Fig 1.
vIRF-1 binds directly to GABARAPL1 in an LIR-independent manner.
(A) Schematic structure of vIRF-1:PD, Proline-rich domain; DBD, DNA-binding domain; and IAD, IRF-association domain. LC3-interacting regions (LIRs) were searched for using a position-specific scoring matrix (PSSM) via the web-based software iLIR, and a conserved LIR of vIRF-1 with the highest PSSM score was predicted at the N-terminal junction of the IAD, as noted. (B) GST pull-down assay to assess in vitro binding between vIRF-1 and ATG8 proteins (LC3A, LC3B, LC3C, GABARAP, GABARAPL1, and GABARAPL2). Purified recombinant T7-tagged vIRF-1 (vIRF-1-T7) was pulled down with 1μg of GST-fused ATG8 proteins immobilized on glutathione beads. GST, GST-SUMO, and GST-Ub were used as controls. (C) Co-immunoprecipitation (co-IP) assays for assessment of the intracellular interactions between vIRF-1 and ATG8 proteins in virus-infected cells. iBCBL-1 cells were reactivated by treatment with 1 μg/ml doxycycline (Dox) for 2 days and fractionated into the nuclear and cytoplasmic fractions. The fractions were immunoprecipitated with normal rabbit IgG (nIgG) and rabbit anti-vIRF-1 antibody, and the immunoprecipitated complex was analyzed by immunoblotting with the indicated antibodies. Lamin B1 and lactate dehydrogenase (LDH) were used as nuclear and cytoplasmic fraction markers, respectively. (D) Immunofluorescence assay of the co-localization of vIRF-1 and GABARAPL1 in mitochondria (TOM20) in lytically virus-infected cells. iBCBL-1 cells were reactivated by treatment with Dox for 2 days and were fixed (Fix) with 4% paraformaldehyde before permeabilization (Per) with 0.5% Triton X-100 or after permeabilization with 25 μg/ml of saponin. A rat anti-vIRF-1 antibody was used. Scale bar, 10 μm. (E) Mutation of the core residues in the predicted LIR of vIRF-1: both tryptophan at position 255 (W255) and leucine at position 258 (L258) were replaced with alanine. (F-G) GST pull-down assays for assessment of in vitro interactions of wild-type (WT) and A255/A258 vIRF-1 with GST-LC3B and GST-GABARAPL1. GST alone was used as a control. The whole-cell lysates of 293T cells expressing Flag-tagged vIRF-1 (vIRF-1-Flag) were used as input for the pull-down assay. (G) The intensities of co-precipitated vIRF-1 proteins relative to input were determined, and the data represent the mean ± SD of three independent experiments. ‘ns’, not significant. (H) Co-IP assay. 293T cells were co-transfected with vIRF-1(WT and A255/A258)-Flag with V5-LC3B or V5-GABARAPL1 and immunoprecipitated with anti-Flag antibody.
Fig 2.
vIRF-1 binds to GABARAPL1 via a short sequence motif in its central region.
(A-C) (A) Diagram of recombinant GST-vIRF-1 (progressively deleted) and vIRF-1 DBD-T7 proteins used in the GST pull-down assays in (B) and (C), respectively. Red asterisks indicate the fusion protein bands with the expected molecular weights. (D) GST pull-down assays with differently deleted vIRF-1 DBD polypeptides fused to T7-intein (I)-chitin-binding domain (CBD). The T7-I-CBD-fused polypeptides were expressed in bacteria, and whole-cell lysates were used for the in vitro binding assays. ‘ns’ indicates nonspecific bands. Red asterisks indicate the full-length T7-I-CBD-fused polypeptides. (E) Generation of the vIRF-1 variants (Δ227–236 and Δ227–246), in which the regions of S227 to I236 and S227 to A246 were deleted. (F) Co-IP assay. 293T cells were co-transfected with V5-GABARAPL1 with vIRF-1-Flag (WT and the variants) or Flag-NIX for 24 h. (G) NanoBiT assay was conducted in 293T cells transfected with the indicated NanoBiT binary (LgB and SmB) fusion proteins. Data are presented as the mean ± SD of three independent experiments. The one-way ANOVA test was used to assess the statistical significance of differences between groups, and the t-test was used for post hoc pairwise comparisons. *, p < 0.05; **, p < 0.01; and ***, p < 0.001. Immunoblots of the cell extracts are shown below the graph.
