Study of membrane protein interactions at single-molecule level with epidermal growth factor receptor and β-2 adrenergic receptor as representative models

Abstract

Membrane proteins interact with each other to send signal downstream, yet the interaction between two or more receptors are of transient nature, which hinders us from understanding the interaction to the full extent. While many different methods have been developed to analyze membrane protein interactions, such as Co-IP, yeast-two-hybrid, and FRET, none were sufficient or suitable methods for analyzing them, because they either disturb the original membrane environment or the status of the protein through genetic tagging or multiple labelling methods, which are enough to create artifacts in the analytical results. Therefore, in trying to solve these problems, many newer methods were introduced to the biological field, one of them being single particle tracking. Yet, single particle tracking alone cannot acquire enough information from a single cell, having to acquire different sets of results from many different cells, which creates ensemble level of understanding of membrane protein interactions. Ensemble level understanding of membrane protein interaction removes noise created, but also disregards heterogeneity of membrane protein interactions. Therefore, there was a call for a new method that can analyze membrane protein interactions in the most natural state possible. In this study, epidermal growth factor receptor (EGFR) and β-¬2 adrenergic receptor (β2AR), which are key receptors in receptor tyrosine kinases (RTKs) and G-protein coupled receptor (GPCR) respectively, were observed as model receptors using single-molecule imaging technique for analysis of membrane protein interaction. Epidermal growth factor receptor (EGFR) is known to have seven ligands which lead to formation of different types of dimers. However, it was never thoroughly observed in different environment with different interaction partners. In order to understand EGFR’s interaction pattern with different receptors, I applied the previously published method named single-molecule diffusional mobility shift assay (smDIMSA). smDIMSA were known to be able to observe ligand-receptor interaction through change in diffusion coefficient after ligand treatment, but I modified this method to observe receptor-receptor interaction. I validated this method using CD86 and CD28, which are constitutive monomer and dimer, respectively, and found that monomers cannot show shift in diffusion coefficient, indicating that the method is dimer specific. Then I went on to check if this can be applied for measuring the interaction between EGFR and ErbB2, and in a non-cancerous non-overexpressed condition of the cell, EGFR and ErbB2 only interacted with each other with the help of the ligand EGF. Lastly, we checked the interaction pattern of EGFR with other receptors that are known to interact with EGFR in different cancerous cell lines to find out that EGFR interacts with different receptors at different strength regardless of the EGFR expression level. Dimerization of GPCRs became general molecular signature, yet the exact mechanism or function of dimerization is not extensively studied. Although β2AR is a model receptor for GPCR because it is the first human GPCR to be unveiled of its structure, the function of dimerization is not known. This is because GPCRs have dynamic interactional pattern, and the transient and dynamic interactional nature hinders us from understanding its molecular mechanism. In order to measure exactly how much interaction occurs, I applied a previously published method that can quantify membrane protein interactions called co-immunoimmobilization (Co-II). Using Co-II, I have unveiled that β2AR forms homodimer during basal status, which is responsible for basal activity. Basal activity can only be disrupted with the inverse agonist ICI-118,551 treatment. ICI-118,551 treatment causes β2AR to go into inactive status, which prohibits the ligand from causing G-protein binding to the receptor. I also experimented on the measurement of interaction between β2AR and Gαs, because the traditional understanding of the GPCR activation mechanism explains GPCR and G-protein interacting with each other under basal condition. Interestingly, I found out that β2AR and Gαs do not interact with each other under basal condition, while dimeric β2AR and Gαs interact with each other. Lastly, I saw that β2AR cannot form homodimers without presence of cholesterol. Together, I have uncovered that β2AR dimerization is responsible for the basal activity which is subjective to inverse agonism or cholesterol.