Functionally Different Agonists Induce Distinct Conformations in the G Protein Coupling Domain of the β2Adrenergic Receptorстатья из журнала
Аннотация: G protein-coupled receptors represent the largest class of drug discovery targets. Drugs that activate G protein-coupled receptors are classified as either agonists or partial agonists. To study the mechanism whereby these different classes of activating ligands modulate receptor function, we directly monitored ligand-induced conformational changes in the G protein-coupling domain of the β2 adrenergic receptor. Fluorescence lifetime analysis of a reporter fluorophore covalently attached to this domain revealed that, in the absence of ligands, this domain oscillates around a single detectable conformation. Binding to an antagonist does not change this conformation but does reduce the flexibility of the domain. However, when the β2 adrenergic receptor is bound to a full agonist, the G protein coupling domain exists in two distinct conformations. Moreover, the conformations induced by a full agonist can be distinguished from those induced by partial agonists. These results provide new insight into the structural consequence of antagonist binding and the basis of agonism and partial agonism. G protein-coupled receptors represent the largest class of drug discovery targets. Drugs that activate G protein-coupled receptors are classified as either agonists or partial agonists. To study the mechanism whereby these different classes of activating ligands modulate receptor function, we directly monitored ligand-induced conformational changes in the G protein-coupling domain of the β2 adrenergic receptor. Fluorescence lifetime analysis of a reporter fluorophore covalently attached to this domain revealed that, in the absence of ligands, this domain oscillates around a single detectable conformation. Binding to an antagonist does not change this conformation but does reduce the flexibility of the domain. However, when the β2 adrenergic receptor is bound to a full agonist, the G protein coupling domain exists in two distinct conformations. Moreover, the conformations induced by a full agonist can be distinguished from those induced by partial agonists. These results provide new insight into the structural consequence of antagonist binding and the basis of agonism and partial agonism. G protein-coupled receptors β2adrenergic receptor fluorescein maleimide transmembrane (−)-isoproterenol (−)-alprenolol salbutamol dobutamine full width at half maximum G protein-coupled receptors (GPCRs)1 are remarkably versatile biological sensors. They are responsible for the majority of cellular responses to hormones and neurotransmitters, as well as for the senses of sight, smell, and taste. Our current models of the mechanism of GPCR activation by diffusible agonists have been deduced from indirect measures of receptor conformation, such as G protein or second messenger activation (1Tota M.R. Schimerlik M.I. Mol. Pharmacol. 1990; 37: 996-1004PubMed Google Scholar, 2Selley D.E. Sim L.J. Xiao R. Liu Q. Childers S.R. Mol. Pharmacol. 1997; 51: 87-96Crossref PubMed Scopus (191) Google Scholar, 3Krumins A.M. Barber R. Mol. Pharmacol. 1997; 52: 144-154Crossref PubMed Scopus (59) Google Scholar, 4Perez D.M. Hwa J. Gaivin R. Mathur M. Brown F. Graham R.M. Mol. Pharmacol. 1996; 49: 112-122PubMed Google Scholar). These indirect assays of GPCR activity provide only limited insight into the agonist-induced structural changes that define the active state of the receptor. To elucidate the mechanism of GPCR activation by diffusible agonists, we developed a means for directly monitoring the active conformation of purified, detergent-solubilized β2 adrenergic receptor (β2AR) by site-specific labeling of an endogenous cysteine (Cys265) with fluorescein maleimide (FM-β2AR) (5Ghanouni P. Steenhuis J.J. Farrens D.L. Kobilka B.K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 5997-6002Crossref PubMed Scopus (305) Google Scholar). Based on homology with rhodopsin (6Palczewski K. Kumasaka T. Hori T. Behnke C.A. Motoshima H. Fox B.A. Le Trong I. Teller D.C. Okada T. Stenkamp R.E. Yamamoto M. Miyano M. Science. 