Fig 3.
Identification of single residues of GABARAPL1 required for vIRF-1 binding.
(A-C) GST pull-down assays were performed with purified vIRF-1-T7 and the following GST-fused LC3B or GABARAPL1 variant proteins: (A) LC3B-GABARAPL1 domain-swap variants in which hexameric blocks of diverged residues of LC3B were replaced with colinear residues of GABARAPL1; (B) LC3B point variants in which each residue in the first and second blocks of LC3B was replaced with the counterpart residue of GABARAPL1; (C) GABARAPL1 variants in which residues glutamic acid at 17 (E17), lysine residues at 20 (K20) and 38 (K38) of GABARAPL1 were replaced with the counterpart residues aspartic acid [D], leucine [L], and glutamic acid [E] of LC3B.
Fig 4.
vIRF-1 binds to GABARAPL1 in a manner different from cellular autophagy receptors.
(A) The schematic structure of GABARAPL1 and its variants, VIRX, HPX, and G116A, are defective in binding to vIRF-1, LIR-containing proteins, and in phosphatidylethanolamine-lipidation, respectively. GIR, GABARAPL1-interacting region; VIR, vIRF-1-interacting region; HP, hydrophobic pockets. (B-C) GST pull-down assays with the GST-fused GABARAPL1 variants and purified recombinant vIRF-1-T7 (B) and p62/SQSTM1 (C). (D) Co-IP assays for assessment of the intracellular interactions of the Flag-tagged mitophagy proteins vIRF-1, p62/SQSTM1, and NIX with V5-GABARAPL1 (WT and variants). 293T cells were transfected with the indicated proteins for 24 h. A longer exposure time was necessary to detect input Flag-NIX.
Fig 5.
The interaction between vIRF-1 and GABARAPL1 is involved in regulating mitochondria content and HHV-8 productive replication.
(A) Agarose gel electrophoresis of the BspHI-digested DNA fragments of BAC16 and the mutant BAC16.vIRF-1ΔGIR (227–236) genomes. The DNA fragments (1,552 and 1,522 base pairs), including and deficient in the vIRF-1 GIR, are indicated with red and white asterisks, respectively. The BAC16 genome is drawn using NEBcutter V3.0. Sharp signs indicate methylation sites that may affect BspHI digestion. (B) DNA sequence verification of the deletion mutation of the vIRF-1 GIR sequences, as noted in the BAC16.vIRF-1ΔGIR genome. (C) Immunoblot analysis of extracts derived from iSLK.BAC16 and iSLK.BAC16.vIRF-1ΔGIR left untreated or treated with Dox and sodium butyrate (SB) for 3 days. MT-ND1 indicates mitochondrially-encoded NADH dehydrogenase 1. (D) RT-qPCR analysis of the encapsidated viral genome copy number in the media of the above iSLK cell cultures. Data represent the mean + SD of three independent experiments. *** < 0.001. (E) Immunoblot verification of CRISPR/Cas9-mediated knockout (KO) of GABARAPL1 in iSLK.BAC16 cells. Control cells were generated by transducing the parental lentiCRISPR v2 vector. (F) Determination of the HHV-8 genome copy number in the media of control and GABARAPL1 KO iSLK.BAC16 cells left untreated or treated with Dox and SB for 3 days. The corresponding cell extracts were immunoblotted with the indicated antibodies. The relative band intensities of MT-ND1 normalized to the loading control LDH are displayed beneath the corresponding panel. The one-way ANOVA test was used to assess the statistical significance of differences between groups, and the t-test was used for post hoc pairwise comparisons. *, p < 0.05 and **, p < 0.01.
