAAR1 cAMP signaling in real time is demonstrated in Fig. 3B, where application of 1 μM ß-PEA, a TAAR1 agonist, produces a sustained reduction in the BRET signal for ßTARR1 (wild-type TAAR1 did not produce changes in the BRET signal; data not shown). Moreover, the addition of IBMX to the incubation media resulted in a significant in- crease of the BRET signal to ß-PEA (Fig. 3B) and isoproter- enol (not shown), without changing the EC

values of re- sponses to either ligands (data not shown). The enhancement of the BRET signal with IBMX is consistent with its role as a phosphodiesterase inhibitor that prevents cAMP degrada- tion. Figure 3C shows that, by increasing the ß-PEA concen- tration, a concomitant increase in the ßTAAR1 BRET signal

occurs and that the resulting signal amplitude remains sta- ble for at least 20 min after induction of cAMP. In contrast, in these HEK-293 cells isoproterenol stimulation of the endog- enous ß2-AR produces a transient change in the BRET signal that is maximal at 400 s and then returns slowly toward baseline (Fig. 3D), an indication of a desensitization of the signaling apparatus and a reduction in cAMP concentration (Violin et al., 2008).

Taken together, the ß-arrestin/ßTAAR1 translocation and the cAMP BRET data indicate that the ßTAAR1 has poorer affinity for ß-arrestin2 than does the ß2-AR. To confirm a ßTAAR1/ß-arrestin desensitization response by BRET, we transiently overexpressed both proteins in the same cell (Fig. 3E). In the presence of overexpressed ß-arrestin2, a return to baseline (i.e., desensitization) of the ß-PEA-mediated BRET signal, as well as a reduction in the magnitude of the BRET signal, is apparent after 5 min.

To demonstrate the general utility of the EPAC BRET sensor for determining cAMP concentrations, we cotrans- fected the sensor with the G

-coupled D2 dopamine (DA) receptor, expecting that prestimulation of the cells by DA would dampen their cAMP responsiveness. Figure 3F dem- onstrates that increasing concentrations of DA diminish the magnitude of an isoproterenol-induced BRET signal toward control levels without significantly changing the time it takes the signal to desensitize.

Specificity of BRET Signaling of ßTAAR1 in HEK-293 Cells. The sensitivity of the BRET EPAC assay enables measurement of the signaling of endogenous receptors (Fig.

Novel cAMP BRET Assay for GPCR Ligands 589

Fig. 2. Expression levels, distribution, and function of the ßTAAR1 in live HEK cells. A, immunofluorescence and transmitted light images of Alexa Fluor 568 immuno-stained HA-ß2AR (leftmost images), ßTAAR1-GFP (middle images), and ßTAAR1 (rightmost images). Images using live HEK-293 cells expressing labeled receptors were acquired on a Zeiss LSM 510 confocal microscope using a 100 /1.4 numerical aperture oil objective. B, overlay of Alexa Fluor 568, anti-HA immunostaining, and GFP imaging of the ßTAAR1-GFP receptor. Images were acquired as above in live HEK cells plated and transfected in 35-mm glass-bottomed dishes (Matek). C, imaging of ßTAAR1-GFP receptor in live U2OS cells transfected using calcium phosphate protocol with ßTAAR1-GFP receptor as in B. Receptor can be seen as enhanced fluorescence at the edge of the cells. D, translocation of ß-arrestin2-GFP was measured in cells transfected with ß2AR (leftmost images) or the ßTAAR1 (rightmost images). Live cells were imaged after labeling with monoclonal anti-HA antibody followed by Alexa Fluor 568 goat anti-mouse antibody (top), and treatment for 30 min at 37°C with either 20 μM isoproterenol (ß2AR) or 20 to 50 μM p-tyramine for the ßTAAR1. ß-Arrestin2-GFP translocation is shown at the bottom (green).

