Conformational fingerprinting of allosteric modulators in metabotropic glutamate receptor 2

  1. Brandon Wey-Hung Liauw
  2. Arash Foroutan
  3. Michael R Schamber
  4. Weifeng Lu
  5. Hamid Samareh Afsari
  6. Reza Vafabakhsh

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    The authors advance our understanding of the molecular underpinnings of allostery in GPCRs by showing the effects of allosteric modulators of mGluR2 on receptor conformation at distinct sites in the presence and absence of orthosteric modulators. This is important as drugs and drug candidates acting outside the site where the orthosteric or endogenous ligands bind are harder to identify. This work provides insights into allosteric changes at the level of individual receptors and provides a new path for drug discovery and is this of interest to colleagues studying GPCRs in health and disease.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1, Reviewer 2 and Reviewer #3 agreed to share their name with the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Activation of G protein-coupled receptors (GPCRs) is an allosteric process. It involves conformational coupling between the orthosteric ligand binding site and the G protein binding site. Factors that bind at non-cognate ligand binding sites to alter the allosteric activation process are classified as allosteric modulators and represent a promising class of therapeutics with distinct modes of binding and action. For many receptors, how modulation of signaling is represented at the structural level is unclear. Here, we developed fluorescence resonance energy transfer (FRET) sensors to quantify receptor modulation at each of the three structural domains of metabotropic glutamate receptor 2 (mGluR2). We identified the conformational fingerprint for several allosteric modulators in live cells. This approach enabled us to derive a receptor-centric representation of allosteric modulation and to correlate structural modulation to the standard signaling modulation metrics. Single-molecule FRET analysis revealed that a NAM (egative allosteric modulator) increases the occupancy of one of the intermediate states while a positive allosteric modulator increases the occupancy of the active state. Moreover, we found that the effect of allosteric modulators on the receptor dynamics is complex and depend on the orthosteric ligand. Collectively, our findings provide a structural mechanism of allosteric modulation in mGluR2 and suggest possible strategies for design of future modulators.

Article activity feed

  1. Version published to 10.7554/elife.78982 on eLife
  2. eLife

    Author Response

    Reviewer #2 (Public Review):

    -Were there any post-translational modifications (phosphorylation etc) or endogenous lipids that need to be quantified to make sense of the data?

    A percentage of receptors could be phosphorylated; therefore, our results represent the average behavior of the population. This is a noteworthy point and we have now explicitly discussed this idea in the revised the manuscript.

    In the in vivo experiments, heterogeneity in PTMs or local lipid environment of receptors could affect conformational change at the individual receptor level. For our analysis we integrate the intensities over the whole cell membrane, so the results represent the average behavior. Likewise, in the single-molecule FRET experiments many individual receptors are included in the analysis. Additionally, since the receptors are purified in the in vitro experiments, there is no further change in PTMs with application of drugs. We have added a sentence in the discussion to highlight the potential heterogeneity in PTMs and local lipid environment. We have also added a sentence to the methods to clarify how in vivo experiments are analyzed.

    Added to line 512 in discussion section: “Potential sources of heterogeneity arising from differences in post-translational modifications or differences in the local lipid environment, may affect receptor conformation. Therefore, our results represent the average of a heterogeneous population of such receptors.”

    Changed line 667 to: “ROIs used for analysis included the whole cell membrane for individual cells.”

    -mGLUR2 is a dimer. I was expecting that at 15 uM of Glutamate, for example, one might see effects of a single protomer-bound receptor. If I'm not mistaken, some class C receptors don't activate their CRDs until both ligand binding sites in the VFT are bound. Looking at all of the profiles in the VFT, CRD, and 7TM, I don't see any evidence of the 2-site binding of glutamate at the VFT. Presumably, there are Hill slopes for all of these profiles?

    Based on our previous work with the wildtype and with the receptor containing one glutamate binding deficient monomer, and available structures, indeed CRD domains do not significantly visit the active state unless both VFT domains are bound to glutamate and in the closed conformation. However, because activations involve progression through 2 intermediate states, we still expect to see FRET change even when both VFT domains are not occupied simultaneously. We have now revised Table 1 to included Hill slope. This data shows that cooperativity is generally observed for the FRET sensors for all the ligands tested.

    Reviewer #3 (Public Review):

    -The main concerns I had were with respect to labelling stoichiometry of the mixed Cy3/Cy5 compounds or SNAP-tag labels. How was this controlled? Clearly, both label cells, as shown in supplemental data and the single molecule FRET data support that both sites are labelled. Are there any concerns about larger molecular complexes such as oligomers that may confound the simple interpretation of interactions between the dimers?

    Among class C GPCRs, only GABA receptors have been shown to be able to potentially form efficient oligomers. Subunit counting experiments have shown that mGluR2 is predominantly dimer (> 90%) on the plasma membrane for the experimental conditions used in this manuscript (Levitz et al., 2016). The same result was obtained from live-cell FRET utilizing a dimer trafficking-control system (Maurel et al., 2008). This work also demonstrated that FRET occurred strictly for dimeric receptors labeled by both donor and acceptor fluorophores and not between neighboring receptors at the plasma membrane. Thus, receptors labeled with donor-only or acceptor-only do not contribute to the relative ΔFRET signal in response to treatment.

    -Some additional context might be a discussion of approaches used and results obtained for other types of conformational biosensors for GPCRs in other classes? Can we learn anything by comparison?

