Fig 1.
Sensitivity of Ubl4A-deficient cells to starvation-induced death is associated with mitochondrial fragmentation.
(A) Cell death assay of wild-type (WT) and Ubl4A-deficient (KO) mouse embryonic fibroblasts (MEFs) in serum free or in HBSS medium for starvation (Starv.) for 16 h. Cell death was measured by the cellular membrane permeabilization to ethidium homodimer-1 (EthD-1). (B) Caspase 3 activation assay of the WT and KO MEFs under HBSS starvation at the indicated time points using NucView® 530 Caspase-3 substrate. (C) Liver tissue sections from WT and Ubl4A KO neonates (< 6 h after birth) immune-stained with anti-Tom 20 antibody. Bar, 20 μm. (D) Quantitative analysis of liver mitochondrial integrity by matrix score. Tiled images with similar number of nuclei from each whole tissue section were manually assigned score based on the integrity of the mitochondria. The integrity of mitochondria is scored as 0 (fragmented), 1 (mostly fragmented), 2 (intermedium), 3 (mostly intact), or 4 (intact); n = 5 for WT and n = 17 for KO neonates. ** p < 0.01. (E) Transmission electron microscopy of the ultrastructure of the mitochondria in the WT and KO neonatal liver hepatocytes (~6 h after birth). Bar, 2 μm.
Fig 2.
Lack of Ubl4A lead to mitochondrial fragmentation under nutrition deprivation.
(A) Morphology of mitochondria in primary WT and KO MEFs under HBSS starvation conditions for 2 h. Mitochondria were immunostained with anti-Tom20 antibody and imaged with confocal fluorescence microscopy; phalloidin-stained actin was used to indicate the cell boundary. Bar, 10 μm. (B) Mitochondrial length analysis based on (A). ** p < 0.01. (C) Quantification of cells with elongated mitochondria (> 5 μm) under starvation. A total of 15 cells for each group was analyzed; *** p < 0.001.
Fig 3.
Fragmented mitochondria in Ubl4A-deficient cells are mainly due to the failure of the fusion process.
(A) Immunoblotting of Mfn1 protein levels in MEFs under HBSS starvation for the indicated times. The intensity of each band was quantitated by ImageJ using actin band as the reference. (B) Immunoblotting with anti-p-Drp1 (S616), anti-p-Drp1 (S637) and anti-Drp1 antibodies in the WT and KO MEFs under starvation for indicated time. Actin was used as a protein loading control. (C) Time-lapse confocal images of WT and KO MEFs co-transfected with mEGFP-Lifeact-7 (green) and DsRed2-Mito-7 (red) under HBSS starvation conditions. A 6-min video of each cell was captured in 15-sec intervals; n = 15 for each group. Bar, 2 μm. (D) Extended time-lapse confocal images of KO MEFs with an interval of 40 sec. Bar, 2 μm. (E) Quantification of the percentage of fusion event for mitochondria in close proximity (< 2 μm) at indicated time points; * p < 0.05 and ** p < 0.01. (F) Non-parametric quantification of mitochondrial fusion events occurred during video acquisition. Seven events per cell for a total of 10 cells in each group were analyzed; * p < 0.05. (G) Quantification of actin fluorescent intensity. Ten areas with similar mitochondrial fluorescent intensities per cell for a total of 10 cells in each group were analyzed.
Fig 4.
The actin-related protein Arp2/3 complex is critical for the mitochondrial fusion process under starvation conditions. (A) Morphology of mitochondria in the WT MEFs under HBSS starvation for 1.5 h with or without actin polymerization inhibitor latrunculin B (LatB, 10 nM). Mitochondria were immunostained with anti-Tom20 antibody, and actin was visualized with phalloidin. (B) Quantification of the cells in (A) with fragmented mitochondria. A total of 25 cells for each group were analyzed; ** p < 0.01. (C) Morphology of mitochondria in the WT MEFs under HBSS starvation for 2 h with or without the Arp2/3 inhibitor CK666 (30 μM). Mitochondria were visualized with MitoTracker Red. CK689 serves as an inactive control. (D) Morphology of mitochondria in the WT MEFs transfected with indicated siRNAs (control, mock siRNA) followed by starvation in HBSS for 2 h. Mitochondria were visualized with MitoTracker Red. (E) Quantification of the cells in (C) with fragmented mitochondria. A total of 25 cells for each group were analyzed; *** p < 0.001. (F) Quantification of cells from (D) with fragmented mitochondria. A total of 15 cells for each group were analyzed; ** p < 0.01 and *** p < 0.001. (G) Immunoblotting of ArpC2 and Arp3 in the WT MEFs from (D). Bar, 10 μm.
