Extended bioluminescence resonance energy transfer (eBRET) for monitoring prolonged protein–protein interactions in live cells

Cellular Signalling 18 (2006) 1664 – 1670 www.elsevier.com/locate/cellsig

Extended bioluminescence resonance energy transfer (eBRET) for monitoring prolonged protein–protein interactions in live cells Kevin D.G. Pfleger a,⁎, Jasmin R. Dromey a , Matthew B. Dalrymple a,b , Esther M.L. Lim a , Walter G. Thomas c , Karin A. Eidne a a

7TM Laboratory/Laboratory for Molecular Endocrinology, Western Australian Institute for Medical Research (WAIMR) and Centre for Medical Research, University of Western Australia, Nedlands, Perth, WA 6009, Australia b Keogh Institute for Medical Research, QEII Medical Centre, Nedlands, Perth, WA 6009, Australia c Baker Heart Research Institute, Melbourne, VIC 3004, Australia Received 19 December 2005; accepted 11 January 2006 Available online 21 February 2006

Abstract Bioluminescence resonance energy transfer (BRET) is an increasingly popular technique for studying protein–protein interactions in live cells. It is particularly suitable for real-time monitoring of such interactions, however, the timescale over which assays can be carried out is currently relatively short (minutes) due to substrate instability. We present a new derivation of the BRET technology, termed ‘extended BRET’ (eBRET), which now enables protein–protein interactions to be monitored in real-time for many hours. This capability has significant benefits for investigating cellular function over extended timescales, as we have illustrated using the agonist-induced G-protein coupled receptor/β-arrestin interaction. The potential for studying the modulation of such interactions by agonists, antagonists, inhibitors, dominant negative mutants and coexpressed accessory proteins is substantial. Furthermore, the advantages of eBRET have important implications for the development of highthroughput BRET screening systems, an ever-expanding area of interest for the pharmaceutical industry. © 2006 Elsevier Inc. All rights reserved. Keywords: Bioluminescence resonance energy transfer; BRET; Protein–protein interaction; G-protein coupled receptor; β-arrestin

1. Introduction Bioluminescence resonance energy transfer (BRET) is an increasingly popular technique for investigating protein– protein interactions in live cells [1,2]. Since the first publication describing its use in 1999 [3], an explosion of interest has occurred, particularly for the study of integral membrane proteins, such as G-protein coupled receptors (GPCRs) [2,4,5]. The BRET technique utilizes fusion proteins consisting of the protein of interest genetically fused to an energy donor, Abbreviations: BRET, bioluminescence resonance energy transfer; GPCRs, G-protein coupled receptors; Rluc, Renilla luciferase; EGFP, enhanced green fluorescent protein; eBRET, extended BRET; GRKs, G-protein coupled receptor kinases; TRHR1, thyrotropin-releasing hormone receptor 1; OxR2, orexin receptor 2; AT1AR, angiotensin II receptor type 1A; ERKs, extracellular signalregulated kinases; JNK3, c-Jun N-terminal kinase 3; FACS, fluorescence activated cell sorting. ⁎ Corresponding author. Tel.: +61 8 9346 1518; fax: +61 8 9346 1818. E-mail address: [email protected] (K.D.G. Pfleger). 0898-6568/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.cellsig.2006.01.004

Renilla luciferase (Rluc), or acceptor, such as enhanced green fluorescent protein (EGFP) [6]. Rluc oxidizes its substrate resulting in the emission of energy. If a suitable energy acceptor is in close proximity and is favourably oriented, this energy can be transferred. The excited acceptor molecule then emits energy at a longer wavelength, detection of which indicates that energy transfer has occurred and, consequently, that donor and acceptor molecules were within close proximity. Indeed the proximitydependence of the energy transfer (to the sixth power) [7] implies that the donor and acceptor, and therefore the proteins of interest fused to the donor and acceptor, interact directly or as part of a complex. BRET is generally carried out using coelenterazine h or DeepBlueC™ as the Rluc substrate. These are unstable in aqueous solutions, particularly at 37 °C. Furthermore, enzymeindependent luminescence (autoluminescence) is greatly increased in media containing serum [8]. Both of these considerations currently limit the use of BRET for prolonged real-time kinetic experiments. This study describes and

