UC Santa Barbara

Oligomerization of G protein-coupled receptors (GPCRs) is a widespread phenomenon whose discovery generates a plethora of alternative targets for new therapeutic approaches towards human diseases. Nevertheless, challenges still exist in the characterization of these complexes, especially in terms of driving factors of formation, interfaces, and functional consequences. Despite their significance, structural and functional studies of GPCR oligomers have been hindered by their dynamic nature and their generally low suitability for biophysical techniques.

Among these receptors, the human adenosine A2A receptor (A2AR) serves as an excellent target to conduct modeling studies into oligomerization of GPCRs, as there is solid evidence that this receptor forms homo- and hetero-oligomers both in vitro and in vivo. Its intrinsically disordered C-terminus is removed in all structural studies of A2AR for stability and homogeneity purposes, but a C-terminal mutation has been shown to prevent A2AR oligomer formation. We first aimed to understand the role of the C-terminus in driving the oligomerization of A2AR. Size-exclusion chromatography (SEC) was applied as the primary method to quantify the oligomer levels of multiple variants of A2AR with strategic mutations and truncations on the C-terminus. We discovered that the C-terminus of A2AR drives receptor homo-oligomerization via multiple types of covalent and non-covalent interactions. Computational analysis revealed that A2AR dimers are formed via multiple interfaces, all involving the C-terminus. Variation of ionic strength of the buffer indicated depletion interactions via the C-terminus to be the main driving force of A2AR oligomerization. Experiments on the C-terminus sans the transmembrane (TM) helices demonstrated that A2AR C-terminus in and of itself can form insoluble aggregate at high salt concentrations.

The inclusion of the C-terminus enables the production and isolation of A2AR oligomers, yet also further complicates biophysical and structural studies of this receptor. Electron paramagnetic resonance (EPR) spectroscopy offers unique capability of probing how the dynamic C-terminus is involved at the multiple interfaces of A2AR oligomers. Much effort had been made using cell sorting in engineering properly folded A2AR variants void of free cysteines to facilitate biophysical characterization by EPR, but the structure and function of these variants needed to be thoroughly investigated. We discovered that these A2AR mutants, selected with an agonist-based assay, showed reduced binding activity to antagonist. Further characterization with EPR power saturation experiments demonstrated that various extracellular disulfide bonds were disrupted in these variants, suggesting that the removed transmembrane cysteines may serve a role in maintaining the proper structure and function of the receptor.

Moving forward, we next sought to visualize the oligomeric interfaces of A2AR and the structural role of the intrinsically disordered C-terminus by combining continuous-wave (CW) EPR with cryogenic electron microscopy (cryo-EM). CW-EPR revealed that a C-terminal residue was immobilized as A2AR formed oligomers, suggesting that the C-terminus is directly involved at the oligomeric interface of the receptor. The related technique double electron electron resonance (DEER) revealed a large intermolecular distance between two C-terminal cysteines, suggesting that A2AR oligomers are not stabilized by direct disulfide bonds between the C-termini. Early cryo-EM data collection yielded a low-resolution 3D structure of A2AR dimers that showed the involvement of the TM regions at the interfaces. Therefore, it appears that both the C-terminus and the TM helices of A2AR contribute to forming the oligomeric interfaces of the receptor.

Finally, in search of a membrane mimetic platform that can retain the native structure and function of A2AR oligomers, styrene maleic acid (SMA) lipid polymers were employed as a promising detergent-free method to isolate transmembrane proteins. We sought to assess the functional impact of extracting directly from the native host environment A2AR and proteorhodopsin (PR), a model bacterial transmembrane proton pump. We discovered that SMA-solubilized A2AR exhibited reduced binding activity to antagonist, likely due to the lack of functional cholesterol. For PR, SMA could only capture the monomeric form of the receptor and could not solubilize the functionally important hexameric form. Further analyses demonstrated that solubilizing PR with SMA severely reduced its active population and disrupt its photocycle properties. Taken together, despite retention of the native host membranes, SMA appeared to have negative impacts on the functional properties of both A2AR and PR.