Almost every aspect of the human body – from our senses to our moods – depends, in one way or another, on a large family of proteins called G-protein-coupled receptors. These receptor proteins, known as GPCRs for short, detect signals from outside the cell and trigger activity within the cell. This allows cells to gather information from their surroundings and to communicate with each other. Importantly, since GPCRs regulate many processes in the body that are involved in disease, it is perhaps unsurprising that over a third of all approved drugs target these receptors.

Like all proteins, GPCRs are long chain-like molecules with a repetitive backbone and short branches called sidechains. Each sidechain has its own chemical properties and electrical charge, which can affect how different parts of the chain interact with each other and what shape the protein can adopt. This in turn can influence how strongly a drug or other molecule can bind to a receptor protein.

Protein crystallography is one technique that has been used to better understand how the different GPCRs are built and how they work. The technique involves growing crystals from pure samples of the protein; this locks millions of copies of the protein in place and provides a snapshot of its shape. However, GPCRs – and especially their sidechains – are flexible and can adopt different shapes, which cannot be seen fully by only looking at protein crystals.

Now, Clark, Dikiy et al. used another technique called nuclear magnetic resonance spectroscopy, or NMR for short, to understand how drugs affect the fast moving sidechains within a GPCR. First, genetically modified yeast was used to create samples of a GPCR called the adenosine receptor A_2A that were labelled with specific markers which made it easier to measure the structure and flexibility of the protein by NMR.

This approach revealed that too much sodium in the sample’s solution supresses the large structural changes that occur in the A_2A receptor when it binds to a drug. Moreover, it showed that the sidechains of several regions on the receptor move in different ways depending on whether the receptor binds to an activating drug or an inhibiting drug.

These findings lay the groundwork for understanding how the movements of sidechains help to activate or inhibit GPCRs, and will complement on-going studies using protein crystals. Moreover, the new approach to producing labelled proteins could be applied to other types of proteins that until now could not be studied with NMR due to practical limitations. In future, this may help scientists to better understand how drugs affect these proteins and to develop new treatments for a whole range of diseases.

Our understanding of the molecular underpinnings of GPCR function has been greatly advanced over the past two decades through a combination of X-ray crystal structures, computational simulations, and spectroscopic studies of protein dynamics. Crystals of bovine rhodopsin provided the first high-resolution picture of a GPCR’s architecture (Palczewski et al., 2000; Li et al., 2004), and structures of photoactivation intermediates (Nakamichi and Okada, 2006; Salom et al., 2006) and retinal-free opsin (Park et al., 2008; Scheerer et al., 2008) further documented the structural transitions involved in rhodopsin activation. For GPCRs activated by diffusible ligands, crystal structures of the human β₂ adrenergic receptor (β₂AR) with inverse agonists (Rosenbaum et al., 2007; Cherezov et al., 2007), agonists (Rosenbaum et al., 2011; Rasmussen et al., 2011a; Warne et al., 2011), and bound G protein (Rasmussen et al., 2011b) provided a molecular basis for understanding how diffusible agonist binding can promote structural changes in a receptor to enhance signaling. Subsequent crystal structures have revealed the ligand binding pockets for GPCRs of diverse function, responding to biogenic amines, purines, lipids, peptides, and proteins. While the sequences of these GPCRs and their orthosteric ligand binding pockets are highly diverse, the overall structures are remarkably similar, with Cα rmsd values between unrelated receptors in the 2–3 Å range. How these diverse ligands activate signaling by a small number of G proteins and arrestins through a common structural scaffold remains a central problem for the GPCR field. Contributing to this problem is the fact that most of the existing X-ray GPCR structures are of antagonist-bound receptors locked in inactive conformations. For the few GPCRs that have been captured in both inactive and active conformations, common structural changes include inward movements of transmembrane (TM) helices at the orthosteric pocket, rearrangements of common core residues through the helical bundle, and rigid-body outward movements of TM6 at the cytoplasmic surface to expose G protein binding epitopes (Katritch et al., 2013).

Beyond the static pictures of GPCR conformations seen in these crystal structures, experimental evidence for complex dynamic behavior has emerged from spectroscopic studies on purified receptors. Unlike rhodopsin, which exhibits an efficient and ordered photon-induced transition from the dark state conformation to metarhodopsin II (Choe et al., 2011a; Choe et al., 2011b), ligand-activated GPCRs generally show considerable basal activity that is reduced by inverse agonists. ¹⁹F NMR and EPR experiments on the β₂AR (Liu et al., 2012a; Manglik et al., 2015) and A_2AR (Ye et al., 2016) provided evidence for at least four different conformational states populated by these receptors, which are differentially stabilized by ligands of different efficacy. ¹H/¹³C 2D NMR studies on ¹³Cε-methionine-labeled β₂AR (Nygaard et al., 2013; Kofuku et al., 2014) showed that a high free-energy barrier exists between inactive and active conformations (with exchange on the millisecond timescale), and agonist binding alone is only capable of weakly populating the active conformation. These studies, in addition to similar measurements on μ-opioid receptor chemically modified to introduce NMR-active ¹³C-methyl labels onto lysine residues (Sounier et al., 2015), suggest that ligand-activated GPCRs are weakly coupled allosteric systems in which multiple energetic inputs (i.e. agonist and G protein binding) are required to predominantly populate the active conformation.

Within the slow millisecond-timescale global interconversion between functional GPCR conformations (Vilardaga et al., 2003), local structural rearrangements of sidechains in the receptor (referred to here as ‘microswitches’) have been observed in comparisons of inactive and active GPCR crystal structures (Katritch et al., 2013), NMR data (Liu et al., 2012a; Manglik et al., 2015; Ye et al., 2016; Bokoch et al., 2010) and MD simulations (Vanni et al., 2009; Dror et al., 2011a). The importance of repacking of sidechains from TM3, TM5, and TM6 beneath the ligand binding pocket, known as the ‘conserved core triad,’ was identified by comparison of structures of β₂AR (Rosenbaum et al., 2011; Rasmussen et al., 2011a) and the μ-opioid receptor (Huang et al., 2015). Further down the TM bundle, an activated microswitch near the site of G protein binding has been found to represent a common feature of GPCRs that have been crystallized in both active and inactive conformations (Venkatakrishnan et al., 2016). A working hypothesis for GPCR activation is that a set of loosely coupled microswitches connecting the orthosteric pocket and G protein binding site are activated by agonists in a non-concerted fashion, and stabilizing subsets of these rearrangements can lead to alternate overall conformations that have different signaling properties.

One of the most powerful ways to probe sidechain dynamics and their contributions to biological processes is provided by NMR-based studies of isotopically-labeled methyl groups (Ruschak and Kay, 2010). To apply these techniques to study the sidechains in these microswitches, one approach would entail isoleucine/leucine/valine (ILV) labeling (Goto et al., 1999; Gardner et al., 1998) and perdeuteration of a wild-type GPCR, to generate proteins with ¹H/¹³C-labeled methyl groups within an otherwise ²H/¹²C-labeled background. Such samples are ideally suited for acquiring ¹H/¹³C methyl TROSY-based (Tugarinov et al., 2003a; Ollerenshaw et al., 2003) relaxation NMR data, allowing the quantitative determination of methyl order parameters and relative motion of sidechains on the ps-ns timescale (Tugarinov et al., 2005; Ishima and Torchia, 2000). Advances in labeling methodology and pulse sequences have enabled such measurements on large macromolecular systems, such as the 80 kDa enzyme malate synthase (Tugarinov and Kay, 2003b; Sandala et al., 2007), the 670 kDa archaeal proteasome core particle (Sprangers and Kay, 2007; Religa et al., 2010), and the 1 MDa GroEL/GroES complex (Fiaux et al., 2002; Horst et al., 2005). NMR spectroscopic interrogation of methyl groups has also been used to dissect the energetic contribution of sidechain entropy to important biological phenomena such as protein-protein interactions (Marlow et al., 2010) and transcription factor binding to double-stranded DNA (Tzeng and Kalodimos, 2012). These previous applications suggest that characterization of sidechain dynamics using methyl NMR approaches should be feasible for a GPCR purified in a detergent micelle, with an aggregate molecular weight in the 100–150 kDa range.

A major hurdle in carrying out such experiments is that perdeuteration and ILV labeling conventionally rely on expression in E. coli (Goto et al., 1999), while wild-type GPCRs typically require a eukaryotic host to ensure proper folding, glycosylation, trafficking, and stability in the plasma membrane. GPCR NMR studies have primarily used supplementation with labeled amino acids (Nygaard et al., 2013; Kofuku et al., 2014; Okude et al., 2015; Isogai et al., 2016; Kofuku et al., 2012) or covalent modification (Liu et al., 2012a; Sounier et al., 2015; Bokoch et al., 2010; Eddy et al., 2016) to incorporate isotopic probes into protein from Sf9 cells, and perdeuteration with incorporation of labeled ILV methyl probes is largely intractable for this established eukaryotic system. Another limitation is that purified wild-type GPCRs are typically prone to aggregation and are not highly thermostable, limiting acquisition times and sample concentration. While high quality 2D NMR spectra of GPCRs have been obtained for receptors with thermostabilizing mutations (Isogai et al., 2016), such mutations can affect the receptor’s dynamics and functional properties, including basal activity and maximal G protein stimulation.