2000; 289: 739-745Crossref PubMed Scopus (5038) Google Scholar), Cys265 is located in the third intracellular loop (IC3) at the cytoplasmic end of the transmembrane 6 (TM6) α helix (Fig.1 A). Mutagenesis studies have shown this region of IC3 to be important for G protein coupling (7O'Dowd B.F. Hnatowich M. Regan J.W. Leader W.M. Caron M.G. Lefkowitz R.J. J. Biol. Chem. 1988; 263: 15985-15992Abstract Full Text PDF PubMed Google Scholar, 8Liggett S.B. Caron M.G. Lefkowitz R.J. Hnatowich M. J. Biol. Chem. 1991; 266: 4816-4821Abstract Full Text PDF PubMed Google Scholar). An environmentally sensitive fluorophore covalently bound to Cys265 is therefore well positioned to detect agonist-induced conformational changes relevant to G protein activation. The effect of agonists and partial agonists on the fluorescence intensity of FM-β2AR correlates well with their biological properties (5Ghanouni P. Steenhuis J.J. Farrens D.L. Kobilka B.K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 5997-6002Crossref PubMed Scopus (305) Google Scholar). Binding of the full agonist isoproterenol induces a conformational change that decreases the fluorescence intensity of FM bound to Cys265 by ∼15% (Fig. 1 B), whereas binding of partial agonists results in smaller changes in intensity, and binding of an antagonist has no effect (5Ghanouni P. Steenhuis J.J. Farrens D.L. Kobilka B.K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 5997-6002Crossref PubMed Scopus (305) Google Scholar). Agonist-induced movement of FM bound to Cys265 was characterized by examining the interaction between the fluorescein at Cys265 and fluorescence quenching reagents localized to different molecular environments of the receptor. The results of these experiments are most consistent with either a clockwise rotation of TM6 and/or a tilting of the cytoplasmic end of TM6 toward TM5 (5Ghanouni P. Steenhuis J.J. Farrens D.L. Kobilka B.K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 5997-6002Crossref PubMed Scopus (305) Google Scholar). Our findings suggest that the conformational changes associated with β2AR activation are similar to those in rhodopsin (9Farrens D.L. Altenbach C. Yang K. Hubbell W.L. Khorana H.G. Science. 1996; 274: 768-770Crossref PubMed Scopus (1113) Google Scholar) and indicate a shared mechanism of GPCR activation. These results provide insight into the nature of the structural changes that occur upon agonist binding. However, several mechanistic questions remain. Using conventional spectroscopy, we observe no change in the fluorescence intensity from FM-β2AR upon antagonist binding. This could indicate that antagonists do not alter receptor structure or that the structural changes are not detectable by FM bound to Cys265. Of greater interest is the structural basis of partial agonism. Partial agonists induce a smaller change in intensity of FM-β2AR than do full agonists. Two models could explain this observation. If we assume that the receptor exists in two functional conformational states, inactive or active, then a partial agonist may simply induce a smaller fraction of receptors to undergo the transition to the active state than does the full agonist. Alternatively, partial agonists may induce a conformation distinct from that induced by full agonists. Conventional fluorescence spectroscopy, which represents an average intensity over a population of fluorescent molecules, does not allow us to distinguish between these two models. We therefore used fluorescence lifetime spectroscopy, which is capable of distinguishing substates within a population of fluorescent molecules, to look for ligand-specific conformational states. Our results indicate that partial agonists and agonists induce distinct conformations. Moreover, we observe structural effects of antagonist binding that could not be detected by conventional spectroscopy. These results help elucidate the structural mechanisms that underlie ligand efficacy. Construction, expression, and purification of human β2AR were performed as described (10Gether U. Lin S. Kobilka B.K. J. Biol. Chem. 1995; 270: 28268-28275Abstract Full Text Full Text PDF PubMed Scopus (258) Google Scholar). Purified, detergent-solubilized, wild-type receptor was labeled with FM (Molecular Probes) as described previously (5Ghanouni P. Steenhuis J.J. Farrens D.L. Kobilka B.K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 5997-6002Crossref PubMed Scopus (305) Google Scholar). The labeling procedure resulted in incorporation of 0.6 mol of FM per mol of receptor. Fluorescence spectroscopy experiments were