Fig 6.
GABARAPL1 is required for vIRF-1/NIX-mediated mitophagy.
(A) Diagram of the mitophagy reporter, mito-mCE. The mCherry-EGFP tandem tag was fused to the mitochondrial targeting signal sequences (amino acids 1 to 33) of TOM20. A glycine-serine linker (L) was introduced between mCherry and EGFP. (B-C) The HeLa.Kyoto cell line (HeLa.Kyotomito-mCE) stably expressing mito-mCE was generated and then transiently transfected with empty vector (EV) or plasmids expressing vIRF-1, GABARAPL1, and NIX for 24 h in the presence of 40 μM leupeptin. (B) Confocal images of the transfected HeLa.Kyotomito-mCE cells. Scale bar, 10 μm. (C) Quantification of mitophagy. When the two microscopic images of EGFP and mCherry (see S2 Fig) were merged, the cells showing a stronger red fluorescence were regarded as positive for mitophagy. The red-to-green fluorescence ratio of more than 50 cells showing mitophagy from ten randomly collected images was determined using the ImageJ (Fiji) software [46]. The cell number (n) counted is noted under the graph column. Cells were cultured in parallel to examine the expression levels of the transfected genes by immunoblot analysis. The one-way ANOVA test was used to assess the statistical significance of differences between groups, and the t-test was used for post hoc comparisons. **, p < 0.01 and ***, p < 0.001. (D) Immunoblots of lysates derived from HeLa.Kyoto cells transfected with plasmids expressing vIRF-1-Flag, V5-GABARAPL1, and HA-NIX for 24 h in the presence and absence of 40 μM leupeptin. The relative intensities (Rel. Int.) of MT-CO2 bands were determined by dividing them with that in empty vector control and noted under each band. (E) Immunoblot analysis of extracts of control and GABARAPL1 KO HeLa.Kyoto cells transfected with plasmids expressing vIRF-1-Flag (500 ng), V5-NIX (200 ng), and HA-GABARAPL1 (50 and 200 ng) for 24 h. (F) Mitophagy assay in HeLa.Kyotomito-mCE cells co-transfected with V5-NIX together with WT or Δ227–236 vIRF-1. ‘n’ indicates the number of cells counted. ***, p < 0.001. (G-H) Co-IP and NanoBiT assays for assessment of the effects of NIX expression on the interaction between vIRF-1 and GABARAPL1. 293T cells were transfected with the indicated plasmids for 24 h. (G) Immunoblot analysis of cell lysates and Flag-IP complexes derived from the transfected cells. Arrowhead indicates the band of V5-GABARAPL1 co-precipitated with vIRF-1, and ‘LC’ indicates the light chain of immunoglobulin G (IgG). (H) The NanoBiT data are presented as the mean ± SD of six independent wells per condition. The one-way ANOVA test was used to assess the statistical significance of differences between groups, and the t-test was used for post hoc pairwise comparisons. **, p < 0.01 and *, p < 0.05. Immunoblots of the cell extracts are shown below the chart.
Fig 7.
NIX promotes the generation of vIRF-1 proteins with higher molecular weight.
(A) vIRF-1’s nuclear localization signal (NLS) and its mutation (NLSX). (B-C) Mitophagy (B) and immunoblot (C) analyses in HeLa.Kyotomito-mCE cells transfected with WT or NLSX vIRF-1 vector along with or without V5-NIX for 24 h. (D) IFA analysis of vIRF-1 and vIRF-1.NLSX in HeLa.Kyoto cells co-transfected with or without V5-NIX plasmid. Cells were fixed (Fix) before or after permeabilization (Per). Arrows indicate the co-localization of vIRF-1 and NIX in mitochondria. Scale bar, 10 μm. (E) Immunoblot analysis of the mitochondrial extracts derived from HeLa.Kyoto cells transfected with the indicated plasmids. (F) NanoBiT assay. 293T cells were transfected with the indicated NanoBiT plasmids with or without V5-NIX. Data are presented as the mean ± SD of six independent wells per condition. The one-way ANOVA test assessed the statistical significance of differences between groups, and the t-test was used for post hoc pairwise comparisons. **, p < 0.01 and *, p < 0.05. Immunoblots of the cell extracts are shown below the chart.