We have addressed many of these practical issues with our generation (Clark et al., 2015) of highly-deuterated ¹H/¹³C Ile δ1-methyl labeled proteins in Pichia pastoris, a methylotrophic yeast (Cereghino and Cregg, 2000) that is adaptable to D₂O-based minimal media, will utilize the Ile precursor α-ketobutyrate, and enables expression of GPCRs at levels sufficient for crystallography (Shimamura et al., 2011; Hino et al., 2012; Yurugi-Kobayashi et al., 2009). Here we apply this labeling method to the wild-type human A_2AR, an important regulator of vascular function with a well-developed pharmacology including the natural hormone adenosine and high-affinity synthetic agonists and antagonists (Preti et al., 2015; de Lera Ruiz et al., 2014). A_2AR has been crystallized in inactive (Jaakola et al., 2008; Liu et al., 2012b), intermediate (Lebon et al., 2011; Xu et al., 2011) and active (Carpenter et al., 2016) conformations, and has been studied by ¹⁹F NMR to characterize its conformational equilibria (Ye et al., 2016). We expressed and purified milligram quantities of labeled wild-type human A_2AR and were able to resolve 20 out of 29 expected peaks in the Ile δ1 region in ¹H/¹³C TROSY HMQC experiments. Through structure-guided mutagenesis, we assigned four of these signals to specific functionally important residues, including I92^3.40 of the conserved core triad (Huang et al., 2015) and I292 at the cytoplasmic surface. We collected spectra in the presence of the high-affinity full agonist NECA and the inverse agonist ZM241385, and also measured the effects of monovalent cations (Na⁺ versus K⁺) on these spectra. We further carried out a modified triple-quantum (3Q) relaxation experiment (Sun et al., 2011) to quantify the relative fast ps-ns motions of these sidechains while the receptor is bound to ligands of opposing efficacy. These data represent an important first step towards understanding how agonists can activate the fast motions of specific sidechains to facilitate conformational changes of a GPCR.

The first challenge in collecting methyl NMR spectra on a wild-type GPCR was achieving high-level expression and milligram scale purification from Pichia pastoris. We initially cloned several GPCRs (including A_2AR) into a modified methanol-inducible expression vector with the wild-type MFα signal sequence at the N-terminus to direct incorporation into the membrane. Several receptors, such as the wild-type M2 muscarinic receptor and A_2AR, showed reliable and reproducible expression, however much of the expressed proteins were present in the immature unprocessed form (without MFα cleavage) indicating that they had not been properly localized to the plasma membrane (Figure 1—figure supplement 1). To improve the efficiency of production of mature, folded receptors, we made several modifications to the MFα sequence, based on biochemical and genetic precedents (Rakestraw et al., 2009; Lin-Cereghino et al., 2013) for improved processing efficiency (Figure 1A, Materials and methods). To validate that Pichia-expressed A_2AR is functional, we carried out radioligand binding assays measuring [³H]-ZM241385 saturation and NECA competition on membranes after cell wall disruption and zymolyase treatment. The K_d for [³H]-ZM241385 was 0.42 nM in NaCl-containing buffer, while the K_i for NECA was 210 nM in low ionic strength and 420 nM in KCl-containing buffer, in agreement with previous binding studies (Xu et al., 2011; Carpenter et al., 2016; Bertheleme et al., 2013) on heterologously expressed A_2AR (Figure 1—figure supplement 2).

Figure 1 with 2 supplements see all

Download asset Open asset

We developed a purification protocol for A_2AR from Pichia-derived membranes, in which we can solubilize and purify ~0.5 mg of monodisperse biochemically pure wild-type A_2AR in dodecylmaltoside (DDM) detergent from 1 L of culture (Figure 1—figure supplement 1). Coupling this protocol with our previously-described labeling method (Clark et al., 2015), we were able to generate samples of wild-type A_2AR that are highly deuterated both within the protein (²H at non-exchangeable sites except Ile δ1 methyl groups) and the surrounding buffer (99% D₂O and deuterated DDM). We prioritized the use of a wild-type A_2AR construct to report on the dynamics of a physiologically and functionally relevant GPCR sample. To confirm that the high level of deuteration did not alter the receptor’s functional properties, we reconstituted the perdeuterated purified A_2AR together with purified G_s heterotrimer in phospholipid vesicles and carried out activation assays by [³⁵S]GTPγS binding. Saturating the receptor with the full agonist NECA led to an 8-fold stimulation of [³⁵S]GTPγS binding, and the inverse agonist ZM241385 produced a 50% reduction in basal activity, similar to values measured for protonated A_2AR (Figure 1B). These data show that the perdeuterated A_2AR is capable of activating G_s at wild-type levels in vitro, and can serve as a valid model for GPCR dynamics.

Our first NMR characterization of A_2AR used ¹H/¹³C TROSY HMQC spectra of the protein in the presence of different ligands or G_s. We initially expressed and solubilized the receptor with the low-affinity antagonist theophylline, and exchanged other ligands onto the receptor during washing in the Co²⁺-affinity and gel filtration steps (Materials and methods). In this way we were able to purify A_2AR samples with the inverse agonist ZM241385, the full agonist NECA, or no ligand. The resulting 2D HMQC spectra, requiring ~100 μM GPCR and a minimum of 2 hr acquisition at 30°C on a cryoprobe-equipped 800 MHz spectrometer, are shown in Figure 2. Overall, in a liganded sample, we could resolve 20 peaks in the Ile δ1 region of the spectrum, out of 29 Ile residues present in our construct (28 from the receptor and one from the C-terminal Protein C tag). The chemical shifts of the agonist-bound and inverse agonist-bound A_2AR are similar (Figure 2A,B), showing subtle changes that are comparable in magnitude to those observed previously for ¹³C-methyl-methionine probes in β₂AR (Kofuku et al., 2014; Kofuku et al., 2012). The spectrum of the unliganded (apo) A_2AR shows a loss of most of the well-dispersed peaks present in the ligand-bound spectra (Figure 2C), which could result from protein instability or conformational exchange on the μs-ms timescale. Lastly, we formed and isolated a high-affinity complex between purified NECA-bound A_2AR and purified G_s heterotrimer in the absence of GDP or GTP (Figure 2—figure supplement 1, Materials and methods). Like the apo spectrum, the G protein complex spectrum had fewer well-dispersed Ile peaks than receptor/ligand complexes (Figure 2D), consistent with the increase in molecular weight of the complex. We also note the appearance of a few distinct peaks (most prominently at 0.85 ppm ¹H, 11.8 ppm ¹³C) in the G protein complex, consistent with the conformation of the complexed receptor being distinguishable from the apo receptor (Rasmussen et al., 2011b; Carpenter et al., 2016).

Figure 2 with 3 supplements see all

Download asset Open asset

Assignment of ¹H/¹³C peaks in the Ile δ1 region by NMR was hampered by the modest dispersion of our spectra and the limited stability of A_2AR during data acquisition. Therefore, we took a site-specific mutagenesis approach, wherein we collected HMQC spectra of samples containing point mutations at a small subset of functionally important Ile sites (Figure 2 and Figure 2—figure supplement 2). This process was complicated by the sensitivity of the receptor to mutations at several of these Iles: changes to Val, Leu, or Met often drastically reduced purification yields and/or led to significantly altered spectra; for an example, see the I106V spectra in Figure 2—figure supplement 2. Three of the assignments (I92^3.40, I238^6.40, and I292) could be made primarily through observation of ZM241385-bound mutant spectra. By comparing WT and I92V spectra in the presence of ZM241385, we see a clear loss of signal at ~0.68ppm/13.6ppm (¹H/¹³C) that we can assign to the δ1 methyl of Ile92. The NECA-bound I92V spectrum is missing a subset of peaks compared to the ZM241385-bound spectrum, possibly due to compromised affinity of NECA for the I92V mutant (Figure 1—figure supplement 2D). Ile238 and Ile292 are well-dispersed peaks, allowing for unambiguous assignment through loss of peaks at 0.72 ppm/10.2 ppm and 1.12 ppm/12.7 ppm, respectively (¹H/¹³C). The mutant spectra of I238^6.40 and I292 have reciprocal effects on each other, where I238V causes a full peak-width shift in the I292 signal (upfield in ¹H), and I292L causes a ~ 0.5 peak-width shift in the I238 signal (upfield in ¹³C). The reciprocity between these mutant spectra is consistent with the proximity of these residues in the A_2AR structure. Mutation of I274^7.39 in the ligand binding pocket (within van der Waals contact of the ligands in the respective crystal structures) to Val or Met led to a disappearance of some peaks, reminiscent of the apo spectrum. However, comparison of ZM241385-bound I274M and I274V spectra narrowed the choice of the assignment of I274^7.39 to one of two peaks, with the greater solvent accessibility as revealed by solvent PRE effect (see below and Figure 4) of one peak serving as the deciding factor of assignment at 0.55 ppm/14.1 ppm (¹H/¹³C). As well as disappearance of many peaks, the spectrum of I274M A_2AR in the presence of ZM241385 displays several apparent peak doublings, suggesting that the mutant protein is undergoing slow conformational exchange, potentially related to the slow interconversion between different inactive states of the receptor (Ye et al., 2016). In addition, we assigned a peak to the Ile residue of the C-terminal Protein C tag in our construct by comparing spectra between differently tagged constructs (not shown). The remaining unassigned peaks are referred to as a-o in the rest of this work (Figure 3A). The chemical shift changes between ZM241385-bound and NECA-bound A_2AR are small overall (Figure 2A,B). In the A_2AR-G_s complex spectrum (Figure 2D), the peak for I92^3.40 appears to shift and simultaneously drop in intensity, strongly suggesting that I92^3.40 is involved in conformational changes on several timescales while the receptor is bound to G_s.