3D). To verify the absence of TAAR1 signaling in HEK/EPAC cells not containing transfected ßTAAR1 cDNA, we compared the isoproterenol and ß-PEA response of non–receptor-trans- fected (mock) cells to cells transfected with ßTAAR1 cDNA. Isoproterenol signaling was the same in both cell populations (Fig. 4A), whereas only the ßTAAR1-transfected cells pro- duced an EPAC signal significantly above baseline after ß-PEA stimulation. The very small rise in the ß-PEA signal in mock cells at high agonist concentrations may be due to activation of endogenous adrenergic receptors.

Activation of the ßTAAR1 by TAAR1 Agonists ß-PEA and d-Amphetamine. The endogenous trace amine ß-PEA is the best established agonist of TAAR1 (Borowsky et al., 2001; Bunzow et al., 2001; Lindemann and Hoener, 2005; Miller et al., 2005; Grandy, 2007; Wainscott et al., 2007; Wolinsky et al., 2007). The effects of amphetamine on behav- ior have historically been attributed to the release of internal stores of dopamine (Jones et al., 1998). However, recent ob- servations indicated that amphetamine derivatives can also directly activate TAAR1 (Bunzow et al., 2001; Miller et al., 2005; Grandy, 2007; Reese et al., 2007; Wainscott et al., 2007; Wolinsky et al., 2007; Xie and Miller, 2007; Xie et al., 2007). Therefore, by using the EPAC cell line, the ability of these two compounds to induce cAMP production through the ßTAAR1 was tested (Fig. 4B). Moreover, to confirm the re- sults, we also measured the cAMP response in HEK cells using the standard cAMP column assay developed by Salomon et al. (1974) (Fig. 4C). The results of the column and BRET measurements are shown in Table 1. They are quali-

590 Barak et al.

Fig. 3. Evaluation of cAMP response in HEK-293 cells using an EPAC BRET biosensor. A, diagram of the postulated molecular rearrangement of the full-length EPAC protein with and without cAMP present. B, time course of the BRET response computed as the ratio of YFP/Rluc emissions (see Materials and Methods) in HEK-293 cells transfected with the ßTAAR1 and permanently expressing the BRET biosensor (clone 8). To generate the ß-PEA curve, the cells were exposed to 10 μM compound in PBS containing calcium and magnesium (see Materials and Methods), and 5 μM coelenterazine at room temperature. The lower ß-PEA curve reflects the increased levels of cAMP in the presence of IBMX (0.5 mM). C, time course of the BRET sensor response to cAMP accumulation after exposure to various doses of ß-PEA in clone 8 cells expressing ßTAAR1 receptor. D, an analogous time course graph is generated for cAMP production secondary to isoproterenol exposure of endogenous ß2-adrenergic receptors in the clone 8 HEK cells. E, the desensitizing effects of ß-arrestin2 (ßarrest2) cotransfection on the cAMP response of 10 μM ß-PEA treated ßTAAR1 results in an upward deflection biosensor and of endogenous the BRET response curve. F, ß-adrenergic receptor. the Increasing G

i -coupled concentrations D2 dopamine receptor of was coexpressed in HEK-293 cells expressing the EPAC BRET dopamine preceding isoproterenol stimulation of the ß-adrenergic receptors results in decreased cAMP production and deflection of the BRET response upwards toward the measured basal cAMP response profile.

values tatively similar, with the BRET assay yielding EC

50 for cAMP activation that are shifted to the left by 3- to 4-fold. BRET Measurement of TAAR1 for Screening. The above results indicated that the TAAR1 BRET assay should be sensitive and specific enough for general screen- ing of TAAR1 ligands using a higher throughput format. Therefore, we measured the dose response of approxi- mately 40 more compounds that could be potential ligands of the TAAR1 in antagonist and agonist assays. Some examples of these studies are plotted in Fig. 4D. Although