performed on a SPEX Fluoromax spectrofluorometer with photon counting mode using an excitation and emission bandpass of 4.2 nm. Approximately 25 pmol of FM-labeled β2AR was diluted into 500 μl of 200 mm Tris, pH 7.5, 500 mm NaCl, 0.1%n·dodecyl-β-d-maltoside, 100 mm mercaptoethanolamine. Excitation was at 490 nm, and emission was measured from 500 to 599 nm with an integration time of 0.3 s/nm for emission scan experiments. For time course experiments, excitation was at 490 nm, and emission was monitored at 517 nm. For anisotropy studies, fluorescence intensities were measured with excitation and emission polarizers in horizontal (H) and vertical (V) combinations. The G factor was calculated from the ratio of the intensities (I) ofIHV/IHH, and the anisotropy (r) was calculated from Equation 1. r=IVV−GIVHIVV+2GIVHEquation 1 For studies measuring ligand effects, no difference was observed when using polarizers in magic angle conditions. Unless otherwise indicated, all experiments were performed at 25 °C, and the sample always underwent constant stirring. The volume of the added ligands was ≤ 1% of total volume, and fluorescence intensity was corrected for this dilution in all experiments shown. All of the compounds tested had an absorbance of less than 0.01 at 490 and 517 nm in the concentrations used, excluding any inner filter effect in the fluorescence experiments. To determine fluorescence lifetimes, ∼250 pmol of FM-β2AR was diluted in 1.5 ml of 200 mm Tris, pH 7.5, 500 mm NaCl, 0.1%n·dodecyl-β-d-maltoside, 100 mmmercaptoethanolamine and incubated for 10 min at 25 °C with or without ligand. Fluorescence lifetimes were measured using a frequency domain 10-GHz fluorometer equipped with a Hamamatsu 6-μm microchannel plate detector as described previously (11Laczko I.G.G. Gryczynski Z. Wiczk W. Malak H. Lakowicz J.R. Rev. Sci. Instrum. 1990; 61: 2331-2337Crossref Scopus (182) Google Scholar). The instrument covered a wide frequency range (4–5000 MHz), which allowed detection of lifetimes ranging from several nanoseconds to a few picoseconds. Samples were placed in a 10-mm path length cuvette. The excitation was provided by the frequency-doubled output of a cavity-dumped pyridine-2 dye laser tuned at 370 nm synchronously pumped by a mode-locked argon ion laser. Sample emission was filtered through Corning 3–72 and 4–96 filters. For the reference signal, 4-(dimethylamino)-4′-cyanostilbene in methanol (463-ps fluorescence lifetime) was observed through the same filter combination. The governing equations for the time-resolved intensity decay data were assumed to be a sum of discrete exponentials as in Equation 2, I(t)=Io∑iαiet/τiEquation 2 where I(t) is the intensity decay,αi is the amplitude (pre-exponential factor), and τi is the fluorescence lifetime of the i-th discrete component or a sum of Gaussian distribution functions as in Equation 3, I(t)=Io∑iαi(τ)et/τiEquation 3 and Equation 4, αi(τ)=1σ2πe−121−τσ3Equation 4 where τ is the center value of the lifetime distribution, and ς is the standard deviation of the Gaussian, which is related to the full width at half maximum by 2.354 ς. In the frequency domain, the measured quantities at each frequency ω are the phase shift (φω) and demodulation factor (m ω) of the emitted light versusthe reference light. Fractional intensity, amplitude, and lifetime parameters were recovered by a non-linear least squares procedure using the software developed at the Center for Fluorescence Spectroscopy. The measured data were compared with calculated values (φ cω,m cω), and the goodness of fit was characterized by Equation 5, χR2=1υ∑ωφω−φcωδφ2+1υ∑ωmω−mcωδm2Equation 5 where υ is the number of degrees of freedom, andδφ and δm are the uncertainties in the measured phase and modulation values, respectively. The sum extends over all frequencies (ω). The β2AR was purified and labeled at Cys265 with fluorescein maleimide to generate FM-β2AR as described previously (5Ghanouni P. Steenhuis J.J. Farrens D.L. Kobilka B.K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 5997-6002Crossref PubMed Scopus (305) Google Scholar). We examined ligand-dependent changes in fluorescence lifetime of FM-β2AR in an effort to identify the existence of agonist-specific conformational states. Fluorescence lifetime analysis can detect discrete conformational states in a population of molecules, whereas fluorescence