Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Quantitative Assessment of β1- and β2-Adrenergic Receptor Homo- and Heterodimerization by Bioluminescence Resonance Energy Transfer

Quantitative Assessment of β1- and β2-Adrenergic Receptor Homo- and Heterodimerization by... THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 277, No. 47, Issue of November 22, pp. 44925–44931, 2002 © 2002 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Quantitative Assessment of  - and  -Adrenergic Receptor 1 2 Homo- and Heterodimerization by Bioluminescence Resonance Energy Transfer* Received for publication, June 11, 2002, and in revised form, August 23, 2002 Published, JBC Papers in Press, September 19, 2002, DOI 10.1074/jbc.M205767200 ¶ ¶ Jean-Franc ¸ ois Mercier‡§, Ali Salahpour‡ , Ste ´ phane Angers , Andreas Breit, and Michel Bouvier From the De ´partement de Biochimie and Groupe de Recherche sur le Syste `me Nerveux Autonome, Universite ´ de Montre ´al, Montre ´al, Quebec H3C 3J7, Canada and 2). In most instances, co-immunoprecipitation was used as Quantitative bioluminescence resonance energy transfer (BRET) analysis was applied to the study of  - the primary experimental evidence supporting the existence of and  -adrenergic receptor homo- and heterodimeriza- such dimers. More recently, however, light resonance energy tion. To assess the relative affinity between each of the transfer techniques such as fluorescence and bioluminescence protomers, BRET saturation experiments were carried resonance energy transfer (FRET and BRET) were also used. out in HEK-293T cells.  - and  -adrenergic receptors 1 2 These “non-invasive” proximity-based assays confirmed that were found to have similar propensity to engage in ho- GPCR dimerization does not represent biochemical artifacts mo- and heterotropic interactions suggesting that, at due to receptor solubilization and can occur in living cells. They equivalent expression levels of the two receptor sub- have been used to demonstrate homodimerization of the  - types, an equal proportion of homo- and heterodimers adrenergic (3), the yeast alpha mating factor (4), the SST5 would form. Analysis of the data also revealed that, at somatostatin (5), the gonadotropin releasing hormone (6), the equimolar expression levels of energy donor and accep- luteinizing hormone (7), the -opioid (8), the thyrotropin-re- tor, more than 80% of the receptor molecules exist as leasing hormone (9), the cholecystokinin (10), and the melato- dimers and that this high incidence of receptor dimer- nin (11) receptors as well as heterodimerization between soma- ization is insensitive to receptor density for expression tostatin receptor subtypes (5), somatostatin and dopamine levels varying between 1.4 and 26.9 pmol of receptor/mg receptors (12), melatonin receptor subtypes (11), and opioid of membrane protein. Taken together, these results in- receptor subtypes (13). dicate that most of the receptors expressed in cells exist An advantage of BRET and FRET over co-immunoprecipita- as constitutive dimers and that, at least in undifferenti- ated fibroblasts, the proportion of homo- and het- tion approaches lies in the more quantitative nature of the erodimers between the closely related  - and  -adre- assay. However, relatively few studies exploited this quantita- 1 2 nergic receptors is determined by their relative levels of tive potential for the study of GPCR dimerization. For the expression. melatonin receptors, Ayoub et al. (11) recently used BRET competition assays to determine that the transfer of energy resulted from the formation of dimers and not of higher order G protein-coupled receptors (GPCRs) represent the largest oligomers. They also showed that ligand binding did not alter family of transmembrane receptors involved in cell signaling. the dimerization state of the receptors. However, other ques- In the past few years, many studies indicated that GPCR tions that could theoretically be addressed by quantitative dimerization can occur between two identical receptors (ho- energy transfer analysis, such as the relative affinity of the modimerization), between two different receptor subtypes of dimer partners for each other and the relative proportion of the same family, or even between receptors that are only dis- receptors engaging in dimer formation, have not yet been tantly related (heterodimerization) (for a review, see Refs. 1 addressed. - and  -adrenergic receptors ( AR and  AR) have pre- 2 1 2 1 viously been shown to exist as homodimers (3, 8, 14, 15). The * This work was supported in part by grants from the Canadian high level of sequence identity existing in domains proposed to Institute for Health Research (CIHR) and the Heart and Stroke Foun- contribute to the dimerization interface (i.e. transmembrane dation of Canada (to M. B.). The costs of publication of this article were helices) (14, 16) makes them a system of choice to study their defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 potential heterodimerization and the relative affinity of the U.S.C. Section 1734 solely to indicate this fact. protomers within homo- and heterotropic complexes. Although ‡ Both authors contributed equally to this work. direct in vivo demonstration for the co-localization of the two § Held a studentship from the Fonds de la Recherche en Sante´du receptor subtypes in the same cell is still lacking, the presence Que ´ bec. Hold studentships from the CIHR. of  and  AR in the same cell types has been taken as 1 2 Holds a Canada Research Chair in Molecular and Cellular Pharma- evidence for their co-expression in transitional and mid-nodal cology. To whom correspondence should be addressed: De ´ pt. de Bio- cells of the atrio-ventricular node, nerve processes of the atrio- chimie, Universite ´ de Montre ´ al, C.P. 6128, Succursale Centre-Ville, ventricular and ventricular conduction systems, as well as in Montre ´ al, Quebec H3C 3J7, Canada. Tel.: 514-343-6372; Fax: 514-343- 2210; E-mail: [email protected]. vascular smooth muscle cells of the kidney (17, 18). The fact The abbreviations used are: GPCR, G protein-coupled receptor; that the two receptor subtypes are also found together in a FRET, fluorescence resonance energy transfer; BRET, bioluminescence large number of tissues, including liver, lung, and fat (19), resonance energy transfer; AR, adrenergic receptor; GFP, green fluo- gives further support to the idea that heterodimerization could rescent protein; PBS, phosphate-buffered saline; Rluc, Renilla lucifer- ase; CYP, cyanopindolol. occur in native tissues and warrants investigations aiming to This paper is available on line at http://www.jbc.org 44925 This is an Open Access article under the CC BY license. 44926 Energy Transfer Analysis of -Adrenergic Receptor Dimers Following harvesting, cells were distributed in 96-well microplates assess the likelihood of such intermolecular complexes. (white Optiplate from Packard Bioscience) at a density of 100,000 Here, quantitative BRET approaches were applied to the cells per well. DeepBlueC was added at a final concentration of 5 M, study of  - and  -adrenergic receptor dimerization in a het- 1 2 and readings were collected using a modified Top-count apparatus erologous mammalian expression system. In particular, BRET (BRETCount, Packard Bioscience) that allows the sequential integra- saturation experiments were carried out to estimate the rela- tion of the signals detected in the 370- to 450-nm and 500- to 530-nm tive affinity of each receptor subtype to engage into homo- and windows using filters with the appropriate band pass (Chroma). The BRET signal is determined by calculating the ratio of the light emitted heterotropic interactions. The influence of receptor density on by the Receptor-GFP (500 –530 nm) over the light emitted by the Re- the proportion of receptor molecules forming dimers was also ceptor-Rluc (370 – 450 nm). The values were corrected by subtracting assessed. Here, we report that  AR and  AR can form homo- 1 2 the background signal detected when the Receptor-Rluc constructs and heterodimers and that the two receptors have similar were expressed alone. affinities for each other and for themselves, suggesting that Fluorescence and Luminescence Measurement heterodimers are likely to form in cells expressing both sub- types. The proportion of receptor molecules forming dimers was Cells were distributed in 96-well microplates (white Costar plate with clear bottom) at a density of 100,000 cells per well. The total also found to be greater than 80% at low receptor density and fluorescence of cells was measured using a FluoroCount (Packard Bio- was constant over a 20-fold expression range. science) with an excitation filter at 400 nm, an emission filter at 510 EXPERIMENTAL PROCEDURES nm, and the following parameters: gain, 1; photo multiplicator tube, 1100 V; time, 1.0 s. After the fluorescence measurement, the same cells Receptor Constructs were incubated for 10 min with Coelenterazine H (Molecular Probe) at AR-GFP10 —The  AR coding sequence without its stop codon was 1 1 a final concentration of 5 M, and the total luminescence of cells was amplified from the pBC12BI-human  AR plasmid (20) using sense and measured using a LumiCount (Packard Bioscience) with the following antisense primers harboring unique SacI and AgeI sites. The fragment parameters: gain, 1; photo multiplicator tube, 700 V; time, 0.5 s. For was then subcloned in-frame into the SacI/AgeI site of the blue variant both measurements, the mean of duplicate wells was calculated. The GFP-sapphire vector (pGFP-N1-Sapphire, Packard Bioscience) to give total fluorescence was then divided by the background determined in the plasmid pGFP-N1- AR-Sapphire. Finally, the GFP-Sapphire was wells containing untransfected cells. Fluorescence was expressed in replaced by a green GFP variant (GFP10) containing the following -fold over background. The background was negligible for the lumines- mutations: P64L, S147P, and S202P. For this purpose, an AgeI/BsrgI cence measurements, so they were expressed as absolute values. fragment of the GFP10 variant was subcloned into the AgeI/BsrgI site of pGFP-N1- AR-Sapphire to finally yield pGFP-N1- AR-GFP10. Radioligand Binding Assay 1 1 AR-GFP10 —The GFP10 AgeI/BsrgI fragment was subcloned into Forty-eight hours after transfection, 10,000 cells (2 g of proteins) the AgeI/BsrgI site of pGFP-N1-His AR-YFP (3). For simplicity, were incubated in a final volume of 500 l of PBS containing 0.1% GFP10 will be referred to as GFP in the remainder of the text. bovine serum albumin with a saturating concentration (250 pM)ofthe AR-Rluc—The pcDNA3.1- AR:6aa:hRluc was a generous gift 1 1 125 125 -adrenergic antagonist [ I]cyanopindolol ([ I]CYP). Nonspecific from BioSignal Packard. This fusion protein contains a linker of six binding was determined as the residual binding observed in the pres- amino acids (YGPPGS) linking the carboxyl tail of the human  AR and ence of 10 M alprenolol (Sigma). Binding reactions were carried out at the humanized Rluc. room temperature for 90 min and stopped by rapid filtration over AR-Rluc—The humanized Rluc coding sequence (pRluc(h), Pack- Whatman GF/C glass-fiber filters. Receptor densities are expressed in ard