K.D.G. Pfleger et al. / Cellular Signalling 18 (2006) 1664–1670

characterizes for the first time a new derivation of the BRET technique that now enables extended real-time experiments to be carried out in live cells, in media containing serum, at 37 °C. Extended BRET (eBRET) utilizes a protected form of coelenterazine h, termed EnduRen™, which is metabolized to coelenterazine h by esterases within cells. Consequently, equilibrium is established between the stable, protected substrate in the media and the free coelenterazine h available for oxidation in the cells. Combining the use of this substrate with sensitive microplate reading technology and advanced kinetic software enables protein–protein interactions to be monitored for several hours in real-time under near-physiological conditions. The potential of eBRET is illustrated here by using the agonist-induced interaction between GPCRs (seven transmembrane-spanning receptors that signal via coupling to heterotrimeric guanine nucleotide binding proteins (G-proteins) [9]) and β-arrestins. The GPCR superfamily includes members that respond with high specificity to a wide range of stimuli from a photon of light through to large proteins [10]. β-arrestins are ubiquitous intracellular adaptor proteins that generally interact with GPCRs following agonist-induced receptor activation and subsequent phosphorylation by G-protein coupled receptor kinases (GRKs) [11]. Their role in ‘arresting’ receptor activity is two-fold: they sterically block binding to G-protein and they target the receptor for internalization into clathrin-coated vesicles [12]. Recent studies have also given evidence for the β-arrestin complex providing an additional signalling platform [11,13], allowing a range of signal transduction pathways to be stimulated and/or modulated in a GPCR-activation dependent manner for a significant period of time after G-protein uncoupling. Identification of such a phenomenon over this timescale highlights the need to be able to monitor prolonged GPCR/β-arrestin interactions in real-time. Three different GPCRs have been used to illustrate the methodology, demonstrating the general applicability of this interaction for studying GPCR function [14]. β-arrestin interactions with the thyrotropin-releasing hormone receptor 1 (TRHR1), orexin receptor 2 (OxR2) and angiotensin II receptor type 1A (AT1AR) have been used to validate various aspects of the methodology, including the advantages over current BRET techniques. These results illustrate the exciting potential of eBRET for real-time monitoring of a multitude of live cell protein–protein interactions over extended time periods.

1665

codon and incorporate a downstream NotI restriction site (5′GTCATATGAATAAGCGGCCGCTCCCAGTTTTGAAG-3′). The resultant PCR product was then cloned in-frame into pcDNA3 containing EGFP. The sequence was confirmed using an ABI Prism 310 Genetic Analyzer (Applied Biosystems). The N-terminally tagged β-arrestin 1 fusion protein was generated by subcloning β-arrestin 1 cDNA kindly provided by Jeffrey L. Benovic (Kimmel Cancer Institute, Philadelphia) into pcDNA3 containing EGFP or Rluc with stop codons mutated out. Agonists used were TRH (Bachem), angiotensin II (Auspep) and orexin A (American Peptide Company). 2.2. Cell culture and transfections COS-7 and HEK293 cells were maintained at 37 °C, 5% CO2 in Complete Media (DMEM containing 0.3 mg/ml glutamine, 100 IU/ml penicillin and 100 μg/ml streptomycin (Gibco)) supplemented with 10% fetal calf serum (FCS; Gibco). Transient transfections were carried out 24 h after seeding using Polyfect (Qiagen) or GeneJuice (Novagen) according to manufacturer instructions. 2.3. Measurement of total inositol phosphate production

2. Experimental procedures

COS-7 cells were seeded in 100 mm dishes at a density of 750,000 cells/dish. 24 h post-transfection, media was removed and replaced with inositol-free Complete Media containing 1% dialyzed FCS before splitting into 24-well plates. Cells were incubated for a further 24 h with [3H]myo-inositol (2 μCi/ml; Amersham Pharmacia Biotech). Media was removed and cells washed twice with buffer A (1 mg/ml fatty acid-free BSA, 140 mM NaCl, 20 mM HEPES, 4 mM KCl, 8 mM D-glucose, 1 mM MgCl2 and 1 mM CaCl2), before incubating at 37 °C for 50 min in buffer A containing 10 mM LiCl with or without addition of orexin A. Following incubation the assay buffer was removed and replaced with 10 mM formic acid for 1 h at 4 °C. The aqueous fraction was subsequently transferred to 5 ml tubes and the total inositol phosphates bound to Dowex (AG 108) anion-exchange resin (Bio-Rad Laboratories). Following washes with water and 60 mM HCOONH4 / 5 mM Na2B4O7, the inositol phosphates were eluted with 1 M HCOONH4 / 0.1 M formic acid. Finally, the cells were solubilized with 0.2 mM NaOH / 1% SDS in order to measure total radioactivity for subsequent normalization. Radioactivity was measured using a 1209 Rackbeta liquid scintillation counter (Wallac). All treatments were performed in triplicate.