Novel cAMP BRET Assay for GPCR Ligands 591

Fig. 4. EPAC response in HEK-293 cells with stimulation of endogenous and transfected receptors. A, clone 8 HEK-293 cells permanently expressing Rluc-EPAC-YFP and endogenous receptors only (mock) or expressing in addition transfected ßTAAR1 were exposed to differing concentrations of isoproterenol or ß-PEA for 5 and 15 min, respectively, at room temperature. Data are presented as the relative change S.E.M. from the baseline level BRET measurements, and results are the average of two (iso Mock) to four independent experiments. iso, isoproterenol. B, clone 8 cells expressing the ßTAAR1 were exposed to varying concentrations of ß-PEA and d-amphetamine for 15 min at room temperature. Results are presented as mean S.E.M. change from baseline BRET signaling and represent two independent experiments performed in duplicate. C, Dowex and Alumina column chromatography was used to measure [3H]cAMP accumulation in HEK-293 cells transfected with the ßTAAR1 receptor and treated with the concentrations of compounds shown in the Figure for 15 min at room temperature. Results are the mean S.E.M. of two (d-amphetamine) or three (ß-PEA) independent experiments performed in duplicate. D, example of TAAR1 BRET assay screening test for TAAR1 ligands. Twenty compounds were tested in this experiment at a single concentration (10 μM). Measurements were performed 10, 20, and 30 min after addition of compounds. b, vehicle baseline; compound 11, ß-PEA; compound 12, l-amphetamine, compound 18, d-methamphetamine. Stippled area around the baseline represents 95% confidence interval for vehicle response. On the basis of the data presented in this report, we estimated Z-score as 0.6.

among these compounds we detected no antagonist [based on their ability to counteract a ß-PEA-induced BRET sig- nal (data not shown)], we found several compounds that induced relatively potent agonist activity (Table 2) sup- porting previous reports on activity of these compounds at TAAR1 receptors (Bunzow et al., 2001; Miller et al., 2005; Reese et al., 2007; Wainscott et al., 2007; Wolinsky et al., 2007; Xie and Miller, 2007; Xie et al., 2007). A list of compounds that exerted no activity at ßTAAR1 in the BRET assay is presented in Table 3.

different species (Lindemann and Hoener, 2005; Grandy, 2007; Wainscott et al., 2007). Accordingly, we observed that although these compounds were active at human TAAR1 with potencies generally similar to that observed in rodent, Rhesus monkey (Macaca mulatta) and chimeric human-ro- dent receptors (Bunzow et al., 2001; Lindemann and Hoener, 2005; Grandy, 2007; Wainscott et al., 2007; Wolinsky et al., 2007; Xie and Miller, 2007; Xie et al., 2007), several important differences were also noted. In particular, the most efficacious compound at human TAAR1 appeared to be l-amphetamine, followed by d-methamphetamine, with activities at least equal to or exceeding that of ß-PEA. It is interesting also that trace amines octopamine and tryptamine showed partial agonist ac- tivity at human TAAR1 with potencies that were somewhat higher than those observed in previous studies with human-rat chimeric receptors (Lindemann and Hoener, 2005; Grandy, 2007; Wainscott et al., 2007). Finally, these studies also con- firmed activity of catecholamine metabolites 3-MT and 4-MT at human TAAR1 (Bunzow et al., 2001; Grandy, 2007; Wainscott et al., 2007), raising an intriguing question on physiological roles that may be mediated by these metabolites via activation of TAAR1 receptors. By assessing the capability of compounds to reduce ß-PEA-induced BRET signaling, we also sought to identify antagonists of TAAR1, but in a small list of compounds tested, no such activity was found. Tests of larger libraries of compounds for TAAR1 activity by using this assay are clearly warranted and are being performed currently in our lab.

