Exploring the dynamics of regulation of G protein‐coupled receptors using green fluorescent protein

New advances in understanding the details of cellular response to agonist ligands require integration of approaches derived from pharmacology and both cell and molecular biology. Direct visualization of alterations in the concentration and distribution gradients of [Ca2+] in individual cells achieved by monitoring differences in fluorescence characteristics of Ca2+ indicator dyes in their Ca2+-bound and unbound states has produced an almost unimaginable increase in our understanding of the dynamics of cellular Ca2+ handling over the last ten years. However, until relatively recently the potential for non-invasive analysis of the dynamics of trafficking of cellular proteins seemed limited. In the recent past the use of mutationally modified forms of the autofluorescent green fluorescent protein (GFP) from the jellyfish Aequorea victoria has begun to make imaging of many proteins in living single cells a reality. In the near future this is likely to result in as profound a revolution in our understanding of cellular targeting, trafficking and interactions between proteins of pharmacological interest as was the introduction and improvement of Ca2+ indicator dyes for the regulation of Ca2+ homeostasis. Although this approach has now been used for a wide range of signalling proteins this review will concentrate on insights gained from the use of GFP-tagging into the kinetics and dynamics of regulation of G protein-coupled receptors (GPCRs) and proteins which interact directly with them. Until the introduction of GFP as an experimental tool single, intact, cell detection and analysis of the potential cellular redistribution of a GPCR or other GPCR-interacting proteins was restricted to two approaches. In the first of these pharmacophores from ligands with appropriate pharmacological selectivity were modified to incorporate fluorescent chemical groups. Key issues for this approach are that pharmacological selectivity and, if possible, potency are preserved in the modified ligand. Validation of the ligands for potential use in physiologically relevant systems could be achieved initially using cell lines which had been transfected with cDNAs for appropriate GPCRs (Daly et al., 1998). However, unless a distinct increase in fluorescence of the ligand is obtained upon binding the GPCR in question, direct visualization is difficult to achieve without substantial washing of the cell or tissue. For GPCRs with small molecule endogenous ligands the ensuing limitation on the affinity of drugs which can be used generally results in the most appropriate ligands being antagonists. As such, a rather static picture of GPCR expression and distribution is the most likely outcome. Many of these issues have recently been reviewed and assessed (McGrath et al., 1996). In the case of GPCRs with high affinity peptide ligands it is often possible to attach a suitable fluorescent dye to a location in the peptide not involved in an integral way in binding the receptor. Ligand affinity is thus often preserved. Furthermore, given the fact that agonist binding to a GPCR frequently results in an alteration in the cellular distribution of the GPCR and the fact that high affinity peptide ligands are routinely co-internalized with the GPCR owing to their slow dissociation rates, internalization of the peptide-GPCR complex can be monitored directly (see Ashworth et al., 1995 for example). Although many early studies utilized fluorescent tags such as rhodamine, improvements in chemistry in this area have allowed generation of fluorescently-tagged peptides which are much brighter and which can thus be used to localize and visualize GPCRs expressed at relatively low levels (see Advanced Bioconcepts at http://www.bioconcept.com for example). The second approach utilizes immunocytochemical detection of the GPCRs. Appropriate antisera, generally to peptide sequences of the GPCR in question, have been produced in many cases and represent the only practical means to immunolocalize GPCRs in tissues and cells which express the GPCR endogenously. However, by far the most widespread strategy in cell transfection studies has been the use of epitope-tagging. Herein, a cDNA encoding a GPCR is modified to incorporate a sequence, either within the N-terminal extracellular region or at the C-terminus, which is known to be identified by a well defined, high affinity, antibody. Widely used epitope tags include HA, c-myc, FLAG™ and VSV. With such modified GPCRs it is important to define that the alterations to the protein sequence do not modify the pharmacology of the GPCR. For GPCRs with small molecule natural ligands, N-terminal epitope tagging is unlikely to modify ligand binding or function. C-terminal tagging would also be anticipated to have little effect on ligand pharmacology or on second messenger regulation by a GPCR. However, recent studies have suggested that not all functions of a GPCR necessarily arise via G protein activation. For example Hall et al. (1998) have recently demonstrated that the β2-adrenoceptor interacts with the Na+/H+ exchanger regulatory factor (NHERF) via a PDZ domain in NHERF. Mutagenesis studies indicate that it is the extreme C-terminal motif of the β2-adrenoceptor which is responsible for this interaction. Although it has not been examined directly, it must be anticipated that addition of a C-terminal epitope tag to the β2-adrenoceptor would modify or potentially ablate this interaction. Immunofluorescence pictures of such epitope-tagged GPCRs before and following treatment with different agonist ligands have provided direct and key insights into the regulation of GPCR internalization and trafficking. However, by their very nature, such experiments must be performed on fixed and permeabilized cells (to allow antibody access to the potentially internalized GPCR) and thus limits use in real time, dynamic, cell imaging. The Aequorea victoria photoprotein GFP is a 238 amino acid protein that emits green light with an emission maximum of 509 nm upon fluorescent excitation