2.1. Materials

2.4. Confocal microscopy

The TRHR1/Rluc [15] and AT1AR/Rluc [16] constructs have been described previously. Human OxR2 cDNA was kindly provided by Masashi Yanagisawa (Howard Hughes Medical Institute, Texas) and was used as a template to generate the OxR2/EGFP construct by PCR amplification. A 5′-primer was designed to incorporate a HindIII restriction site upstream from the start codon (5′-GGACTTGAGAAGCTTATGTCCGGCAC CAAATTG-3′) and a 3′-primer designed to mutate out the stop

HEK293 cells were seeded in 6-well plates at a density of 650,000 cells/well. The cells were plated onto poly-L-lysine coated 8-well chamber slides 24 h post-transfection. Treatments were carried out 48 h post-transfection and cells fixed in 4% paraformaldehyde. Samples were left to dry at room temperature briefly before cover-slipping and sealing. Cells were examined using a confocal laser microscope (BioRad) with an oil immersion 60× objective (Nikon). EGFP-tagged proteins

K.D.G. Pfleger et al. / Cellular Signalling 18 (2006) 1664–1670

were excited at 488 nm and emitted light was detected in the 500–550 nm range. 2.5. Bioluminescence resonance energy transfer (BRET) COS-7 cells were seeded in 6-well plates at a density of 120,000 cells/well. 48 h post-transfection, cells were washed with PBS, detached using 0.05% trypsin / 0.53 mM EDTA and resuspended in HEPES-buffered phenol red free Complete Media containing 10% FCS. An aliquot of cells was taken for analysis of relative EGFP expression using fluorescence activated cell sorting (FACS) using a FACSCalibur (BD Biosciences). For eBRET studies, cells were added to wells of a 96well black and white isoplate (Perkin-Elmer) and EnduRen™ (Promega) added to a final concentration of 30 μM. The plate was then pre-incubated at 37 °C, 5% CO2 for 2 h to ensure equilibrium was reached. For original BRET studies, coelenterazine h (Molecular Probes) was added to a final concentration of 5 μM immediately prior to commencing the assay. BRET measurements were taken at 37 °C using the Victor Light plate reader with Wallac 1420 software (Perkin-Elmer) or the Mithras LB 940 with MikroWin 2000 software (Berthold Technologies). Filtered light emissions were sequentially measured for 5 s in each of the ‘Rluc wavelength window’ (400–475 nM) and ‘EGFP wavelength window’ (N 500 nM). The BRET signal observed between interacting proteins is normalized by subtracting the background BRET ratio. This can be done in two ways; however, the latter method can only be used for ligandinduced interactions. (1) The ratio of N 500 nM emission over the 400–475 nM emission for a cell sample containing only the Rluc construct is subtracted from the same ratio for a sample containing the interacting EGFP and Rluc fusion proteins; (2) the ratio of N500 nM emission over the 400–475 nM emission for a vehicle-treated cell sample containing both EGFP and Rluc fusion proteins is subtracted from the same ratio for a second aliquot of the same cells that is treated with ligand. Note that the former method is used unless the data is labelled as ‘ligand-induced BRET ratio.’

ison of BRET ratios ascertained using coelenterazine h or EnduRen™ at various time points was carried out by one-way ANOVA and differences were not found to be significant using either Tukey or Student–Newman–Keuls multiple comparison post-tests. Comparison of BRET ratios in the presence or absence of dynamin K44A 180 min after addition of TRH was carried out using an unpaired Student's t-test. 3. Results 3.1. Validation of fusion proteins BRET fusion proteins should be validated against wild type proteins before use as the addition of a relatively large luminescent or fluorescent molecule has the potential to affect protein function. The receptor fusion constructs used in this study have been described previously (see Experimental procedures) with the exception of OxR2 fused to EGFP at the C-terminus (OxR2/EGFP). Total inositol phosphate production dose–response data generated for wild type OxR2 and OxR2/

