Final Report Summary - OLIGABA (GPCR oligomers: Facts and function for the GABAB receptor)
G protein-coupled receptors (GPCRs) are a superfamily of proteins encoded by more than 850 genes in the human genome that play a pivotal role in cell communication. It is therefore not surprising that they also constitute the main target of all drugs currently on the market. GPCRs have been shown to assemble together to form homo- and heterocomplexes, resulting in functional cross-talks and trans-modulations between the receptors within a complex. Although this phenomenon is well-characterized in vitro (transiently transfected cell lines), GPCR oligomers remain poorly characterized in native tissues especially in neurons. Among the large GPCR family, the class C GPCRs and in particular the metabotropic glutamate (mGlu) receptors and the GABAB receptor constitute ideal models to study the role of GPCR oligo¬merization. The mGlu receptors are covalently linked homodimers and the GABAB receptor is an obligatory heterodimer composed of the GABAB1 and GABAB2 subunits that has been shown to assemble into larger complexes in heterologous expression systems. Notably, both the mGlu receptors and the GABAB receptor are promising drug targets for several neurological disorders and GPCR oligomers in general are a potential new class of drug targets.
The overall goal of the OLIGABA project (grant agreement number 627227) was to identify and characterize oligomers of class C GPCRs in native tissues using state of the art microscopy techniques.
A first technique called two-photon scanning number and brightness (sN&B) analysis records successive images of a neuron expressing either fluorescently labeled mGlu2 or GABAB receptor over time. Using the fluorescence fluctuations of each pixel of the image, spatial maps of the receptor stoichiometry and the number of receptor complexes are calculated. We transfected hippocampal neurons with tagged receptors and labeled the receptors at the cell surface with a fluorophore. Our results show that the GABAB receptor assembled into large complexes (2 to 8 subunits) already at low density and that the size of the complexes increased with the expression level. This observation is in agreement with previous studies that used non-neuronal cell lines. On the contrary, for the mGlu2 receptor we found that it was primarily a strict dimer in the endogenous expression range, which is consistent with what has been observed in cell lines. Only when the density was very high, larger complexes were observed.
When an agonist and a positive allosteric modulator (PAM) were added to the mGlu2 receptor we saw an increase in the stoichiometry, indicating that complexes larger than dimers were formed. To our surprise, we observed similar results when an antagonist was added to the receptor. These interesting findings led us to focus on elucidating the mechanism behind this ligand-induced oligomerization. We tested first whether intracellular proteins could be involved. Since most intracellular proteins that interact with GPCRs bind to their cytosolic C-terminal domains we truncated the distal part of the mGlu2 C-terminal domain. This did not change the oligomerization propensity of the receptor in absence or presence of ligand, indicating that proteins interacting with the distal part of the mGlu2 C-terminal domain are not involved in the higher order oligomerization of the mGlu2 receptor. We then tested the hypothesis that the ligand-induced oligomerization could be driven by a reduced conformational dynamics. Indeed, mGlu receptors are in equilibrium between various conformations in the basal state. Ligands, such as agonist and antagonists, are known to displace the basal equilibrium and to stabilize a given conformation. To address whether the effects of the ligands on the mGlu2 receptor oligomerization were due to the stabilization of a conformation, we introduced mutations that locked the receptor in the inactive conformation. These mutations reduce the conformational dynamics in a ligand independent way. This mutant mGlu2 receptor locked in the inactive conformation had a stronger tendency to oligomerize than the wild type receptor in the absence of ligand. This indicates that modulation of the conformational dynamics is indeed the mechanism behind ligand-induced oligomerization of the mGlu2 receptor.
We then used a second microscopy technique, time-resolved Förster resonance energy transfer (TR-FRET) microscopy to confirm our sN&B measurements. This technique allows the detection of two fluorescent molecules that are in close proximity. Therefore it could be used to confirm the interaction of two fluorescently labeled mGlu2 dimers expressed in neurons. In the absence of ligand, we could not detect any interaction. In the presence of agonist and PAM and addition of specific mutations to help stabilizing the oligomers, a positive signal could be detected between mGlu2 dimers. This suggests that the mGlu2 receptor does form oligomers when bound to a ligand, which supports our sN&B findings.
In summary we have found that in primary hippocampal neurons in absence of ligand the GABAB receptor has a strong tendency to oligomerize, whereas the mGlu2 receptor is mainly dimeric in the physiologically relevant expression range. However upon addition of activating or inactivating ligands the mGlu2 receptor forms complexes larger than dimers. The mechanism behind this surprising result is most likely a reduction in the conformational dynamics upon ligand binding, which favors oligomerization. One hypothesis regarding the functional role of the oligomerization is that a dimer stabilized in one conformation can favor that a second dimer adopts the same conformation by interacting with it. As a consequence, more receptors will be either active or inactive. This putative phenomenon could be referred to as allostery.