Fig 5.
Lack of Ubl4A results in a smaller pool of primed Arp2/3 surrounding mitochondria.
(A) Schematic representation (not to scale) of the relationship among Ubl4A, the Arp2/3 complex and actin Y-shaped branch. (B) Localization of endogenous Ubl4A, ArpC2 and mitochondria in WT MEFs. The cells were transfected with DsRed2-Mito-7 for 16 h, and then immunostained with anti-Ubl4A and anti-ArpC2 antibodies. Colocalized proteins (pseudocolored white) among Ubl4A (red), ArpC2 (blue) and mitochondria (green) were shown in the enlarged image. The overlay (magenta) of Ubl4A and ArpC2 was extracted to eliminate the background, based on binary image analysis. Line scan of the fluorescent intensity (insert). Bar, 10 μm. (C) and (D) Colocalization of endogenous ArpC2 and mitochondria in WT and KO (C), and KO MEFs transfected with D122A (D). Cells were also transfected with DsRed2-Mito-7 for 16 h and then immunostained with anti-ArpC2 antibody (green). Bar, 10 μm. (E) Quantitation of overlay fluorescent intensities of ArpC2 and mitochondria in selected regions with similar mitochondrial densities based on (C) and (D). Ten fields from a total of 10 cells for each group were analyzed; * p < 0.05 and ** p < 0.01. (F) Immunoblotting of ArpC2 protein in cells shown in (C). Actin was used as a protein loading control.
Fig 6.
Ubl4A WT, but not Arp2/3-binding deficient mutant, can rescue mitochondria-fragmentation phenotype under starvation condition.
(A) Schematic representation of WT Ubl4A and 4 mutant constructs (not to scale) and their abilities to bind the Arp2/3 complex. Enlarged peptide sequences are shown. The conserved “DYD” Arp2/3-binding domain is aligned with N-WASP and cortactin sequences. The acidic domains in N-WASP and cortactin are in boxes with solid lines, and the putative acidic domain in Ubl4A is in a box with dash lines. (B) Rescue assay in WT and KO MEFs transfected with GFP-tagged Ubl4A or D122A constructs for 16 h and starved in HBSS for 2 h. Ctrl, GFP only. Mitochondria were visualized with MitoTracker Red. Enlarged boxed regions shown below. Bar, 10 μm. (C) Quantification of cells with elongated mitochondria as shown in (B). A total of 15 cells for each group was analyzed; *** p < 0.001. (D) Immunoblotting with anti-GFP antibody of cells from (B). The upper bands (Ubl4A*) represent full-length constructs, and the lower bands (GFP*) represent degraded GFP proteins.
Fig 7.
Ubl4A-deficient mediated cell death under starvation stress is through mitochondrial dependent apoptosis.
(A) Mitochondrial transmembrane potential assay. WT and KO MEFs were starved in HBSS for the indicated time. The cells were stained with TMRE and subjected to flow cytometry analysis. (B) Cellular ROS level assay. WT and KO MEFs were in HBSS for indicated times. The cells were stained with H2DCF and subjected to flow cytometry analysis. The value represents the fold difference compared to the level of non-starved cells at the “0” time point for each group. All experiments were repeated independently at least three times for statistical analysis; *** p < 0.001. (C) Visualizations of LC3 puncta and mitochondria in WT and KO MEFs cells that co-transfected with GFP-LC3 and DsRed2-Mito-7 treated with or without chloroquine (CQ, 30 μM) under normal or starvation condition. (D) Quantification of LC3 puncta number per cell based on cells from (C); * p < 0.05 and *** p < 0.001. (E) Quantification of LC3 puncta sizes in cells from (C). Both quantification analysis were performed using ImageJ with the particle analysis function based on binary images. The threshold was set to exclude out particles that smaller than 0.01 μm. Seven whole cell images from each group were analyzed. (F) Immunoblotting of caspase 9 in the WT and KO MEFs under starvation for the indicated time; actin was used as a protein-loading control. (G) Cell death assay under HBSS starvation for 16 h with or without the indicated inhibitors (Inh, inhibitor; BHA, butylated hydroxyanisole).