A Inositol Phosphates (% max)

1666

100 80 wild type OxR2

60

OxR2/EGFP 40 20 0 untreated

-12

-11

-10

-9

-8

-7

-6

[orexin A] (log M)

B

Untreated

Treated

2.6. Cell viability assay Trypan blue dye exclusion was used to assess cell viability. Aliquots of cell samples containing 0.2% trypan blue (final) were counted using a Neubauer haemocytometer and inverted microscope. Viable cells (those excluding the dye) were expressed as a percentage of the total counted. 2.7. Data presentation and statistical analysis Data are presented and analyzed using Prism 4.0 graphing software (GraphPad Software). Sigmoidal dose–response curves were fitted to the data for inositol phosphate production using non-linear regression. Comparison of EC50 values was carried out using an unpaired Student's t-test with Welch's correction (does not assume equal variance). Comparison of percentage cell viability was carried out by one-way ANOVA followed by Dunnett's multiple comparison post-test. Compar-

OxR2/EGFP

EGFP/β-arrestin 1

Fig. 1. Functional validation of BRET fusion proteins. (A) Total inositol phosphate production dose–response curves for wild type OxR2 and OxR2/ EGFP in COS-7 cells. Graph is representative of three independent experiments each carried out in triplicate. (B) Confocal microscopy using HEK293 cells that were untreated or treated with orexin A. OxR2/EGFP was visualized before and after treatment for 30 min. EGFP/β-arrestin 1 was visualized in OxR2expressing cells with and without treatment for 5 min.

K.D.G. Pfleger et al. / Cellular Signalling 18 (2006) 1664–1670

EnduRen™ or coelenterazine h (Fig. 2). The luminescence resulting from EnduRen™ oxidation compared with coelenterazine h oxidation illustrates the difference in substrate stability over time (Fig. 2A). The lack of stability of coelenterazine h is

3.0 2.5 2.0

Addition of TRH or vehicle

A

1.5

0.25 + TRH

1.0 0.20

EnduRen 0.5 Coelenterazine h 0.0 0

1

2

3

4

5

6

7

8

9

B

BRET Ratio

Luminescence (arbitrary units)

A

Coelenterazine h (+ TRH)

0.4

0.15

0.10

0.05 + vehicle

0.3

0.00 0

0.2

1

2

3

4

5

6

7

8

9

3

4

5

6

7

8

9

B

EnduRen (+ TRH)

0.25

0.1

+ TRH EnduRen (+ vehicle)

0.20

Addition of TRH or vehicle

-0.1 0

1

2

3

4

5

6

7

8

9

Time (h) Fig. 2. Comparison of Rluc substrates using the TRH-induced interaction between TRHR1/Rluc and EGFP/β-arrestin 1. Transiently co-transfected COS-7 cells were pre-incubated for 2 h with EnduRen™ or coelenterazine h was added immediately prior to real-time measurements at 37 °C. TRH or vehicle was added after 30 min and measurements recommenced for approximately 9 h. Graphs show luminescence resulting from Rluc substrate oxidation over time (A) and corresponding BRET ratios for the interaction observed over time (B) using both substrates. The BRET ratio observed over time following addition of vehicle using EnduRen™ is included as a baseline. Graphs are representative of three independent experiments.

EGFP indicate that the addition of EGFP does not affect coupling of this receptor to its G-protein, Gq/11 (Fig. 1A). The mean EC50 values for wild type OxR2 (3.0 ± 0.6 nM) and OxR2/ EGFP (4.7 ± 1.1 nM) are not significantly different. Furthermore, confocal microscopy indicates that OxR2/EGFP is correctly localized to the plasma membrane in untreated cells and is internalized into endosomal vesicles following agonist treatment as expected (Fig. 1B). β-arrestin 1 used in this study was fused to either Rluc or EGFP at the N-terminus. Confocal microscopy indicates that EGFP/β-arrestin 1 exhibits diffuse cytoplasmic localization in untreated cells and rapidly translocates to the plasma membrane following GPCR activation. This is illustrated using co-expression and activation of wild type OxR2 (Fig. 1B). 3.2. Comparison of Rluc substrates EnduRen™ and coelenterazine h COS-7 cells transiently transfected with TRHR1/Rluc and EGFP/β-arrestin 1 were assayed in real-time at 37 °C using