at 488 nm. Unlike other bioluminescent reporter proteins no additional substrates or cofactors are required for light emission (Chen et al., 1995). GFP fluorescence is stable and has been measured non-invasively in living cells of many species including mammals, drosophila, C. elegans, yeast and E. coli. GFP fluorescence can be detected by fluorimetry, by fluorescence activated cell sorting (FACS) and by microscopy. As there is no required cofactor the attractiveness of GFP as a reporter protein is obvious (Chalfie & Kain, 1998). The availability of the cDNA sequence for GFP has resulted in the generation and characterization of several GFP mutants with enhanced fluorescence emission. Mutation of the serine at amino acid 65 to threonine resulted in the generation of a protein with a 6 fold increase in the intensity of fluorescence emission (Haas et al., 1996). Furthermore, the presence of the Ser65Thr and mutation of the phenylalanine residue at position 64 to leucine has resulted in a 35 fold increase in fluorescence intensity (Haas et al., 1996). In addition, a number of novel mutants of GFP have also been identified with altered excitation or emission characteristics. For example, mutation of the tyrosine residue at position 66 to histidine generated a protein with altered spectral properties and blue fluorescence emission. This so-called blue fluorescent protein has a λmax for excitation of 458 nm and for emission of 480 nm (Chalfie & Kain, 1998). These and many other variants of GFP are now commercially available from a number of sources. Although often used directly in cellular transient transfection studies simply to identify cells in a field which have taken up cDNA and express the protein, the true benefit of GFP can be utilized following its construction into fusion proteins to tag and thus monitor the distribution of another protein of interest. The modified forms of GFP described above have been widely used in fusion proteins to assess protein trafficking (Ferrer et al., 1997), and subcellular localization of recombinantly expressed proteins (Wang and Hazelrigg, 1994). Introduction of this technology into analysis of GPCR regulation was first reported at the end of 1997 (Barak et al., 1997). Since then a number of groups have described the creation and use of GPCR–GFP fusion proteins to monitor both receptor internalization and recycling following agonist treatment and to explore a number of distinct pharmacological issues. As GFP is a 27 kDa protein it is of the of the of and GPCRs. As such, a key must be to that the of GPCRs are not altered by their tagging with the used to have the GFP to the C-terminal of the GPCR. expressed GFP is a protein which as a is thus unlikely to allow appropriate expression of the fusion protein were GFP to be to the of a GPCR an appropriate sequence was also into the as this would require of the GFP the in all to the GPCRs have been to appropriate second messenger in an Furthermore, for of GPCRs and their with proteins of the has been As with many of GPCR of the on the regulation of a GPCR utilized the β2-adrenoceptor (Barak et al., 1997). transient expression in cells a fusion protein between the β2-adrenoceptor and a of GFP with enhanced agonist and binding characteristics as the Furthermore, the fusion protein also allowed of which was following to a high concentration of no very pharmacological of an GPCR and an of the GPCR has been of GPCR well the β2-adrenoceptor and the GPCR–GFP fusion protein could be in an Although it was not demonstrated that the of in the fusion protein were the as modified in the or the GFP tag also both forms of the GPCR internalized to following a with The fusion protein internalized into in response to a which could be in cells and in real The to of a cell on to a of high intensity light has been widely to the of proteins in fluorescence monitoring of the of the from the area into this region allowed of the of of the GPCR as with transient transfection of many GPCRs into used cell lines is that expression of the protein is by a a of the protein produced to in the or other and is not to the This can be directly using a GPCR–GFP fusion protein (see for example). protein obvious for analysis of agonist regulation of the subcellular distribution of GPCRs in such transfected cells and thus the of cell the fusion protein is to be In the case of the β2-adrenoceptor this has been achieved in both et al., and & cells In cells protein of et al. (1998) a of this in response to which could be monitored in by a in levels of cell binding of the cell derived levels of a were examined by & was in both binding studies and by in a In both systems internalization of by to agonist could be by of agonist and its with the In the cell treatment to with a concentration of resulted in of up to of the receptor binding as monitored by the binding of et al., 1998). treatment of the could be to direct visualization of the cells that much of the internalized was not This was by cell binding ligand is and thus can monitor cellular levels of the fusion protein in & This an of the from the high levels of expression of the in these a effect of of GFP to the GPCR or of a fusion protein in response to The β2-adrenoceptor was with GFP and expressed in field of cells this was with a before and following of (see & for Although only could be it is of interest to that the and of internalization of in cells be by the of and of the agonist ligand of the and of GPCR internalization have been & 1997). However, the to directly visualize these to et al. (1998) much internalization of the fluorescent protein in response to the low β2-adrenoceptor agonist in response to studies have also internalization of a expressed at protein in cells with the in the cell internalization was both and for the fusion protein & The internalized could also be to the following of and its by the & of other GPCR–GFP fusion proteins have been and of cellular distribution and regulation examined following either transient or stable expression in cell These include the receptor et al., 1997), the et al., the et al., and et al., et