Bioscience) was amplified using sense and antisense primers and femtomoles of receptor per milligram of total cell proteins assuming one subcloned into the PCR Blunt II Topo vector (Invitrogen). The hRluc binding site per receptor molecule. Linear regression curves between fragment was excised by digestion with KpnI/XbaI and subcloned into the luminescence and fluorescence signals and the number of receptor the KpnI/XbaI-digested pcDNA3.1 Zeo vector to generate the determined by radio-ligand binding were then generated from cells pcDNA3.1 Zeo/hRluc plasmid. The human His AR coding sequence expressing each of the constructs individually. To determine receptor was amplified without its stop codon using sense and antisense prim- surface density, the surface of HEK-293T cells was determined by ers. The PCR product was subcloned into PCR Blunt II Topo Vector, measuring the average length and width of the cells under phase- then excised by double digestion with HindIII/KpnI and ligated into the contrast microscopy. HindIII/KpnI-digested expression vector pcDNA3.1Zeo/hRluc. The re- sulting construct encodes a six-amino acid linker (GSGTGS) between Light Emission/Receptor Binding Correction Factor the carboxyl-terminal of the  AR and the humanized Rluc sequence. Given that the correlations between receptor numbers and the lumi- Cell Culture and Transfection nescence or fluorescence levels were intrinsic characteristics of each of the constructs, comparison between receptor densities derived from the HEK-293T cells were maintained in Dulbecco’s modified Eagle’s me- light measurements required a correcting factor. This correction was dium supplemented with 10% fetal bovine serum, 100 units/ml penicillin achieved using the linear regression generated for each constructs (see and streptomycin, 2 mML-glutamine (all from Wisent). For transfection legend of Fig. 2) and by normalizing the light emission/receptor number experiments, cells were seeded at a density of 2  10 cells per 100-mm as a function of the steeper slope factor obtained for both Rluc and GFP. dish and cultured for 24 h. Transient transfections were then performed These corrected receptor number values were then used to generate the using the calcium phosphate precipitation protocol (21). 24 h after trans- corrected BRET saturation curves presented in Fig. 3. fection, Dulbecco’s modified Eagle’s medium was replaced, and the cells were then cultured in the same medium for an additional 24 h. RESULTS AND DISCUSSION Forty-height hours post-transfection, cells were washed twice with Homo- and heterodimerization of the  and  AR were PBS, detached with PBS/EDTA and resuspended in PBS/glucose 0.1%. 1 2 On a routine basis, the protein concentration of the samples was deter- investigated by quantitative BRET analysis. For this purpose, mined to control for the number of cells using the Bradford assay kit human  and  AR cDNAs were fused at their carboxyl ter- 1 2 (Bio-Rad) with bovine serum albumin as a standard. minus to the energy donor Rluc and acceptor GFP. The affini- ties of the fusion proteins for the antagonist cyanopindolol and BRET Measurement the agonist isoproterenol as well as the potency of isoproterenol We have used a slight modification of the previously published BRET 2 to stimulate adenylyl cyclase were indistinguishable from assay (3, 22). The new BRET technology (BioSignal Packard) takes those of the wild-type receptors (data not shown). advantage of the spectral properties of a luciferase substrate known as Deep Blue coelenterazine (DeepBlueC, Packard Bioscience), which al- BRET and FRET approaches have been used in several stud- lows a better separation between the Renilla luciferase (Rluc) and the ies to assess GPCR homo- and heterodimerization (2). In most green fluorescent protein (GFP) emission spectra. Upon the catalytic cases, little attention has been paid to the ratio of donor/ degradation of DeepBlueC, the energy donor Rluc emits light with a acceptor molecules that was used in the assays, and thus the peak at 400 nm that allows the excitation of the energy acceptor, GFP. interpretation of the data remained rather qualitative. How- Once excited, GFP then re-emits fluorescence with a peak at 510 nm if ever, controlling this parameter is essential for proper quanti- the donor and acceptor molecules are within BRET-permissive distance (100 Å). tative analysis. Indeed, the level of energy transfer detected for Energy Transfer Analysis of -Adrenergic Receptor Dimers 44927 FIG.1. BRET saturation curves. HEK-293T cells were co-transfected with a constant DNA concentration of  AR- Rluc (A)or  AR-Rluc (B) and increasing DNA concentrations of  AR-GFP,  AR- 1 2 GFP, or soluble GFP. The BRET, total luminescence, and total fluorescence were measured 48 h after transfection. BRET levels are plotted as a function of the total fluorescence signal (-fold over back- ground) used as an index for the concen- tration of receptor-GFP constructs ex- pressed. The results are expressed as the mean  S.E. of 3–10 independent experi- ments carried out in triplicates. The curves were fitted using a non-linear re- gression equation assuming a single bind- ing site (GraphPad Prism). a given concentration of donor should rise with increasing the better resolution between the emission peaks of Rluc and concentration of the acceptor until all donor molecules are GFP with BRET discussed under “Experimental Procedures” engaged by an acceptor. It follows that the energy transfer is not the only difference between the two generations of BRET. should reach a plateau and that saturation curves could theo- In particular, the quantum yield of the Rluc/DeepBlueC coelen- retically be constructed. The maximal level reached should be a terazine couple is lower than that of Rluc/coelenterazine H, and function of the total number of dimers formed and of the dis- the extent of overlap between the emission spectra of Rluc and 2 1 tance between the donor and acceptor within the dimers, the excitation of the GFPs is better for BRET than BRET . whereas the concentration of acceptor giving 50% of energy These parameters can influence the sensitivity of the assays to transfer (BRET ) should be a reflection of the relative affinity detect small changes in distance between energy donors and of the acceptor fusion for the donor fusion proteins. Here, we acceptors. The notion that certain BRET configurations but not applied this theoretical framework to the study of  and  AR others allow the detection of conformational changes induced 1 2 homo- and heterodimerization by constructing BRET satura- by ligands is also well exemplified by the recent observation tion curves in cells co-transfected with a constant amount of that agonist and antagonist binding increased the BRET be- receptor-Rluc construct and increasing concentrations of the tween the melatonin MTR1-Rluc and MTR2-GFP but not receptor-GFP plasmids. As shown in Fig. 1, significant BRET between MTR2-Rluc and MTR1-GFP (11). signals were observed for the  AR/ AR and  AR/ AR pairs Although the curves generated concur with the theoretical 2 2 1 1 confirming previous findings that both receptor subtypes can behavior predicted above, quantitative analysis is complicated form homodimers (3, 14, 15). Albeit to a lower extent, co- by the lack of direct information provided by the fluorescence expression of the two subtypes also led to a sizable BRET signal and luminescence measurements on the precise concentration for the two transfer orientations (i.e.  AR-Rluc/ AR-GFP and of receptor molecules expressed. In the absence of such infor- 1 2 AR-Rluc/ AR-GFP). In all cases, BRET increased as a hy- mation, determining the relative affinity of the protomers for 2 1 perbolic function of the concentration of the GFP fusion con- each other, based on these saturation isotherms, would require struct added (assessed by the fluorescence emitted upon direct that the correlations between light emission and the number of excitation at 400 nm) reaching an asymptote at the highest receptor-GFP and -Rluc fusion molecules are linear and iden- concentrations used. Co-expression of  or  AR-Rluc with tical for the two receptors considered. To directly test this 2 1 soluble GFP led to marginal signals that increased linearly supposition, cells were transfected with increasing concentra- with increasing amount of GFP added. Stimulation with the tions of receptor-GFP and -Rluc constructs. For each DNA agonist isoproterenol did not promote any consistent change in concentrations, the total expression level of the receptors was the BRET saturation curves (data not shown) indicating that determined using the lipophilic ligand [ I]CYP, whereas the the dimers form constitutively and that receptor activation total luminescence and total fluorescence emitted by the Rluc does not affect their oligomerization state. However, one cannot and GFP fusion proteins were measured following addition of exclude the possibility that agonist stimulation could promote the Rluc substrate coelenterazine H and direct excitation of the assembly/disassembly cycles that do not affect the steady-state GFP at 400 nm, respectively. Fig. 2 illustrates the correlation proportion of receptors engaged in dimers. The modest agonist- obtained between the number of total binding sites and either promoted increase in BRET, previously reported for the  AR the luminescence or fluorescence emitted by each of the recep- homodimer using BRET (3), most likely reflected conforma- tor fusion molecules. Even though the regression curves were tional changes that could not be detected using BRET . Indeed, highly linear, their slopes were different for the two receptors 44928 Energy Transfer Analysis of -Adrenergic Receptor Dimers receptor-GFP in a given surface was modeled for the  AR homodimer. For this purpose, the receptor surface density was estimated by microscopic measurements of the HEK-293T cell surface (240 m ) and determining the receptor number using the equations in Fig. 2. The average receptor-Rluc surface density in our BRET saturation experiments was found to be 3000 receptors/m . Fig. 3D shows simulations of bystander BRET carried out for receptor-Rluc levels of 30, 300, and 3000 receptors/m . This was accomplished using a Monte-Carlo approach that assumes a random and uniform distribution of the receptors on the calculated surface. Considering that the diameter of GPCR has been estimated to be 50 Å, assuming that, in the case of bystander BRET, receptor molecules would not be intertwined, and given that the R (the distance at which the energy transfer reaches 50% of its maximum) is 50 Å,we estimated the BRET permissive surface as 50 Å . As shown in Fig. 3D, the predicted bystander BRET curves differed sig- nificantly from the BRET saturation curve obtained experi- mentally. Indeed, they progressed quasi-linearly up to recep- tor-GFP surface density for which experimental BRET values have already reached saturation. It should be noted that the difference between the experimental BRET saturation curve and the bystander BRET modeled for a receptor surface density of 3000/m is most likely being underestimated, because the FIG.2. Linear relationship between total luminescence (A)or surface