BRET Ratio

0.0

0.15

0.10

0.05

+ vehicle

0.00 0

1

2

C 100

90

% Cell Viability

BRET Ratio

1667

*

80 + TRH

*

+ vehicle

70

60

50 0

1

2

3

4

5

6

7

8

9

Time following addition of TRH or vehicle (h) Fig. 3. Validation of cell viability when carrying out eBRET over prolonged time periods. COS-7 cells transiently co-transfected with TRHR1/Rluc and EGFP/ β-arrestin 1 were pre-incubated for 2 h with EnduRen™. Following addition of TRH or vehicle, real-time measurements at 37 °C were carried out for approximately 9 h either continuously (A) or including a 4 h break commencing 2 h posttreatment (B). During the 4 h break, cells were returned to the humidified incubator and maintained at 37 °C, 5% CO2. Graphs are representative of three independent experiments. Cell viability was also assessed using trypan blue dye exclusion 0, 2, 6 and 9 h following treatment with TRH or vehicle and continuous incubation in a BRET instrument at 37 °C (C). The percentage of viable cells was significantly lower after 9 h compared to 0 h in both samples (⁎P b 0.01), but was not significantly lower after 2 or 6 h. Data shown are mean ± SE of three independent experiments.

K.D.G. Pfleger et al. / Cellular Signalling 18 (2006) 1664–1670

BRET Ratio

0.20 0.15 0.10

EnduRen (+ orexin A) EnduRen (+ vehicle) Coelenterazine h (+ orexin A)

0.05 0.00 -0.05 0

30

60

90

120

150

180

210

240

Time following addition of orexin A or vehicle (min) Fig. 4. Real-time eBRET kinetics compared with original BRET time points using the orexin A-induced interaction between Rluc/β-arrestin 1 and OxR2/ EGFP. Transiently co-transfected COS-7 cells were pre-incubated for 2 h with EnduRen™, then assayed for 30 min pre-treatment and 240 min post-treatment with orexin A or vehicle. Measurements were taken in real-time at 37 °C. Data shown are mean ± SE of three independent experiments. Aliquots of similar transfected cells were assayed with coelenterazine h at specific time points, between which the cells were maintained in an incubator at 37 °C, 5% CO2. Data shown are mean ± SE of at least three independent experiments with data values for each experiment being the mean of five sequential measurements.

reflected by the variability in BRET ratio beyond 1 h, which is particularly high after 2 h (Fig. 2B). In contrast, the BRET ratios observed using EnduRen™ exhibit low variability for over 9 h. 3.3. Assessment of cell viability under eBRET assay conditions Continuous real-time monitoring of COS-7 cells transiently co-transfected with TRHR1/Rluc and EGFP/β-arrestin 1 shows a strong and rapid TRH-induced BRET ratio that then decreases over the course of 9 h (Fig. 3A). Similar results were obtained for the 6 to 9 h period post-treatment despite returning the cells to a humidified incubator with 5% CO2 for the 2 to 6 h period post-treatment (Fig. 3B). Cell viability, as assessed by trypan blue dye exclusion, was not significantly reduced after 2 or 6 h, and still exceeded 70% after 9 h (Fig. 3C). 3.4. Comparison between real-time eBRET measurements and BRET time points using coelenterazine h eBRET data obtained in real-time from the same cell population is comparable to prolonged kinetic data obtained with coelenterazine h using separate aliquots of a cell sample incubated with ligand for different periods of time (Fig 4). This is illustrated using the orexin A-induced interaction between Rluc/β-arrestin 1 and OxR2/EGFP. BRET ratios obtained using the two methods are not significantly different at similar time points. 3.5. Modulation of protein–protein interactions monitored by eBRET in real-time The prolonged real-time monitoring of protein–protein interactions in a live cell environment enables a wide range of