Figure 3 with 1 supplement see all

Download asset Open asset

Figure 4

Download asset Open asset

A_2AR contains a well-characterized Na⁺ binding site within the receptor core (Liu et al., 2012b), and Na⁺ has been shown to act as a negative allosteric modulator of A_2AR activation (Gao and Ijzerman, 2000). Due to the presence of 150 mM NaCl in the experiments shown in Figure 2, one possible explanation for the relatively small chemical shift differences between NECA- and ZM241385-bound spectra (Figure 3A,B) is that high [Na⁺] is suppressing structural changes in the agonist-bound receptor. To test this possibility, we acquired HMQC spectra for A_2AR samples with 150 mM KCl substituting for NaCl. The spectrum of ZM241385-bound receptor in KCl is very similar to that in the presence of NaCl (Figure 3C and Figure 3—figure supplement 1B), with the peak assigned to I92^3.40 shifting slightly upfield in the ¹³C dimension. The spectrum of NECA-bound A_2AR, on the other hand, is dramatically different in KCl and NaCl (Figure 3D, Figure 3—figure supplement 1D), with multiple peaks shifting and decreasing in intensity. However, the KCl samples were significantly less stable than those in NaCl, showing markedly decreased spectral quality after 1 (NECA-bound) or 2 (ZM241385-bound) hours of acquisition. Given the weak signal-to-noise and poor stability of the KCl samples, we were only able to perform NMR relaxation experiments (below) on A_2AR in the presence of NaCl.

In addition to ligand-induced chemical shift changes, we used our system to characterize the solvent accessibility of residues within the ZM241385-bound A_2AR (with NaCl) using solvent paramagnetic relaxation enhancements (PREs). To do so, we used the soluble paramagnetic Gd³⁺-DTPA probe (Petros et al., 1990), which enhances the relaxation rates of protons in a distance dependent manner within roughly 15 Å of the protein/solvent interface (Madl et al., 2011). By acquiring ¹H/¹³C TROSY HMQC spectra in the absence and presence of Gd³⁺-DTPA, we could compare peak intensities between the two to establish solvent PRE levels. Examining our data, we see that the average peak intensity in the Gd³⁺-DTPA containing spectrum is about 0.4 of that in the control spectrum, suggesting that the Ile δ1 methyl groups of A_2AR in the DDM micelle are quite close to solvent on average (Figure 4). Of the four most strongly affected peaks, three peaks correspond to assigned residues (I238^6.40, I274^7.39, and I292). The accessibility of I274^7.39 is expected due this amino acid’s presence in the solvent-exposed ligand-binding pocket. Broadening of I292 at the junction between TM7 and Helix 8 is also somewhat expected due to its position at the cytoplasmic surface. The broadening observed for I238^6.40 is more surprising, given that it is further embedded within the membrane and buried within the protein surface of A_2AR. Crystal structures of multiple GPCRs, including A_2AR, β₂AR, μOR, and rhodopsin have all revealed ordered solvent networks within the protein core in contact with the position equivalent to I238^6.40. In ZM241385-bound A_2AR, this residue is also in close proximity to the Na⁺ site (Liu et al., 2012b), one packing layer toward the cytoplasmic surface. The solvent PRE effect seen for this residue could reflect breathing of the structure to expose this region to the Gd³⁺-DTPA complex.

To characterize the motions of the Ile sidechains within A_2AR, we sought to carry out a triple quantum (3Q) relaxation experiment developed by Kay, Tugarinov and colleagues (Sun et al., 2011) to quantify sidechain dynamics in large macromolecules. In this experiment, two related series of 2D ¹H/¹³C HMQC-based spectra were acquired, both of which start by generating transverse ¹H magnetization composed of single quantum (SQ) coherences. During a subsequent variable delay period, intra-methyl ¹H−¹H dipolar cross-correlated relaxation mechanisms lead to a portion of these SQ coherences evolving into 3Q coherences; the rate of this conversion (η) is proportional to the S²_axis methyl order parameter. Practically, the degree of SQ->3Q evolution can be quantitated by examining paired ‘forbidden’ and ‘allowed’ spectra, which report on the fraction of SQ coherences which do and do not evolve into 3Q coherences, respectively. As previously reported (Sun et al., 2011), this allows the simple calculation of the η rate for each specific methyl group from the ratio of peak intensities in the forbidden and allowed spectra. Our initial attempts to apply this method to purified A_2AR were hampered by the fact that the wild-type receptor is insufficiently stable at 30°C to permit the measurement of the requisite paired spectra at many relaxation delays and with a high enough signal-to-noise ratio. We therefore explored the possibility of reducing the number of forbidden spectra required to accurately measure η values. Using similarly ²H,¹²C (¹H,¹³C δ1 methyl) labeled maltose binding protein (MBP) as a test case (Materials and methods), we collected 3Q datasets with different numbers of forbidden spectra paired with a constant number of allowed spectra (Figure 5, Figure 5—figure supplement 1). We found that pairing five allowed spectra with a single forbidden spectrum, we were able to faithfully recapitulate the η values measured using five pairs of forbidden and allowed experiments (Figure 5B,C). This η value is proportional to the S²_axis order parameter, which quantifies the amplitude of motion of the methyl group on the ps-ns timescale that is faster than global molecular tumbling.

Figure 5 with 1 supplement see all

Download asset Open asset

With this modified analysis of 3Q relaxation data and our Ile δ1 ¹H/¹³C-methyl-labeled and deuterated GPCR samples, we measured η values for Ile sidechains in A_2AR bound to either ZM241385 or NECA in NaCl. The resulting values for the 20 Ile peaks and relative changes between the samples with different ligands are shown in Figure 6. The error in these measurements is large due to the limited signal-to-noise of our A_2AR spectra, and we did not attempt to convert these η measurements to S²_axis values. However, in comparing the two datasets some features of the Ile dynamics can be discerned. The average η value for the ZM241385-bound sample is higher than for NECA, indicative of greater overall rigidity of sidechains with inverse agonist. At the individual residue level, we observe a diversity of ligand-dependent changes in η, with some Ile residues becoming more rigid with agonist while most become more flexible. As an internal control, the peak we assigned to the C-terminal Protein C tag shows the lowest η value measured (i.e. greatest flexibility) with little ligand-dependence. Among the other assigned peaks, I292 shows the largest difference between ligands, becoming more flexible in the NECA-bound sample. I92^3.40 and I274^7.39 display more modest increases in flexibility with agonist, while the dynamics of I238^6.40 are largely unchanged by ligand (Figure 6B). Interestingly, these differences in fast timescale dynamics occur in the presence of NaCl, in which chemical shifts for the agonist- and inverse agonist-bound states were nearly identical. To our knowledge, these data represent the first reported effort to experimentally quantify site-specific sidechain dynamics in a GPCR or comparable human integral membrane protein.

Figure 6 with 1 supplement see all

Download asset Open asset

While we were able to measure sidechain dynamics in this challenging A_2AR sample, the associated errors are quite large. As such, we turned to molecular dynamics (MD) simulations to provide independent validation of our dynamics measurements. We extracted S²_axis order parameters for Ile δ1 methyl groups from ~80 ns trajectories (post-equilibrium) of ZM241385- and NECA-bound A_2AR in DDM micelles, using overlapping 30-ns windows to estimate the standard deviation in the order parameters (Figure 6—figure supplement 1). The S²_axis order parameters for some methyl groups (for example, I60 and I80 in ZM241385; I106 and I302 in NECA) show significant variations across different windows, suggesting that slower timescale motions involving these residues contribute to the sidechain dynamics measured on the fast timescale. Interestingly, the differences between ZM241385- and NECA-bound dynamics at three of the four assigned sites (I238^6.40, I274^7.39, and I292) are qualitatively similar to what we see by NMR. The fourth site, I92^3.40, shows the opposite trend, with higher S²_axis, i.e. more rigidity, in the NECA-bound state. The simulations were set up with bulk Na⁺ ions, but none specifically occupying the binding site near I92^3.40, which could account for this discrepancy.

Sidechain dynamics represent an important functional component of protein behavior. In cases of protein-ligand (Marlow et al., 2010; Frederick et al., 2007) and protein-DNA (Tzeng and Kalodimos, 2012) interactions that have been intensively studied, changes in entropy arising from modified sidechain dynamics in complexes were found to substantially contribute toward the overall free energy of binding. For GPCRs, changes in sidechain dynamics may play an energetic role in binding to ligands and G proteins, however we have focused on allosteric mechanisms connecting these two functionally important binding sites. To experimentally assess the roles of sidechain dynamics in a wild-type GPCR, we set out to create a labeled sample that would be amenable to relaxation NMR methods. The spectra we obtained for the labeled Ile δ1 methyl groups of perdeuterated A_2AR are comparable or superior to previously published NMR spectra on other wild-type GPCRs, however the limited dispersion and signal-to-noise contributed to significant error in the relaxation values derived from the data (Figure 6A).