Fig 8.
NIX, but not GABARAPL1, promotes vIRF-1 aggregation for reactivation-induced mitophagy and productive virus replication.
(A) Immunoblot analysis of the RIPA-soluble and insoluble (DRM) fractions derived from iBCBL-1 cells treated with Dox for 0 to 3 days. (NIX)2 indicates dimerized NIX. (B) Immunoblot analysis of the DRM fraction derived from control and NIX KO iBCBL-1 cells treated without or with Dox for 2 days. (TUFM)2 indicates dimerized TUFM, and the red line indicates vIRF-1 proteins with a mass of more than 140 kDa. Relative band intensities of vIRF-1 (> 140 kDa) and (TUFM)2 are graphed. For NIX immunoblotting, whole-cell extracts were used. (C) IFA of control and NIX KO iBCBL-1 cells reactivated by Dox treatment for 2 days. Leupeptin was added to the culture for 24 h before fixation. Arrows indicate autolysosomes containing mitochondria (termed mitolysosomes), the co-localization of LAMP1 and TOM20. The percentages of vIRF-1-expressing cells and mitolysosome-containing cells are shown in the charts. A rat anti-vIRF-1 antibody was used for the triple-color IFA. Scale bar, 10 μm. (D) RT-qPCR analysis of the copy number of the encapsidated viral genome present in the culture media of control and NIX KO iBCBL-1 cells left untreated or treated with Dox for 1 to 3 days. Data represent the mean + SD of three independent experiments. The one-way ANOVA test was used to assess the statistical significance of differences between groups, and the t-test was used for post hoc pairwise comparisons. ** p < 0.01 and * p < 0.05. (E) Immunoblot analysis of the RIPA-soluble and insoluble fractions derived from control and GABARAPL1 KO iBCBL-1 cells left untreated or treated with Dox for 2 days. VDAC was used as a marker of the DRM fraction.
Fig 9.
NIX promotes vIRF-1 dimerization and stabilizes aggregated vIRF-1.
(A) NanoBiT assays of NIX-promoted vIRF-1 dimerization. Data are presented as the mean ± SD of three independent wells per condition. The one-way ANOVA test was used to assess the statistical significance of differences between groups, and the t-test was used for post hoc pairwise comparisons. ***, p < 0.001 and **, p < 0.01. The expression levels of the NanoBiT vIRF-1 (WT and NLSX) and V5-NIX (WT and ΔTA) proteins were examined using immunoblot analysis. (B) Immunoblot analysis of the RIPA-soluble and insoluble fractions derived from HeLa.Kyoto cells co-transfected with vIRF-1 with or without V5-NIX for 24 h and then treated with cycloheximide (CHX) for 0, 2, 4, 6, and 8 h. The red line indicates vIRF-1 proteins with a more than 140 kDa mass. Representative images are shown from three independent experiments. (C) Relative levels of monomeric vIRF-1 (50 kDa) and vIRF-1 aggregates above 140 kDa detected in (B) were quantified and indicated relative to levels at the initiation (0 h) of CHX treatment in the plots. Half-life was determined using a non-linear regression fit.
Fig 10.
Proposed models of vIRF-1-mediated mitophagy.
vIRF-1 is inducibly expressed during lytic replication and targeted to altered or damaged mitochondria, where it plays roles in mitophagy initiation and progression, likely via activation of NIX and TUFM-mediated mitophagy and recruitment of the autophagosomal membrane-bound GABARAPL1 to the mitochondria. Solid and dashed blue arrows indicate the pathways identified by the current study and previously published findings [16,17,18]. Solid and dashed gray arrows indicate the mechanisms identified by others [29,32,47] and unknown, respectively. GL1, GABARAPL1; ‘n’, multimer; ‘2’, dimer.