modulating factors to be investigated. This provides a powerful system by which cellular function can be elucidated, as illustrated by the two examples shown in Fig. 5. Different aliquots of the same cell sample expressing AT1AR/Rluc and EGFP/ β-arrestin 1 were assayed with various doses of angiotensin II, thereby demonstrating the dose-dependency of the interaction (Fig. 5A). In addition to reagent modulation, interactions can be modulated by co-expression with other proteins. The co-expression of dominant negative mutant dynamin (dynamin K44A) resulted in a significant increase (P b 0.05) in the BRET ratio observed for the TRHR1/β-arrestin 1 interaction 180 min after addition of TRH (Fig. 5B). An important component of the BRET ratio calculation is to subtract the ratio of light emissions obtained from a sample where BRET is not occurring. A cell preparation containing only the Rluc construct is such a sample and is used in the ‘standard calculation’ (see Experimental procedures for further details). Such a calculation assumes that the Rluc expression is similar in samples with or without acceptor co-expression and

A Ligand-induced BRET Ratio

0.25

10-6 M 10-7 M 10-8 M 10-9 M

0.15

0.10

0.05

0.00 0

20

40

60

80

100

120

140

160

180

B Ligand-induced BRET Ratio

1668

0.15

0.10 * 0.05

+ dynamin K44A - dynamin K44A

0.00 0

20

40

60

80

100

120

140

160

180

Time following addition of agonist (min) Fig. 5. Modulation of protein–protein interactions monitored by eBRET in realtime. (A) The dose-dependency of eBRET kinetic data illustrated using the angiotensin II-induced interaction between AT1AR/Rluc and EGFP/β-arrestin 1. Data shown are mean ± SE of four independent experiments. (B) The effect of dynamin K44A co-expression on the TRH-induced interaction between TRHR1/Rluc and EGFP/β-arrestin 1. 180 min after addition of TRH, BRET ratios for samples with and without dynamin K44A co-expression were significantly different (⁎P b 0.05). Data shown are mean ± SE of three independent experiments. Transiently co-transfected COS-7 cells were preincubated for 2 h with EnduRen™, then assayed for 20 min pre-treatment and 180 min post-treatment with vehicle or agonist. Measurements were taken in real-time at 37 °C. The ligand-induced BRET ratios are calculated as described in Experimental procedures.

K.D.G. Pfleger et al. / Cellular Signalling 18 (2006) 1664–1670

the ratiometric nature of the resultant data should account for potential differences. However, an alternative and theoretically more accurate calculation when assessing ligand-induced interactions involves using the vehicle-treated sample as the negative control, as it is known to be identical to the ligandtreated sample with regards to protein expression. This method has the additional benefit of not requiring a cell sample containing only the Rluc construct, thereby significantly streamlining the transfection and assaying procedures, an important consideration for high throughput screening. The alternative calculation is used to generate the ligand-induced BRET ratios shown in Fig. 5. 4. Discussion These results clearly validate the eBRET methodology and demonstrate its potential for monitoring protein–protein interactions for prolonged time periods, in live cells, in real-time. We have used a variety of GPCRs to demonstrate the potential of eBRET for investigating the important interaction with β-arrestins. This activation-dependent interaction has recently been identified as resulting, not only in G-protein uncoupling and receptor internalization, but also the formation of secondary scaffolding/signalling complexes after G-protein uncoupling [11,13]. These complexes are able to recruit a range of binding proteins that enable multiple signalling pathways to be activated [13], including c-Src family tyrosine kinases [17], extracellular signal-regulated kinases (ERKs) [18] and c-Jun N-terminal kinase 3 (JNK3) [19]. Furthermore, studies using AT1AR indicate that such complexes are trafficked into endosomal vesicles [20–22], with little reduction in ERK activation after 90 min [22]. We have demonstrated that eBRET is an effective method for prolonged real-time monitoring of β-arrestin interactions with GPCRs, including AT1AR, in live cells. This represents a significant and timely advance in our capacity to elucidate the function of β-arrestin-mediated signalling over the required timescale. The success of the methodology for studying GPCR/ β-arrestin interactions indicates that it can be applied to investigate a wide range of interactions with diverse treatment regimes. Indeed, the primary requirement is the same as that for the current BRET technologies, namely that the proteins of interest can be suitably labelled and functionally validated, as we have shown here for OxR2/EGFP and EGFP/β-arrestin 1. The importance of considering substrate decay when designing high-throughput BRET screening assays has been highlighted recently, with the selection of coelenterazine h over DeepBlueC™ because of its higher quantum yield and significantly slower decay kinetics [14]. EnduRen™ is metabolized to coelenterazine h in cells and so has a similar quantum yield to native coelenterazine h when oxidized by Rluc. However, as we have shown in this study, the decay kinetics of EnduRen™ are substantially slower than those of native coelenterazine h, with important consequences for the design of high-throughput BRET screening assays. The prolonged stability of the substrate enables many 96- or 384-well plates to be screened by BRET without the reduction in luminescence becoming a limiting factor. Consequently, with the aid of readily available robotics/