It is noteworthy that the BRET method has been used for high-throughput screening of chemokine receptor CCR5 an- tagonists using a variant of a BRET-sensitive ß-arrestin2 recruitment assay (Hamdan et al., 2005). During the course of this work, a similar BRET sensor for cAMP, CAMYEL (cAMP sensor using YFP-Epac-RLuc), was developed inde- pendently for characterizing cAMP synthesis stimulated via a sphingosine 1-phosphate/G13 pathway (Jiang et al., 2007). Moreover, the modified EPAC cAMP sensor, described here, has shown recently great utility to monitor dynamics of ß2AR desensitization in live cells by fluorescence resonance energy transfer (Violin et al., 2008). The BRET data in our study indicate that the TAAR1-induced cellular concentration of cAMP is stable over an extended period. In contrast, under similar assay conditions, ß2AR cAMP production decreases relatively rapidly presumably as a result of more pronounced receptor desensitization (Jiang et al., 2007; Violin et al., 2008). The TAAR1 behaves in this regard similarly to the dopamine D3 receptor, which also shows no discernable de- sensitization in HEK 293 cells (Kim et al., 2001, 2005).

In conclusion, we expressed human TAAR1 receptor in HEK cells at the plasma membrane at levels enabling bio- logic characterization and the development of a BRET sig- naling cellular assay for large-scale screening of TAAR1

TABLE 1 Comparison of activity of ß-PEA and d-amphetamine at TAAR1 as measured by column cAMP assay and BRET assay Values are estimated from experiments presented in Fig. 4, B and C.

Compounds

Discussion

Trace amine-associated receptors, like olfactory receptors, are notoriously difficult to express in heterologous cell sys- tems. This poor cell surface expression has been a major limitation for understanding TAAR1 biology and evaluating TAAR1 selective ligands. In fact, up to this point, there has been no compelling direct evidence in the literature to sug- gest that the intact human TAAR1 could be expressed at sufficient amounts in a cell system, particularly at the plasma membrane compartment (Borowsky et al., 2001; Bunzow et al., 2001; Scanlan et al., 2004; Lindemann and Hoener, 2005; Miller et al., 2005; Grandy, 2007; Reese et al., 2007; Wain- scott et al., 2007; Wolinsky et al., 2007; Xie and Miller, 2007; Xie et al., 2007). Our data with HA-tagged and GFP-modified human TAAR1 provides evidence that glycosylation-stabilized expression of the receptor can occur in cells at levels sufficient for characterizing receptor biology and developing reliable in vitro cellular assays without modifying intracellular loops, developing human-rat chimeras (Lindemann and Hoener, 2005; Grandy, 2007), or coexpressing human TAAR1 with rat Gα

(Wainscott et al., 2007). Furthermore, we devised and validated a novel BRET cAMP biosensor cellular assay that permits high-throughput screening for TAAR1 agonists and antagonists. Although ßTAAR1 expression was 5- to 10-fold lower than what can be achieved with the ß2-adrenergic receptor, we observed that an N-ß2-AR peptide modified TAAR1 enables observable plasma membrane expression for which limited ß-arrestin translocation occurs.

The ability of the human TAAR1 to activate adenylyl cyclase with previously identified agonists was assessed by standard methods using column chromatography and by a novel BRET method employing an EPAC biosensor. The two distinct tech- niques were in agreement that human TAAR1 in HEK cells responds to ß-PEA and d-amphetamine, similar to what has been reported for rodent receptors or rodent modified human variants (Borowsky et al., 2001; Bunzow et al., 2001; Linde- mann and Hoener, 2005; Miller et al., 2005; Reese et al., 2007; Wainscott et al., 2007; Wolinsky et al., 2007). Although the measurement of cAMP using chromatography is not well suited to high-throughput screening, we observed that the BRET method described here can be applied to large-scale screening for GPCR ligands. In fact, in our initial BRET cAMP screening, we confirmed agonistic activity at human TAAR1 of several other compounds, including the trace amines octopamine and tryptamine, the amphetamine derivatives l-amphetamine, d-methamphetamine, ( )-MDMA, and phentermine, and the catecholamine metabolites 3-MT and 4-MT (Bunzow et al., 2001; Lindemann and Hoener, 2005; Reese et al., 2007; Wain- scott et al., 2007; Wolinsky et al., 2007; Xie and Miller, 2007; Xie et al., 2007). It has been noted previously that activity of these compounds at TAAR1 can differ significantly among