al., the receptor et al., the receptor et al., the receptor et al., and the receptor & 1998). Many of these studies have the likely of into which the receptor is by of the to as a the internalized GPCR–GFP fusion protein with known to the cells via the 1998). cells with rhodamine, or other forms of by washing to between the and the internalized GPCR–GFP fusion protein is obtained by of the and et al., et al., et al., et al., 1998). with forms of or other which are to and can then be used to the of GPCR–GFP fusion proteins to these with agonist treatment of cells et al., 1998). is to that stable expression of a in widely cell lines resulted in fusion protein internalization in and However, this was in cells et al., 1997). As this was by receptor then the of cell be by the to the of agonist or Although to in a of single fixed a number of now which is on a in with a This an to of alteration in the cellular distribution of a GPCR–GFP fusion protein et al., for allow the of the dynamics of protein and a number of groups also such on In the case of a fusion protein expressed in initially this has distribution the of results in concentration of the GFP to at the This is by internalization of the into a number of small to the many of these small to into et al., The internalization of not its to at the is by et al., As this treatment is routinely used to internalization of proteins via it is to that the represent such cellular redistribution of a of the of the epitope-tagged of the of the was with GFP and expressed in of et al., were on a and in a at and addition of for and The from an distribution to an initially distribution to location in of which (see for are taken from et al. with of imaging of a GPCR–GFP and in fixed cells of G protein it has also been possible to the of of a GPCR and an G cells high levels of both the and the G protein et al. were to use approaches to that treatment with a concentration of resulted in of this G protein from the and into However, by following the internalization of the in cell binding studies the of the G protein was of the GPCR et al., this and assess potential of the GPCR and G protein, et al. examined the cellular distribution of and the G proteins and using a to treatment both of proteins were at the and of the their treatment with the of had been Although a of the cellular and at the was also produced for internalization of these G of the of and and location the internalized proteins to be in the kinetics of this were examined internalization of was it was not possible to G protein internalization within this it is to this that the different proteins are internalized with very different thus potentially or the the high levels of G protein in these cells to is possible that internalization of is with of and that many of GPCR internalization and recycling to the are required to a of the G protein not such that can be studies which with GPCR–GFP recycling this As well as in a which is produced by of the G protein-coupled receptor many GPCRs can be via second to a potential for As the in response to treatment with et al., this is via of protein C. expression in the cell the of pharmacological on the internalization of an was et al., 1998). The was to internalization of the as was the rather protein which is often used as a protein Although not as as the protein was also to internalization of Furthermore, internalization of the was also by treatment of the cells with to the receptor et al., 1998). all GPCRs are to the of et al. have that transient expression of an in cells results in the of the expressed protein the this was not true for an of the used this to the that the can be used as a pharmacological for the of the two GPCR by was in of the in cell thus that is not a between these two GPCRs if provided with that of rather to cellular distribution of the two GPCRs et al., 1997). GPCR can often result in of the modified In experiments to explore and of the agonist binding and the of a GPCR to a of G protein it is important to be to the modified GPCR has been expressed and Although to be in this GFP-tagging of mutationally modified forms of the receptor has been used to explore the of a sequence, and an residue often within such in the C-terminal of this GPCR to the of the expressed protein to from the et al., 1998). potentially in which GFP-tagging of GPCRs has to be utilized in a is to the expression and appropriate of GPCRs for which ligands are are being into ligand The most obvious of GPCR proteins are the G little has been with forms of G This that the extreme of G protein a key for with GPCRs & As such, it has to be demonstrated that a G protein fusion protein, in which the GFP is to the C-terminal of the G protein, would be to interact with a GPCR. all G protein are either or modified in their N-terminal region by of For the G proteins is to the N-terminal of the This if this is no at the as would be the case if GFP was in of the G and other are to As such, the only use to of forms of a G protein has been to explore the of N-terminal in the G protein both to the and to et al. used GFP as a reporter for these following construction of fusion proteins between either or a of forms of this G These were either at their or mutationally modified to which only the N-terminal amino of were to be modified by and and both were the was to the upon transient expression in However, of either or and analysis of subcellular localization of these GFP following expression in cells of a of the G with for a key protein for only a small of the was in the and the to with could be with and this also required the of as no of could be achieved following of the G protein and a of et al., results a key for N-terminal G protein in appropriate cellular of a G protein and for with has been interest in that the subcellular localization of and is very is to the in cells is to the and et al., Although such studies to have been performed by GFP fusion proteins an Furthermore, commercially available are in As such, these would a means to explore potential differences in proteins which interact with these two G proteins to define the for