1669

plate stacking systems, large scale BRET screening can be carried out. Currently, the screening of multiple plates requires administration of substrate approximately every 30 min for coelenterazine h or injection into individual wells prior to DeepBlueC™ measurements [14]. Cell viability is an important consideration if carrying out live cell assays over many hours, whether monitoring a few samples many times to observe interactions over prolonged timescales, or many samples a few times to carry out highthroughput screens. We have demonstrated that by assaying cells in HEPES-buffered media, real-time eBRET measurements can be made for at least 6 h at 37 °C. Indeed, the eBRET data obtained continuously compares favourably with that obtained with an intermediate period in the humidified 37 °C incubator with 5% CO2. The development of detection instrumentation that can maintain cells with CO2 perfusion is likely to further enhance the potential longevity of these assays should this be desired. Prolonged kinetic studies can be carried out with the existing BRET technologies; however, this requires the use of multiple cell samples assayed after set time-points. Data obtained in this fashion is generally comparable to data obtained in real-time using eBRET; however, important events occurring between time-points can be missed. This possibility can be reduced be using frequent time-points; however, this is laborious and requires multiple cell samples. Furthermore, the need to use more than one sample per kinetic profile introduces additional variability. The scope for using agonists, antagonists, inhibitors, dominant negative mutants and co-expressed accessory proteins to dissect out cellular function is considerable. Prolonged real-time assays can be carried out in parallel on aliquots of the same cell sample with a range of different agents at multiple concentrations. We have illustrated this by showing how various doses of angiotensin II differentially affect the interaction between AT1AR and β-arrestin 1. In a similar manner, comparisons between different agonists can be made, antagonists assessed for their ability to compete for agonist binding, or inhibitors/ modulators used to elucidate cellular mechanisms. The alternative to reagent addition is to co-express other proteins, such as dominant negative mutants or accessory proteins. The example we have used to illustrate this is the effect of dominant negative mutant dynamin (dynamin K44A) on the TRH-induced TRHR1 interaction with β-arrestin 1. GPCR/β-arrestin complexes traffic to clathrin-coated pits that become clathrin-coated vesicles when pinched off by the GTPase dynamin [23]. Dynamin K44A is defective in GTP binding and hydrolysis and consequently blocks vesicle formation [23]. We have shown previously that this impairs TRHR1 internalization [15,24]. The results from the present study indicate that the TRHR1/β-arrestin 1 interaction is not initially affected by the impairment of vesicle formation. This is consistent with the notion that GPCR/ β-arrestin interactions do not require receptor internalisation [25]. Indeed, the TRHR1/β-arrestin 1 interaction appears to be prolonged by this impairment, presumably as the processes of dephosphorylation and complex dissociation that occur

1670

K.D.G. Pfleger et al. / Cellular Signalling 18 (2006) 1664–1670

following trafficking into endosomal vesicles [26] are reduced and/or delayed. In conclusion, we have demonstrated the significant advantages of eBRET over current BRET techniques, validated the method for real-time monitoring of live cells at 37 °C for several hours, shown that eBRET provides comparable data to the more laborious time-point assays using BRET with coelenterazine h, and provided examples of possible applications for examining cellular function. eBRET is an enabling technology that has important and exciting implications for enhancing drug discovery and our understanding of cellular biology. Acknowledgements

[4] [5] [6] [7] [8]

[9] [10] [11] [12] [13] [14] [15]