Column Assay BRET Assay

Efficacy (Proportion of ß-PEA)

M % M % ß-PEA 5.8 0.58 10 7 100 1.08 0.29 10 7 100 d-Amphetamine 9.3 2.79 10 7 73.9 12.6 1.36 0.14 10 7 62.6 1.7

592 Barak et al.

ligands. Identification of selective ligands of the TAAR1 would be critical for investigating the functional role of this receptor in mammalian physiology in vivo and/or managing human disorders in which abnormalities in TA physiology may occur. The use of the BRET EPAC biosensor described here may become an effective approach to identify selective agonists and antagonists of a wide range of GPCRs in gen- eral. Compared with other available cAMP assays, the de- scribed BRET assay provides a user-friendly opportunity for real-time dynamic assessment of cAMP production rather than single-endpoint analysis of total accumulated cAMP. In addition, the strategy applied to express TAAR1 in HEK cells may be applicable to other GPCRs, such as olfactory recep- tors that exhibit very limited expression at the plasma mem- brane in model systems.

TABLE 2 List of TAAR1 agonists identified in BRET assay Dose-response curves were generated as in Fig. 4B by measuring responses from 10 4

to 10

M for each compound. Results of at least two independent experiments are presented. Note that no response was observed in mock cells not transfected with ßTAAR1.

Compounds EC

Efficacy (Proportion of ß-PEA)

M % p-Tyramine 3.24 0.39 10 7 101.7 2.3 Octopamine 1.66 0.74 10 6 63.6 1.9 Tryptamine 1.95 0.49 10 6 82.6 4.7 l-Amphetamine 2.45 0.14 10 7 127.9 4.1 ( )-MDMA 3.70 0.54 10 7 89.6 6.3 d-Methamphetamine 1.31 0.19 10 6 107.7 5.1 Phentermine 1.08 0.04 10 6 87.7 1.1 3-MT 2.81 0.24 10 6 63.7 5.4 4-MT 2.87 0.04 10 6 74.4 3.74

TABLE 3 Compounds that did not show activity (agonist or antagonist) at TAAR1 in EPAC BRET assay Dose-response curves were generated as in Fig. 4B by measuring responses from 10

to 10

M for each compound. In antagonist assay, the ability of compounds to counteract BRET signal induced by 10 μM ß-PEA was evaluated. At least two tests were performed.

Compounds

-DOPA D

-DOPA Bupropion Spermidine Spermine Haloperidol α-Methyl-serotonin Morphine Naltrexone Naloxone Methadone L

-Tyrosine Methylphenidate Phentolamine 2-Methyl-6-(phenylethynyl)-pyridine Zonisamide Lamotrigine Modafinil D

-Phenylalanine 5-Methoxy-N,N-dimethyltryptamine 3-Methoxy-

-tyrosine L

-threo-3,4-Dihydroxyphenylserine Pergolide Propranolol Atropine Capsaicin Capsazepine

Acknowledgments

We thank Ava Sweeney, Katherine Clark, and Yushi Bai for excellent technical assistance in performing these experiments.

References Barak LS, Ferguson SS, Zhang J, and Caron MG (1997) A ß-arrestin/green fluores- cent protein biosensor for detecting G protein-coupled receptor activation. J Biol Chem 272:27497–27500. Barak LS, Oakley RH, Laporte SA, and Caron MG (2001) Constitutive arrestin- mediated desensitization of a human vasopressin receptor mutant associated with nephrogenic diabetes insipidus. Proc Natl Acad Sci U S A 98:93–98. Berry MD (2004) Mammalian central nervous system trace amines. Pharmacologic