Our observations suggest a new concept for GPCR oligomerization where the conformational dynamics plays a crucial role in determining the oligomeric state of the receptors. This is another piece in our understanding of this fascinating class of receptors that is of fundamental importance to our physiology. The techniques that we have developed could easily be transposed to other GPCRs. Our study also suggests that drugs that can modulate the conformational dynamics of GPCRs could regulate their oligomerization propensity. Such drugs could be a novel type of biased ligands with the potential to be developed into new drugs with fewer side effects. This would benefit the health of patients and benefit the universities or pharmaceutical companies developing the drugs economically.
The overall goal of the OLIGABA project (grant agreement number 627227) was to identify and characterize oligomers of class C GPCRs in native tissues using state of the art microscopy techniques.
A first technique called two-photon scanning number and brightness (sN&B) analysis records successive images of a neuron expressing either fluorescently labeled mGlu2 or GABAB receptor over time. Using the fluorescence fluctuations of each pixel of the image, spatial maps of the receptor stoichiometry and the number of receptor complexes are calculated. We transfected hippocampal neurons with tagged receptors and labeled the receptors at the cell surface with a fluorophore. Our results show that the GABAB receptor assembled into large complexes (2 to 8 subunits) already at low density and that the size of the complexes increased with the expression level. This observation is in agreement with previous studies that used non-neuronal cell lines. On the contrary, for the mGlu2 receptor we found that it was primarily a strict dimer in the endogenous expression range, which is consistent with what has been observed in cell lines. Only when the density was very high, larger complexes were observed.
When an agonist and a positive allosteric modulator (PAM) were added to the mGlu2 receptor we saw an increase in the stoichiometry, indicating that complexes larger than dimers were formed. To our surprise, we observed similar results when an antagonist was added to the receptor. These interesting findings led us to focus on elucidating the mechanism behind this ligand-induced oligomerization. We tested first whether intracellular proteins could be involved. Since most intracellular proteins that interact with GPCRs bind to their cytosolic C-terminal domains we truncated the distal part of the mGlu2 C-terminal domain. This did not change the oligomerization propensity of the receptor in absence or presence of ligand, indicating that proteins interacting with the distal part of the mGlu2 C-terminal domain are not involved in the higher order oligomerization of the mGlu2 receptor. We then tested the hypothesis that the ligand-induced oligomerization could be driven by a reduced conformational dynamics. Indeed, mGlu receptors are in equilibrium between various conformations in the basal state. Ligands, such as agonist and antagonists, are known to displace the basal equilibrium and to stabilize a given conformation. To address whether the effects of the ligands on the mGlu2 receptor oligomerization were due to the stabilization of a conformation, we introduced mutations that locked the receptor in the inactive conformation. These mutations reduce the conformational dynamics in a ligand independent way. This mutant mGlu2 receptor locked in the inactive conformation had a stronger tendency to oligomerize than the wild type receptor in the absence of ligand. This indicates that modulation of the conformational dynamics is indeed the mechanism behind ligand-induced oligomerization of the mGlu2 receptor.
We then used a second microscopy technique, time-resolved Förster resonance energy transfer (TR-FRET) microscopy to confirm our sN&B measurements. This technique allows the detection of two fluorescent molecules that are in close proximity. Therefore it could be used to confirm the interaction of two fluorescently labeled mGlu2 dimers expressed in neurons. In the absence of ligand, we could not detect any interaction. In the presence of agonist and PAM and addition of specific mutations to help stabilizing the oligomers, a positive signal could be detected between mGlu2 dimers. This suggests that the mGlu2 receptor does form oligomers when bound to a ligand, which supports our sN&B findings.
In summary we have found that in primary hippocampal neurons in absence of ligand the GABAB receptor has a strong tendency to oligomerize, whereas the mGlu2 receptor is mainly dimeric in the physiologically relevant expression range. However upon addition of activating or inactivating ligands the mGlu2 receptor forms complexes larger than dimers. The mechanism behind this surprising result is most likely a reduction in the conformational dynamics upon ligand binding, which favors oligomerization. One hypothesis regarding the functional role of the oligomerization is that a dimer stabilized in one conformation can favor that a second dimer adopts the same conformation by interacting with it. As a consequence, more receptors will be either active or inactive. This putative phenomenon could be referred to as allostery.
Our observations suggest a new concept for GPCR oligomerization where the conformational dynamics plays a crucial role in determining the oligomeric state of the receptors. This is another piece in our understanding of this fascinating class of receptors that is of fundamental importance to our physiology. The techniques that we have developed could easily be transposed to other GPCRs. Our study also suggests that drugs that can modulate the conformational dynamics of GPCRs could regulate their oligomerization propensity. Such drugs could be a novel type of biased ligands with the potential to be developed into new drugs with fewer side effects. This would benefit the health of patients and benefit the universities or pharmaceutical companies developing the drugs economically.