intensity measurements reflect the weighted average of one or more discrete states. Based on the changes in steady-state fluorescence intensity we observed, we predicted that ligand-induced conformational changes in the receptor would alter the fluorescence lifetime of the fluorophore. Fluorescence lifetime, τ, refers to the average time that a fluorophore that has absorbed a photon remains in the excited state before returning to the ground state. The lifetime of fluorescein (nanoseconds) is much faster than the predicted off-rate of the agonists we examined (μs-ms) and much shorter than the half-life of conformational states of bacteriorhodopsin (μs) (12Subramaniam S. Henderson R. Nature. 2000; 406: 653-657Crossref PubMed Scopus (383) Google Scholar) or rhodopsin (ms) (13Farahbakhsh Z.T. Hideg K. Hubbell W.L. Science. 1993; 262: 1416-1419Crossref PubMed Scopus (147) Google Scholar, 14Arnis S. Fahmy K. Hofmann K.P. Sakmar T.P. J. Biol. Chem. 1994; 269: 23879-23881Abstract Full Text PDF PubMed Google Scholar) or of ion channels (μs-ms) (15Hoshi T. Zagotta W.N. Aldrich R.W. J. Gen. Physiol. 1994; 103: 249-278Crossref PubMed Scopus (165) Google Scholar). Therefore, lifetime analysis of fluorescein bound to Cys265 is well suited to capture even short-lived, agonist-induced conformational states. Data from fluorescence lifetime experiments on FM-β2AR bound to different drugs at equilibrium were analyzed in two ways. Traditionally, fluorescence decays are fit to single and multiple discrete exponential functions, and the best fit is determined by χ2 analysis (TableI). This discrete component analysis assumes that the receptor exists in one or a few rigid protein conformations and does not accurately reflect the dynamic nature of proteins. Proteins that are functionally in a single conformational state actually undergo small conformational fluctuations around a minimum energy state (16Frauenfelder H. Sligar S.G. Wolynes P.G. Science. 1991; 254: 1598-1603Crossref PubMed Scopus (2613) Google Scholar), and these small structural perturbations can lead to small changes in the environment around an attached fluorophore. These perturbations are thought to reflect local unfolding reactions within the three-dimensional structure of proteins (17Freire E. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 11680-11682Crossref PubMed Scopus (75) Google Scholar). Such flexibility in protein structure can be modeled using fluorescence lifetime distributions (18Gratton E. Alcala R. Prendergast F.G. Jameson D.M. Fluorescent Biomolecules: Methodologies and Applications. Plenum Press, New York1989: 17-32Crossref Google Scholar), wherein the width of the distributions reflects the conformational flexibility of the protein (Fig.2). The mobility of fluorescein relative to the receptor is minimal, as determined by its high measured anisotropy (r = 0.30 ± 0.02; n = 3) and therefore would be expected to contribute little to the width of the lifetime distribution. Thus, the width of the distribution can be attributed to conformational flexibility in the receptor itself.Table IFluorescent lifetime data for FM-β2 AR in the presence and absence of drugs fit to discrete exponential functionsτ1 (ns)τ2(ns)α2χ2No drug4.22 ± 0.022.9 ± 0.4ALP4.21 ± 0.013.1 ± 0.8ISO4.30 ± 0.010.77 ± 0.050.19 ± 0.033.3 ± 1.0SAL4.35 ± 0.021.45 ± 0.160.08 ± 0.012.0 ± 0.2DOB4.36 ± 0.011.68 ± 0.30.07 ± 0.011.8 ± 0.4The observed fluorescence decay was resolved into one or more exponential components, with each component, i, being described by τi and αi , where αi represents the fractional contribution of τi to the overall decay. The best fit to single or multiple components was determined by χ2 analysis. If different agonists induce a single active state, then the fluorescence lifetime associated with that state (τR* ) should be the same for different drugs, and only the fractional contributions (αDRUG ) should differ. However, if there are agonist-specific conformational states we should observe unique, agonist-specific lifetimes (e.g. τISO , τSAL , and τDOB ). Open table in a new tab The observed fluorescence decay was resolved into one or more exponential components, with each component, i, being described by τi and αi , where αi represents the fractional contribution of τi to the overall decay. The best fit to single or multiple components was determined by χ2 