The Na⁺ site first seen in the high-resolution structure of A_2AR bound to ZM241285 is conserved throughout most Class A GPCRs, and Na⁺ exerts a negative allosteric effect on A_2AR activation by bridging residues on TM3 and TM7 and stabilizing the inactive conformation (Liu et al., 2012b). This effect can be observed pharmacologically as a NaCl-dependent decrease in agonist affinity (Carpenter et al., 2016; Gao and Ijzerman, 2000). For A_2AR expressed in P. pastoris, we observe a 9-fold decrease in NECA affinity when membranes are incubated in 150 mM NaCl buffer versus 150 mM KCl buffer (Figure 1—figure supplement 2), similar to previously reported values (Carpenter et al., 2016). Several of the Ile residues that we assigned (i.e. I92^3.40 and I238^6.40) are in close proximity to the Na⁺ binding site, which should make for sensitive probes of local structure in this part of the receptor core. Our 2D NMR spectra indicate a strong dependency on combined lack of Na⁺ and binding of agonist to stabilize significant structural changes in the receptor core (Figure 3). This observation is also consistent with the active conformation seen in the crystal structure of the A_2AR/mini-G_s complex (Carpenter et al., 2016), which results in collapse of the Na⁺ site due to inward movements of TM3 and TM7. In contrast to our data, the crystal structure of a NECA-bound thermostabilized A_2AR mutant construct (Lebon et al., 2011) showed significant structural changes relative to the inactive conformation even with a high [Na⁺] beyond its EC₅₀ (~50 mM) for negative allosteric modulation (Gao and Ijzerman, 2000). Unfortunately the weak signal and stability of our samples in KCl precluded measurement of 3Q relaxation dynamics under these conditions.

For the Na⁺-bound A_2AR samples, we can correlate our observations of ligand-dependent changes in the relaxation rate of specific sites to previous structural and biophysical studies of GPCRs. I274^7.39 at the ligand binding site makes direct contact with the adenine or adenine analogue rings of both NECA and ZM241385 in their respective crystal structures (Jaakola et al., 2008; Liu et al., 2012b; Lebon et al., 2011; Carpenter et al., 2016) (Figure 7A). In our relaxation datasets, I274^7.39 was more rigid in terms of its fast sidechain motions with inverse agonist bound, and more flexible with NECA occupying the binding pocket. NMR experiments previously showed different ligand-dependent conformations of the extracellular surface region of β₂AR bound to agonists compared to inverse agonists (Bokoch et al., 2010). In addition, MD simulations on β₂AR predicted greater mobility of agonists relative to inverse agonists in the orthosteric pocket of the inactive-state structure (i.e. not bound to G_s) (Dror et al., 2011b). Our data suggests that the fast dynamics of residues at the orthosteric pockets of Class A GPCRs may be correlated with the functional efficacy of the bound ligand. The residue I92^3.40 is one of the three residues of the conserved core triad in A_2AR, along with P189^5.50 and F242^6.44, which interact at a layer beneath the orthosteric pocket further toward the cytoplasmic surface (Figure 7B). Consistent with the rearrangements of this microswitch region between the inactive and active conformations of multiple GPCRs, we observe a lower η value for I92^3.40 in the NECA dataset versus ZM241385, indicating greater sidechain flexibility when A_2AR is agonist-bound. In several GPCRs, including β₂AR and μOR, the conserved core triad is unchanged at a static structural level with an agonist bound but without a G protein or nanobody to further stabilize the active conformation. In contrast, crystal structures of A_2AR bound to agonists alone (Figure 7B) showed an intermediate active-like conformation in the region surrounding I92^3.40 (Lebon et al., 2011; Xu et al., 2011). Our data indicate that low [Na⁺] is required for NECA alone to stabilize structural rearrangements surrounding I92 (Figure 3, Figure 3—figure supplement 1). However agonist alone (even with high [Na⁺]) is enough to at least promote the fast motions of the sidechains in this microswitch, which may reduce the activation energy for the observed packing rearrangement to the active conformation.

Figure 7

Download asset Open asset

The residue I238^6.40 is situated further toward the G protein binding site at a critical region for GPCR activation, where it packs against TM7 and undergoes significant outward movement in the transition to the active conformation (Figure 7C). I238^6.40 is also one helical turn on TM6 above L235^6.37, which participates in a conserved microswitch between inactive and active conformations for multiple GPCRs (Venkatakrishnan et al., 2016). I292 is present at the linker between TM7 and Helix 8, where it packs against Y288^7.53 of the highly conserved NPXXY motif (Katritch et al., 2013) (Figure 7D). NMR studies of ¹³C-dimethyllysine-labeled μOR showed that peak broadening of a lysine probe in Helix 8 (near the position equivalent to I292) was more sensitive to agonist than probes at the cytoplasmic ends of TM5 and TM6 (21). Further, in the structure of A_2AR/mini-G_s (Carpenter et al., 2016), the engineered G_s protein makes direct contact with I292 at the TM7-Helix 8 junction. In our relaxation dataset, the η value of I238^6.40 is low (i.e. more flexible) and essentially independent of the ligand, while I292 undergoes a large change from more rigid to more flexible from ZM241385 to NECA. As mentioned above, the environment surrounding I238^6.40 is loosely packed in A_2AR structures, with an ordered solvent network that may allow for relative freedom of motion for this sidechain. In contrast, despite the solvent exposure of I292 (Figure 4), inverse agonist binding suppresses the fast dynamics of this residue relative to agonist binding (Figure 6). Outward movement of TM6 is one of the hallmarks of GPCR activation seen in crystal structures, and the cytoplasmic ends of TM6 and TM7 are separated from each other in the structures of β₂AR and A_2AR bound to G_s (Rasmussen et al., 2011b; Carpenter et al., 2016). Our data suggest that the responsiveness of sidechain dynamics to ligands in these two regions may be largely uncoupled, and that the inverse agonist activity of ZM241385 could partly arise from its allosteric suppression of dynamics at the cytoplasmic end of TM7. Changes in η for other peaks in our dataset may provide further insights into the regulation of different regions of the receptor by ligands, depending on assignment of the other Ile residues in our spectra.

Beyond our studies of the A_2AR, we can now apply the methods for labeling and NMR spectroscopy described here to other GPCRs and eukaryotic membrane proteins. Since the microswitches discussed above were identified by comparison of different GPCR structures (Huang et al., 2015; Venkatakrishnan et al., 2016), it will be instructive to apply our methods to other receptors to see if the same patterns of ligand-dependent changes apply across the GPCR superfamily or change depending on the cognate ligand or preferred G protein signaling partner. Further, sidechain dynamics may be among the biophysical properties of GPCRs that are altered when receptors are bound to allosteric modulators or biased agonists, classes of ligands that are of increasing value and importance in GPCR pharmacology and drug development. In addition to GPCRs, many other disease-relevant human integral membrane proteins (such as ABC transporters and ion channels) are currently impossible to study by NMR methods, largely due to the challenges of expression, labeling, and perdeuteration in E. coli. Our approach has the potential to make these systems tractable for similar NMR measurements of sidechain dynamics.

1. 1. Lindsay D Clark
   2. Igor Dikiy
   3. Karen Chapman
   4. James Aramini
   5. Michael V LeVine
   6. George Khelashvili
   7. Søren GF Rasmussen
   8. Kevin H Gardner
   9. Daniel M Rosenbaum

   (2017) NMR spectroscopy data, including wild type and mutant spectra

   Publicly available at the Biological Magnetic Resonance Data Bank (accession no. 27261).

   http://www.bmrb.wisc.edu/ftp/pub/bmrb/entry_directories/bmr27261/

1. 1. Lebon G
   2. Warne T
   3. Edwards PC
   4. Bennett K
   5. Langmead CJ
   6. Leslie AGW
   7. Tate CG

   (2011) Thermostabilised HUMAN A2a Receptor with NECA bound

   Publicly available at the RCSB Protein Data Bank (Accession no: 2YDV).

   http://www.rcsb.org/pdb/explore.do?structureId=2ydv
2. 1. Sharff AJ
   2. Quiocho FA

   (1993) REFINED 1.8 ANGSTROMS STRUCTURE REVEALS THE MECHANISM OF BINDING OF A CYCLIC SUGAR, BETA-CYCLODEXTRIN, TO THE MALTODEXTRIN BINDING PROTEIN

   Publicly available at the RCSB Protein Data Bank (Accession no: 1DMB).

   http://www.rcsb.org/pdb/explore.do?structureId=1DMB
3. 1. Liu W
   2. Chun E
   3. Thompson AA
   4. Chubukov P
   5. Xu F
   6. Katritch V
   7. Han GW
   8. Heitman LH
   9. Ijzerman AP
   10. Cherezov V
   11. Stevens RC
   12. GPCR Network

   (2012) Crystal structure of the chimeric protein of A2aAR-BRIL in complex with ZM241385 at 1.8A resolution

   Publicly available at the RCSB Protein Data Bank (Accession no: 4EIY).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=4EIY

1. 1. Abel S
   2. Dupradeau FY
   3. Raman EP
   4. MacKerell AD
   5. Marchi M

   (2011) Molecular simulations of dodecyl-β-maltoside micelles in water: influence of the headgroup conformation and force field parameters

