Single-Molecule Imaging and Computational Microscopy Approaches Clarify the Mechanism of the Dimerization and Membrane Interactions of Green Fluorescent Protein

Abstract

Green fluorescent protein (GFP) is widely used as a biomarker in living systems; however, GFP and its variants are prone to forming low-affinity dimers under physiological conditions. This undesirable tendency is exacerbated when fluorescent proteins (FP) are confined to membranes, fused to naturally-oligomeric proteins, or expressed at high levels in cells. Oligomerization of FPs introduces artifacts into the measurement of subunit stoichiometry, as well as interactions between proteins fused to FPs. Introduction of a single mutation, A206K, has been shown to disrupt hydrophobic interactions in the region responsible for GFP dimerization, thereby contributing to its monomerization. Nevertheless, a detailed understanding of how this single amino acid-dependent inhibition of dimerization in GFP occurs at the atomic level is still lacking. Single-molecule experiments combined with computational microscopy (atomistic molecular dynamics) revealed that the amino group of A206 contributes to GFP dimer formation via a multivalent electrostatic interaction. We further showed that myristoyl modification is an efficient mechanism to promote membrane attachment of GFP. Molecular dynamics-based site-directed mutagenesis has been used to identify the key functional residues in FPs. The data presented here have been utilized as a monomeric control in downstream single-molecule studies, facilitating more accurate stoichiometry quantification of functional protein complexes in living cells.

Effects of small molecule-induced dimerization on the programmed death ligand 1 protein life cycle

“…Long-term effects of PD-L1 dimerization also did not manifest as disappearance of PD-L1 from the plasma membrane, as the protein was still detectable in PD-L1-positive MDA-MB-231 GFP cells after 24 h and 48 h of exposure to GS-4224. Similar to the observations reported by Snapp et al 54 , using EYFP-fused PD-L1 56 resulted in formation of ER whorls or Organized smooth ER (OSER) in our experiments with dimerizing compounds (Suppl. Fig.…”

Influence of nanobody binding on fluorescence emission, mobility, and organization of GFP-tagged proteins

“…An explanation for these observations could be that the (E)GFP tagged proteins appear to a certain extend in aggregates or homo-oligomers that are (partially) disassembled after Nb binding, leading to an average increase in fluorescent particle number ( Figures 2 and 3 ) and mobility ( Figures 3 and 4 ) without majorly affecting the fluorescence lifetime ( Figure S3 ). GPI-EGFP dimers or GFP oligomers have been reported previously ( Aronson et al., 2011 ; Huang et al., 2015 ; Jain et al., 2001 ; Suzuki et al., 2012a ) and could be mediated by the FP tag itself ( Beutel et al., 2015 ; Zacharias et al., 2002 ; Wang et al., 2019 ). However, since aggregates should in principle be brighter, one would in this case expect a decrease in molecular fluorescence brightness (cpm) upon aggregate disassembly.…”

“…Here, we only showed the effect of one specific anti-GFP nanobody (GFP-booster from Chromotek) and we also acknowledge that we only tested one GFP and one EGFP variant expressed on the surface of the living cells (see Transparent methods for protein sequences). Many derivatives of GFP exist that counteract FP-mediated oligomerization especially when passing through oxidative environments ( Aronson et al., 2011 ; Zacharias et al., 2002 ; Suzuki et al., 2012b ; Wang et al., 2019 ). An interesting approach could be to add a short tag to the FP to allow detection with another nanobody ( Braun et al., 2016 ).…”

Plant Sterol Clustering Correlates with Membrane Microdomains as Revealed by Optical and Computational Microscopy

“…However, the detailed understanding of how lipid composition and mobility changes affect the key lipid–lipid interactions and structural changes responsible for the formation of phase separation and microdomain formation still remain as formidable challenges. Despite extensive experimental studies in lipid aggregation and phase behavior, molecular dynamics (MD) simulations at the atomic level provide the highest temporal and spatial resolution of this process, with the potential to capture more detailed and key steps during the formation of microdomains and phase separation [ 38 , 66 , 67 ]. Several well-known MD simulation packages can conduct dynamics simulations on a single GPU and multiple GPUs with performances that outstrip even the most powerful conventional CPU-based supercomputers [ 68 , 69 , 70 ].…”

“…One way to study the dynamic behavior of these molecules at near-atomic resolution is through the use of molecular dynamics (MD) simulations, which provide reasonable and reliable results compared to experimental data [ 33 , 34 , 35 , 36 ]. By using such simulations as a “computational microscope” [ 37 , 38 ], we are able to obtain information on their single molecule dynamics, interactions with other lipids and with membrane proteins, thus characterizing emergent properties and aiding the interpretation of experimental data. Sanchez et al demonstrated that MβCD preferentially removes cholesterol from liquid-disordered (Ld) instead of liquid-ordered (Lo) membrane domains [ 39 ].…”