analysis. If different agonists induce a single active state, then the fluorescence lifetime associated with that state (τR* ) should be the same for different drugs, and only the fractional contributions (αDRUG ) should differ. However, if there are agonist-specific conformational states we should observe unique, agonist-specific lifetimes (e.g. τISO , τSAL , and τDOB ). Lifetime analysis of unliganded FM-β2AR reveals a single, flexible state. This is indicated by both the single, broad Gaussian distribution of lifetimes centered around 4.2 ns (Fig. 2, black trace) and the discrete component analysis, where the fluorescence decay rate of FM-β2AR in the absence of any drug is best fit by a single exponential function (Table I). Binding of the neutral antagonist ALP to FM-β2AR does not significantly change the fluorescent lifetime (Table I) but does narrow the distribution of lifetimes (Fig. 2, red trace), suggesting that ALP stabilizes the receptor and reduces conformational fluctuations. This interpretation is consistent with the results of experiments demonstrating that the β2AR is more resistant to protease digestion when bound to ALP (19Kobilka B.K. J. Biol. Chem. 1990; 265: 7610-7618Abstract Full Text PDF PubMed Google Scholar). Unexpectedly, binding of the full agonist ISO promotes conformational heterogeneity. In the presence of saturating concentrations of ISO, FM-β2AR has two distinguishable fluorescence lifetimes (see Fig. 2, green trace, and TableI) representing at least two distinct conformational states. The long lifetime component is only slightly longer than the lifetime observed in the absence of drugs; however, the distribution is narrower than that observed in the presence of the antagonist ALP (Fig. 2,green and red traces). In contrast, the distribution of the short lifetime component observed in the presence of ISO is relatively broad, suggesting that there is considerable flexibility around Cys265 in this agonist-induced conformation. We next examined the effect of the partial agonists salbutamol (SAL) and dobutamine (DOB) on the fluorescence lifetime of FM-β2AR. Similar to ISO, we observed two lifetimes when the receptor was bound to saturating concentrations of SAL and DOB (see Table I and Fig. 3). The long lifetime component found in the presence of these two partial agonists is indistinguishable from that observed in the ISO-bound receptor; however, the short lifetime component found in both the SAL- and DOB-bound receptor is statistically different from that for the ISO-bound receptor. We observe a strong correlation between a reduction in fluorescence intensity of FM bound to Cys265 and drug efficacy (5Ghanouni P. Steenhuis J.J. Farrens D.L. Kobilka B.K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 5997-6002Crossref PubMed Scopus (305) Google Scholar), and shortening of the average fluorescence lifetime is associated with a reduction in fluorescence intensity. Therefore, the short lifetime, found only in the presence of agonists, likely represents the G protein-activating conformation of FM-β2AR. The different short lifetimes for the full agonist (ISO) and the partial agonists (SAL and DOB) indicate different molecular environments around the fluorophore and therefore represent different, agonist-specific active states. The narrowing and rightward shift of the long lifetime component following binding of both agonists and partial agonists indicate that this lifetime also reflects an agonist-bound state but most likely represents a more abundant intermediate state that would not be expected to alter greatly the intensity of FM bound to Cys265. It is possible that the number of conformations that we observe in these experiments represent only a few of the possible conformations that can be stabilized by drugs. Moreover, whereas the overlapping short lifetime distributions of SAL and DOB (see Fig. 3 B and Table I) suggest that they induce similar conformations, it is possible that a conformationally sensitive probe positioned elsewhere on the receptor could distinguish between DOB- and SAL-bound receptors states. According to the prevailing two-state model of GPCR activation, receptors exist in an equilibrium between a resting (R) state and an active (R*) state that stimulates the G protein (20Samama P. Cotecchia S. Costa T. Lefkowitz R.J. J. Biol. Chem. 1993; 268: 4625-4636Abstract