measurements did not account for plasma membrane total fluorescence (B) and receptor density. HEK-293T cells were details such as microvilli that would contribute to increased transfected with increasing DNA concentrations of  or  AR-Rluc (A) 1 2 cell surface area. Moreover, the number of calculated receptors or -GFP (B) fusions. Total receptor density was determined by radioli- gand binding assays using [ I]CYP as the tracer. The total fluores- reflects the total cellular receptor content and not only those cence was measured following excitation at 400 nm and detection at 510 present at the cell surface. The aspect of the experimental nm, whereas the total luminescence was recorded following the addition BRET saturation curve thus suggests a clustering of the energy of the Rluc substrate, coelenterazine H. The linear regression curve was donor and acceptor molecules resulting form receptor oligomer- generated using GraphPad Prism. The dotted line corresponds to the ization rather than from their random collisions. extrapolation of the linear regression for the  AR-GFP. The linear regression equations used to calculate the receptor amount for a given In an effort to distinguish between dimers and higher order luminescence or fluorescence intensity are as follows:  AR-Rluc: y oligomers, we modeled the BRET saturation curves using an 1.5540(x)  20.00;  AR-Rluc: y  0.3192(x)  20.00;  AR-GFP: y 2 1 equation (modified from Ref. 25) that describes the probability 0.0002902(x)  1.00;  AR-GFP: y  0.0001004(x) 1.00. of forming BRET competent complexes as a function of the number of receptors within a complex (i.e. dimer versus trimer versus tetramer, etc). Fig. 3D shows that the experimental considered. Indeed, the GFP and Rluc signals increased more curve fits better to the theoretical dimer curve than that pre- rapidly with receptor number for the  than the  AR. Al- 1 2 dicted for a trimer, suggesting that the BRET obtained from though the exact cause for this difference remains unknown, co-expression of  AR-Rluc and  AR-GFP results from the this obviously complicates the analysis of the data, because this 2 2 formation of dimeric complexes of this receptor. difference must be taken into account to assess the relative When comparing the BRET saturation curves obtained for affinity of the receptors for each other in BRET saturation the  and  AR homo- and heterodimers (Fig. 3, A and B), curves. The linear regression equations derived from Fig. 2 1 2 were thus used to transform the luminescence and fluorescence similar BRET values were obtained for all pairs considered (Table I), indicating that the receptors had similar relative value in receptor number. Although the BRET saturation affinities for one another. This has important implications, curves were carried out using a fixed concentration of the Rluc fusion partners, co-transfecting an increasing quantity of the because it suggests that, under basal conditions,  and  AR 1 2 homo- and heterodimers have a similar probability of forming GFP constructs introduces some levels of variability in the amount of receptor-Rluc expressed in each case. To rule out the when the two receptors are heterologously expressed in HEK- influence of this variable, the BRET levels were plotted as a 293T cells. Co-localization of the two -adrenergic receptors has been documented in numerous tissues (19, 26, 27). Al- function of the ratio between the receptor-GFP/receptor-Rluc numbers. though it has not been easy to experimentally demonstrate co-localization in the same cells, their expression in the same As shown in Fig. 3, the BRET saturation curves generated following these corrections also behaved as hyperbolic func- cell types has been taken as evidence for the co-existence of the tions reaching a saturation level. The aspect of these curves two receptor subtypes (17, 18). Thus, the equal chance of form- ing homo- and heterodimers found in the present study could greatly contrasts with that of the curve predicted if the ob- served BRET resulted from random collisions promoted by a have important physiological consequences. However, future high receptor density. Indeed, a quasi-linear curve would be studies will be required to determine if heterodimerization can expected if such “bystander” BRET was taking place (23, 24). occur in native tissues. The schematic illustration of the predicted distribution of the The high likelihood of heterodimer formation between ho- energy donor and acceptor in the case of dimerization versus mologous receptors has also been suggested by the recent ob- random collision, presented in Fig. 3C, allows for an intuitive servations of Ramsay et al. (13) showing that homo- and het- appreciation of the difference between the two situations. To erodimerization of the - and -opioid receptors occurs at allow a more quantitative comparison, the progression of by- comparable levels of receptor expression. Although het- stander BRET as a function of increasing concentration of the erodimerization could also be observed between the distantly Energy Transfer Analysis of -Adrenergic Receptor Dimers 44929 FIG.3. A and B, corrected BRET saturation curves. The fluorescence and luminescence data from Fig. 1 were transformed into receptor numbers using equations from Fig. 2 and taking the slope factor into account as indicated under “Experimental Procedures.” The BRET levels are plotted as a function of the ratio of [receptor-GFP]/[receptor-Rluc]. The total number of receptors expressed ranged as follows: AR/ AR, 0.8 –23 pmol/mg; 1 1 AR/ AR, 0.8 –17 pmol/mg;  AR/ AR, 0.3–37 pmol/mg; and  AR/ AR, 0.3– 8 pmol/mg. The curves were fitted using a non-linear regression 1 2 2 1 2 2 equation assuming a single binding site (GraphPad Prism). C, schematic representation of the distribution of the energy donor (Receptor-Rluc) and acceptor (Receptor-GFP) in the case of random collision (upper panel) versus dimerization (lower panel) when the density of energy donor is maintained constant while that of the acceptor is increased. D, modeling of the evolution of BRET resulting form random collision between energy donor and acceptor (“bystander BRET”) as a function of increased acceptor surface density for donor surface densities of 30, 300, and 3000 receptors/ m . We simulated this process using a Monte-Carlo approach based on the assumption of random distribution. For each incremental increase of receptor-GFP, we modeled the probability P of interaction with any receptor-Rluc as P  1  (1  p) , where p is defined as the ratio of the BRET permissive surface of the receptor-Rluc (50A ) over the total calculated surface of the cell and n is the number of remaining receptor-Rluc available. The experimental data of  AR-Rluc/ AR-GFP presented in B are also compared with the expected BRET saturation 2 2 curve for trimeric and dimeric complexes using an equation that describes the probability of forming BRET competent complexes, BRET competent n1 n1 n n1 n1 complex  (nd a  nda )/(nd  nd a  nda ), where n is the number of receptor molecule in the complex, d is the number of receptor-Rluc (energy donor), and a the number of receptor-GFP (energy acceptor). The curves expressed as percent maximal BRET are plotted as a function of the [receptor-GFP]/[receptor-Rluc] ratio. related opioid and  -adrenergic receptors (8, 28), this appar- partners were found to be very similar (Table I), the second ently takes place only at much higher receptor expression hypothesis is more likely. levels, indicating that heterodimerization is more likely to oc- One of the major concerns when considering the physiologi- cur between receptors sharing significant sequence homology. cal relevance of GPCR dimerization is the possibility that the When considering the maximal BRET values obtained in the high expression levels used in most studies could cause spuri- present study, only the  AR-Rluc/ AR-GFP pair was found to ous interactions between receptors. If this were the case, one 1 1 be significantly different from the others. Indeed, the BRET would expect that increasing the total level of receptor expres- max obtained for this pair was 1.5 times that observed for the sion would lead to a proportional increase in dimer formation. other combinations (Table I). As indicated above, this could In contrast, if dimerization can occur independently of receptor indicate either that a larger proportion of  AR than  AR can overexpression, no such relation should be expected. To distin- 1 2 engage in dimerization or that the relative position of Rluc and guish between these two possibilities, BRET between  AR- GFP within the  AR homodimer is more permissive to energy Rluc and  AR-GFP was determined for expression levels rang- 1 2 transfer (shorter distance and/or better orientation of the di- ing from 0.44 to 46.6 pmol/mg of protein. As indicated by the poles). Given that the relative affinities between each of the saturation curves presented above (see also Ref. 23), such com- 44930 Energy Transfer Analysis of -Adrenergic Receptor Dimers TABLE I Parameters obtained from BRET saturation curves The BRET is the maximal BRET obtained for a given pair and the BRET corresponds to the concentration of acceptor giving 50% of the max 50 BRET . The results are expressed as the mean  S.E. of 6 –10 independent experiments and were derived from the data presented in Fig. 3. max BRET  S.E. BRET  S.E. max 60 AR-Rluc/ AR-GFP 0.40  0.015 1.4  0.26 n  10 1 1 AR-Rluc/ AR-GFP 0.26  0.015 0.76  0.16 n  7 1 2 AR-Rluc/ AR-GFP 0.28  0.029 2.2  0.77 n  6 2 1 AR-Rluc/ AR-GFP 0.24  0.015 1.2  0.33 n  7 2 2 FIG.5. Schematic representation of the estimated percentage of  AR dimers in living cells. Corrected BRET saturation curves indicated that a BRET of 0.237 could be obtained for the  AR-Rluc/ max 2 FIG.4. Effect of receptor expression levels on BRET deter- AR-GFP pair. Assuming a free equilibrium between the two con- mined at approximate equimolar [donor]/[acceptor] ratios. structs, only 50% of the dimers should contribute to the BRET signal HEK-293T cells were co-transfected with increasing quantities of  AR- when donor and acceptor are expressed at equimolar concentrations. Rluc and  AR-GFP. The transfection conditions were established so Thus a BRET signal of 0.119 would be expected if 100% of the  AR that both receptor constructs were expressed at approximate equimolar existed as dimers. The average BRET value of 0.0974 experimentally levels, as determined by the total fluorescence and luminescence of the observed upon expression of  AR-Rluc and  AR-GFP at 1:1 ratio 2 2 fusion proteins. BRET is expressed as a function of the total number of (n  20) indicates that 82  10% of the total  AR population exist as receptor expressed. The amounts of  AR-GFP and  AR-Rluc are also 2 2 dimers. indicated (right part of the panel). The data presented were obtained for a total of 23 independent transfections. The BRET values were then over those reported for human heart tissue (0.080 pmol/mg) grouped according to the expression levels determined. When the data from more than two transfections were grouped, BRET values are (29 –33). Taken together, these data therefore suggest that presented as the mean  S.E. (n  3–5). dimerization can occur at