Fig 2: E3 ubiquitin ligases for MAVS degradation are screened and identified in aMPV/C-infected Vero cells. (A) Vero cells were transfected with 40 pmol MARCH5-siRNA (MARCH5), RNF5-siRNA (RNF5), and scrambled-siRNA (Scra) and were detected at 48 h by Western blotting with anti-MARCH5, -RNF5, or β-actin antibody. (B) Vero cells were transfected with 40 pmol MARCH5-siRNA (MARCH5), RNF5-siRNA (RNF5), or scrambled-siRNA (Scra) and then were infected with aMPV/C for 48 h. Cells were harvested and analyzed using Western blotting with anti-MAVS, anti-N, or anti-β-actin antibody. (C) Representative results are displayed with graphs corresponding to the ratios of MAVS: β-actin normalized to the control conditions. (D) 40 pmol RNF5-siRNA (RNF5), or scrambled-siRNA (Scra) transfected Vero cells were co-transfected with Flag, Flag-MAVS3, or Flag-MAVS and HA-Ub. The cells were infected with aMPV/C and analyzed as described in A. Error bars represent mean ± SD of three independent experiments. ns (not significant) p > 0.05; *** p < 0.001, compared with the Scra group.

Fig 3: MARCH5-mediated ubiquitination is altered by the addition of lipids in vitro.(A, B) Coomassie-stained SDS–PAGE gel and (B) Western blot targeting ubiquitin (top row) and MARCH5 (bottom row) of ubiquitination assay of MARCH5 in the presence of lipids (PE, PG, PI(4)P, PC, PA, or CL) or in the absence of lipids (+ control) in a molar ratio of 1:10 protein to lipid. − control included all reaction components except MARCH5. The ubiquitination reactions were terminated at different time points (0, 15, 30, and 60 min). (C, D) Relative band intensities of MARCH5 at mono-ubiquitin at 10 kD (C) and MARCH5 at 25 kD (D) of Coomassie-stained SDS–PAGE gels. Time 0 min of each reaction solution represents 100% and reduction of either MARCH5 or mono-ubiquitin is represented relative to the respective time 0. Data information: In (A, B), experiments were performed as three independent experiments and (C, D) data are represented as mean values ± SEM of these three independent experiments.Source data are available for this figure.

Fig 4: Ubiquitination, de-ubiquitination assay, and crosslinking of MARCH5.(A, B) Coomassie-stained SDS–PAGE gel and (B) Western blot targeting MARCH5 (left) and ubiquitin (right) of ubiquitination and de-ubiquitination assay. U: ubiquitination assay of MARCH5 in presence or absence (+control) of lipids after 120 min, D: deubiquitination of formed ubiquitin chains by MARCH5 in the presence or absence of lipids after 120 min. (C) Crosslinking experiment of MARCH5 in absence of lipids (control) and in presence of CL in a 1:10 M ratio of protein to lipid. Samples were taken after 0, 15, 30, and 60 min of incubation with the crosslinker.

Fig 5: Cardiolipin induces MARCH5 auto-ubiquitination.(A, B) Coomassie-stained SDS–PAGE and (B) Western blot visualised by anti-ubiquitin antibodies (top row) or anti-MARCH5 antibodies (bottom row) of samples taken from the MARCH5 ubiquitination assay in presence of CL at 1:1, 1:5, 1:10, and 1:50 M ratio of protein to lipid or in the absence of lipids (+ control). − control included all reaction components except MARCH5. The ubiquitination reaction was terminated at different time points (0, 15, 30, 60, and 120 min). (C, D) Relative band intensities of MARCH5 at 25 kD (C) and mono-ubiquitin at 10 kD (D) of Coomassie-stained of SDS–PAGE gels. Time 0 min of each reaction solution represents 100%, and reduction of either MARCH5 or mono-ubiquitin is represented relative to their respective band at time 0. Data information: In (A, B), experiments were performed as three independent experiments and (C, D) data were represented as mean values ± SEM of these three independent experiments.Source data are available for this figure.