   The Journal of Physical Chemistry B 115:487–499.

   https://doi.org/10.1021/jp109545v

   • PubMed
   • Google Scholar
2. 1. Bertheleme N
   2. Singh S
   3. Dowell SJ
   4. Hubbard J
   5. Byrne B

   (2013) Loss of constitutive activity is correlated with increased thermostability of the human adenosine A2A receptor

   British Journal of Pharmacology 169:988–998.

   https://doi.org/10.1111/bph.12165

   • PubMed
   • Google Scholar
3. 1. Bokoch MP
   2. Zou Y
   3. Rasmussen SG
   4. Liu CW
   5. Nygaard R
   6. Rosenbaum DM
   7. Fung JJ
   8. Choi HJ
   9. Thian FS
   10. Kobilka TS
   11. Puglisi JD
   12. Weis WI
   13. Pardo L
   14. Prosser RS
   15. Mueller L
   16. Kobilka BK

   (2010) Ligand-specific regulation of the extracellular surface of a G-protein-coupled receptor

   Nature 463:108–112.

   https://doi.org/10.1038/nature08650

   • PubMed
   • Google Scholar
4. 1. Bremi T
   2. Brüschweiler R
   3. Ernst RR

   (1997) A protocol for the interpretation of side-chain dynamics based on nmr relaxation: application to phenylalanines in antamanide

   Journal of the American Chemical Society 119:4272–4284.

   https://doi.org/10.1021/ja9636505

   • Google Scholar
5. 1. Brooks BR
   2. Brooks CL
   3. Mackerell AD
   4. Nilsson L
   5. Petrella RJ
   6. Roux B
   7. Won Y
   8. Archontis G
   9. Bartels C
   10. Boresch S
   11. Caflisch A
   12. Caves L
   13. Cui Q
   14. Dinner AR
   15. Feig M
   16. Fischer S
   17. Gao J
   18. Hodoscek M
   19. Im W
   20. Kuczera K
   21. Lazaridis T
   22. Ma J
   23. Ovchinnikov V
   24. Paci E
   25. Pastor RW
   26. Post CB
   27. Pu JZ
   28. Schaefer M
   29. Tidor B
   30. Venable RM
   31. Woodcock HL
   32. Wu X
   33. Yang W
   34. York DM
   35. Karplus M

   (2009) CHARMM: the biomolecular simulation program

   Journal of Computational Chemistry 30:1545–1614.

   https://doi.org/10.1002/jcc.21287

   • PubMed
   • Google Scholar
6. 1. Carpenter B
   2. Nehmé R
   3. Warne T
   4. Leslie AG
   5. Tate CG

   (2016) Structure of the adenosine A(2A) receptor bound to an engineered G protein

   Nature 536:104–107.

   https://doi.org/10.1038/nature18966

   • PubMed
   • Google Scholar
7. 1. Cereghino JL
   2. Cregg JM

   (2000) Heterologous protein expression in the methylotrophic yeast Pichia pastoris

   FEMS Microbiology Reviews 24:45–66.

   https://doi.org/10.1111/j.1574-6976.2000.tb00532.x

   • PubMed
   • Google Scholar
8. 1. Cherezov V
   2. Rosenbaum DM
   3. Hanson MA
   4. Rasmussen SG
   5. Thian FS
   6. Kobilka TS
   7. Choi HJ
   8. Kuhn P
   9. Weis WI
   10. Kobilka BK
   11. Stevens RC

   (2007) High-resolution crystal structure of an engineered human beta2-adrenergic G protein-coupled receptor

   Science 318:1258–1265.

   https://doi.org/10.1126/science.1150577

   • PubMed
   • Google Scholar
9. 1. Choe HW
   2. Kim YJ
   3. Park JH
   4. Morizumi T
   5. Pai EF
   6. Krauss N
   7. Hofmann KP
   8. Scheerer P
   9. Ernst OP

   (2011b) Crystal structure of metarhodopsin II

   Nature 471:651–655.

   https://doi.org/10.1038/nature09789

   • PubMed
   • Google Scholar
10. 1. Choe HW
    2. Park JH
    3. Kim YJ
    4. Ernst OP

    (2011a) Transmembrane signaling by GPCRs: insight from rhodopsin and opsin structures

    Neuropharmacology 60:52–57.

    https://doi.org/10.1016/j.neuropharm.2010.07.018

    • PubMed
    • Google Scholar
11. 1. Clark L
    2. Zahm JA
    3. Ali R
    4. Kukula M
    5. Bian L
    6. Patrie SM
    7. Gardner KH
    8. Rosen MK
    9. Rosenbaum DM

    (2015) Methyl labeling and TROSY NMR spectroscopy of proteins expressed in the eukaryote Pichia pastoris

    Journal of Biomolecular NMR 62:239–245.

    https://doi.org/10.1007/s10858-015-9939-2

    • PubMed
    • Google Scholar
12. 1. de Lera Ruiz M
    2. Lim YH
    3. Zheng J

    (2014) Adenosine A2A receptor as a drug discovery target

    Journal of Medicinal Chemistry 57:3623–3650.

    https://doi.org/10.1021/jm4011669

    • PubMed
    • Google Scholar
13. 1. Delaglio F
    2. Grzesiek S
    3. Vuister GW
    4. Zhu G
    5. Pfeifer J
    6. Bax A

    (1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes

    Journal of Biomolecular NMR 6:277–293.

    https://doi.org/10.1007/BF00197809

    • PubMed
    • Google Scholar
14. 1. Dror RO
    2. Arlow DH
    3. Maragakis P
    4. Mildorf TJ
    5. Pan AC
    6. Xu H
    7. Borhani DW
    8. Shaw DE

    (2011a) Activation mechanism of the β2-adrenergic receptor

    PNAS 108:18684–18689.

    https://doi.org/10.1073/pnas.1110499108

    • PubMed
    • Google Scholar
15. 1. Dror RO
    2. Pan AC
    3. Arlow DH
    4. Borhani DW
    5. Maragakis P
    6. Shan Y
    7. Xu H
    8. Shaw DE

    (2011b) Pathway and mechanism of drug binding to G-protein-coupled receptors

    PNAS 108:13118–13123.

    https://doi.org/10.1073/pnas.1104614108

    • PubMed
    • Google Scholar
16. 1. Eddy MT
    2. Didenko T
    3. Stevens RC
    4. Wüthrich K

    (2016) β2-adrenergic receptor conformational response to fusion protein in the third intracellular loop

    Structure 24:2190–2197.

    https://doi.org/10.1016/j.str.2016.09.015

    • PubMed
    • Google Scholar
17. 1. Efron B

    (1981) Nonparametric estimates of standard error: The jackknife, the bootstrap and other methods

    Biometrika 68:589–599.

    https://doi.org/10.1093/biomet/68.3.589

    • Google Scholar
18. 1. Essmann U
    2. Perera L
    3. Berkowitz ML
    4. Darden T
    5. Lee H
    6. Pedersen LG

    (1995) A smooth particle mesh Ewald method

    The Journal of Chemical Physics 103:8577–8593.

    https://doi.org/10.1063/1.470117

    • Google Scholar
19. 1. Fiaux J
    2. Bertelsen EB
    3. Horwich AL
    4. Wüthrich K

    (2002) NMR analysis of a 900K GroEL GroES complex

    Nature 418:207–211.

    https://doi.org/10.1038/nature00860

    • PubMed
    • Google Scholar
20. 1. Frederick KK
    2. Marlow MS
    3. Valentine KG
    4. Wand AJ

    (2007) Conformational entropy in molecular recognition by proteins

    Nature 448:325–329.

    https://doi.org/10.1038/nature05959

    • PubMed
    • Google Scholar
21. 1. Gao ZG
    2. Ijzerman AP

    (2000) Allosteric modulation of A(2A) adenosine receptors by amiloride analogues and sodium ions

    Biochemical Pharmacology 60:669–676.

    https://doi.org/10.1016/S0006-2952(00)00360-9

    • PubMed
    • Google Scholar
22. 1. Gardner KH
    2. Zhang X
    3. Gehring K
    4. Kay LE

    (1998) Solution nmr studies of a 42 kda Escherichia Coli maltose binding protein/β-cyclodextrin complex: chemical shift assignments and analysis

    Journal of the American Chemical Society 120:11738–11748.

    https://doi.org/10.1021/ja982019w

    • Google Scholar
23. 1. Goto NK
    2. Gardner KH
    3. Mueller GA
    4. Willis RC
    5. Kay LE

    (1999)

    A robust and cost-effective method for the production of Val, Leu, Ile (delta 1) methyl-protonated 15N-, 13C-, 2H-labeled proteins

    Journal of Biomolecular NMR 13:369–374.