physiologically relevant expression levels and that overexpression is not responsible for this proc- parison can only be carried out if the donor/acceptor ratio is ess. Consistent with this notion, homo- and heterodimerization maintained constant for the different expression levels tested. have been documented in native tissues for a few endogenously For this purpose, the transfection experiments were set up so expressed GPCR using co-immunoprecipitation (34) or Western as to obtain equimolar expression of the  AR-Rluc and  AR- 2 2 blot analysis (35– 40). Unfortunately, the lack of adequate an- GFP. The expression levels were monitored by measuring the tibodies for co-immunoprecipitating native receptors does not total luminescence and fluorescence signals, and the receptor allow such experiments for the endogenously expressed  and number was calculated using the equations derived from Fig. 2. AR. As shown in Fig. 4, constant BRET signals were obtained for In addition to confirming that the BRET signal observed did total  AR levels ranging from 1.4 to 26 pmol/mg when 2 not result from an artifact of overexpression, the data pre- equimolar concentrations of  AR-Rluc and  AR-GFP were 2 2 sented in Fig. 4 make possible some estimates of the proportion expressed. This indicates that a similar percentage of receptors of receptors engaged in dimerization. Indeed, BRET saturation engage in dimer formation over a 20-fold range in expression experiments presented in Fig. 3 have already indicated that the levels. The fact that the BRET signal is largely independent maximal level of BRET that can be obtained when all  AR- from receptor density also confirms that it originates from Rluc are bound to a  AR-GFP partner is 0.237  0.015 (see the dimerization and not random collision events (23). Increases in reported curve in Fig. 5). Assuming a free equilibrium between the BRET signals were observed only when the total  AR was 2 the Rluc and GFP fusion proteins, one would predict that, at expressed at 47 pmol/mg or above. Such increases in BRET at equimolar concentration of the two partners, only 50% of the these very high expression levels may result from random dimers can produce BRET ( AR-Rluc/ AR-GFP dimer) while 2 2 collisions between evenly dispersed donor and acceptor mole- the other half cannot ( AR-Rluc/ AR-Rluc and  AR-GFP/ 2 2 2 cules already engaged in dimers (23), or they could represent AR-GFP dimers). It follows that, if 100% of the receptors artifactual aggregation occurring only at unusually high recep- form dimers, the maximal BRET value observed at equimolar tor numbers. Interestingly, 47 pmol/mg corresponds to a recep- expression levels should be 0.1185 (i.e. 0.237  2). An experi- tor surface density of 2.4 receptors/10,000 Å representing an mental value of 0.0974  0.006 (Fig. 5) was obtained when average distance of less than 100 Å between each receptor averaging 20 BRET values recorded for equimolar expression dimer: a distance that would be permissive to bystander BRET. of  AR-Rluc and -GFP fusions at total expression levels vary- These results emphasize the importance of carefully monitor- ing from 1.3 to 26 pmol/mg. Based on these calculations, ing receptor expression levels in these types of studies. 82  10% of the cellular contingent of  AR exist as dimers. The lowest level of receptor expression allowing the detection CONCLUSION of BRET in the present study was 0.3 pmol/mg of protein (see legend of Fig. 3). Such expression level is comparable to those Our results clearly show that, in addition to forming ho- observed in dog heart tissue (0.5 pmol/mg) and only 3.75-fold modimers,  - and  -adrenergic receptors form heterodimers 1 2 Energy Transfer Analysis of -Adrenergic Receptor Dimers 44931 Y. C. (2000) Science 288, 154 –157 at nearly physiological expression levels. The similar propen- 13. Ramsay, D., Kellett, E., McVey, M., Rees, S., and Milligan, G. (2002) Biochem. sity of the receptor subtypes to form homo- and heterodimers J. 365, 429 – 440 (i.e. the comparable BRET found) could have an important 14. Hebert, T. E., Moffett, S., Morello, J. P., Loisel, T. P., Bichet, D. G., Barret, C., and Bouvier, M. (1996) J. Biol. Chem. 271, 16384 –16392 impact on the  and  AR profile in cells co-expressing the two 1 2 15. Xu, J., Paquet, M., Lau, A. G., Wood, J. D., Ross, C. A., and Hall, R. A. (2001) subtypes. Indeed, relatively modest changes in the expression J. Biol. Chem. 276, 41310 – 41317 16. George, S. R., Lee, S. P., Varghese, G., Zeman, P. R., Seeman, P., Ng, G. Y., and of one subtype should have repercussions not only on the rel- O’Dowd, B. F. (1998) J. Biol. Chem. 273, 30244 –30248 ative proportion of the heterodimer but also on the amount of 17. Petrecca, K., and Shrier, A. (1998) J. Anat. 192, 517–528 homodimers of each subtype. However, the total number of 18. Boivin, V., Jahns, R., Gambaryan, S., Ness, W., Boege, F., and Lohse, M. J. (2001) Kidney Int. 59, 515–531 receptors engaged in dimerization should not be affected, be- 19. Sano, M., Yoshimasa, T., Yagura, T., and Yamamoto, I. (1993) Life Sci. 52, cause we found that the ratio of dimer/total receptor is inde- 1063–1070 pendent of the expression levels and remains stable at 80% 20. Frielle, T., Collins, S., Daniel, K., Caron, M. G., Lefkowitz, R. J., and Kobilka, B. K. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 7920 –7924 for receptor concentrations spanning a 20-fold range. Although 21. Mellon, P. L., Parker, V., Gluzman, Y., and Maniatis, T. (1981) Cell 27, several studies suggested that GPCR can exist as constitutive 279 –288 22. Xu, Y., Piston, D. W., and Johnson, C. H. (1999) Proc. Natl. Acad. Sci. U. S. A. dimers (2), the relatively high proportion of dimeric receptor 96, 151–156 found in the present study is the first indication that dimers 23. Kenworthy, A. K., and Edidin, M. (1998) J. Cell Biol. 142, 69 – 84 may be the predominant species under basal conditions. The 24. Zacharias, D. A., Violin, J. D., Newton, A. C., and Tsien, R. Y. (2002) Science 296, 913–916 currently available techniques do not make assessment possi- 25. Veatch, W., and Stryer, L. (1977) J. Mol. Biol. 113, 89 –102 ble whether constitutive homo- and heterodimerization occurs 26. Brodde, O. E., Karad, K., Zerkowski, H. R., Rohm, N., and Reidemeister, J. C. to a similar extent in cells endogenously expressing  and (1983) Circ. Res. 53, 752–758 27. del Monte, F., Kaumann, A. J., Poole-Wilson, P. A., Wynne, D. G., Pepper, J., AR. However, the fact that various cardiovascular diseases and Harding, S. E. (1993) Circulation 88, 854 – 863 are associated with changes in the relative expressions of the 28. Jordan, B. A., Trapaidze, N., Gomes, I., Nivarthi, R., and Devi, L. A. (2001) two subtypes in heart tissues could have important impacts on Proc. Natl. Acad. Sci. U. S. A. 98, 343–348 29. Laurent, C. E., Cardinal, R., Rousseau, G., Vermeulen, M., Bouchard, C., the repertoire of homo- and heterodimers expressed in a given Wilkinson, M., Armour, J. A., and Bouvier, M. (2001) Am. J. Physiol. 280, cell (41– 44). Future studies will thus be required to assess the R355–R364 possible functional and pathophysiological consequences of 30. Steinfath, M., Danielsen, W., von der, L. H., Mende, U., Meyer, W., Neumann, J., Nose, M., Reich, T., Schmitz, W., and Scholz, H. (1992) Br. J. Pharmacol. AR homo- and heterodimerization. 105, 463– 469 31. Bristow, M. R. (1993) J. Am. Coll. Cardiol. 22, 61A–71A Acknowledgments—We are thankful to BioSignal Packard and in 32. Brodde, O. E., Vogelsang, M., Broede, A., Michel-Reher, M., Beisenbusch- particular to Mireille Caron and Erik Joly for their kind gift of the Schafer, E., Hakim, K., and Zerkowski, H. R. (1998) J. Cardiovasc. Phar- AR-Rluc plasmid. We also thank Dr. Ste ´ phane Aubry for his help in 1 macol. 31, 585–594 the mathematical modeling of bystander BRET as well as Drs. Monique 33. Brodde, O. E., and Michel, M. C. (1999) Pharmacol. Rev. 51, 651– 690 34. Kaupmann, K., Malitschek, B., Schuler, V., Heid, J., Froestl, W., Beck, P., Lagace ´ , Mostafa Mashayekhi, and M. Martin Audet for helpful Mosbacher, J., Bischoff, S., Kulik, A., Shigemoto, R., Karschin, A., and discussions. Bettler, B. (1998) Nature 396, 683– 687 35. Ng, G. Y., O’Dowd, B. F., Lee, S. P., Chung, H. T., Brann, M. R., Seeman, P., REFERENCES and George, S. R. (1996) Biochem. Biophys. Res. Commun. 227, 200 –204 1. Bouvier, M. (2001) Nat. Rev. Neurosci. 2, 274 –286 36. Ciruela, F., Casado, V., Mallol, J., Canela, E. I., Lluis, C., and Franco, R. (1995) 2. Angers, S., Salahpour, A., and Bouvier, M. (2002) Annu. Rev. Pharmacol. J. Neurosci. Res. 42, 818 – 828 Toxicol. 42, 409 – 435 37. Vila-Coro, A. J., Mellado, M., Martin, d. A., Lucas, P., del Real, G., Martinez, 3. Angers, S., Salahpour, A., Joly, E., Hilairet, S., Chelsky, D., Dennis, M., and A., and Rodriguez-Frade, J. M. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, Bouvier, M. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 3684 –3689 3388 –3393 4. Overton, M. C., and Blumer, K. J. (2000) Curr. Biol. 10, 341–344 38. Rodriguez-Frade, J. M., Vila-Coro, A. J., de Ana, A. M., Albar, J. P., Martinez, 5. Rocheville, M., Lange, D. C., Kumar, U., Sasi, R., Patel, R. C., and Patel, Y. C. A., and Mellado, M. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 3628 –3633 (2000) J. Biol. Chem. 275, 7862–7869 39. Vila-Coro, A. J., Rodriguez-Frade, J. M., Martin, d. A., Moreno-Ortiz, M. C., 6. Cornea, A., Janovick, J. A., Maya-Nunez, G., and Conn, P. M. (2001) J. Biol. Martinez, A., and Mellado, M. (1999) FASEB J. 13, 1699 –1710 Chem. 276, 2153–2158 40. Garzon, J., Juarros, J. L., Castro, M. A., and Sanchez-Blazquez, P. (1995) Mol. 7. Roess, D. A., Horvat, R. D., Munnelly, H., and Barisas, B. G. (2000) Endocri- Pharmacol. 47, 738 –744 nology 141, 4518 – 4523 41. Bristow, M. R., Ginsburg, R., Minobe, W., Cubicciotti, R. S., Sageman, W. S., 8. McVey, M., Ramsay, D., Kellett, E., Rees, S., Wilson, S., Pope, A. J., and Lurie, K., Billingham, M. E., Harrison, D. C., and Stinson, E. B. (1982) Milligan, G. (2001) J. Biol. Chem. 276, 14092–14099 N. Engl. J. Med. 307, 205–211 9. Kroeger, K. M., Hanyaloglu, A. C., Seeber, R. M., Miles, L. E., and Eidne, K. A. 42. Bristow, M. R., Ginsburg, R., Umans, V., Fowler, M., Minobe, W., Rasmussen, (2001) J. Biol. Chem. 276, 12736 –12743 R., Zera, P., Menlove, R., Shah, P., Jamieson, S., and Stinson, E. B. (1986) 10. Cheng, Z. J., and Miller, L. J. (2001) J. Biol. Chem. 276, 48040 – 48047 11. Ayoub, M. A., Couturier, C., Lucas-Meunier, E., Angers, S., Fossier, P., Circ. Res. 59, 297–309 Bouvier, M., and Jockers, R. (2002) J. Biol. Chem. 277, 21522–21528 43. Brodde, O. E. (1991) Pharmacol. Rev. 43, 203–242 12. Rocheville, M., Lange, D. C., Kumar, U., Patel, S. C., Patel, R. C., and Patel, 44. Nguyen, C. T., De Champlain, J., and Bouvier, M. (1995) Biomed. J. 2, 34 – 42 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Biological Chemistry Unpaywall

Quantitative Assessment of β1- and β2-Adrenergic Receptor Homo- and Heterodimerization by Bioluminescence Resonance Energy Transfer

Mercier, Jean-François; Salahpour, Ali; Angers, Stéphane; Breit, Andreas; Bouvier, Michel
Journal of Biological Chemistry – Nov 1, 2002
Download PDF
7 pages

Loading next page...