    We have revised the manuscript to include further discussion of results obtained from the use of other conformational sensors.

    Added to line 502: “Recent experiments have shown that GPCRs are dynamic (Nygaard et al., 2013) and undergo transition between multiple conformational states, including multiple intermediate states. For class A GPCRs, studies using conformational biosensors based on nuclear magnetic resonance (NMR) spectroscopy (Huang et al., 2021), double electron-electron resonance (DEER) spectroscopy (Wingler et al., 2019), smFRET (Gregorio et al., 2017), and fluorescent enhancement (Wei et al., 2022) have revealed the importance of conformational dynamics for receptor activation, ligand efficacy, and biased signaling.”

    Added to line 536: “Interestingly, the regulation of intermediate state occupancy has recently been shown to be a mechanism of allosteric modulation for other classes of GPCRs as well. NMR studies on the μ-opioid receptor (Kaneko et al., 2022) and cannabinoid receptor 1 (Wang et al., 2021) revealed that PAMs and NAMs regulate receptor function by acting on intermediate conformations in a manner similar to our findings for BINA and MNI-137. Collectively, these results suggest that designing compounds that regulate intermediate state occupancy is a plausible strategy for the development of allosteric modulators for mGluR2 and other families of GPCRs.”

  3. eLife

    Evaluation Summary:

    The authors advance our understanding of the molecular underpinnings of allostery in GPCRs by showing the effects of allosteric modulators of mGluR2 on receptor conformation at distinct sites in the presence and absence of orthosteric modulators. This is important as drugs and drug candidates acting outside the site where the orthosteric or endogenous ligands bind are harder to identify. This work provides insights into allosteric changes at the level of individual receptors and provides a new path for drug discovery and is this of interest to colleagues studying GPCRs in health and disease.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1, Reviewer 2 and Reviewer #3 agreed to share their name with the authors.)

  4. eLife

    Reviewer #1 (Public Review):

    Liauw et al. examine the conformational effects of positive and negative allosteric modulators (PAMs and NAMs) at mGluR2 GPCRs using FRET pairs in each of three distinct structural domains of the dimer. They show that modulators affect the conformation of all of the domains, albeit in distinct ways. Using single-molecule smFRET they show that the NAM MNI-137 blocks receptor function not by eliminating glutamate-dependent domain motions in general, but specifically by hindering transitions to the active state which traps receptors in intermediate pre-active states. These results shed new insight into the mechanism of class C GPCR allosteric modulators, and specifically MNI-137, highlighting the role of intermediate conformations that may be highly relevant for new therapeutic approaches to human health. Beyond GPCRs, this work will also be of broad interest to the study of allosteric modulation of proteins in general.

  5. eLife

    Reviewer #2 (Public Review):

    This is an exciting advance in our understanding of allostery in class C GPCRs. One could easily envisage a scenario where the FRET signal is dominated by a frequently populated activation intermediate that doesn't correlate with signaling response for example, at least where multiple states constitute the ensemble. Yet, interestingly this doesn't seem to be the case here. The TIRF-based "single molecule" state analysis also helps with our understanding of the modulatory effects of the NAMs and PAMs as discussed by the authors.

    A few issues the authors might wish to consider:

    1. Were there any post-translational modifications (phosphorylation etc) or endogenous lipids that need to be quantified to make sense of the data?
    2. mGLUR2 is a dimer. I was expecting that at 15 uM of Glutamate, for example, one might see effects of a single protomer-bound receptor. If I'm not mistaken, some class C receptors don't activate their CRDs until both ligand binding sites in the VFT are bound. Looking at all of the profiles in the VFT, CRD, and 7TM, I don't see any evidence of the 2-site binding of glutamate at the VFT. Presumably, there are Hill slopes for all of these profiles?
  6. eLife

    Reviewer #3 (Public Review):

    The authors used a combination of site-specific labelling at distinct sites within the mGluR2- the VFT domain in the ligand binding site, ECL2 (newly developed here), and the cysteine-rich domain (CRD) the latter of which is located between the VFT and ECL2. Using live cell FRET based on SNAP-tagged or unnatural amino acids, site-labeled with Cy3 or Cy5 tags, they validate that orthosteric ligands generate FRET changes consistent with their know efficacies and potencies, validating them for use in studying the effects of allosteric modulators. They next use single-molecule FRET to study the effects of the allosteric modulators on the receptor in the presence or absence of the orthosteric ligand, glutamate.

    Major strengths include the careful design, conduct, and analysis of the experiments and the validation of the effects of orthosteric ligands alone before proceeding to measurements of allosteric effects. They produce some very interesting results with the allosteric modulators in both experimental formats - the whole cell FRET consistent with known allosteric effects and the single molecule FRET identifying some independent effects of the allosteric modulators - this was quite striking. The approach is scalable to other GPCRs and to other membrane proteins in general.

    The main concerns I had were with respect to labelling stoichiometry of the mixed Cy3/Cy5 compounds or SNAP-tag labels. How was this controlled? Clearly, both label cells, as shown in supplemental data and the single molecule FRET data support that both sites are labelled. Are there any concerns about larger molecular complexes such as oligomers that may confound the simple interpretation of interactions between the dimers?

    Some additional context might be a discussion of approaches used and results obtained for other types of conformational biosensors for GPCRs in other classes? Can we learn anything by comparison?

  7. Version published to 10.1101/2022.04.27.489706v1 on bioRxiv