    • PubMed
    • Google Scholar
24. 1. Hino T
    2. Arakawa T
    3. Iwanari H
    4. Yurugi-Kobayashi T
    5. Ikeda-Suno C
    6. Nakada-Nakura Y
    7. Kusano-Arai O
    8. Weyand S
    9. Shimamura T
    10. Nomura N
    11. Cameron AD
    12. Kobayashi T
    13. Hamakubo T
    14. Iwata S
    15. Murata T

    (2012) G-protein-coupled receptor inactivation by an allosteric inverse-agonist antibody

    Nature 482:237–240.

    https://doi.org/10.1038/nature10750

    • PubMed
    • Google Scholar
25. 1. Horst R
    2. Bertelsen EB
    3. Fiaux J
    4. Wider G
    5. Horwich AL
    6. Wüthrich K

    (2005) Direct NMR observation of a substrate protein bound to the chaperonin GroEL

    PNAS 102:12748–12753.

    https://doi.org/10.1073/pnas.0505642102

    • PubMed
    • Google Scholar
26. 1. Huang W
    2. Manglik A
    3. Venkatakrishnan AJ
    4. Laeremans T
    5. Feinberg EN
    6. Sanborn AL
    7. Kato HE
    8. Livingston KE
    9. Thorsen TS
    10. Kling RC
    11. Granier S
    12. Gmeiner P
    13. Husbands SM
    14. Traynor JR
    15. Weis WI
    16. Steyaert J
    17. Dror RO
    18. Kobilka BK

    (2015) Structural insights into µ-opioid receptor activation

    Nature 524:315–321.

    https://doi.org/10.1038/nature14886

    • PubMed
    • Google Scholar
27. 1. Humphrey W
    2. Dalke A
    3. Schulten K

    (1996) VMD: visual molecular dynamics

    Journal of Molecular Graphics 14:33–38.

    https://doi.org/10.1016/0263-7855(96)00018-5

    • PubMed
    • Google Scholar
28. 1. Ishima R
    2. Torchia DA

    (2000) Protein dynamics from NMR

    Nature Structural Biology 7:740–743.

    https://doi.org/10.1038/78963

    • PubMed
    • Google Scholar
29. 1. Isogai S
    2. Deupi X
    3. Opitz C
    4. Heydenreich FM
    5. Tsai CJ
    6. Brueckner F
    7. Schertler GF
    8. Veprintsev DB
    9. Grzesiek S

    (2016) Backbone NMR reveals allosteric signal transduction networks in the β1-adrenergic receptor

    Nature 530:237–241.

    https://doi.org/10.1038/nature16577

    • PubMed
    • Google Scholar
30. 1. Jaakola VP
    2. Griffith MT
    3. Hanson MA
    4. Cherezov V
    5. Chien EY
    6. Lane JR
    7. Ijzerman AP
    8. Stevens RC

    (2008) The 2.6 angstrom crystal structure of a human A2A adenosine receptor bound to an antagonist

    Science 322:1211–1217.

    https://doi.org/10.1126/science.1164772

    • PubMed
    • Google Scholar
31. 1. Johnson BA

    (2004) Using NMRView to visualize and analyze the NMR spectra of macromolecules

    Methods in Molecular Biology 278:313–352.

    https://doi.org/10.1385/1-59259-809-9:313

    • PubMed
    • Google Scholar
32. 1. Jones E
    2. Oliphant T
    3. Peterson P

    (2001) SciPy: Open Source Scientific Tools for Python

    SciPy: Open Source Scientific Tools for Python, http://www.scipy.org/.

    http://www.scipy.org/

    • Google Scholar
33. 1. Katritch V
    2. Cherezov V
    3. Stevens RC

    (2013) Structure-function of the G protein-coupled receptor superfamily

    Annual Review of Pharmacology and Toxicology 53:531–556.

    https://doi.org/10.1146/annurev-pharmtox-032112-135923

    • PubMed
    • Google Scholar
34. 1. Kaufmann TC
    2. Engel A
    3. Rémigy HW

    (2006) A novel method for detergent concentration determination

    Biophysical Journal 90:310–317.

    https://doi.org/10.1529/biophysj.105.070193

    • PubMed
    • Google Scholar
35. 1. Khelashvili G
    2. LeVine MV
    3. Shi L
    4. Quick M
    5. Javitch JA
    6. Weinstein H

    (2013) The membrane protein LeuT in micellar systems: aggregation dynamics and detergent binding to the S2 site

    Journal of the American Chemical Society 135:14266–14275.

    https://doi.org/10.1021/ja405984v

    • PubMed
    • Google Scholar
36. 1. Kofuku Y
    2. Ueda T
    3. Okude J
    4. Shiraishi Y
    5. Kondo K
    6. Maeda M
    7. Tsujishita H
    8. Shimada I

    (2012) Efficacy of the β₂-adrenergic receptor is determined by conformational equilibrium in the transmembrane region

    Nature Communications 3:1045.

    https://doi.org/10.1038/ncomms2046

    • PubMed
    • Google Scholar
37. 1. Kofuku Y
    2. Ueda T
    3. Okude J
    4. Shiraishi Y
    5. Kondo K
    6. Mizumura T
    7. Suzuki S
    8. Shimada I

    (2014) Functional dynamics of deuterated β2 -adrenergic receptor in lipid bilayers revealed by NMR spectroscopy

    Angewandte Chemie International Edition 53:13376–13379.

    https://doi.org/10.1002/anie.201406603

    • PubMed
    • Google Scholar
38. 1. Lebon G
    2. Warne T
    3. Edwards PC
    4. Bennett K
    5. Langmead CJ
    6. Leslie AG
    7. Tate CG

    (2011) Agonist-bound adenosine A2A receptor structures reveal common features of GPCR activation

    Nature 474:521–525.

    https://doi.org/10.1038/nature10136

    • PubMed
    • Google Scholar
39. 1. LeVine MV
    2. Khelashvili G
    3. Shi L
    4. Quick M
    5. Javitch JA
    6. Weinstein H

    (2016) Role of Annular Lipids in the Functional Properties of Leucine Transporter LeuT Proteomicelles

    Biochemistry 55:850–859.

    https://doi.org/10.1021/acs.biochem.5b01268

    • PubMed
    • Google Scholar
40. 1. Li J
    2. Edwards PC
    3. Burghammer M
    4. Villa C
    5. Schertler GF

    (2004) Structure of bovine rhodopsin in a trigonal crystal form

    Journal of Molecular Biology 343:1409–1438.

    https://doi.org/10.1016/j.jmb.2004.08.090

    • PubMed
    • Google Scholar
41. 1. Lin-Cereghino GP
    2. Stark CM
    3. Kim D
    4. Chang J
    5. Shaheen N
    6. Poerwanto H
    7. Agari K
    8. Moua P
    9. Low LK
    10. Tran N
    11. Huang AD
    12. Nattestad M
    13. Oshiro KT
    14. Chang JW
    15. Chavan A
    16. Tsai JW
    17. Lin-Cereghino J

    (2013) The effect of α-mating factor secretion signal mutations on recombinant protein expression in Pichia pastoris

    Gene 519:311–317.

    https://doi.org/10.1016/j.gene.2013.01.062

    • PubMed
    • Google Scholar
42. 1. Liu JJ
    2. Horst R
    3. Katritch V
    4. Stevens RC
    5. Wüthrich K

    (2012a) Biased signaling pathways in β2-adrenergic receptor characterized by 19F-NMR

    Science 335:1106–1110.

    https://doi.org/10.1126/science.1215802

    • PubMed
    • Google Scholar
43. 1. Liu W
    2. Chun E
    3. Thompson AA
    4. Chubukov P
    5. Xu F
    6. Katritch V
    7. Han GW
    8. Roth CB
    9. Heitman LH
    10. IJzerman AP
    11. Cherezov V
    12. Stevens RC

    (2012b) Structural basis for allosteric regulation of GPCRs by sodium ions

    Science 337:232–236.

    https://doi.org/10.1126/science.1219218

    • PubMed
    • Google Scholar
44. 1. Madl T
    2. Güttler T
    3. Görlich D
    4. Sattler M

    (2011) Structural analysis of large protein complexes using solvent paramagnetic relaxation enhancements

    Angewandte Chemie International Edition 50:3993–3997.

    https://doi.org/10.1002/anie.201007168

    • PubMed
    • Google Scholar
45. 1. Manglik A
    2. Kim TH
    3. Masureel M
    4. Altenbach C
    5. Yang Z
    6. Hilger D
    7. Lerch MT
    8. Kobilka TS
    9. Thian FS
    10. Hubbell WL
    11. Prosser RS
    12. Kobilka BK

    (2015) Structural Insights into the Dynamic Process of β2-Adrenergic Receptor Signaling

    Cell 161:1101–1111.

    https://doi.org/10.1016/j.cell.2015.04.043

    • PubMed
    • Google Scholar
46. 1. Marlow MS
    2. Dogan J
    3. Frederick KK
    4. Valentine KG
    5. Wand AJ

    (2010) The role of conformational entropy in molecular recognition by calmodulin

    Nature Chemical Biology 6:352–358.

    https://doi.org/10.1038/nchembio.347

    • PubMed
    • Google Scholar
47. 1. Mittermaier A
    2. Kay LE
    3. Forman-Kay JD

    (1999) Analysis of deuterium relaxation-derived methyl axis order parameters and correlation with local structure

    Journal of Biomolecular NMR 13:181–185.

    https://doi.org/10.1023/A:1008387715167

    • PubMed
    • Google Scholar
48. 1. More JJ

    (1977)

    Lecture Notes in Mathematics 630

    105–116, The Levenberg-Marquardt Algorithm: Implementation and Theory, in Numerical Analysis, Lecture Notes in Mathematics 630, Watson GA, Springer Verlag.