 
/lp/unpaywall/quantitative-assessment-of-1-and-2-adrenergic-receptor-homo-and-qGndH7vojN

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0021-9258
DOI
10.1074/jbc.m205767200
Publisher site
See Article on Publisher Site

Abstract

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 277, No. 47, Issue of November 22, pp. 44925–44931, 2002 © 2002 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Quantitative Assessment of  - and  -Adrenergic Receptor 1 2 Homo- and Heterodimerization by Bioluminescence Resonance Energy Transfer* Received for publication, June 11, 2002, and in revised form, August 23, 2002 Published, JBC Papers in Press, September 19, 2002, DOI 10.1074/jbc.M205767200 ¶ ¶ Jean-Franc ¸ ois Mercier‡§, Ali Salahpour‡ , Ste ´ phane Angers , Andreas Breit, and Michel Bouvier From the De ´partement de Biochimie and Groupe de Recherche sur le Syste `me Nerveux Autonome, Universite ´ de Montre ´al, Montre ´al, Quebec H3C 3J7, Canada and 2). In most instances, co-immunoprecipitation was used as Quantitative bioluminescence resonance energy transfer (BRET) analysis was applied to the study of  - the primary experimental evidence supporting the existence of and  -adrenergic receptor homo- and heterodimeriza- such dimers. More recently, however, light resonance energy tion. To assess the relative affinity between each of the transfer techniques such as fluorescence and bioluminescence protomers, BRET saturation experiments were carried resonance energy transfer (FRET and BRET) were also used. out in HEK-293T cells.  - and  -adrenergic receptors 1 2 These “non-invasive” proximity-based assays confirmed that were found to have similar propensity to engage in ho- GPCR dimerization does not represent biochemical artifacts mo- and heterotropic interactions suggesting that, at due to receptor solubilization and can occur in living cells. They equivalent expression levels of the two receptor sub- have been used to demonstrate homodimerization of the  - types, an equal proportion of homo- and heterodimers adrenergic (3), the yeast alpha mating factor (4), the SST5 would form. Analysis of the data also revealed that, at somatostatin (5), the gonadotropin releasing hormone (6), the equimolar expression levels of energy donor and accep- luteinizing hormone (7), the -opioid (8), the thyrotropin-re- tor, more than 80% of the receptor molecules exist as leasing hormone (9), the cholecystokinin (10), and the melato- dimers and that this high incidence of receptor dimer- nin (11) receptors as well as heterodimerization between soma- ization is insensitive to receptor density for expression tostatin receptor subtypes (5), somatostatin and dopamine levels varying between 1.4 and 26.9 pmol of receptor/mg receptors (12), melatonin receptor subtypes (11), and opioid of membrane protein. Taken together, these results in- receptor subtypes (13). dicate that most of the receptors expressed in cells exist An advantage of BRET and FRET over co-immunoprecipita- as constitutive dimers and that, at least in undifferenti- ated fibroblasts, the proportion of homo- and het- tion approaches lies in the more quantitative nature of the erodimers between the closely related  - and  -adre- assay. However, relatively few studies exploited this quantita- 1 2 nergic receptors is determined by their relative levels of tive potential for the study of GPCR dimerization. For the expression. melatonin receptors, Ayoub et al. (11) recently used BRET competition assays to determine that the transfer of energy resulted from the formation of dimers and not of higher order G protein-coupled receptors (GPCRs) represent the largest oligomers. They also showed that ligand binding did not alter family of transmembrane receptors involved in cell signaling. the dimerization state of the receptors. However, other ques- In the past few years, many studies indicated that GPCR tions that could theoretically be addressed by quantitative dimerization can occur between two identical receptors (ho- energy transfer analysis, such as the relative affinity of the modimerization), between two different receptor subtypes of dimer partners for each other and the relative proportion of the same family, or even between receptors that are only dis- receptors engaging in dimer formation, have not yet been tantly related (heterodimerization) (for a review, see Refs. 1 addressed. - and  -adrenergic receptors ( AR and  AR) have pre- 2 1 2 1 viously been shown to exist as homodimers (3, 8, 14, 15). The * This work was supported in part by grants from the Canadian high level of sequence identity existing in domains proposed to Institute for Health Research (CIHR) and the Heart and Stroke Foun- contribute to the dimerization interface (i.e. transmembrane dation of Canada (to M. B.). The costs of publication of this article were helices) (14, 16) makes them a system of choice to study their defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 potential heterodimerization and the relative affinity of the U.S.C. Section 1734 solely to indicate this fact. protomers within homo- and heterotropic complexes. Although ‡ Both authors contributed equally to this work. direct in vivo demonstration for the co-localization of the two § Held a studentship from the Fonds de la Recherche en Sante´du receptor subtypes in the same cell is still lacking, the presence Que ´ bec. Hold studentships from the CIHR. of  and  AR in the same cell types has been taken as 1 2 Holds a Canada Research Chair in Molecular and Cellular Pharma- evidence for their co-expression in transitional and mid-nodal cology. To whom correspondence should be addressed: De ´ pt. de Bio- cells of the atrio-ventricular node, nerve processes of the atrio- chimie, Universite ´ de Montre ´ al, C.P. 6128, Succursale Centre-Ville, ventricular and ventricular conduction systems, as well as in Montre ´ al, Quebec H3C 3J7, Canada. Tel.: 514-343-6372; Fax: 514-343- 2210; E-mail: [email protected]. vascular smooth muscle cells of the kidney (17, 18). The fact The abbreviations used are: GPCR, G protein-coupled receptor; that the two receptor subtypes are also found together in a FRET, fluorescence resonance energy transfer; BRET, bioluminescence large number of tissues, including liver, lung, and fat (19), resonance energy transfer; AR, adrenergic receptor; GFP, green fluo- gives further support to the idea that heterodimerization could rescent protein; PBS, phosphate-buffered saline; Rluc, Renilla lucifer- ase; CYP, cyanopindolol. occur in native tissues and warrants investigations aiming to This paper is available on line at http://www.jbc.org 44925 This is an Open Access article under the CC BY license. 44926 Energy Transfer Analysis of -Adrenergic Receptor Dimers Following harvesting, cells were distributed in 96-well microplates assess the likelihood of such intermolecular complexes. (white Optiplate from Packard Bioscience) at a density of 100,000 Here, quantitative BRET approaches were applied to the cells per well. DeepBlueC was added at a final concentration of 5 M, study of  - and  -adrenergic receptor dimerization in a het- 1 2 and readings were collected using a modified Top-count apparatus erologous mammalian expression system. In particular, BRET (BRETCount, Packard Bioscience) that allows the sequential integra- saturation experiments were carried out to estimate the rela- tion of the signals detected in the 370- to 450-nm and 500- to 530-nm tive affinity of each receptor subtype to engage into homo- and windows using filters with the appropriate band pass (Chroma). The BRET signal is determined by calculating the ratio of the light emitted heterotropic interactions. The influence of receptor density on by the Receptor-GFP (500 –530 nm) over the light emitted by the Re- the proportion of receptor molecules forming dimers was also ceptor-Rluc (370 – 450 nm). The values were corrected by subtracting assessed. Here, we report that  AR and  AR can form homo- 1 2 the background signal detected when the Receptor-Rluc constructs and heterodimers and that the two receptors have similar were expressed alone. affinities for each other and for themselves, suggesting that Fluorescence and Luminescence Measurement heterodimers are likely to form in cells expressing both sub- types. The proportion of receptor molecules forming dimers was Cells were distributed in 96-well microplates (white Costar plate with clear bottom) at a density of 100,000 cells per well. The total also found to be greater than 80% at low receptor density and fluorescence of cells was measured using a FluoroCount (Packard Bio- was constant over a 20-fold expression range. science) with an excitation filter at 400 nm, an emission filter at 510 EXPERIMENTAL PROCEDURES nm, and the following parameters: gain, 1; photo multiplicator tube, 1100 V; time, 1.0 s. After the fluorescence measurement, the same cells Receptor Constructs were incubated for 10 min with Coelenterazine H (Molecular Probe) at AR-GFP10 —The  AR coding sequence without its stop codon was 1 1 a final concentration of 5 M, and the total luminescence of cells was amplified from the pBC12BI-human  AR plasmid (20) using sense and measured using a LumiCount (Packard Bioscience) with the following antisense primers harboring unique SacI and AgeI sites. The fragment parameters: gain, 1; photo multiplicator tube, 700 V; time, 0.5 s. For was then subcloned in-frame into the SacI/AgeI site of the blue variant both measurements, the mean of duplicate wells was calculated. The GFP-sapphire vector (pGFP-N1-Sapphire, Packard Bioscience) to give total fluorescence was then divided by the background determined in the plasmid pGFP-N1- AR-Sapphire. Finally, the GFP-Sapphire was wells containing untransfected cells. Fluorescence was expressed in replaced by a green GFP variant (GFP10) containing the following -fold over background. The background was negligible for the lumines- mutations: P64L, S147P, and S202P. For this purpose, an AgeI/BsrgI cence measurements, so they were expressed as absolute values. fragment of the GFP10 variant was subcloned into the AgeI/BsrgI site of pGFP-N1- AR-Sapphire to finally yield pGFP-N1- AR-GFP10. Radioligand Binding Assay 1 1 AR-GFP10 —The GFP10 AgeI/BsrgI fragment was subcloned into Forty-eight hours after transfection, 10,000 cells (2 g of proteins) the AgeI/BsrgI site of pGFP-N1-His AR-YFP (3). For simplicity, were incubated in a final volume of 500 l of PBS containing 0.1% GFP10 will be referred to as GFP in the remainder of the text. bovine serum albumin with a saturating concentration (250 pM)ofthe AR-Rluc—The pcDNA3.1- AR:6aa:hRluc was a generous gift 1 1 125 125 -adrenergic antagonist [ I]cyanopindolol ([ I]CYP). Nonspecific from BioSignal Packard. This fusion protein contains a linker of six binding was determined as the residual binding observed in the pres- amino acids (YGPPGS) linking the carboxyl tail of the human  AR and ence of 10 M alprenolol (Sigma). Binding reactions were carried out at the humanized Rluc. room temperature for 90 min and stopped by rapid filtration over AR-Rluc—The humanized Rluc coding sequence (pRluc(h), Pack- Whatman GF/C glass-fiber filters. Receptor