Full Text PDF PubMed Google Scholar, 21Lefkowitz R.J. Cotecchia S. Samama P. Costa T. Trends Pharmacol. Sci. 1993; 14: 303-307Abstract Full Text PDF PubMed Scopus (754) Google Scholar, 22Leff P. Trends Pharmacol. Sci. 1995; 16: 89-97Abstract Full Text PDF PubMed Scopus (470) Google Scholar). Agonists preferentially enrich the R* state, whereas inverse agonists select for the R state of the receptor. Neutral antagonists possess an equal affinity for both states and function simply as competitors. In this simple system, functional differences between drugs can be explained by their relative affinity for the single active R* state (Fig.4 A). Alternatively, differences in efficacy between drugs can be because of ligand-specific receptor states (23Kenakin T. Trends Pharmacol. Sci. 1997; 18: 416-417Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar, 24Tucek S. Trends Pharmacol. Sci. 1997; 18: 414-416Abstract Full Text PDF PubMed Google Scholar, 25Strange P.G. Biochem. Pharmacol. 1999; 58: 1081-1088Crossref PubMed Scopus (63) Google Scholar). Based on our lifetime experiments, we propose a model with multiple, agonist-specific receptor states, wherein activation occurs through a sequence of conformational changes. Upon agonist binding, the receptor undergoes a conformational change to an intermediate state (R′) that is associated with a narrowing and rightward shift in the long lifetime distribution. The less abundant active state, represented by the short lifetime, is different for the full agonist ISO (R*) and the partial agonists DOB and SAL (Rx). The relatively slow, temperature-dependent rate of change of fluorescence intensity following agonist binding and the rapid rate of reversal by antagonist (see Ref. 5Ghanouni P. Steenhuis J.J. Farrens D.L. Kobilka B.K. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 5997-6002Crossref PubMed Scopus (305) Google Scholar and Fig. 1 B) suggest that transitions from the intermediate state to the active state are relatively rare high energy events. It is likely that in vivo the active conformation is further stabilized by interactions between the receptor and its cognate G protein Gs. Thus, one might expect the proportion of receptor in the active state to be greater when the receptor is coupled with Gs. Our results have implications for drug discovery and efforts to obtain high resolution crystal structures of GPCRs. The conformational flexibility observed in the ligand-free receptor (Fig.2) may make it particularly challenging to obtain crystals in the absence of a bound ligand. Of greater concern, the existence of two conformational states in the presence of saturating concentrations of full and partial agonists (Fig. 3) will impact efforts to obtain a high resolution structure of the active, agonist-bound receptor. If such an agonist-bound structure is obtained, it will likely represent the more energetically stable of the two conformations and may not be the maximally active conformation. Thus, the use of such a structure for rational drug design efforts may have significant limitations. In support of this contention, recent structural analyses of the intermediate conformational states of the proton pump bacteriorhodopsin have revealed discrepancies attributed to energetic inhibition of conformational states within the three-dimensional crystal (12Subramaniam S. Henderson R. Nature. 2000; 406: 653-657Crossref PubMed Scopus (383) Google Scholar, 26Sass H.J. Buldt G. Gessenich R. Hehn D. Neff D. Schlesinger R. Berendzen J. Ormos P. Nature. 2000; 406: 649-653Crossref PubMed Scopus (318) Google Scholar). Finally, our results indicate that GPCRs are relatively plastic. The number of conformations that we observed in these experiments may represent only a few of a larger spectrum of possible conformations that could be stabilized by drugs. Thus, it may be possible to identify even more potent agonists or agonists that can alter G protein coupling specificity.
Год издания: 2001
Авторы: Pejman Ghanouni, Zygmunt Gryczyński, Jacqueline J. Steenhuis, Tae Weon Lee, David Farrens, Joseph R. Lakowicz, Brian K. Kobilka
Издательство: Elsevier BV
Источник: Journal of Biological Chemistry
Ключевые слова: Receptor Mechanisms and Signaling, Neuropeptides and Animal Physiology, Neuroscience and Neuropharmacology Research
Открытый доступ: hybrid
Том: 276
Выпуск: 27
Страницы: 24433–24436