    • Google Scholar
49. 1. Nakamichi H
    2. Okada T

    (2006) Local peptide movement in the photoreaction intermediate of rhodopsin

    PNAS 103:12729–12734.

    https://doi.org/10.1073/pnas.0601765103

    • PubMed
    • Google Scholar
50. 1. Nygaard R
    2. Zou Y
    3. Dror RO
    4. Mildorf TJ
    5. Arlow DH
    6. Manglik A
    7. Pan AC
    8. Liu CW
    9. Fung JJ
    10. Bokoch MP
    11. Thian FS
    12. Kobilka TS
    13. Shaw DE
    14. Mueller L
    15. Prosser RS
    16. Kobilka BK

    (2013) The dynamic process of β(2)-adrenergic receptor activation

    Cell 152:532–542.

    https://doi.org/10.1016/j.cell.2013.01.008

    • PubMed
    • Google Scholar
51. 1. Okude J
    2. Ueda T
    3. Kofuku Y
    4. Sato M
    5. Nobuyama N
    6. Kondo K
    7. Shiraishi Y
    8. Mizumura T
    9. Onishi K
    10. Natsume M
    11. Maeda M
    12. Tsujishita H
    13. Kuranaga T
    14. Inoue M
    15. Shimada I

    (2015) Identification of a conformational equilibrium that determines the efficacy and functional selectivity of the μ-opioid receptor

    Angewandte Chemie International Edition 54:15771–15776.

    https://doi.org/10.1002/anie.201508794

    • PubMed
    • Google Scholar
52. 1. Ollerenshaw JE
    2. Tugarinov V
    3. Kay LE

    (2003) Methyl TROSY: explanation and experimental verification

    Magnetic Resonance in Chemistry 41:843–852.

    https://doi.org/10.1002/mrc.1256

    • Google Scholar
53. 1. Palczewski K
    2. Kumasaka T
    3. Hori T
    4. Behnke CA
    5. Motoshima H
    6. Fox BA
    7. Le Trong I
    8. Teller DC
    9. Okada T
    10. Stenkamp RE
    11. Yamamoto M
    12. Miyano M

    (2000) Crystal structure of rhodopsin: A G protein-coupled receptor

    Science 289:739–745.

    https://doi.org/10.1126/science.289.5480.739

    • PubMed
    • Google Scholar
54. 1. Park JH
    2. Scheerer P
    3. Hofmann KP
    4. Choe HW
    5. Ernst OP

    (2008) Crystal structure of the ligand-free G-protein-coupled receptor opsin

    Nature 454:183–187.

    https://doi.org/10.1038/nature07063

    • PubMed
    • Google Scholar
55. 1. Petros AM
    2. Mueller L
    3. Kopple KD

    (1990) NMR identification of protein surfaces using paramagnetic probes

    Biochemistry 29:10041–10048.

    https://doi.org/10.1021/bi00495a005

    • PubMed
    • Google Scholar
56. 1. Phillips JC
    2. Braun R
    3. Wang W
    4. Gumbart J
    5. Tajkhorshid E
    6. Villa E
    7. Chipot C
    8. Skeel RD
    9. Kalé L
    10. Schulten K

    (2005) Scalable molecular dynamics with NAMD

    Journal of Computational Chemistry 26:1781–1802.

    https://doi.org/10.1002/jcc.20289

    • PubMed
    • Google Scholar
57. 1. Preti D
    2. Baraldi PG
    3. Moorman AR
    4. Borea PA
    5. Varani K

    (2015) History and perspectives of A2A adenosine receptor antagonists as potential therapeutic agents

    Medicinal Research Reviews 35:790–848.

    https://doi.org/10.1002/med.21344

    • PubMed
    • Google Scholar
58. 1. Rakestraw JA
    2. Sazinsky SL
    3. Piatesi A
    4. Antipov E
    5. Wittrup KD

    (2009) Directed evolution of a secretory leader for the improved expression of heterologous proteins and full-length antibodies in Saccharomyces cerevisiae

    Biotechnology and Bioengineering 103:1192–1201.

    https://doi.org/10.1002/bit.22338

    • PubMed
    • Google Scholar
59. 1. Rasmussen SG
    2. Choi HJ
    3. Fung JJ
    4. Pardon E
    5. Casarosa P
    6. Chae PS
    7. Devree BT
    8. Rosenbaum DM
    9. Thian FS
    10. Kobilka TS
    11. Schnapp A
    12. Konetzki I
    13. Sunahara RK
    14. Gellman SH
    15. Pautsch A
    16. Steyaert J
    17. Weis WI
    18. Kobilka BK

    (2011a) Structure of a nanobody-stabilized active state of the β(2) adrenoceptor

    Nature 469:175–180.

    https://doi.org/10.1038/nature09648

    • PubMed
    • Google Scholar
60. 1. Rasmussen SG
    2. DeVree BT
    3. Zou Y
    4. Kruse AC
    5. Chung KY
    6. Kobilka TS
    7. Thian FS
    8. Chae PS
    9. Pardon E
    10. Calinski D
    11. Mathiesen JM
    12. Shah ST
    13. Lyons JA
    14. Caffrey M
    15. Gellman SH
    16. Steyaert J
    17. Skiniotis G
    18. Weis WI
    19. Sunahara RK
    20. Kobilka BK

    (2011b) Crystal structure of the β2 adrenergic receptor-Gs protein complex

    Nature 477:549–555.

    https://doi.org/10.1038/nature10361

    • PubMed
    • Google Scholar
61. 1. Religa TL
    2. Sprangers R
    3. Kay LE

    (2010) Dynamic regulation of archaeal proteasome gate opening as studied by TROSY NMR

    Science 328:98–102.

    https://doi.org/10.1126/science.1184991

    • PubMed
    • Google Scholar
62. 1. Rosenbaum DM
    2. Cherezov V
    3. Hanson MA
    4. Rasmussen SG
    5. Thian FS
    6. Kobilka TS
    7. Choi HJ
    8. Yao XJ
    9. Weis WI
    10. Stevens RC
    11. Kobilka BK

    (2007) GPCR engineering yields high-resolution structural insights into beta2-adrenergic receptor function

    Science 318:1266–1273.

    https://doi.org/10.1126/science.1150609

    • PubMed
    • Google Scholar
63. 1. Rosenbaum DM
    2. Zhang C
    3. Lyons JA
    4. Holl R
    5. Aragao D
    6. Arlow DH
    7. Rasmussen SG
    8. Choi HJ
    9. Devree BT
    10. Sunahara RK
    11. Chae PS
    12. Gellman SH
    13. Dror RO
    14. Shaw DE
    15. Weis WI
    16. Caffrey M
    17. Gmeiner P
    18. Kobilka BK

    (2011) Structure and function of an irreversible agonist-β(2) adrenoceptor complex

    Nature 469:236–240.

    https://doi.org/10.1038/nature09665

    • PubMed
    • Google Scholar
64. 1. Ruschak AM
    2. Kay LE

    (2010) Methyl groups as probes of supra-molecular structure, dynamics and function

    Journal of Biomolecular NMR 46:75–87.

    https://doi.org/10.1007/s10858-009-9376-1

    • PubMed
    • Google Scholar
65. 1. Sali A
    2. Blundell TL

    (1993) Comparative protein modelling by satisfaction of spatial restraints

    Journal of Molecular Biology 234:779–815.

    https://doi.org/10.1006/jmbi.1993.1626

    • PubMed
    • Google Scholar
66. 1. Salom D
    2. Lodowski DT
    3. Stenkamp RE
    4. Le Trong I
    5. Golczak M
    6. Jastrzebska B
    7. Harris T
    8. Ballesteros JA
    9. Palczewski K

    (2006) Crystal structure of a photoactivated deprotonated intermediate of rhodopsin

    PNAS 103:16123–16128.

    https://doi.org/10.1073/pnas.0608022103

    • PubMed
    • Google Scholar
67. 1. Sandala GM
    2. Smith DM
    3. Marsh EN
    4. Radom L

    (2007) Toward an improved understanding of the glutamate mutase system

    Journal of the American Chemical Society 129:1623–1633.

    https://doi.org/10.1021/ja066432c

    • PubMed
    • Google Scholar
68. 1. Scheerer P
    2. Park JH
    3. Hildebrand PW
    4. Kim YJ
    5. Krauss N
    6. Choe HW
    7. Hofmann KP
    8. Ernst OP

    (2008) Crystal structure of opsin in its G-protein-interacting conformation

    Nature 455:497–502.

    https://doi.org/10.1038/nature07330

    • PubMed
    • Google Scholar
69. 1. Sharff AJ
    2. Rodseth LE
    3. Quiocho FA

    (1993) Refined 1.8-A structure reveals the mode of binding of beta-cyclodextrin to the maltodextrin binding protein

    Biochemistry 32:10553–10559.

    https://doi.org/10.1021/bi00091a004

    • PubMed
    • Google Scholar
70. 1. Shimamura T
    2. Shiroishi M
    3. Weyand S
    4. Tsujimoto H
    5. Winter G
    6. Katritch V
    7. Abagyan R
    8. Cherezov V
    9. Liu W
    10. Han GW
    11. Kobayashi T
    12. Stevens RC
    13. Iwata S