amphetamines, physiologic neuromodulators. J Neurochem 90:257–271. Borowsky B, Adham N, Jones KA, Raddatz R, Artymyshyn R, Ogozalek KL, Durkin MM, Lakhlani PP, Bonini JA, Pathirana S, et al. (2001) Trace amines: identifica- tion of a family of mammalian G protein-coupled receptors. Proc Natl Acad Sci USA 98:8966–8971. Boulton AA (1980) Trace amines and mental disorders. Can J Neurol Sci 7:261–263. Branchek TA and Blackburn TP (2003) Trace amine receptors as targets for novel

therapeutics: legend, myth and fact. Curr Opin Pharmacol 3:90–97. Bunzow JR, Sonders MS, Arttamangkul S, Harrison LM, Zhang G, Quigley DI, Darland T, Suchland KL, Pasumamula S, Kennedy JL, et al. (2001) Amphetamine, 3,4-methylenedioxymethamphetamine, lysergic acid diethylamide, and metabo- lites of the catecholamine neurotransmitters are agonists of a rat trace amine receptor. Mol Pharmacol 60:1181–1188. DiPilato LM, Cheng X, and Zhang J (2004) Fluorescent indicators of cAMP and Epac activation reveal differential dynamics of cAMP signaling within discrete subcel- lular compartments. Proc Natl Acad Sci U S A 101:16513–16518. Fahn S (2003) Description of Parkinson’s disease as a clinical syndrome. Ann N Y

Acad Sci 991:1–14. Grandy DK (2007) Trace amine-associated receptor 1-Family archetype or icono-

clast? Pharmacol Ther 116:355–390. Hamdan FF, Audet M, Garneau P, Pelletier J, and Bouvier M (2005) High- throughput screening of G protein-coupled receptor antagonists using a biolumi- nescence resonance energy transfer 1-based beta-arrestin2 recruitment assay. J Biomol Screen 10:463–475. Holz GG, Kang G, Harbeck M, Roe MW, and Chepurny OG (2006) Cell physiology of

cAMP sensor Epac. J Physiol 577:5–15. Hornykiewicz O (2002)

-DOPA: from a biologically inactive amino acid to a success- ful therapeutic agent. Amino Acids 23:65–70. Jiang LI, Collins J, Davis R, Lin KM, DeCamp D, Roach T, Hsueh R, Rebres RA, Ross EM, Taussig R, et al. (2007) Use of a cAMP BRET sensor to characterize a novel regulation 282:10576–10584.

of cAMP by the sphingosine 1-phosphate/G

pathway. J Biol Chem

Jones SR, Gainetdinov RR, Wightman RM, and Caron MG (1998) Mechanisms of amphetamine action revealed in mice lacking the dopamine transporter. J Neu- rosci 18:1979–1986. Kim KA and von Zastrow M (2001) Old drugs learn new tricks: insights from

mammalian trace amine receptors. Mol Pharmacol 60:1165–1167. Kim KM, Gainetdinov RR, Laporte SA, Caron MG, and Barak LS (2005) G protein- coupled receptor kinase regulates dopamine D3 receptor signaling by modulating the stability of a receptor-filamin-ß-arrestin complex. A case of autoreceptor reg- ulation. J Biol Chem 280:12774–12780. Kim KM, Valenzano KJ, Robinson SR, Yao WD, Barak LS, and Caron MG (2001) Differential regulation of the dopamine D2 and D3 receptors by G protein-coupled receptor kinases and ß-arrestins. J Biol Chem 276:37409–37414. Liberles SD and Buck LB (2006) A second class of chemosensory receptors in the

olfactory epithelium. Nature 442:645–650. Lindemann L and Hoener MC (2005) A renaissance in trace amines inspired by a

novel GPCR family. Trends Pharmacol Sci 26:274–281. Mercuri NB and Bernardi G (2005) The ‘magic’ of