densities are expressed in ard Bioscience) was amplified using sense and antisense primers and femtomoles of receptor per milligram of total cell proteins assuming one subcloned into the PCR Blunt II Topo vector (Invitrogen). The hRluc binding site per receptor molecule. Linear regression curves between fragment was excised by digestion with KpnI/XbaI and subcloned into the luminescence and fluorescence signals and the number of receptor the KpnI/XbaI-digested pcDNA3.1 Zeo vector to generate the determined by radio-ligand binding were then generated from cells pcDNA3.1 Zeo/hRluc plasmid. The human His AR coding sequence expressing each of the constructs individually. To determine receptor was amplified without its stop codon using sense and antisense prim- surface density, the surface of HEK-293T cells was determined by ers. The PCR product was subcloned into PCR Blunt II Topo Vector, measuring the average length and width of the cells under phase- then excised by double digestion with HindIII/KpnI and ligated into the contrast microscopy. HindIII/KpnI-digested expression vector pcDNA3.1Zeo/hRluc. The re- sulting construct encodes a six-amino acid linker (GSGTGS) between Light Emission/Receptor Binding Correction Factor the carboxyl-terminal of the  AR and the humanized Rluc sequence. Given that the correlations between receptor numbers and the lumi- Cell Culture and Transfection nescence or fluorescence levels were intrinsic characteristics of each of the constructs, comparison between receptor densities derived from the HEK-293T cells were maintained in Dulbecco’s modified Eagle’s me- light measurements required a correcting factor. This correction was dium supplemented with 10% fetal bovine serum, 100 units/ml penicillin achieved using the linear regression generated for each constructs (see and streptomycin, 2 mML-glutamine (all from Wisent). For transfection legend of Fig. 2) and by normalizing the light emission/receptor number experiments, cells were seeded at a density of 2  10 cells per 100-mm as a function of the steeper slope factor obtained for both Rluc and GFP. dish and cultured for 24 h. Transient transfections were then performed These corrected receptor number values were then used to generate the using the calcium phosphate precipitation protocol (21). 24 h after trans- corrected BRET saturation curves presented in Fig. 3. fection, Dulbecco’s modified Eagle’s medium was replaced, and the cells were then cultured in the same medium for an additional 24 h. RESULTS AND DISCUSSION Forty-height hours post-transfection, cells were washed twice with Homo- and heterodimerization of the  and  AR were PBS, detached with PBS/EDTA and resuspended in PBS/glucose 0.1%. 1 2 On a routine basis, the protein concentration of the samples was deter- investigated by quantitative BRET analysis. For this purpose, mined to control for the number of cells using the Bradford assay kit human  and  AR cDNAs were fused at their carboxyl ter- 1 2 (Bio-Rad) with bovine serum albumin as a standard. minus to the energy donor Rluc and acceptor GFP. The affini- ties of the fusion proteins for the antagonist cyanopindolol and BRET Measurement the agonist isoproterenol as well as the potency of isoproterenol We have used a slight modification of the previously published BRET 2 to stimulate adenylyl cyclase were indistinguishable from assay (3, 22). The new BRET technology (BioSignal Packard) takes those of the wild-type receptors (data not shown). advantage of the spectral properties of a luciferase substrate known as Deep Blue coelenterazine (DeepBlueC, Packard Bioscience), which al- BRET and FRET approaches have been used in several stud- lows a better separation between the Renilla luciferase (Rluc) and the ies to assess GPCR homo- and heterodimerization (2). In most green fluorescent protein (GFP) emission spectra. Upon the catalytic cases, little attention has been paid to the ratio of donor/ degradation of DeepBlueC, the energy donor Rluc emits light with a acceptor molecules that was used in the assays, and thus the peak at 400 nm that allows the excitation of the energy acceptor, GFP. interpretation of the data remained rather qualitative. How- Once excited, GFP then re-emits fluorescence with a peak at 510 nm if ever, controlling this parameter is essential for proper quanti- the donor and acceptor molecules are within BRET-permissive distance (100 Å). tative analysis. Indeed, the level of energy transfer detected for Energy Transfer Analysis of -Adrenergic Receptor Dimers 44927 FIG.1. BRET saturation curves. HEK-293T cells were co-transfected with a constant DNA concentration of  AR- Rluc (A)or  AR-Rluc (B) and increasing DNA concentrations of  AR-GFP,  AR- 1 2 GFP, or soluble GFP. The BRET, total luminescence, and total fluorescence were measured 48 h after transfection. BRET levels are plotted as a function of the total fluorescence signal (-fold over back- ground) used as an index for the concen- tration of receptor-GFP constructs ex- pressed. The results are expressed as the mean  S.E. of 3–10 independent experi- ments carried out in triplicates. The curves were fitted using a non-linear re- gression equation assuming a single bind- ing site (GraphPad Prism). a given concentration of donor should rise with increasing the better resolution between the emission peaks of Rluc and concentration of the acceptor until all donor molecules are GFP with BRET discussed under “Experimental Procedures” engaged by an acceptor. It follows that the energy transfer is not the only difference between the two generations of BRET. should reach a plateau and that saturation curves could theo- In particular, the quantum yield of the Rluc/DeepBlueC coelen- retically be constructed. The maximal level reached should be a terazine couple is lower than that of Rluc/coelenterazine H, and function of the total number of dimers formed and of the dis- the extent of overlap between the emission spectra of Rluc and 2 1 tance between the donor and acceptor within the dimers, the excitation of the GFPs is better for BRET than BRET . whereas the concentration of acceptor giving 50% of energy These parameters can influence the sensitivity of the assays to transfer (BRET ) should be a reflection of the relative affinity detect small changes in distance between energy donors and of the acceptor fusion for the donor fusion proteins. Here, we acceptors. The notion that certain BRET configurations but not applied this theoretical framework to the study of  and  AR others allow the detection of conformational changes induced 1 2 homo- and heterodimerization by constructing BRET satura- by ligands is also well exemplified by the recent observation tion curves in cells co-transfected with a constant amount of that agonist and antagonist binding increased the BRET be- receptor-Rluc construct and increasing concentrations of the tween the melatonin MTR1-Rluc and MTR2-GFP but not receptor-GFP plasmids. As shown in Fig. 1, significant BRET between MTR2-Rluc and MTR1-GFP (11). signals were observed for the  AR/ AR and  AR/ AR pairs Although the curves generated concur with the theoretical 2 2 1 1 confirming previous findings that both receptor subtypes can behavior predicted above, quantitative analysis is complicated form homodimers (3, 14, 15). Albeit to a lower extent, co- by the lack of direct information provided by the fluorescence expression of the two subtypes also led to a sizable BRET signal and luminescence measurements on the precise concentration for the two transfer orientations (i.e.  AR-Rluc/ AR-GFP and of receptor molecules expressed. In the absence of such infor- 1 2 AR-Rluc/ AR-GFP). In all cases, BRET increased as a hy- mation, determining the relative affinity of the protomers for 2 1 perbolic function of the concentration of the GFP fusion con- each other, based on these saturation isotherms, would require struct added (assessed by the fluorescence emitted upon direct that the correlations between light emission and the number of excitation at 400 nm) reaching an asymptote at the highest receptor-GFP and -Rluc fusion molecules are linear and iden- concentrations used. Co-expression of  or  AR-Rluc with tical for the two receptors considered. To directly test this 2 1 soluble GFP led to marginal signals that increased linearly supposition, cells were transfected with increasing concentra- with increasing amount of GFP added. Stimulation with the tions of receptor-GFP and -Rluc constructs. For each DNA agonist isoproterenol did not promote any consistent change in concentrations, the total expression level of the receptors was the BRET saturation curves (data not shown) indicating that determined using the lipophilic ligand [ I]CYP, whereas the the dimers form constitutively and that receptor activation total luminescence and total fluorescence emitted by the Rluc does not affect their oligomerization state. However, one cannot and GFP fusion proteins were measured following addition of exclude the possibility that agonist stimulation could promote the Rluc substrate coelenterazine H and direct excitation of the assembly/disassembly cycles that do not affect the steady-state GFP at 400 nm, respectively. Fig. 2 illustrates the correlation proportion of receptors engaged in dimers. The modest agonist- obtained between the number of total binding sites and either promoted increase in BRET, previously reported for the  AR the luminescence or fluorescence emitted by each of the recep- homodimer using BRET (3), most likely reflected conforma- tor fusion molecules. Even though the regression curves were tional changes that could not be detected using BRET . Indeed, highly linear, their slopes were different for the two receptors 44928 Energy Transfer Analysis of -Adrenergic Receptor Dimers receptor-GFP in a given surface was modeled for the  AR homodimer. For this purpose, the receptor surface density was estimated by microscopic measurements of the HEK-293T cell surface (240 m ) and determining the receptor number using the equations in Fig. 2. The average receptor-Rluc surface density in our BRET saturation experiments was found to be 3000 receptors/m . Fig. 3D shows simulations of bystander BRET carried out for receptor-Rluc levels of 30, 300, and 3000 receptors/m . This was accomplished using a Monte-Carlo approach that assumes a random and uniform distribution of the receptors on the calculated surface. Considering that the diameter of GPCR has been estimated to be 50 Å, assuming that, in the case of bystander BRET, receptor molecules would not be intertwined, and given that the R (the distance at which the energy transfer reaches 50% of its maximum) is 50 Å,we estimated the BRET permissive surface as 50 Å . As shown in Fig. 3D, the predicted bystander BRET curves differed sig- nificantly from the BRET saturation curve obtained experi- mentally. Indeed, they progressed quasi-linearly up to recep- tor-GFP surface density for which experimental BRET values have already reached saturation. It should be noted that the difference between the experimental BRET saturation curve and the bystander BRET modeled for a receptor surface density of 3000/m