New Techniques to Study Intracellular Receptors in Living Cells: Insights Into RIG-I-Like Receptor Signaling

Abstract

This review discusses new developments in Förster resonance energy transfer (FRET) microscopy and its application to cellular receptors. The method is based on the kinetic theory of FRET, which can be used to predict FRET not only in dimers, but also higher order oligomers of donor and acceptor fluorophores. Models based on such FRET predictions can be fit to observed FRET efficiency histograms (also called FRET spectrograms) and used to estimate intracellular binding constants, free energy values, and stoichiometries. These "FRET spectrometry" methods have been used to analyze oligomers formed by various receptors in cell signaling pathways, but until recently such studies were limited to receptors residing on the cell surface. To study complexes residing inside the cell, a technique called Quantitative Micro-Spectroscopic Imaging (Q-MSI) was developed. Q-MSI combines determination of quaternary structure from pixel-level apparent FRET spectrograms with the determination of both donor and acceptor concentrations at the organelle level. This is done by resolving and analyzing the spectrum of a third fluorescent marker, which does not participate in FRET. Q-MSI was first used to study the interaction of a class of cytoplasmic receptors that bind viral RNA and signal an antiviral response via complexes formed mainly on mitochondrial membranes. Q-MSI revealed previously unknown RNA mitochondrial receptor orientations, and the interaction between the viral RNA receptor called LGP2 with the RNA helicase encoded by the hepatitis virus. The biological importance of these new observations is discussed.

Keywords: ATPase; Antiviral response; FRET; Hepatitis C virus; Innate immunity; LGP2; MDA5; RIG-I; RNA helicase.

Fig. 1. Oligomer formation in RLR signaling.

Various RNA viruses (red, blue) activate antiviral genes after being detected through the RLRs (RIG-I, MDA5, & LGP2) which all converge on MAVS (grey). Ligand binding leads to a conformational change exposing CARDs (purple) that are ubiquitinated (green) and seed oligomers formed by RLRs and a mitochondrial antiviral signaling protein (MAVS)

Fig. 2. Quantitative Micro-Spectroscopic Imaging (Q-MSI).

Q-MSI is a variant of FRET spectrometry, in which pixel-level spectra are used to calculate the FRET efficiency (E_app) at each pixel and the concentrations of donor (D) and acceptor (A) in each region of interest (ROI). Spectral unmixing is used to determine the fluorescence of the donor in the presence of acceptor (F^DA) and the fluorescence of the acceptor in the presence of donor (F^AD), at two different excitation wavelengths (λ_ex,1 and λ_ex,2). These values are used to calculate donor and acceptor concentrations from previously prepared standard curves, and E_app from the equation shown, where Q^D and Q^A are the quantum yields of the donor and acceptor and ρ^D and ρ^A the ratios of fluorescence intensities observed upon excitation of the donor and acceptor at λ_ex,1 and λ_ex,2. The difference between Q-MSI and previous quantitative FRET techniques is that it combines the analysis of a third spectrum used as a marker for a sub-cellular compartment. In the above example, MitoTracker is used to identify mitochondria

Fig. 3. Pixel-level FRET determined from spectrally resolved GFP2-LGP2, Venus-MDA5, and Mitotracker.

Cells were co-transfected with plasmids expressing GFP₂-LGP2 and Venus-MDA5. The mitochondria were stained with MitoTracker-Red and FRET was calculated at a pixel level. (a) The spectrally-resolved 2D fluorescence intensity map for GFP₂-LGP2 (k^DA). (b) The 2D fluorescence intensity map for Venus-MDA5 (k^AD). (c) The 2D fluorescence intensity map for MitoTracker (k^M). (d) The FRET intensity distribution for cells co-expressing GFP₂-LGP2 and Venus-MDA5 (bars = 10 μm)

Fig. 4. Pixel level FRET analysis of the GFP₂-LGP2: Venus-MDA5 interaction.

(a–c) Selection and analysis of E_app in mitochondria regions of cells co-expressing GFP₂-LGP2 and Venus-MDA5 using MitoTracker. The selected mitochondrial ROIs (a) then used as a mask and applied to the E_app intensity map (b), and E_app values in each mitochondrial region were analyzed to identify the most common E_app value within that ROI, which were plotted on a meta-histogram (c) compiling all the peak E_app values across all mitochondrial ROIs selected. (d–f) The same set of images used for analysis in (a–c) were re-analyzed by selecting cytoplasmic regions using a consistent circle comprising 146 pixels on k^DA intensity maps such that most of the cell’s cytoplasm was selected but none of the circles were over-lapping (d). The mask created was applied to the E_app intensity map (e) to select random cytoplasmic regions of a consistent size and most common E_app values in each region was used to generate a meta-histograms comprising all peak E_app values selected (f). Meta histograms in (c) and (f) compare the results of cells analyzed in the presence (red) or absence (blue) of poly(I:C) RNA (bars = 10 μm)

Fig. 5. Models for LGP2: MDA5 oligomers.

(a) Proposed model for LGP2: MDA5 oligomer where LGP2 functions as an endcap or primer for the MDA5 filament. (b) Proposed model for the LGP2:MDA5 oligomer where LGP2 is the predominant protomer in the overall MDA5:LGP2 oligomer

Fig. 6. Model for HCV NS3 interacting with an RLR oligomer to facilitate the cleavage of MAVS.

(a) Helicase domains of RLRs scan the cytoplasm for viral PAMPs like duplex RNA. Upon PAMP binding, RLRs change conformation to expose CARDs, which bind to CARDs on MAVS (green) to form a signaling complex. This triggers a kinase cascade leading to phosphorylation of IRF-3 and transcription of pro-inflammatory cytokine and interferon genes. (b) The viral NS3 helicase (grey), which interacts with LGP2 (red), is covalently tethered to a protease (yellow), which cleaves MAVS and other key proteins needed to initiate the interferon response