-dopa: why is it the gold standard Parkinson’s disease therapy? Trends Pharmacol Sci 26:341–344. Miller GM, Verrico CD, Jassen A, Konar M, Yang H, Panas H, Bahn M, Johnson R, and Madras BK (2005) Primate trace amine receptor 1 modulation by the dopa- mine transporter. J Pharmacol Exp Ther 313:983–994. Ponsioen B, Zhao J, Riedl J, Zwartkruis F, van der Krogt G, Zaccolo M, Moolenaar WH, Bos JL, and Jalink K (2004) Detecting cAMP-induced Epac activation by fluorescence resonance energy transfer: Epac as a novel cAMP indicator. EMBO Rep 5:1176–1180. Premont RT, Gainetdinov RR, and Caron MG (2001) Following the trace of elusive

amines. Proc Natl Acad Sci U S A 98:9474–9475. Rands E, Candelore MR, Cheung AH, Hill WS, Strader CD, and Dixon RA (1990) Mutational analysis of ß-adrenergic receptor glycosylation. J Biol Chem 265: 10759–10764. Reese EA, Bunzow JR, Arttamangkul S, Sonders MS, and Grandy DK (2007) Trace amine-associated receptor 1 displays species-dependent stereoselectivity for iso- mers of methamphetamine, amphetamine, and para-hydroxyamphetamine. J Pharmacol Exp Ther 321:178–186. Sadeghi HM, Innamorati G, and Birnbaumer M (1997) Maturation of receptor proteins in eukaryotic expression systems. J Recept Signal Transduct Res 17:433– 445. Saito H, Kubota M, Roberts RW, Chi Q, and Matsunami H (2004) RTP family members induce functional expression of mammalian odorant receptors. Cell 119: 679–691. Salomon Y, Londos C, and Rodbell M (1974) A highly sensitive adenylate cyclase

assay. Anal Biochem 58:541–548. Scanlan TS, Suchland KL, Hart ME, Chiellini G, Huang Y, Kruzich PJ, Frascarelli S, Crossley DA, Bunzow JR, Ronca-Testoni S, et al. (2004) 3-Iodothyronamine is an endogenous and rapid-acting derivative of thyroid hormone. Nat Med 10:638–642.

Novel cAMP BRET Assay for GPCR Ligands 593

Sotnikova TD, Budygin EA, Jones SR, Dykstra LA, Caron MG, and Gainetdinov RR (2004) Dopamine transporter-dependent and -independent actions of trace amine beta-phenylethylamine. J Neurochem 91:362–373. Violin JD, DiPilato LM, Yildirim N, Elston TC, Zhang J, and Lefkowitz RJ (2008)

ß modeling 2

-Adrenergic of real-time receptor camp signaling dynamics. and desensitization J Biol Chem 283:2949–2961.

elucidated by quantitative

Wainscott DB, Little SP, Yin T, Tu Y, Rocco VP, He JX, and Nelson DL (2007) Pharmacologic characterization of the cloned human trace amine-associated re- ceptor1 (TAAR1) and evidence for species differences with the rat TAAR1. J Phar- macol Exp Ther 320:475–485. Wolinsky TD, Swanson CJ, Smith KE, Zhong H, Borowsky B, Seeman P, Branchek T, and Gerald CP (2007) The Trace Amine 1 receptor knockout mouse: an animal model with relevance to schizophrenia. Genes Brain Behav 6:628–639.

594 Barak et al.

Xie Z and Miller GM (2007) Trace amine-associated receptor 1 is a modulator of the

dopamine transporter. J Pharmacol Exp Ther 321:128–136. Xie Z, Westmoreland SV, Bahn ME, Chen GL, Yang H, Vallender EJ, Yao WD, Madras BK, and Miller GM (2007) Rhesus monkey trace amine-associated recep- tor 1 signaling: enhancement by monoamine transporters and attenuation by the D2 autoreceptor in vitro. J Pharmacol Exp Ther 321:116–127. Zhuang H and Matsunami H (2007) Synergism of accessory factors in functional

expression of mammalian odorant receptors. J Biol Chem 282:15284–15293.

Address correspondence to: Marc G. Caron, Box 3287, Duke University, Durham, NC 27710. E-mail: m.caron@cellbio.duke.edu

Page updated

Report abuse