is most likely being underestimated, because the FIG.2. Linear relationship between total luminescence (A)or surface measurements did not account for plasma membrane total fluorescence (B) and receptor density. HEK-293T cells were details such as microvilli that would contribute to increased transfected with increasing DNA concentrations of  or  AR-Rluc (A) 1 2 cell surface area. Moreover, the number of calculated receptors or -GFP (B) fusions. Total receptor density was determined by radioli- gand binding assays using [ I]CYP as the tracer. The total fluores- reflects the total cellular receptor content and not only those cence was measured following excitation at 400 nm and detection at 510 present at the cell surface. The aspect of the experimental nm, whereas the total luminescence was recorded following the addition BRET saturation curve thus suggests a clustering of the energy of the Rluc substrate, coelenterazine H. The linear regression curve was donor and acceptor molecules resulting form receptor oligomer- generated using GraphPad Prism. The dotted line corresponds to the ization rather than from their random collisions. extrapolation of the linear regression for the  AR-GFP. The linear regression equations used to calculate the receptor amount for a given In an effort to distinguish between dimers and higher order luminescence or fluorescence intensity are as follows:  AR-Rluc: y oligomers, we modeled the BRET saturation curves using an 1.5540(x)  20.00;  AR-Rluc: y  0.3192(x)  20.00;  AR-GFP: y 2 1 equation (modified from Ref. 25) that describes the probability 0.0002902(x)  1.00;  AR-GFP: y  0.0001004(x) 1.00. of forming BRET competent complexes as a function of the number of receptors within a complex (i.e. dimer versus trimer versus tetramer, etc). Fig. 3D shows that the experimental considered. Indeed, the GFP and Rluc signals increased more curve fits better to the theoretical dimer curve than that pre- rapidly with receptor number for the  than the  AR. Al- 1 2 dicted for a trimer, suggesting that the BRET obtained from though the exact cause for this difference remains unknown, co-expression of  AR-Rluc and  AR-GFP results from the this obviously complicates the analysis of the data, because this 2 2 formation of dimeric complexes of this receptor. difference must be taken into account to assess the relative When comparing the BRET saturation curves obtained for affinity of the receptors for each other in BRET saturation the  and  AR homo- and heterodimers (Fig. 3, A and B), curves. The linear regression equations derived from Fig. 2 1 2 were thus used to transform the luminescence and fluorescence similar BRET values were obtained for all pairs considered (Table I), indicating that the receptors had similar relative value in receptor number. Although the BRET saturation affinities for one another. This has important implications, curves were carried out using a fixed concentration of the Rluc fusion partners, co-transfecting an increasing quantity of the because it suggests that, under basal conditions,  and  AR 1 2 homo- and heterodimers have a similar probability of forming GFP constructs introduces some levels of variability in the amount of receptor-Rluc expressed in each case. To rule out the when the two receptors are heterologously expressed in HEK- influence of this variable, the BRET levels were plotted as a 293T cells. Co-localization of the two -adrenergic receptors has been documented in numerous tissues (19, 26, 27). Al- function of the ratio between the receptor-GFP/receptor-Rluc numbers. though it has not been easy to experimentally demonstrate co-localization in the same cells, their expression in the same As shown in Fig. 3, the BRET saturation curves generated following these corrections also behaved as hyperbolic func- cell types has been taken as evidence for the co-existence of the tions reaching a saturation level. The aspect of these curves two receptor subtypes (17, 18). Thus, the equal chance of form- ing homo- and heterodimers found in the present study could greatly contrasts with that of the curve predicted if the ob- served BRET resulted from random collisions promoted by a have important physiological consequences. However, future high receptor density. Indeed, a quasi-linear curve would be studies will be required to determine if heterodimerization can expected if such “bystander” BRET was taking place (23, 24). occur in native tissues. The schematic illustration of the predicted distribution of the The high likelihood of heterodimer formation between ho- energy donor and acceptor in the case of dimerization versus mologous receptors has also been suggested by the recent ob- random collision, presented in Fig. 3C, allows for an intuitive servations of Ramsay et al. (13) showing that homo- and het- appreciation of the difference between the two situations. To erodimerization of the - and -opioid receptors occurs at allow a more quantitative comparison, the progression of by- comparable levels of receptor expression. Although het- stander BRET as a function of increasing concentration of the erodimerization could also be observed between the distantly Energy Transfer Analysis of -Adrenergic Receptor Dimers 44929 FIG.3. A and B, corrected BRET saturation curves. The fluorescence and luminescence data from Fig. 1 were transformed into receptor numbers using equations from Fig. 2 and taking the slope factor into account as indicated under “Experimental Procedures.” The BRET levels are plotted as a function of the ratio of [receptor-GFP]/[receptor-Rluc]. The total number of receptors expressed ranged as follows: AR/ AR, 0.8 –23 pmol/mg; 1 1 AR/ AR, 0.8 –17 pmol/mg;  AR/ AR, 0.3–37 pmol/mg; and  AR/ AR, 0.3– 8 pmol/mg. The curves were fitted using a non-linear regression 1 2 2 1 2 2 equation assuming a single binding site (GraphPad Prism). C, schematic representation of the distribution of the energy donor (Receptor-Rluc) and acceptor (Receptor-GFP) in the case of random collision (upper panel) versus dimerization (lower panel) when the density of energy donor is maintained constant while that of the acceptor is increased. D, modeling of the evolution of BRET resulting form random collision between energy donor and acceptor (“bystander BRET”) as a function of increased acceptor surface density for donor surface densities of 30, 300, and 3000 receptors/ m . We simulated this process using a Monte-Carlo approach based on the assumption of random distribution. For each incremental increase of receptor-GFP, we modeled the probability P of interaction with any receptor-Rluc as P  1  (1  p) , where p is defined as the ratio of the BRET permissive surface of the receptor-Rluc (50A ) over the total calculated surface of the cell and n is the number of remaining receptor-Rluc available. The experimental data of  AR-Rluc/ AR-GFP presented in B are also compared with the expected BRET saturation 2 2 curve for trimeric and dimeric complexes using an equation that describes the probability of forming BRET competent complexes, BRET competent n1 n1 n n1 n1 complex  (nd a  nda )/(nd  nd a  nda ), where n is the number of receptor molecule in the complex, d is the number of receptor-Rluc (energy donor), and a the number of receptor-GFP (energy acceptor). The curves expressed as percent maximal BRET are plotted as a function of the [receptor-GFP]/[receptor-Rluc] ratio. related opioid and  -adrenergic receptors (8, 28), this appar- partners were found to be very similar (Table I), the second ently takes place only at much higher receptor expression hypothesis is more likely. levels, indicating that heterodimerization is more likely to oc- One of the major concerns when considering the physiologi- cur between receptors sharing significant sequence homology. cal relevance of GPCR dimerization is the possibility that the When considering the maximal BRET values obtained in the high expression levels used in most studies could cause spuri- present study, only the  AR-Rluc/ AR-GFP pair was found to ous interactions between receptors. If this were the case, one 1 1 be significantly different from the others. Indeed, the BRET would expect that increasing the total level of receptor expres- max obtained for this pair was 1.5 times that observed for the sion would lead to a proportional increase in dimer formation. other combinations (Table I). As indicated above, this could In contrast, if dimerization can occur independently of receptor indicate either that a larger proportion of  AR than  AR can overexpression, no such relation should be expected. To distin- 1 2 engage in dimerization or that the relative position of Rluc and guish between these two possibilities, BRET between  AR- GFP within the  AR homodimer is more permissive to energy Rluc and  AR-GFP was determined for expression levels rang- 1 2 transfer (shorter distance and/or better orientation of the di- ing from 0.44 to 46.6 pmol/mg of protein. As indicated by the poles). Given that the relative affinities between each of the saturation curves presented above (see also Ref. 23), such com- 44930 Energy Transfer Analysis of -Adrenergic Receptor Dimers TABLE I Parameters obtained from BRET saturation curves The BRET is the maximal BRET obtained for a given pair and the BRET corresponds to the concentration of acceptor giving 50% of the max 50 BRET . The results are expressed as the mean  S.E. of 6 –10 independent experiments and were derived from the data presented in Fig. 3. max BRET  S.E. BRET  S.E. max 60 AR-Rluc/ AR-GFP 0.40  0.015 1.4  0.26 n  10 1 1 AR-Rluc/ AR-GFP 0.26  0.015 0.76  0.16 n  7 1 2 AR-Rluc/ AR-GFP 0.28  0.029 2.2  0.77 n  6 2 1 AR-Rluc/ AR-GFP 0.24  0.015 1.2  0.33 n  7 2 2 FIG.5. Schematic representation of the estimated percentage of  AR dimers in living cells. Corrected BRET saturation curves indicated that a BRET of 0.237 could be obtained for the  AR-Rluc/ max 2 FIG.4. Effect of receptor expression levels on BRET deter- AR-GFP pair. Assuming a free equilibrium between the two con- mined at approximate equimolar [donor]/[acceptor] ratios. structs, only 50% of the dimers should contribute to the BRET signal HEK-293T cells were co-transfected with increasing quantities of  AR- when donor and acceptor are expressed at equimolar concentrations. Rluc and  AR-GFP. The transfection conditions were established so Thus a BRET signal of 0.119 would be expected if 100% of the  AR that both receptor constructs were expressed at approximate equimolar existed as dimers. The average BRET value of 0.0974 experimentally levels, as determined by the total fluorescence and luminescence of the observed upon expression of  AR-Rluc and  AR-GFP at 1:1 ratio 2 2 fusion proteins. BRET is expressed as a function of the total number of (n  20) indicates that 82  10% of the total  AR population exist as receptor expressed. The amounts of  AR-GFP and  AR-Rluc are also 2 2 dimers. indicated (right part of the panel). The data presented were obtained for a total of 23 independent transfections. The BRET values were then over those reported for human heart tissue (0.080 pmol/mg) grouped according to the expression levels determined. When the