We thank Ruth M. Seeber and Paul Rigby for expert technical assistance, the Lotteries Commission of WA for support of the Biomedical Imaging and Analysis Facility and Promega for the initial supply of Enduren. We are grateful to Masashi Yanagisawa (Howard Hughes Medical Institute, Texas) and Jeffrey L. Benovic (Kimmel Cancer Institute, Philadelphia) for providing cDNA constructs. This work was funded by the National Health and Medical Research Council (NHMRC) of Australia (project grants #254646 and #404087). Kevin D.G. Pfleger is supported by an NHMRC Peter Doherty Fellowship (#353709), Jasmin R. Dromey by a University of Western Australia Postgraduate Scholarship, Matthew B. Dalrymple by a Keogh Institute for Medical Research scholarship and Karin A. Eidne by an NHMRC Principal Research Fellowship (#212064). References [1] K.D.G. Pfleger, K.A. Eidne, Nat. Methods in press. [2] K.D.G. Pfleger, K.A. Eidne, Biochem. J. 385 (2005) 625. [3] Y. Xu, D.W. Piston, C.H. Johnson, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 151.

[16] [17]

[18] [19] [20] [21] [22] [23] [24] [25] [26]

G. Milligan, Eur. J. Pharm. Sci. 21 (2004) 397. G. Milligan, M. Bouvier, FEBS J. 272 (2005) 2914. K.D.G. Pfleger, K.A. Eidne, Pituitary 6 (2003) 141. P. Wu, L. Brand, Anal. Biochem. 218 (1994) 1. E.M. Hawkins, M. O'Grady, D. Klaubert, M. Scurria, T. Good, C. Stratford, R. Flemming, D. Simpson, K.V. Wood, 12th International Symposium on Bioluminescence and Chemiluminescence, 2002, Robinson College, University of Cambridge (World Scientific), 149. K.L. Pierce, R.T. Premont, R.J. Lefkowitz, Nat. Rev., Mol. Cell Biol. 3 (2002) 639. T.H. Ji, M. Grossmann, I. Ji, J. Biol. Chem. 273 (1998) 17299. L.M. Luttrell, R.J. Lefkowitz, J. Cell Sci. 115 (2002) 455. S.S. Ferguson, Pharmacol. Rev. 53 (2001) 1. R.J. Lefkowitz, S.K. Shenoy, Science 308 (2005) 512. F.F. Hamdan, M. Audet, P. Garneau, J. Pelletier, M. Bouvier, J. Biomol. Screen. 10 (2005) 463. K.M. Kroeger, A.C. Hanyaloglu, R.M. Seeber, L.E. Miles, K.A. Eidne, J. Biol. Chem. 276 (2001) 12736. D.T. Dinh, H. Qian, R. Seeber, E. Lim, K. Pfleger, K.A. Eidne, W.G. Thomas, Mol. Pharmacol. 67 (2005) 375. L.M. Luttrell, S.S. Ferguson, Y. Daaka, W.E. Miller, S. Maudsley, G.J. Della Rocca, F. Lin, H. Kawakatsu, K. Owada, D.K. Luttrell, M.G. Caron, R.J. Lefkowitz, Science 283 (1999) 655. K.A. DeFea, J. Zalevsky, M.S. Thoma, O. Dery, R.D. Mullins, N.W. Bunnett, J. Cell Biol. 148 (2000) 1267. P.H. McDonald, C.W. Chow, W.E. Miller, S.A. Laporte, M.E. Field, F.T. Lin, R.J. Davis, R.J. Lefkowitz, Science 290 (2000) 1574. L.M. Luttrell, F.L. Roudabush, E.W. Choy, W.E. Miller, M.E. Field, K.L. Pierce, R.J. Lefkowitz, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 2449. A. Tohgo, K.L. Pierce, E.W. Choy, R.J. Lefkowitz, L.M. Luttrell, J. Biol. Chem. 277 (2002) 9429. S. Ahn, S.K. Shenoy, H. Wei, R.J. Lefkowitz, J. Biol. Chem. 279 (2004) 35518. S. Sever, H. Damke, S.L. Schmid, Traffic 1 (2000) 385. A. Heding, M. Vrecl, A.C. Hanyaloglu, R. Sellar, P.L. Taylor, K.A. Eidne, Endocrinology 141 (2000) 299. H. Qian, L. Pipolo, W.G. Thomas, Mol. Endocrinol. 15 (2001) 1706. R.H. Oakley, S.A. Laporte, J.A. Holt, L.S. Barak, M.G. Caron, J. Biol. Chem. 276 (2001) 19452.