the cellular As many GPCRs are in a at a number of and in their C-terminal by of the of protein This only the of GPCR and G protein interactions which results in of agonist function. The GPCR then as a binding for proteins of the et al., In the cell are As such, it be possible to the of a from the to the location of a GPCR at the in response to an of this was achieved for an epitope-tagged of the β2-adrenoceptor and in These studies also provided direct for a of in this et al., to such interactions using forms of were by et al. generated a fusion protein in which a of GFP with enhanced was to the of of the β2-adrenoceptor and into cells by treatment of the cells with resulted in of a of from a fluorescence to a of the and kinetics of were enhanced by of results were produced with transient expression of the receptor (Barak et al., and in review et al., these have that the of can be for a wide range of GPCRs. GPCR which is to is the receptor. This GPCR is in at in it a C-terminal As such, it was anticipated not to be by and transient introduction of a fusion protein into cells high levels of the addition of to of to the et al., 1998). In transfection of cells the resulted in and of to the upon addition of et al., et al., et al., will be of interest to addition of the C-terminal of the to the receptor will allow and with such et al. cell lines and then a epitope-tagged of the As in the of agonist a imaging of the and detection of the GPCR to be in cellular of resulted in a of to the of the of GPCR and their with a of the could be to into This only in cells which could also be to be for the presence of the and not in cells which were not the GPCR. of GPCR and to be results were obtained by the experimental Introduction and of in cells a initially demonstrated of these proteins with the the with resulted in internalization of a of a of the cellular and their as assessed by of the from the individual proteins et al., approaches have also been used to interactions and of with both the receptor et al., and the et al., The to of internalized GPCRs and to be GPCR a very wide range of GPCRs are to of to the For a number of including the β2-adrenoceptor and the the to the to there rather being into the of the cells with the GPCR. GPCRs such as the the receptor et al., and, as the et al., allow of the into in which to be with the GPCR. not the of the GPCR C-terminal a in the is co-internalized with the GPCR. of the C-terminal of the receptor to the of the β2-adrenoceptor results in regulation of the cellular distribution of in a to that for the receptor. Furthermore, the was produced with the GPCR et al., is likely to the for these redistribution of in response to of the et al., was expressed in The of the was then In the of distribution of was and treatment with the distribution of a of the only in a of the cells it the of other studies that the redistribution of was only obtained in cells which had expressed the (see et al., for of expressed to the has also been examined in cells following with and forms of the et al., produced a to cells the GPCR. expression of a of this GPCR a of could be to be at the without addition of This also levels of the GPCR. However, and with the that addition of can the levels of of this addition of resulted in a of et al., The generation of in of second messenger regulation not necessarily with enhanced GPCR of the also a to However, this was not in either the or presence of concentration of was not in the or presence of following with this GPCR in cells et al., is also that the of the is internalized in response to The of and to internalization of the receptor has for a pharmacological which has been to to ligand This has also recently been using as an end et al., transfection of cells with the receptor allowed internalization in response to not This could not be by of was was either with or without In the of additional produced of the in the presence of both ligands were of of the and the of by either ligand was now produced by was not of could be without it was and enhanced by the presence of Furthermore, a of was produced by only upon with et al., results a for differences in the potential of ligands at the receptor and that expression levels of in different cells and tissues be as important in practical as the of the ligand in of GPCR to a ligand. is thus of potential that levels are in including et al., et al., and et al., 1997). of the of using a wide range of signalling with pharmacological and experiments is likely to novel insights into a wide range of important issues in Although the of this forms of factor receptors with tyrosine as well as many proteins involved in from the cell to the have been generated and In of GPCR and regulation two obvious and are likely to be widely forms of GFP which have enhanced the of the have been to the use of GFP tagging and imaging in range of mutants which have resulted in altered spectral for excitation and emission have been and the use of these in with the enhanced the of tagging of two signalling proteins which can then allow imaging of the individual proteins in living cells et al., This would represent a distinct over the use of GFP fluorescence and which the in the GPCR studies in the dynamics of of GPCRs and their recycling to the with of the for cellular of other bioluminescent proteins from a of species is also likely to the available for imaging has also been interest in the of fluorescence to the of proteins within signalling & that modified forms of GFP have been identified which have the to as then as the of the that two are into are it a appropriate way to the of GPCR and potential regulation of such by agonist ligands et al., & 1997). potentially to is which can be achieved between a to a and to an appropriate et al., that of other which a to a GPCR not with the and pharmacology of the protein it be possible to use which a bioluminescent in the region of nm to directly the of interactions in the are to it is that imaging and real pharmacological by ligand binding to a GPCR has a very in the GFP-tagging approaches is by the and and and for their to the studies