data from more than two transfections were grouped, BRET values are (29 –33). Taken together, these data therefore suggest that presented as the mean  S.E. (n  3–5). dimerization can occur at physiologically relevant expression levels and that overexpression is not responsible for this proc- parison can only be carried out if the donor/acceptor ratio is ess. Consistent with this notion, homo- and heterodimerization maintained constant for the different expression levels tested. have been documented in native tissues for a few endogenously For this purpose, the transfection experiments were set up so expressed GPCR using co-immunoprecipitation (34) or Western as to obtain equimolar expression of the  AR-Rluc and  AR- 2 2 blot analysis (35– 40). Unfortunately, the lack of adequate an- GFP. The expression levels were monitored by measuring the tibodies for co-immunoprecipitating native receptors does not total luminescence and fluorescence signals, and the receptor allow such experiments for the endogenously expressed  and number was calculated using the equations derived from Fig. 2. AR. As shown in Fig. 4, constant BRET signals were obtained for In addition to confirming that the BRET signal observed did total  AR levels ranging from 1.4 to 26 pmol/mg when 2 not result from an artifact of overexpression, the data pre- equimolar concentrations of  AR-Rluc and  AR-GFP were 2 2 sented in Fig. 4 make possible some estimates of the proportion expressed. This indicates that a similar percentage of receptors of receptors engaged in dimerization. Indeed, BRET saturation engage in dimer formation over a 20-fold range in expression experiments presented in Fig. 3 have already indicated that the levels. The fact that the BRET signal is largely independent maximal level of BRET that can be obtained when all  AR- from receptor density also confirms that it originates from Rluc are bound to a  AR-GFP partner is 0.237  0.015 (see the dimerization and not random collision events (23). Increases in reported curve in Fig. 5). Assuming a free equilibrium between the BRET signals were observed only when the total  AR was 2 the Rluc and GFP fusion proteins, one would predict that, at expressed at 47 pmol/mg or above. Such increases in BRET at equimolar concentration of the two partners, only 50% of the these very high expression levels may result from random dimers can produce BRET ( AR-Rluc/ AR-GFP dimer) while 2 2 collisions between evenly dispersed donor and acceptor mole- the other half cannot ( AR-Rluc/ AR-Rluc and  AR-GFP/ 2 2 2 cules already engaged in dimers (23), or they could represent AR-GFP dimers). It follows that, if 100% of the receptors artifactual aggregation occurring only at unusually high recep- form dimers, the maximal BRET value observed at equimolar tor numbers. Interestingly, 47 pmol/mg corresponds to a recep- expression levels should be 0.1185 (i.e. 0.237  2). An experi- tor surface density of 2.4 receptors/10,000 Å representing an mental value of 0.0974  0.006 (Fig. 5) was obtained when average distance of less than 100 Å between each receptor averaging 20 BRET values recorded for equimolar expression dimer: a distance that would be permissive to bystander BRET. of  AR-Rluc and -GFP fusions at total expression levels vary- These results emphasize the importance of carefully monitor- ing from 1.3 to 26 pmol/mg. Based on these calculations, ing receptor expression levels in these types of studies. 82  10% of the cellular contingent of  AR exist as dimers. The lowest level of receptor expression allowing the detection CONCLUSION of BRET in the present study was 0.3 pmol/mg of protein (see legend of Fig. 3). Such expression level is comparable to those Our results clearly show that, in addition to forming ho- observed in dog heart tissue (0.5 pmol/mg) and only 3.75-fold modimers,  - and  -adrenergic receptors form heterodimers 1 2 Energy Transfer Analysis of -Adrenergic Receptor Dimers 44931 Y. C. (2000) Science 288, 154 –157 at nearly physiological expression levels. The similar propen- 13. Ramsay, D., Kellett, E., McVey, M., Rees, S., and Milligan, G. (2002) Biochem. sity of the receptor subtypes to form homo- and heterodimers J. 365, 429 – 440 (i.e. the comparable BRET found) could have an important 14. Hebert, T. E., Moffett, S., Morello, J. P., Loisel, T. P., Bichet, D. G., Barret, C., and Bouvier, M. (1996) J. Biol. Chem. 271, 16384 –16392 impact on the  and  AR profile in cells co-expressing the two 1 2 15. Xu, J., Paquet, M., Lau, A. G., Wood, J. D., Ross, C. A., and Hall, R. A. (2001) subtypes. Indeed, relatively modest changes in the expression J. Biol. Chem. 276, 41310 – 41317 16. George, S. R., Lee, S. P., Varghese, G., Zeman, P. R., Seeman, P., Ng, G. Y., and of one subtype should have repercussions not only on the rel- O’Dowd, B. F. (1998) J. Biol. Chem. 273, 30244 –30248 ative proportion of the heterodimer but also on the amount of 17. Petrecca, K., and Shrier, A. (1998) J. Anat. 192, 517–528 homodimers of each subtype. However, the total number of 18. Boivin, V., Jahns, R., Gambaryan, S., Ness, W., Boege, F., and Lohse, M. J. (2001) Kidney Int. 59, 515–531 receptors engaged in dimerization should not be affected, be- 19. Sano, M., Yoshimasa, T., Yagura, T., and Yamamoto, I. (1993) Life Sci. 52, cause we found that the ratio of dimer/total receptor is inde- 1063–1070 pendent of the expression levels and remains stable at 80% 20. Frielle, T., Collins, S., Daniel, K., Caron, M. G., Lefkowitz, R. J., and Kobilka, B. K. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 7920 –7924 for receptor concentrations spanning a 20-fold range. Although 21. Mellon, P. L., Parker, V., Gluzman, Y., and Maniatis, T. (1981) Cell 27, several studies suggested that GPCR can exist as constitutive 279 –288 22. Xu, Y., Piston, D. W., and Johnson, C. H. (1999) Proc. Natl. Acad. Sci. U. S. A. dimers (2), the relatively high proportion of dimeric receptor 96, 151–156 found in the present study is the first indication that dimers 23. Kenworthy, A. K., and Edidin, M. (1998) J. Cell Biol. 142, 69 – 84 may be the predominant species under basal conditions. The 24. Zacharias, D. A., Violin, J. D., Newton, A. C., and Tsien, R. Y. (2002) Science 296, 913–916 currently available techniques do not make assessment possi- 25. Veatch, W., and Stryer, L. (1977) J. Mol. Biol. 113, 89 –102 ble whether constitutive homo- and heterodimerization occurs 26. Brodde, O. E., Karad, K., Zerkowski, H. R., Rohm, N., and Reidemeister, J. C. to a similar extent in cells endogenously expressing  and (1983) Circ. Res. 53, 752–758 27. del Monte, F., Kaumann, A. J., Poole-Wilson, P. A., Wynne, D. G., Pepper, J., AR. However, the fact that various cardiovascular diseases and Harding, S. E. (1993) Circulation 88, 854 – 863 are associated with changes in the relative expressions of the 28. Jordan, B. A., Trapaidze, N., Gomes, I., Nivarthi, R., and Devi, L. A. (2001) two subtypes in heart tissues could have important impacts on Proc. Natl. Acad. Sci. U. S. A. 98, 343–348 29. Laurent, C. E., Cardinal, R., Rousseau, G., Vermeulen, M., Bouchard, C., the repertoire of homo- and heterodimers expressed in a given Wilkinson, M., Armour, J. A., and Bouvier, M. (2001) Am. J. Physiol. 280, cell (41– 44). Future studies will thus be required to assess the R355–R364 possible functional and pathophysiological consequences of 30. Steinfath, M., Danielsen, W., von der, L. H., Mende, U., Meyer, W., Neumann, J., Nose, M., Reich, T., Schmitz, W., and Scholz, H. (1992) Br. J. Pharmacol. AR homo- and heterodimerization. 105, 463– 469 31. Bristow, M. R. (1993) J. Am. Coll. Cardiol. 22, 61A–71A Acknowledgments—We are thankful to BioSignal Packard and in 32. Brodde, O. E., Vogelsang, M., Broede, A., Michel-Reher, M., Beisenbusch- particular to Mireille Caron and Erik Joly for their kind gift of the Schafer, E., Hakim, K., and Zerkowski, H. R. (1998) J. Cardiovasc. Phar- AR-Rluc plasmid. We also thank Dr. Ste ´ phane Aubry for his help in 1 macol. 31, 585–594 the mathematical modeling of bystander BRET as well as Drs. Monique 33. Brodde, O. E., and Michel, M. C. (1999) Pharmacol. Rev. 51, 651– 690 34. Kaupmann, K., Malitschek, B., Schuler, V., Heid, J., Froestl, W., Beck, P., Lagace ´ , Mostafa Mashayekhi, and M. Martin Audet for helpful Mosbacher, J., Bischoff, S., Kulik, A., Shigemoto, R., Karschin, A., and discussions. Bettler, B. (1998) Nature 396, 683– 687 35. Ng, G. Y., O’Dowd, B. F., Lee, S. P., Chung, H. T., Brann, M. R., Seeman, P., REFERENCES and George, S. R. (1996) Biochem. Biophys. Res. Commun. 227, 200 –204 1. Bouvier, M. (2001) Nat. Rev. Neurosci. 2, 274 –286 36. Ciruela, F., Casado, V., Mallol, J., Canela, E. I., Lluis, C., and Franco, R. (1995) 2. Angers, S., Salahpour, A., and Bouvier, M. (2002) Annu. Rev. Pharmacol. J. Neurosci. Res. 42, 818 – 828 Toxicol. 42, 409 – 435 37. Vila-Coro, A. J., Mellado, M., Martin, d. A., Lucas, P., del Real, G., Martinez, 3. Angers, S., Salahpour, A., Joly, E., Hilairet, S., Chelsky, D., Dennis, M., and A., and Rodriguez-Frade, J. M. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, Bouvier, M. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 3684 –3689 3388 –3393 4. Overton, M. C., and Blumer, K. J. (2000) Curr. Biol. 10, 341–344 38. Rodriguez-Frade, J. M., Vila-Coro, A. J., de Ana, A. M., Albar, J. P., Martinez, 5. Rocheville, M., Lange, D. C., Kumar, U., Sasi, R., Patel, R. C., and Patel, Y. C. A., and Mellado, M. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 3628 –3633 (2000) J. Biol. Chem. 275, 7862–7869 39. Vila-Coro, A. J., Rodriguez-Frade, J. M., Martin, d. A., Moreno-Ortiz, M. C., 6. Cornea, A., Janovick, J. A., Maya-Nunez, G., and Conn, P. M. (2001) J. Biol. Martinez, A., and Mellado, M. (1999) FASEB J. 13, 1699 –1710 Chem. 276, 2153–2158 40. Garzon, J., Juarros, J. L., Castro, M. A., and Sanchez-Blazquez, P. (1995) Mol. 7. Roess, D. A., Horvat, R. D., Munnelly, H., and Barisas, B. G. (2000) Endocri- Pharmacol. 47, 738 –744 nology 141, 4518 – 4523 41. Bristow, M. R., Ginsburg, R., Minobe, W., Cubicciotti, R. S., Sageman, W. S., 8. McVey, M., Ramsay, D., Kellett, E., Rees, S., Wilson, S., Pope, A. J., and Lurie, K., Billingham, M. E., Harrison, D. C., and Stinson, E. B. (1982) Milligan, G. (2001) J. Biol. Chem. 276, 14092–14099 N. Engl. J. Med. 307, 205–211 9. Kroeger, K. M., Hanyaloglu, A. C., Seeber, R. M., Miles, L. E., and Eidne, K. A. 42. Bristow, M. R., Ginsburg, R., Umans, V., Fowler, M., Minobe, W., Rasmussen, (2001) J. Biol. Chem. 276, 12736 –12743 R., Zera, P., Menlove, R., Shah, P., Jamieson, S., and Stinson, E. B. (1986) 10. Cheng, Z. J., and Miller, L. J. (2001) J. Biol. Chem. 276, 48040 – 48047 11. Ayoub, M. A., Couturier, C., Lucas-Meunier, E., Angers, S., Fossier, P., Circ. Res. 59, 297–309 Bouvier, M., and Jockers, R. (2002) J. Biol. Chem. 277, 21522–21528 43. Brodde, O. E. (1991) Pharmacol. Rev. 43, 203–242 12. Rocheville, M., Lange, D. C., Kumar, U., Patel, S. C., Patel, R. C., and Patel, 44. Nguyen, C. T., De Champlain, J., and Bouvier, M. (1995) Biomed. J. 2, 34 – 42

Journal

Journal of Biological Chemistry – Unpaywall

Published: Nov 1, 2002

There are no references for this article.