Fig. 3: LGP2 inhibits MDA5 translocation to induce microfilament assembly.

a AlphaFold predicted structural alignment of human MDA5ΔN and LGP2 proteins. b The turnover numbers (k_cat) of LGP2 ATPase using 11.6-kb dsRNA substrate (error bars: mean ± SE). c Representative images (left) and kymographs (right) showing the binding of 60 nM Cy3-LGP2 on dsRNA under various conditions. SYBR Gold stained 11.6-kb dsRNA is shown in red and Cy3-LGP2 is shown in green. Positions of dsRNA are shown adjacent to the right of kymographs. d Representative kymographs showing the formation of Cy3-MDA5 ATM clusters (10 nM) in the absence or presence of 10 nM LGP2. Positions of dsRNA are shown adjacent to the right of kymographs. Fluorescent intensities in kymographs were normalized to generate the heatmaps. White arrowheads indicate the transition of mobile ATM cluster into immobile one. e The frequency of mobile (blue) and immobile (grey) MDA5 ATM cluster in the absence/presence of LGP2 (mean ± SD; n = number of dsRNA molecules). f Representative EM images showing the formation of MDA5–LGP2 filaments with ATP. g A schematic illustration showing the conversion from an ATM cluster to a microfilament induced by LGP2 binding. h Representative kymographs (left) and single-particle trajectories (right) showing a Cy5-LGP2 (3 nM) binding to a Cy3-MDA5 motor (3 nM). Cy3-MDA5 is shown in green and Cy5-LGP2 is shown in red. The merged kymograph is generated by overlaying both channels. Arrowheads indicate the association of LGP2 with MDA5. Positions of dsRNA are shown adjacent to the right of kymographs. i A schematic illustration of LGP2 binding resulting in a stop in MDA5 translocation. j AlphaFold model of an MDA5ΔN–LGP2 complex on a 42-bp dsRNA substrate. The closeup shows the key contacts between MDA5ΔN and LGP2. k Proportion distribution of MDA5ΔN translocation (mobile) and LGP2-induced stop (immobile) events under various conditions (n = total number of events). The results of two-sided pairwise comparisons using Fisher’s exact test are shown (****P < 0.0001). l Histogram of binned MDA5ΔN translocation rates before and after LGP2 binding. Data were fit to Gaussian function to derive the average rates (mean ± SD; n = number of events). m Survival probability of MDA5ΔN and MDA5ΔN–LGP2 molecules on dsRNA (n = number of events examined). n Relative IFNβ mRNA levels showing the MDA5 signaling activities in ΔLGP2 cells after 12 h EMCV infection. IFNβ productions were induced upon EMCV infections. MOI: multiplicity of infection. o Immunoblotting showing the EMCV-induced phosphorylation of TBK1, IRF3 in ΔLGP2 cells. p-TBK1: phosphorylated TBK1, p-IRF3: phosphorylated IRF3. All single-molecule studies were performed at least two separate times. All cell-based assays were repeated independently at three times with similar results. Error bars represent SEM between measurements, centered on the mean of a single experiment.