    (2011) Structure of the human histamine H1 receptor complex with doxepin

    Nature 475:65–70.

    https://doi.org/10.1038/nature10236

    • PubMed
    • Google Scholar
71. 1. Sounier R
    2. Mas C
    3. Steyaert J
    4. Laeremans T
    5. Manglik A
    6. Huang W
    7. Kobilka BK
    8. Déméné H
    9. Granier S

    (2015) Propagation of conformational changes during μ-opioid receptor activation

    Nature 524:375–378.

    https://doi.org/10.1038/nature14680

    • PubMed
    • Google Scholar
72. 1. Sprangers R
    2. Kay LE

    (2007) Quantitative dynamics and binding studies of the 20S proteasome by NMR

    Nature 445:618–622.

    https://doi.org/10.1038/nature05512

    • PubMed
    • Google Scholar
73. 1. Sun B
    2. Bachhawat P
    3. Chu ML
    4. Wood M
    5. Ceska T
    6. Sands ZA
    7. Mercier J
    8. Lebon F
    9. Kobilka TS
    10. Kobilka BK

    (2017) Crystal structure of the adenosine A2A receptor bound to an antagonist reveals a potential allosteric pocket

    PNAS 114:2066–2071.

    https://doi.org/10.1073/pnas.1621423114

    • PubMed
    • Google Scholar
74. 1. Sun H
    2. Kay LE
    3. Tugarinov V

    (2011) An optimized relaxation-based coherence transfer NMR experiment for the measurement of side-chain order in methyl-protonated, highly deuterated proteins

    The Journal of Physical Chemistry B 115:14878–14884.

    https://doi.org/10.1021/jp209049k

    • PubMed
    • Google Scholar
75. 1. Tugarinov V
    2. Hwang PM
    3. Ollerenshaw JE
    4. Kay LE

    (2003a) Cross-correlated relaxation enhanced 1H[bond]13C NMR spectroscopy of methyl groups in very high molecular weight proteins and protein complexes

    Journal of the American Chemical Society 125:10420–10428.

    https://doi.org/10.1021/ja030153x

    • PubMed
    • Google Scholar
76. 1. Tugarinov V
    2. Kay LE

    (2003b) Ile, Leu, and Val methyl assignments of the 723-residue malate synthase G using a new labeling strategy and novel NMR methods

    Journal of the American Chemical Society 125:13868–13878.

    https://doi.org/10.1021/ja030345s

    • PubMed
    • Google Scholar
77. 1. Tugarinov V
    2. Ollerenshaw JE
    3. Kay LE

    (2005) Probing side-chain dynamics in high molecular weight proteins by deuterium NMR spin relaxation: an application to an 82-kDa enzyme

    Journal of the American Chemical Society 127:8214–8225.

    https://doi.org/10.1021/ja0508830

    • PubMed
    • Google Scholar
78. 1. Tzeng SR
    2. Kalodimos CG

    (2012) Protein activity regulation by conformational entropy

    Nature 488:236–240.

    https://doi.org/10.1038/nature11271

    • PubMed
    • Google Scholar
79. 1. Vanni S
    2. Neri M
    3. Tavernelli I
    4. Rothlisberger U

    (2009) Observation of "ionic lock" formation in molecular dynamics simulations of wild-type beta 1 and beta 2 adrenergic receptors

    Biochemistry 48:4789–4797.

    https://doi.org/10.1021/bi900299f

    • PubMed
    • Google Scholar
80. 1. Venkatakrishnan AJ
    2. Deupi X
    3. Lebon G
    4. Heydenreich FM
    5. Flock T
    6. Miljus T
    7. Balaji S
    8. Bouvier M
    9. Veprintsev DB
    10. Tate CG
    11. Schertler GF
    12. Babu MM

    (2016) Diverse activation pathways in class A GPCRs converge near the G-protein-coupling region

    Nature 536:484–487.

    https://doi.org/10.1038/nature19107

    • PubMed
    • Google Scholar
81. 1. Vilardaga JP
    2. Bünemann M
    3. Krasel C
    4. Castro M
    5. Lohse MJ

    (2003) Measurement of the millisecond activation switch of G protein-coupled receptors in living cells

    Nature Biotechnology 21:807–812.

    https://doi.org/10.1038/nbt838

    • PubMed
    • Google Scholar
82. 1. Warne T
    2. Moukhametzianov R
    3. Baker JG
    4. Nehmé R
    5. Edwards PC
    6. Leslie AG
    7. Schertler GF
    8. Tate CG

    (2011) The structural basis for agonist and partial agonist action on a β(1)-adrenergic receptor

    Nature 469:241–244.

    https://doi.org/10.1038/nature09746

    • PubMed
    • Google Scholar
83. 1. Weiss HM
    2. Grisshammer R

    (2002) Purification and characterization of the human adenosine A(2a) receptor functionally expressed in Escherichia coli

    European Journal of Biochemistry 269:82–92.

    https://doi.org/10.1046/j.0014-2956.2002.02618.x

    • PubMed
    • Google Scholar
84. 1. Xu F
    2. Wu H
    3. Katritch V
    4. Han GW
    5. Jacobson KA
    6. Gao ZG
    7. Cherezov V
    8. Stevens RC

    (2011) Structure of an agonist-bound human A2A adenosine receptor

    Science 332:322–327.

    https://doi.org/10.1126/science.1202793

    • PubMed
    • Google Scholar
85. 1. Xue Y
    2. Yuwen T
    3. Zhu F
    4. Skrynnikov NR

    (2014) Role of electrostatic interactions in binding of peptides and intrinsically disordered proteins to their folded targets. 1. NMR and MD characterization of the complex between the c-Crk N-SH3 domain and the peptide Sos

    Biochemistry 53:6473–6495.

    https://doi.org/10.1021/bi500904f

    • PubMed
    • Google Scholar
86. 1. Ye L
    2. Van Eps N
    3. Zimmer M
    4. Ernst OP
    5. Prosser RS

    (2016) Activation of the A2A adenosine G-protein-coupled receptor by conformational selection

    Nature 533:265–268.

    https://doi.org/10.1038/nature17668

    • PubMed
    • Google Scholar
87. 1. Yesselman JD
    2. Price DJ
    3. Knight JL
    4. Brooks CL

    (2012) MATCH: an atom-typing toolset for molecular mechanics force fields

    Journal of Computational Chemistry 33:189–202.

    https://doi.org/10.1002/jcc.21963

    • PubMed
    • Google Scholar
88. 1. Yurugi-Kobayashi T
    2. Asada H
    3. Shiroishi M
    4. Shimamura T
    5. Funamoto S
    6. Katsuta N
    7. Ito K
    8. Sugawara T
    9. Tokuda N
    10. Tsujimoto H
    11. Murata T
    12. Nomura N
    13. Haga K
    14. Haga T
    15. Iwata S
    16. Kobayashi T

    (2009) Comparison of functional non-glycosylated GPCRs expression in Pichia pastoris

    Biochemical and Biophysical Research Communications 380:271–276.

    https://doi.org/10.1016/j.bbrc.2009.01.053

    • PubMed
    • Google Scholar

Author details

1. Lindsay D Clark

   1. Department of Biophysics, The University of Texas Southwestern Medical Center, Dallas, United States
   2. Molecular Biophysics Graduate Program, The University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Conceptualization, Investigation, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Igor Dikiy

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-6879-6883

2. Igor Dikiy

   Structural Biology Initiative, CUNY Advanced Science Research Center, New York, United States

   Contribution

   Conceptualization, Investigation, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Lindsay D Clark

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8155-7788

3. Karen Chapman

   Department of Biophysics, The University of Texas Southwestern Medical Center, Dallas, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

4. Karin EJ Rödström

   Department of Neuroscience and Pharmacology, University of Copenhagen, Copenhagen, Denmark

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

5. James Aramini

   Structural Biology Initiative, CUNY Advanced Science Research Center, New York, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

6. Michael V LeVine

   1. Department of Physiology and Biophysics, Weill Cornell Medical College, New York, United States
   2. Institute for Computational Bioscience, Weill Cornell Medical College, New York, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

7. George Khelashvili

   1. Department of Physiology and Biophysics, Weill Cornell Medical College, New York, United States
   2. Institute for Computational Bioscience, Weill Cornell Medical College, New York, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-7235-8579

8. Søren GF Rasmussen

   Department of Neuroscience and Pharmacology, University of Copenhagen, Copenhagen, Denmark

   Contribution

   Supervision, Investigation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

9. Kevin H Gardner

   1. Structural Biology Initiative, CUNY Advanced Science Research Center, New York, United States
   2. Department of Chemistry and Biochemistry, City College of New York, New York, United States
   3. Biochemistry, Chemistry and Biology PhD Programs, Graduate Center, City University of New York, New York, United States

   Contribution

   Conceptualization, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   Kevin.Gardner@asrc.cuny.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8671-2556

10. Daniel M Rosenbaum

    1. Department of Biophysics, The University of Texas Southwestern Medical Center, Dallas, United States
    2. Molecular Biophysics Graduate Program, The University of Texas Southwestern Medical Center, Dallas, United States

    Contribution

    Conceptualization, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing

    For correspondence

    dan.rosenbaum@utsouthwestern.edu

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-7174-4993

• 3,791

  views

• 755

  downloads

• 74

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.